From 020ac3dca250a327323a539fac52e916bae5dc6c Mon Sep 17 00:00:00 2001 From: Aahlad Date: Thu, 3 Sep 2026 05:28:16 +0000 Subject: [PATCH] Update schedule free adamw jax v2 resnet logs with the new run info that includes the batchnorm fix --- ...magenet_resnet_jax_05-12-2026-19-39-14.log | 1834 ------- ...magenet_resnet_jax_08-31-2026-08-45-50.log | 2584 ++++++++++ .../trial_1/eval_measurements.csv | 76 +- ...t.tfevents.1778614855.ae408ec2d842.13.0.v2 | Bin 360787 -> 0 bytes ...t.tfevents.1788165989.f782bac2bd13.13.0.v2 | Bin 0 -> 520738 bytes .../imagenet_resnet_jax/trial_1/flags_0.json | 5 +- .../trial_1/measurements.csv | 3978 +++++++++------ .../trial_1/meta_data_0.json | 24 +- ...magenet_resnet_jax_05-12-2026-19-45-09.log | 1766 ------- ...magenet_resnet_jax_08-31-2026-09-17-19.log | 2600 ++++++++++ .../trial_1/eval_measurements.csv | 76 +- ...t.tfevents.1778615207.4ea6dc0cbfa9.13.0.v2 | Bin 352423 -> 0 bytes ...t.tfevents.1788167878.1810d6ef6af2.13.0.v2 | Bin 0 -> 524002 bytes .../imagenet_resnet_jax/trial_1/flags_0.json | 5 +- .../trial_1/measurements.csv | 3953 +++++++++------ .../trial_1/meta_data_0.json | 26 +- ...magenet_resnet_jax_05-12-2026-22-28-52.log | 2139 -------- ...magenet_resnet_jax_08-31-2026-09-17-19.log | 2564 ++++++++++ .../trial_1/eval_measurements.csv | 76 +- ...t.tfevents.1778625035.77174c5fce7f.13.0.v2 | Bin 430147 -> 0 bytes ...t.tfevents.1788167878.0ab01cb300d6.13.0.v2 | Bin 0 -> 516862 bytes .../imagenet_resnet_jax/trial_1/flags_0.json | 5 +- .../trial_1/measurements.csv | 4299 +++++++++-------- .../trial_1/meta_data_0.json | 30 +- 24 files changed, 15051 insertions(+), 10989 deletions(-) delete mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/imagenet_resnet_jax_05-12-2026-19-39-14.log create mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-08-45-50.log delete mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/events.out.tfevents.1778614855.ae408ec2d842.13.0.v2 create mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/events.out.tfevents.1788165989.f782bac2bd13.13.0.v2 delete mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/imagenet_resnet_jax_05-12-2026-19-45-09.log create mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-09-17-19.log delete mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/events.out.tfevents.1778615207.4ea6dc0cbfa9.13.0.v2 create mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/events.out.tfevents.1788167878.1810d6ef6af2.13.0.v2 delete mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/imagenet_resnet_jax_05-12-2026-22-28-52.log create mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-09-17-19.log delete mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/events.out.tfevents.1778625035.77174c5fce7f.13.0.v2 create mode 100644 logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/events.out.tfevents.1788167878.0ab01cb300d6.13.0.v2 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/imagenet_resnet_jax_05-12-2026-19-39-14.log b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/imagenet_resnet_jax_05-12-2026-19-39-14.log deleted file mode 100644 index 561a7136a..000000000 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/imagenet_resnet_jax_05-12-2026-19-39-14.log +++ /dev/null @@ -1,1834 +0,0 @@ -python submission_runner.py --framework=jax --workload=imagenet_resnet --submission_path=submissions_algorithms/submissions/self_tuning/schedule_free_adamw_jax_v2/submission.py --data_dir=/data/imagenet/jax --experiment_dir=/experiment_runs --experiment_name=submissions_a100/schedule_free_adamw_jax_v2/study_0 --overwrite=True --save_checkpoints=False --rng_seed=-999461833 --imagenet_v2_data_dir=/data/imagenet/jax --tuning_ruleset=self 2>&1 | tee -a /logs/imagenet_resnet_jax_05-12-2026-19-39-14.log -2026-05-12 19:39:25.029106: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered -WARNING: All log messages before absl::InitializeLog() is called are written to STDERR -E0000 00:00:1778614765.455043 13 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered -E0000 00:00:1778614765.619037 13 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered -W0000 00:00:1778614766.708679 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -W0000 00:00:1778614766.708726 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -W0000 00:00:1778614766.708729 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -W0000 00:00:1778614766.708732 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -/usr/local/lib/python3.11/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.) - _C._set_float32_matmul_precision(precision) -/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'repr' attribute with value False was provided to the `Field()` function, which has no effect in the context it was used. 'repr' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. - warnings.warn( -/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'frozen' attribute with value True was provided to the `Field()` function, which has no effect in the context it was used. 'frozen' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. - warnings.warn( -INFO:2026-05-12 19:40:19,514:jax._src.xla_bridge:830: Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory -I0512 19:40:19.514517 139903809070272 xla_bridge.py:830] Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory -I0512 19:40:20.428745 139903809070272 logger_utils.py:84] Creating experiment directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax. -I0512 19:40:24.094998 139903809070272 submission_runner.py:741] Creating directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1. -I0512 19:40:24.443092 139903809070272 submission_runner.py:242] Initializing dataset. -I0512 19:40:26.192402 139903809070272 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:40:26.257916 139903809070272 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. -I0512 19:40:26.550514 139903809070272 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. -I0512 19:40:26.789263 139903809070272 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:40:28.485011 139903809070272 submission_runner.py:251] Initializing model. -I0512 19:40:53.573942 139903809070272 submission_runner.py:294] Initializing optimizer. -I0512 19:40:55.380751 139903809070272 submission_runner.py:299] Initializing metrics bundle. -I0512 19:40:55.380980 139903809070272 submission_runner.py:321] Initializing checkpoint and logger. -I0512 19:40:55.383667 139903809070272 checkpoints.py:1098] Found no checkpoint files in /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1 with prefix checkpoint_ -I0512 19:40:55.383805 139903809070272 submission_runner.py:344] Saving meta data to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/meta_data_0.json. -I0512 19:40:55.912556 139903809070272 submission_runner.py:348] Saving flags to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/flags_0.json. -I0512 19:40:55.924281 139903809070272 submission_runner.py:359] Starting training loop. -/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. -See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. - warnings.warn("Some donated buffers were not usable:" -I0512 19:41:56.177810 139888644712192 logging_writer.py:48] [0] global_step=0, grad_norm=0.2802160084247589, loss=6.908313751220703 -I0512 19:41:57.962172 139903809070272 spec.py:333] Evaluating on the training split. -I0512 19:41:58.250133 139903809070272 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:41:58.255790 139903809070272 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. -I0512 19:41:58.274577 139903809070272 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. -I0512 19:41:58.318836 139903809070272 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:42:34.805708 139903809070272 spec.py:346] Evaluating on the validation split. -I0512 19:42:34.815381 139903809070272 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:42:34.857994 139903809070272 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. -I0512 19:42:34.861856 139903809070272 reader.py:262] Creating a tf.data.Dataset reading 64 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. -I0512 19:42:35.129204 139903809070272 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split validation, from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:43:14.165922 139903809070272 spec.py:363] Evaluating on the test split. -I0512 19:43:14.366281 139903809070272 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 -I0512 19:43:14.446761 139903809070272 reader.py:262] Creating a tf.data.Dataset reading 16 files located in folders: /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0. -I0512 19:43:14.489205 139903809070272 logging_logger.py:49] Constructing tf.data.Dataset imagenet_v2 for split test, from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 -I0512 19:43:21.197924 139903809070272 submission_runner.py:516] Time since start: 145.27s, Step: 1, {'train/accuracy': Array(0.00107621, dtype=float32), 'train/loss': Array(6.9101987, dtype=float32), 'validation/accuracy': Array(0.0011, dtype=float32), 'validation/loss': Array(6.910505, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(6.910612, dtype=float32), 'test/num_examples': 10000, 'score': 62.0377836227417, 'total_duration': 145.27190852165222, 'accumulated_submission_time': 62.0377836227417, 'accumulated_eval_time': 83.23402738571167, 'accumulated_logging_time': 0} -I0512 19:43:21.216632 139687746324224 logging_writer.py:48] [1] accumulated_eval_time=83.234, accumulated_logging_time=0, accumulated_submission_time=62.0378, global_step=1, preemption_count=0, score=62.0378, test/accuracy=0.0013000001199543476, test/loss=6.910612106323242, test/num_examples=10000, total_duration=145.272, train/accuracy=0.0010762116871774197, train/loss=6.91019868850708, validation/accuracy=0.0010999999940395355, validation/loss=6.9105048179626465, validation/num_examples=50000 -/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. -See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. - warnings.warn("Some donated buffers were not usable:" -I0512 19:44:16.217670 139687360456448 logging_writer.py:48] [100] global_step=100, grad_norm=0.9382998943328857, loss=6.756219863891602 -I0512 19:44:57.416836 139687368849152 logging_writer.py:48] [200] global_step=200, grad_norm=1.7159137725830078, loss=6.43281364440918 -I0512 19:45:45.414412 139687360456448 logging_writer.py:48] [300] global_step=300, grad_norm=1.9524415731430054, loss=6.266709327697754 -I0512 19:46:32.927969 139687368849152 logging_writer.py:48] [400] global_step=400, grad_norm=3.24735951423645, loss=6.178126335144043 -I0512 19:47:12.749411 139687360456448 logging_writer.py:48] [500] global_step=500, grad_norm=4.8911638259887695, loss=6.262415885925293 -I0512 19:47:56.981137 139687368849152 logging_writer.py:48] [600] global_step=600, grad_norm=2.3864083290100098, loss=6.08144474029541 -I0512 19:48:36.665714 139687360456448 logging_writer.py:48] [700] global_step=700, grad_norm=2.7533810138702393, loss=5.971892356872559 -I0512 19:49:16.379648 139687368849152 logging_writer.py:48] [800] global_step=800, grad_norm=3.687201976776123, loss=5.987758159637451 -I0512 19:49:56.421373 139687360456448 logging_writer.py:48] [900] global_step=900, grad_norm=2.628094434738159, loss=5.933117866516113 -I0512 19:50:36.246309 139687368849152 logging_writer.py:48] [1000] global_step=1000, grad_norm=2.3546266555786133, loss=5.840041160583496 -I0512 19:51:15.669799 139687360456448 logging_writer.py:48] [1100] global_step=1100, grad_norm=3.599294900894165, loss=5.881869792938232 -I0512 19:51:55.499728 139687368849152 logging_writer.py:48] [1200] global_step=1200, grad_norm=3.974897861480713, loss=5.886465072631836 -I0512 19:52:41.853332 139687360456448 logging_writer.py:48] [1300] global_step=1300, grad_norm=5.0108442306518555, loss=5.775548458099365 -I0512 19:53:30.965494 139687368849152 logging_writer.py:48] [1400] global_step=1400, grad_norm=3.1748340129852295, loss=5.806589126586914 -I0512 19:54:18.604654 139687360456448 logging_writer.py:48] [1500] global_step=1500, grad_norm=4.364927291870117, loss=5.697990417480469 -I0512 19:54:59.727285 139687368849152 logging_writer.py:48] [1600] global_step=1600, grad_norm=2.647134780883789, loss=5.748484134674072 -I0512 19:55:39.680126 139687360456448 logging_writer.py:48] [1700] global_step=1700, grad_norm=2.4704806804656982, loss=5.695845127105713 -I0512 19:56:27.253078 139687368849152 logging_writer.py:48] [1800] global_step=1800, grad_norm=1.746597170829773, loss=5.610013961791992 -I0512 19:57:06.608548 139687360456448 logging_writer.py:48] [1900] global_step=1900, grad_norm=6.677021026611328, loss=5.8874101638793945 -I0512 19:57:45.978434 139687368849152 logging_writer.py:48] [2000] global_step=2000, grad_norm=5.062442302703857, loss=5.7971625328063965 -I0512 19:58:32.254596 139687360456448 logging_writer.py:48] [2100] global_step=2100, grad_norm=3.3601138591766357, loss=5.8100056648254395 -I0512 19:59:14.069373 139687368849152 logging_writer.py:48] [2200] global_step=2200, grad_norm=2.5220048427581787, loss=5.525139331817627 -I0512 19:59:55.983500 139687360456448 logging_writer.py:48] [2300] global_step=2300, grad_norm=3.1667466163635254, loss=5.560049533843994 -I0512 20:00:42.109352 139687368849152 logging_writer.py:48] [2400] global_step=2400, grad_norm=3.263444423675537, loss=5.671502113342285 -I0512 20:01:24.267102 139687360456448 logging_writer.py:48] [2500] global_step=2500, grad_norm=2.5383379459381104, loss=5.541540145874023 -I0512 20:02:10.073902 139687368849152 logging_writer.py:48] [2600] global_step=2600, grad_norm=3.044522285461426, loss=5.448047161102295 -I0512 20:02:56.019405 139687360456448 logging_writer.py:48] [2700] global_step=2700, grad_norm=2.3397135734558105, loss=5.385077953338623 -I0512 20:03:41.771748 139687368849152 logging_writer.py:48] [2800] global_step=2800, grad_norm=2.659743547439575, loss=5.375866889953613 -I0512 20:04:27.394889 139687360456448 logging_writer.py:48] [2900] global_step=2900, grad_norm=2.9278173446655273, loss=5.387120246887207 -I0512 20:05:13.179934 139687368849152 logging_writer.py:48] [3000] global_step=3000, grad_norm=3.331817865371704, loss=5.322810173034668 -I0512 20:06:02.445156 139687360456448 logging_writer.py:48] [3100] global_step=3100, grad_norm=3.513012409210205, loss=5.5266876220703125 -I0512 20:06:48.107108 139687368849152 logging_writer.py:48] [3200] global_step=3200, grad_norm=3.0216310024261475, loss=5.354588508605957 -I0512 20:07:43.798170 139687360456448 logging_writer.py:48] [3300] global_step=3300, grad_norm=1.2741576433181763, loss=5.287280559539795 -I0512 20:08:26.581706 139687368849152 logging_writer.py:48] [3400] global_step=3400, grad_norm=3.2336976528167725, loss=5.4634552001953125 -I0512 20:09:09.107233 139687360456448 logging_writer.py:48] [3500] global_step=3500, grad_norm=2.3230338096618652, loss=5.291337013244629 -I0512 20:09:51.626270 139687368849152 logging_writer.py:48] [3600] global_step=3600, grad_norm=2.6491072177886963, loss=5.3013505935668945 -I0512 20:10:34.820938 139687360456448 logging_writer.py:48] [3700] global_step=3700, grad_norm=3.639723062515259, loss=5.378150939941406 -I0512 20:11:17.914213 139687368849152 logging_writer.py:48] [3800] global_step=3800, grad_norm=3.0273232460021973, loss=5.203970432281494 -I0512 20:12:03.717263 139687360456448 logging_writer.py:48] [3900] global_step=3900, grad_norm=2.9177939891815186, loss=5.185932636260986 -I0512 20:12:48.956612 139687368849152 logging_writer.py:48] [4000] global_step=4000, grad_norm=3.2125682830810547, loss=5.222461700439453 -I0512 20:13:34.379122 139687360456448 logging_writer.py:48] [4100] global_step=4100, grad_norm=2.945589780807495, loss=5.10198974609375 -I0512 20:14:15.770854 139687368849152 logging_writer.py:48] [4200] global_step=4200, grad_norm=3.40757155418396, loss=5.250633716583252 -I0512 20:15:00.762333 139687360456448 logging_writer.py:48] [4300] global_step=4300, grad_norm=2.1821987628936768, loss=5.077275276184082 -I0512 20:15:45.982752 139687368849152 logging_writer.py:48] [4400] global_step=4400, grad_norm=1.9292452335357666, loss=5.111266613006592 -I0512 20:16:31.418749 139687360456448 logging_writer.py:48] [4500] global_step=4500, grad_norm=3.077381134033203, loss=5.138885498046875 -I0512 20:16:37.278876 139903809070272 spec.py:333] Evaluating on the training split. -I0512 20:16:49.626265 139903809070272 spec.py:346] Evaluating on the validation split. -I0512 20:17:11.306954 139903809070272 spec.py:363] Evaluating on the test split. -I0512 20:17:12.218195 139903809070272 submission_runner.py:516] Time since start: 2176.29s, Step: 4518, {'train/accuracy': Array(0.03958068, dtype=float32), 'train/loss': Array(5.9223266, dtype=float32), 'validation/accuracy': Array(0.03606, dtype=float32), 'validation/loss': Array(5.994765, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0295, dtype=float32), 'test/loss': Array(6.248613, dtype=float32), 'test/num_examples': 10000, 'score': 2058.049063682556, 'total_duration': 2176.2921516895294, 'accumulated_submission_time': 2058.049063682556, 'accumulated_eval_time': 118.17159128189087, 'accumulated_logging_time': 0.027879953384399414} -I0512 20:17:12.240303 139687754716928 logging_writer.py:48] [4518] accumulated_eval_time=118.172, accumulated_logging_time=0.02788, accumulated_submission_time=2058.05, global_step=4518, preemption_count=0, score=2058.05, test/accuracy=0.029500002041459084, test/loss=6.248612880706787, test/num_examples=10000, total_duration=2176.29, train/accuracy=0.03958067670464516, train/loss=5.922326564788818, validation/accuracy=0.03605999797582626, validation/loss=5.994764804840088, validation/num_examples=50000 -I0512 20:17:47.479588 139687763109632 logging_writer.py:48] [4600] global_step=4600, grad_norm=3.0757808685302734, loss=5.022922039031982 -I0512 20:18:29.114845 139687754716928 logging_writer.py:48] [4700] global_step=4700, grad_norm=2.1589345932006836, loss=5.158249855041504 -I0512 20:19:06.450168 139687763109632 logging_writer.py:48] [4800] global_step=4800, grad_norm=2.1429059505462646, loss=5.04161262512207 -I0512 20:19:43.210133 139687754716928 logging_writer.py:48] [4900] global_step=4900, grad_norm=1.5161939859390259, loss=5.01307487487793 -I0512 20:20:20.344309 139687763109632 logging_writer.py:48] [5000] global_step=5000, grad_norm=3.9396579265594482, loss=5.104777812957764 -I0512 20:21:02.106507 139687754716928 logging_writer.py:48] [5100] global_step=5100, grad_norm=4.218153476715088, loss=5.143609523773193 -I0512 20:21:39.222205 139687763109632 logging_writer.py:48] [5200] global_step=5200, grad_norm=3.3677444458007812, loss=5.036825656890869 -I0512 20:22:15.976858 139687754716928 logging_writer.py:48] [5300] global_step=5300, grad_norm=3.7020740509033203, loss=4.969170570373535 -I0512 20:22:53.226783 139687763109632 logging_writer.py:48] [5400] global_step=5400, grad_norm=2.8452134132385254, loss=5.026850700378418 -I0512 20:23:29.952333 139687754716928 logging_writer.py:48] [5500] global_step=5500, grad_norm=4.970737457275391, loss=4.811808109283447 -I0512 20:24:06.933721 139687763109632 logging_writer.py:48] [5600] global_step=5600, grad_norm=3.414165496826172, loss=4.902349472045898 -I0512 20:24:44.125248 139687754716928 logging_writer.py:48] [5700] global_step=5700, grad_norm=2.45456862449646, loss=4.897881507873535 -I0512 20:25:21.301983 139687763109632 logging_writer.py:48] [5800] global_step=5800, grad_norm=5.332715034484863, loss=4.974867820739746 -I0512 20:25:58.056898 139687754716928 logging_writer.py:48] [5900] global_step=5900, grad_norm=3.269678831100464, loss=4.857068061828613 -I0512 20:26:35.257316 139687763109632 logging_writer.py:48] [6000] global_step=6000, grad_norm=2.5486977100372314, loss=4.811395645141602 -I0512 20:27:11.961945 139687754716928 logging_writer.py:48] [6100] global_step=6100, grad_norm=4.340679168701172, loss=4.983715534210205 -I0512 20:27:49.051329 139687763109632 logging_writer.py:48] [6200] global_step=6200, grad_norm=2.0728766918182373, loss=4.827754974365234 -I0512 20:28:26.350684 139687754716928 logging_writer.py:48] [6300] global_step=6300, grad_norm=1.8308902978897095, loss=4.71298885345459 -I0512 20:29:03.139892 139687763109632 logging_writer.py:48] [6400] global_step=6400, grad_norm=3.7294921875, loss=4.928774356842041 -I0512 20:29:44.239579 139687754716928 logging_writer.py:48] [6500] global_step=6500, grad_norm=3.407163619995117, loss=4.760442733764648 -I0512 20:30:25.574839 139687763109632 logging_writer.py:48] [6600] global_step=6600, grad_norm=2.967484712600708, loss=4.7801594734191895 -I0512 20:31:06.624220 139687754716928 logging_writer.py:48] [6700] global_step=6700, grad_norm=1.866814136505127, loss=4.697668075561523 -I0512 20:31:47.904933 139687763109632 logging_writer.py:48] [6800] global_step=6800, grad_norm=3.706573486328125, loss=4.767075538635254 -I0512 20:32:29.467040 139687754716928 logging_writer.py:48] [6900] global_step=6900, grad_norm=4.648370742797852, loss=4.785259246826172 -I0512 20:33:06.239376 139687763109632 logging_writer.py:48] [7000] global_step=7000, grad_norm=3.8844687938690186, loss=4.786125659942627 -I0512 20:33:43.360861 139687754716928 logging_writer.py:48] [7100] global_step=7100, grad_norm=1.2563434839248657, loss=4.713617324829102 -I0512 20:34:20.530880 139687763109632 logging_writer.py:48] [7200] global_step=7200, grad_norm=3.5405759811401367, loss=4.678468704223633 -I0512 20:34:57.302222 139687754716928 logging_writer.py:48] [7300] global_step=7300, grad_norm=3.221348285675049, loss=4.840146064758301 -I0512 20:35:34.481682 139687763109632 logging_writer.py:48] [7400] global_step=7400, grad_norm=3.2844977378845215, loss=4.826122760772705 -I0512 20:36:11.695938 139687754716928 logging_writer.py:48] [7500] global_step=7500, grad_norm=4.426513671875, loss=4.890480995178223 -I0512 20:36:52.870333 139687763109632 logging_writer.py:48] [7600] global_step=7600, grad_norm=3.5370514392852783, loss=4.669809341430664 -I0512 20:37:29.889665 139687754716928 logging_writer.py:48] [7700] global_step=7700, grad_norm=3.928209066390991, loss=4.813296318054199 -I0512 20:38:07.131660 139687763109632 logging_writer.py:48] [7800] global_step=7800, grad_norm=4.109675407409668, loss=4.658947944641113 -I0512 20:38:43.902080 139687754716928 logging_writer.py:48] [7900] global_step=7900, grad_norm=2.5258688926696777, loss=4.547046184539795 -I0512 20:39:21.017734 139687763109632 logging_writer.py:48] [8000] global_step=8000, grad_norm=5.185024261474609, loss=4.82103157043457 -I0512 20:39:58.353317 139687754716928 logging_writer.py:48] [8100] global_step=8100, grad_norm=4.452086448669434, loss=4.842930793762207 -I0512 20:40:35.227110 139687763109632 logging_writer.py:48] [8200] global_step=8200, grad_norm=1.3208221197128296, loss=4.551862716674805 -I0512 20:41:12.005470 139687754716928 logging_writer.py:48] [8300] global_step=8300, grad_norm=3.1019811630249023, loss=4.562349319458008 -I0512 20:41:49.516621 139687763109632 logging_writer.py:48] [8400] global_step=8400, grad_norm=2.722907781600952, loss=4.578224182128906 -I0512 20:42:26.346344 139687754716928 logging_writer.py:48] [8500] global_step=8500, grad_norm=2.8841350078582764, loss=4.61148738861084 -I0512 20:43:03.570422 139687763109632 logging_writer.py:48] [8600] global_step=8600, grad_norm=2.633077621459961, loss=4.473111152648926 -I0512 20:43:40.837627 139687754716928 logging_writer.py:48] [8700] global_step=8700, grad_norm=3.822070598602295, loss=4.606818675994873 -I0512 20:44:17.645178 139687763109632 logging_writer.py:48] [8800] global_step=8800, grad_norm=3.3098480701446533, loss=4.614909648895264 -I0512 20:44:54.717713 139687754716928 logging_writer.py:48] [8900] global_step=8900, grad_norm=4.846261501312256, loss=4.680874824523926 -I0512 20:45:31.967572 139687763109632 logging_writer.py:48] [9000] global_step=9000, grad_norm=3.0733392238616943, loss=4.596996784210205 -I0512 20:46:08.876970 139687754716928 logging_writer.py:48] [9100] global_step=9100, grad_norm=4.547897815704346, loss=4.743021011352539 -I0512 20:46:46.048499 139687763109632 logging_writer.py:48] [9200] global_step=9200, grad_norm=3.2441205978393555, loss=4.6345624923706055 -I0512 20:47:23.342687 139687754716928 logging_writer.py:48] [9300] global_step=9300, grad_norm=4.205641269683838, loss=4.600100994110107 -I0512 20:48:00.212854 139687763109632 logging_writer.py:48] [9400] global_step=9400, grad_norm=3.1424062252044678, loss=4.475790977478027 -I0512 20:48:37.376146 139687754716928 logging_writer.py:48] [9500] global_step=9500, grad_norm=3.134442090988159, loss=4.515594005584717 -I0512 20:49:14.635312 139687763109632 logging_writer.py:48] [9600] global_step=9600, grad_norm=3.334294080734253, loss=4.450712203979492 -I0512 20:49:51.405991 139687754716928 logging_writer.py:48] [9700] global_step=9700, grad_norm=2.7007834911346436, loss=4.489980697631836 -I0512 20:50:28.402466 139687763109632 logging_writer.py:48] [9800] global_step=9800, grad_norm=8.787710189819336, loss=4.575618743896484 -I0512 20:50:28.413539 139903809070272 spec.py:333] Evaluating on the training split. -I0512 20:50:41.338789 139903809070272 spec.py:346] Evaluating on the validation split. -I0512 20:51:32.382508 139903809070272 spec.py:363] Evaluating on the test split. -I0512 20:51:33.278759 139903809070272 submission_runner.py:516] Time since start: 4237.35s, Step: 9801, {'train/accuracy': Array(0.00974569, dtype=float32), 'train/loss': Array(7.5413218, dtype=float32), 'validation/accuracy': Array(0.01064, dtype=float32), 'validation/loss': Array(7.568091, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0079, dtype=float32), 'test/loss': Array(7.6627584, dtype=float32), 'test/num_examples': 10000, 'score': 4054.1717386245728, 'total_duration': 4237.352629899979, 'accumulated_submission_time': 4054.1717386245728, 'accumulated_eval_time': 183.03495979309082, 'accumulated_logging_time': 0.0581820011138916} -I0512 20:51:33.307406 139687754716928 logging_writer.py:48] [9801] accumulated_eval_time=183.035, accumulated_logging_time=0.058182, accumulated_submission_time=4054.17, global_step=9801, preemption_count=0, score=4054.17, test/accuracy=0.007900000549852848, test/loss=7.6627583503723145, test/num_examples=10000, total_duration=4237.35, train/accuracy=0.009745694696903229, train/loss=7.541321754455566, validation/accuracy=0.010639999993145466, validation/loss=7.568090915679932, validation/num_examples=50000 -I0512 20:52:20.224866 139687763109632 logging_writer.py:48] [9900] global_step=9900, grad_norm=4.0356059074401855, loss=4.556339740753174 -I0512 20:53:06.017261 139687754716928 logging_writer.py:48] [10000] global_step=10000, grad_norm=2.649406671524048, loss=4.472428321838379 -I0512 20:53:47.712953 139687763109632 logging_writer.py:48] [10100] global_step=10100, grad_norm=4.734405040740967, loss=4.51960563659668 -I0512 20:54:29.417647 139687754716928 logging_writer.py:48] [10200] global_step=10200, grad_norm=1.7634782791137695, loss=4.45598840713501 -I0512 20:55:10.742923 139687763109632 logging_writer.py:48] [10300] global_step=10300, grad_norm=1.9574803113937378, loss=4.424410343170166 -I0512 20:55:51.682655 139687754716928 logging_writer.py:48] [10400] global_step=10400, grad_norm=4.531286239624023, loss=4.361980438232422 -I0512 20:56:33.477789 139687763109632 logging_writer.py:48] [10500] global_step=10500, grad_norm=6.095785140991211, loss=4.57425594329834 -I0512 20:57:14.504872 139687754716928 logging_writer.py:48] [10600] global_step=10600, grad_norm=5.7735419273376465, loss=4.59458065032959 -I0512 20:57:55.950217 139687763109632 logging_writer.py:48] [10700] global_step=10700, grad_norm=8.404025077819824, loss=4.700727939605713 -I0512 20:58:37.691206 139687754716928 logging_writer.py:48] [10800] global_step=10800, grad_norm=3.036003828048706, loss=4.426248550415039 -I0512 20:59:19.035314 139687763109632 logging_writer.py:48] [10900] global_step=10900, grad_norm=2.9663846492767334, loss=4.350268363952637 -I0512 21:00:00.586247 139687754716928 logging_writer.py:48] [11000] global_step=11000, grad_norm=8.538077354431152, loss=4.614936828613281 -I0512 21:00:46.950387 139687763109632 logging_writer.py:48] [11100] global_step=11100, grad_norm=2.5892300605773926, loss=4.356301784515381 -I0512 21:01:28.256157 139687754716928 logging_writer.py:48] [11200] global_step=11200, grad_norm=4.808152675628662, loss=4.49868631362915 -I0512 21:02:09.927021 139687763109632 logging_writer.py:48] [11300] global_step=11300, grad_norm=3.9248297214508057, loss=4.443073272705078 -I0512 21:02:51.638608 139687754716928 logging_writer.py:48] [11400] global_step=11400, grad_norm=2.778954267501831, loss=4.425868034362793 -I0512 21:03:37.439811 139687763109632 logging_writer.py:48] [11500] global_step=11500, grad_norm=7.622842311859131, loss=4.56803035736084 -I0512 21:04:19.189204 139687754716928 logging_writer.py:48] [11600] global_step=11600, grad_norm=3.2570579051971436, loss=4.40865421295166 -I0512 21:05:00.949764 139687763109632 logging_writer.py:48] [11700] global_step=11700, grad_norm=3.3706181049346924, loss=4.343086242675781 -I0512 21:05:42.328101 139687754716928 logging_writer.py:48] [11800] global_step=11800, grad_norm=5.211455821990967, loss=4.545670509338379 -I0512 21:06:23.912988 139687763109632 logging_writer.py:48] [11900] global_step=11900, grad_norm=2.577214241027832, loss=4.376078128814697 -I0512 21:07:05.697006 139687754716928 logging_writer.py:48] [12000] global_step=12000, grad_norm=4.953535556793213, loss=4.559054374694824 -I0512 21:07:46.971395 139687763109632 logging_writer.py:48] [12100] global_step=12100, grad_norm=3.306490898132324, loss=4.4554643630981445 -I0512 21:08:33.244001 139687754716928 logging_writer.py:48] [12200] global_step=12200, grad_norm=3.197166919708252, loss=4.450170516967773 -I0512 21:09:14.998593 139687763109632 logging_writer.py:48] [12300] global_step=12300, grad_norm=2.8215527534484863, loss=4.278886318206787 -I0512 21:09:56.312023 139687754716928 logging_writer.py:48] [12400] global_step=12400, grad_norm=3.85644793510437, loss=4.34965705871582 -I0512 21:10:37.649062 139687763109632 logging_writer.py:48] [12500] global_step=12500, grad_norm=2.738300085067749, loss=4.313940048217773 -I0512 21:11:24.334533 139687754716928 logging_writer.py:48] [12600] global_step=12600, grad_norm=3.078697919845581, loss=4.4544453620910645 -I0512 21:12:05.672915 139687763109632 logging_writer.py:48] [12700] global_step=12700, grad_norm=5.460382461547852, loss=4.436374664306641 -I0512 21:12:47.299516 139687754716928 logging_writer.py:48] [12800] global_step=12800, grad_norm=3.106762170791626, loss=4.331503868103027 -I0512 21:13:29.098515 139687763109632 logging_writer.py:48] [12900] global_step=12900, grad_norm=2.612004041671753, loss=4.351551532745361 -I0512 21:14:10.426566 139687754716928 logging_writer.py:48] [13000] global_step=13000, grad_norm=3.872493267059326, loss=4.2877936363220215 -I0512 21:14:52.075899 139687763109632 logging_writer.py:48] [13100] global_step=13100, grad_norm=2.412848711013794, loss=4.177600383758545 -I0512 21:15:33.852538 139687754716928 logging_writer.py:48] [13200] global_step=13200, grad_norm=4.551445007324219, loss=4.2214460372924805 -I0512 21:16:15.205266 139687763109632 logging_writer.py:48] [13300] global_step=13300, grad_norm=12.106030464172363, loss=5.087636947631836 -I0512 21:17:01.408210 139687754716928 logging_writer.py:48] [13400] global_step=13400, grad_norm=4.094177722930908, loss=4.307053565979004 -I0512 21:17:43.224365 139687763109632 logging_writer.py:48] [13500] global_step=13500, grad_norm=4.601501941680908, loss=4.293885231018066 -I0512 21:18:29.227870 139687754716928 logging_writer.py:48] [13600] global_step=13600, grad_norm=5.7885236740112305, loss=4.47043514251709 -I0512 21:19:10.857241 139687763109632 logging_writer.py:48] [13700] global_step=13700, grad_norm=3.5543041229248047, loss=4.331947326660156 -I0512 21:19:57.146904 139687754716928 logging_writer.py:48] [13800] global_step=13800, grad_norm=3.262827157974243, loss=4.282590866088867 -I0512 21:20:38.482297 139687763109632 logging_writer.py:48] [13900] global_step=13900, grad_norm=3.0661590099334717, loss=4.330014705657959 -I0512 21:21:20.032089 139687754716928 logging_writer.py:48] [14000] global_step=14000, grad_norm=5.992323398590088, loss=4.254425048828125 -I0512 21:22:01.876201 139687763109632 logging_writer.py:48] [14100] global_step=14100, grad_norm=3.724125385284424, loss=4.340271949768066 -I0512 21:22:43.083638 139687754716928 logging_writer.py:48] [14200] global_step=14200, grad_norm=7.024749755859375, loss=4.561751842498779 -I0512 21:23:24.672060 139687763109632 logging_writer.py:48] [14300] global_step=14300, grad_norm=18.445703506469727, loss=4.336794853210449 -I0512 21:24:06.354096 139687754716928 logging_writer.py:48] [14400] global_step=14400, grad_norm=3.81604266166687, loss=4.209531307220459 -I0512 21:24:47.586064 139687763109632 logging_writer.py:48] [14500] global_step=14500, grad_norm=4.374428749084473, loss=4.188451766967773 -I0512 21:24:49.521304 139903809070272 spec.py:333] Evaluating on the training split. -I0512 21:25:01.812958 139903809070272 spec.py:346] Evaluating on the validation split. -I0512 21:27:01.068653 139903809070272 spec.py:363] Evaluating on the test split. -I0512 21:27:01.959974 139903809070272 submission_runner.py:516] Time since start: 6366.03s, Step: 14506, {'train/accuracy': Array(0.00585938, dtype=float32), 'train/loss': Array(8.495525, dtype=float32), 'validation/accuracy': Array(0.00528, dtype=float32), 'validation/loss': Array(8.498083, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0042, dtype=float32), 'test/loss': Array(8.537536, dtype=float32), 'test/num_examples': 10000, 'score': 6050.326933860779, 'total_duration': 6366.033954143524, 'accumulated_submission_time': 6050.326933860779, 'accumulated_eval_time': 315.471892118454, 'accumulated_logging_time': 0.1032261848449707} -I0512 21:27:01.984766 139687754716928 logging_writer.py:48] [14506] accumulated_eval_time=315.472, accumulated_logging_time=0.103226, accumulated_submission_time=6050.33, global_step=14506, preemption_count=0, score=6050.33, test/accuracy=0.004200000315904617, test/loss=8.537535667419434, test/num_examples=10000, total_duration=6366.03, train/accuracy=0.005859375, train/loss=8.495525360107422, validation/accuracy=0.005279999691992998, validation/loss=8.498083114624023, validation/num_examples=50000 -I0512 21:27:36.865387 139687763109632 logging_writer.py:48] [14600] global_step=14600, grad_norm=4.467906951904297, loss=4.266951560974121 -I0512 21:28:14.415077 139687754716928 logging_writer.py:48] [14700] global_step=14700, grad_norm=3.3470304012298584, loss=4.346274375915527 -I0512 21:28:51.210108 139687763109632 logging_writer.py:48] [14800] global_step=14800, grad_norm=4.785727500915527, loss=4.325855255126953 -I0512 21:29:28.369802 139687754716928 logging_writer.py:48] [14900] global_step=14900, grad_norm=5.4956464767456055, loss=4.244760990142822 -I0512 21:30:05.600263 139687763109632 logging_writer.py:48] [15000] global_step=15000, grad_norm=6.085444450378418, loss=4.3275556564331055 -I0512 21:30:42.381095 139687754716928 logging_writer.py:48] [15100] global_step=15100, grad_norm=7.097640514373779, loss=4.310112953186035 -I0512 21:31:19.112363 139687763109632 logging_writer.py:48] [15200] global_step=15200, grad_norm=10.452030181884766, loss=4.288189888000488 -I0512 21:31:56.738276 139687754716928 logging_writer.py:48] [15300] global_step=15300, grad_norm=2.343078136444092, loss=4.353918075561523 -I0512 21:32:33.643425 139687763109632 logging_writer.py:48] [15400] global_step=15400, grad_norm=3.6630425453186035, loss=4.374762058258057 -I0512 21:33:10.805676 139687754716928 logging_writer.py:48] [15500] global_step=15500, grad_norm=3.7355542182922363, loss=4.314619541168213 -I0512 21:33:48.042123 139687763109632 logging_writer.py:48] [15600] global_step=15600, grad_norm=4.433247089385986, loss=4.285011291503906 -I0512 21:34:24.783172 139687754716928 logging_writer.py:48] [15700] global_step=15700, grad_norm=3.806504011154175, loss=4.203555583953857 -I0512 21:35:01.877728 139687763109632 logging_writer.py:48] [15800] global_step=15800, grad_norm=3.1342339515686035, loss=4.219532012939453 -I0512 21:35:39.202394 139687754716928 logging_writer.py:48] [15900] global_step=15900, grad_norm=5.624528408050537, loss=4.309572696685791 -I0512 21:36:15.970752 139687763109632 logging_writer.py:48] [16000] global_step=16000, grad_norm=5.142672538757324, loss=4.389681816101074 -I0512 21:36:57.497771 139687754716928 logging_writer.py:48] [16100] global_step=16100, grad_norm=4.930809020996094, loss=4.163058280944824 -I0512 21:37:35.005177 139687763109632 logging_writer.py:48] [16200] global_step=16200, grad_norm=3.6823551654815674, loss=4.166629791259766 -I0512 21:38:11.813781 139687754716928 logging_writer.py:48] [16300] global_step=16300, grad_norm=3.59070086479187, loss=4.139599800109863 -I0512 21:38:48.946840 139687763109632 logging_writer.py:48] [16400] global_step=16400, grad_norm=6.197380065917969, loss=4.2903289794921875 -I0512 21:39:26.230979 139687754716928 logging_writer.py:48] [16500] global_step=16500, grad_norm=5.625287055969238, loss=4.281441688537598 -I0512 21:40:02.979725 139687763109632 logging_writer.py:48] [16600] global_step=16600, grad_norm=4.578366756439209, loss=4.142776012420654 -I0512 21:40:39.835848 139687754716928 logging_writer.py:48] [16700] global_step=16700, grad_norm=3.791961193084717, loss=4.329822540283203 -I0512 21:41:17.310419 139687763109632 logging_writer.py:48] [16800] global_step=16800, grad_norm=3.7367794513702393, loss=4.252202987670898 -I0512 21:41:54.046045 139687754716928 logging_writer.py:48] [16900] global_step=16900, grad_norm=3.255845069885254, loss=4.158737659454346 -I0512 21:42:31.148539 139687763109632 logging_writer.py:48] [17000] global_step=17000, grad_norm=3.4928836822509766, loss=4.172212600708008 -I0512 21:43:08.376846 139687754716928 logging_writer.py:48] [17100] global_step=17100, grad_norm=4.996946334838867, loss=4.104815483093262 -I0512 21:43:45.146441 139687763109632 logging_writer.py:48] [17200] global_step=17200, grad_norm=3.4076035022735596, loss=4.154796600341797 -I0512 21:44:22.012897 139687754716928 logging_writer.py:48] [17300] global_step=17300, grad_norm=4.8138298988342285, loss=4.170757293701172 -I0512 21:44:59.638448 139687763109632 logging_writer.py:48] [17400] global_step=17400, grad_norm=3.3729805946350098, loss=4.182557106018066 -I0512 21:45:36.419302 139687754716928 logging_writer.py:48] [17500] global_step=17500, grad_norm=5.335815906524658, loss=4.226658344268799 -I0512 21:46:13.225486 139687763109632 logging_writer.py:48] [17600] global_step=17600, grad_norm=3.7623515129089355, loss=4.210519313812256 -I0512 21:46:50.750223 139687754716928 logging_writer.py:48] [17700] global_step=17700, grad_norm=4.07235860824585, loss=4.233376502990723 -I0512 21:47:27.477945 139687763109632 logging_writer.py:48] [17800] global_step=17800, grad_norm=3.9344775676727295, loss=4.257962703704834 -I0512 21:48:04.415408 139687754716928 logging_writer.py:48] [17900] global_step=17900, grad_norm=2.547037363052368, loss=4.0837202072143555 -I0512 21:48:41.987360 139687763109632 logging_writer.py:48] [18000] global_step=18000, grad_norm=8.017416954040527, loss=4.349206447601318 -I0512 21:49:18.744427 139687754716928 logging_writer.py:48] [18100] global_step=18100, grad_norm=5.6542134284973145, loss=4.268446922302246 -I0512 21:49:55.546835 139687763109632 logging_writer.py:48] [18200] global_step=18200, grad_norm=4.561398029327393, loss=4.034845352172852 -I0512 21:50:33.073621 139687754716928 logging_writer.py:48] [18300] global_step=18300, grad_norm=3.614933967590332, loss=4.1169233322143555 -I0512 21:51:09.853354 139687763109632 logging_writer.py:48] [18400] global_step=18400, grad_norm=5.384772300720215, loss=4.1801300048828125 -I0512 21:51:46.599603 139687754716928 logging_writer.py:48] [18500] global_step=18500, grad_norm=4.814983367919922, loss=4.084420204162598 -I0512 21:52:24.117339 139687763109632 logging_writer.py:48] [18600] global_step=18600, grad_norm=3.638984203338623, loss=4.1951799392700195 -I0512 21:53:00.901017 139687754716928 logging_writer.py:48] [18700] global_step=18700, grad_norm=2.190462350845337, loss=4.089112281799316 -I0512 21:53:37.734268 139687763109632 logging_writer.py:48] [18800] global_step=18800, grad_norm=2.1416358947753906, loss=4.068811416625977 -I0512 21:54:15.008185 139687754716928 logging_writer.py:48] [18900] global_step=18900, grad_norm=4.326211929321289, loss=4.161866664886475 -I0512 21:54:52.095619 139687763109632 logging_writer.py:48] [19000] global_step=19000, grad_norm=5.230624198913574, loss=4.149301052093506 -I0512 21:55:28.859814 139687754716928 logging_writer.py:48] [19100] global_step=19100, grad_norm=5.5730180740356445, loss=4.181936264038086 -I0512 21:56:06.417297 139687763109632 logging_writer.py:48] [19200] global_step=19200, grad_norm=5.422348976135254, loss=4.191165924072266 -I0512 21:56:43.239219 139687754716928 logging_writer.py:48] [19300] global_step=19300, grad_norm=4.952305316925049, loss=4.2868194580078125 -I0512 21:57:20.003896 139687763109632 logging_writer.py:48] [19400] global_step=19400, grad_norm=4.717532634735107, loss=4.112893581390381 -I0512 21:57:57.627012 139687754716928 logging_writer.py:48] [19500] global_step=19500, grad_norm=6.789829254150391, loss=4.218955039978027 -I0512 21:58:34.472833 139687763109632 logging_writer.py:48] [19600] global_step=19600, grad_norm=5.659283638000488, loss=4.224696159362793 -I0512 21:59:11.276229 139687754716928 logging_writer.py:48] [19700] global_step=19700, grad_norm=7.346611499786377, loss=4.184014320373535 -I0512 21:59:48.896883 139687763109632 logging_writer.py:48] [19800] global_step=19800, grad_norm=4.838632106781006, loss=4.325345516204834 -I0512 22:00:18.625651 139903809070272 spec.py:333] Evaluating on the training split. -I0512 22:00:34.938104 139903809070272 spec.py:346] Evaluating on the validation split. -I0512 22:01:47.733700 139903809070272 spec.py:363] Evaluating on the test split. -I0512 22:01:48.626481 139903809070272 submission_runner.py:516] Time since start: 8452.70s, Step: 19882, {'train/accuracy': Array(0.00348772, dtype=float32), 'train/loss': Array(8.590018, dtype=float32), 'validation/accuracy': Array(0.00326, dtype=float32), 'validation/loss': Array(8.605198, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0032, dtype=float32), 'test/loss': Array(8.626876, dtype=float32), 'test/num_examples': 10000, 'score': 8046.917476177216, 'total_duration': 8452.700548648834, 'accumulated_submission_time': 8046.917476177216, 'accumulated_eval_time': 405.47107315063477, 'accumulated_logging_time': 0.136307954788208} -I0512 22:01:48.650059 139687754716928 logging_writer.py:48] [19882] accumulated_eval_time=405.471, accumulated_logging_time=0.136308, accumulated_submission_time=8046.92, global_step=19882, preemption_count=0, score=8046.92, test/accuracy=0.003200000151991844, test/loss=8.626875877380371, test/num_examples=10000, total_duration=8452.7, train/accuracy=0.0034877231810241938, train/loss=8.590018272399902, validation/accuracy=0.0032599999103695154, validation/loss=8.60519790649414, validation/num_examples=50000 -I0512 22:01:55.604315 139687763109632 logging_writer.py:48] [19900] global_step=19900, grad_norm=8.406874656677246, loss=4.280713081359863 -I0512 22:02:32.344230 139687754716928 logging_writer.py:48] [20000] global_step=20000, grad_norm=3.053067445755005, loss=4.073483943939209 -I0512 22:03:09.892919 139687763109632 logging_writer.py:48] [20100] global_step=20100, grad_norm=9.283489227294922, loss=4.2329325675964355 -I0512 22:03:46.740668 139687754716928 logging_writer.py:48] [20200] global_step=20200, grad_norm=5.709168434143066, loss=4.194241523742676 -I0512 22:04:23.484967 139687763109632 logging_writer.py:48] [20300] global_step=20300, grad_norm=6.525169849395752, loss=4.078738212585449 -I0512 22:05:01.164392 139687754716928 logging_writer.py:48] [20400] global_step=20400, grad_norm=3.783874750137329, loss=4.030940055847168 -I0512 22:05:37.945447 139687763109632 logging_writer.py:48] [20500] global_step=20500, grad_norm=3.389150857925415, loss=4.022976398468018 -I0512 22:06:15.083303 139687754716928 logging_writer.py:48] [20600] global_step=20600, grad_norm=4.952849864959717, loss=4.132309913635254 -I0512 22:06:52.302178 139687763109632 logging_writer.py:48] [20700] global_step=20700, grad_norm=4.070734024047852, loss=3.974364995956421 -I0512 22:07:29.091099 139687754716928 logging_writer.py:48] [20800] global_step=20800, grad_norm=5.434362411499023, loss=4.146421432495117 -I0512 22:08:05.895843 139687763109632 logging_writer.py:48] [20900] global_step=20900, grad_norm=3.2595744132995605, loss=3.9693593978881836 -I0512 22:08:43.411136 139687754716928 logging_writer.py:48] [21000] global_step=21000, grad_norm=6.054633617401123, loss=4.160525798797607 -I0512 22:09:20.173166 139687763109632 logging_writer.py:48] [21100] global_step=21100, grad_norm=9.128397941589355, loss=4.139852523803711 -I0512 22:09:57.222661 139687754716928 logging_writer.py:48] [21200] global_step=21200, grad_norm=2.679887294769287, loss=4.096500873565674 -I0512 22:10:34.504651 139687763109632 logging_writer.py:48] [21300] global_step=21300, grad_norm=7.4985671043396, loss=4.174956798553467 -I0512 22:11:11.210394 139687754716928 logging_writer.py:48] [21400] global_step=21400, grad_norm=9.177841186523438, loss=4.279783725738525 -I0512 22:11:47.983658 139687763109632 logging_writer.py:48] [21500] global_step=21500, grad_norm=5.297593116760254, loss=4.127745628356934 -I0512 22:12:25.446873 139687754716928 logging_writer.py:48] [21600] global_step=21600, grad_norm=3.7307286262512207, loss=4.087367057800293 -I0512 22:13:02.275417 139687763109632 logging_writer.py:48] [21700] global_step=21700, grad_norm=5.655477046966553, loss=4.020288467407227 -I0512 22:13:39.091591 139687754716928 logging_writer.py:48] [21800] global_step=21800, grad_norm=35.790287017822266, loss=4.108758449554443 -I0512 22:14:16.708454 139687763109632 logging_writer.py:48] [21900] global_step=21900, grad_norm=3.843329906463623, loss=4.021250247955322 -I0512 22:14:53.456973 139687754716928 logging_writer.py:48] [22000] global_step=22000, grad_norm=5.458951950073242, loss=4.042703151702881 -I0512 22:15:30.254833 139687763109632 logging_writer.py:48] [22100] global_step=22100, grad_norm=2.7214105129241943, loss=3.9247922897338867 -I0512 22:16:07.798324 139687754716928 logging_writer.py:48] [22200] global_step=22200, grad_norm=2.6311631202697754, loss=3.9167001247406006 -I0512 22:16:44.520392 139687763109632 logging_writer.py:48] [22300] global_step=22300, grad_norm=3.7071094512939453, loss=4.010440349578857 -I0512 22:17:21.316984 139687754716928 logging_writer.py:48] [22400] global_step=22400, grad_norm=3.873833179473877, loss=3.9798355102539062 -I0512 22:17:58.863017 139687763109632 logging_writer.py:48] [22500] global_step=22500, grad_norm=8.183263778686523, loss=4.199581146240234 -I0512 22:18:35.668168 139687754716928 logging_writer.py:48] [22600] global_step=22600, grad_norm=4.083929061889648, loss=4.1011247634887695 -I0512 22:19:12.486654 139687763109632 logging_writer.py:48] [22700] global_step=22700, grad_norm=6.538111209869385, loss=3.981410026550293 -I0512 22:19:50.123474 139687754716928 logging_writer.py:48] [22800] global_step=22800, grad_norm=10.706646919250488, loss=4.174378395080566 -I0512 22:20:26.887380 139687763109632 logging_writer.py:48] [22900] global_step=22900, grad_norm=4.356204509735107, loss=4.07528829574585 -I0512 22:21:03.667746 139687754716928 logging_writer.py:48] [23000] global_step=23000, grad_norm=7.315605640411377, loss=3.974761724472046 -I0512 22:21:41.190844 139687763109632 logging_writer.py:48] [23100] global_step=23100, grad_norm=9.27894115447998, loss=4.069298267364502 -I0512 22:22:17.953858 139687754716928 logging_writer.py:48] [23200] global_step=23200, grad_norm=4.462464332580566, loss=3.9956350326538086 -I0512 22:22:54.665745 139687763109632 logging_writer.py:48] [23300] global_step=23300, grad_norm=6.267051696777344, loss=4.070705413818359 -I0512 22:23:32.281850 139687754716928 logging_writer.py:48] [23400] global_step=23400, grad_norm=3.6902379989624023, loss=4.033260822296143 -I0512 22:24:09.110009 139687763109632 logging_writer.py:48] [23500] global_step=23500, grad_norm=3.887242317199707, loss=3.962902307510376 -I0512 22:24:45.932952 139687754716928 logging_writer.py:48] [23600] global_step=23600, grad_norm=12.84101676940918, loss=4.23712158203125 -I0512 22:25:23.249515 139687763109632 logging_writer.py:48] [23700] global_step=23700, grad_norm=3.023109197616577, loss=3.99177622795105 -I0512 22:26:00.302374 139687754716928 logging_writer.py:48] [23800] global_step=23800, grad_norm=4.021589279174805, loss=3.9973108768463135 -I0512 22:26:37.093438 139687763109632 logging_writer.py:48] [23900] global_step=23900, grad_norm=11.097367286682129, loss=4.028170585632324 -I0512 22:27:14.684562 139687754716928 logging_writer.py:48] [24000] global_step=24000, grad_norm=5.87770414352417, loss=4.119693756103516 -I0512 22:27:51.469004 139687763109632 logging_writer.py:48] [24100] global_step=24100, grad_norm=6.8114519119262695, loss=3.992546558380127 -I0512 22:28:28.197327 139687754716928 logging_writer.py:48] [24200] global_step=24200, grad_norm=4.102545261383057, loss=3.9565720558166504 -I0512 22:29:05.830046 139687763109632 logging_writer.py:48] [24300] global_step=24300, grad_norm=12.081990242004395, loss=8.69340705871582 -I0512 22:29:42.647812 139687754716928 logging_writer.py:48] [24400] global_step=24400, grad_norm=3.1757678985595703, loss=4.020027160644531 -I0512 22:30:19.484155 139687763109632 logging_writer.py:48] [24500] global_step=24500, grad_norm=7.491662502288818, loss=4.034582614898682 -I0512 22:30:57.080311 139687754716928 logging_writer.py:48] [24600] global_step=24600, grad_norm=5.102310657501221, loss=4.033141613006592 -I0512 22:31:33.808819 139687763109632 logging_writer.py:48] [24700] global_step=24700, grad_norm=4.769066333770752, loss=4.0572919845581055 -I0512 22:32:10.537499 139687754716928 logging_writer.py:48] [24800] global_step=24800, grad_norm=5.653899192810059, loss=3.8886618614196777 -I0512 22:32:48.084204 139687763109632 logging_writer.py:48] [24900] global_step=24900, grad_norm=7.136525630950928, loss=4.245449066162109 -I0512 22:33:24.893081 139687754716928 logging_writer.py:48] [25000] global_step=25000, grad_norm=4.1434855461120605, loss=4.0414252281188965 -I0512 22:34:01.694511 139687763109632 logging_writer.py:48] [25100] global_step=25100, grad_norm=5.5106520652771, loss=4.079558372497559 -I0512 22:34:39.196274 139687754716928 logging_writer.py:48] [25200] global_step=25200, grad_norm=5.802320957183838, loss=4.017122268676758 -I0512 22:35:04.664408 139903809070272 spec.py:333] Evaluating on the training split. -I0512 22:35:16.771551 139903809070272 spec.py:346] Evaluating on the validation split. -I0512 22:36:23.093469 139903809070272 spec.py:363] Evaluating on the test split. -I0512 22:36:23.986086 139903809070272 submission_runner.py:516] Time since start: 10528.06s, Step: 25270, {'train/accuracy': Array(0.00209263, dtype=float32), 'train/loss': Array(8.377973, dtype=float32), 'validation/accuracy': Array(0.00232, dtype=float32), 'validation/loss': Array(8.369723, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0025, dtype=float32), 'test/loss': Array(8.380475, dtype=float32), 'test/num_examples': 10000, 'score': 10042.881111860275, 'total_duration': 10528.060062170029, 'accumulated_submission_time': 10042.881111860275, 'accumulated_eval_time': 484.79101037979126, 'accumulated_logging_time': 0.16821861267089844} -I0512 22:36:24.010486 139687763109632 logging_writer.py:48] [25270] accumulated_eval_time=484.791, accumulated_logging_time=0.168219, accumulated_submission_time=10042.9, global_step=25270, preemption_count=0, score=10042.9, test/accuracy=0.002500000176951289, test/loss=8.380475044250488, test/num_examples=10000, total_duration=10528.1, train/accuracy=0.0020926338620483875, train/loss=8.377972602844238, validation/accuracy=0.002319999970495701, validation/loss=8.369723320007324, validation/num_examples=50000 -I0512 22:36:35.367029 139687754716928 logging_writer.py:48] [25300] global_step=25300, grad_norm=5.0090413093566895, loss=4.075660705566406 -I0512 22:37:28.105850 139687763109632 logging_writer.py:48] [25400] global_step=25400, grad_norm=4.783086776733398, loss=4.096097946166992 -I0512 22:38:09.866961 139687754716928 logging_writer.py:48] [25500] global_step=25500, grad_norm=4.429019927978516, loss=3.8388516902923584 -I0512 22:38:50.563983 139687763109632 logging_writer.py:48] [25600] global_step=25600, grad_norm=3.5773208141326904, loss=3.9284706115722656 -I0512 22:39:35.330314 139687754716928 logging_writer.py:48] [25700] global_step=25700, grad_norm=5.636041641235352, loss=3.854092597961426 -I0512 22:40:16.566765 139687763109632 logging_writer.py:48] [25800] global_step=25800, grad_norm=84.76251983642578, loss=4.006638050079346 -I0512 22:40:57.590240 139687754716928 logging_writer.py:48] [25900] global_step=25900, grad_norm=4.267861366271973, loss=3.9146504402160645 -I0512 22:41:38.352638 139687763109632 logging_writer.py:48] [26000] global_step=26000, grad_norm=9.864656448364258, loss=4.1194987297058105 -I0512 22:42:19.836794 139687754716928 logging_writer.py:48] [26100] global_step=26100, grad_norm=5.344272613525391, loss=4.043724536895752 -I0512 22:43:00.610912 139687763109632 logging_writer.py:48] [26200] global_step=26200, grad_norm=6.116037368774414, loss=4.044600009918213 -I0512 22:43:49.363410 139687754716928 logging_writer.py:48] [26300] global_step=26300, grad_norm=6.744386196136475, loss=4.144931793212891 -I0512 22:44:30.547307 139687763109632 logging_writer.py:48] [26400] global_step=26400, grad_norm=4.537485122680664, loss=4.0968852043151855 -I0512 22:45:11.630976 139687754716928 logging_writer.py:48] [26500] global_step=26500, grad_norm=4.212287425994873, loss=3.9606616497039795 -I0512 22:45:52.413232 139687763109632 logging_writer.py:48] [26600] global_step=26600, grad_norm=5.0336594581604, loss=4.035872459411621 -I0512 22:46:34.010747 139687754716928 logging_writer.py:48] [26700] global_step=26700, grad_norm=3.8216371536254883, loss=3.9123623371124268 -I0512 22:47:14.758947 139687763109632 logging_writer.py:48] [26800] global_step=26800, grad_norm=5.373470783233643, loss=3.977769374847412 -I0512 22:47:55.585614 139687754716928 logging_writer.py:48] [26900] global_step=26900, grad_norm=4.339493274688721, loss=3.8900046348571777 -I0512 22:48:45.273163 139687763109632 logging_writer.py:48] [27000] global_step=27000, grad_norm=7.179563045501709, loss=4.071037292480469 -I0512 22:49:26.519295 139687754716928 logging_writer.py:48] [27100] global_step=27100, grad_norm=5.7475666999816895, loss=4.146737098693848 -I0512 22:50:07.425355 139687763109632 logging_writer.py:48] [27200] global_step=27200, grad_norm=3.610961437225342, loss=3.8999686241149902 -I0512 22:50:52.832624 139687754716928 logging_writer.py:48] [27300] global_step=27300, grad_norm=5.5209550857543945, loss=3.945978879928589 -I0512 22:51:33.976927 139687763109632 logging_writer.py:48] [27400] global_step=27400, grad_norm=10.086233139038086, loss=4.044466972351074 -I0512 22:52:14.863880 139687754716928 logging_writer.py:48] [27500] global_step=27500, grad_norm=5.707131385803223, loss=4.130337715148926 -I0512 22:52:56.131576 139687763109632 logging_writer.py:48] [27600] global_step=27600, grad_norm=4.818155288696289, loss=4.052228927612305 -I0512 22:53:37.392756 139687754716928 logging_writer.py:48] [27700] global_step=27700, grad_norm=6.16942024230957, loss=4.108067989349365 -I0512 22:54:18.261059 139687763109632 logging_writer.py:48] [27800] global_step=27800, grad_norm=10.649847984313965, loss=4.188987731933594 -I0512 22:54:59.624612 139687754716928 logging_writer.py:48] [27900] global_step=27900, grad_norm=5.7529473304748535, loss=3.971198081970215 -I0512 22:55:40.673037 139687763109632 logging_writer.py:48] [28000] global_step=28000, grad_norm=3.9842119216918945, loss=3.9278066158294678 -I0512 22:56:21.499412 139687754716928 logging_writer.py:48] [28100] global_step=28100, grad_norm=5.341738700866699, loss=3.795717477798462 -I0512 22:57:02.873680 139687763109632 logging_writer.py:48] [28200] global_step=28200, grad_norm=5.5505146980285645, loss=3.8924903869628906 -I0512 22:57:44.184056 139687754716928 logging_writer.py:48] [28300] global_step=28300, grad_norm=3.417536497116089, loss=3.962368965148926 -I0512 22:58:25.076368 139687763109632 logging_writer.py:48] [28400] global_step=28400, grad_norm=3.3096542358398438, loss=4.052375793457031 -I0512 22:59:06.479977 139687754716928 logging_writer.py:48] [28500] global_step=28500, grad_norm=3.623269557952881, loss=3.8696789741516113 -I0512 22:59:47.406164 139687763109632 logging_writer.py:48] [28600] global_step=28600, grad_norm=12.655169486999512, loss=4.1518096923828125 -I0512 23:00:32.783698 139687754716928 logging_writer.py:48] [28700] global_step=28700, grad_norm=4.246352195739746, loss=4.0519328117370605 -I0512 23:01:14.078061 139687763109632 logging_writer.py:48] [28800] global_step=28800, grad_norm=6.436347484588623, loss=3.9160284996032715 -I0512 23:01:55.264981 139687754716928 logging_writer.py:48] [28900] global_step=28900, grad_norm=3.4846153259277344, loss=3.920990467071533 -I0512 23:02:36.077592 139687763109632 logging_writer.py:48] [29000] global_step=29000, grad_norm=3.8023717403411865, loss=3.944115161895752 -I0512 23:03:17.413938 139687754716928 logging_writer.py:48] [29100] global_step=29100, grad_norm=4.863331317901611, loss=3.846749782562256 -I0512 23:04:02.761474 139687763109632 logging_writer.py:48] [29200] global_step=29200, grad_norm=6.018075466156006, loss=4.077815532684326 -I0512 23:04:43.638253 139687754716928 logging_writer.py:48] [29300] global_step=29300, grad_norm=3.918243646621704, loss=3.97365665435791 -I0512 23:05:25.049821 139687763109632 logging_writer.py:48] [29400] global_step=29400, grad_norm=4.738921165466309, loss=4.004408359527588 -I0512 23:06:06.367094 139687754716928 logging_writer.py:48] [29500] global_step=29500, grad_norm=21.797780990600586, loss=3.8951947689056396 -I0512 23:06:47.352305 139687763109632 logging_writer.py:48] [29600] global_step=29600, grad_norm=5.425933837890625, loss=4.175122261047363 -I0512 23:07:28.825625 139687754716928 logging_writer.py:48] [29700] global_step=29700, grad_norm=3.7484278678894043, loss=3.8656957149505615 -I0512 23:08:14.445381 139687763109632 logging_writer.py:48] [29800] global_step=29800, grad_norm=7.528026103973389, loss=4.079552173614502 -I0512 23:08:52.422938 139687754716928 logging_writer.py:48] [29900] global_step=29900, grad_norm=17.990604400634766, loss=4.018243312835693 -I0512 23:09:32.439730 139687763109632 logging_writer.py:48] [30000] global_step=30000, grad_norm=5.313564300537109, loss=3.7454404830932617 -I0512 23:09:40.074586 139903809070272 spec.py:333] Evaluating on the training split. -I0512 23:09:55.746531 139903809070272 spec.py:346] Evaluating on the validation split. -I0512 23:10:40.805971 139903809070272 spec.py:363] Evaluating on the test split. -I0512 23:10:41.702571 139903809070272 submission_runner.py:516] Time since start: 12585.78s, Step: 30022, {'train/accuracy': Array(0.00233179, dtype=float32), 'train/loss': Array(8.121338, dtype=float32), 'validation/accuracy': Array(0.00188, dtype=float32), 'validation/loss': Array(8.11584, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(8.119603, dtype=float32), 'test/num_examples': 10000, 'score': 12038.886108875275, 'total_duration': 12585.776431322098, 'accumulated_submission_time': 12038.886108875275, 'accumulated_eval_time': 546.4171376228333, 'accumulated_logging_time': 0.2008829116821289} -I0512 23:10:41.732445 139687754716928 logging_writer.py:48] [30022] accumulated_eval_time=546.417, accumulated_logging_time=0.200883, accumulated_submission_time=12038.9, global_step=30022, preemption_count=0, score=12038.9, test/accuracy=0.0019000000320374966, test/loss=8.119603157043457, test/num_examples=10000, total_duration=12585.8, train/accuracy=0.0023317921441048384, train/loss=8.121337890625, validation/accuracy=0.001879999996162951, validation/loss=8.115839958190918, validation/num_examples=50000 -I0512 23:11:19.448091 139687763109632 logging_writer.py:48] [30100] global_step=30100, grad_norm=5.670923709869385, loss=4.107810974121094 -I0512 23:12:00.356763 139687754716928 logging_writer.py:48] [30200] global_step=30200, grad_norm=4.054909706115723, loss=3.872443437576294 -I0512 23:12:41.680143 139687763109632 logging_writer.py:48] [30300] global_step=30300, grad_norm=3.2703683376312256, loss=3.8781092166900635 -I0512 23:13:22.914584 139687754716928 logging_writer.py:48] [30400] global_step=30400, grad_norm=3.773303508758545, loss=3.8501124382019043 -I0512 23:14:03.830232 139687763109632 logging_writer.py:48] [30500] global_step=30500, grad_norm=3.8438799381256104, loss=3.8641955852508545 -I0512 23:14:46.914761 139687754716928 logging_writer.py:48] [30600] global_step=30600, grad_norm=9.286420822143555, loss=4.028032302856445 -I0512 23:15:29.050661 139687763109632 logging_writer.py:48] [30700] global_step=30700, grad_norm=7.415320873260498, loss=3.9021546840667725 -I0512 23:16:10.229323 139687754716928 logging_writer.py:48] [30800] global_step=30800, grad_norm=2.6745450496673584, loss=3.8964734077453613 -I0512 23:16:51.579301 139687763109632 logging_writer.py:48] [30900] global_step=30900, grad_norm=6.043990135192871, loss=3.979768753051758 -I0512 23:17:32.841269 139687754716928 logging_writer.py:48] [31000] global_step=31000, grad_norm=3.903710126876831, loss=3.9740214347839355 -I0512 23:18:13.762779 139687763109632 logging_writer.py:48] [31100] global_step=31100, grad_norm=6.759665012359619, loss=3.968963861465454 -I0512 23:18:55.170271 139687754716928 logging_writer.py:48] [31200] global_step=31200, grad_norm=5.356132507324219, loss=3.872446060180664 -I0512 23:19:36.084198 139687763109632 logging_writer.py:48] [31300] global_step=31300, grad_norm=4.229129791259766, loss=3.950345993041992 -I0512 23:20:17.281273 139687754716928 logging_writer.py:48] [31400] global_step=31400, grad_norm=5.462503433227539, loss=3.923957109451294 -I0512 23:20:58.579639 139687763109632 logging_writer.py:48] [31500] global_step=31500, grad_norm=3.5583322048187256, loss=3.732351779937744 -I0512 23:21:43.906646 139687754716928 logging_writer.py:48] [31600] global_step=31600, grad_norm=3.913525104522705, loss=3.8800182342529297 -I0512 23:22:24.734016 139687763109632 logging_writer.py:48] [31700] global_step=31700, grad_norm=6.797950267791748, loss=3.906998634338379 -I0512 23:23:06.135627 139687754716928 logging_writer.py:48] [31800] global_step=31800, grad_norm=43.56948471069336, loss=4.001133918762207 -I0512 23:23:47.296673 139687763109632 logging_writer.py:48] [31900] global_step=31900, grad_norm=3.223294734954834, loss=3.8622961044311523 -I0512 23:24:28.170179 139687754716928 logging_writer.py:48] [32000] global_step=32000, grad_norm=9.97351360321045, loss=4.068391799926758 -I0512 23:25:09.542005 139687763109632 logging_writer.py:48] [32100] global_step=32100, grad_norm=5.164907932281494, loss=3.9653573036193848 -I0512 23:25:50.821169 139687754716928 logging_writer.py:48] [32200] global_step=32200, grad_norm=6.021162033081055, loss=3.937924861907959 -I0512 23:26:31.678599 139687763109632 logging_writer.py:48] [32300] global_step=32300, grad_norm=4.305586338043213, loss=3.8091025352478027 -I0512 23:27:13.119444 139687754716928 logging_writer.py:48] [32400] global_step=32400, grad_norm=4.115222930908203, loss=3.907789945602417 -I0512 23:27:54.305066 139687763109632 logging_writer.py:48] [32500] global_step=32500, grad_norm=3.7513461112976074, loss=3.657170295715332 -I0512 23:28:35.286437 139687754716928 logging_writer.py:48] [32600] global_step=32600, grad_norm=4.7049880027771, loss=3.954026699066162 -I0512 23:29:16.633292 139687763109632 logging_writer.py:48] [32700] global_step=32700, grad_norm=3.884251594543457, loss=3.774423122406006 -I0512 23:29:57.687087 139687754716928 logging_writer.py:48] [32800] global_step=32800, grad_norm=12.891077041625977, loss=4.078519344329834 -I0512 23:30:38.684149 139687763109632 logging_writer.py:48] [32900] global_step=32900, grad_norm=4.596135139465332, loss=3.9260666370391846 -I0512 23:31:20.060675 139687754716928 logging_writer.py:48] [33000] global_step=33000, grad_norm=5.231578826904297, loss=3.926236152648926 -I0512 23:32:05.527028 139687763109632 logging_writer.py:48] [33100] global_step=33100, grad_norm=3.33107852935791, loss=3.8254241943359375 -I0512 23:32:46.453140 139687754716928 logging_writer.py:48] [33200] global_step=33200, grad_norm=15.77064323425293, loss=4.12331485748291 -I0512 23:33:27.811812 139687763109632 logging_writer.py:48] [33300] global_step=33300, grad_norm=5.116666793823242, loss=3.8323798179626465 -I0512 23:34:04.951390 139687754716928 logging_writer.py:48] [33400] global_step=33400, grad_norm=6.238740921020508, loss=4.100395679473877 -I0512 23:34:42.265621 139687763109632 logging_writer.py:48] [33500] global_step=33500, grad_norm=4.191104412078857, loss=3.987335681915283 -I0512 23:35:19.781349 139687754716928 logging_writer.py:48] [33600] global_step=33600, grad_norm=8.030035972595215, loss=3.9302358627319336 -I0512 23:35:57.134279 139687763109632 logging_writer.py:48] [33700] global_step=33700, grad_norm=29.971622467041016, loss=4.00416374206543 -I0512 23:36:34.107347 139687754716928 logging_writer.py:48] [33800] global_step=33800, grad_norm=4.8936591148376465, loss=3.9212613105773926 -I0512 23:37:11.545238 139687763109632 logging_writer.py:48] [33900] global_step=33900, grad_norm=5.059073448181152, loss=4.1556830406188965 -I0512 23:37:48.798057 139687754716928 logging_writer.py:48] [34000] global_step=34000, grad_norm=6.436087131500244, loss=3.9019112586975098 -I0512 23:38:25.767031 139687763109632 logging_writer.py:48] [34100] global_step=34100, grad_norm=8.254064559936523, loss=3.873213291168213 -I0512 23:39:03.228880 139687754716928 logging_writer.py:48] [34200] global_step=34200, grad_norm=5.018177509307861, loss=3.990841865539551 -I0512 23:39:40.488522 139687763109632 logging_writer.py:48] [34300] global_step=34300, grad_norm=4.086965084075928, loss=3.837604522705078 -I0512 23:40:17.434119 139687754716928 logging_writer.py:48] [34400] global_step=34400, grad_norm=5.8384528160095215, loss=3.918893814086914 -I0512 23:40:54.796059 139687763109632 logging_writer.py:48] [34500] global_step=34500, grad_norm=3.3457095623016357, loss=3.807539939880371 -I0512 23:41:32.082294 139687754716928 logging_writer.py:48] [34600] global_step=34600, grad_norm=4.995266437530518, loss=3.817324161529541 -I0512 23:42:09.081244 139687763109632 logging_writer.py:48] [34700] global_step=34700, grad_norm=7.972383499145508, loss=3.9199867248535156 -I0512 23:42:46.450663 139687754716928 logging_writer.py:48] [34800] global_step=34800, grad_norm=9.742805480957031, loss=4.074171543121338 -I0512 23:43:23.610672 139687763109632 logging_writer.py:48] [34900] global_step=34900, grad_norm=3.49131441116333, loss=3.8093838691711426 -I0512 23:43:57.748115 139903809070272 spec.py:333] Evaluating on the training split. -I0512 23:44:14.923178 139903809070272 spec.py:346] Evaluating on the validation split. -I0512 23:45:04.108200 139903809070272 spec.py:363] Evaluating on the test split. -I0512 23:45:05.001738 139903809070272 submission_runner.py:516] Time since start: 14649.08s, Step: 34981, {'train/accuracy': Array(0.00185348, dtype=float32), 'train/loss': Array(8.153365, dtype=float32), 'validation/accuracy': Array(0.00166, dtype=float32), 'validation/loss': Array(8.154742, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.002, dtype=float32), 'test/loss': Array(8.15415, dtype=float32), 'test/num_examples': 10000, 'score': 14034.850373506546, 'total_duration': 14649.07592201233, 'accumulated_submission_time': 14034.850373506546, 'accumulated_eval_time': 613.6692273616791, 'accumulated_logging_time': 0.23974204063415527} -I0512 23:45:05.026824 139687754716928 logging_writer.py:48] [34981] accumulated_eval_time=613.669, accumulated_logging_time=0.239742, accumulated_submission_time=14034.9, global_step=34981, preemption_count=0, score=14034.9, test/accuracy=0.0020000000949949026, test/loss=8.154150009155273, test/num_examples=10000, total_duration=14649.1, train/accuracy=0.0018534756964072585, train/loss=8.153365135192871, validation/accuracy=0.0016599999507889152, validation/loss=8.154742240905762, validation/num_examples=50000 -I0512 23:45:18.101655 139687763109632 logging_writer.py:48] [35000] global_step=35000, grad_norm=3.7252049446105957, loss=3.818173408508301 -I0512 23:46:01.319994 139687754716928 logging_writer.py:48] [35100] global_step=35100, grad_norm=4.11764669418335, loss=3.84157133102417 -I0512 23:46:42.590371 139687763109632 logging_writer.py:48] [35200] global_step=35200, grad_norm=3.202796697616577, loss=3.7598161697387695 -I0512 23:47:23.616325 139687754716928 logging_writer.py:48] [35300] global_step=35300, grad_norm=3.2224950790405273, loss=3.97312068939209 -I0512 23:48:05.027643 139687763109632 logging_writer.py:48] [35400] global_step=35400, grad_norm=5.209256649017334, loss=3.9930968284606934 -I0512 23:48:46.707527 139687754716928 logging_writer.py:48] [35500] global_step=35500, grad_norm=6.290946006774902, loss=3.8703932762145996 -I0512 23:49:31.348678 139687763109632 logging_writer.py:48] [35600] global_step=35600, grad_norm=7.147085189819336, loss=3.7950210571289062 -I0512 23:50:16.810872 139687754716928 logging_writer.py:48] [35700] global_step=35700, grad_norm=3.325148344039917, loss=3.7409274578094482 -I0512 23:50:58.118310 139687763109632 logging_writer.py:48] [35800] global_step=35800, grad_norm=5.83747673034668, loss=3.85790753364563 -I0512 23:51:40.370658 139687754716928 logging_writer.py:48] [35900] global_step=35900, grad_norm=8.08106517791748, loss=4.094920635223389 -I0512 23:52:24.569780 139687763109632 logging_writer.py:48] [36000] global_step=36000, grad_norm=4.552894115447998, loss=3.827995538711548 -I0512 23:53:07.264368 139687754716928 logging_writer.py:48] [36100] global_step=36100, grad_norm=7.85722017288208, loss=3.9836642742156982 -I0512 23:53:50.224024 139687763109632 logging_writer.py:48] [36200] global_step=36200, grad_norm=4.099224090576172, loss=3.924478530883789 -I0512 23:54:31.755629 139687754716928 logging_writer.py:48] [36300] global_step=36300, grad_norm=5.83288049697876, loss=3.9188013076782227 -I0512 23:55:13.005635 139687763109632 logging_writer.py:48] [36400] global_step=36400, grad_norm=3.3009631633758545, loss=3.724641799926758 -I0512 23:55:53.944053 139687754716928 logging_writer.py:48] [36500] global_step=36500, grad_norm=6.3063249588012695, loss=3.824336051940918 -I0512 23:56:39.384067 139687763109632 logging_writer.py:48] [36600] global_step=36600, grad_norm=5.548352241516113, loss=4.018687725067139 -I0512 23:57:20.228551 139687754716928 logging_writer.py:48] [36700] global_step=36700, grad_norm=4.460649013519287, loss=3.9003400802612305 -I0512 23:58:05.644439 139687763109632 logging_writer.py:48] [36800] global_step=36800, grad_norm=4.213024139404297, loss=3.8723621368408203 -I0512 23:58:47.060394 139687754716928 logging_writer.py:48] [36900] global_step=36900, grad_norm=3.955843687057495, loss=3.7579007148742676 -I0512 23:59:27.957088 139687763109632 logging_writer.py:48] [37000] global_step=37000, grad_norm=3.433434247970581, loss=3.8330416679382324 -I0513 00:00:09.135447 139687754716928 logging_writer.py:48] [37100] global_step=37100, grad_norm=4.332891464233398, loss=3.802766799926758 -I0513 00:00:54.526563 139687763109632 logging_writer.py:48] [37200] global_step=37200, grad_norm=6.355403423309326, loss=3.822744846343994 -I0513 00:01:38.988969 139687754716928 logging_writer.py:48] [37300] global_step=37300, grad_norm=3.4918911457061768, loss=3.902833938598633 -I0513 00:02:20.426353 139687763109632 logging_writer.py:48] [37400] global_step=37400, grad_norm=4.3078789710998535, loss=3.9332375526428223 -I0513 00:03:01.915688 139687754716928 logging_writer.py:48] [37500] global_step=37500, grad_norm=5.115251541137695, loss=3.8975119590759277 -I0513 00:03:42.855021 139687763109632 logging_writer.py:48] [37600] global_step=37600, grad_norm=5.082361698150635, loss=3.741316080093384 -I0513 00:04:24.079688 139687754716928 logging_writer.py:48] [37700] global_step=37700, grad_norm=4.148218154907227, loss=3.720444917678833 -I0513 00:05:05.415965 139687763109632 logging_writer.py:48] [37800] global_step=37800, grad_norm=5.041350841522217, loss=3.996617317199707 -I0513 00:05:51.681115 139687754716928 logging_writer.py:48] [37900] global_step=37900, grad_norm=4.970189571380615, loss=3.7904515266418457 -I0513 00:06:33.017193 139687763109632 logging_writer.py:48] [38000] global_step=38000, grad_norm=4.6872968673706055, loss=3.916959762573242 -I0513 00:07:15.264746 139687754716928 logging_writer.py:48] [38100] global_step=38100, grad_norm=10.095328330993652, loss=3.8814010620117188 -I0513 00:08:00.332817 139687763109632 logging_writer.py:48] [38200] global_step=38200, grad_norm=4.3700947761535645, loss=3.7748427391052246 -I0513 00:08:42.566693 139687754716928 logging_writer.py:48] [38300] global_step=38300, grad_norm=7.689476013183594, loss=3.8940072059631348 -I0513 00:09:24.543385 139687763109632 logging_writer.py:48] [38400] global_step=38400, grad_norm=3.9452762603759766, loss=3.9027624130249023 -I0513 00:10:05.931041 139687754716928 logging_writer.py:48] [38500] global_step=38500, grad_norm=4.0301947593688965, loss=3.8300538063049316 -I0513 00:10:47.685419 139687763109632 logging_writer.py:48] [38600] global_step=38600, grad_norm=3.366262435913086, loss=3.7661118507385254 -I0513 00:11:29.657194 139687754716928 logging_writer.py:48] [38700] global_step=38700, grad_norm=4.849618434906006, loss=3.842303991317749 -I0513 00:12:11.043352 139687763109632 logging_writer.py:48] [38800] global_step=38800, grad_norm=4.880417823791504, loss=3.923431158065796 -I0513 00:12:52.760677 139687754716928 logging_writer.py:48] [38900] global_step=38900, grad_norm=5.403744697570801, loss=3.7762560844421387 -I0513 00:13:34.697524 139687763109632 logging_writer.py:48] [39000] global_step=39000, grad_norm=5.294404983520508, loss=3.7983107566833496 -I0513 00:14:16.170652 139687754716928 logging_writer.py:48] [39100] global_step=39100, grad_norm=5.625661373138428, loss=3.883674383163452 -I0513 00:14:57.900660 139687763109632 logging_writer.py:48] [39200] global_step=39200, grad_norm=8.314431190490723, loss=3.8155646324157715 -I0513 00:15:39.805950 139687754716928 logging_writer.py:48] [39300] global_step=39300, grad_norm=8.6240816116333, loss=3.7700130939483643 -I0513 00:16:21.173452 139687763109632 logging_writer.py:48] [39400] global_step=39400, grad_norm=4.902342319488525, loss=3.7117700576782227 -I0513 00:17:02.893278 139687754716928 logging_writer.py:48] [39500] global_step=39500, grad_norm=5.408238887786865, loss=3.8540539741516113 -I0513 00:17:44.843076 139687763109632 logging_writer.py:48] [39600] global_step=39600, grad_norm=4.661646366119385, loss=3.854827404022217 -I0513 00:18:21.486323 139903809070272 spec.py:333] Evaluating on the training split. -I0513 00:18:32.637863 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 00:19:41.470993 139903809070272 spec.py:363] Evaluating on the test split. -I0513 00:19:42.365323 139903809070272 submission_runner.py:516] Time since start: 16726.44s, Step: 39688, {'train/accuracy': Array(0.00203284, dtype=float32), 'train/loss': Array(8.241737, dtype=float32), 'validation/accuracy': Array(0.00196, dtype=float32), 'validation/loss': Array(8.220464, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0018, dtype=float32), 'test/loss': Array(8.218177, dtype=float32), 'test/num_examples': 10000, 'score': 16031.259686946869, 'total_duration': 16726.439479112625, 'accumulated_submission_time': 16031.259686946869, 'accumulated_eval_time': 694.5466666221619, 'accumulated_logging_time': 0.27304887771606445} -I0513 00:19:42.390630 139687754716928 logging_writer.py:48] [39688] accumulated_eval_time=694.547, accumulated_logging_time=0.273049, accumulated_submission_time=16031.3, global_step=39688, preemption_count=0, score=16031.3, test/accuracy=0.0018000000854954123, test/loss=8.21817684173584, test/num_examples=10000, total_duration=16726.4, train/accuracy=0.0020328443497419357, train/loss=8.241737365722656, validation/accuracy=0.0019600000232458115, validation/loss=8.220463752746582, validation/num_examples=50000 -I0513 00:19:47.141470 139687763109632 logging_writer.py:48] [39700] global_step=39700, grad_norm=5.738367080688477, loss=3.7730040550231934 -I0513 00:20:33.057861 139687754716928 logging_writer.py:48] [39800] global_step=39800, grad_norm=5.091378211975098, loss=3.8936028480529785 -I0513 00:21:14.673001 139687763109632 logging_writer.py:48] [39900] global_step=39900, grad_norm=3.8261427879333496, loss=3.8260934352874756 -I0513 00:21:55.776410 139687754716928 logging_writer.py:48] [40000] global_step=40000, grad_norm=6.349472522735596, loss=3.888667583465576 -I0513 00:22:37.424350 139687763109632 logging_writer.py:48] [40100] global_step=40100, grad_norm=2.558600664138794, loss=3.7873971462249756 -I0513 00:23:19.099745 139687754716928 logging_writer.py:48] [40200] global_step=40200, grad_norm=4.754927158355713, loss=3.7953648567199707 -I0513 00:24:00.328468 139687763109632 logging_writer.py:48] [40300] global_step=40300, grad_norm=3.9491238594055176, loss=3.857673406600952 -I0513 00:24:41.933254 139687754716928 logging_writer.py:48] [40400] global_step=40400, grad_norm=5.373605728149414, loss=3.8455991744995117 -I0513 00:25:23.655463 139687763109632 logging_writer.py:48] [40500] global_step=40500, grad_norm=3.3648273944854736, loss=3.807971715927124 -I0513 00:26:04.953991 139687754716928 logging_writer.py:48] [40600] global_step=40600, grad_norm=3.0873358249664307, loss=3.8526113033294678 -I0513 00:26:46.297162 139687763109632 logging_writer.py:48] [40700] global_step=40700, grad_norm=6.00152063369751, loss=4.032992839813232 -I0513 00:27:32.002857 139687754716928 logging_writer.py:48] [40800] global_step=40800, grad_norm=4.5797319412231445, loss=3.781627655029297 -I0513 00:28:17.379044 139687763109632 logging_writer.py:48] [40900] global_step=40900, grad_norm=5.601802825927734, loss=3.9036099910736084 -I0513 00:28:58.556667 139687754716928 logging_writer.py:48] [41000] global_step=41000, grad_norm=3.8403189182281494, loss=3.663151502609253 -I0513 00:29:40.282965 139687763109632 logging_writer.py:48] [41100] global_step=41100, grad_norm=5.629751205444336, loss=3.696437358856201 -I0513 00:30:21.346511 139687754716928 logging_writer.py:48] [41200] global_step=41200, grad_norm=5.469810485839844, loss=3.9539601802825928 -I0513 00:31:02.779274 139687763109632 logging_writer.py:48] [41300] global_step=41300, grad_norm=4.22484827041626, loss=3.8048253059387207 -I0513 00:31:44.296067 139687754716928 logging_writer.py:48] [41400] global_step=41400, grad_norm=3.8463854789733887, loss=3.8221073150634766 -I0513 00:32:25.260880 139687763109632 logging_writer.py:48] [41500] global_step=41500, grad_norm=3.970649480819702, loss=3.911717414855957 -I0513 00:33:11.615679 139687754716928 logging_writer.py:48] [41600] global_step=41600, grad_norm=4.668820381164551, loss=3.815760612487793 -I0513 00:33:53.612981 139687763109632 logging_writer.py:48] [41700] global_step=41700, grad_norm=3.5994646549224854, loss=3.812967300415039 -I0513 00:34:35.033764 139687754716928 logging_writer.py:48] [41800] global_step=41800, grad_norm=4.705358982086182, loss=3.715324878692627 -I0513 00:35:16.869094 139687763109632 logging_writer.py:48] [41900] global_step=41900, grad_norm=4.423020839691162, loss=3.82956600189209 -I0513 00:35:58.773392 139687754716928 logging_writer.py:48] [42000] global_step=42000, grad_norm=4.74512243270874, loss=3.9103710651397705 -I0513 00:36:40.177193 139687763109632 logging_writer.py:48] [42100] global_step=42100, grad_norm=4.49418306350708, loss=3.796219825744629 -I0513 00:37:21.918662 139687754716928 logging_writer.py:48] [42200] global_step=42200, grad_norm=6.2629265785217285, loss=3.689824104309082 -I0513 00:38:03.873128 139687763109632 logging_writer.py:48] [42300] global_step=42300, grad_norm=2.9851889610290527, loss=3.744157552719116 -I0513 00:38:45.342635 139687754716928 logging_writer.py:48] [42400] global_step=42400, grad_norm=3.915278673171997, loss=3.809074878692627 -I0513 00:39:27.030568 139687763109632 logging_writer.py:48] [42500] global_step=42500, grad_norm=4.015066146850586, loss=3.7186825275421143 -I0513 00:40:08.873977 139687754716928 logging_writer.py:48] [42600] global_step=42600, grad_norm=3.108705759048462, loss=3.7514994144439697 -I0513 00:40:50.281358 139687763109632 logging_writer.py:48] [42700] global_step=42700, grad_norm=8.181145668029785, loss=3.8451027870178223 -I0513 00:41:32.032207 139687754716928 logging_writer.py:48] [42800] global_step=42800, grad_norm=5.713146209716797, loss=3.819605588912964 -I0513 00:42:13.934782 139687763109632 logging_writer.py:48] [42900] global_step=42900, grad_norm=10.42215347290039, loss=3.7484965324401855 -I0513 00:42:59.890441 139687754716928 logging_writer.py:48] [43000] global_step=43000, grad_norm=4.037586212158203, loss=3.7140612602233887 -I0513 00:43:41.269757 139687763109632 logging_writer.py:48] [43100] global_step=43100, grad_norm=6.044254302978516, loss=4.029401779174805 -I0513 00:44:23.580536 139687754716928 logging_writer.py:48] [43200] global_step=43200, grad_norm=4.870251178741455, loss=3.7339015007019043 -I0513 00:45:04.987516 139687763109632 logging_writer.py:48] [43300] global_step=43300, grad_norm=4.949679374694824, loss=3.635988712310791 -I0513 00:45:46.761410 139687754716928 logging_writer.py:48] [43400] global_step=43400, grad_norm=4.766444683074951, loss=3.888214111328125 -I0513 00:46:28.624111 139687763109632 logging_writer.py:48] [43500] global_step=43500, grad_norm=5.873553276062012, loss=3.8929004669189453 -I0513 00:47:10.010315 139687754716928 logging_writer.py:48] [43600] global_step=43600, grad_norm=5.1403069496154785, loss=3.852839231491089 -I0513 00:47:51.769552 139687763109632 logging_writer.py:48] [43700] global_step=43700, grad_norm=3.9341909885406494, loss=3.816638946533203 -I0513 00:48:33.716391 139687754716928 logging_writer.py:48] [43800] global_step=43800, grad_norm=4.757513999938965, loss=3.769291877746582 -I0513 00:49:15.266456 139687763109632 logging_writer.py:48] [43900] global_step=43900, grad_norm=3.845001697540283, loss=3.7140095233917236 -I0513 00:49:57.147035 139687754716928 logging_writer.py:48] [44000] global_step=44000, grad_norm=2.83164381980896, loss=3.8117716312408447 -I0513 00:50:39.072417 139687763109632 logging_writer.py:48] [44100] global_step=44100, grad_norm=4.5005879402160645, loss=3.788820266723633 -I0513 00:51:20.460139 139687754716928 logging_writer.py:48] [44200] global_step=44200, grad_norm=3.7706799507141113, loss=3.802673816680908 -I0513 00:52:02.245436 139687763109632 logging_writer.py:48] [44300] global_step=44300, grad_norm=3.290562152862549, loss=3.7193961143493652 -I0513 00:52:44.163243 139687754716928 logging_writer.py:48] [44400] global_step=44400, grad_norm=4.760210037231445, loss=3.6966323852539062 -I0513 00:52:58.586481 139903809070272 spec.py:333] Evaluating on the training split. -I0513 00:53:09.884393 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 00:54:13.617699 139903809070272 spec.py:363] Evaluating on the test split. -I0513 00:54:14.512925 139903809070272 submission_runner.py:516] Time since start: 18798.59s, Step: 44428, {'train/accuracy': Array(0.00143495, dtype=float32), 'train/loss': Array(8.262915, dtype=float32), 'validation/accuracy': Array(0.00184, dtype=float32), 'validation/loss': Array(8.247698, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(8.245352, dtype=float32), 'test/num_examples': 10000, 'score': 18027.405270814896, 'total_duration': 18798.586426973343, 'accumulated_submission_time': 18027.405270814896, 'accumulated_eval_time': 770.4708964824677, 'accumulated_logging_time': 0.30670833587646484} -I0513 00:54:14.541291 139687763109632 logging_writer.py:48] [44428] accumulated_eval_time=770.471, accumulated_logging_time=0.306708, accumulated_submission_time=18027.4, global_step=44428, preemption_count=0, score=18027.4, test/accuracy=0.001500000013038516, test/loss=8.245351791381836, test/num_examples=10000, total_duration=18798.6, train/accuracy=0.001434948993846774, train/loss=8.262914657592773, validation/accuracy=0.0018399999244138598, validation/loss=8.247697830200195, validation/num_examples=50000 -I0513 00:54:42.823231 139687754716928 logging_writer.py:48] [44500] global_step=44500, grad_norm=4.495846271514893, loss=3.719083786010742 -I0513 00:55:26.128515 139687763109632 logging_writer.py:48] [44600] global_step=44600, grad_norm=3.3361692428588867, loss=3.7332382202148438 -I0513 00:56:08.416478 139687754716928 logging_writer.py:48] [44700] global_step=44700, grad_norm=6.68978214263916, loss=3.9469921588897705 -I0513 00:56:50.319611 139687763109632 logging_writer.py:48] [44800] global_step=44800, grad_norm=3.378696918487549, loss=3.8290510177612305 -I0513 00:57:32.010876 139687754716928 logging_writer.py:48] [44900] global_step=44900, grad_norm=4.061112880706787, loss=3.7271244525909424 -I0513 00:58:16.337429 139687763109632 logging_writer.py:48] [45000] global_step=45000, grad_norm=5.553672790527344, loss=3.852018356323242 -I0513 00:58:58.423820 139687754716928 logging_writer.py:48] [45100] global_step=45100, grad_norm=7.022552967071533, loss=3.7395622730255127 -I0513 00:59:40.505710 139687763109632 logging_writer.py:48] [45200] global_step=45200, grad_norm=4.597343444824219, loss=3.756193161010742 -I0513 01:00:22.278794 139687754716928 logging_writer.py:48] [45300] global_step=45300, grad_norm=5.633834362030029, loss=3.8174800872802734 -I0513 01:01:03.468956 139687763109632 logging_writer.py:48] [45400] global_step=45400, grad_norm=3.572523832321167, loss=3.7917628288269043 -I0513 01:01:48.287785 139687754716928 logging_writer.py:48] [45500] global_step=45500, grad_norm=6.70773983001709, loss=3.7050490379333496 -I0513 01:02:30.536043 139687763109632 logging_writer.py:48] [45600] global_step=45600, grad_norm=4.136594772338867, loss=3.893301248550415 -I0513 01:03:11.576071 139687754716928 logging_writer.py:48] [45700] global_step=45700, grad_norm=4.269682884216309, loss=3.755326271057129 -I0513 01:03:54.429070 139687763109632 logging_writer.py:48] [45800] global_step=45800, grad_norm=3.559001922607422, loss=3.6119606494903564 -I0513 01:04:38.287125 139687754716928 logging_writer.py:48] [45900] global_step=45900, grad_norm=3.980013608932495, loss=3.806084632873535 -I0513 01:05:19.310055 139687763109632 logging_writer.py:48] [46000] global_step=46000, grad_norm=5.036451816558838, loss=3.7876765727996826 -I0513 01:06:00.706825 139687754716928 logging_writer.py:48] [46100] global_step=46100, grad_norm=4.7088303565979, loss=3.690720558166504 -I0513 01:06:44.279980 139687763109632 logging_writer.py:48] [46200] global_step=46200, grad_norm=3.0130040645599365, loss=3.749631881713867 -I0513 01:07:27.347799 139687754716928 logging_writer.py:48] [46300] global_step=46300, grad_norm=5.657476902008057, loss=3.825711250305176 -I0513 01:08:13.574694 139687763109632 logging_writer.py:48] [46400] global_step=46400, grad_norm=3.141327381134033, loss=3.863184690475464 -I0513 01:08:56.600575 139687754716928 logging_writer.py:48] [46500] global_step=46500, grad_norm=4.333033084869385, loss=3.6074869632720947 -I0513 01:09:39.539684 139687763109632 logging_writer.py:48] [46600] global_step=46600, grad_norm=5.907912254333496, loss=4.0293097496032715 -I0513 01:10:21.005425 139687754716928 logging_writer.py:48] [46700] global_step=46700, grad_norm=5.03055477142334, loss=3.8685927391052246 -I0513 01:11:03.406821 139687763109632 logging_writer.py:48] [46800] global_step=46800, grad_norm=6.78891658782959, loss=3.7499351501464844 -I0513 01:11:44.674085 139687754716928 logging_writer.py:48] [46900] global_step=46900, grad_norm=3.878190517425537, loss=3.651235580444336 -I0513 01:12:26.332351 139687763109632 logging_writer.py:48] [47000] global_step=47000, grad_norm=5.011298656463623, loss=3.803100109100342 -I0513 01:13:08.086618 139687754716928 logging_writer.py:48] [47100] global_step=47100, grad_norm=4.358531475067139, loss=3.704285144805908 -I0513 01:13:49.460487 139687763109632 logging_writer.py:48] [47200] global_step=47200, grad_norm=4.522953987121582, loss=3.6638951301574707 -I0513 01:14:34.500658 139687754716928 logging_writer.py:48] [47300] global_step=47300, grad_norm=5.360713481903076, loss=3.66890287399292 -I0513 01:15:16.916203 139687763109632 logging_writer.py:48] [47400] global_step=47400, grad_norm=4.05458402633667, loss=3.6959948539733887 -I0513 01:15:58.299402 139687754716928 logging_writer.py:48] [47500] global_step=47500, grad_norm=3.3621556758880615, loss=3.7703471183776855 -I0513 01:16:39.999463 139687763109632 logging_writer.py:48] [47600] global_step=47600, grad_norm=3.694612503051758, loss=3.7792608737945557 -I0513 01:17:21.751904 139687754716928 logging_writer.py:48] [47700] global_step=47700, grad_norm=4.776402950286865, loss=4.095537185668945 -I0513 01:18:07.626296 139687763109632 logging_writer.py:48] [47800] global_step=47800, grad_norm=6.280882835388184, loss=3.814283609390259 -I0513 01:18:48.899086 139687754716928 logging_writer.py:48] [47900] global_step=47900, grad_norm=4.524450302124023, loss=3.684497356414795 -I0513 01:19:31.077409 139687763109632 logging_writer.py:48] [48000] global_step=48000, grad_norm=5.259480953216553, loss=3.7923879623413086 -I0513 01:20:12.431755 139687754716928 logging_writer.py:48] [48100] global_step=48100, grad_norm=3.2222986221313477, loss=3.768596649169922 -I0513 01:20:54.151207 139687763109632 logging_writer.py:48] [48200] global_step=48200, grad_norm=4.037336826324463, loss=3.8624589443206787 -I0513 01:21:40.472534 139687754716928 logging_writer.py:48] [48300] global_step=48300, grad_norm=4.105864524841309, loss=3.7245090007781982 -I0513 01:22:21.797787 139687763109632 logging_writer.py:48] [48400] global_step=48400, grad_norm=4.117225170135498, loss=3.769376277923584 -I0513 01:23:03.642882 139687754716928 logging_writer.py:48] [48500] global_step=48500, grad_norm=5.883687973022461, loss=3.884206771850586 -I0513 01:23:45.547940 139687763109632 logging_writer.py:48] [48600] global_step=48600, grad_norm=3.87267804145813, loss=3.7335643768310547 -I0513 01:24:26.960174 139687754716928 logging_writer.py:48] [48700] global_step=48700, grad_norm=4.875817775726318, loss=3.8303635120391846 -I0513 01:25:13.251258 139687763109632 logging_writer.py:48] [48800] global_step=48800, grad_norm=4.487644672393799, loss=3.777116298675537 -I0513 01:25:55.152424 139687754716928 logging_writer.py:48] [48900] global_step=48900, grad_norm=5.419808864593506, loss=3.668461322784424 -I0513 01:26:36.505803 139687763109632 logging_writer.py:48] [49000] global_step=49000, grad_norm=3.966418743133545, loss=3.8208088874816895 -I0513 01:27:18.282113 139687754716928 logging_writer.py:48] [49100] global_step=49100, grad_norm=6.176051139831543, loss=3.914729595184326 -I0513 01:27:31.060623 139903809070272 spec.py:333] Evaluating on the training split. -I0513 01:27:41.667263 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 01:29:09.642996 139903809070272 spec.py:363] Evaluating on the test split. -I0513 01:29:10.537894 139903809070272 submission_runner.py:516] Time since start: 20894.61s, Step: 49135, {'train/accuracy': Array(0.00163425, dtype=float32), 'train/loss': Array(8.264104, dtype=float32), 'validation/accuracy': Array(0.00176, dtype=float32), 'validation/loss': Array(8.257346, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(8.255711, dtype=float32), 'test/num_examples': 10000, 'score': 20023.873562812805, 'total_duration': 20894.611929893494, 'accumulated_submission_time': 20023.873562812805, 'accumulated_eval_time': 869.9464857578278, 'accumulated_logging_time': 0.3443114757537842} -I0513 01:29:10.562589 139687763109632 logging_writer.py:48] [49135] accumulated_eval_time=869.946, accumulated_logging_time=0.344311, accumulated_submission_time=20023.9, global_step=49135, preemption_count=0, score=20023.9, test/accuracy=0.0013000001199543476, test/loss=8.25571060180664, test/num_examples=10000, total_duration=20894.6, train/accuracy=0.0016342473682016134, train/loss=8.264103889465332, validation/accuracy=0.0017599998973309994, validation/loss=8.257346153259277, validation/num_examples=50000 -I0513 01:29:43.237081 139687754716928 logging_writer.py:48] [49200] global_step=49200, grad_norm=5.365841388702393, loss=3.8717103004455566 -I0513 01:30:24.188430 139687763109632 logging_writer.py:48] [49300] global_step=49300, grad_norm=7.160286903381348, loss=3.8151583671569824 -I0513 01:31:05.432227 139687754716928 logging_writer.py:48] [49400] global_step=49400, grad_norm=4.511857032775879, loss=3.8213579654693604 -I0513 01:31:47.009154 139687763109632 logging_writer.py:48] [49500] global_step=49500, grad_norm=5.561305522918701, loss=3.8931565284729004 -I0513 01:32:27.920399 139687754716928 logging_writer.py:48] [49600] global_step=49600, grad_norm=7.927204132080078, loss=3.9708521366119385 -I0513 01:33:09.209846 139687763109632 logging_writer.py:48] [49700] global_step=49700, grad_norm=5.550274848937988, loss=3.816206932067871 -I0513 01:33:50.590485 139687754716928 logging_writer.py:48] [49800] global_step=49800, grad_norm=13.793540954589844, loss=4.068350791931152 -I0513 01:34:31.519564 139687763109632 logging_writer.py:48] [49900] global_step=49900, grad_norm=3.174260139465332, loss=3.758592128753662 -I0513 01:35:12.750936 139687754716928 logging_writer.py:48] [50000] global_step=50000, grad_norm=6.174293518066406, loss=3.953754425048828 -I0513 01:35:54.235727 139687763109632 logging_writer.py:48] [50100] global_step=50100, grad_norm=5.81603479385376, loss=3.688891887664795 -I0513 01:36:35.124359 139687754716928 logging_writer.py:48] [50200] global_step=50200, grad_norm=6.294396877288818, loss=4.001725673675537 -I0513 01:37:16.367098 139687763109632 logging_writer.py:48] [50300] global_step=50300, grad_norm=5.034473896026611, loss=3.840576171875 -I0513 01:37:57.816768 139687754716928 logging_writer.py:48] [50400] global_step=50400, grad_norm=4.5284318923950195, loss=3.702695608139038 -I0513 01:38:38.809403 139687763109632 logging_writer.py:48] [50500] global_step=50500, grad_norm=5.817103862762451, loss=3.776585340499878 -I0513 01:39:19.203775 139687754716928 logging_writer.py:48] [50600] global_step=50600, grad_norm=5.02617073059082, loss=3.6592016220092773 -I0513 01:40:00.916080 139687763109632 logging_writer.py:48] [50700] global_step=50700, grad_norm=4.625972270965576, loss=3.7639756202697754 -I0513 01:40:41.841172 139687754716928 logging_writer.py:48] [50800] global_step=50800, grad_norm=4.592901229858398, loss=3.6707706451416016 -I0513 01:41:23.150539 139687763109632 logging_writer.py:48] [50900] global_step=50900, grad_norm=2.9204978942871094, loss=3.5999560356140137 -I0513 01:42:04.530428 139687754716928 logging_writer.py:48] [51000] global_step=51000, grad_norm=3.7707228660583496, loss=3.7798962593078613 -I0513 01:42:45.567348 139687763109632 logging_writer.py:48] [51100] global_step=51100, grad_norm=3.652191162109375, loss=3.757239580154419 -I0513 01:43:26.834130 139687754716928 logging_writer.py:48] [51200] global_step=51200, grad_norm=5.330374240875244, loss=3.6340603828430176 -I0513 01:44:08.284938 139687763109632 logging_writer.py:48] [51300] global_step=51300, grad_norm=4.922441482543945, loss=3.559295654296875 -I0513 01:44:49.393868 139687754716928 logging_writer.py:48] [51400] global_step=51400, grad_norm=3.3839271068573, loss=3.7486958503723145 -I0513 01:45:33.969844 139687763109632 logging_writer.py:48] [51500] global_step=51500, grad_norm=3.516151189804077, loss=3.814251661300659 -I0513 01:46:15.565124 139687754716928 logging_writer.py:48] [51600] global_step=51600, grad_norm=3.336361885070801, loss=3.932081937789917 -I0513 01:46:56.482944 139687763109632 logging_writer.py:48] [51700] global_step=51700, grad_norm=5.712367057800293, loss=4.035030364990234 -I0513 01:47:37.846495 139687754716928 logging_writer.py:48] [51800] global_step=51800, grad_norm=6.018395900726318, loss=3.632411241531372 -I0513 01:48:19.332431 139687763109632 logging_writer.py:48] [51900] global_step=51900, grad_norm=4.898831844329834, loss=3.7984817028045654 -I0513 01:49:00.271512 139687754716928 logging_writer.py:48] [52000] global_step=52000, grad_norm=4.509390354156494, loss=3.749276876449585 -I0513 01:49:41.474128 139687763109632 logging_writer.py:48] [52100] global_step=52100, grad_norm=6.878928184509277, loss=3.671360492706299 -I0513 01:50:22.875661 139687754716928 logging_writer.py:48] [52200] global_step=52200, grad_norm=7.803660869598389, loss=3.9464991092681885 -I0513 01:51:03.868227 139687763109632 logging_writer.py:48] [52300] global_step=52300, grad_norm=3.543653964996338, loss=3.7522478103637695 -I0513 01:51:48.802837 139687754716928 logging_writer.py:48] [52400] global_step=52400, grad_norm=3.937105655670166, loss=3.7056188583374023 -I0513 01:52:30.453898 139687763109632 logging_writer.py:48] [52500] global_step=52500, grad_norm=4.671799182891846, loss=3.687063217163086 -I0513 01:53:11.409454 139687754716928 logging_writer.py:48] [52600] global_step=52600, grad_norm=8.388178825378418, loss=3.8928215503692627 -I0513 01:53:52.421000 139687763109632 logging_writer.py:48] [52700] global_step=52700, grad_norm=6.45353364944458, loss=3.8706116676330566 -I0513 01:54:38.317304 139687754716928 logging_writer.py:48] [52800] global_step=52800, grad_norm=4.004523754119873, loss=3.613154649734497 -I0513 01:55:19.314821 139687763109632 logging_writer.py:48] [52900] global_step=52900, grad_norm=6.072022438049316, loss=3.632161855697632 -I0513 01:56:00.679431 139687754716928 logging_writer.py:48] [53000] global_step=53000, grad_norm=4.211053848266602, loss=3.764756441116333 -I0513 01:56:46.168184 139687763109632 logging_writer.py:48] [53100] global_step=53100, grad_norm=6.6874847412109375, loss=3.6929025650024414 -I0513 01:57:35.184982 139687754716928 logging_writer.py:48] [53200] global_step=53200, grad_norm=5.552595615386963, loss=3.9167282581329346 -I0513 01:58:16.591524 139687763109632 logging_writer.py:48] [53300] global_step=53300, grad_norm=10.889020919799805, loss=3.9836020469665527 -I0513 01:58:58.017224 139687754716928 logging_writer.py:48] [53400] global_step=53400, grad_norm=4.990148544311523, loss=3.8310883045196533 -I0513 01:59:39.115092 139687763109632 logging_writer.py:48] [53500] global_step=53500, grad_norm=6.350290775299072, loss=3.649461269378662 -I0513 02:00:20.437976 139687754716928 logging_writer.py:48] [53600] global_step=53600, grad_norm=2.9172637462615967, loss=3.6889214515686035 -I0513 02:01:01.895761 139687763109632 logging_writer.py:48] [53700] global_step=53700, grad_norm=5.431403160095215, loss=3.8153786659240723 -I0513 02:01:42.901274 139687754716928 logging_writer.py:48] [53800] global_step=53800, grad_norm=5.747957229614258, loss=3.760408878326416 -I0513 02:02:26.677706 139903809070272 spec.py:333] Evaluating on the training split. -I0513 02:02:37.413235 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 02:03:46.274865 139903809070272 spec.py:363] Evaluating on the test split. -I0513 02:03:47.161908 139903809070272 submission_runner.py:516] Time since start: 22971.24s, Step: 53896, {'train/accuracy': Array(0.00171397, dtype=float32), 'train/loss': Array(8.253426, dtype=float32), 'validation/accuracy': Array(0.00174, dtype=float32), 'validation/loss': Array(8.235537, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(8.235645, dtype=float32), 'test/num_examples': 10000, 'score': 22019.93824338913, 'total_duration': 22971.235975027084, 'accumulated_submission_time': 22019.93824338913, 'accumulated_eval_time': 950.4290370941162, 'accumulated_logging_time': 0.3771963119506836} -I0513 02:03:47.190310 139687763109632 logging_writer.py:48] [53896] accumulated_eval_time=950.429, accumulated_logging_time=0.377196, accumulated_submission_time=22019.9, global_step=53896, preemption_count=0, score=22019.9, test/accuracy=0.0012000000569969416, test/loss=8.235645294189453, test/num_examples=10000, total_duration=22971.2, train/accuracy=0.001713966834358871, train/loss=8.253425598144531, validation/accuracy=0.0017399999778717756, validation/loss=8.235536575317383, validation/num_examples=50000 -I0513 02:03:48.834240 139687754716928 logging_writer.py:48] [53900] global_step=53900, grad_norm=8.254786491394043, loss=3.887938976287842 -I0513 02:04:34.124188 139687763109632 logging_writer.py:48] [54000] global_step=54000, grad_norm=4.050253868103027, loss=3.7449822425842285 -I0513 02:05:14.914795 139687754716928 logging_writer.py:48] [54100] global_step=54100, grad_norm=3.4184212684631348, loss=3.642448902130127 -I0513 02:05:55.969343 139687763109632 logging_writer.py:48] [54200] global_step=54200, grad_norm=23.046079635620117, loss=3.8561248779296875 -I0513 02:06:37.223494 139687754716928 logging_writer.py:48] [54300] global_step=54300, grad_norm=4.293612957000732, loss=3.696910858154297 -I0513 02:07:18.043880 139687763109632 logging_writer.py:48] [54400] global_step=54400, grad_norm=2.62320613861084, loss=3.713134765625 -I0513 02:07:59.597890 139687754716928 logging_writer.py:48] [54500] global_step=54500, grad_norm=5.221779823303223, loss=3.6465604305267334 -I0513 02:08:41.058540 139687763109632 logging_writer.py:48] [54600] global_step=54600, grad_norm=3.2922797203063965, loss=3.712770938873291 -I0513 02:09:21.983999 139687754716928 logging_writer.py:48] [54700] global_step=54700, grad_norm=3.780437469482422, loss=3.8474087715148926 -I0513 02:10:03.175013 139687763109632 logging_writer.py:48] [54800] global_step=54800, grad_norm=4.711358547210693, loss=3.7810935974121094 -I0513 02:10:44.554153 139687754716928 logging_writer.py:48] [54900] global_step=54900, grad_norm=6.993558406829834, loss=3.801097869873047 -I0513 02:11:25.553742 139687763109632 logging_writer.py:48] [55000] global_step=55000, grad_norm=5.08542013168335, loss=3.7087271213531494 -I0513 02:12:06.860477 139687754716928 logging_writer.py:48] [55100] global_step=55100, grad_norm=5.367064952850342, loss=3.7550082206726074 -I0513 02:12:48.290386 139687763109632 logging_writer.py:48] [55200] global_step=55200, grad_norm=5.362349987030029, loss=3.783287525177002 -I0513 02:13:29.251242 139687754716928 logging_writer.py:48] [55300] global_step=55300, grad_norm=3.3240387439727783, loss=3.7667407989501953 -I0513 02:14:10.281978 139687763109632 logging_writer.py:48] [55400] global_step=55400, grad_norm=5.192694187164307, loss=3.5870330333709717 -I0513 02:14:51.639704 139687754716928 logging_writer.py:48] [55500] global_step=55500, grad_norm=4.367573261260986, loss=3.6223227977752686 -I0513 02:15:40.827154 139687763109632 logging_writer.py:48] [55600] global_step=55600, grad_norm=7.926640510559082, loss=3.9210903644561768 -I0513 02:16:22.023155 139687754716928 logging_writer.py:48] [55700] global_step=55700, grad_norm=5.150660037994385, loss=3.74173641204834 -I0513 02:17:07.532995 139687763109632 logging_writer.py:48] [55800] global_step=55800, grad_norm=6.285862922668457, loss=3.8508567810058594 -I0513 02:17:48.509212 139687754716928 logging_writer.py:48] [55900] global_step=55900, grad_norm=5.781080722808838, loss=3.715615749359131 -I0513 02:18:29.835566 139687763109632 logging_writer.py:48] [56000] global_step=56000, grad_norm=5.723334789276123, loss=3.764068126678467 -I0513 02:19:11.239970 139687754716928 logging_writer.py:48] [56100] global_step=56100, grad_norm=6.508880138397217, loss=3.8194267749786377 -I0513 02:19:52.231631 139687763109632 logging_writer.py:48] [56200] global_step=56200, grad_norm=2.6089086532592773, loss=3.8350095748901367 -I0513 02:20:33.493191 139687754716928 logging_writer.py:48] [56300] global_step=56300, grad_norm=3.25722599029541, loss=3.617736339569092 -I0513 02:21:14.954880 139687763109632 logging_writer.py:48] [56400] global_step=56400, grad_norm=17.7414493560791, loss=4.5096330642700195 -I0513 02:21:55.841074 139687754716928 logging_writer.py:48] [56500] global_step=56500, grad_norm=4.025692939758301, loss=3.7712442874908447 -I0513 02:22:41.255688 139687763109632 logging_writer.py:48] [56600] global_step=56600, grad_norm=3.490370035171509, loss=3.7073159217834473 -I0513 02:23:22.666406 139687754716928 logging_writer.py:48] [56700] global_step=56700, grad_norm=3.779212713241577, loss=3.6467432975769043 -I0513 02:24:03.604401 139687763109632 logging_writer.py:48] [56800] global_step=56800, grad_norm=3.19504451751709, loss=3.666728973388672 -I0513 02:24:44.831405 139687754716928 logging_writer.py:48] [56900] global_step=56900, grad_norm=4.512517929077148, loss=3.7533352375030518 -I0513 02:25:26.273535 139687763109632 logging_writer.py:48] [57000] global_step=57000, grad_norm=9.118008613586426, loss=3.781275510787964 -I0513 02:26:07.287144 139687754716928 logging_writer.py:48] [57100] global_step=57100, grad_norm=4.379135608673096, loss=3.726134777069092 -I0513 02:26:51.915030 139687763109632 logging_writer.py:48] [57200] global_step=57200, grad_norm=4.042726039886475, loss=3.8502979278564453 -I0513 02:27:35.962476 139687754716928 logging_writer.py:48] [57300] global_step=57300, grad_norm=5.283278465270996, loss=3.8037421703338623 -I0513 02:28:18.355124 139687763109632 logging_writer.py:48] [57400] global_step=57400, grad_norm=3.762319564819336, loss=3.7398059368133545 -I0513 02:29:00.190063 139687754716928 logging_writer.py:48] [57500] global_step=57500, grad_norm=3.3871002197265625, loss=3.629765033721924 -I0513 02:29:41.886537 139687763109632 logging_writer.py:48] [57600] global_step=57600, grad_norm=3.869986057281494, loss=3.647819995880127 -I0513 02:30:22.986505 139687754716928 logging_writer.py:48] [57700] global_step=57700, grad_norm=5.122893333435059, loss=3.7367048263549805 -I0513 02:31:04.448567 139687763109632 logging_writer.py:48] [57800] global_step=57800, grad_norm=3.3173305988311768, loss=3.725592613220215 -I0513 02:31:50.216735 139687754716928 logging_writer.py:48] [57900] global_step=57900, grad_norm=3.6714208126068115, loss=3.7291860580444336 -I0513 02:32:31.289960 139687763109632 logging_writer.py:48] [58000] global_step=58000, grad_norm=10.119243621826172, loss=4.06237268447876 -I0513 02:33:16.942769 139687754716928 logging_writer.py:48] [58100] global_step=58100, grad_norm=5.583806991577148, loss=3.724050283432007 -I0513 02:33:58.435019 139687763109632 logging_writer.py:48] [58200] global_step=58200, grad_norm=4.453329086303711, loss=3.6045289039611816 -I0513 02:34:39.471009 139687754716928 logging_writer.py:48] [58300] global_step=58300, grad_norm=6.116827011108398, loss=3.7900729179382324 -I0513 02:35:20.938193 139687763109632 logging_writer.py:48] [58400] global_step=58400, grad_norm=3.2840781211853027, loss=3.7023377418518066 -I0513 02:36:02.480310 139687754716928 logging_writer.py:48] [58500] global_step=58500, grad_norm=4.162327766418457, loss=3.7367324829101562 -I0513 02:36:43.596779 139687763109632 logging_writer.py:48] [58600] global_step=58600, grad_norm=4.8851165771484375, loss=3.7628173828125 -I0513 02:37:03.599272 139903809070272 spec.py:333] Evaluating on the training split. -I0513 02:37:13.832328 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 02:38:12.043301 139903809070272 spec.py:363] Evaluating on the test split. -I0513 02:38:12.939385 139903809070272 submission_runner.py:516] Time since start: 25037.01s, Step: 58654, {'train/accuracy': Array(0.00169404, dtype=float32), 'train/loss': Array(8.188425, dtype=float32), 'validation/accuracy': Array(0.0017, dtype=float32), 'validation/loss': Array(8.179385, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0014, dtype=float32), 'test/loss': Array(8.181261, dtype=float32), 'test/num_examples': 10000, 'score': 24016.297024965286, 'total_duration': 25037.013209342957, 'accumulated_submission_time': 24016.297024965286, 'accumulated_eval_time': 1019.7672593593597, 'accumulated_logging_time': 0.41385316848754883} -I0513 02:38:12.969896 139687754716928 logging_writer.py:48] [58654] accumulated_eval_time=1019.77, accumulated_logging_time=0.413853, accumulated_submission_time=24016.3, global_step=58654, preemption_count=0, score=24016.3, test/accuracy=0.0014000000664964318, test/loss=8.18126106262207, test/num_examples=10000, total_duration=25037, train/accuracy=0.001694036996923387, train/loss=8.188425064086914, validation/accuracy=0.0016999999061226845, validation/loss=8.1793851852417, validation/num_examples=50000 -I0513 02:38:35.484652 139687763109632 logging_writer.py:48] [58700] global_step=58700, grad_norm=11.201581001281738, loss=3.8576273918151855 -I0513 02:39:12.807054 139687754716928 logging_writer.py:48] [58800] global_step=58800, grad_norm=6.166787624359131, loss=3.9082250595092773 -I0513 02:39:49.729188 139687763109632 logging_writer.py:48] [58900] global_step=58900, grad_norm=4.364743709564209, loss=3.751093864440918 -I0513 02:40:26.850817 139687754716928 logging_writer.py:48] [59000] global_step=59000, grad_norm=2.839359998703003, loss=3.7826929092407227 -I0513 02:41:04.266967 139687763109632 logging_writer.py:48] [59100] global_step=59100, grad_norm=4.022916316986084, loss=3.540834426879883 -I0513 02:41:41.164169 139687754716928 logging_writer.py:48] [59200] global_step=59200, grad_norm=4.4122819900512695, loss=3.6593525409698486 -I0513 02:42:18.408392 139687763109632 logging_writer.py:48] [59300] global_step=59300, grad_norm=3.978407382965088, loss=3.726576328277588 -I0513 02:42:55.822129 139687754716928 logging_writer.py:48] [59400] global_step=59400, grad_norm=3.785562038421631, loss=3.8316659927368164 -I0513 02:43:32.784528 139687763109632 logging_writer.py:48] [59500] global_step=59500, grad_norm=3.1105008125305176, loss=3.6872024536132812 -I0513 02:44:09.975831 139687754716928 logging_writer.py:48] [59600] global_step=59600, grad_norm=8.06755256652832, loss=3.7192866802215576 -I0513 02:44:47.623104 139687763109632 logging_writer.py:48] [59700] global_step=59700, grad_norm=7.80720329284668, loss=3.859403133392334 -I0513 02:45:24.568631 139687754716928 logging_writer.py:48] [59800] global_step=59800, grad_norm=3.0368053913116455, loss=3.747211217880249 -I0513 02:46:01.807213 139687763109632 logging_writer.py:48] [59900] global_step=59900, grad_norm=6.284008502960205, loss=3.918862819671631 -I0513 02:46:39.251975 139687754716928 logging_writer.py:48] [60000] global_step=60000, grad_norm=5.494969367980957, loss=3.736943244934082 -I0513 02:47:16.140891 139687763109632 logging_writer.py:48] [60100] global_step=60100, grad_norm=7.196032524108887, loss=3.660900115966797 -I0513 02:47:53.521225 139687754716928 logging_writer.py:48] [60200] global_step=60200, grad_norm=4.569591522216797, loss=3.806845188140869 -I0513 02:48:31.004385 139687763109632 logging_writer.py:48] [60300] global_step=60300, grad_norm=4.44943380355835, loss=3.7252519130706787 -I0513 02:49:07.896763 139687754716928 logging_writer.py:48] [60400] global_step=60400, grad_norm=5.508068084716797, loss=3.796757698059082 -I0513 02:49:45.129824 139687763109632 logging_writer.py:48] [60500] global_step=60500, grad_norm=8.081077575683594, loss=3.9384331703186035 -I0513 02:50:22.454580 139687754716928 logging_writer.py:48] [60600] global_step=60600, grad_norm=4.194653511047363, loss=3.9076387882232666 -I0513 02:50:59.380352 139687763109632 logging_writer.py:48] [60700] global_step=60700, grad_norm=4.5420241355896, loss=3.7204830646514893 -I0513 02:51:36.655647 139687754716928 logging_writer.py:48] [60800] global_step=60800, grad_norm=24.93932342529297, loss=3.873307228088379 -I0513 02:52:14.021280 139687763109632 logging_writer.py:48] [60900] global_step=60900, grad_norm=4.58526611328125, loss=3.7948784828186035 -I0513 02:52:51.000335 139687754716928 logging_writer.py:48] [61000] global_step=61000, grad_norm=4.128470420837402, loss=3.8336832523345947 -I0513 02:53:28.328835 139687763109632 logging_writer.py:48] [61100] global_step=61100, grad_norm=3.1275475025177, loss=3.5531368255615234 -I0513 02:54:05.666374 139687754716928 logging_writer.py:48] [61200] global_step=61200, grad_norm=7.289842128753662, loss=3.7774648666381836 -I0513 02:54:42.490871 139687763109632 logging_writer.py:48] [61300] global_step=61300, grad_norm=5.137810230255127, loss=3.619879961013794 -I0513 02:55:19.757186 139687754716928 logging_writer.py:48] [61400] global_step=61400, grad_norm=5.786778450012207, loss=3.7474780082702637 -I0513 02:55:57.142409 139687763109632 logging_writer.py:48] [61500] global_step=61500, grad_norm=3.3874270915985107, loss=3.7106058597564697 -I0513 02:56:34.066053 139687754716928 logging_writer.py:48] [61600] global_step=61600, grad_norm=3.3103740215301514, loss=3.7420902252197266 -I0513 02:57:11.326475 139687763109632 logging_writer.py:48] [61700] global_step=61700, grad_norm=5.45706033706665, loss=3.7426698207855225 -I0513 02:57:48.796757 139687754716928 logging_writer.py:48] [61800] global_step=61800, grad_norm=3.1062095165252686, loss=3.6449480056762695 -I0513 02:58:25.705620 139687763109632 logging_writer.py:48] [61900] global_step=61900, grad_norm=7.463752269744873, loss=3.8270926475524902 -I0513 02:59:03.063236 139687754716928 logging_writer.py:48] [62000] global_step=62000, grad_norm=4.642884254455566, loss=3.830510139465332 -I0513 02:59:40.449442 139687763109632 logging_writer.py:48] [62100] global_step=62100, grad_norm=6.257845878601074, loss=3.869779586791992 -I0513 03:00:17.366396 139687754716928 logging_writer.py:48] [62200] global_step=62200, grad_norm=5.086574077606201, loss=3.7534589767456055 -I0513 03:00:54.601855 139687763109632 logging_writer.py:48] [62300] global_step=62300, grad_norm=3.255706787109375, loss=3.712271213531494 -I0513 03:01:32.023935 139687754716928 logging_writer.py:48] [62400] global_step=62400, grad_norm=12.972366333007812, loss=3.809220552444458 -I0513 03:02:08.932769 139687763109632 logging_writer.py:48] [62500] global_step=62500, grad_norm=5.828891277313232, loss=3.883704900741577 -I0513 03:02:46.187655 139687754716928 logging_writer.py:48] [62600] global_step=62600, grad_norm=5.344369411468506, loss=3.537379741668701 -I0513 03:03:23.593572 139687763109632 logging_writer.py:48] [62700] global_step=62700, grad_norm=4.453258514404297, loss=3.603692054748535 -I0513 03:04:00.513911 139687754716928 logging_writer.py:48] [62800] global_step=62800, grad_norm=3.844236373901367, loss=3.697650671005249 -I0513 03:04:37.732836 139687763109632 logging_writer.py:48] [62900] global_step=62900, grad_norm=4.523203372955322, loss=3.6880288124084473 -I0513 03:05:15.123934 139687754716928 logging_writer.py:48] [63000] global_step=63000, grad_norm=3.9698939323425293, loss=3.8623175621032715 -I0513 03:05:51.991163 139687763109632 logging_writer.py:48] [63100] global_step=63100, grad_norm=5.4277024269104, loss=3.753875732421875 -I0513 03:06:29.187859 139687754716928 logging_writer.py:48] [63200] global_step=63200, grad_norm=6.095491886138916, loss=3.6877684593200684 -I0513 03:07:06.627465 139687763109632 logging_writer.py:48] [63300] global_step=63300, grad_norm=5.261175632476807, loss=3.6736197471618652 -I0513 03:07:43.921986 139687754716928 logging_writer.py:48] [63400] global_step=63400, grad_norm=4.9305901527404785, loss=3.8247551918029785 -I0513 03:08:20.933807 139687763109632 logging_writer.py:48] [63500] global_step=63500, grad_norm=6.388390064239502, loss=3.785383939743042 -I0513 03:08:58.410153 139687754716928 logging_writer.py:48] [63600] global_step=63600, grad_norm=3.6646506786346436, loss=3.669736862182617 -I0513 03:09:35.263775 139687763109632 logging_writer.py:48] [63700] global_step=63700, grad_norm=6.811042785644531, loss=3.89323091506958 -I0513 03:10:12.418566 139687754716928 logging_writer.py:48] [63800] global_step=63800, grad_norm=4.432936668395996, loss=3.867218255996704 -I0513 03:10:49.863259 139687763109632 logging_writer.py:48] [63900] global_step=63900, grad_norm=3.641993999481201, loss=3.743807792663574 -I0513 03:11:27.093539 139687754716928 logging_writer.py:48] [64000] global_step=64000, grad_norm=4.29096794128418, loss=3.771710157394409 -I0513 03:11:29.014851 139903809070272 spec.py:333] Evaluating on the training split. -I0513 03:11:39.292842 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 03:12:37.017171 139903809070272 spec.py:363] Evaluating on the test split. -I0513 03:12:37.910851 139903809070272 submission_runner.py:516] Time since start: 27101.98s, Step: 64006, {'train/accuracy': Array(0.00179369, dtype=float32), 'train/loss': Array(8.101958, dtype=float32), 'validation/accuracy': Array(0.00168, dtype=float32), 'validation/loss': Array(8.089027, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0016, dtype=float32), 'test/loss': Array(8.0927105, dtype=float32), 'test/num_examples': 10000, 'score': 26012.28999519348, 'total_duration': 27101.98451256752, 'accumulated_submission_time': 26012.28999519348, 'accumulated_eval_time': 1088.661203622818, 'accumulated_logging_time': 0.4541494846343994} -I0513 03:12:37.951874 139687763109632 logging_writer.py:48] [64006] accumulated_eval_time=1088.66, accumulated_logging_time=0.454149, accumulated_submission_time=26012.3, global_step=64006, preemption_count=0, score=26012.3, test/accuracy=0.001600000075995922, test/loss=8.092710494995117, test/num_examples=10000, total_duration=27102, train/accuracy=0.0017936861841008067, train/loss=8.101958274841309, validation/accuracy=0.0016799999866634607, validation/loss=8.089027404785156, validation/num_examples=50000 -I0513 03:13:17.994961 139687754716928 logging_writer.py:48] [64100] global_step=64100, grad_norm=4.030590057373047, loss=3.7392213344573975 -I0513 03:13:55.387794 139687763109632 logging_writer.py:48] [64200] global_step=64200, grad_norm=3.826037883758545, loss=3.687376022338867 -I0513 03:14:32.244590 139687754716928 logging_writer.py:48] [64300] global_step=64300, grad_norm=3.8314058780670166, loss=3.6277883052825928 -I0513 03:15:09.296774 139687763109632 logging_writer.py:48] [64400] global_step=64400, grad_norm=5.706965446472168, loss=3.8561158180236816 -I0513 03:15:46.689631 139687754716928 logging_writer.py:48] [64500] global_step=64500, grad_norm=6.917937755584717, loss=4.011709690093994 -I0513 03:16:32.669526 139687763109632 logging_writer.py:48] [64600] global_step=64600, grad_norm=5.238850116729736, loss=3.7526626586914062 -I0513 03:17:14.462648 139687754716928 logging_writer.py:48] [64700] global_step=64700, grad_norm=4.667111396789551, loss=3.9492297172546387 -I0513 03:18:01.112894 139687763109632 logging_writer.py:48] [64800] global_step=64800, grad_norm=3.566570997238159, loss=3.7210023403167725 -I0513 03:18:37.956321 139687754716928 logging_writer.py:48] [64900] global_step=64900, grad_norm=4.71935510635376, loss=3.522326946258545 -I0513 03:19:15.291687 139687763109632 logging_writer.py:48] [65000] global_step=65000, grad_norm=3.8931164741516113, loss=3.8704774379730225 -I0513 03:19:56.807083 139687754716928 logging_writer.py:48] [65100] global_step=65100, grad_norm=3.453197479248047, loss=3.6740100383758545 -I0513 03:20:33.725769 139687763109632 logging_writer.py:48] [65200] global_step=65200, grad_norm=5.698867321014404, loss=3.7660434246063232 -I0513 03:21:10.958997 139687754716928 logging_writer.py:48] [65300] global_step=65300, grad_norm=9.512503623962402, loss=3.952834129333496 -I0513 03:21:48.366362 139687763109632 logging_writer.py:48] [65400] global_step=65400, grad_norm=3.464953899383545, loss=3.7395873069763184 -I0513 03:22:25.253700 139687754716928 logging_writer.py:48] [65500] global_step=65500, grad_norm=2.901662588119507, loss=3.7379069328308105 -I0513 03:23:02.408374 139687763109632 logging_writer.py:48] [65600] global_step=65600, grad_norm=3.564404249191284, loss=3.6590564250946045 -I0513 03:23:39.708108 139687754716928 logging_writer.py:48] [65700] global_step=65700, grad_norm=4.4541521072387695, loss=3.6865599155426025 -I0513 03:24:20.751821 139687763109632 logging_writer.py:48] [65800] global_step=65800, grad_norm=18.0384521484375, loss=3.7878706455230713 -I0513 03:24:58.054371 139687754716928 logging_writer.py:48] [65900] global_step=65900, grad_norm=4.660128593444824, loss=3.6610984802246094 -I0513 03:25:35.396188 139687763109632 logging_writer.py:48] [66000] global_step=66000, grad_norm=4.548684120178223, loss=3.6421475410461426 -I0513 03:26:12.597899 139687754716928 logging_writer.py:48] [66100] global_step=66100, grad_norm=7.01583194732666, loss=3.9274187088012695 -I0513 03:26:49.430962 139687763109632 logging_writer.py:48] [66200] global_step=66200, grad_norm=3.004728317260742, loss=3.741170883178711 -I0513 03:27:30.863424 139687754716928 logging_writer.py:48] [66300] global_step=66300, grad_norm=4.24213981628418, loss=3.6540935039520264 -I0513 03:28:15.934303 139687763109632 logging_writer.py:48] [66400] global_step=66400, grad_norm=19.151277542114258, loss=3.735459566116333 -I0513 03:28:53.117040 139687754716928 logging_writer.py:48] [66500] global_step=66500, grad_norm=4.3734259605407715, loss=3.7781403064727783 -I0513 03:29:34.661979 139687763109632 logging_writer.py:48] [66600] global_step=66600, grad_norm=34.828712463378906, loss=3.79734468460083 -I0513 03:30:11.971175 139687754716928 logging_writer.py:48] [66700] global_step=66700, grad_norm=4.487275123596191, loss=3.8496830463409424 -I0513 03:30:48.900022 139687763109632 logging_writer.py:48] [66800] global_step=66800, grad_norm=6.110284805297852, loss=3.8515377044677734 -I0513 03:31:26.240409 139687754716928 logging_writer.py:48] [66900] global_step=66900, grad_norm=3.898533344268799, loss=3.7125442028045654 -I0513 03:32:03.169913 139687763109632 logging_writer.py:48] [67000] global_step=67000, grad_norm=4.015486717224121, loss=3.838724136352539 -I0513 03:32:44.491525 139687754716928 logging_writer.py:48] [67100] global_step=67100, grad_norm=3.7285027503967285, loss=3.601334571838379 -I0513 03:33:25.898082 139687763109632 logging_writer.py:48] [67200] global_step=67200, grad_norm=8.06448745727539, loss=3.8656153678894043 -I0513 03:34:06.856369 139687754716928 logging_writer.py:48] [67300] global_step=67300, grad_norm=2.764652967453003, loss=3.6415305137634277 -I0513 03:34:52.231377 139687763109632 logging_writer.py:48] [67400] global_step=67400, grad_norm=4.752501010894775, loss=3.8310046195983887 -I0513 03:35:33.737146 139687754716928 logging_writer.py:48] [67500] global_step=67500, grad_norm=13.809173583984375, loss=4.103457450866699 -I0513 03:36:10.643738 139687763109632 logging_writer.py:48] [67600] global_step=67600, grad_norm=6.790821552276611, loss=3.8764967918395996 -I0513 03:36:47.842384 139687754716928 logging_writer.py:48] [67700] global_step=67700, grad_norm=6.655921459197998, loss=3.8610496520996094 -I0513 03:37:29.335839 139687763109632 logging_writer.py:48] [67800] global_step=67800, grad_norm=4.688515663146973, loss=3.770561695098877 -I0513 03:38:10.414616 139687754716928 logging_writer.py:48] [67900] global_step=67900, grad_norm=10.376776695251465, loss=3.7260684967041016 -I0513 03:38:47.604699 139687763109632 logging_writer.py:48] [68000] global_step=68000, grad_norm=3.2727389335632324, loss=3.765468120574951 -I0513 03:39:24.961693 139687754716928 logging_writer.py:48] [68100] global_step=68100, grad_norm=6.955250263214111, loss=3.816838264465332 -I0513 03:40:01.898186 139687763109632 logging_writer.py:48] [68200] global_step=68200, grad_norm=4.9848127365112305, loss=3.650938034057617 -I0513 03:40:39.173383 139687754716928 logging_writer.py:48] [68300] global_step=68300, grad_norm=7.124767303466797, loss=3.766458511352539 -I0513 03:41:16.494020 139687763109632 logging_writer.py:48] [68400] global_step=68400, grad_norm=3.5716617107391357, loss=3.723963975906372 -I0513 03:41:53.405871 139687754716928 logging_writer.py:48] [68500] global_step=68500, grad_norm=4.700095176696777, loss=3.7076175212860107 -I0513 03:42:30.486258 139687763109632 logging_writer.py:48] [68600] global_step=68600, grad_norm=6.56386137008667, loss=3.8598670959472656 -I0513 03:43:07.826823 139687754716928 logging_writer.py:48] [68700] global_step=68700, grad_norm=6.646833419799805, loss=3.8042752742767334 -I0513 03:43:44.751529 139687763109632 logging_writer.py:48] [68800] global_step=68800, grad_norm=5.086121559143066, loss=3.7159767150878906 -I0513 03:44:21.973252 139687754716928 logging_writer.py:48] [68900] global_step=68900, grad_norm=4.6214919090271, loss=3.751352071762085 -I0513 03:44:59.395376 139687763109632 logging_writer.py:48] [69000] global_step=69000, grad_norm=3.985478162765503, loss=3.7047171592712402 -I0513 03:45:36.324247 139687754716928 logging_writer.py:48] [69100] global_step=69100, grad_norm=3.9880130290985107, loss=3.7583045959472656 -I0513 03:45:54.049696 139903809070272 spec.py:333] Evaluating on the training split. -I0513 03:46:08.016969 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 03:47:10.451671 139903809070272 spec.py:363] Evaluating on the test split. -I0513 03:47:11.345744 139903809070272 submission_runner.py:516] Time since start: 29175.42s, Step: 69149, {'train/accuracy': Array(0.00159439, dtype=float32), 'train/loss': Array(8.021358, dtype=float32), 'validation/accuracy': Array(0.00172, dtype=float32), 'validation/loss': Array(8.019661, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0016, dtype=float32), 'test/loss': Array(8.025281, dtype=float32), 'test/num_examples': 10000, 'score': 28008.3186814785, 'total_duration': 29175.41971540451, 'accumulated_submission_time': 28008.3186814785, 'accumulated_eval_time': 1165.955510854721, 'accumulated_logging_time': 0.5219497680664062} -I0513 03:47:11.379261 139687763109632 logging_writer.py:48] [69149] accumulated_eval_time=1165.96, accumulated_logging_time=0.52195, accumulated_submission_time=28008.3, global_step=69149, preemption_count=0, score=28008.3, test/accuracy=0.001600000075995922, test/loss=8.025280952453613, test/num_examples=10000, total_duration=29175.4, train/accuracy=0.0015943876933306456, train/loss=8.021357536315918, validation/accuracy=0.00171999994199723, validation/loss=8.019660949707031, validation/num_examples=50000 -I0513 03:47:30.629128 139687754716928 logging_writer.py:48] [69200] global_step=69200, grad_norm=5.306702136993408, loss=3.754940986633301 -I0513 03:48:17.731521 139687763109632 logging_writer.py:48] [69300] global_step=69300, grad_norm=4.085259914398193, loss=3.6303534507751465 -I0513 03:48:54.636847 139687754716928 logging_writer.py:48] [69400] global_step=69400, grad_norm=18.792741775512695, loss=3.789513111114502 -I0513 03:49:36.384330 139687763109632 logging_writer.py:48] [69500] global_step=69500, grad_norm=5.176723480224609, loss=3.9641401767730713 -I0513 03:50:22.900968 139687754716928 logging_writer.py:48] [69600] global_step=69600, grad_norm=3.663153648376465, loss=3.7188801765441895 -I0513 03:51:04.447124 139687763109632 logging_writer.py:48] [69700] global_step=69700, grad_norm=6.02840518951416, loss=3.973231792449951 -I0513 03:51:50.404557 139687754716928 logging_writer.py:48] [69800] global_step=69800, grad_norm=3.8252248764038086, loss=3.727902412414551 -I0513 03:52:37.183639 139687763109632 logging_writer.py:48] [69900] global_step=69900, grad_norm=7.37995719909668, loss=3.767214298248291 -I0513 03:53:18.881579 139687754716928 logging_writer.py:48] [70000] global_step=70000, grad_norm=4.974602699279785, loss=3.70261812210083 -I0513 03:53:56.171545 139687763109632 logging_writer.py:48] [70100] global_step=70100, grad_norm=5.421108245849609, loss=3.687434196472168 -I0513 03:54:33.511364 139687754716928 logging_writer.py:48] [70200] global_step=70200, grad_norm=5.935712814331055, loss=3.8266892433166504 -I0513 03:55:10.331348 139687763109632 logging_writer.py:48] [70300] global_step=70300, grad_norm=3.9207327365875244, loss=3.6833205223083496 -I0513 03:55:47.629682 139687754716928 logging_writer.py:48] [70400] global_step=70400, grad_norm=4.359708309173584, loss=3.6687774658203125 -I0513 03:56:24.985654 139687763109632 logging_writer.py:48] [70500] global_step=70500, grad_norm=4.833279609680176, loss=3.8540220260620117 -I0513 03:57:06.082619 139687754716928 logging_writer.py:48] [70600] global_step=70600, grad_norm=2.832794189453125, loss=3.691422939300537 -I0513 03:57:47.145566 139687763109632 logging_writer.py:48] [70700] global_step=70700, grad_norm=4.827691078186035, loss=3.8247017860412598 -I0513 03:58:24.871616 139687754716928 logging_writer.py:48] [70800] global_step=70800, grad_norm=3.690662384033203, loss=3.655728578567505 -I0513 03:59:01.782052 139687763109632 logging_writer.py:48] [70900] global_step=70900, grad_norm=4.863901615142822, loss=3.7457680702209473 -I0513 03:59:39.092246 139687754716928 logging_writer.py:48] [71000] global_step=71000, grad_norm=6.010776519775391, loss=3.658968448638916 -I0513 04:00:20.578174 139687763109632 logging_writer.py:48] [71100] global_step=71100, grad_norm=3.585256814956665, loss=3.7620131969451904 -I0513 04:00:57.515112 139687754716928 logging_writer.py:48] [71200] global_step=71200, grad_norm=3.343449831008911, loss=3.6127138137817383 -I0513 04:01:34.430493 139687763109632 logging_writer.py:48] [71300] global_step=71300, grad_norm=7.281449794769287, loss=3.798952579498291 -I0513 04:02:16.183268 139687754716928 logging_writer.py:48] [71400] global_step=71400, grad_norm=5.109241485595703, loss=3.8709230422973633 -I0513 04:02:57.128647 139687763109632 logging_writer.py:48] [71500] global_step=71500, grad_norm=3.3054211139678955, loss=3.697056293487549 -I0513 04:03:38.162150 139687754716928 logging_writer.py:48] [71600] global_step=71600, grad_norm=11.261923789978027, loss=3.888277053833008 -I0513 04:04:20.033641 139687763109632 logging_writer.py:48] [71700] global_step=71700, grad_norm=4.706143379211426, loss=3.6927242279052734 -I0513 04:05:01.148424 139687754716928 logging_writer.py:48] [71800] global_step=71800, grad_norm=4.710484504699707, loss=3.7248380184173584 -I0513 04:05:42.273637 139687763109632 logging_writer.py:48] [71900] global_step=71900, grad_norm=4.517912864685059, loss=3.725945472717285 -I0513 04:06:23.982488 139687754716928 logging_writer.py:48] [72000] global_step=72000, grad_norm=5.608820915222168, loss=3.7892253398895264 -I0513 04:07:04.993422 139687763109632 logging_writer.py:48] [72100] global_step=72100, grad_norm=5.425504684448242, loss=3.51816463470459 -I0513 04:07:50.267889 139687754716928 logging_writer.py:48] [72200] global_step=72200, grad_norm=4.031135082244873, loss=3.7068252563476562 -I0513 04:08:27.838741 139687763109632 logging_writer.py:48] [72300] global_step=72300, grad_norm=4.028103351593018, loss=3.636160373687744 -I0513 04:09:12.987416 139687754716928 logging_writer.py:48] [72400] global_step=72400, grad_norm=3.7963147163391113, loss=3.634131669998169 -I0513 04:09:49.913715 139687763109632 logging_writer.py:48] [72500] global_step=72500, grad_norm=6.4693684577941895, loss=3.7034177780151367 -I0513 04:10:27.576132 139687754716928 logging_writer.py:48] [72600] global_step=72600, grad_norm=4.926512718200684, loss=3.801273822784424 -I0513 04:11:04.513990 139687763109632 logging_writer.py:48] [72700] global_step=72700, grad_norm=4.696149826049805, loss=3.6061344146728516 -I0513 04:11:41.423671 139687754716928 logging_writer.py:48] [72800] global_step=72800, grad_norm=5.651340007781982, loss=3.8129777908325195 -I0513 04:12:18.971425 139687763109632 logging_writer.py:48] [72900] global_step=72900, grad_norm=5.109965801239014, loss=3.667635440826416 -I0513 04:13:00.077010 139687754716928 logging_writer.py:48] [73000] global_step=73000, grad_norm=8.042091369628906, loss=3.9227042198181152 -I0513 04:13:45.318465 139687763109632 logging_writer.py:48] [73100] global_step=73100, grad_norm=3.846865653991699, loss=3.9130213260650635 -I0513 04:14:27.260626 139687754716928 logging_writer.py:48] [73200] global_step=73200, grad_norm=4.8641557693481445, loss=3.607619524002075 -I0513 04:15:08.337087 139687763109632 logging_writer.py:48] [73300] global_step=73300, grad_norm=3.925283432006836, loss=3.836091995239258 -I0513 04:15:45.161854 139687754716928 logging_writer.py:48] [73400] global_step=73400, grad_norm=3.7207236289978027, loss=3.652616500854492 -I0513 04:16:26.770192 139687763109632 logging_writer.py:48] [73500] global_step=73500, grad_norm=4.19284725189209, loss=3.866607189178467 -I0513 04:17:08.091065 139687754716928 logging_writer.py:48] [73600] global_step=73600, grad_norm=3.4965531826019287, loss=3.7043139934539795 -I0513 04:17:49.884943 139687763109632 logging_writer.py:48] [73700] global_step=73700, grad_norm=5.1119489669799805, loss=3.814199924468994 -I0513 04:18:30.834842 139687754716928 logging_writer.py:48] [73800] global_step=73800, grad_norm=4.591399669647217, loss=3.7854831218719482 -I0513 04:19:11.884196 139687763109632 logging_writer.py:48] [73900] global_step=73900, grad_norm=20.22456932067871, loss=3.852537155151367 -I0513 04:19:52.965245 139687754716928 logging_writer.py:48] [74000] global_step=74000, grad_norm=5.517630577087402, loss=3.73464035987854 -I0513 04:20:27.605533 139903809070272 spec.py:333] Evaluating on the training split. -I0513 04:20:38.030676 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 04:21:38.742443 139903809070272 spec.py:363] Evaluating on the test split. -I0513 04:21:39.634954 139903809070272 submission_runner.py:516] Time since start: 31243.71s, Step: 74083, {'train/accuracy': Array(0.0017339, dtype=float32), 'train/loss': Array(7.922397, dtype=float32), 'validation/accuracy': Array(0.00178, dtype=float32), 'validation/loss': Array(7.906208, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0016, dtype=float32), 'test/loss': Array(7.9134455, dtype=float32), 'test/num_examples': 10000, 'score': 30004.491594314575, 'total_duration': 31243.708920955658, 'accumulated_submission_time': 30004.491594314575, 'accumulated_eval_time': 1237.9831821918488, 'accumulated_logging_time': 0.5660214424133301} -I0513 04:21:39.665676 139687763109632 logging_writer.py:48] [74083] accumulated_eval_time=1237.98, accumulated_logging_time=0.566021, accumulated_submission_time=30004.5, global_step=74083, preemption_count=0, score=30004.5, test/accuracy=0.001600000075995922, test/loss=7.913445472717285, test/num_examples=10000, total_duration=31243.7, train/accuracy=0.001733896671794355, train/loss=7.922397136688232, validation/accuracy=0.001779999933205545, validation/loss=7.906208038330078, validation/num_examples=50000 -I0513 04:21:46.748739 139687754716928 logging_writer.py:48] [74100] global_step=74100, grad_norm=3.696587562561035, loss=3.816995859146118 -I0513 04:22:23.557995 139687763109632 logging_writer.py:48] [74200] global_step=74200, grad_norm=3.931011199951172, loss=3.6546401977539062 -I0513 04:23:04.656805 139687754716928 logging_writer.py:48] [74300] global_step=74300, grad_norm=5.11812686920166, loss=3.7213449478149414 -I0513 04:23:46.336438 139687763109632 logging_writer.py:48] [74400] global_step=74400, grad_norm=8.482263565063477, loss=3.7857861518859863 -I0513 04:24:27.233314 139687754716928 logging_writer.py:48] [74500] global_step=74500, grad_norm=3.7454543113708496, loss=3.7504117488861084 -I0513 04:25:08.115726 139687763109632 logging_writer.py:48] [74600] global_step=74600, grad_norm=5.7869553565979, loss=3.6668639183044434 -I0513 04:25:49.871312 139687754716928 logging_writer.py:48] [74700] global_step=74700, grad_norm=4.551324367523193, loss=3.9294753074645996 -I0513 04:26:30.787429 139687763109632 logging_writer.py:48] [74800] global_step=74800, grad_norm=2.7466039657592773, loss=3.7471697330474854 -I0513 04:27:11.790240 139687754716928 logging_writer.py:48] [74900] global_step=74900, grad_norm=4.2196574211120605, loss=3.7441413402557373 -I0513 04:27:53.224572 139687763109632 logging_writer.py:48] [75000] global_step=75000, grad_norm=17.691320419311523, loss=3.6279053688049316 -I0513 04:28:30.391528 139687754716928 logging_writer.py:48] [75100] global_step=75100, grad_norm=3.8422834873199463, loss=3.7112202644348145 -I0513 04:29:14.962228 139687763109632 logging_writer.py:48] [75200] global_step=75200, grad_norm=4.25221061706543, loss=3.769153594970703 -I0513 04:29:56.847985 139687754716928 logging_writer.py:48] [75300] global_step=75300, grad_norm=5.108579635620117, loss=3.73201584815979 -I0513 04:30:41.848664 139687763109632 logging_writer.py:48] [75400] global_step=75400, grad_norm=4.323288440704346, loss=3.716930866241455 -I0513 04:31:22.776936 139687754716928 logging_writer.py:48] [75500] global_step=75500, grad_norm=8.185215950012207, loss=3.7979910373687744 -I0513 04:32:04.310924 139687763109632 logging_writer.py:48] [75600] global_step=75600, grad_norm=3.0675158500671387, loss=3.811805248260498 -I0513 04:32:45.394242 139687754716928 logging_writer.py:48] [75700] global_step=75700, grad_norm=4.543263912200928, loss=3.658810615539551 -I0513 04:33:31.684240 139687763109632 logging_writer.py:48] [75800] global_step=75800, grad_norm=3.840331792831421, loss=3.8873867988586426 -I0513 04:34:14.048708 139687754716928 logging_writer.py:48] [75900] global_step=75900, grad_norm=3.8007187843322754, loss=3.6675500869750977 -I0513 04:35:00.214841 139687763109632 logging_writer.py:48] [76000] global_step=76000, grad_norm=5.232848644256592, loss=3.833155632019043 -I0513 04:35:41.856791 139687754716928 logging_writer.py:48] [76100] global_step=76100, grad_norm=4.129121780395508, loss=3.6824400424957275 -I0513 04:36:23.910207 139687763109632 logging_writer.py:48] [76200] global_step=76200, grad_norm=5.218644142150879, loss=3.8238728046417236 -I0513 04:37:05.921591 139687754716928 logging_writer.py:48] [76300] global_step=76300, grad_norm=5.4879560470581055, loss=3.9197707176208496 -I0513 04:37:47.643528 139687763109632 logging_writer.py:48] [76400] global_step=76400, grad_norm=6.472754001617432, loss=3.8573076725006104 -I0513 04:38:25.048589 139687754716928 logging_writer.py:48] [76500] global_step=76500, grad_norm=5.15786600112915, loss=3.5851941108703613 -I0513 04:39:02.244777 139687763109632 logging_writer.py:48] [76600] global_step=76600, grad_norm=7.1468634605407715, loss=4.014366626739502 -I0513 04:39:39.111174 139687754716928 logging_writer.py:48] [76700] global_step=76700, grad_norm=6.552323341369629, loss=3.711240530014038 -I0513 04:40:20.980348 139687763109632 logging_writer.py:48] [76800] global_step=76800, grad_norm=5.8571062088012695, loss=3.854034900665283 -I0513 04:40:58.219192 139687754716928 logging_writer.py:48] [76900] global_step=76900, grad_norm=4.977330207824707, loss=3.748668670654297 -I0513 04:41:39.641491 139687763109632 logging_writer.py:48] [77000] global_step=77000, grad_norm=3.4422969818115234, loss=3.627591371536255 -I0513 04:42:17.072228 139687754716928 logging_writer.py:48] [77100] global_step=77100, grad_norm=4.839957237243652, loss=3.6837785243988037 -I0513 04:42:54.244621 139687763109632 logging_writer.py:48] [77200] global_step=77200, grad_norm=4.6235127449035645, loss=3.783010482788086 -I0513 04:43:31.110712 139687754716928 logging_writer.py:48] [77300] global_step=77300, grad_norm=5.986728668212891, loss=3.80226993560791 -I0513 04:44:08.476222 139687763109632 logging_writer.py:48] [77400] global_step=77400, grad_norm=2.0977659225463867, loss=3.6201581954956055 -I0513 04:44:45.648968 139687754716928 logging_writer.py:48] [77500] global_step=77500, grad_norm=7.811527252197266, loss=3.852595567703247 -I0513 04:45:27.151043 139687763109632 logging_writer.py:48] [77600] global_step=77600, grad_norm=4.759081840515137, loss=3.7891783714294434 -I0513 04:46:10.977020 139687754716928 logging_writer.py:48] [77700] global_step=77700, grad_norm=4.0051374435424805, loss=3.8065431118011475 -I0513 04:46:50.360161 139687763109632 logging_writer.py:48] [77800] global_step=77800, grad_norm=3.4594359397888184, loss=3.6744863986968994 -I0513 04:47:32.188264 139687754716928 logging_writer.py:48] [77900] global_step=77900, grad_norm=5.619726657867432, loss=3.7515532970428467 -I0513 04:48:14.091310 139687763109632 logging_writer.py:48] [78000] global_step=78000, grad_norm=4.065164566040039, loss=3.743877410888672 -I0513 04:48:51.285283 139687754716928 logging_writer.py:48] [78100] global_step=78100, grad_norm=6.501830577850342, loss=3.8045644760131836 -I0513 04:49:28.133442 139687763109632 logging_writer.py:48] [78200] global_step=78200, grad_norm=4.1934614181518555, loss=3.775304079055786 -I0513 04:50:05.508218 139687754716928 logging_writer.py:48] [78300] global_step=78300, grad_norm=3.4277589321136475, loss=3.709949493408203 -I0513 04:50:42.714861 139687763109632 logging_writer.py:48] [78400] global_step=78400, grad_norm=5.607649803161621, loss=3.745128870010376 -I0513 04:51:28.876314 139687754716928 logging_writer.py:48] [78500] global_step=78500, grad_norm=18.03079605102539, loss=3.8331871032714844 -I0513 04:52:12.970727 139687763109632 logging_writer.py:48] [78600] global_step=78600, grad_norm=3.556236982345581, loss=3.6804370880126953 -I0513 04:52:56.413807 139687754716928 logging_writer.py:48] [78700] global_step=78700, grad_norm=4.853753089904785, loss=3.884110450744629 -I0513 04:53:37.865248 139687763109632 logging_writer.py:48] [78800] global_step=78800, grad_norm=11.598162651062012, loss=3.681364059448242 -I0513 04:54:19.839126 139687754716928 logging_writer.py:48] [78900] global_step=78900, grad_norm=3.5925111770629883, loss=3.6197612285614014 -I0513 04:54:55.781183 139903809070272 spec.py:333] Evaluating on the training split. -I0513 04:55:05.863082 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 04:56:01.597747 139903809070272 spec.py:363] Evaluating on the test split. -I0513 04:56:02.493306 139903809070272 submission_runner.py:516] Time since start: 33306.57s, Step: 78998, {'train/accuracy': Array(0.00201291, dtype=float32), 'train/loss': Array(7.8298244, dtype=float32), 'validation/accuracy': Array(0.002, dtype=float32), 'validation/loss': Array(7.8241224, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0018, dtype=float32), 'test/loss': Array(7.8329372, dtype=float32), 'test/num_examples': 10000, 'score': 32000.55635213852, 'total_duration': 33306.56731748581, 'accumulated_submission_time': 32000.55635213852, 'accumulated_eval_time': 1304.693614244461, 'accumulated_logging_time': 0.6050024032592773} -I0513 04:56:02.523547 139687763109632 logging_writer.py:48] [78998] accumulated_eval_time=1304.69, accumulated_logging_time=0.605002, accumulated_submission_time=32000.6, global_step=78998, preemption_count=0, score=32000.6, test/accuracy=0.0018000000854954123, test/loss=7.832937240600586, test/num_examples=10000, total_duration=33306.6, train/accuracy=0.002012914512306452, train/loss=7.829824447631836, validation/accuracy=0.001999999862164259, validation/loss=7.824122428894043, validation/num_examples=50000 -I0513 04:56:03.747354 139687754716928 logging_writer.py:48] [79000] global_step=79000, grad_norm=4.021106719970703, loss=3.710181474685669 -I0513 04:56:40.712427 139687763109632 logging_writer.py:48] [79100] global_step=79100, grad_norm=4.4491353034973145, loss=3.824151039123535 -I0513 04:57:18.566381 139687754716928 logging_writer.py:48] [79200] global_step=79200, grad_norm=4.045343399047852, loss=3.7108495235443115 -I0513 04:58:00.226487 139687763109632 logging_writer.py:48] [79300] global_step=79300, grad_norm=4.000266075134277, loss=3.6974456310272217 -I0513 04:58:37.205353 139687754716928 logging_writer.py:48] [79400] global_step=79400, grad_norm=3.6480021476745605, loss=3.628716230392456 -I0513 04:59:14.673404 139687763109632 logging_writer.py:48] [79500] global_step=79500, grad_norm=5.4434428215026855, loss=3.9547457695007324 -I0513 04:59:52.025651 139687754716928 logging_writer.py:48] [79600] global_step=79600, grad_norm=5.321559906005859, loss=3.9649364948272705 -I0513 05:00:33.015294 139687763109632 logging_writer.py:48] [79700] global_step=79700, grad_norm=9.374236106872559, loss=3.72347354888916 -I0513 05:01:10.520588 139687754716928 logging_writer.py:48] [79800] global_step=79800, grad_norm=5.131307601928711, loss=3.656935214996338 -I0513 05:01:47.740612 139687763109632 logging_writer.py:48] [79900] global_step=79900, grad_norm=3.6012301445007324, loss=3.6748557090759277 -I0513 05:02:24.716174 139687754716928 logging_writer.py:48] [80000] global_step=80000, grad_norm=4.468842029571533, loss=3.564425230026245 -I0513 05:03:02.386880 139687763109632 logging_writer.py:48] [80100] global_step=80100, grad_norm=5.765340805053711, loss=3.719902753829956 -I0513 05:03:39.745652 139687754716928 logging_writer.py:48] [80200] global_step=80200, grad_norm=5.464033603668213, loss=3.871671199798584 -I0513 05:04:16.648150 139687763109632 logging_writer.py:48] [80300] global_step=80300, grad_norm=3.892550468444824, loss=3.7248075008392334 -I0513 05:04:54.035664 139687754716928 logging_writer.py:48] [80400] global_step=80400, grad_norm=3.8439583778381348, loss=3.5787971019744873 -I0513 05:05:30.938423 139687763109632 logging_writer.py:48] [80500] global_step=80500, grad_norm=4.649686813354492, loss=3.886613607406616 -I0513 05:06:12.090916 139687754716928 logging_writer.py:48] [80600] global_step=80600, grad_norm=3.801652431488037, loss=3.755026340484619 -I0513 05:06:49.446513 139687763109632 logging_writer.py:48] [80700] global_step=80700, grad_norm=3.3088977336883545, loss=3.6975557804107666 -I0513 05:07:26.695087 139687754716928 logging_writer.py:48] [80800] global_step=80800, grad_norm=3.08111572265625, loss=3.802811622619629 -I0513 05:08:07.737676 139687763109632 logging_writer.py:48] [80900] global_step=80900, grad_norm=3.750645637512207, loss=3.696290969848633 -I0513 05:08:45.060894 139687754716928 logging_writer.py:48] [81000] global_step=81000, grad_norm=4.108607769012451, loss=3.6531448364257812 -I0513 05:09:22.265210 139687763109632 logging_writer.py:48] [81100] global_step=81100, grad_norm=4.126155853271484, loss=3.651721477508545 -I0513 05:09:59.058708 139687754716928 logging_writer.py:48] [81200] global_step=81200, grad_norm=7.018787860870361, loss=3.776949644088745 -I0513 05:10:39.601318 139687763109632 logging_writer.py:48] [81300] global_step=81300, grad_norm=4.060508728027344, loss=3.6988863945007324 -I0513 05:11:21.434932 139687754716928 logging_writer.py:48] [81400] global_step=81400, grad_norm=3.68337082862854, loss=3.738966464996338 -I0513 05:12:07.396572 139687763109632 logging_writer.py:48] [81500] global_step=81500, grad_norm=3.880741596221924, loss=3.650499105453491 -I0513 05:12:49.285540 139687754716928 logging_writer.py:48] [81600] global_step=81600, grad_norm=4.316303253173828, loss=3.7095718383789062 -I0513 05:13:30.982569 139687763109632 logging_writer.py:48] [81700] global_step=81700, grad_norm=4.954980373382568, loss=3.8093485832214355 -I0513 05:14:12.373542 139687754716928 logging_writer.py:48] [81800] global_step=81800, grad_norm=3.854111909866333, loss=3.72955322265625 -I0513 05:14:54.184734 139687763109632 logging_writer.py:48] [81900] global_step=81900, grad_norm=2.5804994106292725, loss=3.6572937965393066 -I0513 05:15:35.825661 139687754716928 logging_writer.py:48] [82000] global_step=82000, grad_norm=15.087784767150879, loss=3.63972806930542 -I0513 05:16:21.033038 139687763109632 logging_writer.py:48] [82100] global_step=82100, grad_norm=8.31338119506836, loss=3.8191564083099365 -I0513 05:17:00.091399 139687754716928 logging_writer.py:48] [82200] global_step=82200, grad_norm=4.265251159667969, loss=3.799915313720703 -I0513 05:17:44.266142 139687763109632 logging_writer.py:48] [82300] global_step=82300, grad_norm=5.545849323272705, loss=3.770085334777832 -I0513 05:18:25.782516 139687754716928 logging_writer.py:48] [82400] global_step=82400, grad_norm=3.4135382175445557, loss=3.7258217334747314 -I0513 05:19:07.613044 139687763109632 logging_writer.py:48] [82500] global_step=82500, grad_norm=3.7948663234710693, loss=3.6742143630981445 -I0513 05:19:49.071393 139687754716928 logging_writer.py:48] [82600] global_step=82600, grad_norm=7.294571399688721, loss=3.7484331130981445 -I0513 05:20:30.773587 139687763109632 logging_writer.py:48] [82700] global_step=82700, grad_norm=4.471800327301025, loss=3.7154083251953125 -I0513 05:21:12.659042 139687754716928 logging_writer.py:48] [82800] global_step=82800, grad_norm=9.75173568725586, loss=4.117370128631592 -I0513 05:21:54.377877 139687763109632 logging_writer.py:48] [82900] global_step=82900, grad_norm=8.064929962158203, loss=3.8888378143310547 -I0513 05:22:35.744275 139687754716928 logging_writer.py:48] [83000] global_step=83000, grad_norm=5.980416774749756, loss=3.7389822006225586 -I0513 05:23:13.087012 139687763109632 logging_writer.py:48] [83100] global_step=83100, grad_norm=4.623166084289551, loss=3.791919708251953 -I0513 05:23:50.355633 139687754716928 logging_writer.py:48] [83200] global_step=83200, grad_norm=4.530366897583008, loss=3.702111005783081 -I0513 05:24:31.699156 139687763109632 logging_writer.py:48] [83300] global_step=83300, grad_norm=4.3807830810546875, loss=3.7214884757995605 -I0513 05:25:13.524852 139687754716928 logging_writer.py:48] [83400] global_step=83400, grad_norm=4.154781818389893, loss=3.698267936706543 -I0513 05:25:55.232144 139687763109632 logging_writer.py:48] [83500] global_step=83500, grad_norm=3.7099661827087402, loss=3.7450034618377686 -I0513 05:26:36.742040 139687754716928 logging_writer.py:48] [83600] global_step=83600, grad_norm=2.1874239444732666, loss=3.6890947818756104 -I0513 05:27:18.681507 139687763109632 logging_writer.py:48] [83700] global_step=83700, grad_norm=3.9286623001098633, loss=3.7358787059783936 -I0513 05:28:00.432723 139687754716928 logging_writer.py:48] [83800] global_step=83800, grad_norm=3.6030776500701904, loss=3.8510851860046387 -I0513 05:28:41.935054 139687763109632 logging_writer.py:48] [83900] global_step=83900, grad_norm=5.4002227783203125, loss=3.688281774520874 -I0513 05:29:19.319223 139687754716928 logging_writer.py:48] [84000] global_step=84000, grad_norm=4.020452499389648, loss=3.7614879608154297 -I0513 05:29:19.333402 139903809070272 spec.py:333] Evaluating on the training split. -I0513 05:29:28.909674 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 05:30:44.453424 139903809070272 spec.py:363] Evaluating on the test split. -I0513 05:30:45.348510 139903809070272 submission_runner.py:516] Time since start: 35389.42s, Step: 84001, {'train/accuracy': Array(0.00195312, dtype=float32), 'train/loss': Array(7.742617, dtype=float32), 'validation/accuracy': Array(0.00208, dtype=float32), 'validation/loss': Array(7.741793, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(7.752259, dtype=float32), 'test/num_examples': 10000, 'score': 33997.315504550934, 'total_duration': 35389.42226982117, 'accumulated_submission_time': 33997.315504550934, 'accumulated_eval_time': 1390.706758737564, 'accumulated_logging_time': 0.6433432102203369} -I0513 05:30:45.379862 139687763109632 logging_writer.py:48] [84001] accumulated_eval_time=1390.71, accumulated_logging_time=0.643343, accumulated_submission_time=33997.3, global_step=84001, preemption_count=0, score=33997.3, test/accuracy=0.0019000000320374966, test/loss=7.752258777618408, test/num_examples=10000, total_duration=35389.4, train/accuracy=0.001953125, train/loss=7.742617130279541, validation/accuracy=0.0020800000056624413, validation/loss=7.741793155670166, validation/num_examples=50000 -I0513 05:31:23.473627 139687754716928 logging_writer.py:48] [84100] global_step=84100, grad_norm=7.55328369140625, loss=3.7580602169036865 -I0513 05:32:03.012924 139687763109632 logging_writer.py:48] [84200] global_step=84200, grad_norm=3.9432058334350586, loss=3.7173566818237305 -I0513 05:32:44.354043 139687754716928 logging_writer.py:48] [84300] global_step=84300, grad_norm=5.370382785797119, loss=3.7508950233459473 -I0513 05:33:22.197653 139687763109632 logging_writer.py:48] [84400] global_step=84400, grad_norm=4.089113712310791, loss=3.8355252742767334 -I0513 05:33:59.438491 139687754716928 logging_writer.py:48] [84500] global_step=84500, grad_norm=5.105422019958496, loss=3.705106496810913 -I0513 05:34:36.795109 139687763109632 logging_writer.py:48] [84600] global_step=84600, grad_norm=3.110154867172241, loss=3.6448323726654053 -I0513 05:35:13.927181 139687754716928 logging_writer.py:48] [84700] global_step=84700, grad_norm=3.443512201309204, loss=3.7893311977386475 -I0513 05:35:50.921194 139687763109632 logging_writer.py:48] [84800] global_step=84800, grad_norm=6.473030090332031, loss=3.9054017066955566 -I0513 05:36:28.306729 139687754716928 logging_writer.py:48] [84900] global_step=84900, grad_norm=4.874751091003418, loss=3.818531036376953 -I0513 05:37:05.568743 139687763109632 logging_writer.py:48] [85000] global_step=85000, grad_norm=3.658778667449951, loss=3.7459540367126465 -I0513 05:37:42.507008 139687754716928 logging_writer.py:48] [85100] global_step=85100, grad_norm=5.413975238800049, loss=3.7489013671875 -I0513 05:38:19.954231 139687763109632 logging_writer.py:48] [85200] global_step=85200, grad_norm=4.599834442138672, loss=3.739680051803589 -I0513 05:38:56.897384 139687754716928 logging_writer.py:48] [85300] global_step=85300, grad_norm=5.484665393829346, loss=3.56758451461792 -I0513 05:39:34.178903 139687763109632 logging_writer.py:48] [85400] global_step=85400, grad_norm=3.779583692550659, loss=3.740471839904785 -I0513 05:40:11.496320 139687754716928 logging_writer.py:48] [85500] global_step=85500, grad_norm=3.9007351398468018, loss=3.7400858402252197 -I0513 05:40:48.785992 139687763109632 logging_writer.py:48] [85600] global_step=85600, grad_norm=4.118542671203613, loss=3.9590888023376465 -I0513 05:41:25.650681 139687754716928 logging_writer.py:48] [85700] global_step=85700, grad_norm=5.297441482543945, loss=3.9031548500061035 -I0513 05:42:02.967468 139687763109632 logging_writer.py:48] [85800] global_step=85800, grad_norm=32.51613998413086, loss=3.9326202869415283 -I0513 05:42:40.228915 139687754716928 logging_writer.py:48] [85900] global_step=85900, grad_norm=4.045722007751465, loss=3.7188029289245605 -I0513 05:43:17.161682 139687763109632 logging_writer.py:48] [86000] global_step=86000, grad_norm=4.867930889129639, loss=3.6247305870056152 -I0513 05:43:54.576155 139687754716928 logging_writer.py:48] [86100] global_step=86100, grad_norm=3.8269338607788086, loss=3.7825634479522705 -I0513 05:44:36.122668 139687763109632 logging_writer.py:48] [86200] global_step=86200, grad_norm=10.42080307006836, loss=3.618262529373169 -I0513 05:45:12.990963 139687754716928 logging_writer.py:48] [86300] global_step=86300, grad_norm=4.942002773284912, loss=3.837939977645874 -I0513 05:45:54.669927 139687763109632 logging_writer.py:48] [86400] global_step=86400, grad_norm=3.6761374473571777, loss=3.812711715698242 -I0513 05:46:36.116093 139687754716928 logging_writer.py:48] [86500] global_step=86500, grad_norm=3.9985268115997314, loss=3.6732237339019775 -I0513 05:47:17.239797 139687763109632 logging_writer.py:48] [86600] global_step=86600, grad_norm=4.869739532470703, loss=3.7111258506774902 -I0513 05:47:58.998490 139687754716928 logging_writer.py:48] [86700] global_step=86700, grad_norm=3.046722173690796, loss=3.7269818782806396 -I0513 05:48:36.091946 139687763109632 logging_writer.py:48] [86800] global_step=86800, grad_norm=4.350020408630371, loss=3.674987554550171 -I0513 05:49:12.930180 139687754716928 logging_writer.py:48] [86900] global_step=86900, grad_norm=3.872469902038574, loss=3.838085174560547 -I0513 05:49:50.326902 139687763109632 logging_writer.py:48] [87000] global_step=87000, grad_norm=5.204840660095215, loss=3.782299041748047 -I0513 05:50:27.517282 139687754716928 logging_writer.py:48] [87100] global_step=87100, grad_norm=5.171623706817627, loss=3.999424934387207 -I0513 05:51:08.821216 139687763109632 logging_writer.py:48] [87200] global_step=87200, grad_norm=4.0187530517578125, loss=3.7867395877838135 -I0513 05:51:50.457051 139687754716928 logging_writer.py:48] [87300] global_step=87300, grad_norm=6.306530952453613, loss=3.6776437759399414 -I0513 05:52:27.301726 139687763109632 logging_writer.py:48] [87400] global_step=87400, grad_norm=4.656188011169434, loss=3.7544002532958984 -I0513 05:53:04.432259 139687754716928 logging_writer.py:48] [87500] global_step=87500, grad_norm=4.4258246421813965, loss=3.8112988471984863 -I0513 05:53:41.770452 139687763109632 logging_writer.py:48] [87600] global_step=87600, grad_norm=4.473336219787598, loss=3.8432085514068604 -I0513 05:54:18.930697 139687754716928 logging_writer.py:48] [87700] global_step=87700, grad_norm=3.483896017074585, loss=3.6533632278442383 -I0513 05:54:55.802183 139687763109632 logging_writer.py:48] [87800] global_step=87800, grad_norm=3.2544803619384766, loss=3.78635311126709 -I0513 05:55:33.104928 139687754716928 logging_writer.py:48] [87900] global_step=87900, grad_norm=2.782379627227783, loss=3.72377610206604 -I0513 05:56:10.273621 139687763109632 logging_writer.py:48] [88000] global_step=88000, grad_norm=4.171748161315918, loss=3.7442374229431152 -I0513 05:56:47.156424 139687754716928 logging_writer.py:48] [88100] global_step=88100, grad_norm=5.782363414764404, loss=3.7264509201049805 -I0513 05:57:24.527137 139687763109632 logging_writer.py:48] [88200] global_step=88200, grad_norm=3.291214942932129, loss=3.7182369232177734 -I0513 05:58:01.795062 139687754716928 logging_writer.py:48] [88300] global_step=88300, grad_norm=10.135679244995117, loss=3.8133902549743652 -I0513 05:58:38.632496 139687763109632 logging_writer.py:48] [88400] global_step=88400, grad_norm=4.347157001495361, loss=3.6833624839782715 -I0513 05:59:15.956338 139687754716928 logging_writer.py:48] [88500] global_step=88500, grad_norm=13.583102226257324, loss=3.816171169281006 -I0513 05:59:53.222369 139687763109632 logging_writer.py:48] [88600] global_step=88600, grad_norm=5.213527679443359, loss=3.834340810775757 -I0513 06:00:30.012052 139687754716928 logging_writer.py:48] [88700] global_step=88700, grad_norm=5.15724515914917, loss=3.680812358856201 -I0513 06:01:07.415254 139687763109632 logging_writer.py:48] [88800] global_step=88800, grad_norm=3.9589874744415283, loss=3.637986183166504 -I0513 06:01:44.508160 139687754716928 logging_writer.py:48] [88900] global_step=88900, grad_norm=5.978626728057861, loss=3.8194069862365723 -I0513 06:02:21.367484 139687763109632 logging_writer.py:48] [89000] global_step=89000, grad_norm=3.4327266216278076, loss=3.704106330871582 -I0513 06:02:58.743329 139687754716928 logging_writer.py:48] [89100] global_step=89100, grad_norm=4.097993850708008, loss=3.8290226459503174 -I0513 06:03:35.978379 139687763109632 logging_writer.py:48] [89200] global_step=89200, grad_norm=4.746204376220703, loss=3.814223051071167 -I0513 06:04:01.351544 139903809070272 spec.py:333] Evaluating on the training split. -I0513 06:04:11.655478 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 06:05:00.670964 139903809070272 spec.py:363] Evaluating on the test split. -I0513 06:05:01.566745 139903809070272 submission_runner.py:516] Time since start: 37445.64s, Step: 89270, {'train/accuracy': Array(0.00205277, dtype=float32), 'train/loss': Array(7.701437, dtype=float32), 'validation/accuracy': Array(0.00222, dtype=float32), 'validation/loss': Array(7.7050667, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(7.7178745, dtype=float32), 'test/num_examples': 10000, 'score': 35993.23620557785, 'total_duration': 37445.64071536064, 'accumulated_submission_time': 35993.23620557785, 'accumulated_eval_time': 1450.920217514038, 'accumulated_logging_time': 0.6834356784820557} -I0513 06:05:01.600498 139687754716928 logging_writer.py:48] [89270] accumulated_eval_time=1450.92, accumulated_logging_time=0.683436, accumulated_submission_time=35993.2, global_step=89270, preemption_count=0, score=35993.2, test/accuracy=0.0019000000320374966, test/loss=7.717874526977539, test/num_examples=10000, total_duration=37445.6, train/accuracy=0.0020527741871774197, train/loss=7.701436996459961, validation/accuracy=0.002219999907538295, validation/loss=7.705066680908203, validation/num_examples=50000 -I0513 06:05:13.069705 139687763109632 logging_writer.py:48] [89300] global_step=89300, grad_norm=3.8505990505218506, loss=3.680452585220337 -I0513 06:05:54.712724 139687754716928 logging_writer.py:48] [89400] global_step=89400, grad_norm=2.8794517517089844, loss=3.7205636501312256 -I0513 06:06:35.609851 139687763109632 logging_writer.py:48] [89500] global_step=89500, grad_norm=4.069445610046387, loss=3.7814645767211914 -I0513 06:07:16.849178 139687754716928 logging_writer.py:48] [89600] global_step=89600, grad_norm=3.067946434020996, loss=3.8672256469726562 -I0513 06:07:58.582643 139687763109632 logging_writer.py:48] [89700] global_step=89700, grad_norm=3.982511281967163, loss=3.9328062534332275 -I0513 06:08:41.012457 139687754716928 logging_writer.py:48] [89800] global_step=89800, grad_norm=3.7572145462036133, loss=3.8066935539245605 -I0513 06:09:17.858498 139687763109632 logging_writer.py:48] [89900] global_step=89900, grad_norm=2.7074007987976074, loss=3.7679104804992676 -I0513 06:09:55.257833 139687754716928 logging_writer.py:48] [90000] global_step=90000, grad_norm=5.139956951141357, loss=3.7176876068115234 -I0513 06:10:32.178045 139687763109632 logging_writer.py:48] [90100] global_step=90100, grad_norm=2.949829339981079, loss=3.7700586318969727 -I0513 06:11:09.418323 139687754716928 logging_writer.py:48] [90200] global_step=90200, grad_norm=6.139378070831299, loss=3.796140670776367 -I0513 06:11:50.911125 139687763109632 logging_writer.py:48] [90300] global_step=90300, grad_norm=5.150022029876709, loss=3.7076916694641113 -I0513 06:12:32.978114 139687754716928 logging_writer.py:48] [90400] global_step=90400, grad_norm=2.9947092533111572, loss=3.6556406021118164 -I0513 06:13:09.764947 139687763109632 logging_writer.py:48] [90500] global_step=90500, grad_norm=5.591921806335449, loss=3.8734381198883057 -I0513 06:13:47.068437 139687754716928 logging_writer.py:48] [90600] global_step=90600, grad_norm=4.661604881286621, loss=3.7834393978118896 -I0513 06:14:24.283749 139687763109632 logging_writer.py:48] [90700] global_step=90700, grad_norm=3.0296759605407715, loss=3.638990879058838 -I0513 06:15:01.133841 139687754716928 logging_writer.py:48] [90800] global_step=90800, grad_norm=5.179849624633789, loss=3.7577571868896484 -I0513 06:15:38.561088 139687763109632 logging_writer.py:48] [90900] global_step=90900, grad_norm=9.363432884216309, loss=3.6611297130584717 -I0513 06:16:15.420556 139687754716928 logging_writer.py:48] [91000] global_step=91000, grad_norm=4.752152919769287, loss=3.716167688369751 -I0513 06:16:52.664390 139687763109632 logging_writer.py:48] [91100] global_step=91100, grad_norm=4.937257289886475, loss=3.7749457359313965 -I0513 06:17:30.038599 139687754716928 logging_writer.py:48] [91200] global_step=91200, grad_norm=7.168136119842529, loss=3.679614543914795 -I0513 06:18:07.330261 139687763109632 logging_writer.py:48] [91300] global_step=91300, grad_norm=3.387678623199463, loss=3.947810649871826 -I0513 06:18:44.209482 139687754716928 logging_writer.py:48] [91400] global_step=91400, grad_norm=4.504289150238037, loss=3.6985111236572266 -I0513 06:19:21.477025 139687763109632 logging_writer.py:48] [91500] global_step=91500, grad_norm=5.968008995056152, loss=3.7972850799560547 -I0513 06:19:58.301323 139687754716928 logging_writer.py:48] [91600] global_step=91600, grad_norm=8.734095573425293, loss=3.7522974014282227 -I0513 06:20:35.549128 139687763109632 logging_writer.py:48] [91700] global_step=91700, grad_norm=6.992635250091553, loss=3.727034091949463 -I0513 06:21:12.912655 139687754716928 logging_writer.py:48] [91800] global_step=91800, grad_norm=5.626187801361084, loss=3.8925585746765137 -I0513 06:21:50.056940 139687763109632 logging_writer.py:48] [91900] global_step=91900, grad_norm=4.888877868652344, loss=3.803011894226074 -I0513 06:22:26.877046 139687754716928 logging_writer.py:48] [92000] global_step=92000, grad_norm=6.231159210205078, loss=3.7174501419067383 -I0513 06:23:04.158683 139687763109632 logging_writer.py:48] [92100] global_step=92100, grad_norm=3.9198811054229736, loss=3.8386049270629883 -I0513 06:23:40.935051 139687754716928 logging_writer.py:48] [92200] global_step=92200, grad_norm=6.8617024421691895, loss=3.776503801345825 -I0513 06:24:18.106847 139687763109632 logging_writer.py:48] [92300] global_step=92300, grad_norm=5.418927192687988, loss=3.7680013179779053 -I0513 06:24:55.343113 139687754716928 logging_writer.py:48] [92400] global_step=92400, grad_norm=4.013881206512451, loss=3.8469505310058594 -I0513 06:25:32.117633 139687763109632 logging_writer.py:48] [92500] global_step=92500, grad_norm=2.490809202194214, loss=3.5705184936523438 -I0513 06:26:09.305657 139687754716928 logging_writer.py:48] [92600] global_step=92600, grad_norm=4.893141746520996, loss=3.9030792713165283 -I0513 06:26:46.641130 139687763109632 logging_writer.py:48] [92700] global_step=92700, grad_norm=3.648906946182251, loss=3.5375795364379883 -I0513 06:27:23.509501 139687754716928 logging_writer.py:48] [92800] global_step=92800, grad_norm=4.844003677368164, loss=3.7166409492492676 -I0513 06:28:00.813382 139687763109632 logging_writer.py:48] [92900] global_step=92900, grad_norm=4.258030891418457, loss=3.917231559753418 -I0513 06:28:38.225126 139687754716928 logging_writer.py:48] [93000] global_step=93000, grad_norm=3.370234251022339, loss=3.7809746265411377 -I0513 06:29:15.028522 139687763109632 logging_writer.py:48] [93100] global_step=93100, grad_norm=3.731139659881592, loss=3.610816478729248 -I0513 06:29:52.209125 139687754716928 logging_writer.py:48] [93200] global_step=93200, grad_norm=4.09384822845459, loss=3.6165645122528076 -I0513 06:30:29.499027 139687763109632 logging_writer.py:48] [93300] global_step=93300, grad_norm=5.425533294677734, loss=3.774170398712158 -I0513 06:31:06.463732 139687754716928 logging_writer.py:48] [93400] global_step=93400, grad_norm=4.7932963371276855, loss=3.8074324131011963 -I0513 06:31:43.731928 139687763109632 logging_writer.py:48] [93500] global_step=93500, grad_norm=3.4705705642700195, loss=3.6655778884887695 -I0513 06:32:21.070512 139687754716928 logging_writer.py:48] [93600] global_step=93600, grad_norm=3.1339187622070312, loss=3.7078182697296143 -I0513 06:32:57.866044 139687763109632 logging_writer.py:48] [93700] global_step=93700, grad_norm=5.676608562469482, loss=3.7883143424987793 -I0513 06:33:34.730331 139687754716928 logging_writer.py:48] [93800] global_step=93800, grad_norm=3.6876909732818604, loss=3.653183937072754 -I0513 06:34:12.318483 139687763109632 logging_writer.py:48] [93900] global_step=93900, grad_norm=3.286033868789673, loss=3.622105360031128 -I0513 06:34:49.197796 139687754716928 logging_writer.py:48] [94000] global_step=94000, grad_norm=4.78407096862793, loss=3.733415126800537 -I0513 06:35:26.747758 139687763109632 logging_writer.py:48] [94100] global_step=94100, grad_norm=3.138641119003296, loss=3.592461347579956 -I0513 06:36:04.130137 139687754716928 logging_writer.py:48] [94200] global_step=94200, grad_norm=5.464801788330078, loss=3.658163070678711 -I0513 06:36:41.006907 139687763109632 logging_writer.py:48] [94300] global_step=94300, grad_norm=4.975419521331787, loss=3.904366970062256 -I0513 06:37:18.344825 139687754716928 logging_writer.py:48] [94400] global_step=94400, grad_norm=4.250817775726318, loss=3.7946712970733643 -I0513 06:37:55.805081 139687763109632 logging_writer.py:48] [94500] global_step=94500, grad_norm=3.3583648204803467, loss=3.6458168029785156 -I0513 06:38:17.919889 139903809070272 spec.py:333] Evaluating on the training split. -I0513 06:38:28.394732 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 06:39:11.501953 139903809070272 spec.py:363] Evaluating on the test split. -I0513 06:39:12.393386 139903809070272 submission_runner.py:516] Time since start: 39496.47s, Step: 94561, {'train/accuracy': Array(0.00181362, dtype=float32), 'train/loss': Array(7.658918, dtype=float32), 'validation/accuracy': Array(0.00264, dtype=float32), 'validation/loss': Array(7.6551237, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(7.670427, dtype=float32), 'test/num_examples': 10000, 'score': 37989.504944086075, 'total_duration': 39496.46735692024, 'accumulated_submission_time': 37989.504944086075, 'accumulated_eval_time': 1505.3919687271118, 'accumulated_logging_time': 0.7256448268890381} -I0513 06:39:12.425568 139687754716928 logging_writer.py:48] [94561] accumulated_eval_time=1505.39, accumulated_logging_time=0.725645, accumulated_submission_time=37989.5, global_step=94561, preemption_count=0, score=37989.5, test/accuracy=0.0019000000320374966, test/loss=7.670426845550537, test/num_examples=10000, total_duration=39496.5, train/accuracy=0.0018136160215362906, train/loss=7.6589179039001465, validation/accuracy=0.002639999845996499, validation/loss=7.655123710632324, validation/num_examples=50000 -I0513 06:39:27.270431 139687763109632 logging_writer.py:48] [94600] global_step=94600, grad_norm=3.703909158706665, loss=3.8781142234802246 -I0513 06:40:04.587382 139687754716928 logging_writer.py:48] [94700] global_step=94700, grad_norm=12.766131401062012, loss=3.8258934020996094 -I0513 06:40:41.929373 139687763109632 logging_writer.py:48] [94800] global_step=94800, grad_norm=4.438161373138428, loss=3.7057266235351562 -I0513 06:41:18.852665 139687754716928 logging_writer.py:48] [94900] global_step=94900, grad_norm=12.796070098876953, loss=3.8659939765930176 -I0513 06:42:00.102376 139687763109632 logging_writer.py:48] [95000] global_step=95000, grad_norm=6.741845607757568, loss=3.7234621047973633 -I0513 06:42:37.531025 139687754716928 logging_writer.py:48] [95100] global_step=95100, grad_norm=5.626379013061523, loss=3.6420092582702637 -I0513 06:43:15.596575 139687763109632 logging_writer.py:48] [95200] global_step=95200, grad_norm=24.623329162597656, loss=3.8588132858276367 -I0513 06:43:53.429952 139687754716928 logging_writer.py:48] [95300] global_step=95300, grad_norm=4.399186134338379, loss=3.7811670303344727 -I0513 06:44:31.281400 139687763109632 logging_writer.py:48] [95400] global_step=95400, grad_norm=3.988271951675415, loss=3.864511489868164 -I0513 06:45:08.964326 139687754716928 logging_writer.py:48] [95500] global_step=95500, grad_norm=3.225506067276001, loss=3.7032737731933594 -I0513 06:45:47.041934 139687763109632 logging_writer.py:48] [95600] global_step=95600, grad_norm=4.847471714019775, loss=3.8203041553497314 -I0513 06:46:25.261284 139687754716928 logging_writer.py:48] [95700] global_step=95700, grad_norm=2.5323994159698486, loss=3.745797872543335 -I0513 06:47:02.888123 139687763109632 logging_writer.py:48] [95800] global_step=95800, grad_norm=4.304119110107422, loss=3.6176066398620605 -I0513 06:47:40.718801 139687754716928 logging_writer.py:48] [95900] global_step=95900, grad_norm=5.248993873596191, loss=3.7649459838867188 -I0513 06:48:22.810630 139687763109632 logging_writer.py:48] [96000] global_step=96000, grad_norm=2.998384714126587, loss=3.692094326019287 -I0513 06:48:59.760532 139687754716928 logging_writer.py:48] [96100] global_step=96100, grad_norm=4.023457050323486, loss=3.771495819091797 -I0513 06:49:42.595899 139687763109632 logging_writer.py:48] [96200] global_step=96200, grad_norm=4.86956787109375, loss=3.7292516231536865 -I0513 06:50:25.652423 139687754716928 logging_writer.py:48] [96300] global_step=96300, grad_norm=4.999479293823242, loss=3.700017213821411 -I0513 06:51:14.086517 139687763109632 logging_writer.py:48] [96400] global_step=96400, grad_norm=36.604976654052734, loss=3.927724838256836 -I0513 06:51:55.895933 139687754716928 logging_writer.py:48] [96500] global_step=96500, grad_norm=3.2503793239593506, loss=3.81792950630188 -I0513 06:52:33.336014 139687763109632 logging_writer.py:48] [96600] global_step=96600, grad_norm=3.2179625034332275, loss=3.714346408843994 -I0513 06:53:14.518846 139687754716928 logging_writer.py:48] [96700] global_step=96700, grad_norm=4.199678421020508, loss=3.8640036582946777 -I0513 06:53:55.976755 139687763109632 logging_writer.py:48] [96800] global_step=96800, grad_norm=5.025609016418457, loss=3.784147024154663 -I0513 06:54:37.712737 139687754716928 logging_writer.py:48] [96900] global_step=96900, grad_norm=3.6968889236450195, loss=3.8399059772491455 -I0513 06:55:18.738225 139687763109632 logging_writer.py:48] [97000] global_step=97000, grad_norm=3.9724557399749756, loss=3.6676583290100098 -I0513 06:56:00.249450 139687754716928 logging_writer.py:48] [97100] global_step=97100, grad_norm=6.457381248474121, loss=3.8307268619537354 -I0513 06:56:41.875123 139687763109632 logging_writer.py:48] [97200] global_step=97200, grad_norm=15.202038764953613, loss=3.8486833572387695 -I0513 06:57:23.066547 139687754716928 logging_writer.py:48] [97300] global_step=97300, grad_norm=3.295957088470459, loss=3.648118495941162 -I0513 06:58:04.546071 139687763109632 logging_writer.py:48] [97400] global_step=97400, grad_norm=5.11083984375, loss=3.801076650619507 -I0513 06:58:48.175428 139687754716928 logging_writer.py:48] [97500] global_step=97500, grad_norm=3.7680201530456543, loss=3.740736484527588 -I0513 06:59:31.034635 139687763109632 logging_writer.py:48] [97600] global_step=97600, grad_norm=3.7731950283050537, loss=3.628119468688965 -I0513 07:00:19.362910 139687754716928 logging_writer.py:48] [97700] global_step=97700, grad_norm=6.798786640167236, loss=3.9346425533294678 -I0513 07:01:01.314581 139687763109632 logging_writer.py:48] [97800] global_step=97800, grad_norm=3.533423662185669, loss=3.792713165283203 -I0513 07:01:42.758376 139687754716928 logging_writer.py:48] [97900] global_step=97900, grad_norm=3.629615306854248, loss=3.7033119201660156 -I0513 07:02:24.615814 139687763109632 logging_writer.py:48] [98000] global_step=98000, grad_norm=6.921454906463623, loss=3.879488468170166 -I0513 07:03:06.621750 139687754716928 logging_writer.py:48] [98100] global_step=98100, grad_norm=5.56049108505249, loss=3.8285961151123047 -I0513 07:03:48.123935 139687763109632 logging_writer.py:48] [98200] global_step=98200, grad_norm=5.798943519592285, loss=3.784205913543701 -I0513 07:04:34.365125 139687754716928 logging_writer.py:48] [98300] global_step=98300, grad_norm=4.619037628173828, loss=3.7910115718841553 -I0513 07:05:16.229334 139687763109632 logging_writer.py:48] [98400] global_step=98400, grad_norm=3.9360058307647705, loss=3.758904457092285 -I0513 07:05:57.611659 139687754716928 logging_writer.py:48] [98500] global_step=98500, grad_norm=3.0322351455688477, loss=3.6895761489868164 -I0513 07:06:34.789449 139687763109632 logging_writer.py:48] [98600] global_step=98600, grad_norm=2.9373536109924316, loss=3.8033998012542725 -I0513 07:07:12.153285 139687754716928 logging_writer.py:48] [98700] global_step=98700, grad_norm=3.3407788276672363, loss=3.7084126472473145 -I0513 07:07:53.620360 139687763109632 logging_writer.py:48] [98800] global_step=98800, grad_norm=4.587729454040527, loss=3.7536163330078125 -I0513 07:08:35.351803 139687754716928 logging_writer.py:48] [98900] global_step=98900, grad_norm=3.6661689281463623, loss=3.578075885772705 -I0513 07:09:17.207605 139687763109632 logging_writer.py:48] [99000] global_step=99000, grad_norm=5.084210395812988, loss=3.884312152862549 -I0513 07:09:58.610620 139687754716928 logging_writer.py:48] [99100] global_step=99100, grad_norm=4.985972881317139, loss=3.808479070663452 -I0513 07:10:40.389219 139687763109632 logging_writer.py:48] [99200] global_step=99200, grad_norm=4.305146217346191, loss=3.7884440422058105 -I0513 07:11:22.236365 139687754716928 logging_writer.py:48] [99300] global_step=99300, grad_norm=3.0768301486968994, loss=3.618441104888916 -I0513 07:12:03.611924 139687763109632 logging_writer.py:48] [99400] global_step=99400, grad_norm=3.149554491043091, loss=3.629627227783203 -I0513 07:12:28.506308 139903809070272 spec.py:333] Evaluating on the training split. -I0513 07:12:42.526957 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 07:13:52.086938 139903809070272 spec.py:363] Evaluating on the test split. -I0513 07:13:52.978745 139903809070272 submission_runner.py:516] Time since start: 41577.05s, Step: 99468, {'train/accuracy': Array(0.00163425, dtype=float32), 'train/loss': Array(7.64568, dtype=float32), 'validation/accuracy': Array(0.00204, dtype=float32), 'validation/loss': Array(7.6454577, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0018, dtype=float32), 'test/loss': Array(7.6621127, dtype=float32), 'test/num_examples': 10000, 'score': 39985.52833747864, 'total_duration': 41577.052481651306, 'accumulated_submission_time': 39985.52833747864, 'accumulated_eval_time': 1589.862429857254, 'accumulated_logging_time': 0.773186445236206} -I0513 07:13:53.009495 139687754716928 logging_writer.py:48] [99468] accumulated_eval_time=1589.86, accumulated_logging_time=0.773186, accumulated_submission_time=39985.5, global_step=99468, preemption_count=0, score=39985.5, test/accuracy=0.0018000000854954123, test/loss=7.662112712860107, test/num_examples=10000, total_duration=41577.1, train/accuracy=0.0016342473682016134, train/loss=7.645679950714111, validation/accuracy=0.00203999993391335, validation/loss=7.645457744598389, validation/num_examples=50000 -I0513 07:14:05.333490 139687763109632 logging_writer.py:48] [99500] global_step=99500, grad_norm=5.89680290222168, loss=3.9320316314697266 -I0513 07:14:44.015402 139687754716928 logging_writer.py:48] [99600] global_step=99600, grad_norm=5.127388954162598, loss=3.7925925254821777 -I0513 07:15:22.900619 139687763109632 logging_writer.py:48] [99700] global_step=99700, grad_norm=4.696542263031006, loss=3.870345115661621 -I0513 07:16:01.638728 139687754716928 logging_writer.py:48] [99800] global_step=99800, grad_norm=3.9106686115264893, loss=3.6467907428741455 -I0513 07:16:39.292660 139687763109632 logging_writer.py:48] [99900] global_step=99900, grad_norm=3.7363011837005615, loss=3.755709171295166 -I0513 07:17:16.273618 139687754716928 logging_writer.py:48] [100000] global_step=100000, grad_norm=3.792367935180664, loss=3.6697068214416504 -I0513 07:17:54.011425 139687763109632 logging_writer.py:48] [100100] global_step=100100, grad_norm=5.456989288330078, loss=3.7061567306518555 -I0513 07:18:31.560497 139687754716928 logging_writer.py:48] [100200] global_step=100200, grad_norm=5.057387351989746, loss=3.8636231422424316 -I0513 07:19:08.963673 139687763109632 logging_writer.py:48] [100300] global_step=100300, grad_norm=4.1727190017700195, loss=3.8779098987579346 -I0513 07:19:45.914177 139687754716928 logging_writer.py:48] [100400] global_step=100400, grad_norm=4.745175838470459, loss=3.825899600982666 -I0513 07:20:23.408974 139687763109632 logging_writer.py:48] [100500] global_step=100500, grad_norm=6.56605339050293, loss=3.736830234527588 -I0513 07:21:00.432682 139687754716928 logging_writer.py:48] [100600] global_step=100600, grad_norm=4.786130905151367, loss=3.7488484382629395 -I0513 07:21:37.801390 139687763109632 logging_writer.py:48] [100700] global_step=100700, grad_norm=20.541248321533203, loss=3.7330307960510254 -I0513 07:22:15.159969 139687754716928 logging_writer.py:48] [100800] global_step=100800, grad_norm=5.5016584396362305, loss=3.6011452674865723 -I0513 07:22:52.483414 139687763109632 logging_writer.py:48] [100900] global_step=100900, grad_norm=4.455763339996338, loss=3.711522340774536 -I0513 07:23:29.501981 139687754716928 logging_writer.py:48] [101000] global_step=101000, grad_norm=6.271005630493164, loss=3.858475685119629 -I0513 07:24:06.880737 139687763109632 logging_writer.py:48] [101100] global_step=101100, grad_norm=4.170464038848877, loss=3.6536941528320312 -I0513 07:24:47.923672 139687754716928 logging_writer.py:48] [101200] global_step=101200, grad_norm=3.901453971862793, loss=3.705207109451294 -I0513 07:25:25.103624 139687763109632 logging_writer.py:48] [101300] global_step=101300, grad_norm=6.175779342651367, loss=3.8873682022094727 -I0513 07:26:02.531193 139687754716928 logging_writer.py:48] [101400] global_step=101400, grad_norm=7.37399435043335, loss=3.6051688194274902 -I0513 07:26:39.358865 139687763109632 logging_writer.py:48] [101500] global_step=101500, grad_norm=4.433084964752197, loss=3.7269303798675537 -I0513 07:27:19.961722 139687754716928 logging_writer.py:48] [101600] global_step=101600, grad_norm=4.545472145080566, loss=3.6889545917510986 -I0513 07:28:00.213342 139687763109632 logging_writer.py:48] [101700] global_step=101700, grad_norm=4.99125862121582, loss=3.7542662620544434 -I0513 07:28:37.109558 139687754716928 logging_writer.py:48] [101800] global_step=101800, grad_norm=4.445478916168213, loss=3.653172254562378 -I0513 07:29:14.302682 139687763109632 logging_writer.py:48] [101900] global_step=101900, grad_norm=3.330256223678589, loss=3.8209147453308105 -I0513 07:29:51.660226 139687754716928 logging_writer.py:48] [102000] global_step=102000, grad_norm=33.32331085205078, loss=3.729184150695801 -I0513 07:30:28.567790 139687763109632 logging_writer.py:48] [102100] global_step=102100, grad_norm=6.682192802429199, loss=3.644392967224121 -I0513 07:31:05.685641 139687754716928 logging_writer.py:48] [102200] global_step=102200, grad_norm=3.8319015502929688, loss=3.7488369941711426 -I0513 07:31:47.431105 139687763109632 logging_writer.py:48] [102300] global_step=102300, grad_norm=4.994502067565918, loss=3.776306629180908 -I0513 07:32:33.077534 139687754716928 logging_writer.py:48] [102400] global_step=102400, grad_norm=4.4694600105285645, loss=3.8610172271728516 -I0513 07:33:10.324187 139687763109632 logging_writer.py:48] [102500] global_step=102500, grad_norm=7.183643341064453, loss=3.948726177215576 -I0513 07:33:51.972741 139687754716928 logging_writer.py:48] [102600] global_step=102600, grad_norm=4.961597919464111, loss=3.865356206893921 -I0513 07:34:33.134918 139687763109632 logging_writer.py:48] [102700] global_step=102700, grad_norm=6.295807838439941, loss=3.809837818145752 -I0513 07:35:10.299949 139687754716928 logging_writer.py:48] [102800] global_step=102800, grad_norm=4.743649005889893, loss=3.912973165512085 -I0513 07:35:47.720950 139687763109632 logging_writer.py:48] [102900] global_step=102900, grad_norm=5.166382789611816, loss=3.824899673461914 -I0513 07:36:28.813378 139687754716928 logging_writer.py:48] [103000] global_step=103000, grad_norm=6.142827033996582, loss=3.7141408920288086 -I0513 07:37:14.727876 139687763109632 logging_writer.py:48] [103100] global_step=103100, grad_norm=11.059510231018066, loss=3.820427417755127 -I0513 07:37:53.999861 139687754716928 logging_writer.py:48] [103200] global_step=103200, grad_norm=3.9270360469818115, loss=3.787768840789795 -I0513 07:38:33.723064 139687763109632 logging_writer.py:48] [103300] global_step=103300, grad_norm=3.366436243057251, loss=3.7589800357818604 -I0513 07:39:10.610377 139687754716928 logging_writer.py:48] [103400] global_step=103400, grad_norm=4.323198318481445, loss=3.8141727447509766 -I0513 07:39:48.199999 139687763109632 logging_writer.py:48] [103500] global_step=103500, grad_norm=5.847237586975098, loss=3.837212085723877 -I0513 07:40:25.035674 139687754716928 logging_writer.py:48] [103600] global_step=103600, grad_norm=4.90206241607666, loss=3.80224609375 -I0513 07:41:02.325386 139687763109632 logging_writer.py:48] [103700] global_step=103700, grad_norm=3.7248265743255615, loss=3.8341972827911377 -I0513 07:41:39.691123 139687754716928 logging_writer.py:48] [103800] global_step=103800, grad_norm=6.629977703094482, loss=3.816354274749756 -I0513 07:42:16.457932 139687763109632 logging_writer.py:48] [103900] global_step=103900, grad_norm=4.132185935974121, loss=3.8516862392425537 -I0513 07:42:53.319390 139687754716928 logging_writer.py:48] [104000] global_step=104000, grad_norm=5.598031997680664, loss=3.6790378093719482 -I0513 07:43:31.046778 139687763109632 logging_writer.py:48] [104100] global_step=104100, grad_norm=4.284043312072754, loss=3.908501148223877 -I0513 07:44:07.945015 139687754716928 logging_writer.py:48] [104200] global_step=104200, grad_norm=3.288975477218628, loss=3.7162022590637207 -I0513 07:44:44.847792 139687763109632 logging_writer.py:48] [104300] global_step=104300, grad_norm=5.2575273513793945, loss=3.781266212463379 -I0513 07:45:22.499548 139687754716928 logging_writer.py:48] [104400] global_step=104400, grad_norm=6.402026176452637, loss=3.811129093170166 -I0513 07:45:59.378200 139687763109632 logging_writer.py:48] [104500] global_step=104500, grad_norm=3.1554386615753174, loss=3.6744394302368164 -I0513 07:46:36.308157 139687754716928 logging_writer.py:48] [104600] global_step=104600, grad_norm=3.103302478790283, loss=3.766770839691162 -I0513 07:47:09.484096 139903809070272 spec.py:333] Evaluating on the training split. -I0513 07:47:20.028833 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 07:47:30.259700 139903809070272 spec.py:363] Evaluating on the test split. -I0513 07:47:31.148776 139903809070272 submission_runner.py:516] Time since start: 43595.22s, Step: 104689, {'train/accuracy': Array(0.00161432, dtype=float32), 'train/loss': Array(7.6252575, dtype=float32), 'validation/accuracy': Array(0.00212, dtype=float32), 'validation/loss': Array(7.635452, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0018, dtype=float32), 'test/loss': Array(7.65344, dtype=float32), 'test/num_examples': 10000, 'score': 41981.95154380798, 'total_duration': 43595.22164225578, 'accumulated_submission_time': 41981.95154380798, 'accumulated_eval_time': 1611.5242595672607, 'accumulated_logging_time': 0.8123252391815186} -I0513 07:47:31.189334 139687763109632 logging_writer.py:48] [104689] accumulated_eval_time=1611.52, accumulated_logging_time=0.812325, accumulated_submission_time=41982, global_step=104689, preemption_count=0, score=41982, test/accuracy=0.0018000000854954123, test/loss=7.653439998626709, test/num_examples=10000, total_duration=43595.2, train/accuracy=0.0016143175307661295, train/loss=7.62525749206543, validation/accuracy=0.0021199998445808887, validation/loss=7.635451793670654, validation/num_examples=50000 -I0513 07:47:36.213948 139687754716928 logging_writer.py:48] [104700] global_step=104700, grad_norm=3.2510292530059814, loss=3.7815017700195312 -I0513 07:48:27.876893 139687763109632 logging_writer.py:48] [104800] global_step=104800, grad_norm=3.76088809967041, loss=3.783773183822632 -I0513 07:49:10.735136 139687754716928 logging_writer.py:48] [104900] global_step=104900, grad_norm=5.106197357177734, loss=3.6924948692321777 -I0513 07:49:48.644960 139687763109632 logging_writer.py:48] [105000] global_step=105000, grad_norm=17.579076766967773, loss=3.862417697906494 -I0513 07:50:41.380240 139687754716928 logging_writer.py:48] [105100] global_step=105100, grad_norm=3.4353830814361572, loss=3.737675189971924 -I0513 07:51:22.068758 139687763109632 logging_writer.py:48] [105200] global_step=105200, grad_norm=4.442686080932617, loss=3.777113437652588 -I0513 07:52:00.397878 139687754716928 logging_writer.py:48] [105300] global_step=105300, grad_norm=3.931931734085083, loss=3.817312479019165 -I0513 07:52:37.338463 139687763109632 logging_writer.py:48] [105400] global_step=105400, grad_norm=4.413466930389404, loss=3.8777737617492676 -I0513 07:53:14.265121 139687754716928 logging_writer.py:48] [105500] global_step=105500, grad_norm=5.526045322418213, loss=3.7998979091644287 -I0513 07:53:58.699404 139687763109632 logging_writer.py:48] [105600] global_step=105600, grad_norm=3.6940481662750244, loss=3.7890052795410156 -I0513 07:54:37.969223 139687754716928 logging_writer.py:48] [105700] global_step=105700, grad_norm=3.288703203201294, loss=3.7253997325897217 -I0513 07:55:16.661096 139687763109632 logging_writer.py:48] [105800] global_step=105800, grad_norm=3.8688642978668213, loss=3.750140905380249 -I0513 07:55:56.283862 139687754716928 logging_writer.py:48] [105900] global_step=105900, grad_norm=3.915168523788452, loss=3.850710391998291 -I0513 07:56:35.517480 139687763109632 logging_writer.py:48] [106000] global_step=106000, grad_norm=3.672391653060913, loss=3.869386911392212 -I0513 07:57:14.968445 139687754716928 logging_writer.py:48] [106100] global_step=106100, grad_norm=3.0799262523651123, loss=3.7116520404815674 -I0513 07:57:54.917973 139687763109632 logging_writer.py:48] [106200] global_step=106200, grad_norm=4.5108323097229, loss=3.8725242614746094 -I0513 07:58:34.493152 139687754716928 logging_writer.py:48] [106300] global_step=106300, grad_norm=4.462536811828613, loss=3.6396219730377197 -I0513 07:59:17.404389 139687763109632 logging_writer.py:48] [106400] global_step=106400, grad_norm=7.345053672790527, loss=3.931995391845703 -I0513 08:00:26.897816 139687754716928 logging_writer.py:48] [106500] global_step=106500, grad_norm=15.922429084777832, loss=3.8099992275238037 -I0513 08:01:33.669044 139687763109632 logging_writer.py:48] [106600] global_step=106600, grad_norm=6.613266944885254, loss=3.9996538162231445 -I0513 08:02:22.105711 139687754716928 logging_writer.py:48] [106700] global_step=106700, grad_norm=3.361079216003418, loss=3.7452783584594727 -I0513 08:03:02.122812 139687763109632 logging_writer.py:48] [106800] global_step=106800, grad_norm=5.353756427764893, loss=3.6334986686706543 -I0513 08:03:47.808030 139687754716928 logging_writer.py:48] [106900] global_step=106900, grad_norm=5.467444896697998, loss=3.7954277992248535 -I0513 08:04:28.100423 139687763109632 logging_writer.py:48] [107000] global_step=107000, grad_norm=6.998638153076172, loss=3.913753032684326 -I0513 08:05:13.392621 139687754716928 logging_writer.py:48] [107100] global_step=107100, grad_norm=5.831449031829834, loss=3.622070789337158 -I0513 08:05:55.858018 139687763109632 logging_writer.py:48] [107200] global_step=107200, grad_norm=11.557222366333008, loss=3.991921901702881 -I0513 08:06:36.215467 139687754716928 logging_writer.py:48] [107300] global_step=107300, grad_norm=4.460023880004883, loss=3.8960299491882324 -I0513 08:07:27.348878 139687763109632 logging_writer.py:48] [107400] global_step=107400, grad_norm=4.135594844818115, loss=3.736508846282959 -I0513 08:08:20.555383 139687754716928 logging_writer.py:48] [107500] global_step=107500, grad_norm=5.457708358764648, loss=3.687166213989258 -I0513 08:09:04.234848 139687763109632 logging_writer.py:48] [107600] global_step=107600, grad_norm=3.8390932083129883, loss=3.769501209259033 -I0513 08:10:04.224482 139687754716928 logging_writer.py:48] [107700] global_step=107700, grad_norm=4.4817914962768555, loss=3.7791569232940674 -I0513 08:10:58.759375 139687763109632 logging_writer.py:48] [107800] global_step=107800, grad_norm=4.019161224365234, loss=3.8200526237487793 -I0513 08:11:45.912255 139687754716928 logging_writer.py:48] [107900] global_step=107900, grad_norm=2.7714931964874268, loss=3.7639713287353516 -I0513 08:12:39.824278 139687763109632 logging_writer.py:48] [108000] global_step=108000, grad_norm=4.6806535720825195, loss=3.737016439437866 -I0513 08:13:34.582451 139687754716928 logging_writer.py:48] [108100] global_step=108100, grad_norm=3.037707805633545, loss=3.9234962463378906 -I0513 08:14:14.255735 139687763109632 logging_writer.py:48] [108200] global_step=108200, grad_norm=4.945133209228516, loss=3.876303195953369 -I0513 08:14:53.892762 139687754716928 logging_writer.py:48] [108300] global_step=108300, grad_norm=3.8655920028686523, loss=3.725841522216797 -I0513 08:15:33.494817 139687763109632 logging_writer.py:48] [108400] global_step=108400, grad_norm=7.269028186798096, loss=3.9254531860351562 -I0513 08:16:29.987423 139687754716928 logging_writer.py:48] [108500] global_step=108500, grad_norm=5.321829319000244, loss=3.783844470977783 -I0513 08:17:27.827201 139687763109632 logging_writer.py:48] [108600] global_step=108600, grad_norm=10.783356666564941, loss=3.806241989135742 -I0513 08:18:23.826370 139687754716928 logging_writer.py:48] [108700] global_step=108700, grad_norm=11.696349143981934, loss=3.6098814010620117 -I0513 08:19:21.050325 139687763109632 logging_writer.py:48] [108800] global_step=108800, grad_norm=4.0714592933654785, loss=3.769871234893799 -I0513 08:20:22.008624 139687754716928 logging_writer.py:48] [108900] global_step=108900, grad_norm=4.563718318939209, loss=3.724813461303711 -I0513 08:20:47.770216 139903809070272 spec.py:333] Evaluating on the training split. -I0513 08:21:03.149894 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 08:22:07.264907 139903809070272 spec.py:363] Evaluating on the test split. -I0513 08:22:08.529819 139903809070272 submission_runner.py:516] Time since start: 45672.23s, Step: 108937, {'train/accuracy': Array(0.00175383, dtype=float32), 'train/loss': Array(7.6281743, dtype=float32), 'validation/accuracy': Array(0.00204, dtype=float32), 'validation/loss': Array(7.6303515, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0018, dtype=float32), 'test/loss': Array(7.649656, dtype=float32), 'test/num_examples': 10000, 'score': 43978.46745610237, 'total_duration': 45672.228942632675, 'accumulated_submission_time': 43978.46745610237, 'accumulated_eval_time': 1691.907369852066, 'accumulated_logging_time': 0.8714404106140137} -I0513 08:22:08.995190 139687763109632 logging_writer.py:48] [108937] accumulated_eval_time=1691.91, accumulated_logging_time=0.87144, accumulated_submission_time=43978.5, global_step=108937, preemption_count=0, score=43978.5, test/accuracy=0.0018000000854954123, test/loss=7.649655818939209, test/num_examples=10000, total_duration=45672.2, train/accuracy=0.0017538265092298388, train/loss=7.628174304962158, validation/accuracy=0.00203999993391335, validation/loss=7.630351543426514, validation/num_examples=50000 -I0513 08:22:50.695119 139687754716928 logging_writer.py:48] [109000] global_step=109000, grad_norm=5.376040935516357, loss=3.7603280544281006 -I0513 08:24:01.575752 139687763109632 logging_writer.py:48] [109100] global_step=109100, grad_norm=5.657344818115234, loss=3.820467472076416 -I0513 08:25:07.529922 139687754716928 logging_writer.py:48] [109200] global_step=109200, grad_norm=9.227174758911133, loss=3.875277042388916 -I0513 08:26:13.356856 139687763109632 logging_writer.py:48] [109300] global_step=109300, grad_norm=2.3602027893066406, loss=3.5669336318969727 -I0513 08:27:18.558662 139687754716928 logging_writer.py:48] [109400] global_step=109400, grad_norm=3.8874096870422363, loss=4.048623085021973 -I0513 08:28:25.586556 139687763109632 logging_writer.py:48] [109500] global_step=109500, grad_norm=4.342096328735352, loss=3.9647741317749023 -I0513 08:29:35.909601 139687754716928 logging_writer.py:48] [109600] global_step=109600, grad_norm=4.491919040679932, loss=3.850578784942627 -I0513 08:30:41.564221 139687763109632 logging_writer.py:48] [109700] global_step=109700, grad_norm=4.997833251953125, loss=3.935580253601074 -I0513 08:31:51.464561 139687754716928 logging_writer.py:48] [109800] global_step=109800, grad_norm=3.8086423873901367, loss=3.9143073558807373 -I0513 08:32:57.923812 139687763109632 logging_writer.py:48] [109900] global_step=109900, grad_norm=4.30916166305542, loss=3.672621726989746 -I0513 08:34:02.474002 139687754716928 logging_writer.py:48] [110000] global_step=110000, grad_norm=9.459060668945312, loss=3.7844951152801514 -I0513 08:35:09.750319 139687763109632 logging_writer.py:48] [110100] global_step=110100, grad_norm=4.740483283996582, loss=3.917182445526123 -I0513 08:36:09.118781 139687754716928 logging_writer.py:48] [110200] global_step=110200, grad_norm=9.231329917907715, loss=3.8291678428649902 -I0513 08:37:16.916166 139687763109632 logging_writer.py:48] [110300] global_step=110300, grad_norm=3.6373510360717773, loss=3.842376947402954 -I0513 08:38:26.268148 139687754716928 logging_writer.py:48] [110400] global_step=110400, grad_norm=7.411046504974365, loss=3.811601161956787 -I0513 08:39:29.378890 139687763109632 logging_writer.py:48] [110500] global_step=110500, grad_norm=3.132138252258301, loss=3.6745035648345947 -I0513 08:40:34.519426 139687754716928 logging_writer.py:48] [110600] global_step=110600, grad_norm=5.307441711425781, loss=3.5823419094085693 -I0513 08:41:42.732717 139687763109632 logging_writer.py:48] [110700] global_step=110700, grad_norm=2.6436421871185303, loss=3.6574835777282715 -I0513 08:42:51.549124 139687754716928 logging_writer.py:48] [110800] global_step=110800, grad_norm=19.19878387451172, loss=3.92822003364563 -I0513 08:43:57.470424 139687763109632 logging_writer.py:48] [110900] global_step=110900, grad_norm=3.57155442237854, loss=3.7600321769714355 -I0513 08:45:09.542351 139687754716928 logging_writer.py:48] [111000] global_step=111000, grad_norm=4.153966426849365, loss=3.7033374309539795 -I0513 08:46:13.539672 139687763109632 logging_writer.py:48] [111100] global_step=111100, grad_norm=5.543024063110352, loss=3.871777057647705 -I0513 08:47:24.359289 139687754716928 logging_writer.py:48] [111200] global_step=111200, grad_norm=4.63667631149292, loss=3.652104377746582 -I0513 08:48:32.089760 139687763109632 logging_writer.py:48] [111300] global_step=111300, grad_norm=4.046331405639648, loss=3.7886769771575928 -I0513 08:49:40.373560 139687754716928 logging_writer.py:48] [111400] global_step=111400, grad_norm=4.123543739318848, loss=3.7942094802856445 -I0513 08:50:47.950905 139687763109632 logging_writer.py:48] [111500] global_step=111500, grad_norm=7.0936126708984375, loss=3.691082239151001 -I0513 08:52:00.675891 139687754716928 logging_writer.py:48] [111600] global_step=111600, grad_norm=18.019807815551758, loss=3.9077935218811035 -I0513 08:53:06.163442 139687763109632 logging_writer.py:48] [111700] global_step=111700, grad_norm=3.397520065307617, loss=3.9913716316223145 -I0513 08:54:15.885077 139687754716928 logging_writer.py:48] [111800] global_step=111800, grad_norm=10.30646800994873, loss=3.9151318073272705 -I0513 08:55:25.074404 139903809070272 spec.py:333] Evaluating on the training split. -I0513 08:55:40.450950 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 08:57:28.984284 139903809070272 spec.py:363] Evaluating on the test split. -I0513 08:57:30.047386 139903809070272 submission_runner.py:516] Time since start: 47793.90s, Step: 111895, {'train/accuracy': Array(0.00213249, dtype=float32), 'train/loss': Array(7.6150737, dtype=float32), 'validation/accuracy': Array(0.00194, dtype=float32), 'validation/loss': Array(7.6296186, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0018, dtype=float32), 'test/loss': Array(7.648998, dtype=float32), 'test/num_examples': 10000, 'score': 45974.4246840477, 'total_duration': 47793.9038836956, 'accumulated_submission_time': 45974.4246840477, 'accumulated_eval_time': 1816.6612305641174, 'accumulated_logging_time': 1.4030964374542236} -I0513 08:57:30.353026 139687763109632 logging_writer.py:48] [111895] accumulated_eval_time=1816.66, accumulated_logging_time=1.4031, accumulated_submission_time=45974.4, global_step=111895, preemption_count=0, score=45974.4, test/accuracy=0.0018000000854954123, test/loss=7.648997783660889, test/num_examples=10000, total_duration=47793.9, train/accuracy=0.0021324935369193554, train/loss=7.6150736808776855, validation/accuracy=0.0019399999873712659, validation/loss=7.6296186447143555, validation/num_examples=50000 -I0513 08:57:33.442073 139687754716928 logging_writer.py:48] [111900] global_step=111900, grad_norm=3.9275574684143066, loss=3.8042986392974854 -I0513 08:58:35.556745 139687763109632 logging_writer.py:48] [112000] global_step=112000, grad_norm=4.138191223144531, loss=3.738027572631836 -I0513 08:59:41.530930 139687754716928 logging_writer.py:48] [112100] global_step=112100, grad_norm=4.603808403015137, loss=3.7808923721313477 -I0513 09:00:49.280425 139687763109632 logging_writer.py:48] [112200] global_step=112200, grad_norm=4.335636615753174, loss=3.9187803268432617 -I0513 09:01:52.989485 139687754716928 logging_writer.py:48] [112300] global_step=112300, grad_norm=3.078798770904541, loss=3.6989147663116455 -I0513 09:02:53.326941 139687763109632 logging_writer.py:48] [112400] global_step=112400, grad_norm=3.8041462898254395, loss=3.877378225326538 -I0513 09:03:57.921770 139687754716928 logging_writer.py:48] [112500] global_step=112500, grad_norm=16.200010299682617, loss=3.833498477935791 -2026-05-13 09:04:31.878520: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 09:04:48.947590 139687763109632 logging_writer.py:48] [112600] global_step=112600, grad_norm=8.872913360595703, loss=3.8494200706481934 -I0513 09:05:54.055995 139687754716928 logging_writer.py:48] [112700] global_step=112700, grad_norm=3.0200469493865967, loss=3.7359578609466553 -I0513 09:06:59.856325 139687763109632 logging_writer.py:48] [112800] global_step=112800, grad_norm=7.824386119842529, loss=3.814516067504883 -I0513 09:08:01.404734 139687754716928 logging_writer.py:48] [112900] global_step=112900, grad_norm=7.209688663482666, loss=3.877319574356079 -I0513 09:09:07.191588 139687763109632 logging_writer.py:48] [113000] global_step=113000, grad_norm=6.990146160125732, loss=3.910600185394287 -I0513 09:10:14.907585 139687754716928 logging_writer.py:48] [113100] global_step=113100, grad_norm=11.177308082580566, loss=3.8120627403259277 -I0513 09:11:24.450673 139687763109632 logging_writer.py:48] [113200] global_step=113200, grad_norm=3.3114168643951416, loss=3.947348117828369 -I0513 09:12:32.762606 139687754716928 logging_writer.py:48] [113300] global_step=113300, grad_norm=4.960545539855957, loss=3.7409136295318604 -I0513 09:13:39.607371 139687763109632 logging_writer.py:48] [113400] global_step=113400, grad_norm=4.188526153564453, loss=3.8735268115997314 -I0513 09:14:45.250764 139687754716928 logging_writer.py:48] [113500] global_step=113500, grad_norm=6.301297187805176, loss=3.835031509399414 -I0513 09:15:49.672093 139687763109632 logging_writer.py:48] [113600] global_step=113600, grad_norm=3.3167946338653564, loss=3.7566356658935547 -I0513 09:16:52.211071 139687754716928 logging_writer.py:48] [113700] global_step=113700, grad_norm=2.9355976581573486, loss=3.805861234664917 -I0513 09:17:59.712845 139687763109632 logging_writer.py:48] [113800] global_step=113800, grad_norm=4.48390531539917, loss=3.7768759727478027 -2026-05-13 09:18:18.609669: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 09:19:07.266314 139687754716928 logging_writer.py:48] [113900] global_step=113900, grad_norm=3.7172391414642334, loss=3.6984591484069824 -I0513 09:20:14.254797 139687763109632 logging_writer.py:48] [114000] global_step=114000, grad_norm=11.397680282592773, loss=3.744321823120117 -I0513 09:21:17.434268 139687754716928 logging_writer.py:48] [114100] global_step=114100, grad_norm=4.159839153289795, loss=3.8136022090911865 -I0513 09:22:24.861263 139687763109632 logging_writer.py:48] [114200] global_step=114200, grad_norm=3.752556085586548, loss=3.7094502449035645 -I0513 09:23:27.331802 139687754716928 logging_writer.py:48] [114300] global_step=114300, grad_norm=7.379518508911133, loss=4.011590480804443 -I0513 09:24:34.392681 139687763109632 logging_writer.py:48] [114400] global_step=114400, grad_norm=2.7422313690185547, loss=3.8343300819396973 -I0513 09:25:38.814752 139687754716928 logging_writer.py:48] [114500] global_step=114500, grad_norm=7.15902853012085, loss=3.826964855194092 -I0513 09:26:48.647122 139687763109632 logging_writer.py:48] [114600] global_step=114600, grad_norm=5.699357986450195, loss=3.8478589057922363 -I0513 09:27:53.416008 139687754716928 logging_writer.py:48] [114700] global_step=114700, grad_norm=10.639412879943848, loss=3.8803086280822754 -I0513 09:28:55.686928 139687763109632 logging_writer.py:48] [114800] global_step=114800, grad_norm=5.991600513458252, loss=3.8578884601593018 -I0513 09:29:59.987307 139687754716928 logging_writer.py:48] [114900] global_step=114900, grad_norm=5.52059268951416, loss=3.769235610961914 -I0513 09:30:46.089848 139903809070272 spec.py:333] Evaluating on the training split. -I0513 09:30:59.659030 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 09:31:16.333846 139903809070272 spec.py:363] Evaluating on the test split. -I0513 09:31:17.405534 139903809070272 submission_runner.py:516] Time since start: 49821.30s, Step: 114967, {'train/accuracy': Array(0.0020727, dtype=float32), 'train/loss': Array(7.616824, dtype=float32), 'validation/accuracy': Array(0.00188, dtype=float32), 'validation/loss': Array(7.625837, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0017, dtype=float32), 'test/loss': Array(7.6456165, dtype=float32), 'test/num_examples': 10000, 'score': 47970.05360388756, 'total_duration': 49821.3028011322, 'accumulated_submission_time': 47970.05360388756, 'accumulated_eval_time': 1847.798466205597, 'accumulated_logging_time': 1.7638790607452393} -I0513 09:31:17.782290 139687763109632 logging_writer.py:48] [114967] accumulated_eval_time=1847.8, accumulated_logging_time=1.76388, accumulated_submission_time=47970.1, global_step=114967, preemption_count=0, score=47970.1, test/accuracy=0.0017000001389533281, test/loss=7.64561653137207, test/num_examples=10000, total_duration=49821.3, train/accuracy=0.0020727040246129036, train/loss=7.616824150085449, validation/accuracy=0.001879999996162951, validation/loss=7.6258368492126465, validation/num_examples=50000 -I0513 09:31:35.727245 139687754716928 logging_writer.py:48] [115000] global_step=115000, grad_norm=4.448126316070557, loss=3.842202663421631 -2026-05-13 09:32:26.264721: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 09:32:46.006645 139687763109632 logging_writer.py:48] [115100] global_step=115100, grad_norm=4.40342378616333, loss=3.866725444793701 -I0513 09:33:57.623267 139687754716928 logging_writer.py:48] [115200] global_step=115200, grad_norm=21.89187240600586, loss=3.9015438556671143 -I0513 09:35:05.340777 139687763109632 logging_writer.py:48] [115300] global_step=115300, grad_norm=3.2971532344818115, loss=3.8190054893493652 -I0513 09:36:10.231581 139687754716928 logging_writer.py:48] [115400] global_step=115400, grad_norm=3.460507392883301, loss=3.772664785385132 -I0513 09:37:20.134662 139687763109632 logging_writer.py:48] [115500] global_step=115500, grad_norm=55.11227035522461, loss=3.8955678939819336 -I0513 09:38:26.855165 139687754716928 logging_writer.py:48] [115600] global_step=115600, grad_norm=2.929187297821045, loss=3.711915969848633 -I0513 09:39:32.461666 139687763109632 logging_writer.py:48] [115700] global_step=115700, grad_norm=5.431930065155029, loss=3.9283447265625 -I0513 09:40:44.386435 139687754716928 logging_writer.py:48] [115800] global_step=115800, grad_norm=4.4711198806762695, loss=3.767794132232666 -I0513 09:41:48.274777 139687763109632 logging_writer.py:48] [115900] global_step=115900, grad_norm=4.5517497062683105, loss=3.7989234924316406 -I0513 09:42:53.367541 139687754716928 logging_writer.py:48] [116000] global_step=116000, grad_norm=3.651970863342285, loss=3.878032684326172 -I0513 09:43:59.841168 139687763109632 logging_writer.py:48] [116100] global_step=116100, grad_norm=4.106604099273682, loss=3.7333598136901855 -I0513 09:45:01.887596 139687754716928 logging_writer.py:48] [116200] global_step=116200, grad_norm=5.714157581329346, loss=3.968327045440674 -I0513 09:46:06.630780 139687763109632 logging_writer.py:48] [116300] global_step=116300, grad_norm=2.672926902770996, loss=3.7668402194976807 -2026-05-13 09:46:26.074223: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 09:47:13.844173 139687754716928 logging_writer.py:48] [116400] global_step=116400, grad_norm=3.8861982822418213, loss=3.9910411834716797 -I0513 09:48:18.236461 139687763109632 logging_writer.py:48] [116500] global_step=116500, grad_norm=6.089529037475586, loss=3.698751211166382 -I0513 09:49:20.667387 139687754716928 logging_writer.py:48] [116600] global_step=116600, grad_norm=2.6865110397338867, loss=3.710942268371582 -I0513 09:50:33.168061 139687763109632 logging_writer.py:48] [116700] global_step=116700, grad_norm=4.908960819244385, loss=3.811631441116333 -I0513 09:51:37.315514 139687754716928 logging_writer.py:48] [116800] global_step=116800, grad_norm=4.274679183959961, loss=3.837118148803711 -I0513 09:52:45.502802 139687763109632 logging_writer.py:48] [116900] global_step=116900, grad_norm=2.6250956058502197, loss=3.8440568447113037 -I0513 09:53:54.597345 139687754716928 logging_writer.py:48] [117000] global_step=117000, grad_norm=4.186766147613525, loss=3.678527593612671 -I0513 09:55:01.709641 139687763109632 logging_writer.py:48] [117100] global_step=117100, grad_norm=3.3089892864227295, loss=3.8470113277435303 -I0513 09:56:11.011793 139687754716928 logging_writer.py:48] [117200] global_step=117200, grad_norm=3.795294761657715, loss=3.8503036499023438 -I0513 09:57:20.487131 139687763109632 logging_writer.py:48] [117300] global_step=117300, grad_norm=2.507814407348633, loss=3.740316152572632 -I0513 09:58:27.989596 139687754716928 logging_writer.py:48] [117400] global_step=117400, grad_norm=3.857512950897217, loss=3.8079452514648438 -I0513 09:59:38.544517 139687763109632 logging_writer.py:48] [117500] global_step=117500, grad_norm=5.1463212966918945, loss=3.765735149383545 -2026-05-13 10:00:32.128949: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 10:00:46.031689 139687754716928 logging_writer.py:48] [117600] global_step=117600, grad_norm=14.902926445007324, loss=3.7564468383789062 -I0513 10:01:50.825833 139687763109632 logging_writer.py:48] [117700] global_step=117700, grad_norm=3.0766639709472656, loss=3.717747688293457 -I0513 10:03:00.419538 139687754716928 logging_writer.py:48] [117800] global_step=117800, grad_norm=3.4450430870056152, loss=3.83012056350708 -I0513 10:04:11.070042 139687763109632 logging_writer.py:48] [117900] global_step=117900, grad_norm=5.425838947296143, loss=4.120442867279053 -I0513 10:04:33.471840 139903809070272 spec.py:333] Evaluating on the training split. -I0513 10:04:40.448392 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 10:05:10.277969 139903809070272 spec.py:363] Evaluating on the test split. -I0513 10:05:11.345599 139903809070272 submission_runner.py:516] Time since start: 51855.24s, Step: 117934, {'train/accuracy': Array(0.00185348, dtype=float32), 'train/loss': Array(7.623401, dtype=float32), 'validation/accuracy': Array(0.00182, dtype=float32), 'validation/loss': Array(7.628771, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(7.6490345, dtype=float32), 'test/num_examples': 10000, 'score': 49965.64317059517, 'total_duration': 51855.24300432205, 'accumulated_submission_time': 49965.64317059517, 'accumulated_eval_time': 1885.4940078258514, 'accumulated_logging_time': 2.1830015182495117} -I0513 10:05:11.676729 139687754716928 logging_writer.py:48] [117934] accumulated_eval_time=1885.49, accumulated_logging_time=2.183, accumulated_submission_time=49965.6, global_step=117934, preemption_count=0, score=49965.6, test/accuracy=0.0019000000320374966, test/loss=7.64903450012207, test/num_examples=10000, total_duration=51855.2, train/accuracy=0.0018534756964072585, train/loss=7.623401165008545, validation/accuracy=0.001820000004954636, validation/loss=7.62877082824707, validation/num_examples=50000 -I0513 10:05:52.469070 139687763109632 logging_writer.py:48] [118000] global_step=118000, grad_norm=9.385499000549316, loss=3.8958511352539062 -I0513 10:06:58.135482 139687754716928 logging_writer.py:48] [118100] global_step=118100, grad_norm=2.937913179397583, loss=3.6514029502868652 -I0513 10:08:07.183374 139687763109632 logging_writer.py:48] [118200] global_step=118200, grad_norm=4.623687744140625, loss=3.9199445247650146 -I0513 10:09:15.104660 139687754716928 logging_writer.py:48] [118300] global_step=118300, grad_norm=2.7990782260894775, loss=3.7207889556884766 -I0513 10:10:28.886340 139687763109632 logging_writer.py:48] [118400] global_step=118400, grad_norm=3.9721271991729736, loss=3.649758815765381 -I0513 10:11:41.330292 139687754716928 logging_writer.py:48] [118500] global_step=118500, grad_norm=5.5147385597229, loss=3.776904582977295 -I0513 10:12:47.153513 139687763109632 logging_writer.py:48] [118600] global_step=118600, grad_norm=5.439055919647217, loss=3.842003583908081 -I0513 10:13:58.643160 139687754716928 logging_writer.py:48] [118700] global_step=118700, grad_norm=8.089831352233887, loss=3.962508201599121 -I0513 10:15:12.101438 139687763109632 logging_writer.py:48] [118800] global_step=118800, grad_norm=6.541725158691406, loss=3.878880500793457 -2026-05-13 10:15:30.985342: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 10:16:19.201940 139687754716928 logging_writer.py:48] [118900] global_step=118900, grad_norm=4.258546352386475, loss=3.8112611770629883 -I0513 10:17:28.783435 139687763109632 logging_writer.py:48] [119000] global_step=119000, grad_norm=4.312628269195557, loss=3.858954668045044 -I0513 10:18:35.439197 139687754716928 logging_writer.py:48] [119100] global_step=119100, grad_norm=10.229043006896973, loss=3.9746620655059814 -I0513 10:19:45.576012 139687763109632 logging_writer.py:48] [119200] global_step=119200, grad_norm=11.070023536682129, loss=3.9006264209747314 -I0513 10:20:53.282153 139687754716928 logging_writer.py:48] [119300] global_step=119300, grad_norm=11.379998207092285, loss=3.939807891845703 -I0513 10:22:01.224261 139687763109632 logging_writer.py:48] [119400] global_step=119400, grad_norm=3.5441818237304688, loss=3.953808307647705 -I0513 10:23:09.732521 139687754716928 logging_writer.py:48] [119500] global_step=119500, grad_norm=3.0569496154785156, loss=3.739339590072632 -I0513 10:24:12.250463 139687763109632 logging_writer.py:48] [119600] global_step=119600, grad_norm=5.335159778594971, loss=3.998969316482544 -I0513 10:25:20.964979 139687754716928 logging_writer.py:48] [119700] global_step=119700, grad_norm=7.362301826477051, loss=3.875570297241211 -I0513 10:26:33.916867 139687763109632 logging_writer.py:48] [119800] global_step=119800, grad_norm=6.28887414932251, loss=3.816551446914673 -I0513 10:27:43.808085 139687754716928 logging_writer.py:48] [119900] global_step=119900, grad_norm=4.81928014755249, loss=3.9429931640625 -I0513 10:28:48.157867 139687763109632 logging_writer.py:48] [120000] global_step=120000, grad_norm=3.286750316619873, loss=3.736171245574951 -2026-05-13 10:29:43.537263: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 10:29:56.188822 139687754716928 logging_writer.py:48] [120100] global_step=120100, grad_norm=5.218892574310303, loss=3.807651996612549 -I0513 10:31:05.729163 139687763109632 logging_writer.py:48] [120200] global_step=120200, grad_norm=3.576721668243408, loss=3.817460775375366 -I0513 10:32:08.822005 139687754716928 logging_writer.py:48] [120300] global_step=120300, grad_norm=4.1150054931640625, loss=3.8690505027770996 -I0513 10:33:16.509081 139687763109632 logging_writer.py:48] [120400] global_step=120400, grad_norm=3.899745225906372, loss=3.792226791381836 -I0513 10:34:26.894164 139687754716928 logging_writer.py:48] [120500] global_step=120500, grad_norm=3.3161087036132812, loss=3.854116439819336 -I0513 10:35:34.290475 139687763109632 logging_writer.py:48] [120600] global_step=120600, grad_norm=4.229435920715332, loss=3.9411606788635254 -I0513 10:36:42.189429 139687754716928 logging_writer.py:48] [120700] global_step=120700, grad_norm=3.6027894020080566, loss=3.8828125 -I0513 10:37:47.069415 139687763109632 logging_writer.py:48] [120800] global_step=120800, grad_norm=5.321989059448242, loss=3.9399218559265137 -I0513 10:38:27.459262 139903809070272 spec.py:333] Evaluating on the training split. -I0513 10:38:33.695808 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 10:38:55.231405 139903809070272 spec.py:363] Evaluating on the test split. -I0513 10:38:56.280691 139903809070272 submission_runner.py:516] Time since start: 53880.20s, Step: 120883, {'train/accuracy': Array(0.00131537, dtype=float32), 'train/loss': Array(7.611876, dtype=float32), 'validation/accuracy': Array(0.00178, dtype=float32), 'validation/loss': Array(7.623189, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0021, dtype=float32), 'test/loss': Array(7.644687, dtype=float32), 'test/num_examples': 10000, 'score': 51961.32361125946, 'total_duration': 53880.19545173645, 'accumulated_submission_time': 51961.32361125946, 'accumulated_eval_time': 1914.1544938087463, 'accumulated_logging_time': 2.561258316040039} -I0513 10:38:56.621073 139687754716928 logging_writer.py:48] [120883] accumulated_eval_time=1914.15, accumulated_logging_time=2.56126, accumulated_submission_time=51961.3, global_step=120883, preemption_count=0, score=51961.3, test/accuracy=0.0021000001579523087, test/loss=7.644687175750732, test/num_examples=10000, total_duration=53880.2, train/accuracy=0.0013153698528185487, train/loss=7.611876010894775, validation/accuracy=0.001779999933205545, validation/loss=7.6231889724731445, validation/num_examples=50000 -I0513 10:39:05.619055 139687763109632 logging_writer.py:48] [120900] global_step=120900, grad_norm=3.5195841789245605, loss=3.8635146617889404 -I0513 10:40:14.087387 139687754716928 logging_writer.py:48] [121000] global_step=121000, grad_norm=3.942357301712036, loss=3.850008010864258 -I0513 10:41:23.594339 139687763109632 logging_writer.py:48] [121100] global_step=121100, grad_norm=4.546257495880127, loss=3.9812710285186768 -I0513 10:42:28.880029 139687754716928 logging_writer.py:48] [121200] global_step=121200, grad_norm=3.233799934387207, loss=3.778258800506592 -I0513 10:43:32.484766 139687763109632 logging_writer.py:48] [121300] global_step=121300, grad_norm=4.964593887329102, loss=3.8536155223846436 -2026-05-13 10:43:55.637069: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 10:44:39.391443 139687754716928 logging_writer.py:48] [121400] global_step=121400, grad_norm=2.58447527885437, loss=3.7594428062438965 -I0513 10:45:51.565302 139687763109632 logging_writer.py:48] [121500] global_step=121500, grad_norm=3.575542688369751, loss=3.7236175537109375 -I0513 10:47:01.578415 139687754716928 logging_writer.py:48] [121600] global_step=121600, grad_norm=4.909750461578369, loss=3.8425068855285645 -I0513 10:48:06.015015 139687763109632 logging_writer.py:48] [121700] global_step=121700, grad_norm=5.095527648925781, loss=3.9218850135803223 -I0513 10:49:13.770134 139687754716928 logging_writer.py:48] [121800] global_step=121800, grad_norm=5.524242401123047, loss=3.859065294265747 -I0513 10:50:21.812762 139687763109632 logging_writer.py:48] [121900] global_step=121900, grad_norm=7.943033218383789, loss=3.7824018001556396 -I0513 10:51:33.469705 139687754716928 logging_writer.py:48] [122000] global_step=122000, grad_norm=2.6752731800079346, loss=3.75932240486145 -I0513 10:52:37.636797 139687763109632 logging_writer.py:48] [122100] global_step=122100, grad_norm=3.4225659370422363, loss=3.731212615966797 -I0513 10:53:42.529544 139687754716928 logging_writer.py:48] [122200] global_step=122200, grad_norm=3.2170941829681396, loss=3.835641622543335 -I0513 10:54:45.875796 139687763109632 logging_writer.py:48] [122300] global_step=122300, grad_norm=6.434548854827881, loss=3.9361352920532227 -I0513 10:55:54.143086 139687754716928 logging_writer.py:48] [122400] global_step=122400, grad_norm=4.343682289123535, loss=3.7894227504730225 -I0513 10:57:01.807804 139687763109632 logging_writer.py:48] [122500] global_step=122500, grad_norm=27.08906364440918, loss=3.985386610031128 -2026-05-13 10:58:00.204026: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 10:58:11.746751 139687754716928 logging_writer.py:48] [122600] global_step=122600, grad_norm=5.17721700668335, loss=3.716869831085205 -I0513 10:59:23.210741 139687763109632 logging_writer.py:48] [122700] global_step=122700, grad_norm=11.616267204284668, loss=3.7101919651031494 -I0513 11:00:34.878913 139687754716928 logging_writer.py:48] [122800] global_step=122800, grad_norm=15.874013900756836, loss=3.939215660095215 -I0513 11:01:45.544109 139687763109632 logging_writer.py:48] [122900] global_step=122900, grad_norm=3.7724404335021973, loss=3.7477667331695557 -I0513 11:02:49.557573 139687754716928 logging_writer.py:48] [123000] global_step=123000, grad_norm=5.122750282287598, loss=3.834712505340576 -I0513 11:03:58.493150 139687763109632 logging_writer.py:48] [123100] global_step=123100, grad_norm=2.88747501373291, loss=3.7349963188171387 -I0513 11:05:06.667425 139687754716928 logging_writer.py:48] [123200] global_step=123200, grad_norm=3.8362483978271484, loss=3.846237897872925 -I0513 11:06:09.729883 139687763109632 logging_writer.py:48] [123300] global_step=123300, grad_norm=6.313994884490967, loss=3.8585920333862305 -I0513 11:07:18.850728 139687754716928 logging_writer.py:48] [123400] global_step=123400, grad_norm=3.9272093772888184, loss=3.8692800998687744 -I0513 11:08:28.702476 139687763109632 logging_writer.py:48] [123500] global_step=123500, grad_norm=10.609228134155273, loss=3.916779041290283 -I0513 11:09:37.270421 139687754716928 logging_writer.py:48] [123600] global_step=123600, grad_norm=10.902921676635742, loss=3.950090169906616 -I0513 11:10:41.837012 139687763109632 logging_writer.py:48] [123700] global_step=123700, grad_norm=12.306964874267578, loss=3.8594112396240234 -I0513 11:11:48.684572 139687754716928 logging_writer.py:48] [123800] global_step=123800, grad_norm=4.174154281616211, loss=3.825528144836426 -I0513 11:12:12.244357 139903809070272 spec.py:333] Evaluating on the training split. -2026-05-13 11:12:12.507369: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 11:12:20.242026 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 11:12:46.743680 139903809070272 spec.py:363] Evaluating on the test split. -I0513 11:12:47.822109 139903809070272 submission_runner.py:516] Time since start: 55911.71s, Step: 123836, {'train/accuracy': Array(0.00157446, dtype=float32), 'train/loss': Array(7.5977, dtype=float32), 'validation/accuracy': Array(0.00174, dtype=float32), 'validation/loss': Array(7.615778, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(7.6381392, dtype=float32), 'test/num_examples': 10000, 'score': 53956.820291996, 'total_duration': 55911.70957875252, 'accumulated_submission_time': 53956.820291996, 'accumulated_eval_time': 1949.5440051555634, 'accumulated_logging_time': 2.9746978282928467} -I0513 11:12:48.223452 139687763109632 logging_writer.py:48] [123836] accumulated_eval_time=1949.54, accumulated_logging_time=2.9747, accumulated_submission_time=53956.8, global_step=123836, preemption_count=0, score=53956.8, test/accuracy=0.0019000000320374966, test/loss=7.638139247894287, test/num_examples=10000, total_duration=55911.7, train/accuracy=0.0015744578558951616, train/loss=7.597700119018555, validation/accuracy=0.0017399999778717756, validation/loss=7.615777969360352, validation/num_examples=50000 -I0513 11:13:35.333002 139687754716928 logging_writer.py:48] [123900] global_step=123900, grad_norm=10.455455780029297, loss=3.743910312652588 -I0513 11:14:45.076770 139687763109632 logging_writer.py:48] [124000] global_step=124000, grad_norm=3.051199197769165, loss=3.8120224475860596 -I0513 11:15:52.328147 139687754716928 logging_writer.py:48] [124100] global_step=124100, grad_norm=4.071479797363281, loss=3.8573455810546875 -I0513 11:17:00.010420 139687763109632 logging_writer.py:48] [124200] global_step=124200, grad_norm=4.92400598526001, loss=3.8181962966918945 -I0513 11:18:03.966283 139687754716928 logging_writer.py:48] [124300] global_step=124300, grad_norm=12.579404830932617, loss=3.876626968383789 -I0513 11:19:09.378968 139687763109632 logging_writer.py:48] [124400] global_step=124400, grad_norm=3.1175878047943115, loss=3.7407772541046143 -I0513 11:20:22.718158 139687754716928 logging_writer.py:48] [124500] global_step=124500, grad_norm=9.269796371459961, loss=3.845768690109253 -I0513 11:21:34.849729 139687763109632 logging_writer.py:48] [124600] global_step=124600, grad_norm=3.082162380218506, loss=3.63718318939209 -I0513 11:22:42.429950 139687754716928 logging_writer.py:48] [124700] global_step=124700, grad_norm=21.21193504333496, loss=3.868161201477051 -I0513 11:23:53.632206 139687763109632 logging_writer.py:48] [124800] global_step=124800, grad_norm=5.950034141540527, loss=3.8787364959716797 -I0513 11:25:02.298879 139687754716928 logging_writer.py:48] [124900] global_step=124900, grad_norm=4.496840000152588, loss=3.846446990966797 -I0513 11:26:06.504552 139687763109632 logging_writer.py:48] [125000] global_step=125000, grad_norm=3.7403392791748047, loss=3.9419875144958496 -2026-05-13 11:27:03.625398: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 11:27:13.509331 139687754716928 logging_writer.py:48] [125100] global_step=125100, grad_norm=4.447742938995361, loss=3.7944674491882324 -I0513 11:28:21.271747 139687763109632 logging_writer.py:48] [125200] global_step=125200, grad_norm=3.3972320556640625, loss=3.7243635654449463 -I0513 11:29:29.954883 139687754716928 logging_writer.py:48] [125300] global_step=125300, grad_norm=11.197061538696289, loss=3.8184287548065186 -I0513 11:30:37.070448 139687763109632 logging_writer.py:48] [125400] global_step=125400, grad_norm=2.690697193145752, loss=3.762277841567993 -I0513 11:31:43.025434 139687754716928 logging_writer.py:48] [125500] global_step=125500, grad_norm=3.7688205242156982, loss=3.8464903831481934 -I0513 11:32:53.959339 139687763109632 logging_writer.py:48] [125600] global_step=125600, grad_norm=3.3049323558807373, loss=3.764789342880249 -I0513 11:34:01.902549 139687754716928 logging_writer.py:48] [125700] global_step=125700, grad_norm=4.22227668762207, loss=3.718139171600342 -I0513 11:35:16.489995 139687763109632 logging_writer.py:48] [125800] global_step=125800, grad_norm=4.557240962982178, loss=3.836404323577881 -I0513 11:36:26.145100 139687754716928 logging_writer.py:48] [125900] global_step=125900, grad_norm=4.884913921356201, loss=3.6980175971984863 -I0513 11:37:35.514548 139687763109632 logging_writer.py:48] [126000] global_step=126000, grad_norm=6.722773551940918, loss=3.923069953918457 -I0513 11:38:40.431383 139687754716928 logging_writer.py:48] [126100] global_step=126100, grad_norm=4.241055011749268, loss=3.6775474548339844 -I0513 11:39:46.654480 139687763109632 logging_writer.py:48] [126200] global_step=126200, grad_norm=3.49995493888855, loss=3.750203847885132 -I0513 11:40:57.852031 139687754716928 logging_writer.py:48] [126300] global_step=126300, grad_norm=4.542597770690918, loss=3.7865655422210693 -2026-05-13 11:41:24.990125: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 11:42:07.771849 139687763109632 logging_writer.py:48] [126400] global_step=126400, grad_norm=4.361297607421875, loss=3.967942237854004 -I0513 11:43:16.782180 139687754716928 logging_writer.py:48] [126500] global_step=126500, grad_norm=10.771360397338867, loss=4.007995128631592 -I0513 11:44:24.499958 139687763109632 logging_writer.py:48] [126600] global_step=126600, grad_norm=2.984830379486084, loss=3.773300886154175 -I0513 11:45:31.121047 139687754716928 logging_writer.py:48] [126700] global_step=126700, grad_norm=4.078250885009766, loss=3.8257579803466797 -I0513 11:46:03.868256 139903809070272 spec.py:333] Evaluating on the training split. -I0513 11:46:11.388448 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 11:46:38.492519 139903809070272 spec.py:363] Evaluating on the test split. -I0513 11:46:39.588339 139903809070272 submission_runner.py:516] Time since start: 57943.46s, Step: 126748, {'train/accuracy': Array(0.00219228, dtype=float32), 'train/loss': Array(7.5961895, dtype=float32), 'validation/accuracy': Array(0.00178, dtype=float32), 'validation/loss': Array(7.6027384, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(7.6265955, dtype=float32), 'test/num_examples': 10000, 'score': 55952.328662633896, 'total_duration': 57943.46052527428, 'accumulated_submission_time': 55952.328662633896, 'accumulated_eval_time': 1985.0605597496033, 'accumulated_logging_time': 3.459739923477173} -I0513 11:46:39.950093 139687763109632 logging_writer.py:48] [126748] accumulated_eval_time=1985.06, accumulated_logging_time=3.45974, accumulated_submission_time=55952.3, global_step=126748, preemption_count=0, score=55952.3, test/accuracy=0.0019000000320374966, test/loss=7.626595497131348, test/num_examples=10000, total_duration=57943.5, train/accuracy=0.002192283049225807, train/loss=7.596189498901367, validation/accuracy=0.001779999933205545, validation/loss=7.602738380432129, validation/num_examples=50000 -I0513 11:47:11.491497 139687754716928 logging_writer.py:48] [126800] global_step=126800, grad_norm=4.606532573699951, loss=3.8750064373016357 -I0513 11:48:32.128759 139687763109632 logging_writer.py:48] [126900] global_step=126900, grad_norm=3.9841883182525635, loss=3.8666985034942627 -I0513 11:49:46.740780 139687754716928 logging_writer.py:48] [127000] global_step=127000, grad_norm=4.1958699226379395, loss=3.9528965950012207 -I0513 11:50:51.944499 139687763109632 logging_writer.py:48] [127100] global_step=127100, grad_norm=5.826695919036865, loss=3.8324782848358154 -I0513 11:51:55.886265 139687754716928 logging_writer.py:48] [127200] global_step=127200, grad_norm=17.807783126831055, loss=3.827130079269409 -I0513 11:53:00.730909 139687763109632 logging_writer.py:48] [127300] global_step=127300, grad_norm=4.149832248687744, loss=3.7210633754730225 -I0513 11:54:06.581568 139687754716928 logging_writer.py:48] [127400] global_step=127400, grad_norm=4.049962520599365, loss=3.8029251098632812 -I0513 11:55:15.112465 139687763109632 logging_writer.py:48] [127500] global_step=127500, grad_norm=4.833665370941162, loss=3.7859160900115967 -2026-05-13 11:56:16.519169: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 11:56:24.204265 139687754716928 logging_writer.py:48] [127600] global_step=127600, grad_norm=2.4317705631256104, loss=3.9065184593200684 -I0513 11:57:35.163806 139687763109632 logging_writer.py:48] [127700] global_step=127700, grad_norm=3.3372576236724854, loss=3.861889362335205 -I0513 11:58:44.881264 139687754716928 logging_writer.py:48] [127800] global_step=127800, grad_norm=9.56240177154541, loss=3.962143898010254 -I0513 11:59:49.630620 139687763109632 logging_writer.py:48] [127900] global_step=127900, grad_norm=3.500061511993408, loss=3.910555839538574 -I0513 12:00:55.492291 139687754716928 logging_writer.py:48] [128000] global_step=128000, grad_norm=6.892001628875732, loss=3.924448013305664 -I0513 12:02:03.221343 139687763109632 logging_writer.py:48] [128100] global_step=128100, grad_norm=12.777950286865234, loss=3.8254148960113525 -I0513 12:03:12.248109 139687754716928 logging_writer.py:48] [128200] global_step=128200, grad_norm=19.338077545166016, loss=3.851771831512451 -I0513 12:04:23.214755 139687763109632 logging_writer.py:48] [128300] global_step=128300, grad_norm=5.131955623626709, loss=3.9616007804870605 -I0513 12:05:30.926232 139687754716928 logging_writer.py:48] [128400] global_step=128400, grad_norm=24.7841854095459, loss=3.933617115020752 -I0513 12:06:37.096023 139687763109632 logging_writer.py:48] [128500] global_step=128500, grad_norm=8.217951774597168, loss=3.9373788833618164 -I0513 12:07:47.080520 139687754716928 logging_writer.py:48] [128600] global_step=128600, grad_norm=5.615487575531006, loss=3.9927263259887695 -I0513 12:08:55.604752 139687763109632 logging_writer.py:48] [128700] global_step=128700, grad_norm=25.944843292236328, loss=3.913140296936035 -I0513 12:10:01.417479 139687754716928 logging_writer.py:48] [128800] global_step=128800, grad_norm=6.115100383758545, loss=3.8181145191192627 -2026-05-13 12:10:28.321852: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 12:11:07.392245 139687763109632 logging_writer.py:48] [128900] global_step=128900, grad_norm=3.8512988090515137, loss=3.8204195499420166 -I0513 12:12:16.741657 139687754716928 logging_writer.py:48] [129000] global_step=129000, grad_norm=5.82735538482666, loss=3.9977965354919434 -I0513 12:13:26.362483 139687763109632 logging_writer.py:48] [129100] global_step=129100, grad_norm=4.9329986572265625, loss=3.8429203033447266 -I0513 12:14:36.002873 139687754716928 logging_writer.py:48] [129200] global_step=129200, grad_norm=18.143943786621094, loss=3.8571548461914062 -I0513 12:15:44.725834 139687763109632 logging_writer.py:48] [129300] global_step=129300, grad_norm=5.595682144165039, loss=3.9041507244110107 -I0513 12:16:49.949076 139687754716928 logging_writer.py:48] [129400] global_step=129400, grad_norm=4.007479190826416, loss=3.857821226119995 -I0513 12:18:01.823556 139687763109632 logging_writer.py:48] [129500] global_step=129500, grad_norm=5.656755447387695, loss=3.9158129692077637 -I0513 12:19:12.002124 139687754716928 logging_writer.py:48] [129600] global_step=129600, grad_norm=9.009785652160645, loss=4.060838222503662 -I0513 12:19:57.238784 139903809070272 spec.py:333] Evaluating on the training split. -I0513 12:20:03.825531 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 12:20:24.167303 139903809070272 spec.py:363] Evaluating on the test split. -I0513 12:20:25.289858 139903809070272 submission_runner.py:516] Time since start: 59969.13s, Step: 129664, {'train/accuracy': Array(0.0017339, dtype=float32), 'train/loss': Array(7.580449, dtype=float32), 'validation/accuracy': Array(0.00182, dtype=float32), 'validation/loss': Array(7.594372, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(7.61915, dtype=float32), 'test/num_examples': 10000, 'score': 57949.51488518715, 'total_duration': 59969.133659124374, 'accumulated_submission_time': 57949.51488518715, 'accumulated_eval_time': 2012.879814863205, 'accumulated_logging_time': 3.8696322441101074} -I0513 12:20:25.671332 139687763109632 logging_writer.py:48] [129664] accumulated_eval_time=2012.88, accumulated_logging_time=3.86963, accumulated_submission_time=57949.5, global_step=129664, preemption_count=0, score=57949.5, test/accuracy=0.0019000000320374966, test/loss=7.619150161743164, test/num_examples=10000, total_duration=59969.1, train/accuracy=0.001733896671794355, train/loss=7.580449104309082, validation/accuracy=0.001820000004954636, validation/loss=7.594371795654297, validation/num_examples=50000 -I0513 12:20:50.076023 139687754716928 logging_writer.py:48] [129700] global_step=129700, grad_norm=3.3517041206359863, loss=3.8489534854888916 -I0513 12:22:00.034149 139687763109632 logging_writer.py:48] [129800] global_step=129800, grad_norm=4.2091851234436035, loss=3.9517598152160645 -I0513 12:23:10.080735 139687754716928 logging_writer.py:48] [129900] global_step=129900, grad_norm=3.575305461883545, loss=3.7060327529907227 -I0513 12:24:19.759184 139687763109632 logging_writer.py:48] [130000] global_step=130000, grad_norm=5.282336235046387, loss=3.874969720840454 -2026-05-13 12:25:20.765679: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 12:25:26.615506 139687754716928 logging_writer.py:48] [130100] global_step=130100, grad_norm=37.60280227661133, loss=4.0482258796691895 -I0513 12:26:29.761362 139687763109632 logging_writer.py:48] [130200] global_step=130200, grad_norm=3.634320020675659, loss=3.8021938800811768 -I0513 12:27:37.126093 139687754716928 logging_writer.py:48] [130300] global_step=130300, grad_norm=4.035384654998779, loss=3.9017956256866455 -I0513 12:28:41.472706 139687763109632 logging_writer.py:48] [130400] global_step=130400, grad_norm=6.218775272369385, loss=4.133018493652344 -I0513 12:29:48.789074 139687754716928 logging_writer.py:48] [130500] global_step=130500, grad_norm=8.197781562805176, loss=4.0526227951049805 -I0513 12:30:54.432006 139687763109632 logging_writer.py:48] [130600] global_step=130600, grad_norm=6.804259300231934, loss=3.933678388595581 -I0513 12:32:07.617242 139687754716928 logging_writer.py:48] [130700] global_step=130700, grad_norm=5.4732255935668945, loss=3.832442283630371 -I0513 12:33:17.134093 139687763109632 logging_writer.py:48] [130800] global_step=130800, grad_norm=3.016866445541382, loss=3.6159777641296387 -I0513 12:34:23.139551 139687754716928 logging_writer.py:48] [130900] global_step=130900, grad_norm=5.9715423583984375, loss=3.8426806926727295 -I0513 12:35:29.829675 139687763109632 logging_writer.py:48] [131000] global_step=131000, grad_norm=5.156772613525391, loss=3.8743629455566406 -I0513 12:36:36.653619 139687754716928 logging_writer.py:48] [131100] global_step=131100, grad_norm=2.889143943786621, loss=3.752737522125244 -I0513 12:37:44.225588 139687763109632 logging_writer.py:48] [131200] global_step=131200, grad_norm=5.513582706451416, loss=3.812678098678589 -I0513 12:38:50.365015 139687754716928 logging_writer.py:48] [131300] global_step=131300, grad_norm=7.140570640563965, loss=3.864816427230835 -2026-05-13 12:39:16.612286: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 12:39:56.642698 139687763109632 logging_writer.py:48] [131400] global_step=131400, grad_norm=5.273253440856934, loss=3.916078567504883 -I0513 12:41:06.287019 139687754716928 logging_writer.py:48] [131500] global_step=131500, grad_norm=3.816934823989868, loss=3.7993600368499756 -I0513 12:42:12.124858 139687763109632 logging_writer.py:48] [131600] global_step=131600, grad_norm=4.67458963394165, loss=3.779062271118164 -I0513 12:43:22.070429 139687754716928 logging_writer.py:48] [131700] global_step=131700, grad_norm=3.432312488555908, loss=3.7781577110290527 -I0513 12:44:28.773259 139687763109632 logging_writer.py:48] [131800] global_step=131800, grad_norm=3.6761341094970703, loss=3.837067127227783 -I0513 12:45:36.394600 139687754716928 logging_writer.py:48] [131900] global_step=131900, grad_norm=3.499526023864746, loss=3.744359016418457 -I0513 12:46:47.000185 139687763109632 logging_writer.py:48] [132000] global_step=132000, grad_norm=4.909304618835449, loss=3.942880392074585 -I0513 12:47:55.775829 139687754716928 logging_writer.py:48] [132100] global_step=132100, grad_norm=4.4625115394592285, loss=3.766652822494507 -I0513 12:49:06.263387 139687763109632 logging_writer.py:48] [132200] global_step=132200, grad_norm=8.08125114440918, loss=3.7677574157714844 -I0513 12:50:14.178820 139687754716928 logging_writer.py:48] [132300] global_step=132300, grad_norm=4.330085754394531, loss=3.9088375568389893 -I0513 12:51:19.223485 139687763109632 logging_writer.py:48] [132400] global_step=132400, grad_norm=3.6348166465759277, loss=3.744764804840088 -I0513 12:52:27.109218 139687754716928 logging_writer.py:48] [132500] global_step=132500, grad_norm=4.631178379058838, loss=3.924250364303589 -2026-05-13 12:53:30.342944: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 12:53:35.778607 139687763109632 logging_writer.py:48] [132600] global_step=132600, grad_norm=5.528410911560059, loss=3.917715549468994 -I0513 12:53:42.252264 139903809070272 spec.py:333] Evaluating on the training split. -I0513 12:53:49.488960 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 12:54:23.117773 139903809070272 spec.py:363] Evaluating on the test split. -I0513 12:54:24.156534 139903809070272 submission_runner.py:516] Time since start: 62008.08s, Step: 132610, {'train/accuracy': Array(0.00181362, dtype=float32), 'train/loss': Array(7.5625515, dtype=float32), 'validation/accuracy': Array(0.0018, dtype=float32), 'validation/loss': Array(7.579831, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0018, dtype=float32), 'test/loss': Array(7.605457, dtype=float32), 'test/num_examples': 10000, 'score': 59946.0051176548, 'total_duration': 62008.08043074608, 'accumulated_submission_time': 59946.0051176548, 'accumulated_eval_time': 2054.632348537445, 'accumulated_logging_time': 4.288118839263916} -I0513 12:54:24.480548 139687754716928 logging_writer.py:48] [132610] accumulated_eval_time=2054.63, accumulated_logging_time=4.28812, accumulated_submission_time=59946, global_step=132610, preemption_count=0, score=59946, test/accuracy=0.0018000000854954123, test/loss=7.605456829071045, test/num_examples=10000, total_duration=62008.1, train/accuracy=0.0018136160215362906, train/loss=7.562551498413086, validation/accuracy=0.0017999999690800905, validation/loss=7.579831123352051, validation/num_examples=50000 -I0513 12:55:25.015340 139687763109632 logging_writer.py:48] [132700] global_step=132700, grad_norm=4.634973526000977, loss=3.7707347869873047 -I0513 12:56:40.422108 139687754716928 logging_writer.py:48] [132800] global_step=132800, grad_norm=4.797515392303467, loss=3.7419161796569824 -I0513 12:57:54.855727 139687763109632 logging_writer.py:48] [132900] global_step=132900, grad_norm=4.546232223510742, loss=3.9256958961486816 -I0513 12:59:06.126979 139687754716928 logging_writer.py:48] [133000] global_step=133000, grad_norm=4.129562854766846, loss=3.981171131134033 -I0513 13:00:10.629151 139687763109632 logging_writer.py:48] [133100] global_step=133100, grad_norm=27.676130294799805, loss=3.9268150329589844 -I0513 13:01:20.969399 139687754716928 logging_writer.py:48] [133200] global_step=133200, grad_norm=6.153794765472412, loss=3.9577839374542236 -I0513 13:02:27.485227 139687763109632 logging_writer.py:48] [133300] global_step=133300, grad_norm=3.319084405899048, loss=3.8944592475891113 -I0513 13:03:34.521988 139687754716928 logging_writer.py:48] [133400] global_step=133400, grad_norm=6.172501564025879, loss=3.933821201324463 -I0513 13:04:40.475426 139687763109632 logging_writer.py:48] [133500] global_step=133500, grad_norm=6.1541337966918945, loss=3.99470591545105 -I0513 13:05:49.549925 139687754716928 logging_writer.py:48] [133600] global_step=133600, grad_norm=11.731232643127441, loss=3.846670150756836 -I0513 13:07:06.019594 139687763109632 logging_writer.py:48] [133700] global_step=133700, grad_norm=6.1500563621521, loss=3.9219419956207275 -I0513 13:08:14.313781 139687754716928 logging_writer.py:48] [133800] global_step=133800, grad_norm=15.117558479309082, loss=3.839243173599243 -2026-05-13 13:08:43.809210: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 13:09:27.052400 139687763109632 logging_writer.py:48] [133900] global_step=133900, grad_norm=6.694962024688721, loss=3.846773147583008 -I0513 13:10:36.665037 139687754716928 logging_writer.py:48] [134000] global_step=134000, grad_norm=7.819849014282227, loss=4.093023300170898 -I0513 13:11:45.996554 139687763109632 logging_writer.py:48] [134100] global_step=134100, grad_norm=4.031583786010742, loss=3.91951584815979 -I0513 13:12:50.775900 139687754716928 logging_writer.py:48] [134200] global_step=134200, grad_norm=13.05477523803711, loss=3.978976249694824 -I0513 13:13:57.623489 139687763109632 logging_writer.py:48] [134300] global_step=134300, grad_norm=6.653199195861816, loss=3.9441490173339844 -I0513 13:15:07.574597 139687754716928 logging_writer.py:48] [134400] global_step=134400, grad_norm=6.414670944213867, loss=3.8172335624694824 -I0513 13:16:23.401602 139687763109632 logging_writer.py:48] [134500] global_step=134500, grad_norm=7.899640083312988, loss=3.905886650085449 -I0513 13:17:33.613043 139687754716928 logging_writer.py:48] [134600] global_step=134600, grad_norm=5.46203088760376, loss=3.898900270462036 -I0513 13:18:44.534525 139687763109632 logging_writer.py:48] [134700] global_step=134700, grad_norm=3.0623080730438232, loss=3.7560372352600098 -I0513 13:19:51.361374 139687754716928 logging_writer.py:48] [134800] global_step=134800, grad_norm=4.73820161819458, loss=3.814734935760498 -I0513 13:21:02.024520 139687763109632 logging_writer.py:48] [134900] global_step=134900, grad_norm=4.217526435852051, loss=3.7313404083251953 -I0513 13:22:08.071847 139687754716928 logging_writer.py:48] [135000] global_step=135000, grad_norm=3.93574857711792, loss=3.8367722034454346 -2026-05-13 13:23:19.536120: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 13:23:21.746493 139687763109632 logging_writer.py:48] [135100] global_step=135100, grad_norm=4.870224475860596, loss=3.7801713943481445 -I0513 13:24:37.535194 139687754716928 logging_writer.py:48] [135200] global_step=135200, grad_norm=4.392237186431885, loss=3.7925894260406494 -I0513 13:25:50.333842 139687763109632 logging_writer.py:48] [135300] global_step=135300, grad_norm=3.627432346343994, loss=3.7758002281188965 -I0513 13:27:05.646705 139687754716928 logging_writer.py:48] [135400] global_step=135400, grad_norm=11.15189266204834, loss=3.898169755935669 -I0513 13:27:42.236047 139903809070272 spec.py:333] Evaluating on the training split. -I0513 13:27:49.978787 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 13:28:23.086386 139903809070272 spec.py:363] Evaluating on the test split. -I0513 13:28:24.404938 139903809070272 submission_runner.py:516] Time since start: 64048.06s, Step: 135448, {'train/accuracy': Array(0.00177376, dtype=float32), 'train/loss': Array(7.552586, dtype=float32), 'validation/accuracy': Array(0.00204, dtype=float32), 'validation/loss': Array(7.5581985, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(7.5847716, dtype=float32), 'test/num_examples': 10000, 'score': 61943.662088871, 'total_duration': 64048.056831121445, 'accumulated_submission_time': 61943.662088871, 'accumulated_eval_time': 2096.377536058426, 'accumulated_logging_time': 4.6587910652160645} -I0513 13:28:24.830971 139687763109632 logging_writer.py:48] [135448] accumulated_eval_time=2096.38, accumulated_logging_time=4.65879, accumulated_submission_time=61943.7, global_step=135448, preemption_count=0, score=61943.7, test/accuracy=0.0019000000320374966, test/loss=7.584771633148193, test/num_examples=10000, total_duration=64048.1, train/accuracy=0.0017737563466653228, train/loss=7.552586078643799, validation/accuracy=0.00203999993391335, validation/loss=7.55819845199585, validation/num_examples=50000 -I0513 13:29:00.474167 139687754716928 logging_writer.py:48] [135500] global_step=135500, grad_norm=2.251721143722534, loss=3.7219882011413574 -I0513 13:30:16.216642 139687763109632 logging_writer.py:48] [135600] global_step=135600, grad_norm=3.530466318130493, loss=3.9474120140075684 -I0513 13:31:29.799575 139687754716928 logging_writer.py:48] [135700] global_step=135700, grad_norm=3.649196147918701, loss=3.847079277038574 -I0513 13:32:41.305064 139687763109632 logging_writer.py:48] [135800] global_step=135800, grad_norm=3.504812002182007, loss=3.970182418823242 -I0513 13:33:51.401178 139687754716928 logging_writer.py:48] [135900] global_step=135900, grad_norm=5.99545431137085, loss=3.998887300491333 -I0513 13:34:59.626659 139687763109632 logging_writer.py:48] [136000] global_step=136000, grad_norm=25.90404510498047, loss=3.860492706298828 -I0513 13:36:09.229794 139687754716928 logging_writer.py:48] [136100] global_step=136100, grad_norm=8.213357925415039, loss=4.018126487731934 -I0513 13:37:17.422393 139687763109632 logging_writer.py:48] [136200] global_step=136200, grad_norm=4.917113304138184, loss=3.8106865882873535 -I0513 13:38:23.026824 139687754716928 logging_writer.py:48] [136300] global_step=136300, grad_norm=6.219893932342529, loss=3.931882381439209 -2026-05-13 13:38:56.594962: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 13:39:30.824506 139687763109632 logging_writer.py:48] [136400] global_step=136400, grad_norm=7.151308059692383, loss=3.9189653396606445 -I0513 13:40:41.923156 139687754716928 logging_writer.py:48] [136500] global_step=136500, grad_norm=3.9935061931610107, loss=3.852660655975342 -I0513 13:41:47.449702 139687763109632 logging_writer.py:48] [136600] global_step=136600, grad_norm=4.939479827880859, loss=3.8053665161132812 -I0513 13:42:59.728563 139687754716928 logging_writer.py:48] [136700] global_step=136700, grad_norm=3.2404212951660156, loss=3.8049747943878174 -I0513 13:44:06.825987 139687763109632 logging_writer.py:48] [136800] global_step=136800, grad_norm=4.219491004943848, loss=3.717860221862793 -I0513 13:45:15.723404 139687754716928 logging_writer.py:48] [136900] global_step=136900, grad_norm=5.423902988433838, loss=3.8966310024261475 -I0513 13:46:22.623265 139687763109632 logging_writer.py:48] [137000] global_step=137000, grad_norm=6.568713665008545, loss=3.7315526008605957 -I0513 13:47:34.273607 139687754716928 logging_writer.py:48] [137100] global_step=137100, grad_norm=4.994969844818115, loss=3.9801855087280273 -I0513 13:48:37.741432 139687763109632 logging_writer.py:48] [137200] global_step=137200, grad_norm=5.715720176696777, loss=3.822418689727783 -I0513 13:49:44.648984 139687754716928 logging_writer.py:48] [137300] global_step=137300, grad_norm=2.990103244781494, loss=3.8554129600524902 -I0513 13:50:50.557484 139687763109632 logging_writer.py:48] [137400] global_step=137400, grad_norm=3.6423773765563965, loss=3.8682897090911865 -I0513 13:51:54.547658 139687754716928 logging_writer.py:48] [137500] global_step=137500, grad_norm=8.655993461608887, loss=4.03466796875 -2026-05-13 13:52:58.536060: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 13:52:58.854356 139687763109632 logging_writer.py:48] [137600] global_step=137600, grad_norm=8.74659252166748, loss=4.1233954429626465 -I0513 13:54:07.326411 139687754716928 logging_writer.py:48] [137700] global_step=137700, grad_norm=4.8780412673950195, loss=3.8968520164489746 -I0513 13:55:09.769671 139687763109632 logging_writer.py:48] [137800] global_step=137800, grad_norm=3.712815761566162, loss=3.8314337730407715 -I0513 13:56:20.651696 139687754716928 logging_writer.py:48] [137900] global_step=137900, grad_norm=4.84600305557251, loss=3.9532673358917236 -I0513 13:57:28.406516 139687763109632 logging_writer.py:48] [138000] global_step=138000, grad_norm=4.772759437561035, loss=3.8014814853668213 -I0513 13:58:36.873993 139687754716928 logging_writer.py:48] [138100] global_step=138100, grad_norm=6.478760242462158, loss=3.8645901679992676 -I0513 13:59:48.791750 139687763109632 logging_writer.py:48] [138200] global_step=138200, grad_norm=4.562845230102539, loss=3.8523616790771484 -I0513 14:01:04.418574 139687754716928 logging_writer.py:48] [138300] global_step=138300, grad_norm=3.5587213039398193, loss=3.7978322505950928 -I0513 14:01:40.115527 139903809070272 spec.py:333] Evaluating on the training split. -I0513 14:01:47.690632 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 14:02:06.185770 139903809070272 spec.py:363] Evaluating on the test split. -I0513 14:02:07.246491 139903809070272 submission_runner.py:516] Time since start: 66071.16s, Step: 138357, {'train/accuracy': Array(0.00213249, dtype=float32), 'train/loss': Array(7.5125346, dtype=float32), 'validation/accuracy': Array(0.00206, dtype=float32), 'validation/loss': Array(7.53964, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.002, dtype=float32), 'test/loss': Array(7.5671916, dtype=float32), 'test/num_examples': 10000, 'score': 63938.85355734825, 'total_duration': 66071.15640091896, 'accumulated_submission_time': 63938.85355734825, 'accumulated_eval_time': 2123.342755794525, 'accumulated_logging_time': 5.124191522598267} -I0513 14:02:07.628733 139687763109632 logging_writer.py:48] [138357] accumulated_eval_time=2123.34, accumulated_logging_time=5.12419, accumulated_submission_time=63938.9, global_step=138357, preemption_count=0, score=63938.9, test/accuracy=0.0020000000949949026, test/loss=7.5671916007995605, test/num_examples=10000, total_duration=66071.2, train/accuracy=0.0021324935369193554, train/loss=7.5125346183776855, validation/accuracy=0.002059999853372574, validation/loss=7.539639949798584, validation/num_examples=50000 -I0513 14:02:36.304134 139687754716928 logging_writer.py:48] [138400] global_step=138400, grad_norm=5.556369781494141, loss=3.775578737258911 -I0513 14:03:48.777481 139687763109632 logging_writer.py:48] [138500] global_step=138500, grad_norm=3.3718373775482178, loss=3.8791933059692383 -I0513 14:04:59.836325 139687754716928 logging_writer.py:48] [138600] global_step=138600, grad_norm=22.835023880004883, loss=3.897369623184204 -I0513 14:06:12.282043 139687763109632 logging_writer.py:48] [138700] global_step=138700, grad_norm=3.365523338317871, loss=3.8858745098114014 -I0513 14:07:20.070285 139687754716928 logging_writer.py:48] [138800] global_step=138800, grad_norm=7.226608753204346, loss=3.9636268615722656 -2026-05-13 14:07:54.509423: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 14:08:30.197064 139687763109632 logging_writer.py:48] [138900] global_step=138900, grad_norm=20.09050178527832, loss=3.910851001739502 -I0513 14:09:36.069742 139687754716928 logging_writer.py:48] [139000] global_step=139000, grad_norm=6.673020362854004, loss=4.03193473815918 -I0513 14:10:46.203717 139687763109632 logging_writer.py:48] [139100] global_step=139100, grad_norm=5.4175214767456055, loss=3.9209084510803223 -I0513 14:11:53.443101 139687754716928 logging_writer.py:48] [139200] global_step=139200, grad_norm=5.785498142242432, loss=3.899226665496826 -I0513 14:13:01.076937 139687763109632 logging_writer.py:48] [139300] global_step=139300, grad_norm=10.16620922088623, loss=3.8999686241149902 -I0513 14:14:04.891974 139687754716928 logging_writer.py:48] [139400] global_step=139400, grad_norm=9.924324035644531, loss=4.029358863830566 -I0513 14:15:11.490022 139687763109632 logging_writer.py:48] [139500] global_step=139500, grad_norm=5.184852123260498, loss=3.724806308746338 -I0513 14:16:14.787827 139687754716928 logging_writer.py:48] [139600] global_step=139600, grad_norm=4.376903057098389, loss=3.7715139389038086 -I0513 14:17:20.549621 139687763109632 logging_writer.py:48] [139700] global_step=139700, grad_norm=5.826095104217529, loss=3.7473230361938477 -I0513 14:18:31.930478 139687754716928 logging_writer.py:48] [139800] global_step=139800, grad_norm=10.620084762573242, loss=4.004089832305908 -I0513 14:19:39.527781 139687763109632 logging_writer.py:48] [139900] global_step=139900, grad_norm=6.008485794067383, loss=3.7739477157592773 -I0513 14:20:47.184607 139687754716928 logging_writer.py:48] [140000] global_step=140000, grad_norm=5.750551223754883, loss=3.8834943771362305 -2026-05-13 14:21:59.195976: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 14:22:00.072216 139687763109632 logging_writer.py:48] [140100] global_step=140100, grad_norm=11.466198921203613, loss=3.8390493392944336 -I0513 14:23:07.221661 139687754716928 logging_writer.py:48] [140200] global_step=140200, grad_norm=4.500518321990967, loss=3.9482409954071045 -I0513 14:24:22.351985 139687763109632 logging_writer.py:48] [140300] global_step=140300, grad_norm=16.083974838256836, loss=3.9149060249328613 -I0513 14:25:32.233165 139687754716928 logging_writer.py:48] [140400] global_step=140400, grad_norm=3.3167359828948975, loss=3.8398661613464355 -I0513 14:26:38.604107 139687763109632 logging_writer.py:48] [140500] global_step=140500, grad_norm=16.605504989624023, loss=3.6711275577545166 -I0513 14:27:52.041372 139687754716928 logging_writer.py:48] [140600] global_step=140600, grad_norm=30.90733528137207, loss=3.966285228729248 -I0513 14:28:59.852411 139687763109632 logging_writer.py:48] [140700] global_step=140700, grad_norm=4.094808578491211, loss=3.7786147594451904 -I0513 14:30:03.894812 139687754716928 logging_writer.py:48] [140800] global_step=140800, grad_norm=10.223880767822266, loss=4.088576316833496 -I0513 14:31:09.884247 139687763109632 logging_writer.py:48] [140900] global_step=140900, grad_norm=5.088144779205322, loss=3.8142409324645996 -I0513 14:32:18.768498 139687754716928 logging_writer.py:48] [141000] global_step=141000, grad_norm=6.465155601501465, loss=4.042456150054932 -I0513 14:33:22.287211 139687763109632 logging_writer.py:48] [141100] global_step=141100, grad_norm=6.886141777038574, loss=3.7587594985961914 -I0513 14:34:30.098670 139687754716928 logging_writer.py:48] [141200] global_step=141200, grad_norm=51.13761901855469, loss=4.2609100341796875 -I0513 14:35:23.187962 139903809070272 spec.py:333] Evaluating on the training split. -I0513 14:35:30.927047 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 14:35:55.437857 139903809070272 spec.py:363] Evaluating on the test split. -I0513 14:35:56.495940 139903809070272 submission_runner.py:516] Time since start: 68100.41s, Step: 141279, {'train/accuracy': Array(0.00197305, dtype=float32), 'train/loss': Array(7.5101995, dtype=float32), 'validation/accuracy': Array(0.00222, dtype=float32), 'validation/loss': Array(7.5270987, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(7.5543923, dtype=float32), 'test/num_examples': 10000, 'score': 65934.31674814224, 'total_duration': 68100.40691971779, 'accumulated_submission_time': 65934.31674814224, 'accumulated_eval_time': 2156.486003637314, 'accumulated_logging_time': 5.550399541854858} -I0513 14:35:56.951151 139687763109632 logging_writer.py:48] [141279] accumulated_eval_time=2156.49, accumulated_logging_time=5.5504, accumulated_submission_time=65934.3, global_step=141279, preemption_count=0, score=65934.3, test/accuracy=0.0019000000320374966, test/loss=7.554392337799072, test/num_examples=10000, total_duration=68100.4, train/accuracy=0.001973054837435484, train/loss=7.510199546813965, validation/accuracy=0.002219999907538295, validation/loss=7.527098655700684, validation/num_examples=50000 -I0513 14:36:08.339671 139687754716928 logging_writer.py:48] [141300] global_step=141300, grad_norm=4.9166951179504395, loss=3.9202539920806885 -2026-05-13 14:36:45.813061: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 14:37:19.999255 139687763109632 logging_writer.py:48] [141400] global_step=141400, grad_norm=5.344111442565918, loss=4.035045146942139 -I0513 14:38:31.659342 139687754716928 logging_writer.py:48] [141500] global_step=141500, grad_norm=6.056302070617676, loss=3.6611530780792236 -I0513 14:39:41.972214 139687763109632 logging_writer.py:48] [141600] global_step=141600, grad_norm=11.051179885864258, loss=3.96002459526062 -I0513 14:40:47.446480 139687754716928 logging_writer.py:48] [141700] global_step=141700, grad_norm=5.760129451751709, loss=3.97357439994812 -I0513 14:41:53.431101 139687763109632 logging_writer.py:48] [141800] global_step=141800, grad_norm=3.679607391357422, loss=3.876610279083252 -I0513 14:43:02.272421 139687754716928 logging_writer.py:48] [141900] global_step=141900, grad_norm=5.914991855621338, loss=4.043649673461914 -I0513 14:44:11.216372 139687763109632 logging_writer.py:48] [142000] global_step=142000, grad_norm=3.86784029006958, loss=3.9334168434143066 -I0513 14:45:19.194869 139687754716928 logging_writer.py:48] [142100] global_step=142100, grad_norm=4.647952556610107, loss=3.8349125385284424 -I0513 14:46:26.996871 139687763109632 logging_writer.py:48] [142200] global_step=142200, grad_norm=5.142815113067627, loss=4.038658618927002 -I0513 14:47:30.710227 139687754716928 logging_writer.py:48] [142300] global_step=142300, grad_norm=9.013766288757324, loss=3.997434139251709 -I0513 14:48:34.549887 139687763109632 logging_writer.py:48] [142400] global_step=142400, grad_norm=3.598330497741699, loss=3.7415642738342285 -I0513 14:49:39.852652 139687754716928 logging_writer.py:48] [142500] global_step=142500, grad_norm=45.369972229003906, loss=3.8918814659118652 -I0513 14:50:45.467102 139687763109632 logging_writer.py:48] [142600] global_step=142600, grad_norm=4.870360374450684, loss=3.8174307346343994 -2026-05-13 14:50:46.617498: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 14:51:53.446866 139687754716928 logging_writer.py:48] [142700] global_step=142700, grad_norm=3.637164354324341, loss=3.8061270713806152 -I0513 14:53:03.258074 139687763109632 logging_writer.py:48] [142800] global_step=142800, grad_norm=4.315118789672852, loss=3.953026294708252 -I0513 14:54:17.934706 139687754716928 logging_writer.py:48] [142900] global_step=142900, grad_norm=8.890266418457031, loss=3.971794843673706 -I0513 14:55:28.258156 139687763109632 logging_writer.py:48] [143000] global_step=143000, grad_norm=10.819942474365234, loss=3.92189359664917 -I0513 14:56:38.372321 139687754716928 logging_writer.py:48] [143100] global_step=143100, grad_norm=4.489664554595947, loss=3.802966594696045 -I0513 14:57:45.404612 139687763109632 logging_writer.py:48] [143200] global_step=143200, grad_norm=3.0291764736175537, loss=3.7971749305725098 -I0513 14:58:50.354196 139687754716928 logging_writer.py:48] [143300] global_step=143300, grad_norm=3.111781358718872, loss=3.7122299671173096 -I0513 14:59:58.366989 139687763109632 logging_writer.py:48] [143400] global_step=143400, grad_norm=4.758915424346924, loss=3.863041877746582 -I0513 15:01:08.638515 139687754716928 logging_writer.py:48] [143500] global_step=143500, grad_norm=7.975854396820068, loss=3.862150192260742 -I0513 15:02:16.787504 139687763109632 logging_writer.py:48] [143600] global_step=143600, grad_norm=4.1292924880981445, loss=3.8461058139801025 -I0513 15:03:29.955826 139687754716928 logging_writer.py:48] [143700] global_step=143700, grad_norm=5.006647109985352, loss=3.84997296333313 -I0513 15:04:40.678902 139687763109632 logging_writer.py:48] [143800] global_step=143800, grad_norm=7.449295997619629, loss=3.8256430625915527 -2026-05-13 15:05:17.112957: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 15:05:48.471089 139687754716928 logging_writer.py:48] [143900] global_step=143900, grad_norm=10.432781219482422, loss=3.914327621459961 -I0513 15:07:03.055650 139687763109632 logging_writer.py:48] [144000] global_step=144000, grad_norm=4.502899169921875, loss=3.8878417015075684 -I0513 15:08:09.855858 139687754716928 logging_writer.py:48] [144100] global_step=144100, grad_norm=3.0637621879577637, loss=3.9866912364959717 -I0513 15:09:12.565316 139903809070272 spec.py:333] Evaluating on the training split. -I0513 15:09:20.218045 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 15:09:51.146143 139903809070272 spec.py:363] Evaluating on the test split. -I0513 15:09:52.183347 139903809070272 submission_runner.py:516] Time since start: 70136.12s, Step: 144193, {'train/accuracy': Array(0.00235172, dtype=float32), 'train/loss': Array(7.513111, dtype=float32), 'validation/accuracy': Array(0.00256, dtype=float32), 'validation/loss': Array(7.514347, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(7.5415044, dtype=float32), 'test/num_examples': 10000, 'score': 67929.82773637772, 'total_duration': 70136.11622142792, 'accumulated_submission_time': 67929.82773637772, 'accumulated_eval_time': 2195.9613015651703, 'accumulated_logging_time': 6.055431127548218} -I0513 15:09:52.560454 139687763109632 logging_writer.py:48] [144193] accumulated_eval_time=2195.96, accumulated_logging_time=6.05543, accumulated_submission_time=67929.8, global_step=144193, preemption_count=0, score=67929.8, test/accuracy=0.0019000000320374966, test/loss=7.541504383087158, test/num_examples=10000, total_duration=70136.1, train/accuracy=0.0023517219815403223, train/loss=7.513111114501953, validation/accuracy=0.0025599999353289604, validation/loss=7.514347076416016, validation/num_examples=50000 -I0513 15:09:55.725873 139687754716928 logging_writer.py:48] [144200] global_step=144200, grad_norm=4.148107051849365, loss=3.889343738555908 -I0513 15:11:04.988821 139687763109632 logging_writer.py:48] [144300] global_step=144300, grad_norm=26.39380645751953, loss=3.9268622398376465 -I0513 15:12:12.856963 139687754716928 logging_writer.py:48] [144400] global_step=144400, grad_norm=3.846163034439087, loss=3.825794219970703 -I0513 15:13:18.892936 139687763109632 logging_writer.py:48] [144500] global_step=144500, grad_norm=7.60176420211792, loss=3.8658041954040527 -I0513 15:14:29.005422 139687754716928 logging_writer.py:48] [144600] global_step=144600, grad_norm=8.486637115478516, loss=3.8613698482513428 -I0513 15:15:37.530104 139687763109632 logging_writer.py:48] [144700] global_step=144700, grad_norm=4.230263710021973, loss=3.8233046531677246 -I0513 15:16:47.610922 139687754716928 logging_writer.py:48] [144800] global_step=144800, grad_norm=2.453479051589966, loss=3.9084200859069824 -I0513 15:17:57.445372 139687763109632 logging_writer.py:48] [144900] global_step=144900, grad_norm=4.661877155303955, loss=3.966707468032837 -I0513 15:19:04.209729 139687754716928 logging_writer.py:48] [145000] global_step=145000, grad_norm=3.5486996173858643, loss=3.8808932304382324 -I0513 15:20:10.793933 139687763109632 logging_writer.py:48] [145100] global_step=145100, grad_norm=4.595849990844727, loss=3.997016429901123 -2026-05-13 15:20:13.485565: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 15:21:15.550291 139687754716928 logging_writer.py:48] [145200] global_step=145200, grad_norm=4.136438846588135, loss=3.801668167114258 -I0513 15:22:25.120341 139687763109632 logging_writer.py:48] [145300] global_step=145300, grad_norm=6.004209995269775, loss=3.840820074081421 -I0513 15:23:35.815160 139687754716928 logging_writer.py:48] [145400] global_step=145400, grad_norm=4.6914286613464355, loss=3.857877731323242 -I0513 15:24:47.972366 139687763109632 logging_writer.py:48] [145500] global_step=145500, grad_norm=8.417786598205566, loss=3.839351177215576 -I0513 15:25:56.816296 139687754716928 logging_writer.py:48] [145600] global_step=145600, grad_norm=5.604111671447754, loss=3.9657788276672363 -I0513 15:27:02.821231 139687763109632 logging_writer.py:48] [145700] global_step=145700, grad_norm=25.089435577392578, loss=4.028815269470215 -I0513 15:28:19.862309 139687754716928 logging_writer.py:48] [145800] global_step=145800, grad_norm=6.852769374847412, loss=4.037015914916992 -I0513 15:29:29.665288 139687763109632 logging_writer.py:48] [145900] global_step=145900, grad_norm=4.516221523284912, loss=4.027734756469727 -I0513 15:30:40.298885 139687754716928 logging_writer.py:48] [146000] global_step=146000, grad_norm=10.67943000793457, loss=3.9644775390625 -I0513 15:31:50.364467 139687763109632 logging_writer.py:48] [146100] global_step=146100, grad_norm=3.4641830921173096, loss=3.8558645248413086 -I0513 15:33:03.463014 139687754716928 logging_writer.py:48] [146200] global_step=146200, grad_norm=5.498815536499023, loss=3.886444091796875 -I0513 15:34:09.358096 139687763109632 logging_writer.py:48] [146300] global_step=146300, grad_norm=3.670952558517456, loss=3.852908134460449 -2026-05-13 15:34:46.711802: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 15:35:16.870870 139687754716928 logging_writer.py:48] [146400] global_step=146400, grad_norm=12.727400779724121, loss=3.969163417816162 -I0513 15:36:18.755497 139687763109632 logging_writer.py:48] [146500] global_step=146500, grad_norm=5.42717170715332, loss=3.9526758193969727 -I0513 15:37:27.623495 139687754716928 logging_writer.py:48] [146600] global_step=146600, grad_norm=12.044346809387207, loss=4.180284023284912 -I0513 15:38:34.372785 139687763109632 logging_writer.py:48] [146700] global_step=146700, grad_norm=3.3213958740234375, loss=3.8617756366729736 -I0513 15:39:42.086342 139687754716928 logging_writer.py:48] [146800] global_step=146800, grad_norm=2.926464557647705, loss=3.962233543395996 -I0513 15:40:50.781092 139687763109632 logging_writer.py:48] [146900] global_step=146900, grad_norm=6.3585896492004395, loss=3.8714356422424316 -I0513 15:42:04.609282 139687754716928 logging_writer.py:48] [147000] global_step=147000, grad_norm=5.088845729827881, loss=3.8260703086853027 -I0513 15:43:08.863953 139903809070272 spec.py:333] Evaluating on the training split. -I0513 15:43:14.960413 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 15:44:03.243064 139903809070272 spec.py:363] Evaluating on the test split. -I0513 15:44:04.330333 139903809070272 submission_runner.py:516] Time since start: 72188.21s, Step: 147094, {'train/accuracy': Array(0.00187341, dtype=float32), 'train/loss': Array(7.492316, dtype=float32), 'validation/accuracy': Array(0.00244, dtype=float32), 'validation/loss': Array(7.5076427, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0021, dtype=float32), 'test/loss': Array(7.5356674, dtype=float32), 'test/num_examples': 10000, 'score': 69926.05036187172, 'total_duration': 72188.20590472221, 'accumulated_submission_time': 69926.05036187172, 'accumulated_eval_time': 2251.2275352478027, 'accumulated_logging_time': 6.4613258838653564} -I0513 15:44:04.712582 139687763109632 logging_writer.py:48] [147094] accumulated_eval_time=2251.23, accumulated_logging_time=6.46133, accumulated_submission_time=69926.1, global_step=147094, preemption_count=0, score=69926.1, test/accuracy=0.0021000001579523087, test/loss=7.535667419433594, test/num_examples=10000, total_duration=72188.2, train/accuracy=0.0018734055338427424, train/loss=7.492315769195557, validation/accuracy=0.0024399999529123306, validation/loss=7.50764274597168, validation/num_examples=50000 -I0513 15:44:07.698813 139687754716928 logging_writer.py:48] [147100] global_step=147100, grad_norm=12.511688232421875, loss=3.979996681213379 -I0513 15:45:17.020889 139687763109632 logging_writer.py:48] [147200] global_step=147200, grad_norm=3.8981714248657227, loss=3.7231695652008057 -I0513 15:46:31.078465 139687754716928 logging_writer.py:48] [147300] global_step=147300, grad_norm=6.674967288970947, loss=4.110592842102051 -I0513 15:47:39.314797 139687763109632 logging_writer.py:48] [147400] global_step=147400, grad_norm=5.20046329498291, loss=3.6665241718292236 -I0513 15:48:44.457706 139687754716928 logging_writer.py:48] [147500] global_step=147500, grad_norm=4.690314292907715, loss=3.8882620334625244 -I0513 15:49:57.878504 139687763109632 logging_writer.py:48] [147600] global_step=147600, grad_norm=4.412166595458984, loss=3.9818410873413086 -2026-05-13 15:50:02.026588: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 15:51:02.897823 139687754716928 logging_writer.py:48] [147700] global_step=147700, grad_norm=4.063638210296631, loss=3.8720703125 -I0513 15:52:10.090137 139687763109632 logging_writer.py:48] [147800] global_step=147800, grad_norm=4.387138366699219, loss=3.7562694549560547 -I0513 15:53:21.190680 139687754716928 logging_writer.py:48] [147900] global_step=147900, grad_norm=4.195677280426025, loss=3.751518726348877 -I0513 15:54:36.545141 139687763109632 logging_writer.py:48] [148000] global_step=148000, grad_norm=3.69515061378479, loss=3.7246389389038086 -I0513 15:55:45.044183 139687754716928 logging_writer.py:48] [148100] global_step=148100, grad_norm=3.383946657180786, loss=3.6832375526428223 -I0513 15:56:55.481303 139687763109632 logging_writer.py:48] [148200] global_step=148200, grad_norm=4.207132339477539, loss=3.900099992752075 -I0513 15:58:01.043886 139687754716928 logging_writer.py:48] [148300] global_step=148300, grad_norm=9.754958152770996, loss=3.9203672409057617 -I0513 15:59:06.794017 139687763109632 logging_writer.py:48] [148400] global_step=148400, grad_norm=5.101773262023926, loss=3.919267177581787 -I0513 16:00:24.114238 139687754716928 logging_writer.py:48] [148500] global_step=148500, grad_norm=4.554518699645996, loss=3.686781406402588 -I0513 16:01:33.870666 139687763109632 logging_writer.py:48] [148600] global_step=148600, grad_norm=3.0557968616485596, loss=3.8194761276245117 -I0513 16:02:40.523891 139687754716928 logging_writer.py:48] [148700] global_step=148700, grad_norm=5.14883279800415, loss=3.764087677001953 -I0513 16:03:46.985398 139687763109632 logging_writer.py:48] [148800] global_step=148800, grad_norm=3.271817445755005, loss=3.762169361114502 -2026-05-13 16:04:27.039465: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 16:04:56.864307 139687754716928 logging_writer.py:48] [148900] global_step=148900, grad_norm=4.229083061218262, loss=3.881613254547119 -I0513 16:06:13.175965 139687763109632 logging_writer.py:48] [149000] global_step=149000, grad_norm=4.859543323516846, loss=3.875035285949707 -I0513 16:07:27.652992 139687754716928 logging_writer.py:48] [149100] global_step=149100, grad_norm=2.9685616493225098, loss=3.8511810302734375 -I0513 16:08:31.837098 139687763109632 logging_writer.py:48] [149200] global_step=149200, grad_norm=1.9591914415359497, loss=3.5909671783447266 -I0513 16:09:44.039083 139687754716928 logging_writer.py:48] [149300] global_step=149300, grad_norm=5.084506034851074, loss=4.031536102294922 -I0513 16:10:52.577812 139687763109632 logging_writer.py:48] [149400] global_step=149400, grad_norm=4.532956600189209, loss=3.7850687503814697 -I0513 16:11:57.539200 139687754716928 logging_writer.py:48] [149500] global_step=149500, grad_norm=3.1831021308898926, loss=3.897552013397217 -I0513 16:13:04.263117 139687763109632 logging_writer.py:48] [149600] global_step=149600, grad_norm=75.85710906982422, loss=3.9243428707122803 -I0513 16:14:13.293015 139687754716928 logging_writer.py:48] [149700] global_step=149700, grad_norm=6.58159065246582, loss=3.8394367694854736 -I0513 16:15:22.416765 139687763109632 logging_writer.py:48] [149800] global_step=149800, grad_norm=2.840411901473999, loss=3.7560226917266846 -I0513 16:16:29.892978 139687754716928 logging_writer.py:48] [149900] global_step=149900, grad_norm=5.337430000305176, loss=3.884756088256836 -I0513 16:17:21.719715 139903809070272 spec.py:333] Evaluating on the training split. -I0513 16:17:29.405679 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 16:17:54.569986 139903809070272 spec.py:363] Evaluating on the test split. -I0513 16:17:55.614884 139903809070272 submission_runner.py:516] Time since start: 74219.54s, Step: 149980, {'train/accuracy': Array(0.0024713, dtype=float32), 'train/loss': Array(7.497335, dtype=float32), 'validation/accuracy': Array(0.0025, dtype=float32), 'validation/loss': Array(7.5091887, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0018, dtype=float32), 'test/loss': Array(7.536077, dtype=float32), 'test/num_examples': 10000, 'score': 71922.94839406013, 'total_duration': 74219.54028224945, 'accumulated_submission_time': 71922.94839406013, 'accumulated_eval_time': 2284.97248506546, 'accumulated_logging_time': 6.8999550342559814} -I0513 16:17:56.016473 139687763109632 logging_writer.py:48] [149980] accumulated_eval_time=2284.97, accumulated_logging_time=6.89996, accumulated_submission_time=71922.9, global_step=149980, preemption_count=0, score=71922.9, test/accuracy=0.0018000000854954123, test/loss=7.53607702255249, test/num_examples=10000, total_duration=74219.5, train/accuracy=0.002471301006153226, train/loss=7.497334957122803, validation/accuracy=0.0024999999441206455, validation/loss=7.509188652038574, validation/num_examples=50000 -I0513 16:18:09.054718 139687754716928 logging_writer.py:48] [150000] global_step=150000, grad_norm=5.35140323638916, loss=3.9775969982147217 -I0513 16:19:24.090821 139687763109632 logging_writer.py:48] [150100] global_step=150100, grad_norm=4.142479419708252, loss=3.927765369415283 -2026-05-13 16:19:29.530787: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 16:20:31.145710 139687754716928 logging_writer.py:48] [150200] global_step=150200, grad_norm=13.88120174407959, loss=3.9683947563171387 -I0513 16:21:47.088719 139687763109632 logging_writer.py:48] [150300] global_step=150300, grad_norm=2.865879774093628, loss=3.808566093444824 -I0513 16:22:54.935615 139687754716928 logging_writer.py:48] [150400] global_step=150400, grad_norm=11.834768295288086, loss=3.976593255996704 -I0513 16:24:04.584400 139687763109632 logging_writer.py:48] [150500] global_step=150500, grad_norm=3.5319480895996094, loss=3.745652198791504 -I0513 16:25:16.661589 139687754716928 logging_writer.py:48] [150600] global_step=150600, grad_norm=4.929830551147461, loss=3.842797040939331 -I0513 16:26:26.358871 139687763109632 logging_writer.py:48] [150700] global_step=150700, grad_norm=2.938366413116455, loss=3.815891742706299 -I0513 16:27:31.161199 139687754716928 logging_writer.py:48] [150800] global_step=150800, grad_norm=3.687251091003418, loss=3.9081478118896484 -I0513 16:28:38.578025 139687763109632 logging_writer.py:48] [150900] global_step=150900, grad_norm=24.976579666137695, loss=3.8210854530334473 -I0513 16:29:47.370573 139687754716928 logging_writer.py:48] [151000] global_step=151000, grad_norm=11.597968101501465, loss=3.8551480770111084 -I0513 16:30:58.492720 139687763109632 logging_writer.py:48] [151100] global_step=151100, grad_norm=3.5757510662078857, loss=3.689439296722412 -I0513 16:32:05.086080 139687754716928 logging_writer.py:48] [151200] global_step=151200, grad_norm=4.606236934661865, loss=3.949587345123291 -I0513 16:33:14.180388 139687763109632 logging_writer.py:48] [151300] global_step=151300, grad_norm=11.574318885803223, loss=3.936323642730713 -2026-05-13 16:33:54.605174: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 16:34:22.982350 139687754716928 logging_writer.py:48] [151400] global_step=151400, grad_norm=3.442448139190674, loss=3.7879083156585693 -I0513 16:35:31.652761 139687763109632 logging_writer.py:48] [151500] global_step=151500, grad_norm=5.850992679595947, loss=3.8674917221069336 -I0513 16:36:37.923699 139687754716928 logging_writer.py:48] [151600] global_step=151600, grad_norm=4.484780788421631, loss=3.927189826965332 -I0513 16:37:44.424001 139687763109632 logging_writer.py:48] [151700] global_step=151700, grad_norm=3.676269054412842, loss=3.833505392074585 -I0513 16:38:44.620744 139687754716928 logging_writer.py:48] [151800] global_step=151800, grad_norm=3.7917354106903076, loss=3.904345750808716 -I0513 16:39:44.452841 139687763109632 logging_writer.py:48] [151900] global_step=151900, grad_norm=4.094565391540527, loss=3.818464756011963 -I0513 16:40:50.648673 139687754716928 logging_writer.py:48] [152000] global_step=152000, grad_norm=3.5094549655914307, loss=3.863084077835083 -I0513 16:41:56.083091 139687763109632 logging_writer.py:48] [152100] global_step=152100, grad_norm=2.9403560161590576, loss=3.709188938140869 -I0513 16:43:01.441246 139687754716928 logging_writer.py:48] [152200] global_step=152200, grad_norm=5.310137748718262, loss=3.976884603500366 -I0513 16:44:08.031468 139687763109632 logging_writer.py:48] [152300] global_step=152300, grad_norm=4.602383613586426, loss=3.8807883262634277 -I0513 16:45:19.064689 139687754716928 logging_writer.py:48] [152400] global_step=152400, grad_norm=5.9561285972595215, loss=3.8538031578063965 -I0513 16:46:25.346712 139687763109632 logging_writer.py:48] [152500] global_step=152500, grad_norm=4.6170973777771, loss=3.946331024169922 -I0513 16:47:31.012945 139687754716928 logging_writer.py:48] [152600] global_step=152600, grad_norm=4.88385534286499, loss=3.9479641914367676 -2026-05-13 16:47:38.783729: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 16:48:36.850704 139687763109632 logging_writer.py:48] [152700] global_step=152700, grad_norm=3.898998498916626, loss=3.8737728595733643 -I0513 16:49:45.941119 139687754716928 logging_writer.py:48] [152800] global_step=152800, grad_norm=5.07167911529541, loss=3.782308578491211 -I0513 16:50:50.208293 139687763109632 logging_writer.py:48] [152900] global_step=152900, grad_norm=4.237201690673828, loss=3.6983323097229004 -I0513 16:51:11.548806 139903809070272 spec.py:333] Evaluating on the training split. -I0513 16:51:17.937734 139903809070272 spec.py:346] Evaluating on the validation split. -I0513 16:51:43.985398 139903809070272 spec.py:363] Evaluating on the test split. -I0513 16:51:45.063807 139903809070272 submission_runner.py:516] Time since start: 76248.95s, Step: 152933, {'train/accuracy': Array(0.0026706, dtype=float32), 'train/loss': Array(7.502706, dtype=float32), 'validation/accuracy': Array(0.00278, dtype=float32), 'validation/loss': Array(7.5088005, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.002, dtype=float32), 'test/loss': Array(7.536695, dtype=float32), 'test/num_examples': 10000, 'score': 73918.37983560562, 'total_duration': 76248.95145559311, 'accumulated_submission_time': 73918.37983560562, 'accumulated_eval_time': 2318.299463748932, 'accumulated_logging_time': 7.3489696979522705} -I0513 16:51:45.379731 139687754716928 logging_writer.py:48] [152933] accumulated_eval_time=2318.3, accumulated_logging_time=7.34897, accumulated_submission_time=73918.4, global_step=152933, preemption_count=0, score=73918.4, test/accuracy=0.0020000000949949026, test/loss=7.5366950035095215, test/num_examples=10000, total_duration=76249, train/accuracy=0.0026705993805080652, train/loss=7.502706050872803, validation/accuracy=0.0027799999807029963, validation/loss=7.508800506591797, validation/num_examples=50000 -I0513 16:52:31.907673 139687763109632 logging_writer.py:48] [153000] global_step=153000, grad_norm=3.741344690322876, loss=3.824896812438965 -I0513 16:53:46.807878 139687754716928 logging_writer.py:48] [153100] global_step=153100, grad_norm=3.741760730743408, loss=3.8072686195373535 -I0513 16:54:59.247668 139687763109632 logging_writer.py:48] [153200] global_step=153200, grad_norm=3.2579946517944336, loss=3.7740604877471924 -I0513 16:56:16.046027 139687754716928 logging_writer.py:48] [153300] global_step=153300, grad_norm=3.876110315322876, loss=3.730680227279663 -I0513 16:57:29.115822 139687763109632 logging_writer.py:48] [153400] global_step=153400, grad_norm=4.552067756652832, loss=3.824369192123413 -I0513 16:58:42.169894 139687754716928 logging_writer.py:48] [153500] global_step=153500, grad_norm=3.168260097503662, loss=3.8680336475372314 -I0513 16:59:54.405544 139687763109632 logging_writer.py:48] [153600] global_step=153600, grad_norm=4.868819713592529, loss=3.746490478515625 -I0513 17:01:02.669850 139687754716928 logging_writer.py:48] [153700] global_step=153700, grad_norm=6.4317216873168945, loss=3.914503574371338 -I0513 17:02:08.190501 139687763109632 logging_writer.py:48] [153800] global_step=153800, grad_norm=8.59994888305664, loss=3.919564723968506 -2026-05-13 17:02:51.555110: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 17:03:15.264073 139687754716928 logging_writer.py:48] [153900] global_step=153900, grad_norm=4.560880184173584, loss=3.8131439685821533 -I0513 17:04:25.354224 139687763109632 logging_writer.py:48] [154000] global_step=154000, grad_norm=2.7458622455596924, loss=3.6829123497009277 -I0513 17:05:27.383045 139687754716928 logging_writer.py:48] [154100] global_step=154100, grad_norm=9.315206527709961, loss=4.050850868225098 -I0513 17:06:31.737194 139687763109632 logging_writer.py:48] [154200] global_step=154200, grad_norm=4.950337886810303, loss=3.8840558528900146 -I0513 17:07:41.941112 139687754716928 logging_writer.py:48] [154300] global_step=154300, grad_norm=10.534607887268066, loss=3.773994207382202 -I0513 17:09:00.252845 139687763109632 logging_writer.py:48] [154400] global_step=154400, grad_norm=3.7198245525360107, loss=3.8218724727630615 -I0513 17:10:12.339599 139687754716928 logging_writer.py:48] [154500] global_step=154500, grad_norm=5.419073104858398, loss=3.9301095008850098 -I0513 17:11:25.166809 139687763109632 logging_writer.py:48] [154600] global_step=154600, grad_norm=2.5316293239593506, loss=3.760789155960083 -I0513 17:12:36.981391 139687754716928 logging_writer.py:48] [154700] global_step=154700, grad_norm=3.285118579864502, loss=3.6156809329986572 -I0513 17:13:45.058838 139687763109632 logging_writer.py:48] [154800] global_step=154800, grad_norm=3.31827449798584, loss=3.7218990325927734 -I0513 17:15:00.704227 139687754716928 logging_writer.py:48] [154900] global_step=154900, grad_norm=11.172554016113281, loss=3.8823890686035156 -I0513 17:16:11.235091 139687763109632 logging_writer.py:48] [155000] global_step=155000, grad_norm=6.34672212600708, loss=3.8571784496307373 -I0513 17:17:24.217732 139687754716928 logging_writer.py:48] [155100] global_step=155100, grad_norm=4.832947254180908, loss=3.8364570140838623 -2026-05-13 17:17:32.648641: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 17:18:32.135396 139687763109632 logging_writer.py:48] [155200] global_step=155200, grad_norm=8.382080078125, loss=3.8235442638397217 -I0513 17:19:35.586720 139687754716928 logging_writer.py:48] [155300] global_step=155300, grad_norm=2.8770084381103516, loss=3.8691883087158203 -I0513 17:20:48.467298 139687763109632 logging_writer.py:48] [155400] global_step=155400, grad_norm=24.73626708984375, loss=4.035109043121338 -I0513 17:22:00.509332 139687754716928 logging_writer.py:48] [155500] global_step=155500, grad_norm=3.1201658248901367, loss=3.8010940551757812 -I0513 17:23:20.724888 139687763109632 logging_writer.py:48] [155600] global_step=155600, grad_norm=3.328146457672119, loss=3.9718940258026123 -I0513 17:24:32.190037 139687754716928 logging_writer.py:48] [155700] global_step=155700, grad_norm=3.185468912124634, loss=3.795300245285034 -I0513 17:25:02.944737 139687763109632 logging_writer.py:48] [155746] global_step=155746, preemption_count=0, score=75915.6 -I0513 17:25:03.837218 139903809070272 submission_runner.py:857] Final imagenet_resnet score: 75915.5999929905 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-08-45-50.log b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-08-45-50.log new file mode 100644 index 000000000..800289f4a --- /dev/null +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-08-45-50.log @@ -0,0 +1,2584 @@ +python submission_runner.py --framework=jax --workload=imagenet_resnet --submission_path=submissions_algorithms/submissions/self_tuning/schedule_free_adamw_jax_v2_bn_fix/submission.py --data_dir=/data/imagenet/jax --experiment_dir=/experiment_runs --experiment_name=submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_0 --overwrite=True --save_checkpoints=False --rng_seed=-999461833 --imagenet_v2_data_dir=/data/imagenet/jax --tuning_ruleset=self 2>&1 | tee -a /logs/imagenet_resnet_jax_08-31-2026-08-45-50.log +2026-08-31 08:45:50.942166: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered +WARNING: All log messages before absl::InitializeLog() is called are written to STDERR +E0000 00:00:1788165950.965208 13 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered +E0000 00:00:1788165950.972781 13 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered +W0000 00:00:1788165950.991219 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +W0000 00:00:1788165950.991245 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +W0000 00:00:1788165950.991248 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +W0000 00:00:1788165950.991250 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +/usr/local/lib/python3.11/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.) + _C._set_float32_matmul_precision(precision) +/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'repr' attribute with value False was provided to the `Field()` function, which has no effect in the context it was used. 'repr' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. + warnings.warn( +/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'frozen' attribute with value True was provided to the `Field()` function, which has no effect in the context it was used. 'frozen' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. + warnings.warn( +INFO:2026-08-31 08:46:01,638:jax._src.xla_bridge:830: Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory +I0831 08:46:01.638224 139757377230016 xla_bridge.py:830] Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory +I0831 08:46:02.395941 139757377230016 logger_utils.py:59] Removing existing experiment directory /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_0/imagenet_resnet_jax because --overwrite was set. +I0831 08:46:02.401387 139757377230016 logger_utils.py:84] Creating experiment directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_0/imagenet_resnet_jax. +I0831 08:46:02.996536 139757377230016 submission_runner.py:741] Creating directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_0/imagenet_resnet_jax/trial_1. +I0831 08:46:03.261393 139757377230016 submission_runner.py:242] Initializing dataset. +I0831 08:46:03.571712 139757377230016 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 08:46:03.636668 139757377230016 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. +I0831 08:46:03.914165 139757377230016 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. +I0831 08:46:03.993771 139757377230016 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 08:46:04.868155 139757377230016 submission_runner.py:251] Initializing model. +I0831 08:46:27.853143 139757377230016 submission_runner.py:294] Initializing optimizer. +I0831 08:46:29.413419 139757377230016 submission_runner.py:299] Initializing metrics bundle. +I0831 08:46:29.413624 139757377230016 submission_runner.py:321] Initializing checkpoint and logger. +I0831 08:46:29.420755 139757377230016 checkpoints.py:1098] Found no checkpoint files in /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_0/imagenet_resnet_jax/trial_1 with prefix checkpoint_ +I0831 08:46:29.420875 139757377230016 submission_runner.py:344] Saving meta data to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_0/imagenet_resnet_jax/trial_1/meta_data_0.json. +I0831 08:46:29.647591 139757377230016 submission_runner.py:348] Saving flags to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_0/imagenet_resnet_jax/trial_1/flags_0.json. +I0831 08:46:29.662222 139757377230016 submission_runner.py:359] Starting training loop. +/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. +See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. + warnings.warn("Some donated buffers were not usable:" +I0831 08:47:22.276690 139598934832896 logging_writer.py:48] [0] global_step=0, grad_norm=0.6744329333305359, loss=6.914268493652344 +I0831 08:47:23.463624 139757377230016 spec.py:333] Evaluating on the training split. +I0831 08:47:23.730556 139757377230016 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 08:47:23.735954 139757377230016 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. +I0831 08:47:23.754035 139757377230016 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. +I0831 08:47:23.793460 139757377230016 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 08:47:49.825342 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 08:47:49.829616 139757377230016 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 08:47:49.851871 139757377230016 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. +I0831 08:47:49.854423 139757377230016 reader.py:262] Creating a tf.data.Dataset reading 64 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. +I0831 08:47:49.890127 139757377230016 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split validation, from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 08:48:22.311747 139757377230016 spec.py:363] Evaluating on the test split. +I0831 08:48:22.417903 139757377230016 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 +I0831 08:48:22.510527 139757377230016 reader.py:262] Creating a tf.data.Dataset reading 16 files located in folders: /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0. +I0831 08:48:22.555934 139757377230016 logging_logger.py:49] Constructing tf.data.Dataset imagenet_v2 for split test, from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 +I0831 08:48:28.164529 139757377230016 submission_runner.py:516] Time since start: 118.50s, Step: 1, {'train/accuracy': Array(0.00107621, dtype=float32), 'train/loss': Array(6.9101987, dtype=float32), 'validation/accuracy': Array(0.0011, dtype=float32), 'validation/loss': Array(6.910505, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(6.910612, dtype=float32), 'test/num_examples': 10000, 'score': 53.801310300827026, 'total_duration': 118.50065970420837, 'accumulated_submission_time': 53.801310300827026, 'accumulated_eval_time': 64.69926476478577, 'accumulated_logging_time': 0} +I0831 08:48:28.195801 139540710803200 logging_writer.py:48] [1] accumulated_eval_time=64.6993, accumulated_logging_time=0, accumulated_submission_time=53.8013, global_step=1, preemption_count=0, score=53.8013, test/accuracy=0.0013000001199543476, test/loss=6.910612106323242, test/num_examples=10000, total_duration=118.501, train/accuracy=0.0010762116871774197, train/loss=6.91019868850708, validation/accuracy=0.0010999999940395355, validation/loss=6.9105048179626465, validation/num_examples=50000 +/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. +See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. + warnings.warn("Some donated buffers were not usable:" +I0831 08:49:10.299362 139540459153152 logging_writer.py:48] [100] global_step=100, grad_norm=0.6860280632972717, loss=6.796332359313965 +I0831 08:49:59.633524 139540467545856 logging_writer.py:48] [200] global_step=200, grad_norm=0.8489809632301331, loss=6.484612464904785 +I0831 08:50:48.453057 139540459153152 logging_writer.py:48] [300] global_step=300, grad_norm=1.0689995288848877, loss=6.201939582824707 +I0831 08:51:37.227531 139540467545856 logging_writer.py:48] [400] global_step=400, grad_norm=2.035437822341919, loss=5.8884477615356445 +I0831 08:52:27.273710 139540459153152 logging_writer.py:48] [500] global_step=500, grad_norm=1.8697954416275024, loss=5.808591842651367 +I0831 08:53:15.209136 139540467545856 logging_writer.py:48] [600] global_step=600, grad_norm=3.970616579055786, loss=5.630148887634277 +I0831 08:54:04.635902 139540459153152 logging_writer.py:48] [700] global_step=700, grad_norm=4.3386640548706055, loss=5.43958854675293 +I0831 08:54:53.071150 139540467545856 logging_writer.py:48] [800] global_step=800, grad_norm=3.2051055431365967, loss=5.24249792098999 +I0831 08:55:39.954963 139540459153152 logging_writer.py:48] [900] global_step=900, grad_norm=4.9889445304870605, loss=5.143521308898926 +I0831 08:56:27.081810 139540467545856 logging_writer.py:48] [1000] global_step=1000, grad_norm=4.439692974090576, loss=4.97685432434082 +I0831 08:57:16.693206 139540459153152 logging_writer.py:48] [1100] global_step=1100, grad_norm=5.249515056610107, loss=4.953073501586914 +I0831 08:58:05.986883 139540467545856 logging_writer.py:48] [1200] global_step=1200, grad_norm=5.5972113609313965, loss=4.812962532043457 +I0831 08:58:41.234569 139540459153152 logging_writer.py:48] [1300] global_step=1300, grad_norm=4.05490255355835, loss=4.6063947677612305 +I0831 08:59:08.306896 139540467545856 logging_writer.py:48] [1400] global_step=1400, grad_norm=4.379157543182373, loss=4.607489585876465 +I0831 08:59:35.480682 139540459153152 logging_writer.py:48] [1500] global_step=1500, grad_norm=3.1159729957580566, loss=4.377635955810547 +I0831 09:00:02.579827 139540467545856 logging_writer.py:48] [1600] global_step=1600, grad_norm=4.363247871398926, loss=4.354098796844482 +I0831 09:00:29.663279 139540459153152 logging_writer.py:48] [1700] global_step=1700, grad_norm=5.260931491851807, loss=4.3621296882629395 +I0831 09:00:56.839418 139540467545856 logging_writer.py:48] [1800] global_step=1800, grad_norm=5.4478230476379395, loss=4.281497001647949 +I0831 09:01:23.936638 139540459153152 logging_writer.py:48] [1900] global_step=1900, grad_norm=4.826054573059082, loss=4.091533660888672 +I0831 09:01:51.256285 139540467545856 logging_writer.py:48] [2000] global_step=2000, grad_norm=4.0353264808654785, loss=4.148013591766357 +I0831 09:02:18.421770 139540459153152 logging_writer.py:48] [2100] global_step=2100, grad_norm=5.712440013885498, loss=4.054986000061035 +I0831 09:02:45.563501 139540467545856 logging_writer.py:48] [2200] global_step=2200, grad_norm=4.179815769195557, loss=3.828669786453247 +I0831 09:03:12.686919 139540459153152 logging_writer.py:48] [2300] global_step=2300, grad_norm=3.295295238494873, loss=3.900071144104004 +I0831 09:03:39.874315 139540467545856 logging_writer.py:48] [2400] global_step=2400, grad_norm=5.145644664764404, loss=3.8688621520996094 +I0831 09:04:06.944254 139540459153152 logging_writer.py:48] [2500] global_step=2500, grad_norm=3.3030648231506348, loss=3.7171175479888916 +I0831 09:04:34.034398 139540467545856 logging_writer.py:48] [2600] global_step=2600, grad_norm=4.079652309417725, loss=3.558654308319092 +I0831 09:05:01.180208 139540459153152 logging_writer.py:48] [2700] global_step=2700, grad_norm=4.294636249542236, loss=3.553633689880371 +I0831 09:05:28.297868 139540467545856 logging_writer.py:48] [2800] global_step=2800, grad_norm=3.4494917392730713, loss=3.404123306274414 +I0831 09:05:55.396698 139540459153152 logging_writer.py:48] [2900] global_step=2900, grad_norm=3.6309683322906494, loss=3.449347496032715 +I0831 09:06:22.797955 139540467545856 logging_writer.py:48] [3000] global_step=3000, grad_norm=4.545529365539551, loss=3.322761297225952 +I0831 09:06:49.915161 139540459153152 logging_writer.py:48] [3100] global_step=3100, grad_norm=3.627488851547241, loss=3.417109251022339 +I0831 09:07:17.033184 139540467545856 logging_writer.py:48] [3200] global_step=3200, grad_norm=2.8654136657714844, loss=3.220179319381714 +I0831 09:07:44.195625 139540459153152 logging_writer.py:48] [3300] global_step=3300, grad_norm=3.651508331298828, loss=3.187009811401367 +I0831 09:08:11.306868 139540467545856 logging_writer.py:48] [3400] global_step=3400, grad_norm=3.55690598487854, loss=3.1201210021972656 +I0831 09:08:38.411463 139540459153152 logging_writer.py:48] [3500] global_step=3500, grad_norm=3.3634865283966064, loss=3.0035715103149414 +I0831 09:09:05.595075 139540467545856 logging_writer.py:48] [3600] global_step=3600, grad_norm=3.388749361038208, loss=3.0160908699035645 +I0831 09:09:32.691494 139540459153152 logging_writer.py:48] [3700] global_step=3700, grad_norm=2.7773454189300537, loss=3.1409807205200195 +I0831 09:09:59.780807 139540467545856 logging_writer.py:48] [3800] global_step=3800, grad_norm=3.021484851837158, loss=2.8881890773773193 +I0831 09:10:26.945915 139540459153152 logging_writer.py:48] [3900] global_step=3900, grad_norm=3.2017390727996826, loss=2.949516773223877 +I0831 09:10:54.061642 139540467545856 logging_writer.py:48] [4000] global_step=4000, grad_norm=2.9251468181610107, loss=2.765871286392212 +I0831 09:11:21.382022 139540459153152 logging_writer.py:48] [4100] global_step=4100, grad_norm=3.714607000350952, loss=2.847548484802246 +I0831 09:11:48.572063 139540467545856 logging_writer.py:48] [4200] global_step=4200, grad_norm=2.4411838054656982, loss=2.8455421924591064 +I0831 09:12:15.676466 139540459153152 logging_writer.py:48] [4300] global_step=4300, grad_norm=3.9159207344055176, loss=2.766733169555664 +I0831 09:12:42.794403 139540467545856 logging_writer.py:48] [4400] global_step=4400, grad_norm=2.2171683311462402, loss=2.6653366088867188 +I0831 09:13:10.003541 139540459153152 logging_writer.py:48] [4500] global_step=4500, grad_norm=3.2516491413116455, loss=2.638270378112793 +I0831 09:13:37.118839 139540467545856 logging_writer.py:48] [4600] global_step=4600, grad_norm=2.954125165939331, loss=2.5777575969696045 +I0831 09:14:04.209929 139540459153152 logging_writer.py:48] [4700] global_step=4700, grad_norm=2.444253444671631, loss=2.624563694000244 +I0831 09:14:31.387999 139540467545856 logging_writer.py:48] [4800] global_step=4800, grad_norm=2.2688074111938477, loss=2.644932746887207 +I0831 09:14:58.508377 139540459153152 logging_writer.py:48] [4900] global_step=4900, grad_norm=2.913524866104126, loss=2.5804240703582764 +I0831 09:15:25.619944 139540467545856 logging_writer.py:48] [5000] global_step=5000, grad_norm=3.028923749923706, loss=2.500077247619629 +I0831 09:15:53.013469 139540459153152 logging_writer.py:48] [5100] global_step=5100, grad_norm=2.381420135498047, loss=2.496798038482666 +I0831 09:16:20.112979 139540467545856 logging_writer.py:48] [5200] global_step=5200, grad_norm=2.5973732471466064, loss=2.3642289638519287 +I0831 09:16:47.220714 139540459153152 logging_writer.py:48] [5300] global_step=5300, grad_norm=2.116661787033081, loss=2.403353214263916 +I0831 09:17:14.395443 139540467545856 logging_writer.py:48] [5400] global_step=5400, grad_norm=2.644618034362793, loss=2.4623560905456543 +I0831 09:17:41.510859 139540459153152 logging_writer.py:48] [5500] global_step=5500, grad_norm=2.4153525829315186, loss=2.2745778560638428 +I0831 09:18:08.608169 139540467545856 logging_writer.py:48] [5600] global_step=5600, grad_norm=2.4079883098602295, loss=2.4579966068267822 +I0831 09:18:35.785987 139540459153152 logging_writer.py:48] [5700] global_step=5700, grad_norm=2.4223523139953613, loss=2.3152859210968018 +I0831 09:19:02.893011 139540467545856 logging_writer.py:48] [5800] global_step=5800, grad_norm=2.574902296066284, loss=2.3781397342681885 +I0831 09:19:30.018755 139540459153152 logging_writer.py:48] [5900] global_step=5900, grad_norm=2.499323606491089, loss=2.262944459915161 +I0831 09:19:57.198277 139540467545856 logging_writer.py:48] [6000] global_step=6000, grad_norm=2.5902411937713623, loss=2.2581727504730225 +I0831 09:20:24.288435 139540459153152 logging_writer.py:48] [6100] global_step=6100, grad_norm=2.0480172634124756, loss=2.238314390182495 +I0831 09:20:51.688616 139540467545856 logging_writer.py:48] [6200] global_step=6200, grad_norm=2.4927284717559814, loss=2.2575788497924805 +I0831 09:21:18.927880 139540459153152 logging_writer.py:48] [6300] global_step=6300, grad_norm=2.9313101768493652, loss=2.211747646331787 +I0831 09:21:44.277741 139757377230016 spec.py:333] Evaluating on the training split. +I0831 09:21:54.956246 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 09:22:05.031423 139757377230016 spec.py:363] Evaluating on the test split. +I0831 09:22:06.031125 139757377230016 submission_runner.py:516] Time since start: 2136.37s, Step: 6395, {'train/accuracy': Array(0.6256776, dtype=float32), 'train/loss': Array(1.5689651, dtype=float32), 'validation/accuracy': Array(0.56868, dtype=float32), 'validation/loss': Array(1.8507421, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.43960002, dtype=float32), 'test/loss': Array(2.5965824, dtype=float32), 'test/num_examples': 10000, 'score': 2049.826003551483, 'total_duration': 2136.365298271179, 'accumulated_submission_time': 2049.826003551483, 'accumulated_eval_time': 86.44905090332031, 'accumulated_logging_time': 0.03977036476135254} +I0831 09:22:06.087952 139540719195904 logging_writer.py:48] [6395] accumulated_eval_time=86.4491, accumulated_logging_time=0.0397704, accumulated_submission_time=2049.83, global_step=6395, preemption_count=0, score=2049.83, test/accuracy=0.43960002064704895, test/loss=2.5965824127197266, test/num_examples=10000, total_duration=2136.37, train/accuracy=0.6256775856018066, train/loss=1.5689650774002075, validation/accuracy=0.5686799883842468, validation/loss=1.8507421016693115, validation/num_examples=50000 +I0831 09:22:07.849803 139540727588608 logging_writer.py:48] [6400] global_step=6400, grad_norm=2.3953588008880615, loss=2.3829550743103027 +I0831 09:22:35.075583 139540719195904 logging_writer.py:48] [6500] global_step=6500, grad_norm=2.4519715309143066, loss=2.1553525924682617 +I0831 09:23:02.258405 139540727588608 logging_writer.py:48] [6600] global_step=6600, grad_norm=2.8765344619750977, loss=2.18791127204895 +I0831 09:23:29.391064 139540719195904 logging_writer.py:48] [6700] global_step=6700, grad_norm=2.5946249961853027, loss=2.205749988555908 +I0831 09:23:56.512482 139540727588608 logging_writer.py:48] [6800] global_step=6800, grad_norm=1.8977155685424805, loss=2.088545799255371 +I0831 09:24:23.705001 139540719195904 logging_writer.py:48] [6900] global_step=6900, grad_norm=2.2655646800994873, loss=2.188692569732666 +I0831 09:24:50.862033 139540727588608 logging_writer.py:48] [7000] global_step=7000, grad_norm=2.6860148906707764, loss=2.096611499786377 +I0831 09:25:18.003448 139540719195904 logging_writer.py:48] [7100] global_step=7100, grad_norm=2.7385330200195312, loss=2.0667901039123535 +I0831 09:25:45.202214 139540727588608 logging_writer.py:48] [7200] global_step=7200, grad_norm=2.54921817779541, loss=2.0317254066467285 +I0831 09:26:12.540767 139540719195904 logging_writer.py:48] [7300] global_step=7300, grad_norm=2.3847198486328125, loss=2.1349599361419678 +I0831 09:26:39.639805 139540727588608 logging_writer.py:48] [7400] global_step=7400, grad_norm=3.0261738300323486, loss=2.1404995918273926 +I0831 09:27:06.837857 139540719195904 logging_writer.py:48] [7500] global_step=7500, grad_norm=2.2785348892211914, loss=2.051191568374634 +I0831 09:27:33.953276 139540727588608 logging_writer.py:48] [7600] global_step=7600, grad_norm=2.905258893966675, loss=2.0209150314331055 +I0831 09:28:01.042492 139540719195904 logging_writer.py:48] [7700] global_step=7700, grad_norm=2.252206802368164, loss=2.1314549446105957 +I0831 09:28:28.207807 139540727588608 logging_writer.py:48] [7800] global_step=7800, grad_norm=2.3922119140625, loss=2.136021375656128 +I0831 09:28:55.287560 139540719195904 logging_writer.py:48] [7900] global_step=7900, grad_norm=2.025501251220703, loss=1.976367473602295 +I0831 09:29:22.414036 139540727588608 logging_writer.py:48] [8000] global_step=8000, grad_norm=2.2612087726593018, loss=1.9049280881881714 +I0831 09:29:49.580972 139540719195904 logging_writer.py:48] [8100] global_step=8100, grad_norm=3.101289749145508, loss=2.07089900970459 +I0831 09:30:16.702055 139540727588608 logging_writer.py:48] [8200] global_step=8200, grad_norm=2.6022586822509766, loss=1.9978675842285156 +I0831 09:30:44.050298 139540719195904 logging_writer.py:48] [8300] global_step=8300, grad_norm=2.656414031982422, loss=1.9280245304107666 +I0831 09:31:11.205246 139540727588608 logging_writer.py:48] [8400] global_step=8400, grad_norm=2.0691564083099365, loss=2.0323984622955322 +I0831 09:31:38.309885 139540719195904 logging_writer.py:48] [8500] global_step=8500, grad_norm=2.704132080078125, loss=1.9894758462905884 +I0831 09:32:05.398619 139540727588608 logging_writer.py:48] [8600] global_step=8600, grad_norm=2.2918899059295654, loss=1.8938233852386475 +I0831 09:32:32.558248 139540719195904 logging_writer.py:48] [8700] global_step=8700, grad_norm=2.8639109134674072, loss=1.8969879150390625 +I0831 09:32:59.689872 139540727588608 logging_writer.py:48] [8800] global_step=8800, grad_norm=2.5532844066619873, loss=2.0150439739227295 +I0831 09:33:26.809526 139540719195904 logging_writer.py:48] [8900] global_step=8900, grad_norm=3.2368154525756836, loss=1.959276556968689 +I0831 09:33:53.986835 139540727588608 logging_writer.py:48] [9000] global_step=9000, grad_norm=2.3949527740478516, loss=1.8857829570770264 +I0831 09:34:21.090270 139540719195904 logging_writer.py:48] [9100] global_step=9100, grad_norm=2.3167433738708496, loss=1.8994563817977905 +I0831 09:34:48.174309 139540727588608 logging_writer.py:48] [9200] global_step=9200, grad_norm=2.2682154178619385, loss=1.915427327156067 +I0831 09:35:15.372652 139540719195904 logging_writer.py:48] [9300] global_step=9300, grad_norm=1.789599895477295, loss=1.7792565822601318 +I0831 09:35:42.707122 139540727588608 logging_writer.py:48] [9400] global_step=9400, grad_norm=2.2247719764709473, loss=1.8861719369888306 +I0831 09:36:09.808723 139540719195904 logging_writer.py:48] [9500] global_step=9500, grad_norm=2.167947769165039, loss=1.9237910509109497 +I0831 09:36:37.043987 139540727588608 logging_writer.py:48] [9600] global_step=9600, grad_norm=2.227323293685913, loss=1.800553798675537 +I0831 09:37:04.181745 139540719195904 logging_writer.py:48] [9700] global_step=9700, grad_norm=1.8557896614074707, loss=1.7982640266418457 +I0831 09:37:31.339177 139540727588608 logging_writer.py:48] [9800] global_step=9800, grad_norm=2.230602502822876, loss=1.812801480293274 +I0831 09:37:58.865085 139540719195904 logging_writer.py:48] [9900] global_step=9900, grad_norm=1.8674036264419556, loss=1.8508939743041992 +I0831 09:38:26.429589 139540727588608 logging_writer.py:48] [10000] global_step=10000, grad_norm=2.016143321990967, loss=1.8326280117034912 +I0831 09:38:53.786630 139540719195904 logging_writer.py:48] [10100] global_step=10100, grad_norm=2.6098203659057617, loss=1.8939179182052612 +I0831 09:39:21.197757 139540727588608 logging_writer.py:48] [10200] global_step=10200, grad_norm=2.486398220062256, loss=1.829833745956421 +I0831 09:39:48.288702 139540719195904 logging_writer.py:48] [10300] global_step=10300, grad_norm=2.1197168827056885, loss=1.787002682685852 +I0831 09:40:15.606040 139540727588608 logging_writer.py:48] [10400] global_step=10400, grad_norm=2.551349401473999, loss=1.8501476049423218 +I0831 09:40:42.761992 139540719195904 logging_writer.py:48] [10500] global_step=10500, grad_norm=2.1618452072143555, loss=1.7364336252212524 +I0831 09:41:09.869085 139540727588608 logging_writer.py:48] [10600] global_step=10600, grad_norm=2.595633029937744, loss=1.7848546504974365 +I0831 09:41:36.970292 139540719195904 logging_writer.py:48] [10700] global_step=10700, grad_norm=2.30145525932312, loss=1.739357829093933 +I0831 09:42:04.128305 139540727588608 logging_writer.py:48] [10800] global_step=10800, grad_norm=2.1686511039733887, loss=1.7244539260864258 +I0831 09:42:31.240156 139540719195904 logging_writer.py:48] [10900] global_step=10900, grad_norm=2.149010419845581, loss=1.7005544900894165 +I0831 09:42:58.351098 139540727588608 logging_writer.py:48] [11000] global_step=11000, grad_norm=2.283306837081909, loss=1.7669949531555176 +I0831 09:43:25.528100 139540719195904 logging_writer.py:48] [11100] global_step=11100, grad_norm=2.230611801147461, loss=1.705134630203247 +I0831 09:43:52.602515 139540727588608 logging_writer.py:48] [11200] global_step=11200, grad_norm=2.1108078956604004, loss=1.836393117904663 +I0831 09:44:19.738616 139540719195904 logging_writer.py:48] [11300] global_step=11300, grad_norm=2.2637734413146973, loss=1.7684381008148193 +I0831 09:44:46.915490 139540727588608 logging_writer.py:48] [11400] global_step=11400, grad_norm=2.2170448303222656, loss=1.7492544651031494 +I0831 09:45:14.233112 139540719195904 logging_writer.py:48] [11500] global_step=11500, grad_norm=2.292778730392456, loss=1.7742791175842285 +I0831 09:45:41.351654 139540727588608 logging_writer.py:48] [11600] global_step=11600, grad_norm=1.9491803646087646, loss=1.691994547843933 +I0831 09:46:08.526179 139540719195904 logging_writer.py:48] [11700] global_step=11700, grad_norm=2.3708863258361816, loss=1.637040376663208 +I0831 09:46:35.639300 139540727588608 logging_writer.py:48] [11800] global_step=11800, grad_norm=2.954817533493042, loss=1.7830240726470947 +I0831 09:47:02.767083 139540719195904 logging_writer.py:48] [11900] global_step=11900, grad_norm=2.252830982208252, loss=1.7220581769943237 +I0831 09:47:29.932803 139540727588608 logging_writer.py:48] [12000] global_step=12000, grad_norm=2.3853113651275635, loss=1.7560423612594604 +I0831 09:47:57.025314 139540719195904 logging_writer.py:48] [12100] global_step=12100, grad_norm=2.207202196121216, loss=1.7237346172332764 +I0831 09:48:24.142941 139540727588608 logging_writer.py:48] [12200] global_step=12200, grad_norm=2.3745133876800537, loss=1.6723917722702026 +I0831 09:48:51.354018 139540719195904 logging_writer.py:48] [12300] global_step=12300, grad_norm=2.9911558628082275, loss=1.6431628465652466 +I0831 09:49:18.466163 139540727588608 logging_writer.py:48] [12400] global_step=12400, grad_norm=2.4188966751098633, loss=1.7261488437652588 +I0831 09:49:45.775169 139540719195904 logging_writer.py:48] [12500] global_step=12500, grad_norm=2.525672435760498, loss=1.680429458618164 +I0831 09:50:12.946755 139540727588608 logging_writer.py:48] [12600] global_step=12600, grad_norm=2.2774715423583984, loss=1.749212384223938 +I0831 09:50:40.058662 139540719195904 logging_writer.py:48] [12700] global_step=12700, grad_norm=2.005841016769409, loss=1.6487700939178467 +I0831 09:51:07.136183 139540727588608 logging_writer.py:48] [12800] global_step=12800, grad_norm=2.4042255878448486, loss=1.798248291015625 +I0831 09:51:34.289035 139540719195904 logging_writer.py:48] [12900] global_step=12900, grad_norm=2.135181188583374, loss=1.6885521411895752 +I0831 09:52:01.390072 139540727588608 logging_writer.py:48] [13000] global_step=13000, grad_norm=2.6341793537139893, loss=1.670894742012024 +I0831 09:52:28.562095 139540719195904 logging_writer.py:48] [13100] global_step=13100, grad_norm=2.189192295074463, loss=1.5263959169387817 +I0831 09:52:55.703136 139540727588608 logging_writer.py:48] [13200] global_step=13200, grad_norm=2.112290382385254, loss=1.6569252014160156 +I0831 09:53:22.840547 139540719195904 logging_writer.py:48] [13300] global_step=13300, grad_norm=2.6336147785186768, loss=1.635023593902588 +I0831 09:53:49.956808 139540727588608 logging_writer.py:48] [13400] global_step=13400, grad_norm=2.468625068664551, loss=1.568206548690796 +I0831 09:54:17.143648 139540719195904 logging_writer.py:48] [13500] global_step=13500, grad_norm=2.2449703216552734, loss=1.5940901041030884 +I0831 09:54:44.496119 139540727588608 logging_writer.py:48] [13600] global_step=13600, grad_norm=2.6835951805114746, loss=1.667227029800415 +I0831 09:55:11.605710 139540719195904 logging_writer.py:48] [13700] global_step=13700, grad_norm=2.4242160320281982, loss=1.6533634662628174 +I0831 09:55:22.103456 139757377230016 spec.py:333] Evaluating on the training split. +I0831 09:55:33.517985 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 09:55:43.287672 139757377230016 spec.py:363] Evaluating on the test split. +I0831 09:55:44.160509 139757377230016 submission_runner.py:516] Time since start: 4154.49s, Step: 13740, {'train/accuracy': Array(0.75713485, dtype=float32), 'train/loss': Array(0.9526457, dtype=float32), 'validation/accuracy': Array(0.66587996, dtype=float32), 'validation/loss': Array(1.383565, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.5343, dtype=float32), 'test/loss': Array(2.1114912, dtype=float32), 'test/num_examples': 10000, 'score': 4045.7496836185455, 'total_duration': 4154.4920818805695, 'accumulated_submission_time': 4045.7496836185455, 'accumulated_eval_time': 108.49990367889404, 'accumulated_logging_time': 0.1349964141845703} +I0831 09:55:44.198232 139540727588608 logging_writer.py:48] [13740] accumulated_eval_time=108.5, accumulated_logging_time=0.134996, accumulated_submission_time=4045.75, global_step=13740, preemption_count=0, score=4045.75, test/accuracy=0.5343000292778015, test/loss=2.1114912033081055, test/num_examples=10000, total_duration=4154.49, train/accuracy=0.7571348547935486, train/loss=0.9526457190513611, validation/accuracy=0.6658799648284912, validation/loss=1.3835649490356445, validation/num_examples=50000 +I0831 09:56:00.866657 139540727588608 logging_writer.py:48] [13800] global_step=13800, grad_norm=2.9247355461120605, loss=1.6157665252685547 +I0831 09:56:27.978312 139540719195904 logging_writer.py:48] [13900] global_step=13900, grad_norm=2.487575054168701, loss=1.7263990640640259 +I0831 09:56:55.058404 139540727588608 logging_writer.py:48] [14000] global_step=14000, grad_norm=2.314270496368408, loss=1.5998975038528442 +I0831 09:57:22.223715 139540719195904 logging_writer.py:48] [14100] global_step=14100, grad_norm=2.5474886894226074, loss=1.5780583620071411 +I0831 09:57:49.315496 139540727588608 logging_writer.py:48] [14200] global_step=14200, grad_norm=2.240269660949707, loss=1.6310772895812988 +I0831 09:58:16.403185 139540719195904 logging_writer.py:48] [14300] global_step=14300, grad_norm=2.3761141300201416, loss=1.6258517503738403 +I0831 09:58:43.572210 139540727588608 logging_writer.py:48] [14400] global_step=14400, grad_norm=2.4127097129821777, loss=1.5797104835510254 +I0831 09:59:10.675673 139540719195904 logging_writer.py:48] [14500] global_step=14500, grad_norm=2.6421759128570557, loss=1.604823112487793 +I0831 09:59:37.766547 139540727588608 logging_writer.py:48] [14600] global_step=14600, grad_norm=2.765028953552246, loss=1.6098766326904297 +I0831 10:00:05.133627 139540719195904 logging_writer.py:48] [14700] global_step=14700, grad_norm=2.5833792686462402, loss=1.6327356100082397 +I0831 10:00:32.238925 139540727588608 logging_writer.py:48] [14800] global_step=14800, grad_norm=2.3662049770355225, loss=1.5503499507904053 +I0831 10:00:59.347378 139540719195904 logging_writer.py:48] [14900] global_step=14900, grad_norm=2.9328174591064453, loss=1.5539522171020508 +I0831 10:01:26.505471 139540727588608 logging_writer.py:48] [15000] global_step=15000, grad_norm=2.5427913665771484, loss=1.5619895458221436 +I0831 10:01:53.632505 139540719195904 logging_writer.py:48] [15100] global_step=15100, grad_norm=2.7473244667053223, loss=1.6411395072937012 +I0831 10:02:20.736697 139540727588608 logging_writer.py:48] [15200] global_step=15200, grad_norm=2.8432211875915527, loss=1.5915170907974243 +I0831 10:02:47.899092 139540719195904 logging_writer.py:48] [15300] global_step=15300, grad_norm=2.464256525039673, loss=1.650564432144165 +I0831 10:03:15.010691 139540727588608 logging_writer.py:48] [15400] global_step=15400, grad_norm=2.587932586669922, loss=1.6801741123199463 +I0831 10:03:42.126036 139540719195904 logging_writer.py:48] [15500] global_step=15500, grad_norm=2.929628610610962, loss=1.7180150747299194 +I0831 10:04:09.276400 139540727588608 logging_writer.py:48] [15600] global_step=15600, grad_norm=2.6715075969696045, loss=1.587929368019104 +I0831 10:04:36.591128 139540719195904 logging_writer.py:48] [15700] global_step=15700, grad_norm=2.594147205352783, loss=1.510210633277893 +I0831 10:05:03.711326 139540727588608 logging_writer.py:48] [15800] global_step=15800, grad_norm=2.56847882270813, loss=1.664489984512329 +I0831 10:05:30.880602 139540719195904 logging_writer.py:48] [15900] global_step=15900, grad_norm=2.9320390224456787, loss=1.5827500820159912 +I0831 10:05:57.989542 139540727588608 logging_writer.py:48] [16000] global_step=16000, grad_norm=2.91140079498291, loss=1.6061723232269287 +I0831 10:06:25.110658 139540719195904 logging_writer.py:48] [16100] global_step=16100, grad_norm=3.0662012100219727, loss=1.5521304607391357 +I0831 10:06:52.277197 139540727588608 logging_writer.py:48] [16200] global_step=16200, grad_norm=2.7371127605438232, loss=1.5358161926269531 +I0831 10:07:19.396139 139540719195904 logging_writer.py:48] [16300] global_step=16300, grad_norm=2.9295170307159424, loss=1.5253418684005737 +I0831 10:07:46.501785 139540727588608 logging_writer.py:48] [16400] global_step=16400, grad_norm=2.804436683654785, loss=1.4761290550231934 +I0831 10:08:13.676750 139540719195904 logging_writer.py:48] [16500] global_step=16500, grad_norm=3.475895643234253, loss=1.6188033819198608 +I0831 10:08:40.780055 139540727588608 logging_writer.py:48] [16600] global_step=16600, grad_norm=2.843618631362915, loss=1.440528154373169 +I0831 10:09:07.872628 139540719195904 logging_writer.py:48] [16700] global_step=16700, grad_norm=2.9456183910369873, loss=1.6323065757751465 +I0831 10:09:35.224353 139540727588608 logging_writer.py:48] [16800] global_step=16800, grad_norm=3.1221704483032227, loss=1.5898852348327637 +I0831 10:10:02.321914 139540719195904 logging_writer.py:48] [16900] global_step=16900, grad_norm=3.4443001747131348, loss=1.6145122051239014 +I0831 10:10:29.415445 139540727588608 logging_writer.py:48] [17000] global_step=17000, grad_norm=3.114633083343506, loss=1.609286904335022 +I0831 10:10:56.582131 139540719195904 logging_writer.py:48] [17100] global_step=17100, grad_norm=2.8715944290161133, loss=1.415501594543457 +I0831 10:11:23.675252 139540727588608 logging_writer.py:48] [17200] global_step=17200, grad_norm=3.337390661239624, loss=1.5409317016601562 +I0831 10:11:50.795716 139540719195904 logging_writer.py:48] [17300] global_step=17300, grad_norm=2.9945621490478516, loss=1.453870177268982 +I0831 10:12:17.974304 139540727588608 logging_writer.py:48] [17400] global_step=17400, grad_norm=3.0368294715881348, loss=1.4843043088912964 +I0831 10:12:45.067491 139540719195904 logging_writer.py:48] [17500] global_step=17500, grad_norm=3.077991247177124, loss=1.4751956462860107 +I0831 10:13:12.227694 139540727588608 logging_writer.py:48] [17600] global_step=17600, grad_norm=3.211374282836914, loss=1.5701603889465332 +I0831 10:13:39.401212 139540719195904 logging_writer.py:48] [17700] global_step=17700, grad_norm=3.43701434135437, loss=1.6494673490524292 +I0831 10:14:06.650123 139540727588608 logging_writer.py:48] [17800] global_step=17800, grad_norm=3.425670623779297, loss=1.5263237953186035 +I0831 10:14:33.740127 139540719195904 logging_writer.py:48] [17900] global_step=17900, grad_norm=3.518460512161255, loss=1.5226722955703735 +I0831 10:15:00.892085 139540727588608 logging_writer.py:48] [18000] global_step=18000, grad_norm=3.443199872970581, loss=1.5015571117401123 +I0831 10:15:28.006851 139540719195904 logging_writer.py:48] [18100] global_step=18100, grad_norm=3.6929986476898193, loss=1.4996833801269531 +I0831 10:15:55.123241 139540727588608 logging_writer.py:48] [18200] global_step=18200, grad_norm=3.5989232063293457, loss=1.419275164604187 +I0831 10:16:22.295714 139540719195904 logging_writer.py:48] [18300] global_step=18300, grad_norm=3.999492883682251, loss=1.5664268732070923 +I0831 10:16:49.395874 139540727588608 logging_writer.py:48] [18400] global_step=18400, grad_norm=4.027578353881836, loss=1.5067591667175293 +I0831 10:17:16.497887 139540719195904 logging_writer.py:48] [18500] global_step=18500, grad_norm=3.6771106719970703, loss=1.4374092817306519 +I0831 10:17:43.733538 139540727588608 logging_writer.py:48] [18600] global_step=18600, grad_norm=4.272151947021484, loss=1.6163733005523682 +I0831 10:18:10.834214 139540719195904 logging_writer.py:48] [18700] global_step=18700, grad_norm=4.046473979949951, loss=1.5734844207763672 +I0831 10:18:37.959108 139540727588608 logging_writer.py:48] [18800] global_step=18800, grad_norm=4.362948417663574, loss=1.5825697183609009 +I0831 10:19:05.368608 139540719195904 logging_writer.py:48] [18900] global_step=18900, grad_norm=4.670261383056641, loss=1.452459692955017 +I0831 10:19:32.503671 139540727588608 logging_writer.py:48] [19000] global_step=19000, grad_norm=4.418466567993164, loss=1.5249208211898804 +I0831 10:19:59.633135 139540719195904 logging_writer.py:48] [19100] global_step=19100, grad_norm=4.4973931312561035, loss=1.42304527759552 +I0831 10:20:26.787700 139540727588608 logging_writer.py:48] [19200] global_step=19200, grad_norm=4.5855631828308105, loss=1.4709618091583252 +I0831 10:20:53.882232 139540719195904 logging_writer.py:48] [19300] global_step=19300, grad_norm=5.289451599121094, loss=1.6386706829071045 +I0831 10:21:21.004482 139540727588608 logging_writer.py:48] [19400] global_step=19400, grad_norm=4.795241832733154, loss=1.467435598373413 +I0831 10:21:48.175781 139540719195904 logging_writer.py:48] [19500] global_step=19500, grad_norm=5.48272180557251, loss=1.4683492183685303 +I0831 10:22:15.296750 139540727588608 logging_writer.py:48] [19600] global_step=19600, grad_norm=5.387566089630127, loss=1.4379559755325317 +I0831 10:22:42.413926 139540719195904 logging_writer.py:48] [19700] global_step=19700, grad_norm=5.492058277130127, loss=1.4219332933425903 +I0831 10:23:09.587232 139540727588608 logging_writer.py:48] [19800] global_step=19800, grad_norm=5.954037189483643, loss=1.548330307006836 +I0831 10:23:36.874991 139540719195904 logging_writer.py:48] [19900] global_step=19900, grad_norm=6.324367046356201, loss=1.49332857131958 +I0831 10:24:04.030058 139540727588608 logging_writer.py:48] [20000] global_step=20000, grad_norm=6.676005840301514, loss=1.525114893913269 +I0831 10:24:31.201773 139540719195904 logging_writer.py:48] [20100] global_step=20100, grad_norm=7.180425643920898, loss=1.6376086473464966 +I0831 10:24:58.312022 139540727588608 logging_writer.py:48] [20200] global_step=20200, grad_norm=6.891312599182129, loss=1.5392935276031494 +I0831 10:25:25.428666 139540719195904 logging_writer.py:48] [20300] global_step=20300, grad_norm=6.966064453125, loss=1.477125644683838 +I0831 10:25:52.586144 139540727588608 logging_writer.py:48] [20400] global_step=20400, grad_norm=7.211143493652344, loss=1.5515682697296143 +I0831 10:26:19.673496 139540719195904 logging_writer.py:48] [20500] global_step=20500, grad_norm=7.457107067108154, loss=1.4731723070144653 +I0831 10:26:46.818024 139540727588608 logging_writer.py:48] [20600] global_step=20600, grad_norm=7.430365562438965, loss=1.4786523580551147 +I0831 10:27:14.032456 139540719195904 logging_writer.py:48] [20700] global_step=20700, grad_norm=7.629820823669434, loss=1.4592504501342773 +I0831 10:27:41.139577 139540727588608 logging_writer.py:48] [20800] global_step=20800, grad_norm=8.020670890808105, loss=1.4321132898330688 +I0831 10:28:08.270686 139540719195904 logging_writer.py:48] [20900] global_step=20900, grad_norm=8.064123153686523, loss=1.4215633869171143 +I0831 10:28:35.673534 139540727588608 logging_writer.py:48] [21000] global_step=21000, grad_norm=8.3737211227417, loss=1.435939073562622 +I0831 10:29:00.189251 139757377230016 spec.py:333] Evaluating on the training split. +I0831 10:29:11.464420 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 10:29:22.143092 139757377230016 spec.py:363] Evaluating on the test split. +I0831 10:29:23.018645 139757377230016 submission_runner.py:516] Time since start: 6173.35s, Step: 21092, {'train/accuracy': Array(0.80626196, dtype=float32), 'train/loss': Array(0.7468492, dtype=float32), 'validation/accuracy': Array(0.69604, dtype=float32), 'validation/loss': Array(1.2423426, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.5665, dtype=float32), 'test/loss': Array(1.966288, dtype=float32), 'test/num_examples': 10000, 'score': 6041.659068584442, 'total_duration': 6173.354122161865, 'accumulated_submission_time': 6041.659068584442, 'accumulated_eval_time': 131.3270001411438, 'accumulated_logging_time': 0.20162391662597656} +I0831 10:29:23.054018 139540719195904 logging_writer.py:48] [21092] accumulated_eval_time=131.327, accumulated_logging_time=0.201624, accumulated_submission_time=6041.66, global_step=21092, preemption_count=0, score=6041.66, test/accuracy=0.5665000081062317, test/loss=1.9662879705429077, test/num_examples=10000, total_duration=6173.35, train/accuracy=0.8062619566917419, train/loss=0.7468491792678833, validation/accuracy=0.6960399746894836, validation/loss=1.2423425912857056, validation/num_examples=50000 +I0831 10:29:25.682582 139540727588608 logging_writer.py:48] [21100] global_step=21100, grad_norm=9.557382583618164, loss=1.4255931377410889 +I0831 10:29:52.888407 139540719195904 logging_writer.py:48] [21200] global_step=21200, grad_norm=9.634325981140137, loss=1.5469609498977661 +I0831 10:30:20.019183 139540727588608 logging_writer.py:48] [21300] global_step=21300, grad_norm=9.685497283935547, loss=1.6118385791778564 +I0831 10:30:47.111061 139540719195904 logging_writer.py:48] [21400] global_step=21400, grad_norm=9.886861801147461, loss=1.5240147113800049 +I0831 10:31:14.244115 139540727588608 logging_writer.py:48] [21500] global_step=21500, grad_norm=9.797523498535156, loss=1.4754592180252075 +I0831 10:31:41.443229 139540719195904 logging_writer.py:48] [21600] global_step=21600, grad_norm=10.45247745513916, loss=1.5948452949523926 +I0831 10:32:08.602532 139540727588608 logging_writer.py:48] [21700] global_step=21700, grad_norm=11.104936599731445, loss=1.4480738639831543 +I0831 10:32:35.705932 139540719195904 logging_writer.py:48] [21800] global_step=21800, grad_norm=11.44325065612793, loss=1.567525029182434 +I0831 10:33:02.876957 139540727588608 logging_writer.py:48] [21900] global_step=21900, grad_norm=11.143261909484863, loss=1.4766201972961426 +I0831 10:33:29.989309 139540719195904 logging_writer.py:48] [22000] global_step=22000, grad_norm=11.234718322753906, loss=1.4688124656677246 +I0831 10:33:57.296616 139540727588608 logging_writer.py:48] [22100] global_step=22100, grad_norm=11.976441383361816, loss=1.5527771711349487 +I0831 10:34:24.469956 139540719195904 logging_writer.py:48] [22200] global_step=22200, grad_norm=11.653144836425781, loss=1.5262765884399414 +I0831 10:34:51.597905 139540727588608 logging_writer.py:48] [22300] global_step=22300, grad_norm=12.515901565551758, loss=1.4616196155548096 +I0831 10:35:18.740513 139540719195904 logging_writer.py:48] [22400] global_step=22400, grad_norm=12.398321151733398, loss=1.5515942573547363 +I0831 10:35:45.931721 139540727588608 logging_writer.py:48] [22500] global_step=22500, grad_norm=11.966618537902832, loss=1.543088674545288 +I0831 10:36:13.057742 139540719195904 logging_writer.py:48] [22600] global_step=22600, grad_norm=12.671228408813477, loss=1.516204595565796 +I0831 10:36:40.145552 139540727588608 logging_writer.py:48] [22700] global_step=22700, grad_norm=12.06314754486084, loss=1.4373699426651 +I0831 10:37:07.319505 139540719195904 logging_writer.py:48] [22800] global_step=22800, grad_norm=12.62319278717041, loss=1.5195612907409668 +I0831 10:37:34.423801 139540727588608 logging_writer.py:48] [22900] global_step=22900, grad_norm=12.841315269470215, loss=1.5209558010101318 +I0831 10:38:01.535426 139540719195904 logging_writer.py:48] [23000] global_step=23000, grad_norm=12.794351577758789, loss=1.5264403820037842 +I0831 10:38:28.711149 139540727588608 logging_writer.py:48] [23100] global_step=23100, grad_norm=14.13135051727295, loss=1.5459315776824951 +I0831 10:38:56.041190 139540719195904 logging_writer.py:48] [23200] global_step=23200, grad_norm=13.100569725036621, loss=1.546830177307129 +I0831 10:39:23.128574 139540727588608 logging_writer.py:48] [23300] global_step=23300, grad_norm=13.003190040588379, loss=1.4513229131698608 +I0831 10:39:50.279952 139540719195904 logging_writer.py:48] [23400] global_step=23400, grad_norm=12.78990650177002, loss=1.4974379539489746 +I0831 10:40:17.371847 139540727588608 logging_writer.py:48] [23500] global_step=23500, grad_norm=13.094698905944824, loss=1.4613187313079834 +I0831 10:40:44.495095 139540719195904 logging_writer.py:48] [23600] global_step=23600, grad_norm=12.847939491271973, loss=1.6071805953979492 +I0831 10:41:11.667814 139540727588608 logging_writer.py:48] [23700] global_step=23700, grad_norm=13.217056274414062, loss=1.4518870115280151 +I0831 10:41:38.805086 139540719195904 logging_writer.py:48] [23800] global_step=23800, grad_norm=13.09594440460205, loss=1.5262119770050049 +I0831 10:42:05.895063 139540727588608 logging_writer.py:48] [23900] global_step=23900, grad_norm=12.245737075805664, loss=1.4961656332015991 +I0831 10:42:33.075735 139540719195904 logging_writer.py:48] [24000] global_step=24000, grad_norm=13.137687683105469, loss=1.4849153757095337 +I0831 10:43:00.164798 139540727588608 logging_writer.py:48] [24100] global_step=24100, grad_norm=12.375476837158203, loss=1.4814996719360352 +I0831 10:43:27.272675 139540719195904 logging_writer.py:48] [24200] global_step=24200, grad_norm=12.37205982208252, loss=1.4268970489501953 +I0831 10:43:54.646221 139540727588608 logging_writer.py:48] [24300] global_step=24300, grad_norm=12.036982536315918, loss=1.3881593942642212 +I0831 10:44:21.808842 139540719195904 logging_writer.py:48] [24400] global_step=24400, grad_norm=11.926409721374512, loss=1.4728968143463135 +I0831 10:44:48.927291 139540727588608 logging_writer.py:48] [24500] global_step=24500, grad_norm=11.771098136901855, loss=1.4304404258728027 +I0831 10:45:16.127245 139540719195904 logging_writer.py:48] [24600] global_step=24600, grad_norm=11.649035453796387, loss=1.5236186981201172 +I0831 10:45:43.227485 139540727588608 logging_writer.py:48] [24700] global_step=24700, grad_norm=11.907681465148926, loss=1.486174464225769 +I0831 10:46:10.313423 139540719195904 logging_writer.py:48] [24800] global_step=24800, grad_norm=10.682490348815918, loss=1.3228707313537598 +I0831 10:46:37.524952 139540727588608 logging_writer.py:48] [24900] global_step=24900, grad_norm=11.951889038085938, loss=1.578093409538269 +I0831 10:47:04.671285 139540719195904 logging_writer.py:48] [25000] global_step=25000, grad_norm=11.263921737670898, loss=1.515159249305725 +I0831 10:47:31.784247 139540727588608 logging_writer.py:48] [25100] global_step=25100, grad_norm=10.91555118560791, loss=1.4973716735839844 +I0831 10:47:59.002391 139540719195904 logging_writer.py:48] [25200] global_step=25200, grad_norm=11.098624229431152, loss=1.4847348928451538 +I0831 10:48:26.120682 139540727588608 logging_writer.py:48] [25300] global_step=25300, grad_norm=10.891446113586426, loss=1.4778203964233398 +I0831 10:48:53.457036 139540719195904 logging_writer.py:48] [25400] global_step=25400, grad_norm=10.664565086364746, loss=1.4708545207977295 +I0831 10:49:20.623734 139540727588608 logging_writer.py:48] [25500] global_step=25500, grad_norm=10.126751899719238, loss=1.411015272140503 +I0831 10:49:47.725658 139540719195904 logging_writer.py:48] [25600] global_step=25600, grad_norm=10.253349304199219, loss=1.3837159872055054 +I0831 10:50:14.813283 139540727588608 logging_writer.py:48] [25700] global_step=25700, grad_norm=9.889931678771973, loss=1.3123438358306885 +I0831 10:50:41.979097 139540719195904 logging_writer.py:48] [25800] global_step=25800, grad_norm=9.558504104614258, loss=1.494045615196228 +I0831 10:51:09.105386 139540727588608 logging_writer.py:48] [25900] global_step=25900, grad_norm=9.501413345336914, loss=1.4486650228500366 +I0831 10:51:36.266505 139540719195904 logging_writer.py:48] [26000] global_step=26000, grad_norm=9.698800086975098, loss=1.446463704109192 +I0831 10:52:03.436760 139540727588608 logging_writer.py:48] [26100] global_step=26100, grad_norm=9.408885955810547, loss=1.4476759433746338 +I0831 10:52:30.550791 139540719195904 logging_writer.py:48] [26200] global_step=26200, grad_norm=9.650245666503906, loss=1.3825984001159668 +I0831 10:52:57.635385 139540727588608 logging_writer.py:48] [26300] global_step=26300, grad_norm=9.363245010375977, loss=1.4882676601409912 +I0831 10:53:25.006984 139540719195904 logging_writer.py:48] [26400] global_step=26400, grad_norm=9.1680908203125, loss=1.5331602096557617 +I0831 10:53:52.119494 139540727588608 logging_writer.py:48] [26500] global_step=26500, grad_norm=9.169416427612305, loss=1.4329626560211182 +I0831 10:54:19.210794 139540719195904 logging_writer.py:48] [26600] global_step=26600, grad_norm=9.049760818481445, loss=1.4842700958251953 +I0831 10:54:46.384382 139540727588608 logging_writer.py:48] [26700] global_step=26700, grad_norm=8.756246566772461, loss=1.4040155410766602 +I0831 10:55:13.496634 139540719195904 logging_writer.py:48] [26800] global_step=26800, grad_norm=9.146037101745605, loss=1.5803797245025635 +I0831 10:55:40.642151 139540727588608 logging_writer.py:48] [26900] global_step=26900, grad_norm=8.69566822052002, loss=1.4477241039276123 +I0831 10:56:07.823172 139540719195904 logging_writer.py:48] [27000] global_step=27000, grad_norm=8.426907539367676, loss=1.4665682315826416 +I0831 10:56:34.938295 139540727588608 logging_writer.py:48] [27100] global_step=27100, grad_norm=8.21655559539795, loss=1.4632482528686523 +I0831 10:57:02.058788 139540719195904 logging_writer.py:48] [27200] global_step=27200, grad_norm=8.094533920288086, loss=1.3652199506759644 +I0831 10:57:29.236750 139540727588608 logging_writer.py:48] [27300] global_step=27300, grad_norm=8.046775817871094, loss=1.3511769771575928 +I0831 10:57:56.364717 139540719195904 logging_writer.py:48] [27400] global_step=27400, grad_norm=7.874637126922607, loss=1.3922624588012695 +I0831 10:58:23.691699 139540727588608 logging_writer.py:48] [27500] global_step=27500, grad_norm=7.043196678161621, loss=1.3832042217254639 +I0831 10:58:50.852096 139540719195904 logging_writer.py:48] [27600] global_step=27600, grad_norm=7.395362854003906, loss=1.451237440109253 +I0831 10:59:17.940412 139540727588608 logging_writer.py:48] [27700] global_step=27700, grad_norm=7.603185653686523, loss=1.4768431186676025 +I0831 10:59:45.060351 139540719195904 logging_writer.py:48] [27800] global_step=27800, grad_norm=7.7160820960998535, loss=1.4309496879577637 +I0831 11:00:12.193082 139540727588608 logging_writer.py:48] [27900] global_step=27900, grad_norm=7.284078121185303, loss=1.3665826320648193 +I0831 11:00:39.278562 139540719195904 logging_writer.py:48] [28000] global_step=28000, grad_norm=7.05027961730957, loss=1.332709789276123 +I0831 11:01:06.407034 139540727588608 logging_writer.py:48] [28100] global_step=28100, grad_norm=6.550110816955566, loss=1.2366924285888672 +I0831 11:01:33.594225 139540719195904 logging_writer.py:48] [28200] global_step=28200, grad_norm=6.895545959472656, loss=1.3160228729248047 +I0831 11:02:00.697011 139540727588608 logging_writer.py:48] [28300] global_step=28300, grad_norm=7.538845062255859, loss=1.4293622970581055 +I0831 11:02:27.807030 139540719195904 logging_writer.py:48] [28400] global_step=28400, grad_norm=6.912667274475098, loss=1.3876631259918213 +I0831 11:02:39.064034 139757377230016 spec.py:333] Evaluating on the training split. +I0831 11:02:50.022603 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 11:02:59.738879 139757377230016 spec.py:363] Evaluating on the test split. +I0831 11:03:00.607123 139757377230016 submission_runner.py:516] Time since start: 8190.94s, Step: 28443, {'train/accuracy': Array(0.82924104, dtype=float32), 'train/loss': Array(0.63326895, dtype=float32), 'validation/accuracy': Array(0.71349996, dtype=float32), 'validation/loss': Array(1.1654677, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.58580005, dtype=float32), 'test/loss': Array(1.8870754, dtype=float32), 'test/num_examples': 10000, 'score': 8037.608729124069, 'total_duration': 8190.942818403244, 'accumulated_submission_time': 8037.608729124069, 'accumulated_eval_time': 152.86800980567932, 'accumulated_logging_time': 0.2448570728302002} +I0831 11:03:00.648617 139540727588608 logging_writer.py:48] [28443] accumulated_eval_time=152.868, accumulated_logging_time=0.244857, accumulated_submission_time=8037.61, global_step=28443, preemption_count=0, score=8037.61, test/accuracy=0.585800051689148, test/loss=1.887075424194336, test/num_examples=10000, total_duration=8190.94, train/accuracy=0.8292410373687744, train/loss=0.6332689523696899, validation/accuracy=0.7134999632835388, validation/loss=1.1654677391052246, validation/num_examples=50000 +I0831 11:03:16.643344 139540719195904 logging_writer.py:48] [28500] global_step=28500, grad_norm=6.856855392456055, loss=1.334416151046753 +I0831 11:03:43.922574 139540727588608 logging_writer.py:48] [28600] global_step=28600, grad_norm=7.024263381958008, loss=1.4856795072555542 +I0831 11:04:10.983700 139540719195904 logging_writer.py:48] [28700] global_step=28700, grad_norm=6.6532206535339355, loss=1.4719828367233276 +I0831 11:04:38.165450 139540727588608 logging_writer.py:48] [28800] global_step=28800, grad_norm=6.259854316711426, loss=1.366386890411377 +I0831 11:05:05.261640 139540719195904 logging_writer.py:48] [28900] global_step=28900, grad_norm=6.5984015464782715, loss=1.3714625835418701 +I0831 11:05:32.394308 139540727588608 logging_writer.py:48] [29000] global_step=29000, grad_norm=6.480096340179443, loss=1.375211238861084 +I0831 11:05:59.558579 139540719195904 logging_writer.py:48] [29100] global_step=29100, grad_norm=6.551892280578613, loss=1.385138988494873 +I0831 11:06:26.686451 139540727588608 logging_writer.py:48] [29200] global_step=29200, grad_norm=6.235160827636719, loss=1.389631986618042 +I0831 11:06:53.816206 139540719195904 logging_writer.py:48] [29300] global_step=29300, grad_norm=6.632320404052734, loss=1.4930956363677979 +I0831 11:07:20.961253 139540727588608 logging_writer.py:48] [29400] global_step=29400, grad_norm=5.8751540184021, loss=1.32743239402771 +I0831 11:07:48.064663 139540719195904 logging_writer.py:48] [29500] global_step=29500, grad_norm=6.234432697296143, loss=1.3566339015960693 +I0831 11:08:15.318711 139540727588608 logging_writer.py:48] [29600] global_step=29600, grad_norm=6.388964653015137, loss=1.4252231121063232 +I0831 11:08:42.459514 139540719195904 logging_writer.py:48] [29700] global_step=29700, grad_norm=5.952165126800537, loss=1.3801355361938477 +I0831 11:09:09.547132 139540727588608 logging_writer.py:48] [29800] global_step=29800, grad_norm=6.346652030944824, loss=1.450993299484253 +I0831 11:09:36.635507 139540719195904 logging_writer.py:48] [29900] global_step=29900, grad_norm=5.932229995727539, loss=1.3031558990478516 +I0831 11:10:03.815738 139540727588608 logging_writer.py:48] [30000] global_step=30000, grad_norm=5.505983352661133, loss=1.1938716173171997 +I0831 11:10:30.924921 139540719195904 logging_writer.py:48] [30100] global_step=30100, grad_norm=5.755011558532715, loss=1.3663808107376099 +I0831 11:10:58.031605 139540727588608 logging_writer.py:48] [30200] global_step=30200, grad_norm=5.739498615264893, loss=1.3611278533935547 +I0831 11:11:25.180492 139540719195904 logging_writer.py:48] [30300] global_step=30300, grad_norm=5.950647354125977, loss=1.3464634418487549 +I0831 11:11:52.273781 139540727588608 logging_writer.py:48] [30400] global_step=30400, grad_norm=5.556553363800049, loss=1.3372281789779663 +I0831 11:12:19.374334 139540719195904 logging_writer.py:48] [30500] global_step=30500, grad_norm=5.97622013092041, loss=1.312080979347229 +I0831 11:12:46.539577 139540727588608 logging_writer.py:48] [30600] global_step=30600, grad_norm=5.419106960296631, loss=1.291137933731079 +I0831 11:13:13.843331 139540719195904 logging_writer.py:48] [30700] global_step=30700, grad_norm=5.577155113220215, loss=1.2933436632156372 +I0831 11:13:40.949983 139540727588608 logging_writer.py:48] [30800] global_step=30800, grad_norm=5.6605730056762695, loss=1.408380389213562 +I0831 11:14:08.107430 139540719195904 logging_writer.py:48] [30900] global_step=30900, grad_norm=5.130184650421143, loss=1.2073397636413574 +I0831 11:14:35.189364 139540727588608 logging_writer.py:48] [31000] global_step=31000, grad_norm=5.48284387588501, loss=1.416611671447754 +I0831 11:15:02.294664 139540719195904 logging_writer.py:48] [31100] global_step=31100, grad_norm=5.173871040344238, loss=1.2612526416778564 +I0831 11:15:29.477540 139540727588608 logging_writer.py:48] [31200] global_step=31200, grad_norm=5.459846496582031, loss=1.3581262826919556 +I0831 11:15:56.606070 139540719195904 logging_writer.py:48] [31300] global_step=31300, grad_norm=5.2973504066467285, loss=1.3610316514968872 +I0831 11:16:23.706544 139540727588608 logging_writer.py:48] [31400] global_step=31400, grad_norm=5.414575099945068, loss=1.3468780517578125 +I0831 11:16:50.884352 139540719195904 logging_writer.py:48] [31500] global_step=31500, grad_norm=5.036263942718506, loss=1.2264533042907715 +I0831 11:17:17.982735 139540727588608 logging_writer.py:48] [31600] global_step=31600, grad_norm=5.269961357116699, loss=1.3461639881134033 +I0831 11:17:45.122669 139540719195904 logging_writer.py:48] [31700] global_step=31700, grad_norm=5.440232276916504, loss=1.2917888164520264 +I0831 11:18:12.498792 139540727588608 logging_writer.py:48] [31800] global_step=31800, grad_norm=5.33486795425415, loss=1.3250828981399536 +I0831 11:18:39.628047 139540719195904 logging_writer.py:48] [31900] global_step=31900, grad_norm=5.132400989532471, loss=1.444732904434204 +I0831 11:19:06.713125 139540727588608 logging_writer.py:48] [32000] global_step=32000, grad_norm=5.207421779632568, loss=1.3273389339447021 +I0831 11:19:33.871004 139540719195904 logging_writer.py:48] [32100] global_step=32100, grad_norm=4.952977180480957, loss=1.2982454299926758 +I0831 11:20:00.968861 139540727588608 logging_writer.py:48] [32200] global_step=32200, grad_norm=4.778348445892334, loss=1.2664382457733154 +I0831 11:20:28.079846 139540719195904 logging_writer.py:48] [32300] global_step=32300, grad_norm=4.9364190101623535, loss=1.2475731372833252 +I0831 11:20:55.248096 139540727588608 logging_writer.py:48] [32400] global_step=32400, grad_norm=5.110878944396973, loss=1.384222388267517 +I0831 11:21:22.353686 139540719195904 logging_writer.py:48] [32500] global_step=32500, grad_norm=4.868292331695557, loss=1.2670761346817017 +I0831 11:21:49.469213 139540727588608 logging_writer.py:48] [32600] global_step=32600, grad_norm=5.067262649536133, loss=1.3139801025390625 +I0831 11:22:16.675864 139540719195904 logging_writer.py:48] [32700] global_step=32700, grad_norm=5.01043176651001, loss=1.3385252952575684 +I0831 11:22:44.024305 139540727588608 logging_writer.py:48] [32800] global_step=32800, grad_norm=4.581556797027588, loss=1.243595004081726 +I0831 11:23:11.117251 139540719195904 logging_writer.py:48] [32900] global_step=32900, grad_norm=5.027096271514893, loss=1.3686928749084473 +I0831 11:23:38.279569 139540727588608 logging_writer.py:48] [33000] global_step=33000, grad_norm=5.23756742477417, loss=1.3736300468444824 +I0831 11:24:05.356337 139540719195904 logging_writer.py:48] [33100] global_step=33100, grad_norm=4.798322677612305, loss=1.4210114479064941 +I0831 11:24:32.429423 139540727588608 logging_writer.py:48] [33200] global_step=33200, grad_norm=4.609452247619629, loss=1.3469798564910889 +I0831 11:24:59.637112 139540719195904 logging_writer.py:48] [33300] global_step=33300, grad_norm=4.531521320343018, loss=1.3296074867248535 +I0831 11:25:26.760348 139540727588608 logging_writer.py:48] [33400] global_step=33400, grad_norm=4.706089496612549, loss=1.3575105667114258 +I0831 11:25:53.855592 139540719195904 logging_writer.py:48] [33500] global_step=33500, grad_norm=4.7591447830200195, loss=1.4054676294326782 +I0831 11:26:21.020307 139540727588608 logging_writer.py:48] [33600] global_step=33600, grad_norm=4.575411796569824, loss=1.2726943492889404 +I0831 11:26:48.141492 139540719195904 logging_writer.py:48] [33700] global_step=33700, grad_norm=4.580747604370117, loss=1.2526943683624268 +I0831 11:27:15.248939 139540727588608 logging_writer.py:48] [33800] global_step=33800, grad_norm=4.494745254516602, loss=1.31716787815094 +I0831 11:27:42.623288 139540719195904 logging_writer.py:48] [33900] global_step=33900, grad_norm=4.5195183753967285, loss=1.2414218187332153 +I0831 11:28:09.714022 139540727588608 logging_writer.py:48] [34000] global_step=34000, grad_norm=4.661245822906494, loss=1.2922687530517578 +I0831 11:28:36.802186 139540719195904 logging_writer.py:48] [34100] global_step=34100, grad_norm=4.629753112792969, loss=1.3563445806503296 +I0831 11:29:03.961892 139540727588608 logging_writer.py:48] [34200] global_step=34200, grad_norm=4.885883808135986, loss=1.3018529415130615 +I0831 11:29:31.050252 139540719195904 logging_writer.py:48] [34300] global_step=34300, grad_norm=4.390519142150879, loss=1.2564465999603271 +I0831 11:29:58.313432 139540727588608 logging_writer.py:48] [34400] global_step=34400, grad_norm=4.209048271179199, loss=1.2834184169769287 +I0831 11:30:25.501081 139540719195904 logging_writer.py:48] [34500] global_step=34500, grad_norm=4.776627540588379, loss=1.2690743207931519 +I0831 11:30:52.639477 139540727588608 logging_writer.py:48] [34600] global_step=34600, grad_norm=4.430150985717773, loss=1.234391689300537 +I0831 11:31:19.734762 139540719195904 logging_writer.py:48] [34700] global_step=34700, grad_norm=4.621642112731934, loss=1.2821805477142334 +I0831 11:31:46.906653 139540727588608 logging_writer.py:48] [34800] global_step=34800, grad_norm=4.689769268035889, loss=1.3922069072723389 +I0831 11:32:14.270456 139540719195904 logging_writer.py:48] [34900] global_step=34900, grad_norm=4.467526912689209, loss=1.3079547882080078 +I0831 11:32:41.389118 139540727588608 logging_writer.py:48] [35000] global_step=35000, grad_norm=4.358373641967773, loss=1.3244608640670776 +I0831 11:33:08.547978 139540719195904 logging_writer.py:48] [35100] global_step=35100, grad_norm=4.333826541900635, loss=1.2724456787109375 +I0831 11:33:35.650083 139540727588608 logging_writer.py:48] [35200] global_step=35200, grad_norm=4.3171610832214355, loss=1.2240955829620361 +I0831 11:34:02.759742 139540719195904 logging_writer.py:48] [35300] global_step=35300, grad_norm=4.5857625007629395, loss=1.2443599700927734 +I0831 11:34:29.923554 139540727588608 logging_writer.py:48] [35400] global_step=35400, grad_norm=4.247714042663574, loss=1.3141417503356934 +I0831 11:34:57.002305 139540719195904 logging_writer.py:48] [35500] global_step=35500, grad_norm=4.356078624725342, loss=1.2321714162826538 +I0831 11:35:24.110615 139540727588608 logging_writer.py:48] [35600] global_step=35600, grad_norm=4.124521732330322, loss=1.188905119895935 +I0831 11:35:51.284962 139540719195904 logging_writer.py:48] [35700] global_step=35700, grad_norm=4.432794094085693, loss=1.2831470966339111 +I0831 11:36:16.824327 139757377230016 spec.py:333] Evaluating on the training split. +I0831 11:36:27.667526 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 11:36:38.080013 139757377230016 spec.py:363] Evaluating on the test split. +I0831 11:36:38.954117 139757377230016 submission_runner.py:516] Time since start: 10209.29s, Step: 35796, {'train/accuracy': Array(0.8547313, dtype=float32), 'train/loss': Array(0.5380873, dtype=float32), 'validation/accuracy': Array(0.72433996, dtype=float32), 'validation/loss': Array(1.1205891, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6003, dtype=float32), 'test/loss': Array(1.8511142, dtype=float32), 'test/num_examples': 10000, 'score': 10033.702872037888, 'total_duration': 10209.286883592606, 'accumulated_submission_time': 10033.702872037888, 'accumulated_eval_time': 174.9927954673767, 'accumulated_logging_time': 0.31510424613952637} +I0831 11:36:39.027848 139540727588608 logging_writer.py:48] [35796] accumulated_eval_time=174.993, accumulated_logging_time=0.315104, accumulated_submission_time=10033.7, global_step=35796, preemption_count=0, score=10033.7, test/accuracy=0.6003000140190125, test/loss=1.8511141538619995, test/num_examples=10000, total_duration=10209.3, train/accuracy=0.8547313213348389, train/loss=0.5380873084068298, validation/accuracy=0.7243399620056152, validation/loss=1.1205891370773315, validation/num_examples=50000 +I0831 11:36:40.532334 139540719195904 logging_writer.py:48] [35800] global_step=35800, grad_norm=4.7202630043029785, loss=1.3348482847213745 +I0831 11:37:07.685735 139540727588608 logging_writer.py:48] [35900] global_step=35900, grad_norm=4.244609832763672, loss=1.2421958446502686 +I0831 11:37:35.050549 139540719195904 logging_writer.py:48] [36000] global_step=36000, grad_norm=4.5500898361206055, loss=1.3328924179077148 +I0831 11:38:02.158855 139540727588608 logging_writer.py:48] [36100] global_step=36100, grad_norm=4.172214984893799, loss=1.2736306190490723 +I0831 11:38:29.302474 139540719195904 logging_writer.py:48] [36200] global_step=36200, grad_norm=4.507602691650391, loss=1.4264013767242432 +I0831 11:38:56.479331 139540727588608 logging_writer.py:48] [36300] global_step=36300, grad_norm=4.254380226135254, loss=1.2949914932250977 +I0831 11:39:23.598407 139540719195904 logging_writer.py:48] [36400] global_step=36400, grad_norm=4.281058311462402, loss=1.3259429931640625 +I0831 11:39:50.683337 139540727588608 logging_writer.py:48] [36500] global_step=36500, grad_norm=4.119572639465332, loss=1.1970545053482056 +I0831 11:40:17.838210 139540719195904 logging_writer.py:48] [36600] global_step=36600, grad_norm=4.170992374420166, loss=1.268265962600708 +I0831 11:40:44.934122 139540727588608 logging_writer.py:48] [36700] global_step=36700, grad_norm=4.217509746551514, loss=1.2188754081726074 +I0831 11:41:12.069585 139540719195904 logging_writer.py:48] [36800] global_step=36800, grad_norm=4.251855850219727, loss=1.283942461013794 +I0831 11:41:39.251447 139540727588608 logging_writer.py:48] [36900] global_step=36900, grad_norm=4.392795085906982, loss=1.274078369140625 +I0831 11:42:06.581817 139540719195904 logging_writer.py:48] [37000] global_step=37000, grad_norm=4.3447465896606445, loss=1.3322309255599976 +I0831 11:42:33.690864 139540727588608 logging_writer.py:48] [37100] global_step=37100, grad_norm=4.149899959564209, loss=1.227020502090454 +I0831 11:43:00.866869 139540719195904 logging_writer.py:48] [37200] global_step=37200, grad_norm=4.309436321258545, loss=1.271711826324463 +I0831 11:43:27.976668 139540727588608 logging_writer.py:48] [37300] global_step=37300, grad_norm=4.196052551269531, loss=1.426633596420288 +I0831 11:43:55.098424 139540719195904 logging_writer.py:48] [37400] global_step=37400, grad_norm=4.161314487457275, loss=1.2849328517913818 +I0831 11:44:22.333451 139540727588608 logging_writer.py:48] [37500] global_step=37500, grad_norm=3.8893167972564697, loss=1.2044968605041504 +I0831 11:44:49.422745 139540719195904 logging_writer.py:48] [37600] global_step=37600, grad_norm=4.477601051330566, loss=1.2356882095336914 +I0831 11:45:16.550478 139540727588608 logging_writer.py:48] [37700] global_step=37700, grad_norm=4.040524959564209, loss=1.260939359664917 +I0831 11:45:43.699750 139540719195904 logging_writer.py:48] [37800] global_step=37800, grad_norm=4.063314914703369, loss=1.3086317777633667 +I0831 11:46:10.815043 139540727588608 logging_writer.py:48] [37900] global_step=37900, grad_norm=4.1932854652404785, loss=1.2433574199676514 +I0831 11:46:37.924953 139540719195904 logging_writer.py:48] [38000] global_step=38000, grad_norm=4.101161003112793, loss=1.3293920755386353 +I0831 11:47:05.285003 139540727588608 logging_writer.py:48] [38100] global_step=38100, grad_norm=3.9735195636749268, loss=1.2837492227554321 +I0831 11:47:32.399626 139540719195904 logging_writer.py:48] [38200] global_step=38200, grad_norm=3.9089362621307373, loss=1.184288501739502 +I0831 11:47:59.498619 139540727588608 logging_writer.py:48] [38300] global_step=38300, grad_norm=4.22376823425293, loss=1.2818279266357422 +I0831 11:48:26.675985 139540719195904 logging_writer.py:48] [38400] global_step=38400, grad_norm=4.141959190368652, loss=1.2876839637756348 +I0831 11:48:53.777054 139540727588608 logging_writer.py:48] [38500] global_step=38500, grad_norm=4.308426380157471, loss=1.3033928871154785 +I0831 11:49:20.885647 139540719195904 logging_writer.py:48] [38600] global_step=38600, grad_norm=4.090238571166992, loss=1.2827363014221191 +I0831 11:49:48.062012 139540727588608 logging_writer.py:48] [38700] global_step=38700, grad_norm=4.012562274932861, loss=1.2983996868133545 +I0831 11:50:15.175712 139540719195904 logging_writer.py:48] [38800] global_step=38800, grad_norm=4.216050624847412, loss=1.3068649768829346 +I0831 11:50:42.306340 139540727588608 logging_writer.py:48] [38900] global_step=38900, grad_norm=3.934626579284668, loss=1.1754285097122192 +I0831 11:51:09.504460 139540719195904 logging_writer.py:48] [39000] global_step=39000, grad_norm=3.8839404582977295, loss=1.2506990432739258 +I0831 11:51:36.830153 139540727588608 logging_writer.py:48] [39100] global_step=39100, grad_norm=4.1025614738464355, loss=1.3252112865447998 +I0831 11:52:04.010630 139540719195904 logging_writer.py:48] [39200] global_step=39200, grad_norm=3.685434341430664, loss=1.1950883865356445 +I0831 11:52:31.164944 139540727588608 logging_writer.py:48] [39300] global_step=39300, grad_norm=3.9448537826538086, loss=1.2676030397415161 +I0831 11:52:58.283696 139540719195904 logging_writer.py:48] [39400] global_step=39400, grad_norm=4.269414901733398, loss=1.2683179378509521 +I0831 11:53:25.392218 139540727588608 logging_writer.py:48] [39500] global_step=39500, grad_norm=3.9312424659729004, loss=1.193228006362915 +I0831 11:53:52.568399 139540719195904 logging_writer.py:48] [39600] global_step=39600, grad_norm=4.028475761413574, loss=1.2327141761779785 +I0831 11:54:19.651744 139540727588608 logging_writer.py:48] [39700] global_step=39700, grad_norm=3.725679874420166, loss=1.2113289833068848 +I0831 11:54:46.779476 139540719195904 logging_writer.py:48] [39800] global_step=39800, grad_norm=3.8251304626464844, loss=1.2429364919662476 +I0831 11:55:14.000859 139540727588608 logging_writer.py:48] [39900] global_step=39900, grad_norm=3.7393782138824463, loss=1.2069926261901855 +I0831 11:55:41.106040 139540719195904 logging_writer.py:48] [40000] global_step=40000, grad_norm=4.148505210876465, loss=1.274507999420166 +I0831 11:56:08.243225 139540727588608 logging_writer.py:48] [40100] global_step=40100, grad_norm=4.037038803100586, loss=1.235325813293457 +I0831 11:56:35.636312 139540719195904 logging_writer.py:48] [40200] global_step=40200, grad_norm=3.896759510040283, loss=1.2146053314208984 +I0831 11:57:02.727898 139540727588608 logging_writer.py:48] [40300] global_step=40300, grad_norm=4.270120620727539, loss=1.3373217582702637 +I0831 11:57:29.840425 139540719195904 logging_writer.py:48] [40400] global_step=40400, grad_norm=3.673239231109619, loss=1.1482446193695068 +I0831 11:57:57.015678 139540727588608 logging_writer.py:48] [40500] global_step=40500, grad_norm=3.8309826850891113, loss=1.219042181968689 +I0831 11:58:24.101074 139540719195904 logging_writer.py:48] [40600] global_step=40600, grad_norm=4.034950256347656, loss=1.2846187353134155 +I0831 11:58:51.209084 139540727588608 logging_writer.py:48] [40700] global_step=40700, grad_norm=3.9151432514190674, loss=1.2480207681655884 +I0831 11:59:18.396447 139540719195904 logging_writer.py:48] [40800] global_step=40800, grad_norm=3.947235345840454, loss=1.2184481620788574 +I0831 11:59:45.537173 139540727588608 logging_writer.py:48] [40900] global_step=40900, grad_norm=3.8022079467773438, loss=1.154207706451416 +I0831 12:00:12.634277 139540719195904 logging_writer.py:48] [41000] global_step=41000, grad_norm=3.997093915939331, loss=1.250453233718872 +I0831 12:00:39.793740 139540727588608 logging_writer.py:48] [41100] global_step=41100, grad_norm=4.1447319984436035, loss=1.2411139011383057 +I0831 12:01:07.112896 139540719195904 logging_writer.py:48] [41200] global_step=41200, grad_norm=3.891000509262085, loss=1.324763536453247 +I0831 12:01:34.202043 139540727588608 logging_writer.py:48] [41300] global_step=41300, grad_norm=3.8678011894226074, loss=1.32218599319458 +I0831 12:02:01.341271 139540719195904 logging_writer.py:48] [41400] global_step=41400, grad_norm=3.790781021118164, loss=1.239753246307373 +I0831 12:02:28.480658 139540727588608 logging_writer.py:48] [41500] global_step=41500, grad_norm=3.8935763835906982, loss=1.3166730403900146 +I0831 12:02:55.579226 139540719195904 logging_writer.py:48] [41600] global_step=41600, grad_norm=3.819061279296875, loss=1.211667537689209 +I0831 12:03:22.732768 139540727588608 logging_writer.py:48] [41700] global_step=41700, grad_norm=4.105715274810791, loss=1.2219505310058594 +I0831 12:03:49.830610 139540719195904 logging_writer.py:48] [41800] global_step=41800, grad_norm=3.958251714706421, loss=1.2603569030761719 +I0831 12:04:16.931814 139540727588608 logging_writer.py:48] [41900] global_step=41900, grad_norm=3.997368335723877, loss=1.3004412651062012 +I0831 12:04:44.093357 139540719195904 logging_writer.py:48] [42000] global_step=42000, grad_norm=3.8785512447357178, loss=1.3307416439056396 +I0831 12:05:11.173076 139540727588608 logging_writer.py:48] [42100] global_step=42100, grad_norm=3.905547857284546, loss=1.239331603050232 +I0831 12:05:38.271915 139540719195904 logging_writer.py:48] [42200] global_step=42200, grad_norm=4.06660795211792, loss=1.2006256580352783 +I0831 12:06:05.643071 139540727588608 logging_writer.py:48] [42300] global_step=42300, grad_norm=3.8144569396972656, loss=1.227912187576294 +I0831 12:06:32.732711 139540719195904 logging_writer.py:48] [42400] global_step=42400, grad_norm=4.192952632904053, loss=1.171286702156067 +I0831 12:06:59.837933 139540727588608 logging_writer.py:48] [42500] global_step=42500, grad_norm=3.8545830249786377, loss=1.1758785247802734 +I0831 12:07:26.980470 139540719195904 logging_writer.py:48] [42600] global_step=42600, grad_norm=3.980395555496216, loss=1.202991008758545 +I0831 12:07:54.108528 139540727588608 logging_writer.py:48] [42700] global_step=42700, grad_norm=3.9088146686553955, loss=1.239029049873352 +I0831 12:08:21.208687 139540719195904 logging_writer.py:48] [42800] global_step=42800, grad_norm=4.031144142150879, loss=1.2118124961853027 +I0831 12:08:48.401324 139540727588608 logging_writer.py:48] [42900] global_step=42900, grad_norm=3.92457914352417, loss=1.161218285560608 +I0831 12:09:15.512140 139540719195904 logging_writer.py:48] [43000] global_step=43000, grad_norm=3.662536382675171, loss=1.1592092514038086 +I0831 12:09:42.625682 139540727588608 logging_writer.py:48] [43100] global_step=43100, grad_norm=4.054052829742432, loss=1.1741749048233032 +I0831 12:09:55.015928 139757377230016 spec.py:333] Evaluating on the training split. +I0831 12:10:05.534105 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 12:10:15.495197 139757377230016 spec.py:363] Evaluating on the test split. +I0831 12:10:16.380347 139757377230016 submission_runner.py:516] Time since start: 12226.72s, Step: 43147, {'train/accuracy': Array(0.871313, dtype=float32), 'train/loss': Array(0.46765828, dtype=float32), 'validation/accuracy': Array(0.72818, dtype=float32), 'validation/loss': Array(1.1009811, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6056, dtype=float32), 'test/loss': Array(1.8409466, dtype=float32), 'test/num_examples': 10000, 'score': 12029.607981204987, 'total_duration': 12226.716282129288, 'accumulated_submission_time': 12029.607981204987, 'accumulated_eval_time': 196.35537314414978, 'accumulated_logging_time': 0.4192686080932617} +I0831 12:10:16.430464 139540719195904 logging_writer.py:48] [43147] accumulated_eval_time=196.355, accumulated_logging_time=0.419269, accumulated_submission_time=12029.6, global_step=43147, preemption_count=0, score=12029.6, test/accuracy=0.6055999994277954, test/loss=1.8409465551376343, test/num_examples=10000, total_duration=12226.7, train/accuracy=0.8713129758834839, train/loss=0.46765828132629395, validation/accuracy=0.7281799912452698, validation/loss=1.1009811162948608, validation/num_examples=50000 +I0831 12:10:31.227133 139540727588608 logging_writer.py:48] [43200] global_step=43200, grad_norm=3.878725051879883, loss=1.196995496749878 +I0831 12:10:58.289913 139540719195904 logging_writer.py:48] [43300] global_step=43300, grad_norm=3.9285032749176025, loss=1.186333417892456 +I0831 12:11:25.560173 139540727588608 logging_writer.py:48] [43400] global_step=43400, grad_norm=3.9330484867095947, loss=1.210917353630066 +I0831 12:11:52.690369 139540719195904 logging_writer.py:48] [43500] global_step=43500, grad_norm=4.333321571350098, loss=1.2290897369384766 +I0831 12:12:19.781125 139540727588608 logging_writer.py:48] [43600] global_step=43600, grad_norm=4.000523090362549, loss=1.3087717294692993 +I0831 12:12:46.893382 139540719195904 logging_writer.py:48] [43700] global_step=43700, grad_norm=3.876587390899658, loss=1.2098255157470703 +I0831 12:13:14.054736 139540727588608 logging_writer.py:48] [43800] global_step=43800, grad_norm=3.9283854961395264, loss=1.3042746782302856 +I0831 12:13:41.162425 139540719195904 logging_writer.py:48] [43900] global_step=43900, grad_norm=3.8389394283294678, loss=1.2098431587219238 +I0831 12:14:08.254004 139540727588608 logging_writer.py:48] [44000] global_step=44000, grad_norm=3.8067798614501953, loss=1.1577210426330566 +I0831 12:14:35.416293 139540719195904 logging_writer.py:48] [44100] global_step=44100, grad_norm=3.965766191482544, loss=1.206108570098877 +I0831 12:15:02.498712 139540727588608 logging_writer.py:48] [44200] global_step=44200, grad_norm=4.240312099456787, loss=1.208968162536621 +I0831 12:15:29.598157 139540719195904 logging_writer.py:48] [44300] global_step=44300, grad_norm=3.58960223197937, loss=1.1802445650100708 +I0831 12:15:56.739521 139540727588608 logging_writer.py:48] [44400] global_step=44400, grad_norm=3.479884147644043, loss=1.0906481742858887 +I0831 12:16:24.022304 139540719195904 logging_writer.py:48] [44500] global_step=44500, grad_norm=3.727334976196289, loss=1.1844298839569092 +I0831 12:16:51.123938 139540727588608 logging_writer.py:48] [44600] global_step=44600, grad_norm=3.646181344985962, loss=1.1912580728530884 +I0831 12:17:18.282775 139540719195904 logging_writer.py:48] [44700] global_step=44700, grad_norm=4.152027606964111, loss=1.2284400463104248 +I0831 12:17:45.386218 139540727588608 logging_writer.py:48] [44800] global_step=44800, grad_norm=3.6859235763549805, loss=1.2416346073150635 +I0831 12:18:12.475402 139540719195904 logging_writer.py:48] [44900] global_step=44900, grad_norm=4.000932216644287, loss=1.2352441549301147 +I0831 12:18:39.655351 139540727588608 logging_writer.py:48] [45000] global_step=45000, grad_norm=4.428884506225586, loss=1.321946382522583 +I0831 12:19:06.746152 139540719195904 logging_writer.py:48] [45100] global_step=45100, grad_norm=3.59700345993042, loss=1.1054151058197021 +I0831 12:19:33.895883 139540727588608 logging_writer.py:48] [45200] global_step=45200, grad_norm=3.581268548965454, loss=1.144853115081787 +I0831 12:20:01.068797 139540719195904 logging_writer.py:48] [45300] global_step=45300, grad_norm=4.20071268081665, loss=1.2616429328918457 +I0831 12:20:28.150524 139540727588608 logging_writer.py:48] [45400] global_step=45400, grad_norm=3.523730993270874, loss=1.1928092241287231 +I0831 12:20:55.478504 139540719195904 logging_writer.py:48] [45500] global_step=45500, grad_norm=3.867340564727783, loss=1.2411229610443115 +I0831 12:21:22.652992 139540727588608 logging_writer.py:48] [45600] global_step=45600, grad_norm=3.750519037246704, loss=1.2503901720046997 +I0831 12:21:49.737785 139540719195904 logging_writer.py:48] [45700] global_step=45700, grad_norm=3.8615806102752686, loss=1.176875114440918 +I0831 12:22:16.820963 139540727588608 logging_writer.py:48] [45800] global_step=45800, grad_norm=3.8128325939178467, loss=1.171188473701477 +I0831 12:22:43.981503 139540719195904 logging_writer.py:48] [45900] global_step=45900, grad_norm=3.7311201095581055, loss=1.2256351709365845 +I0831 12:23:11.052111 139540727588608 logging_writer.py:48] [46000] global_step=46000, grad_norm=3.6924757957458496, loss=1.1618521213531494 +I0831 12:23:38.172645 139540719195904 logging_writer.py:48] [46100] global_step=46100, grad_norm=3.80656099319458, loss=1.1767730712890625 +I0831 12:24:05.330730 139540727588608 logging_writer.py:48] [46200] global_step=46200, grad_norm=3.9275805950164795, loss=1.22255539894104 +I0831 12:24:32.421322 139540719195904 logging_writer.py:48] [46300] global_step=46300, grad_norm=3.907480001449585, loss=1.1760849952697754 +I0831 12:24:59.494417 139540727588608 logging_writer.py:48] [46400] global_step=46400, grad_norm=3.705510139465332, loss=1.1467382907867432 +I0831 12:25:26.691468 139540719195904 logging_writer.py:48] [46500] global_step=46500, grad_norm=3.757244348526001, loss=1.1796646118164062 +I0831 12:25:53.992217 139540727588608 logging_writer.py:48] [46600] global_step=46600, grad_norm=3.9603328704833984, loss=1.3142532110214233 +I0831 12:26:21.089591 139540719195904 logging_writer.py:48] [46700] global_step=46700, grad_norm=3.718492031097412, loss=1.335824966430664 +I0831 12:26:48.246034 139540727588608 logging_writer.py:48] [46800] global_step=46800, grad_norm=4.099957466125488, loss=1.2506394386291504 +I0831 12:27:15.351977 139540719195904 logging_writer.py:48] [46900] global_step=46900, grad_norm=3.6402297019958496, loss=1.1925692558288574 +I0831 12:27:42.466343 139540727588608 logging_writer.py:48] [47000] global_step=47000, grad_norm=3.858234167098999, loss=1.1832149028778076 +I0831 12:28:09.600526 139540719195904 logging_writer.py:48] [47100] global_step=47100, grad_norm=3.6134397983551025, loss=1.1921443939208984 +I0831 12:28:36.676377 139540727588608 logging_writer.py:48] [47200] global_step=47200, grad_norm=3.486380100250244, loss=1.1003851890563965 +I0831 12:29:03.788163 139540719195904 logging_writer.py:48] [47300] global_step=47300, grad_norm=3.9076380729675293, loss=1.1034411191940308 +I0831 12:29:30.944924 139540727588608 logging_writer.py:48] [47400] global_step=47400, grad_norm=3.609618902206421, loss=1.1048024892807007 +I0831 12:29:58.011694 139540719195904 logging_writer.py:48] [47500] global_step=47500, grad_norm=4.064023971557617, loss=1.2876694202423096 +I0831 12:30:25.169609 139540727588608 logging_writer.py:48] [47600] global_step=47600, grad_norm=3.8037753105163574, loss=1.2342020273208618 +I0831 12:30:52.542477 139540719195904 logging_writer.py:48] [47700] global_step=47700, grad_norm=3.840196371078491, loss=1.287545919418335 +I0831 12:31:19.626679 139540727588608 logging_writer.py:48] [47800] global_step=47800, grad_norm=3.74434757232666, loss=1.2515287399291992 +I0831 12:31:46.683367 139540719195904 logging_writer.py:48] [47900] global_step=47900, grad_norm=3.771639347076416, loss=1.1102864742279053 +I0831 12:32:13.859468 139540727588608 logging_writer.py:48] [48000] global_step=48000, grad_norm=3.796393871307373, loss=1.2691946029663086 +I0831 12:32:40.951302 139540719195904 logging_writer.py:48] [48100] global_step=48100, grad_norm=3.634566307067871, loss=1.2292393445968628 +I0831 12:33:08.045721 139540727588608 logging_writer.py:48] [48200] global_step=48200, grad_norm=3.944098711013794, loss=1.237140417098999 +I0831 12:33:35.236737 139540719195904 logging_writer.py:48] [48300] global_step=48300, grad_norm=3.700312376022339, loss=1.111405372619629 +I0831 12:34:02.332248 139540727588608 logging_writer.py:48] [48400] global_step=48400, grad_norm=3.8034729957580566, loss=1.2181332111358643 +I0831 12:34:29.432133 139540719195904 logging_writer.py:48] [48500] global_step=48500, grad_norm=3.853167772293091, loss=1.154242992401123 +I0831 12:34:56.587351 139540727588608 logging_writer.py:48] [48600] global_step=48600, grad_norm=3.830934762954712, loss=1.2079558372497559 +I0831 12:35:23.970284 139540719195904 logging_writer.py:48] [48700] global_step=48700, grad_norm=3.870253801345825, loss=1.1418249607086182 +I0831 12:35:51.055382 139540727588608 logging_writer.py:48] [48800] global_step=48800, grad_norm=3.627871513366699, loss=1.229339838027954 +I0831 12:36:18.192724 139540719195904 logging_writer.py:48] [48900] global_step=48900, grad_norm=3.80617356300354, loss=1.1667430400848389 +I0831 12:36:45.295702 139540727588608 logging_writer.py:48] [49000] global_step=49000, grad_norm=4.032338619232178, loss=1.2822518348693848 +I0831 12:37:12.382266 139540719195904 logging_writer.py:48] [49100] global_step=49100, grad_norm=4.05773401260376, loss=1.3354737758636475 +I0831 12:37:39.560772 139540727588608 logging_writer.py:48] [49200] global_step=49200, grad_norm=3.873044490814209, loss=1.2915927171707153 +I0831 12:38:06.652205 139540719195904 logging_writer.py:48] [49300] global_step=49300, grad_norm=3.5548551082611084, loss=1.0813219547271729 +I0831 12:38:33.773818 139540727588608 logging_writer.py:48] [49400] global_step=49400, grad_norm=3.8711817264556885, loss=1.2754942178726196 +I0831 12:39:00.926746 139540719195904 logging_writer.py:48] [49500] global_step=49500, grad_norm=3.7301387786865234, loss=1.2191672325134277 +I0831 12:39:28.020856 139540727588608 logging_writer.py:48] [49600] global_step=49600, grad_norm=3.7637341022491455, loss=1.1661289930343628 +I0831 12:39:55.104529 139540719195904 logging_writer.py:48] [49700] global_step=49700, grad_norm=3.755552291870117, loss=1.163621187210083 +I0831 12:40:22.506520 139540727588608 logging_writer.py:48] [49800] global_step=49800, grad_norm=4.094992637634277, loss=1.2108473777770996 +I0831 12:40:49.595893 139540719195904 logging_writer.py:48] [49900] global_step=49900, grad_norm=3.79190993309021, loss=1.192185640335083 +I0831 12:41:16.717260 139540727588608 logging_writer.py:48] [50000] global_step=50000, grad_norm=3.902987003326416, loss=1.2289087772369385 +I0831 12:41:43.882025 139540719195904 logging_writer.py:48] [50100] global_step=50100, grad_norm=3.8318843841552734, loss=1.2076377868652344 +I0831 12:42:10.980606 139540727588608 logging_writer.py:48] [50200] global_step=50200, grad_norm=3.9105801582336426, loss=1.2144370079040527 +I0831 12:42:38.063223 139540719195904 logging_writer.py:48] [50300] global_step=50300, grad_norm=4.017617225646973, loss=1.242980718612671 +I0831 12:43:05.241966 139540727588608 logging_writer.py:48] [50400] global_step=50400, grad_norm=4.053542613983154, loss=1.1752004623413086 +I0831 12:43:32.344485 139540719195904 logging_writer.py:48] [50500] global_step=50500, grad_norm=3.8965697288513184, loss=1.1887019872665405 +I0831 12:43:32.522243 139757377230016 spec.py:333] Evaluating on the training split. +I0831 12:43:42.076637 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 12:43:52.373755 139757377230016 spec.py:363] Evaluating on the test split. +I0831 12:43:53.266558 139757377230016 submission_runner.py:516] Time since start: 14243.60s, Step: 50502, {'train/accuracy': Array(0.8854432, dtype=float32), 'train/loss': Array(0.41532198, dtype=float32), 'validation/accuracy': Array(0.7328, dtype=float32), 'validation/loss': Array(1.0901387, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.61230004, dtype=float32), 'test/loss': Array(1.8371395, dtype=float32), 'test/num_examples': 10000, 'score': 14025.61386179924, 'total_duration': 14243.602019548416, 'accumulated_submission_time': 14025.61386179924, 'accumulated_eval_time': 217.09737825393677, 'accumulated_logging_time': 0.5025744438171387} +I0831 12:43:53.320607 139540727588608 logging_writer.py:48] [50502] accumulated_eval_time=217.097, accumulated_logging_time=0.502574, accumulated_submission_time=14025.6, global_step=50502, preemption_count=0, score=14025.6, test/accuracy=0.6123000383377075, test/loss=1.8371394872665405, test/num_examples=10000, total_duration=14243.6, train/accuracy=0.8854432106018066, train/loss=0.4153219759464264, validation/accuracy=0.7328000068664551, validation/loss=1.0901386737823486, validation/num_examples=50000 +I0831 12:44:20.255620 139540719195904 logging_writer.py:48] [50600] global_step=50600, grad_norm=3.8640666007995605, loss=1.1660219430923462 +I0831 12:44:47.406877 139540727588608 logging_writer.py:48] [50700] global_step=50700, grad_norm=4.028095722198486, loss=1.264628529548645 +I0831 12:45:14.758065 139540719195904 logging_writer.py:48] [50800] global_step=50800, grad_norm=3.800403594970703, loss=1.2110083103179932 +I0831 12:45:41.862168 139540727588608 logging_writer.py:48] [50900] global_step=50900, grad_norm=4.171287536621094, loss=1.2216659784317017 +I0831 12:46:09.009771 139540719195904 logging_writer.py:48] [51000] global_step=51000, grad_norm=3.679856300354004, loss=1.2483497858047485 +I0831 12:46:36.110512 139540727588608 logging_writer.py:48] [51100] global_step=51100, grad_norm=4.079906463623047, loss=1.2520487308502197 +I0831 12:47:03.228552 139540719195904 logging_writer.py:48] [51200] global_step=51200, grad_norm=3.7118425369262695, loss=1.1630172729492188 +I0831 12:47:30.383319 139540727588608 logging_writer.py:48] [51300] global_step=51300, grad_norm=3.9609529972076416, loss=1.1060419082641602 +I0831 12:47:57.492629 139540719195904 logging_writer.py:48] [51400] global_step=51400, grad_norm=3.825989246368408, loss=1.2153922319412231 +I0831 12:48:24.573240 139540727588608 logging_writer.py:48] [51500] global_step=51500, grad_norm=3.918325901031494, loss=1.1542541980743408 +I0831 12:48:51.774119 139540719195904 logging_writer.py:48] [51600] global_step=51600, grad_norm=3.7876622676849365, loss=1.2338354587554932 +I0831 12:49:18.890168 139540727588608 logging_writer.py:48] [51700] global_step=51700, grad_norm=3.6581404209136963, loss=1.1320767402648926 +I0831 12:49:45.987556 139540719195904 logging_writer.py:48] [51800] global_step=51800, grad_norm=3.8201615810394287, loss=1.064109444618225 +I0831 12:50:13.366817 139540727588608 logging_writer.py:48] [51900] global_step=51900, grad_norm=4.051779270172119, loss=1.162381649017334 +I0831 12:50:40.485703 139540719195904 logging_writer.py:48] [52000] global_step=52000, grad_norm=4.00862979888916, loss=1.2615082263946533 +I0831 12:51:07.595764 139540727588608 logging_writer.py:48] [52100] global_step=52100, grad_norm=4.110744476318359, loss=1.1734048128128052 +I0831 12:51:34.763385 139540719195904 logging_writer.py:48] [52200] global_step=52200, grad_norm=4.136712074279785, loss=1.2130181789398193 +I0831 12:52:01.873131 139540727588608 logging_writer.py:48] [52300] global_step=52300, grad_norm=4.260636329650879, loss=1.2912178039550781 +I0831 12:52:29.033252 139540719195904 logging_writer.py:48] [52400] global_step=52400, grad_norm=3.800905227661133, loss=1.0987062454223633 +I0831 12:52:56.192470 139540727588608 logging_writer.py:48] [52500] global_step=52500, grad_norm=3.7023284435272217, loss=1.2087016105651855 +I0831 12:53:23.295372 139540719195904 logging_writer.py:48] [52600] global_step=52600, grad_norm=4.256481647491455, loss=1.2762980461120605 +I0831 12:53:50.393065 139540727588608 logging_writer.py:48] [52700] global_step=52700, grad_norm=3.839111089706421, loss=1.2562456130981445 +I0831 12:54:17.572826 139540719195904 logging_writer.py:48] [52800] global_step=52800, grad_norm=3.785646438598633, loss=1.1561301946640015 +I0831 12:54:44.922188 139540727588608 logging_writer.py:48] [52900] global_step=52900, grad_norm=3.926473379135132, loss=1.17299222946167 +I0831 12:55:12.018752 139540719195904 logging_writer.py:48] [53000] global_step=53000, grad_norm=4.021918773651123, loss=1.2395172119140625 +I0831 12:55:39.194250 139540727588608 logging_writer.py:48] [53100] global_step=53100, grad_norm=3.881983995437622, loss=1.0995144844055176 +I0831 12:56:06.292862 139540719195904 logging_writer.py:48] [53200] global_step=53200, grad_norm=4.120935916900635, loss=1.2103831768035889 +I0831 12:56:33.402919 139540727588608 logging_writer.py:48] [53300] global_step=53300, grad_norm=4.020929336547852, loss=1.311069369316101 +I0831 12:57:00.560995 139540719195904 logging_writer.py:48] [53400] global_step=53400, grad_norm=4.068737983703613, loss=1.1980650424957275 +I0831 12:57:27.669223 139540727588608 logging_writer.py:48] [53500] global_step=53500, grad_norm=3.921393871307373, loss=1.1122959852218628 +I0831 12:57:54.730311 139540719195904 logging_writer.py:48] [53600] global_step=53600, grad_norm=3.740293025970459, loss=1.145302653312683 +I0831 12:58:21.898789 139540727588608 logging_writer.py:48] [53700] global_step=53700, grad_norm=3.9754703044891357, loss=1.1674790382385254 +I0831 12:58:48.994636 139540719195904 logging_writer.py:48] [53800] global_step=53800, grad_norm=3.9218270778656006, loss=1.1551556587219238 +I0831 12:59:16.102398 139540727588608 logging_writer.py:48] [53900] global_step=53900, grad_norm=3.86102294921875, loss=1.2060766220092773 +I0831 12:59:43.485126 139540719195904 logging_writer.py:48] [54000] global_step=54000, grad_norm=3.702805519104004, loss=1.1905566453933716 +I0831 13:00:10.592657 139540727588608 logging_writer.py:48] [54100] global_step=54100, grad_norm=3.8269619941711426, loss=1.1583553552627563 +I0831 13:00:37.709592 139540719195904 logging_writer.py:48] [54200] global_step=54200, grad_norm=3.970186471939087, loss=1.1822278499603271 +I0831 13:01:04.897457 139540727588608 logging_writer.py:48] [54300] global_step=54300, grad_norm=3.6424434185028076, loss=1.0892633199691772 +I0831 13:01:32.020994 139540719195904 logging_writer.py:48] [54400] global_step=54400, grad_norm=4.111021041870117, loss=1.2418785095214844 +I0831 13:01:59.140436 139540727588608 logging_writer.py:48] [54500] global_step=54500, grad_norm=3.932718515396118, loss=1.1298924684524536 +I0831 13:02:26.308677 139540719195904 logging_writer.py:48] [54600] global_step=54600, grad_norm=3.884824752807617, loss=1.151043176651001 +I0831 13:02:53.403856 139540727588608 logging_writer.py:48] [54700] global_step=54700, grad_norm=3.979407548904419, loss=1.134543538093567 +I0831 13:03:20.488641 139540719195904 logging_writer.py:48] [54800] global_step=54800, grad_norm=4.484163284301758, loss=1.201373815536499 +I0831 13:03:47.676948 139540727588608 logging_writer.py:48] [54900] global_step=54900, grad_norm=3.740485668182373, loss=1.0867016315460205 +I0831 13:04:15.013966 139540719195904 logging_writer.py:48] [55000] global_step=55000, grad_norm=3.91239070892334, loss=1.223646879196167 +I0831 13:04:42.145814 139540727588608 logging_writer.py:48] [55100] global_step=55100, grad_norm=4.371642589569092, loss=1.2296147346496582 +I0831 13:05:09.316781 139540719195904 logging_writer.py:48] [55200] global_step=55200, grad_norm=3.7047321796417236, loss=1.147697925567627 +I0831 13:05:36.429258 139540727588608 logging_writer.py:48] [55300] global_step=55300, grad_norm=3.816472053527832, loss=1.2530736923217773 +I0831 13:06:03.515468 139540719195904 logging_writer.py:48] [55400] global_step=55400, grad_norm=3.7710280418395996, loss=1.1119556427001953 +I0831 13:06:30.703094 139540727588608 logging_writer.py:48] [55500] global_step=55500, grad_norm=3.895153522491455, loss=1.1224584579467773 +I0831 13:06:57.823288 139540719195904 logging_writer.py:48] [55600] global_step=55600, grad_norm=4.360355854034424, loss=1.230055570602417 +I0831 13:07:24.924249 139540727588608 logging_writer.py:48] [55700] global_step=55700, grad_norm=3.8989036083221436, loss=1.2129848003387451 +I0831 13:07:52.078378 139540719195904 logging_writer.py:48] [55800] global_step=55800, grad_norm=3.8862464427948, loss=1.1935923099517822 +I0831 13:08:19.193713 139540727588608 logging_writer.py:48] [55900] global_step=55900, grad_norm=3.786190986633301, loss=1.1556484699249268 +I0831 13:08:46.280599 139540719195904 logging_writer.py:48] [56000] global_step=56000, grad_norm=4.165676116943359, loss=1.213923692703247 +I0831 13:09:13.641499 139540727588608 logging_writer.py:48] [56100] global_step=56100, grad_norm=4.0112223625183105, loss=1.1515460014343262 +I0831 13:09:40.749574 139540719195904 logging_writer.py:48] [56200] global_step=56200, grad_norm=4.13128137588501, loss=1.2806158065795898 +I0831 13:10:07.843030 139540727588608 logging_writer.py:48] [56300] global_step=56300, grad_norm=3.8533663749694824, loss=1.0993503332138062 +I0831 13:10:34.990199 139540719195904 logging_writer.py:48] [56400] global_step=56400, grad_norm=4.043814182281494, loss=1.1235060691833496 +I0831 13:11:02.109277 139540727588608 logging_writer.py:48] [56500] global_step=56500, grad_norm=4.0894880294799805, loss=1.2926568984985352 +I0831 13:11:29.206444 139540719195904 logging_writer.py:48] [56600] global_step=56600, grad_norm=3.721463680267334, loss=1.2155168056488037 +I0831 13:11:56.382789 139540727588608 logging_writer.py:48] [56700] global_step=56700, grad_norm=3.8828091621398926, loss=1.141677737236023 +I0831 13:12:23.504581 139540719195904 logging_writer.py:48] [56800] global_step=56800, grad_norm=4.081435203552246, loss=1.201361894607544 +I0831 13:12:50.596766 139540727588608 logging_writer.py:48] [56900] global_step=56900, grad_norm=3.9855315685272217, loss=1.2298650741577148 +I0831 13:13:17.787920 139540719195904 logging_writer.py:48] [57000] global_step=57000, grad_norm=3.9699361324310303, loss=1.1900684833526611 +I0831 13:13:45.104512 139540727588608 logging_writer.py:48] [57100] global_step=57100, grad_norm=4.228451728820801, loss=1.2142601013183594 +I0831 13:14:12.220935 139540719195904 logging_writer.py:48] [57200] global_step=57200, grad_norm=4.07476282119751, loss=1.2213060855865479 +I0831 13:14:39.387118 139540727588608 logging_writer.py:48] [57300] global_step=57300, grad_norm=4.171145915985107, loss=1.1951603889465332 +I0831 13:15:06.498758 139540719195904 logging_writer.py:48] [57400] global_step=57400, grad_norm=3.9170422554016113, loss=1.1416330337524414 +I0831 13:15:33.612405 139540727588608 logging_writer.py:48] [57500] global_step=57500, grad_norm=3.879352569580078, loss=1.143683671951294 +I0831 13:16:00.746222 139540719195904 logging_writer.py:48] [57600] global_step=57600, grad_norm=4.299371242523193, loss=1.2836177349090576 +I0831 13:16:27.861755 139540727588608 logging_writer.py:48] [57700] global_step=57700, grad_norm=3.9988062381744385, loss=1.1603652238845825 +I0831 13:16:54.958745 139540719195904 logging_writer.py:48] [57800] global_step=57800, grad_norm=3.765183925628662, loss=1.1006163358688354 +I0831 13:17:09.468734 139757377230016 spec.py:333] Evaluating on the training split. +I0831 13:17:18.918017 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 13:17:28.801949 139757377230016 spec.py:363] Evaluating on the test split. +I0831 13:17:29.699454 139757377230016 submission_runner.py:516] Time since start: 16260.03s, Step: 57855, {'train/accuracy': Array(0.89253825, dtype=float32), 'train/loss': Array(0.3850234, dtype=float32), 'validation/accuracy': Array(0.73662, dtype=float32), 'validation/loss': Array(1.0837384, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.609, dtype=float32), 'test/loss': Array(1.840022, dtype=float32), 'test/num_examples': 10000, 'score': 16021.685701847076, 'total_duration': 16260.028512001038, 'accumulated_submission_time': 16021.685701847076, 'accumulated_eval_time': 237.31938290596008, 'accumulated_logging_time': 0.5800423622131348} +I0831 13:17:29.748150 139540727588608 logging_writer.py:48] [57855] accumulated_eval_time=237.319, accumulated_logging_time=0.580042, accumulated_submission_time=16021.7, global_step=57855, preemption_count=0, score=16021.7, test/accuracy=0.609000027179718, test/loss=1.8400219678878784, test/num_examples=10000, total_duration=16260, train/accuracy=0.8925382494926453, train/loss=0.3850233852863312, validation/accuracy=0.7366200089454651, validation/loss=1.0837384462356567, validation/num_examples=50000 +I0831 13:17:42.646984 139540727588608 logging_writer.py:48] [57900] global_step=57900, grad_norm=4.041118621826172, loss=1.2027671337127686 +I0831 13:18:09.735868 139540719195904 logging_writer.py:48] [58000] global_step=58000, grad_norm=4.164193630218506, loss=1.2515063285827637 +I0831 13:18:36.814246 139540727588608 logging_writer.py:48] [58100] global_step=58100, grad_norm=3.8585994243621826, loss=1.1270262002944946 +I0831 13:19:04.186744 139540719195904 logging_writer.py:48] [58200] global_step=58200, grad_norm=3.9867732524871826, loss=1.1456878185272217 +I0831 13:19:31.256504 139540727588608 logging_writer.py:48] [58300] global_step=58300, grad_norm=4.2866106033325195, loss=1.2606315612792969 +I0831 13:19:58.345583 139540719195904 logging_writer.py:48] [58400] global_step=58400, grad_norm=3.9092772006988525, loss=1.1484383344650269 +I0831 13:20:25.508113 139540727588608 logging_writer.py:48] [58500] global_step=58500, grad_norm=4.108815670013428, loss=1.1697138547897339 +I0831 13:20:52.639593 139540719195904 logging_writer.py:48] [58600] global_step=58600, grad_norm=4.694065093994141, loss=1.2007215023040771 +I0831 13:21:19.742998 139540727588608 logging_writer.py:48] [58700] global_step=58700, grad_norm=3.8600072860717773, loss=1.1486828327178955 +I0831 13:21:46.895214 139540719195904 logging_writer.py:48] [58800] global_step=58800, grad_norm=3.7497315406799316, loss=1.0931408405303955 +I0831 13:22:13.988959 139540727588608 logging_writer.py:48] [58900] global_step=58900, grad_norm=4.020484447479248, loss=1.245805263519287 +I0831 13:22:41.104136 139540719195904 logging_writer.py:48] [59000] global_step=59000, grad_norm=4.556707382202148, loss=1.228139042854309 +I0831 13:23:08.277716 139540727588608 logging_writer.py:48] [59100] global_step=59100, grad_norm=4.143918514251709, loss=1.0857951641082764 +I0831 13:23:35.559405 139540719195904 logging_writer.py:48] [59200] global_step=59200, grad_norm=4.044656753540039, loss=1.122304916381836 +I0831 13:24:02.639070 139540727588608 logging_writer.py:48] [59300] global_step=59300, grad_norm=4.29014778137207, loss=1.1408071517944336 +I0831 13:24:29.785770 139540719195904 logging_writer.py:48] [59400] global_step=59400, grad_norm=4.181191921234131, loss=1.2643784284591675 +I0831 13:24:56.883589 139540727588608 logging_writer.py:48] [59500] global_step=59500, grad_norm=4.127740859985352, loss=1.2178974151611328 +I0831 13:25:23.980302 139540719195904 logging_writer.py:48] [59600] global_step=59600, grad_norm=3.998202323913574, loss=1.0730668306350708 +I0831 13:25:51.155747 139540727588608 logging_writer.py:48] [59700] global_step=59700, grad_norm=4.511320114135742, loss=1.213517665863037 +I0831 13:26:18.266366 139540719195904 logging_writer.py:48] [59800] global_step=59800, grad_norm=4.17498254776001, loss=1.2108474969863892 +I0831 13:26:45.360296 139540727588608 logging_writer.py:48] [59900] global_step=59900, grad_norm=4.146834850311279, loss=1.1923496723175049 +I0831 13:27:12.503471 139540719195904 logging_writer.py:48] [60000] global_step=60000, grad_norm=3.7286126613616943, loss=1.1063026189804077 +I0831 13:27:39.628671 139540727588608 logging_writer.py:48] [60100] global_step=60100, grad_norm=4.054551601409912, loss=1.1324818134307861 +I0831 13:28:06.746386 139540719195904 logging_writer.py:48] [60200] global_step=60200, grad_norm=4.186339855194092, loss=1.1639115810394287 +I0831 13:28:34.121102 139540727588608 logging_writer.py:48] [60300] global_step=60300, grad_norm=4.027463436126709, loss=1.142842411994934 +I0831 13:29:01.242747 139540719195904 logging_writer.py:48] [60400] global_step=60400, grad_norm=4.097710609436035, loss=1.1667375564575195 +I0831 13:29:28.328428 139540727588608 logging_writer.py:48] [60500] global_step=60500, grad_norm=4.1667304039001465, loss=1.2456499338150024 +I0831 13:29:55.503316 139540719195904 logging_writer.py:48] [60600] global_step=60600, grad_norm=4.43645715713501, loss=1.1989213228225708 +I0831 13:30:22.594155 139540727588608 logging_writer.py:48] [60700] global_step=60700, grad_norm=4.265892028808594, loss=1.2106387615203857 +I0831 13:30:49.733455 139540719195904 logging_writer.py:48] [60800] global_step=60800, grad_norm=4.073123931884766, loss=1.114328384399414 +I0831 13:31:16.895074 139540727588608 logging_writer.py:48] [60900] global_step=60900, grad_norm=4.290305137634277, loss=1.1338751316070557 +I0831 13:31:43.997188 139540719195904 logging_writer.py:48] [61000] global_step=61000, grad_norm=4.189687728881836, loss=1.118222713470459 +I0831 13:32:11.077241 139540727588608 logging_writer.py:48] [61100] global_step=61100, grad_norm=4.0874505043029785, loss=1.1407793760299683 +I0831 13:32:38.245408 139540719195904 logging_writer.py:48] [61200] global_step=61200, grad_norm=3.8766753673553467, loss=1.065988540649414 +I0831 13:33:05.341774 139540727588608 logging_writer.py:48] [61300] global_step=61300, grad_norm=4.286856174468994, loss=1.1069066524505615 +I0831 13:33:32.670951 139540719195904 logging_writer.py:48] [61400] global_step=61400, grad_norm=4.276535987854004, loss=1.1756739616394043 +I0831 13:33:59.825260 139540727588608 logging_writer.py:48] [61500] global_step=61500, grad_norm=4.273529529571533, loss=1.1471271514892578 +I0831 13:34:26.931727 139540719195904 logging_writer.py:48] [61600] global_step=61600, grad_norm=4.194718837738037, loss=1.1037203073501587 +I0831 13:34:54.021187 139540727588608 logging_writer.py:48] [61700] global_step=61700, grad_norm=4.0157270431518555, loss=1.048763394355774 +I0831 13:35:21.178122 139540719195904 logging_writer.py:48] [61800] global_step=61800, grad_norm=4.572597503662109, loss=1.1253948211669922 +I0831 13:35:48.275742 139540727588608 logging_writer.py:48] [61900] global_step=61900, grad_norm=4.2384748458862305, loss=1.107635498046875 +I0831 13:36:15.393282 139540719195904 logging_writer.py:48] [62000] global_step=62000, grad_norm=4.157260894775391, loss=1.1408571004867554 +I0831 13:36:42.549543 139540727588608 logging_writer.py:48] [62100] global_step=62100, grad_norm=4.371570110321045, loss=1.263484001159668 +I0831 13:37:09.628035 139540719195904 logging_writer.py:48] [62200] global_step=62200, grad_norm=4.365884780883789, loss=1.1170647144317627 +I0831 13:37:36.706985 139540727588608 logging_writer.py:48] [62300] global_step=62300, grad_norm=4.32310676574707, loss=1.1869494915008545 +I0831 13:38:04.108423 139540719195904 logging_writer.py:48] [62400] global_step=62400, grad_norm=4.643028736114502, loss=1.219673752784729 +I0831 13:38:31.203308 139540727588608 logging_writer.py:48] [62500] global_step=62500, grad_norm=4.538817405700684, loss=1.2222492694854736 +I0831 13:38:58.307030 139540719195904 logging_writer.py:48] [62600] global_step=62600, grad_norm=3.968566656112671, loss=1.058457851409912 +I0831 13:39:25.451693 139540727588608 logging_writer.py:48] [62700] global_step=62700, grad_norm=4.538475036621094, loss=1.1615204811096191 +I0831 13:39:52.561463 139540719195904 logging_writer.py:48] [62800] global_step=62800, grad_norm=4.822351455688477, loss=1.1555804014205933 +I0831 13:40:19.646280 139540727588608 logging_writer.py:48] [62900] global_step=62900, grad_norm=4.345246315002441, loss=1.0453455448150635 +I0831 13:40:46.794142 139540719195904 logging_writer.py:48] [63000] global_step=63000, grad_norm=4.433428764343262, loss=1.2152522802352905 +I0831 13:41:13.917238 139540727588608 logging_writer.py:48] [63100] global_step=63100, grad_norm=4.222021579742432, loss=1.1546388864517212 +I0831 13:41:41.023684 139540719195904 logging_writer.py:48] [63200] global_step=63200, grad_norm=4.490721702575684, loss=1.11820387840271 +I0831 13:42:08.183294 139540727588608 logging_writer.py:48] [63300] global_step=63300, grad_norm=4.523038864135742, loss=1.0728404521942139 +I0831 13:42:35.264720 139540719195904 logging_writer.py:48] [63400] global_step=63400, grad_norm=4.373589992523193, loss=1.1755000352859497 +I0831 13:43:02.614817 139540727588608 logging_writer.py:48] [63500] global_step=63500, grad_norm=4.24800443649292, loss=1.090347170829773 +I0831 13:43:29.753198 139540719195904 logging_writer.py:48] [63600] global_step=63600, grad_norm=4.273742198944092, loss=1.13649582862854 +I0831 13:43:56.862274 139540727588608 logging_writer.py:48] [63700] global_step=63700, grad_norm=4.640417098999023, loss=1.1735934019088745 +I0831 13:44:23.980981 139540719195904 logging_writer.py:48] [63800] global_step=63800, grad_norm=4.234874725341797, loss=1.0893020629882812 +I0831 13:44:51.127120 139540727588608 logging_writer.py:48] [63900] global_step=63900, grad_norm=4.455885887145996, loss=1.1877307891845703 +I0831 13:45:18.220582 139540719195904 logging_writer.py:48] [64000] global_step=64000, grad_norm=4.438656330108643, loss=1.139983057975769 +I0831 13:45:45.333924 139540727588608 logging_writer.py:48] [64100] global_step=64100, grad_norm=4.40826416015625, loss=1.1444511413574219 +I0831 13:46:12.500951 139540719195904 logging_writer.py:48] [64200] global_step=64200, grad_norm=4.3170061111450195, loss=1.1207876205444336 +I0831 13:46:39.596391 139540727588608 logging_writer.py:48] [64300] global_step=64300, grad_norm=4.467159748077393, loss=1.1381728649139404 +I0831 13:47:06.688423 139540719195904 logging_writer.py:48] [64400] global_step=64400, grad_norm=4.606790542602539, loss=1.0970185995101929 +I0831 13:47:33.869280 139540727588608 logging_writer.py:48] [64500] global_step=64500, grad_norm=4.512940883636475, loss=1.2085509300231934 +I0831 13:48:01.174889 139540719195904 logging_writer.py:48] [64600] global_step=64600, grad_norm=4.579649448394775, loss=1.1796603202819824 +I0831 13:48:28.248419 139540727588608 logging_writer.py:48] [64700] global_step=64700, grad_norm=4.497619152069092, loss=1.2076114416122437 +I0831 13:48:55.415782 139540719195904 logging_writer.py:48] [64800] global_step=64800, grad_norm=4.4055585861206055, loss=1.1563019752502441 +I0831 13:49:22.513969 139540727588608 logging_writer.py:48] [64900] global_step=64900, grad_norm=4.502171993255615, loss=1.0322864055633545 +I0831 13:49:49.621854 139540719195904 logging_writer.py:48] [65000] global_step=65000, grad_norm=4.534034729003906, loss=1.241274118423462 +I0831 13:50:16.795896 139540727588608 logging_writer.py:48] [65100] global_step=65100, grad_norm=4.625748634338379, loss=1.1614686250686646 +I0831 13:50:43.925496 139540719195904 logging_writer.py:48] [65200] global_step=65200, grad_norm=4.7350382804870605, loss=1.1474820375442505 +I0831 13:50:45.711474 139757377230016 spec.py:333] Evaluating on the training split. +I0831 13:50:54.790838 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 13:51:05.298954 139757377230016 spec.py:363] Evaluating on the test split. +I0831 13:51:06.200670 139757377230016 submission_runner.py:516] Time since start: 18276.52s, Step: 65208, {'train/accuracy': Array(0.90124756, dtype=float32), 'train/loss': Array(0.35447967, dtype=float32), 'validation/accuracy': Array(0.73748, dtype=float32), 'validation/loss': Array(1.0785924, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6104, dtype=float32), 'test/loss': Array(1.8365499, dtype=float32), 'test/num_examples': 10000, 'score': 18017.576412677765, 'total_duration': 18276.523310661316, 'accumulated_submission_time': 18017.576412677765, 'accumulated_eval_time': 257.79344487190247, 'accumulated_logging_time': 0.6483263969421387} +I0831 13:51:06.267806 139540727588608 logging_writer.py:48] [65208] accumulated_eval_time=257.793, accumulated_logging_time=0.648326, accumulated_submission_time=18017.6, global_step=65208, preemption_count=0, score=18017.6, test/accuracy=0.6104000210762024, test/loss=1.8365498781204224, test/num_examples=10000, total_duration=18276.5, train/accuracy=0.9012475609779358, train/loss=0.35447967052459717, validation/accuracy=0.7374799847602844, validation/loss=1.0785924196243286, validation/num_examples=50000 +I0831 13:51:31.637896 139540719195904 logging_writer.py:48] [65300] global_step=65300, grad_norm=4.4084672927856445, loss=1.1686173677444458 +I0831 13:51:58.793514 139540727588608 logging_writer.py:48] [65400] global_step=65400, grad_norm=4.625330924987793, loss=1.1611039638519287 +I0831 13:52:25.924852 139540719195904 logging_writer.py:48] [65500] global_step=65500, grad_norm=4.851144790649414, loss=1.2499616146087646 +I0831 13:52:53.211605 139540727588608 logging_writer.py:48] [65600] global_step=65600, grad_norm=4.840524673461914, loss=1.1195294857025146 +I0831 13:53:20.370573 139540719195904 logging_writer.py:48] [65700] global_step=65700, grad_norm=4.62559175491333, loss=1.1100071668624878 +I0831 13:53:47.492867 139540727588608 logging_writer.py:48] [65800] global_step=65800, grad_norm=4.627632141113281, loss=1.1573272943496704 +I0831 13:54:14.599057 139540719195904 logging_writer.py:48] [65900] global_step=65900, grad_norm=4.686713695526123, loss=1.0334947109222412 +I0831 13:54:41.765657 139540727588608 logging_writer.py:48] [66000] global_step=66000, grad_norm=4.385500907897949, loss=1.037397027015686 +I0831 13:55:08.841567 139540719195904 logging_writer.py:48] [66100] global_step=66100, grad_norm=4.738503456115723, loss=1.173659324645996 +I0831 13:55:35.955669 139540727588608 logging_writer.py:48] [66200] global_step=66200, grad_norm=4.786446571350098, loss=1.1743108034133911 +I0831 13:56:03.112194 139540719195904 logging_writer.py:48] [66300] global_step=66300, grad_norm=4.637209415435791, loss=1.1685290336608887 +I0831 13:56:30.223822 139540727588608 logging_writer.py:48] [66400] global_step=66400, grad_norm=4.615949630737305, loss=1.0552440881729126 +I0831 13:56:57.303147 139540719195904 logging_writer.py:48] [66500] global_step=66500, grad_norm=4.283204555511475, loss=1.1156580448150635 +I0831 13:57:24.496188 139540727588608 logging_writer.py:48] [66600] global_step=66600, grad_norm=4.391711711883545, loss=1.0595004558563232 +I0831 13:57:51.804424 139540719195904 logging_writer.py:48] [66700] global_step=66700, grad_norm=5.078994274139404, loss=1.220777988433838 +I0831 13:58:18.929701 139540727588608 logging_writer.py:48] [66800] global_step=66800, grad_norm=4.648497581481934, loss=1.1116327047348022 +I0831 13:58:46.069276 139540719195904 logging_writer.py:48] [66900] global_step=66900, grad_norm=4.754629611968994, loss=1.1105635166168213 +I0831 13:59:13.172781 139540727588608 logging_writer.py:48] [67000] global_step=67000, grad_norm=5.155903339385986, loss=1.2200498580932617 +I0831 13:59:40.268464 139540719195904 logging_writer.py:48] [67100] global_step=67100, grad_norm=4.485778331756592, loss=1.0552616119384766 +I0831 14:00:07.442929 139540727588608 logging_writer.py:48] [67200] global_step=67200, grad_norm=4.500012397766113, loss=1.057807207107544 +I0831 14:00:34.550737 139540719195904 logging_writer.py:48] [67300] global_step=67300, grad_norm=4.672736644744873, loss=1.1248643398284912 +I0831 14:01:01.640295 139540727588608 logging_writer.py:48] [67400] global_step=67400, grad_norm=4.845271587371826, loss=1.164858341217041 +I0831 14:01:28.799671 139540719195904 logging_writer.py:48] [67500] global_step=67500, grad_norm=4.822690010070801, loss=1.1714894771575928 +I0831 14:01:55.901929 139540727588608 logging_writer.py:48] [67600] global_step=67600, grad_norm=5.0031232833862305, loss=1.1086652278900146 +I0831 14:02:23.234045 139540719195904 logging_writer.py:48] [67700] global_step=67700, grad_norm=4.75212287902832, loss=1.1819580793380737 +I0831 14:02:50.390040 139540727588608 logging_writer.py:48] [67800] global_step=67800, grad_norm=4.759058952331543, loss=1.1221896409988403 +I0831 14:03:17.517119 139540719195904 logging_writer.py:48] [67900] global_step=67900, grad_norm=4.81304931640625, loss=1.1828184127807617 +I0831 14:03:44.612629 139540727588608 logging_writer.py:48] [68000] global_step=68000, grad_norm=4.950571060180664, loss=1.1001051664352417 +I0831 14:04:11.785838 139540719195904 logging_writer.py:48] [68100] global_step=68100, grad_norm=4.800511360168457, loss=1.1944143772125244 +I0831 14:04:38.899561 139540727588608 logging_writer.py:48] [68200] global_step=68200, grad_norm=4.589383125305176, loss=1.0434989929199219 +I0831 14:05:05.992502 139540719195904 logging_writer.py:48] [68300] global_step=68300, grad_norm=4.951854228973389, loss=1.102988839149475 +I0831 14:05:33.156585 139540727588608 logging_writer.py:48] [68400] global_step=68400, grad_norm=4.9022440910339355, loss=1.056746006011963 +I0831 14:06:00.259508 139540719195904 logging_writer.py:48] [68500] global_step=68500, grad_norm=4.846834659576416, loss=1.1349437236785889 +I0831 14:06:27.387410 139540727588608 logging_writer.py:48] [68600] global_step=68600, grad_norm=4.818789958953857, loss=1.0974410772323608 +I0831 14:06:54.572856 139540719195904 logging_writer.py:48] [68700] global_step=68700, grad_norm=4.931778430938721, loss=1.199153184890747 +I0831 14:07:21.856224 139540727588608 logging_writer.py:48] [68800] global_step=68800, grad_norm=5.051036357879639, loss=1.1476305723190308 +I0831 14:07:48.925763 139540719195904 logging_writer.py:48] [68900] global_step=68900, grad_norm=4.879879474639893, loss=1.1401822566986084 +I0831 14:08:16.073468 139540727588608 logging_writer.py:48] [69000] global_step=69000, grad_norm=4.779599189758301, loss=1.1636688709259033 +I0831 14:08:43.156609 139540719195904 logging_writer.py:48] [69100] global_step=69100, grad_norm=4.733686447143555, loss=1.1386079788208008 +I0831 14:09:10.226541 139540727588608 logging_writer.py:48] [69200] global_step=69200, grad_norm=4.913372993469238, loss=1.0821601152420044 +I0831 14:09:37.413309 139540719195904 logging_writer.py:48] [69300] global_step=69300, grad_norm=4.538520812988281, loss=1.005618691444397 +I0831 14:10:04.530532 139540727588608 logging_writer.py:48] [69400] global_step=69400, grad_norm=4.685912609100342, loss=1.0542535781860352 +I0831 14:10:31.652286 139540719195904 logging_writer.py:48] [69500] global_step=69500, grad_norm=5.225014686584473, loss=1.222835659980774 +I0831 14:10:58.815046 139540727588608 logging_writer.py:48] [69600] global_step=69600, grad_norm=4.832304954528809, loss=1.0613253116607666 +I0831 14:11:25.913750 139540719195904 logging_writer.py:48] [69700] global_step=69700, grad_norm=4.997440814971924, loss=1.0843690633773804 +I0831 14:11:53.212703 139540727588608 logging_writer.py:48] [69800] global_step=69800, grad_norm=5.0918426513671875, loss=1.136114478111267 +I0831 14:12:20.409056 139540719195904 logging_writer.py:48] [69900] global_step=69900, grad_norm=5.058177471160889, loss=1.0965068340301514 +I0831 14:12:47.498684 139540727588608 logging_writer.py:48] [70000] global_step=70000, grad_norm=5.116286754608154, loss=1.1150177717208862 +I0831 14:13:14.568920 139540719195904 logging_writer.py:48] [70100] global_step=70100, grad_norm=5.034357070922852, loss=1.082290530204773 +I0831 14:13:41.733715 139540727588608 logging_writer.py:48] [70200] global_step=70200, grad_norm=5.078695297241211, loss=1.201764464378357 +I0831 14:14:08.800408 139540719195904 logging_writer.py:48] [70300] global_step=70300, grad_norm=4.842857837677002, loss=1.0396071672439575 +I0831 14:14:35.909703 139540727588608 logging_writer.py:48] [70400] global_step=70400, grad_norm=5.281187534332275, loss=1.1442655324935913 +I0831 14:15:03.084229 139540719195904 logging_writer.py:48] [70500] global_step=70500, grad_norm=4.997383117675781, loss=1.1571868658065796 +I0831 14:15:30.169057 139540727588608 logging_writer.py:48] [70600] global_step=70600, grad_norm=5.002779960632324, loss=1.129809021949768 +I0831 14:15:57.276254 139540719195904 logging_writer.py:48] [70700] global_step=70700, grad_norm=5.060858726501465, loss=1.177739143371582 +I0831 14:16:24.450411 139540727588608 logging_writer.py:48] [70800] global_step=70800, grad_norm=4.895029067993164, loss=1.0943540334701538 +I0831 14:16:51.762460 139540719195904 logging_writer.py:48] [70900] global_step=70900, grad_norm=4.892037868499756, loss=1.1062815189361572 +I0831 14:17:18.850089 139540727588608 logging_writer.py:48] [71000] global_step=71000, grad_norm=4.909911632537842, loss=1.0888062715530396 +I0831 14:17:46.019329 139540719195904 logging_writer.py:48] [71100] global_step=71100, grad_norm=4.904439926147461, loss=1.0683060884475708 +I0831 14:18:13.117469 139540727588608 logging_writer.py:48] [71200] global_step=71200, grad_norm=5.02208137512207, loss=1.1104037761688232 +I0831 14:18:40.266189 139540719195904 logging_writer.py:48] [71300] global_step=71300, grad_norm=4.902232646942139, loss=1.0629605054855347 +I0831 14:19:07.430469 139540727588608 logging_writer.py:48] [71400] global_step=71400, grad_norm=4.865591526031494, loss=1.059987187385559 +I0831 14:19:34.527427 139540719195904 logging_writer.py:48] [71500] global_step=71500, grad_norm=4.850372314453125, loss=1.0402113199234009 +I0831 14:20:01.642715 139540727588608 logging_writer.py:48] [71600] global_step=71600, grad_norm=5.13913631439209, loss=1.1360918283462524 +I0831 14:20:28.849527 139540719195904 logging_writer.py:48] [71700] global_step=71700, grad_norm=5.084836006164551, loss=1.1192049980163574 +I0831 14:20:55.973710 139540727588608 logging_writer.py:48] [71800] global_step=71800, grad_norm=5.142754554748535, loss=1.0479192733764648 +I0831 14:21:23.276573 139540719195904 logging_writer.py:48] [71900] global_step=71900, grad_norm=5.119273662567139, loss=1.052194595336914 +I0831 14:21:50.453713 139540727588608 logging_writer.py:48] [72000] global_step=72000, grad_norm=5.223738670349121, loss=1.1881577968597412 +I0831 14:22:17.546402 139540719195904 logging_writer.py:48] [72100] global_step=72100, grad_norm=4.810418605804443, loss=0.9835866093635559 +I0831 14:22:44.652518 139540727588608 logging_writer.py:48] [72200] global_step=72200, grad_norm=4.916736602783203, loss=1.084503412246704 +I0831 14:23:11.807532 139540719195904 logging_writer.py:48] [72300] global_step=72300, grad_norm=4.785176753997803, loss=1.0774383544921875 +I0831 14:23:38.929244 139540727588608 logging_writer.py:48] [72400] global_step=72400, grad_norm=5.186130523681641, loss=1.0957170724868774 +I0831 14:24:06.055676 139540719195904 logging_writer.py:48] [72500] global_step=72500, grad_norm=5.191675186157227, loss=1.2011656761169434 +I0831 14:24:22.203474 139757377230016 spec.py:333] Evaluating on the training split. +I0831 14:24:30.337248 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 14:24:40.415546 139757377230016 spec.py:363] Evaluating on the test split. +I0831 14:24:41.299163 139757377230016 submission_runner.py:516] Time since start: 20291.63s, Step: 72561, {'train/accuracy': Array(0.90824294, dtype=float32), 'train/loss': Array(0.32824045, dtype=float32), 'validation/accuracy': Array(0.74074, dtype=float32), 'validation/loss': Array(1.0726177, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6108, dtype=float32), 'test/loss': Array(1.8346378, dtype=float32), 'test/num_examples': 10000, 'score': 20013.385497570038, 'total_duration': 20291.634605646133, 'accumulated_submission_time': 20013.385497570038, 'accumulated_eval_time': 276.8868010044098, 'accumulated_logging_time': 0.7893733978271484} +I0831 14:24:41.353234 139540727588608 logging_writer.py:48] [72561] accumulated_eval_time=276.887, accumulated_logging_time=0.789373, accumulated_submission_time=20013.4, global_step=72561, preemption_count=0, score=20013.4, test/accuracy=0.61080002784729, test/loss=1.8346377611160278, test/num_examples=10000, total_duration=20291.6, train/accuracy=0.90824294090271, train/loss=0.32824045419692993, validation/accuracy=0.7407400012016296, validation/loss=1.0726176500320435, validation/num_examples=50000 +I0831 14:24:52.310816 139540719195904 logging_writer.py:48] [72600] global_step=72600, grad_norm=5.258336067199707, loss=1.1007555723190308 +I0831 14:25:19.402232 139540727588608 logging_writer.py:48] [72700] global_step=72700, grad_norm=5.065289497375488, loss=1.0494859218597412 +I0831 14:25:46.480303 139540719195904 logging_writer.py:48] [72800] global_step=72800, grad_norm=5.373915672302246, loss=1.0879582166671753 +I0831 14:26:13.632205 139540727588608 logging_writer.py:48] [72900] global_step=72900, grad_norm=4.803307056427002, loss=1.0320947170257568 +I0831 14:26:40.954724 139540719195904 logging_writer.py:48] [73000] global_step=73000, grad_norm=5.073699951171875, loss=1.1762635707855225 +I0831 14:27:08.019201 139540727588608 logging_writer.py:48] [73100] global_step=73100, grad_norm=5.1512932777404785, loss=1.1518657207489014 +I0831 14:27:35.194759 139540719195904 logging_writer.py:48] [73200] global_step=73200, grad_norm=4.900606155395508, loss=1.0374982357025146 +I0831 14:28:02.294033 139540727588608 logging_writer.py:48] [73300] global_step=73300, grad_norm=5.006901741027832, loss=1.0804922580718994 +I0831 14:28:29.381299 139540719195904 logging_writer.py:48] [73400] global_step=73400, grad_norm=5.176017761230469, loss=1.1005182266235352 +I0831 14:28:56.542944 139540727588608 logging_writer.py:48] [73500] global_step=73500, grad_norm=4.877220630645752, loss=1.0573527812957764 +I0831 14:29:23.641581 139540719195904 logging_writer.py:48] [73600] global_step=73600, grad_norm=5.238534450531006, loss=1.1672782897949219 +I0831 14:29:50.731818 139540727588608 logging_writer.py:48] [73700] global_step=73700, grad_norm=5.4541425704956055, loss=1.1482524871826172 +I0831 14:30:17.896106 139540719195904 logging_writer.py:48] [73800] global_step=73800, grad_norm=4.8899712562561035, loss=1.0302114486694336 +I0831 14:30:44.980780 139540727588608 logging_writer.py:48] [73900] global_step=73900, grad_norm=5.284416675567627, loss=1.1735618114471436 +I0831 14:31:12.276559 139540719195904 logging_writer.py:48] [74000] global_step=74000, grad_norm=4.71981143951416, loss=0.9710990786552429 +I0831 14:31:39.448508 139540727588608 logging_writer.py:48] [74100] global_step=74100, grad_norm=5.173820972442627, loss=1.1262990236282349 +I0831 14:32:06.559592 139540719195904 logging_writer.py:48] [74200] global_step=74200, grad_norm=5.029001712799072, loss=1.0381803512573242 +I0831 14:32:33.619798 139540727588608 logging_writer.py:48] [74300] global_step=74300, grad_norm=4.889930248260498, loss=1.1397886276245117 +I0831 14:33:00.797139 139540719195904 logging_writer.py:48] [74400] global_step=74400, grad_norm=5.22511625289917, loss=1.0749008655548096 +I0831 14:33:27.880789 139540727588608 logging_writer.py:48] [74500] global_step=74500, grad_norm=5.099572658538818, loss=1.022437572479248 +I0831 14:33:55.003522 139540719195904 logging_writer.py:48] [74600] global_step=74600, grad_norm=5.02353048324585, loss=1.0671968460083008 +I0831 14:34:22.177782 139540727588608 logging_writer.py:48] [74700] global_step=74700, grad_norm=5.156341552734375, loss=1.1270580291748047 +I0831 14:34:49.324909 139540719195904 logging_writer.py:48] [74800] global_step=74800, grad_norm=5.060338497161865, loss=1.1326285600662231 +I0831 14:35:16.469672 139540727588608 logging_writer.py:48] [74900] global_step=74900, grad_norm=5.070630073547363, loss=1.0973758697509766 +I0831 14:35:43.631304 139540719195904 logging_writer.py:48] [75000] global_step=75000, grad_norm=4.683627128601074, loss=0.9957688450813293 +I0831 14:36:10.738088 139540727588608 logging_writer.py:48] [75100] global_step=75100, grad_norm=5.094455718994141, loss=1.1114228963851929 +I0831 14:36:38.065063 139540719195904 logging_writer.py:48] [75200] global_step=75200, grad_norm=4.899845123291016, loss=1.0775337219238281 +I0831 14:37:05.211656 139540727588608 logging_writer.py:48] [75300] global_step=75300, grad_norm=5.001206398010254, loss=1.1131665706634521 +I0831 14:37:32.313997 139540719195904 logging_writer.py:48] [75400] global_step=75400, grad_norm=5.352885723114014, loss=1.0816067457199097 +I0831 14:37:59.403317 139540727588608 logging_writer.py:48] [75500] global_step=75500, grad_norm=5.19547176361084, loss=1.1168345212936401 +I0831 14:38:26.541747 139540719195904 logging_writer.py:48] [75600] global_step=75600, grad_norm=5.0340986251831055, loss=1.0960524082183838 +I0831 14:38:53.652051 139540727588608 logging_writer.py:48] [75700] global_step=75700, grad_norm=5.151155948638916, loss=1.1211767196655273 +I0831 14:39:20.784143 139540719195904 logging_writer.py:48] [75800] global_step=75800, grad_norm=5.041630744934082, loss=1.1007633209228516 +I0831 14:39:47.952102 139540727588608 logging_writer.py:48] [75900] global_step=75900, grad_norm=5.421639442443848, loss=1.099066972732544 +I0831 14:40:15.066986 139540719195904 logging_writer.py:48] [76000] global_step=76000, grad_norm=5.105546474456787, loss=1.1071828603744507 +I0831 14:40:42.183245 139540727588608 logging_writer.py:48] [76100] global_step=76100, grad_norm=5.046034812927246, loss=1.1315319538116455 +I0831 14:41:09.588495 139540719195904 logging_writer.py:48] [76200] global_step=76200, grad_norm=5.160421371459961, loss=1.1579453945159912 +I0831 14:41:36.717707 139540727588608 logging_writer.py:48] [76300] global_step=76300, grad_norm=5.071197986602783, loss=1.09730863571167 +I0831 14:42:03.831100 139540719195904 logging_writer.py:48] [76400] global_step=76400, grad_norm=5.100929260253906, loss=1.087558627128601 +I0831 14:42:31.027359 139540727588608 logging_writer.py:48] [76500] global_step=76500, grad_norm=4.890820503234863, loss=1.0608569383621216 +I0831 14:42:58.144211 139540719195904 logging_writer.py:48] [76600] global_step=76600, grad_norm=4.929311275482178, loss=1.1246623992919922 +I0831 14:43:25.233380 139540727588608 logging_writer.py:48] [76700] global_step=76700, grad_norm=4.898760795593262, loss=1.038846492767334 +I0831 14:43:52.428675 139540719195904 logging_writer.py:48] [76800] global_step=76800, grad_norm=5.133862495422363, loss=1.1213703155517578 +I0831 14:44:19.513286 139540727588608 logging_writer.py:48] [76900] global_step=76900, grad_norm=5.241682529449463, loss=1.078906774520874 +I0831 14:44:46.595087 139540719195904 logging_writer.py:48] [77000] global_step=77000, grad_norm=4.978885173797607, loss=1.0525554418563843 +I0831 14:45:13.772156 139540727588608 logging_writer.py:48] [77100] global_step=77100, grad_norm=5.049511909484863, loss=1.1005327701568604 +I0831 14:45:40.897891 139540719195904 logging_writer.py:48] [77200] global_step=77200, grad_norm=4.948261737823486, loss=1.0380823612213135 +I0831 14:46:08.227655 139540727588608 logging_writer.py:48] [77300] global_step=77300, grad_norm=4.855027675628662, loss=1.0551496744155884 +I0831 14:46:35.387624 139540719195904 logging_writer.py:48] [77400] global_step=77400, grad_norm=5.110690116882324, loss=1.0439121723175049 +I0831 14:47:02.466961 139540727588608 logging_writer.py:48] [77500] global_step=77500, grad_norm=5.24586820602417, loss=1.1588771343231201 +I0831 14:47:29.579693 139540719195904 logging_writer.py:48] [77600] global_step=77600, grad_norm=5.253383159637451, loss=1.1526169776916504 +I0831 14:47:56.732844 139540727588608 logging_writer.py:48] [77700] global_step=77700, grad_norm=5.149881839752197, loss=1.1658427715301514 +I0831 14:48:23.843462 139540719195904 logging_writer.py:48] [77800] global_step=77800, grad_norm=5.53518009185791, loss=1.1289994716644287 +I0831 14:48:50.958380 139540727588608 logging_writer.py:48] [77900] global_step=77900, grad_norm=4.778284549713135, loss=1.0319041013717651 +I0831 14:49:18.111934 139540719195904 logging_writer.py:48] [78000] global_step=78000, grad_norm=5.195228576660156, loss=1.0831815004348755 +I0831 14:49:45.180809 139540727588608 logging_writer.py:48] [78100] global_step=78100, grad_norm=5.241842746734619, loss=1.1263651847839355 +I0831 14:50:12.294131 139540719195904 logging_writer.py:48] [78200] global_step=78200, grad_norm=5.098332405090332, loss=1.1317367553710938 +I0831 14:50:39.696551 139540727588608 logging_writer.py:48] [78300] global_step=78300, grad_norm=4.9898552894592285, loss=1.0627614259719849 +I0831 14:51:06.763930 139540719195904 logging_writer.py:48] [78400] global_step=78400, grad_norm=5.159222602844238, loss=1.1542898416519165 +I0831 14:51:33.875002 139540727588608 logging_writer.py:48] [78500] global_step=78500, grad_norm=5.451019763946533, loss=1.182709813117981 +I0831 14:52:01.008587 139540719195904 logging_writer.py:48] [78600] global_step=78600, grad_norm=4.751550674438477, loss=1.0387800931930542 +I0831 14:52:28.122160 139540727588608 logging_writer.py:48] [78700] global_step=78700, grad_norm=5.461548805236816, loss=1.1650280952453613 +I0831 14:52:55.232805 139540719195904 logging_writer.py:48] [78800] global_step=78800, grad_norm=5.0468831062316895, loss=1.063409447669983 +I0831 14:53:22.408478 139540727588608 logging_writer.py:48] [78900] global_step=78900, grad_norm=5.135944366455078, loss=1.1858068704605103 +I0831 14:53:49.495729 139540719195904 logging_writer.py:48] [79000] global_step=79000, grad_norm=5.133060455322266, loss=1.1476011276245117 +I0831 14:54:16.569549 139540727588608 logging_writer.py:48] [79100] global_step=79100, grad_norm=5.210805892944336, loss=1.0379137992858887 +I0831 14:54:43.732961 139540719195904 logging_writer.py:48] [79200] global_step=79200, grad_norm=5.22809362411499, loss=1.0903551578521729 +I0831 14:55:10.835960 139540727588608 logging_writer.py:48] [79300] global_step=79300, grad_norm=5.242574691772461, loss=1.0695445537567139 +I0831 14:55:38.166122 139540719195904 logging_writer.py:48] [79400] global_step=79400, grad_norm=4.6560235023498535, loss=0.9886162281036377 +I0831 14:56:05.321842 139540727588608 logging_writer.py:48] [79500] global_step=79500, grad_norm=5.13606595993042, loss=1.1352734565734863 +I0831 14:56:32.425243 139540719195904 logging_writer.py:48] [79600] global_step=79600, grad_norm=5.278378009796143, loss=1.1470916271209717 +I0831 14:56:59.522086 139540727588608 logging_writer.py:48] [79700] global_step=79700, grad_norm=4.837241172790527, loss=1.0685762166976929 +I0831 14:57:26.683513 139540719195904 logging_writer.py:48] [79800] global_step=79800, grad_norm=5.013140678405762, loss=1.0356762409210205 +I0831 14:57:53.770994 139540727588608 logging_writer.py:48] [79900] global_step=79900, grad_norm=5.156036376953125, loss=1.08138906955719 +I0831 14:57:57.493708 139757377230016 spec.py:333] Evaluating on the training split. +I0831 14:58:05.472194 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 14:58:15.244947 139757377230016 spec.py:363] Evaluating on the test split. +I0831 14:58:16.158501 139757377230016 submission_runner.py:516] Time since start: 22306.47s, Step: 79915, {'train/accuracy': Array(0.9136041, dtype=float32), 'train/loss': Array(0.3116816, dtype=float32), 'validation/accuracy': Array(0.74292, dtype=float32), 'validation/loss': Array(1.0638095, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.61490005, dtype=float32), 'test/loss': Array(1.8260653, dtype=float32), 'test/num_examples': 10000, 'score': 22009.430178642273, 'total_duration': 22306.47315478325, 'accumulated_submission_time': 22009.430178642273, 'accumulated_eval_time': 295.5284707546234, 'accumulated_logging_time': 0.8865344524383545} +I0831 14:58:16.223726 139540719195904 logging_writer.py:48] [79915] accumulated_eval_time=295.528, accumulated_logging_time=0.886534, accumulated_submission_time=22009.4, global_step=79915, preemption_count=0, score=22009.4, test/accuracy=0.6149000525474548, test/loss=1.8260653018951416, test/num_examples=10000, total_duration=22306.5, train/accuracy=0.9136040806770325, train/loss=0.3116815984249115, validation/accuracy=0.7429199814796448, validation/loss=1.0638095140457153, validation/num_examples=50000 +I0831 14:58:39.600140 139540727588608 logging_writer.py:48] [80000] global_step=80000, grad_norm=4.744093418121338, loss=0.9367897510528564 +I0831 14:59:06.709165 139540719195904 logging_writer.py:48] [80100] global_step=80100, grad_norm=5.226950168609619, loss=1.0818908214569092 +I0831 14:59:33.763787 139540727588608 logging_writer.py:48] [80200] global_step=80200, grad_norm=5.130753993988037, loss=1.0885624885559082 +I0831 15:00:00.892218 139540719195904 logging_writer.py:48] [80300] global_step=80300, grad_norm=5.093482494354248, loss=1.0991146564483643 +I0831 15:00:28.046098 139540727588608 logging_writer.py:48] [80400] global_step=80400, grad_norm=5.289251804351807, loss=1.1376756429672241 +I0831 15:00:55.411067 139540719195904 logging_writer.py:48] [80500] global_step=80500, grad_norm=5.154627799987793, loss=1.0619131326675415 +I0831 15:01:22.517303 139540727588608 logging_writer.py:48] [80600] global_step=80600, grad_norm=5.198306560516357, loss=1.1179152727127075 +I0831 15:01:49.701807 139540719195904 logging_writer.py:48] [80700] global_step=80700, grad_norm=5.296383857727051, loss=1.1098949909210205 +I0831 15:02:16.813965 139540727588608 logging_writer.py:48] [80800] global_step=80800, grad_norm=5.339659690856934, loss=1.142543911933899 +I0831 15:02:43.923091 139540719195904 logging_writer.py:48] [80900] global_step=80900, grad_norm=5.13142728805542, loss=1.0713517665863037 +I0831 15:03:11.075199 139540727588608 logging_writer.py:48] [81000] global_step=81000, grad_norm=5.270387172698975, loss=1.1071650981903076 +I0831 15:03:38.217443 139540719195904 logging_writer.py:48] [81100] global_step=81100, grad_norm=4.758582592010498, loss=1.0342024564743042 +I0831 15:04:05.315285 139540727588608 logging_writer.py:48] [81200] global_step=81200, grad_norm=5.376667499542236, loss=1.1756317615509033 +I0831 15:04:32.493022 139540719195904 logging_writer.py:48] [81300] global_step=81300, grad_norm=4.837802886962891, loss=1.0759117603302002 +I0831 15:04:59.594568 139540727588608 logging_writer.py:48] [81400] global_step=81400, grad_norm=5.163736343383789, loss=1.1063644886016846 +I0831 15:05:26.905103 139540719195904 logging_writer.py:48] [81500] global_step=81500, grad_norm=5.180639743804932, loss=1.111250877380371 +I0831 15:05:54.061866 139540727588608 logging_writer.py:48] [81600] global_step=81600, grad_norm=5.208388805389404, loss=1.107893466949463 +I0831 15:06:21.173560 139540719195904 logging_writer.py:48] [81700] global_step=81700, grad_norm=5.025547504425049, loss=1.0409092903137207 +I0831 15:06:48.253910 139540727588608 logging_writer.py:48] [81800] global_step=81800, grad_norm=5.025572299957275, loss=1.0438916683197021 +I0831 15:07:15.455354 139540719195904 logging_writer.py:48] [81900] global_step=81900, grad_norm=5.105414390563965, loss=1.1133625507354736 +I0831 15:07:42.555746 139540727588608 logging_writer.py:48] [82000] global_step=82000, grad_norm=4.910178184509277, loss=0.953773021697998 +I0831 15:08:09.677469 139540719195904 logging_writer.py:48] [82100] global_step=82100, grad_norm=4.815548419952393, loss=1.06009840965271 +I0831 15:08:36.825794 139540727588608 logging_writer.py:48] [82200] global_step=82200, grad_norm=5.293375492095947, loss=1.1484216451644897 +I0831 15:09:03.927998 139540719195904 logging_writer.py:48] [82300] global_step=82300, grad_norm=5.31777286529541, loss=1.145188808441162 +I0831 15:09:31.036367 139540727588608 logging_writer.py:48] [82400] global_step=82400, grad_norm=5.161777496337891, loss=1.0146054029464722 +I0831 15:09:58.180765 139540719195904 logging_writer.py:48] [82500] global_step=82500, grad_norm=5.2499847412109375, loss=1.1282031536102295 +I0831 15:10:25.495724 139540727588608 logging_writer.py:48] [82600] global_step=82600, grad_norm=5.494203090667725, loss=1.1245166063308716 +I0831 15:10:52.606284 139540719195904 logging_writer.py:48] [82700] global_step=82700, grad_norm=4.862037181854248, loss=1.0237562656402588 +I0831 15:11:19.775637 139540727588608 logging_writer.py:48] [82800] global_step=82800, grad_norm=5.1293158531188965, loss=1.0669690370559692 +I0831 15:11:46.879893 139540719195904 logging_writer.py:48] [82900] global_step=82900, grad_norm=5.1616926193237305, loss=1.1265203952789307 +I0831 15:12:13.964505 139540727588608 logging_writer.py:48] [83000] global_step=83000, grad_norm=4.930270195007324, loss=1.028162956237793 +I0831 15:12:41.108870 139540719195904 logging_writer.py:48] [83100] global_step=83100, grad_norm=5.094121932983398, loss=1.0673696994781494 +I0831 15:13:08.210945 139540727588608 logging_writer.py:48] [83200] global_step=83200, grad_norm=5.020478248596191, loss=1.107775092124939 +I0831 15:13:35.311998 139540719195904 logging_writer.py:48] [83300] global_step=83300, grad_norm=5.273924350738525, loss=1.1261659860610962 +I0831 15:14:02.449509 139540727588608 logging_writer.py:48] [83400] global_step=83400, grad_norm=4.917331218719482, loss=1.100678563117981 +I0831 15:14:29.557577 139540719195904 logging_writer.py:48] [83500] global_step=83500, grad_norm=4.800319671630859, loss=1.0908207893371582 +I0831 15:14:56.893798 139540727588608 logging_writer.py:48] [83600] global_step=83600, grad_norm=5.255965709686279, loss=1.140325903892517 +I0831 15:15:24.058232 139540719195904 logging_writer.py:48] [83700] global_step=83700, grad_norm=4.909034252166748, loss=1.0991261005401611 +I0831 15:15:51.163793 139540727588608 logging_writer.py:48] [83800] global_step=83800, grad_norm=5.041651248931885, loss=1.0956569910049438 +I0831 15:16:18.251625 139540719195904 logging_writer.py:48] [83900] global_step=83900, grad_norm=4.911158084869385, loss=1.0243654251098633 +I0831 15:16:45.431208 139540727588608 logging_writer.py:48] [84000] global_step=84000, grad_norm=5.0859456062316895, loss=1.1317462921142578 +I0831 15:17:12.541075 139540719195904 logging_writer.py:48] [84100] global_step=84100, grad_norm=4.957993507385254, loss=1.010432481765747 +I0831 15:17:39.631086 139540727588608 logging_writer.py:48] [84200] global_step=84200, grad_norm=5.366715431213379, loss=1.0703991651535034 +I0831 15:18:06.796607 139540719195904 logging_writer.py:48] [84300] global_step=84300, grad_norm=5.201183319091797, loss=1.0934944152832031 +I0831 15:18:33.910831 139540727588608 logging_writer.py:48] [84400] global_step=84400, grad_norm=5.126269817352295, loss=1.1197023391723633 +I0831 15:19:00.984234 139540719195904 logging_writer.py:48] [84500] global_step=84500, grad_norm=5.112061500549316, loss=1.103268027305603 +I0831 15:19:28.141565 139540727588608 logging_writer.py:48] [84600] global_step=84600, grad_norm=5.0100579261779785, loss=1.0136958360671997 +I0831 15:19:55.404945 139540719195904 logging_writer.py:48] [84700] global_step=84700, grad_norm=5.16511869430542, loss=1.0557440519332886 +I0831 15:20:22.498792 139540727588608 logging_writer.py:48] [84800] global_step=84800, grad_norm=5.157710552215576, loss=1.038612961769104 +I0831 15:20:49.666828 139540719195904 logging_writer.py:48] [84900] global_step=84900, grad_norm=4.876095771789551, loss=1.1368645429611206 +I0831 15:21:16.765154 139540727588608 logging_writer.py:48] [85000] global_step=85000, grad_norm=5.159566402435303, loss=1.0806019306182861 +I0831 15:21:43.886749 139540719195904 logging_writer.py:48] [85100] global_step=85100, grad_norm=5.064005374908447, loss=1.072684645652771 +I0831 15:22:11.043895 139540727588608 logging_writer.py:48] [85200] global_step=85200, grad_norm=4.852941989898682, loss=1.0412896871566772 +I0831 15:22:38.139861 139540719195904 logging_writer.py:48] [85300] global_step=85300, grad_norm=4.9304585456848145, loss=0.9937907457351685 +I0831 15:23:05.264745 139540727588608 logging_writer.py:48] [85400] global_step=85400, grad_norm=5.173736095428467, loss=1.1275479793548584 +I0831 15:23:32.427345 139540719195904 logging_writer.py:48] [85500] global_step=85500, grad_norm=5.238888263702393, loss=1.0882760286331177 +I0831 15:23:59.553194 139540727588608 logging_writer.py:48] [85600] global_step=85600, grad_norm=5.312877655029297, loss=1.155517816543579 +I0831 15:24:26.869541 139540719195904 logging_writer.py:48] [85700] global_step=85700, grad_norm=4.987231731414795, loss=1.1068980693817139 +I0831 15:24:54.062950 139540727588608 logging_writer.py:48] [85800] global_step=85800, grad_norm=5.158888816833496, loss=1.0830414295196533 +I0831 15:25:21.163216 139540719195904 logging_writer.py:48] [85900] global_step=85900, grad_norm=5.0162763595581055, loss=1.0491430759429932 +I0831 15:25:48.249259 139540727588608 logging_writer.py:48] [86000] global_step=86000, grad_norm=4.799073219299316, loss=1.0986053943634033 +I0831 15:26:15.401664 139540719195904 logging_writer.py:48] [86100] global_step=86100, grad_norm=4.954151630401611, loss=1.09287428855896 +I0831 15:26:42.482722 139540727588608 logging_writer.py:48] [86200] global_step=86200, grad_norm=5.214901924133301, loss=1.0880508422851562 +I0831 15:27:09.584585 139540719195904 logging_writer.py:48] [86300] global_step=86300, grad_norm=5.2477617263793945, loss=1.0971214771270752 +I0831 15:27:36.739965 139540727588608 logging_writer.py:48] [86400] global_step=86400, grad_norm=5.169947624206543, loss=1.1557714939117432 +I0831 15:28:03.827989 139540719195904 logging_writer.py:48] [86500] global_step=86500, grad_norm=5.3332200050354, loss=1.1058385372161865 +I0831 15:28:30.918766 139540727588608 logging_writer.py:48] [86600] global_step=86600, grad_norm=5.159790992736816, loss=1.0301131010055542 +I0831 15:28:58.079639 139540719195904 logging_writer.py:48] [86700] global_step=86700, grad_norm=5.1269426345825195, loss=1.1231868267059326 +I0831 15:29:25.459433 139540727588608 logging_writer.py:48] [86800] global_step=86800, grad_norm=5.1101765632629395, loss=1.0677825212478638 +I0831 15:29:52.547476 139540719195904 logging_writer.py:48] [86900] global_step=86900, grad_norm=5.483833312988281, loss=1.1991355419158936 +I0831 15:30:19.704907 139540727588608 logging_writer.py:48] [87000] global_step=87000, grad_norm=4.789300918579102, loss=0.9257583618164062 +I0831 15:30:46.797233 139540719195904 logging_writer.py:48] [87100] global_step=87100, grad_norm=5.1198225021362305, loss=1.1940183639526367 +I0831 15:31:13.899950 139540727588608 logging_writer.py:48] [87200] global_step=87200, grad_norm=5.058359146118164, loss=1.0590507984161377 +I0831 15:31:32.193545 139757377230016 spec.py:333] Evaluating on the training split. +I0831 15:31:39.393924 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 15:31:49.478786 139757377230016 spec.py:363] Evaluating on the test split. +I0831 15:31:50.365302 139757377230016 submission_runner.py:516] Time since start: 24320.70s, Step: 87269, {'train/accuracy': Array(0.9151985, dtype=float32), 'train/loss': Array(0.29835552, dtype=float32), 'validation/accuracy': Array(0.74513996, dtype=float32), 'validation/loss': Array(1.0610034, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6208, dtype=float32), 'test/loss': Array(1.8269247, dtype=float32), 'test/num_examples': 10000, 'score': 24005.31016421318, 'total_duration': 24320.700825214386, 'accumulated_submission_time': 24005.31016421318, 'accumulated_eval_time': 313.69797587394714, 'accumulated_logging_time': 0.9892408847808838} +I0831 15:31:50.423866 139540719195904 logging_writer.py:48] [87269] accumulated_eval_time=313.698, accumulated_logging_time=0.989241, accumulated_submission_time=24005.3, global_step=87269, preemption_count=0, score=24005.3, test/accuracy=0.6208000183105469, test/loss=1.8269246816635132, test/num_examples=10000, total_duration=24320.7, train/accuracy=0.9151985049247742, train/loss=0.2983555197715759, validation/accuracy=0.7451399564743042, validation/loss=1.0610034465789795, validation/num_examples=50000 +I0831 15:31:59.233253 139540727588608 logging_writer.py:48] [87300] global_step=87300, grad_norm=4.8629279136657715, loss=1.1477861404418945 +I0831 15:32:26.320140 139540719195904 logging_writer.py:48] [87400] global_step=87400, grad_norm=5.109825611114502, loss=1.054623007774353 +I0831 15:32:53.415584 139540727588608 logging_writer.py:48] [87500] global_step=87500, grad_norm=5.0553154945373535, loss=1.0578076839447021 +I0831 15:33:20.539156 139540719195904 logging_writer.py:48] [87600] global_step=87600, grad_norm=5.151459693908691, loss=1.055330514907837 +I0831 15:33:47.645319 139540727588608 logging_writer.py:48] [87700] global_step=87700, grad_norm=5.1655778884887695, loss=0.9961017370223999 +I0831 15:34:15.001339 139540719195904 logging_writer.py:48] [87800] global_step=87800, grad_norm=5.200411319732666, loss=1.1454986333847046 +I0831 15:34:42.166128 139540727588608 logging_writer.py:48] [87900] global_step=87900, grad_norm=4.959214687347412, loss=1.0364348888397217 +I0831 15:35:09.269785 139540719195904 logging_writer.py:48] [88000] global_step=88000, grad_norm=5.038551330566406, loss=1.030794620513916 +I0831 15:35:36.359434 139540727588608 logging_writer.py:48] [88100] global_step=88100, grad_norm=5.188911437988281, loss=1.078804612159729 +I0831 15:36:03.538005 139540719195904 logging_writer.py:48] [88200] global_step=88200, grad_norm=5.192467212677002, loss=1.0401575565338135 +I0831 15:36:30.648172 139540727588608 logging_writer.py:48] [88300] global_step=88300, grad_norm=4.8265910148620605, loss=1.087786316871643 +I0831 15:36:57.744390 139540719195904 logging_writer.py:48] [88400] global_step=88400, grad_norm=4.98877477645874, loss=1.023648738861084 +I0831 15:37:24.940158 139540727588608 logging_writer.py:48] [88500] global_step=88500, grad_norm=4.843820095062256, loss=1.0654933452606201 +I0831 15:37:52.023149 139540719195904 logging_writer.py:48] [88600] global_step=88600, grad_norm=5.0491180419921875, loss=1.1243466138839722 +I0831 15:38:19.174173 139540727588608 logging_writer.py:48] [88700] global_step=88700, grad_norm=4.995851516723633, loss=1.0561825037002563 +I0831 15:38:46.336572 139540719195904 logging_writer.py:48] [88800] global_step=88800, grad_norm=5.079293251037598, loss=1.1354633569717407 +I0831 15:39:13.674318 139540727588608 logging_writer.py:48] [88900] global_step=88900, grad_norm=4.874382495880127, loss=1.0750494003295898 +I0831 15:39:40.803162 139540719195904 logging_writer.py:48] [89000] global_step=89000, grad_norm=5.3274712562561035, loss=1.0396728515625 +I0831 15:40:08.001009 139540727588608 logging_writer.py:48] [89100] global_step=89100, grad_norm=5.026710033416748, loss=1.107635736465454 +I0831 15:40:35.076878 139540719195904 logging_writer.py:48] [89200] global_step=89200, grad_norm=5.340487957000732, loss=1.1309958696365356 +I0831 15:41:02.157513 139540727588608 logging_writer.py:48] [89300] global_step=89300, grad_norm=4.976593017578125, loss=1.054032325744629 +I0831 15:41:29.330208 139540719195904 logging_writer.py:48] [89400] global_step=89400, grad_norm=5.027141571044922, loss=0.9904265999794006 +I0831 15:41:56.421278 139540727588608 logging_writer.py:48] [89500] global_step=89500, grad_norm=5.018741130828857, loss=1.0614213943481445 +I0831 15:42:23.522044 139540719195904 logging_writer.py:48] [89600] global_step=89600, grad_norm=5.032220840454102, loss=1.0859384536743164 +I0831 15:42:50.671598 139540727588608 logging_writer.py:48] [89700] global_step=89700, grad_norm=5.296989917755127, loss=1.1314268112182617 +I0831 15:43:17.757259 139540719195904 logging_writer.py:48] [89800] global_step=89800, grad_norm=4.984623432159424, loss=1.0617742538452148 +I0831 15:43:44.823745 139540727588608 logging_writer.py:48] [89900] global_step=89900, grad_norm=5.439864158630371, loss=1.1386750936508179 +I0831 15:44:12.204118 139540719195904 logging_writer.py:48] [90000] global_step=90000, grad_norm=5.298142910003662, loss=1.0591130256652832 +I0831 15:44:39.303791 139540727588608 logging_writer.py:48] [90100] global_step=90100, grad_norm=5.362574100494385, loss=1.1434476375579834 +I0831 15:45:06.406528 139540719195904 logging_writer.py:48] [90200] global_step=90200, grad_norm=4.93883752822876, loss=1.004425048828125 +I0831 15:45:33.577515 139540727588608 logging_writer.py:48] [90300] global_step=90300, grad_norm=4.978388786315918, loss=1.0280625820159912 +I0831 15:46:00.686330 139540719195904 logging_writer.py:48] [90400] global_step=90400, grad_norm=4.818467617034912, loss=1.014555811882019 +I0831 15:46:27.792379 139540727588608 logging_writer.py:48] [90500] global_step=90500, grad_norm=4.926148414611816, loss=1.0545144081115723 +I0831 15:46:54.958017 139540719195904 logging_writer.py:48] [90600] global_step=90600, grad_norm=4.867527484893799, loss=1.0103247165679932 +I0831 15:47:22.070003 139540727588608 logging_writer.py:48] [90700] global_step=90700, grad_norm=4.868866443634033, loss=1.0130285024642944 +I0831 15:47:49.241351 139540719195904 logging_writer.py:48] [90800] global_step=90800, grad_norm=5.372440338134766, loss=1.1501595973968506 +I0831 15:48:16.426282 139540727588608 logging_writer.py:48] [90900] global_step=90900, grad_norm=4.859857559204102, loss=0.9395514726638794 +I0831 15:48:43.707505 139540719195904 logging_writer.py:48] [91000] global_step=91000, grad_norm=4.980844020843506, loss=1.0504813194274902 +I0831 15:49:10.822660 139540727588608 logging_writer.py:48] [91100] global_step=91100, grad_norm=4.975203037261963, loss=1.0149034261703491 +I0831 15:49:37.977198 139540719195904 logging_writer.py:48] [91200] global_step=91200, grad_norm=5.175693988800049, loss=1.0313645601272583 +I0831 15:50:05.096682 139540727588608 logging_writer.py:48] [91300] global_step=91300, grad_norm=5.218865394592285, loss=1.1493602991104126 +I0831 15:50:32.214063 139540719195904 logging_writer.py:48] [91400] global_step=91400, grad_norm=5.179586887359619, loss=1.0151209831237793 +I0831 15:50:59.392093 139540727588608 logging_writer.py:48] [91500] global_step=91500, grad_norm=5.208112716674805, loss=1.016558051109314 +I0831 15:51:26.486468 139540719195904 logging_writer.py:48] [91600] global_step=91600, grad_norm=5.002886772155762, loss=1.054046630859375 +I0831 15:51:53.566093 139540727588608 logging_writer.py:48] [91700] global_step=91700, grad_norm=4.868175029754639, loss=1.022315502166748 +I0831 15:52:20.750478 139540719195904 logging_writer.py:48] [91800] global_step=91800, grad_norm=5.098211765289307, loss=1.1375863552093506 +I0831 15:52:47.853440 139540727588608 logging_writer.py:48] [91900] global_step=91900, grad_norm=5.053808689117432, loss=1.114558219909668 +I0831 15:53:14.959598 139540719195904 logging_writer.py:48] [92000] global_step=92000, grad_norm=4.597872257232666, loss=0.9242792129516602 +I0831 15:53:42.348778 139540727588608 logging_writer.py:48] [92100] global_step=92100, grad_norm=5.125771522521973, loss=1.0540127754211426 +I0831 15:54:09.440109 139540719195904 logging_writer.py:48] [92200] global_step=92200, grad_norm=5.0782551765441895, loss=1.0456550121307373 +I0831 15:54:36.559237 139540727588608 logging_writer.py:48] [92300] global_step=92300, grad_norm=4.904450416564941, loss=1.0198277235031128 +I0831 15:55:03.732775 139540719195904 logging_writer.py:48] [92400] global_step=92400, grad_norm=5.063963890075684, loss=1.078228235244751 +I0831 15:55:30.835637 139540727588608 logging_writer.py:48] [92500] global_step=92500, grad_norm=4.838295936584473, loss=0.9733790755271912 +I0831 15:55:57.940678 139540719195904 logging_writer.py:48] [92600] global_step=92600, grad_norm=5.093165874481201, loss=1.0777785778045654 +I0831 15:56:25.135798 139540727588608 logging_writer.py:48] [92700] global_step=92700, grad_norm=4.637843132019043, loss=0.9443497657775879 +I0831 15:56:52.272713 139540719195904 logging_writer.py:48] [92800] global_step=92800, grad_norm=4.997882843017578, loss=1.1145237684249878 +I0831 15:57:19.370851 139540727588608 logging_writer.py:48] [92900] global_step=92900, grad_norm=5.226167678833008, loss=1.1063461303710938 +I0831 15:57:46.539185 139540719195904 logging_writer.py:48] [93000] global_step=93000, grad_norm=4.769576549530029, loss=1.0636565685272217 +I0831 15:58:13.851995 139540727588608 logging_writer.py:48] [93100] global_step=93100, grad_norm=4.958584785461426, loss=1.0632445812225342 +I0831 15:58:40.936592 139540719195904 logging_writer.py:48] [93200] global_step=93200, grad_norm=4.811371326446533, loss=1.0100610256195068 +I0831 15:59:08.098115 139540727588608 logging_writer.py:48] [93300] global_step=93300, grad_norm=5.17028284072876, loss=1.170006513595581 +I0831 15:59:35.234508 139540719195904 logging_writer.py:48] [93400] global_step=93400, grad_norm=4.854822635650635, loss=1.0168758630752563 +I0831 16:00:02.322394 139540727588608 logging_writer.py:48] [93500] global_step=93500, grad_norm=5.0155510902404785, loss=1.0143861770629883 +I0831 16:00:29.481657 139540719195904 logging_writer.py:48] [93600] global_step=93600, grad_norm=5.473679065704346, loss=1.0423561334609985 +I0831 16:00:56.544569 139540727588608 logging_writer.py:48] [93700] global_step=93700, grad_norm=4.9615349769592285, loss=1.019302248954773 +I0831 16:01:23.648015 139540719195904 logging_writer.py:48] [93800] global_step=93800, grad_norm=4.892245292663574, loss=1.0526599884033203 +I0831 16:01:50.789263 139540727588608 logging_writer.py:48] [93900] global_step=93900, grad_norm=5.51822566986084, loss=1.1046515703201294 +I0831 16:02:17.889570 139540719195904 logging_writer.py:48] [94000] global_step=94000, grad_norm=4.8373799324035645, loss=1.022350549697876 +I0831 16:02:44.980992 139540727588608 logging_writer.py:48] [94100] global_step=94100, grad_norm=4.932233810424805, loss=0.958229660987854 +I0831 16:03:12.348929 139540719195904 logging_writer.py:48] [94200] global_step=94200, grad_norm=5.02880334854126, loss=0.9583393335342407 +I0831 16:03:39.430106 139540727588608 logging_writer.py:48] [94300] global_step=94300, grad_norm=4.7607340812683105, loss=0.9972282648086548 +I0831 16:04:06.552648 139540719195904 logging_writer.py:48] [94400] global_step=94400, grad_norm=4.854794025421143, loss=0.9720350503921509 +I0831 16:04:33.719804 139540727588608 logging_writer.py:48] [94500] global_step=94500, grad_norm=4.755732536315918, loss=0.9436827301979065 +I0831 16:05:00.818746 139540719195904 logging_writer.py:48] [94600] global_step=94600, grad_norm=4.999763488769531, loss=1.1419841051101685 +I0831 16:05:06.403109 139757377230016 spec.py:333] Evaluating on the training split. +I0831 16:05:14.202847 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 16:05:24.495621 139757377230016 spec.py:363] Evaluating on the test split. +I0831 16:05:25.385649 139757377230016 submission_runner.py:516] Time since start: 26335.72s, Step: 94622, {'train/accuracy': Array(0.92352915, dtype=float32), 'train/loss': Array(0.27080315, dtype=float32), 'validation/accuracy': Array(0.74688, dtype=float32), 'validation/loss': Array(1.0591964, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62310004, dtype=float32), 'test/loss': Array(1.8336954, dtype=float32), 'test/num_examples': 10000, 'score': 26001.21587538719, 'total_duration': 26335.716311454773, 'accumulated_submission_time': 26001.21587538719, 'accumulated_eval_time': 332.67340207099915, 'accumulated_logging_time': 1.0685644149780273} +I0831 16:05:25.455092 139540727588608 logging_writer.py:48] [94622] accumulated_eval_time=332.673, accumulated_logging_time=1.06856, accumulated_submission_time=26001.2, global_step=94622, preemption_count=0, score=26001.2, test/accuracy=0.6231000423431396, test/loss=1.833695411682129, test/num_examples=10000, total_duration=26335.7, train/accuracy=0.9235291481018066, train/loss=0.27080315351486206, validation/accuracy=0.7468799948692322, validation/loss=1.0591963529586792, validation/num_examples=50000 +I0831 16:05:46.943280 139540719195904 logging_writer.py:48] [94700] global_step=94700, grad_norm=4.969169616699219, loss=1.085073709487915 +I0831 16:06:14.069524 139540727588608 logging_writer.py:48] [94800] global_step=94800, grad_norm=4.727773666381836, loss=1.0311708450317383 +I0831 16:06:41.166256 139540719195904 logging_writer.py:48] [94900] global_step=94900, grad_norm=4.923572063446045, loss=1.0282840728759766 +I0831 16:07:08.276054 139540727588608 logging_writer.py:48] [95000] global_step=95000, grad_norm=5.050494194030762, loss=1.049731731414795 +I0831 16:07:35.408810 139540719195904 logging_writer.py:48] [95100] global_step=95100, grad_norm=4.819559097290039, loss=1.0059688091278076 +I0831 16:08:02.527615 139540727588608 logging_writer.py:48] [95200] global_step=95200, grad_norm=4.904679775238037, loss=1.0477782487869263 +I0831 16:08:29.805633 139540719195904 logging_writer.py:48] [95300] global_step=95300, grad_norm=5.053136348724365, loss=1.071906328201294 +I0831 16:08:56.977482 139540727588608 logging_writer.py:48] [95400] global_step=95400, grad_norm=4.947798728942871, loss=1.0192687511444092 +I0831 16:09:24.092738 139540719195904 logging_writer.py:48] [95500] global_step=95500, grad_norm=4.880970001220703, loss=1.0594422817230225 +I0831 16:09:51.197499 139540727588608 logging_writer.py:48] [95600] global_step=95600, grad_norm=4.917267799377441, loss=1.1019580364227295 +I0831 16:10:18.352635 139540719195904 logging_writer.py:48] [95700] global_step=95700, grad_norm=4.710488796234131, loss=0.9927248358726501 +I0831 16:10:45.486602 139540727588608 logging_writer.py:48] [95800] global_step=95800, grad_norm=4.823322296142578, loss=1.068575382232666 +I0831 16:11:12.585715 139540719195904 logging_writer.py:48] [95900] global_step=95900, grad_norm=5.090761184692383, loss=1.0772392749786377 +I0831 16:11:39.798937 139540727588608 logging_writer.py:48] [96000] global_step=96000, grad_norm=4.746954917907715, loss=1.0085936784744263 +I0831 16:12:06.930952 139540719195904 logging_writer.py:48] [96100] global_step=96100, grad_norm=4.627437591552734, loss=1.0695043802261353 +I0831 16:12:34.027774 139540727588608 logging_writer.py:48] [96200] global_step=96200, grad_norm=4.847785472869873, loss=1.0295919179916382 +I0831 16:13:01.421063 139540719195904 logging_writer.py:48] [96300] global_step=96300, grad_norm=4.771572113037109, loss=0.9809989333152771 +I0831 16:13:28.519160 139540727588608 logging_writer.py:48] [96400] global_step=96400, grad_norm=4.886623859405518, loss=1.027539849281311 +I0831 16:13:55.650367 139540719195904 logging_writer.py:48] [96500] global_step=96500, grad_norm=5.030670166015625, loss=1.0479248762130737 +I0831 16:14:22.966934 139540727588608 logging_writer.py:48] [96600] global_step=96600, grad_norm=5.052288055419922, loss=1.0560775995254517 +I0831 16:14:50.090753 139540719195904 logging_writer.py:48] [96700] global_step=96700, grad_norm=5.12360954284668, loss=1.0479873418807983 +I0831 16:15:17.184178 139540727588608 logging_writer.py:48] [96800] global_step=96800, grad_norm=5.0657267570495605, loss=1.1321897506713867 +I0831 16:15:44.357843 139540719195904 logging_writer.py:48] [96900] global_step=96900, grad_norm=4.683468818664551, loss=0.9537731409072876 +I0831 16:16:11.509917 139540727588608 logging_writer.py:48] [97000] global_step=97000, grad_norm=4.904040813446045, loss=1.0468429327011108 +I0831 16:16:38.639991 139540719195904 logging_writer.py:48] [97100] global_step=97100, grad_norm=4.719038486480713, loss=1.0360827445983887 +I0831 16:17:05.843890 139540727588608 logging_writer.py:48] [97200] global_step=97200, grad_norm=5.009496212005615, loss=1.04325532913208 +I0831 16:17:32.950872 139540719195904 logging_writer.py:48] [97300] global_step=97300, grad_norm=4.81718635559082, loss=0.9815117716789246 +I0831 16:18:00.256120 139540727588608 logging_writer.py:48] [97400] global_step=97400, grad_norm=5.045836448669434, loss=1.1241886615753174 +I0831 16:18:27.389396 139540719195904 logging_writer.py:48] [97500] global_step=97500, grad_norm=5.0923871994018555, loss=1.1006451845169067 +I0831 16:18:54.479058 139540727588608 logging_writer.py:48] [97600] global_step=97600, grad_norm=5.074033260345459, loss=1.0501879453659058 +I0831 16:19:21.578035 139540719195904 logging_writer.py:48] [97700] global_step=97700, grad_norm=5.3151092529296875, loss=1.067679762840271 +I0831 16:19:48.708284 139540727588608 logging_writer.py:48] [97800] global_step=97800, grad_norm=4.849456787109375, loss=1.0440030097961426 +I0831 16:20:15.828990 139540719195904 logging_writer.py:48] [97900] global_step=97900, grad_norm=5.065085411071777, loss=1.1033505201339722 +I0831 16:20:42.920597 139540727588608 logging_writer.py:48] [98000] global_step=98000, grad_norm=4.736199855804443, loss=0.9836573600769043 +I0831 16:21:10.124824 139540719195904 logging_writer.py:48] [98100] global_step=98100, grad_norm=5.020537853240967, loss=0.9930371642112732 +I0831 16:21:37.244208 139540727588608 logging_writer.py:48] [98200] global_step=98200, grad_norm=5.194108009338379, loss=1.0423190593719482 +I0831 16:22:04.363484 139540719195904 logging_writer.py:48] [98300] global_step=98300, grad_norm=5.040644645690918, loss=1.0631153583526611 +I0831 16:22:31.534114 139540727588608 logging_writer.py:48] [98400] global_step=98400, grad_norm=4.684607982635498, loss=0.9492117166519165 +I0831 16:22:58.849356 139540719195904 logging_writer.py:48] [98500] global_step=98500, grad_norm=4.94767427444458, loss=1.062397837638855 +I0831 16:23:25.964593 139540727588608 logging_writer.py:48] [98600] global_step=98600, grad_norm=5.144436359405518, loss=1.0962319374084473 +I0831 16:23:53.121670 139540719195904 logging_writer.py:48] [98700] global_step=98700, grad_norm=4.696563243865967, loss=0.9837020635604858 +I0831 16:24:20.203274 139540727588608 logging_writer.py:48] [98800] global_step=98800, grad_norm=4.990105628967285, loss=1.107542872428894 +I0831 16:24:47.309614 139540719195904 logging_writer.py:48] [98900] global_step=98900, grad_norm=4.737512111663818, loss=0.9301794171333313 +I0831 16:25:14.464158 139540727588608 logging_writer.py:48] [99000] global_step=99000, grad_norm=5.161459922790527, loss=1.1184072494506836 +I0831 16:25:41.545909 139540719195904 logging_writer.py:48] [99100] global_step=99100, grad_norm=5.052849769592285, loss=1.0586446523666382 +I0831 16:26:08.652418 139540727588608 logging_writer.py:48] [99200] global_step=99200, grad_norm=4.991428375244141, loss=0.9945316314697266 +I0831 16:26:35.829212 139540719195904 logging_writer.py:48] [99300] global_step=99300, grad_norm=4.913521766662598, loss=1.0111422538757324 +I0831 16:27:02.925789 139540727588608 logging_writer.py:48] [99400] global_step=99400, grad_norm=4.920143127441406, loss=0.9721000790596008 +I0831 16:27:30.269571 139540719195904 logging_writer.py:48] [99500] global_step=99500, grad_norm=5.242607116699219, loss=1.0570776462554932 +I0831 16:27:57.446014 139540727588608 logging_writer.py:48] [99600] global_step=99600, grad_norm=5.134932994842529, loss=1.0076546669006348 +I0831 16:28:24.552374 139540719195904 logging_writer.py:48] [99700] global_step=99700, grad_norm=4.983876705169678, loss=1.0705803632736206 +I0831 16:28:51.645261 139540727588608 logging_writer.py:48] [99800] global_step=99800, grad_norm=4.49092435836792, loss=0.9185853004455566 +I0831 16:29:18.805963 139540719195904 logging_writer.py:48] [99900] global_step=99900, grad_norm=4.947582244873047, loss=1.061302900314331 +I0831 16:29:45.910203 139540727588608 logging_writer.py:48] [100000] global_step=100000, grad_norm=5.179671287536621, loss=1.046684980392456 +I0831 16:30:12.991812 139540719195904 logging_writer.py:48] [100100] global_step=100100, grad_norm=5.167572975158691, loss=1.0297633409500122 +I0831 16:30:40.164970 139540727588608 logging_writer.py:48] [100200] global_step=100200, grad_norm=4.947312831878662, loss=1.0366803407669067 +I0831 16:31:07.292695 139540719195904 logging_writer.py:48] [100300] global_step=100300, grad_norm=5.108885765075684, loss=1.1061465740203857 +I0831 16:31:34.385422 139540727588608 logging_writer.py:48] [100400] global_step=100400, grad_norm=5.1959452629089355, loss=1.1637647151947021 +I0831 16:32:01.535565 139540719195904 logging_writer.py:48] [100500] global_step=100500, grad_norm=5.007674217224121, loss=1.0667779445648193 +I0831 16:32:28.889877 139540727588608 logging_writer.py:48] [100600] global_step=100600, grad_norm=4.846701622009277, loss=1.0005419254302979 +I0831 16:32:55.968062 139540719195904 logging_writer.py:48] [100700] global_step=100700, grad_norm=4.866466999053955, loss=1.0246458053588867 +I0831 16:33:23.137314 139540727588608 logging_writer.py:48] [100800] global_step=100800, grad_norm=4.846804618835449, loss=0.9701067209243774 +I0831 16:33:50.216752 139540719195904 logging_writer.py:48] [100900] global_step=100900, grad_norm=4.773890018463135, loss=1.035614252090454 +I0831 16:34:17.302778 139540727588608 logging_writer.py:48] [101000] global_step=101000, grad_norm=5.132377624511719, loss=1.0424933433532715 +I0831 16:34:44.461640 139540719195904 logging_writer.py:48] [101100] global_step=101100, grad_norm=5.115495681762695, loss=1.0250136852264404 +I0831 16:35:11.593574 139540727588608 logging_writer.py:48] [101200] global_step=101200, grad_norm=4.720860958099365, loss=0.9821350574493408 +I0831 16:35:38.680662 139540719195904 logging_writer.py:48] [101300] global_step=101300, grad_norm=5.073286056518555, loss=1.146422266960144 +I0831 16:36:05.824313 139540727588608 logging_writer.py:48] [101400] global_step=101400, grad_norm=4.447088241577148, loss=0.9089692831039429 +I0831 16:36:32.900595 139540719195904 logging_writer.py:48] [101500] global_step=101500, grad_norm=5.157322883605957, loss=0.9695748090744019 +I0831 16:36:59.982389 139540727588608 logging_writer.py:48] [101600] global_step=101600, grad_norm=4.663829803466797, loss=0.9595123529434204 +I0831 16:37:27.378483 139540719195904 logging_writer.py:48] [101700] global_step=101700, grad_norm=5.1077165603637695, loss=1.0426952838897705 +I0831 16:37:54.503994 139540727588608 logging_writer.py:48] [101800] global_step=101800, grad_norm=4.885904312133789, loss=1.0386500358581543 +I0831 16:38:21.596487 139540719195904 logging_writer.py:48] [101900] global_step=101900, grad_norm=4.8586320877075195, loss=0.9904083013534546 +I0831 16:38:41.528341 139757377230016 spec.py:333] Evaluating on the training split. +I0831 16:38:48.568138 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 16:38:59.131909 139757377230016 spec.py:363] Evaluating on the test split. +I0831 16:39:00.023081 139757377230016 submission_runner.py:516] Time since start: 28350.36s, Step: 101975, {'train/accuracy': Array(0.92490435, dtype=float32), 'train/loss': Array(0.26681402, dtype=float32), 'validation/accuracy': Array(0.74736, dtype=float32), 'validation/loss': Array(1.0602218, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62320006, dtype=float32), 'test/loss': Array(1.8342712, dtype=float32), 'test/num_examples': 10000, 'score': 27997.165731191635, 'total_duration': 28350.357009649277, 'accumulated_submission_time': 27997.165731191635, 'accumulated_eval_time': 351.16429591178894, 'accumulated_logging_time': 1.2089686393737793} +I0831 16:39:00.077483 139540727588608 logging_writer.py:48] [101975] accumulated_eval_time=351.164, accumulated_logging_time=1.20897, accumulated_submission_time=27997.2, global_step=101975, preemption_count=0, score=27997.2, test/accuracy=0.6232000589370728, test/loss=1.8342711925506592, test/num_examples=10000, total_duration=28350.4, train/accuracy=0.9249043464660645, train/loss=0.2668140232563019, validation/accuracy=0.7473599910736084, validation/loss=1.060221791267395, validation/num_examples=50000 +I0831 16:39:07.268846 139540719195904 logging_writer.py:48] [102000] global_step=102000, grad_norm=4.645664215087891, loss=0.9264371991157532 +I0831 16:39:34.352988 139540727588608 logging_writer.py:48] [102100] global_step=102100, grad_norm=4.621421813964844, loss=1.036573052406311 +I0831 16:40:01.427694 139540719195904 logging_writer.py:48] [102200] global_step=102200, grad_norm=5.00708532333374, loss=0.9960358142852783 +I0831 16:40:28.603183 139540727588608 logging_writer.py:48] [102300] global_step=102300, grad_norm=4.633608341217041, loss=0.9069032073020935 +I0831 16:40:55.718580 139540719195904 logging_writer.py:48] [102400] global_step=102400, grad_norm=4.9907331466674805, loss=1.0679248571395874 +I0831 16:41:22.833091 139540727588608 logging_writer.py:48] [102500] global_step=102500, grad_norm=5.270980358123779, loss=1.088824987411499 +I0831 16:41:49.984489 139540719195904 logging_writer.py:48] [102600] global_step=102600, grad_norm=4.9403862953186035, loss=0.9794908761978149 +I0831 16:42:17.088757 139540727588608 logging_writer.py:48] [102700] global_step=102700, grad_norm=4.885138034820557, loss=0.9967936277389526 +I0831 16:42:44.418525 139540719195904 logging_writer.py:48] [102800] global_step=102800, grad_norm=5.239289283752441, loss=1.1094340085983276 +I0831 16:43:11.575778 139540727588608 logging_writer.py:48] [102900] global_step=102900, grad_norm=5.300139904022217, loss=1.0616694688796997 +I0831 16:43:38.675004 139540719195904 logging_writer.py:48] [103000] global_step=103000, grad_norm=4.951668739318848, loss=0.9739927053451538 +I0831 16:44:05.776321 139540727588608 logging_writer.py:48] [103100] global_step=103100, grad_norm=5.1962409019470215, loss=1.0499143600463867 +I0831 16:44:32.945577 139540719195904 logging_writer.py:48] [103200] global_step=103200, grad_norm=4.512073516845703, loss=0.9437228441238403 +I0831 16:45:00.034543 139540727588608 logging_writer.py:48] [103300] global_step=103300, grad_norm=4.836172103881836, loss=1.1054214239120483 +I0831 16:45:27.109035 139540719195904 logging_writer.py:48] [103400] global_step=103400, grad_norm=4.734205722808838, loss=0.9271208643913269 +I0831 16:45:54.261632 139540727588608 logging_writer.py:48] [103500] global_step=103500, grad_norm=5.192615032196045, loss=1.0347824096679688 +I0831 16:46:21.348295 139540719195904 logging_writer.py:48] [103600] global_step=103600, grad_norm=4.911137104034424, loss=1.0092694759368896 +I0831 16:46:48.434552 139540727588608 logging_writer.py:48] [103700] global_step=103700, grad_norm=4.896384239196777, loss=0.9757914543151855 +I0831 16:47:15.795584 139540719195904 logging_writer.py:48] [103800] global_step=103800, grad_norm=4.996157646179199, loss=0.9974159002304077 +I0831 16:47:42.899654 139540727588608 logging_writer.py:48] [103900] global_step=103900, grad_norm=4.741395473480225, loss=1.0038204193115234 +I0831 16:48:10.011745 139540719195904 logging_writer.py:48] [104000] global_step=104000, grad_norm=5.213545799255371, loss=1.0410833358764648 +I0831 16:48:37.177361 139540727588608 logging_writer.py:48] [104100] global_step=104100, grad_norm=5.306631565093994, loss=1.1378026008605957 +I0831 16:49:04.288126 139540719195904 logging_writer.py:48] [104200] global_step=104200, grad_norm=5.152668476104736, loss=1.0453546047210693 +I0831 16:49:31.382550 139540727588608 logging_writer.py:48] [104300] global_step=104300, grad_norm=5.3145833015441895, loss=1.0001747608184814 +I0831 16:49:58.522059 139540719195904 logging_writer.py:48] [104400] global_step=104400, grad_norm=5.121182918548584, loss=1.0586731433868408 +I0831 16:50:25.617025 139540727588608 logging_writer.py:48] [104500] global_step=104500, grad_norm=4.836161136627197, loss=0.9595435857772827 +I0831 16:50:52.704445 139540719195904 logging_writer.py:48] [104600] global_step=104600, grad_norm=5.1765923500061035, loss=1.1110246181488037 +I0831 16:51:19.858588 139540727588608 logging_writer.py:48] [104700] global_step=104700, grad_norm=4.946266174316406, loss=0.9800402522087097 +I0831 16:51:46.957661 139540719195904 logging_writer.py:48] [104800] global_step=104800, grad_norm=5.329427242279053, loss=1.0546252727508545 +I0831 16:52:14.222987 139540727588608 logging_writer.py:48] [104900] global_step=104900, grad_norm=5.004871368408203, loss=0.9639538526535034 +I0831 16:52:41.378633 139540719195904 logging_writer.py:48] [105000] global_step=105000, grad_norm=4.712414264678955, loss=0.9864548444747925 +I0831 16:53:08.482191 139540727588608 logging_writer.py:48] [105100] global_step=105100, grad_norm=4.942111492156982, loss=1.0402873754501343 +I0831 16:53:35.557294 139540719195904 logging_writer.py:48] [105200] global_step=105200, grad_norm=4.7874932289123535, loss=1.033484697341919 +I0831 16:54:02.718307 139540727588608 logging_writer.py:48] [105300] global_step=105300, grad_norm=5.003039360046387, loss=1.0254104137420654 +I0831 16:54:29.818155 139540719195904 logging_writer.py:48] [105400] global_step=105400, grad_norm=5.204089641571045, loss=1.1215821504592896 +I0831 16:54:56.923418 139540727588608 logging_writer.py:48] [105500] global_step=105500, grad_norm=4.9027814865112305, loss=1.0522323846817017 +I0831 16:55:24.060464 139540719195904 logging_writer.py:48] [105600] global_step=105600, grad_norm=4.746074199676514, loss=0.9647051095962524 +I0831 16:55:51.151500 139540727588608 logging_writer.py:48] [105700] global_step=105700, grad_norm=4.659722328186035, loss=1.0360445976257324 +I0831 16:56:18.238730 139540719195904 logging_writer.py:48] [105800] global_step=105800, grad_norm=5.023189544677734, loss=1.0234198570251465 +I0831 16:56:45.385236 139540727588608 logging_writer.py:48] [105900] global_step=105900, grad_norm=5.107797145843506, loss=1.090755820274353 +I0831 16:57:12.688092 139540719195904 logging_writer.py:48] [106000] global_step=106000, grad_norm=5.265072345733643, loss=1.082385778427124 +I0831 16:57:39.758769 139540727588608 logging_writer.py:48] [106100] global_step=106100, grad_norm=4.701483249664307, loss=0.9684467315673828 +I0831 16:58:06.930704 139540719195904 logging_writer.py:48] [106200] global_step=106200, grad_norm=4.87825345993042, loss=1.0006011724472046 +I0831 16:58:34.012195 139540727588608 logging_writer.py:48] [106300] global_step=106300, grad_norm=4.766509056091309, loss=0.9941439628601074 +I0831 16:59:01.086994 139540719195904 logging_writer.py:48] [106400] global_step=106400, grad_norm=4.907028675079346, loss=1.0326787233352661 +I0831 16:59:28.246299 139540727588608 logging_writer.py:48] [106500] global_step=106500, grad_norm=4.8752360343933105, loss=1.0046732425689697 +I0831 16:59:55.322720 139540719195904 logging_writer.py:48] [106600] global_step=106600, grad_norm=5.215755462646484, loss=1.0618667602539062 +I0831 17:00:22.433940 139540727588608 logging_writer.py:48] [106700] global_step=106700, grad_norm=4.768545627593994, loss=1.0029346942901611 +I0831 17:00:49.561756 139540719195904 logging_writer.py:48] [106800] global_step=106800, grad_norm=4.507075786590576, loss=0.8717372417449951 +I0831 17:01:16.686484 139540727588608 logging_writer.py:48] [106900] global_step=106900, grad_norm=4.951524257659912, loss=1.019490122795105 +I0831 17:01:43.768046 139540719195904 logging_writer.py:48] [107000] global_step=107000, grad_norm=4.987048149108887, loss=1.0884076356887817 +I0831 17:02:11.133291 139540727588608 logging_writer.py:48] [107100] global_step=107100, grad_norm=4.901164531707764, loss=0.9717271327972412 +I0831 17:02:38.245088 139540719195904 logging_writer.py:48] [107200] global_step=107200, grad_norm=4.888548851013184, loss=1.0147302150726318 +I0831 17:03:05.353119 139540727588608 logging_writer.py:48] [107300] global_step=107300, grad_norm=4.986262321472168, loss=1.03008234500885 +I0831 17:03:32.492737 139540719195904 logging_writer.py:48] [107400] global_step=107400, grad_norm=5.101861476898193, loss=1.0646132230758667 +I0831 17:03:59.570688 139540727588608 logging_writer.py:48] [107500] global_step=107500, grad_norm=4.7701735496521, loss=0.9912980794906616 +I0831 17:04:26.649103 139540719195904 logging_writer.py:48] [107600] global_step=107600, grad_norm=5.021111011505127, loss=1.0799189805984497 +I0831 17:04:53.810723 139540727588608 logging_writer.py:48] [107700] global_step=107700, grad_norm=5.038873195648193, loss=0.9219493269920349 +I0831 17:05:20.932728 139540719195904 logging_writer.py:48] [107800] global_step=107800, grad_norm=4.963326930999756, loss=1.024501919746399 +I0831 17:05:48.012313 139540727588608 logging_writer.py:48] [107900] global_step=107900, grad_norm=5.329412460327148, loss=1.0346424579620361 +I0831 17:06:15.157060 139540719195904 logging_writer.py:48] [108000] global_step=108000, grad_norm=5.004898548126221, loss=1.0173414945602417 +I0831 17:06:42.487087 139540727588608 logging_writer.py:48] [108100] global_step=108100, grad_norm=4.996073246002197, loss=1.030786156654358 +I0831 17:07:09.587914 139540719195904 logging_writer.py:48] [108200] global_step=108200, grad_norm=5.000154495239258, loss=0.9875686168670654 +I0831 17:07:36.703140 139540727588608 logging_writer.py:48] [108300] global_step=108300, grad_norm=4.57889461517334, loss=0.9480152130126953 +I0831 17:08:03.798787 139540719195904 logging_writer.py:48] [108400] global_step=108400, grad_norm=5.116833686828613, loss=1.1127240657806396 +I0831 17:08:30.907722 139540727588608 logging_writer.py:48] [108500] global_step=108500, grad_norm=4.964255332946777, loss=1.0086930990219116 +I0831 17:08:58.106080 139540719195904 logging_writer.py:48] [108600] global_step=108600, grad_norm=5.029326438903809, loss=1.071121335029602 +I0831 17:09:25.199064 139540727588608 logging_writer.py:48] [108700] global_step=108700, grad_norm=4.657139778137207, loss=0.8972160220146179 +I0831 17:09:52.287435 139540719195904 logging_writer.py:48] [108800] global_step=108800, grad_norm=4.812421798706055, loss=0.9355010390281677 +I0831 17:10:19.461660 139540727588608 logging_writer.py:48] [108900] global_step=108900, grad_norm=4.844290733337402, loss=0.9831030368804932 +I0831 17:10:46.556585 139540719195904 logging_writer.py:48] [109000] global_step=109000, grad_norm=5.0133819580078125, loss=1.0371623039245605 +I0831 17:11:13.634555 139540727588608 logging_writer.py:48] [109100] global_step=109100, grad_norm=5.143393516540527, loss=1.0966556072235107 +I0831 17:11:41.012585 139540719195904 logging_writer.py:48] [109200] global_step=109200, grad_norm=5.0216240882873535, loss=0.9975601434707642 +I0831 17:12:08.106847 139540727588608 logging_writer.py:48] [109300] global_step=109300, grad_norm=4.777124404907227, loss=0.985363781452179 +I0831 17:12:16.112768 139757377230016 spec.py:333] Evaluating on the training split. +I0831 17:12:22.956387 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 17:12:33.481148 139757377230016 spec.py:363] Evaluating on the test split. +I0831 17:12:34.370992 139757377230016 submission_runner.py:516] Time since start: 30364.71s, Step: 109331, {'train/accuracy': Array(0.9286312, dtype=float32), 'train/loss': Array(0.24832252, dtype=float32), 'validation/accuracy': Array(0.7486, dtype=float32), 'validation/loss': Array(1.0602936, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6223, dtype=float32), 'test/loss': Array(1.8452394, dtype=float32), 'test/num_examples': 10000, 'score': 29993.137478351593, 'total_duration': 30364.70661520958, 'accumulated_submission_time': 29993.137478351593, 'accumulated_eval_time': 369.42036747932434, 'accumulated_logging_time': 1.274552583694458} +I0831 17:12:34.427276 139540719195904 logging_writer.py:48] [109331] accumulated_eval_time=369.42, accumulated_logging_time=1.27455, accumulated_submission_time=29993.1, global_step=109331, preemption_count=0, score=29993.1, test/accuracy=0.6223000288009644, test/loss=1.8452394008636475, test/num_examples=10000, total_duration=30364.7, train/accuracy=0.9286311864852905, train/loss=0.2483225166797638, validation/accuracy=0.7486000061035156, validation/loss=1.0602935552597046, validation/num_examples=50000 +I0831 17:12:53.528442 139540727588608 logging_writer.py:48] [109400] global_step=109400, grad_norm=4.96727991104126, loss=1.004333257675171 +I0831 17:13:20.646306 139540719195904 logging_writer.py:48] [109500] global_step=109500, grad_norm=4.957625389099121, loss=1.0130369663238525 +I0831 17:13:47.707601 139540727588608 logging_writer.py:48] [109600] global_step=109600, grad_norm=5.1412787437438965, loss=0.997840166091919 +I0831 17:14:14.833042 139540719195904 logging_writer.py:48] [109700] global_step=109700, grad_norm=4.83319091796875, loss=1.0193136930465698 +I0831 17:14:41.978091 139540727588608 logging_writer.py:48] [109800] global_step=109800, grad_norm=4.864040851593018, loss=0.9834492802619934 +I0831 17:15:09.066032 139540719195904 logging_writer.py:48] [109900] global_step=109900, grad_norm=4.790924549102783, loss=0.9928283095359802 +I0831 17:15:36.169919 139540727588608 logging_writer.py:48] [110000] global_step=110000, grad_norm=4.717838764190674, loss=1.0193734169006348 +I0831 17:16:03.358652 139540719195904 logging_writer.py:48] [110100] global_step=110100, grad_norm=4.740447998046875, loss=0.9752176403999329 +I0831 17:16:30.442852 139540727588608 logging_writer.py:48] [110200] global_step=110200, grad_norm=4.813823223114014, loss=1.047119140625 +I0831 17:16:57.761467 139540719195904 logging_writer.py:48] [110300] global_step=110300, grad_norm=5.02108907699585, loss=1.0582752227783203 +I0831 17:17:24.932603 139540727588608 logging_writer.py:48] [110400] global_step=110400, grad_norm=4.8421831130981445, loss=1.0451836585998535 +I0831 17:17:52.018300 139540719195904 logging_writer.py:48] [110500] global_step=110500, grad_norm=4.8494954109191895, loss=0.9751569032669067 +I0831 17:18:19.129918 139540727588608 logging_writer.py:48] [110600] global_step=110600, grad_norm=4.3335795402526855, loss=0.817509651184082 +I0831 17:18:46.298301 139540719195904 logging_writer.py:48] [110700] global_step=110700, grad_norm=4.839282035827637, loss=1.011615514755249 +I0831 17:19:13.396821 139540727588608 logging_writer.py:48] [110800] global_step=110800, grad_norm=4.907031059265137, loss=1.0448641777038574 +I0831 17:19:40.536155 139540719195904 logging_writer.py:48] [110900] global_step=110900, grad_norm=5.065103054046631, loss=1.057963490486145 +I0831 17:20:07.686022 139540727588608 logging_writer.py:48] [111000] global_step=111000, grad_norm=5.044705390930176, loss=0.9934645891189575 +I0831 17:20:34.801557 139540719195904 logging_writer.py:48] [111100] global_step=111100, grad_norm=4.817972660064697, loss=1.059714913368225 +I0831 17:21:01.937311 139540727588608 logging_writer.py:48] [111200] global_step=111200, grad_norm=4.928931713104248, loss=0.9584872126579285 +I0831 17:21:29.109231 139540719195904 logging_writer.py:48] [111300] global_step=111300, grad_norm=5.229252815246582, loss=1.0294744968414307 +I0831 17:21:56.430039 139540727588608 logging_writer.py:48] [111400] global_step=111400, grad_norm=5.017659664154053, loss=1.0524282455444336 +I0831 17:22:23.530830 139540719195904 logging_writer.py:48] [111500] global_step=111500, grad_norm=4.848768711090088, loss=0.956535816192627 +I0831 17:22:50.741851 139540727588608 logging_writer.py:48] [111600] global_step=111600, grad_norm=4.630928993225098, loss=1.0005667209625244 +I0831 17:23:17.822840 139540719195904 logging_writer.py:48] [111700] global_step=111700, grad_norm=5.052680015563965, loss=1.10763418674469 +I0831 17:23:44.921309 139540727588608 logging_writer.py:48] [111800] global_step=111800, grad_norm=5.122849464416504, loss=1.0270577669143677 +I0831 17:24:12.073651 139540719195904 logging_writer.py:48] [111900] global_step=111900, grad_norm=4.810207843780518, loss=0.957984447479248 +I0831 17:24:39.194191 139540727588608 logging_writer.py:48] [112000] global_step=112000, grad_norm=4.718740940093994, loss=1.058193564414978 +I0831 17:25:06.293487 139540719195904 logging_writer.py:48] [112100] global_step=112100, grad_norm=4.772493839263916, loss=0.9249016046524048 +I0831 17:25:33.454821 139540727588608 logging_writer.py:48] [112200] global_step=112200, grad_norm=4.739798545837402, loss=1.0316896438598633 +I0831 17:26:00.568971 139540719195904 logging_writer.py:48] [112300] global_step=112300, grad_norm=4.75501012802124, loss=0.9556118249893188 +I0831 17:26:27.903811 139540727588608 logging_writer.py:48] [112400] global_step=112400, grad_norm=5.378325462341309, loss=1.0141996145248413 +I0831 17:26:55.039561 139540719195904 logging_writer.py:48] [112500] global_step=112500, grad_norm=4.8861212730407715, loss=0.9980276226997375 +I0831 17:27:22.122788 139540727588608 logging_writer.py:48] [112600] global_step=112600, grad_norm=4.767139911651611, loss=0.9572533369064331 +I0831 17:27:49.237812 139540719195904 logging_writer.py:48] [112700] global_step=112700, grad_norm=5.038192272186279, loss=1.0260449647903442 +I0831 17:28:16.375740 139540727588608 logging_writer.py:48] [112800] global_step=112800, grad_norm=5.116514205932617, loss=1.0376782417297363 +I0831 17:28:43.503288 139540719195904 logging_writer.py:48] [112900] global_step=112900, grad_norm=4.992646217346191, loss=1.0483359098434448 +I0831 17:29:10.590700 139540727588608 logging_writer.py:48] [113000] global_step=113000, grad_norm=4.848923206329346, loss=1.041832447052002 +I0831 17:29:37.757535 139540719195904 logging_writer.py:48] [113100] global_step=113100, grad_norm=5.080832481384277, loss=1.032609224319458 +I0831 17:30:04.846027 139540727588608 logging_writer.py:48] [113200] global_step=113200, grad_norm=5.121710300445557, loss=1.0551822185516357 +I0831 17:30:31.918457 139540719195904 logging_writer.py:48] [113300] global_step=113300, grad_norm=4.8500823974609375, loss=0.9852249622344971 +I0831 17:30:59.051364 139540727588608 logging_writer.py:48] [113400] global_step=113400, grad_norm=5.15411376953125, loss=1.0434041023254395 +I0831 17:31:26.329117 139540719195904 logging_writer.py:48] [113500] global_step=113500, grad_norm=4.929070949554443, loss=0.9733931422233582 +I0831 17:31:53.407024 139540727588608 logging_writer.py:48] [113600] global_step=113600, grad_norm=4.786120891571045, loss=1.0046652555465698 +I0831 17:32:20.540459 139540719195904 logging_writer.py:48] [113700] global_step=113700, grad_norm=5.234891891479492, loss=1.0781701803207397 +I0831 17:32:47.637764 139540727588608 logging_writer.py:48] [113800] global_step=113800, grad_norm=4.796046257019043, loss=1.0695703029632568 +I0831 17:33:14.728340 139540719195904 logging_writer.py:48] [113900] global_step=113900, grad_norm=5.1159162521362305, loss=1.0030916929244995 +I0831 17:33:41.874886 139540727588608 logging_writer.py:48] [114000] global_step=114000, grad_norm=4.933363914489746, loss=1.0189522504806519 +I0831 17:34:08.965359 139540719195904 logging_writer.py:48] [114100] global_step=114100, grad_norm=4.623847007751465, loss=1.0023763179779053 +I0831 17:34:36.049445 139540727588608 logging_writer.py:48] [114200] global_step=114200, grad_norm=4.900040149688721, loss=1.012617588043213 +I0831 17:35:03.199331 139540719195904 logging_writer.py:48] [114300] global_step=114300, grad_norm=5.068504810333252, loss=1.1597883701324463 +I0831 17:35:30.270987 139540727588608 logging_writer.py:48] [114400] global_step=114400, grad_norm=4.55355167388916, loss=0.9316093325614929 +I0831 17:35:57.357628 139540719195904 logging_writer.py:48] [114500] global_step=114500, grad_norm=4.863308429718018, loss=1.040431261062622 +I0831 17:36:24.717590 139540727588608 logging_writer.py:48] [114600] global_step=114600, grad_norm=5.022075176239014, loss=1.0157123804092407 +I0831 17:36:51.823427 139540719195904 logging_writer.py:48] [114700] global_step=114700, grad_norm=5.101999759674072, loss=1.029576063156128 +I0831 17:37:18.908256 139540727588608 logging_writer.py:48] [114800] global_step=114800, grad_norm=4.789543628692627, loss=0.9683778882026672 +I0831 17:37:46.045563 139540719195904 logging_writer.py:48] [114900] global_step=114900, grad_norm=4.870083332061768, loss=0.9704713821411133 +I0831 17:38:13.182768 139540727588608 logging_writer.py:48] [115000] global_step=115000, grad_norm=5.077556133270264, loss=1.0500140190124512 +I0831 17:38:40.273367 139540719195904 logging_writer.py:48] [115100] global_step=115100, grad_norm=4.827365398406982, loss=0.9857801198959351 +I0831 17:39:07.420332 139540727588608 logging_writer.py:48] [115200] global_step=115200, grad_norm=4.799469947814941, loss=0.9691098928451538 +I0831 17:39:34.507575 139540719195904 logging_writer.py:48] [115300] global_step=115300, grad_norm=5.004415035247803, loss=1.0687587261199951 +I0831 17:40:01.611094 139540727588608 logging_writer.py:48] [115400] global_step=115400, grad_norm=4.708354949951172, loss=0.9987789392471313 +I0831 17:40:28.771785 139540719195904 logging_writer.py:48] [115500] global_step=115500, grad_norm=4.955265522003174, loss=0.9781686067581177 +I0831 17:40:55.825788 139540727588608 logging_writer.py:48] [115600] global_step=115600, grad_norm=4.895991802215576, loss=0.9518639445304871 +I0831 17:41:23.148327 139540719195904 logging_writer.py:48] [115700] global_step=115700, grad_norm=5.162551403045654, loss=0.9962859749794006 +I0831 17:41:50.318349 139540727588608 logging_writer.py:48] [115800] global_step=115800, grad_norm=4.937767028808594, loss=1.0064547061920166 +I0831 17:42:17.423976 139540719195904 logging_writer.py:48] [115900] global_step=115900, grad_norm=4.809168338775635, loss=1.021673321723938 +I0831 17:42:44.539261 139540727588608 logging_writer.py:48] [116000] global_step=116000, grad_norm=5.339272499084473, loss=1.0284873247146606 +I0831 17:43:11.692478 139540719195904 logging_writer.py:48] [116100] global_step=116100, grad_norm=5.020646572113037, loss=1.0193743705749512 +I0831 17:43:38.812531 139540727588608 logging_writer.py:48] [116200] global_step=116200, grad_norm=4.895806312561035, loss=1.060762643814087 +I0831 17:44:05.938578 139540719195904 logging_writer.py:48] [116300] global_step=116300, grad_norm=4.915205001831055, loss=1.0027375221252441 +I0831 17:44:33.099130 139540727588608 logging_writer.py:48] [116400] global_step=116400, grad_norm=4.765228748321533, loss=1.0136722326278687 +I0831 17:45:00.217566 139540719195904 logging_writer.py:48] [116500] global_step=116500, grad_norm=4.974015712738037, loss=0.9715111255645752 +I0831 17:45:27.340415 139540727588608 logging_writer.py:48] [116600] global_step=116600, grad_norm=4.9411163330078125, loss=0.9743790626525879 +I0831 17:45:50.478220 139757377230016 spec.py:333] Evaluating on the training split. +I0831 17:45:57.419199 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 17:46:07.647902 139757377230016 spec.py:363] Evaluating on the test split. +I0831 17:46:08.538273 139757377230016 submission_runner.py:516] Time since start: 32378.87s, Step: 116686, {'train/accuracy': Array(0.93215877, dtype=float32), 'train/loss': Array(0.23511541, dtype=float32), 'validation/accuracy': Array(0.74862, dtype=float32), 'validation/loss': Array(1.062171, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62310004, dtype=float32), 'test/loss': Array(1.8545954, dtype=float32), 'test/num_examples': 10000, 'score': 31989.126221895218, 'total_duration': 32378.873791217804, 'accumulated_submission_time': 31989.126221895218, 'accumulated_eval_time': 387.4781656265259, 'accumulated_logging_time': 1.3403329849243164} +I0831 17:46:08.595034 139540719195904 logging_writer.py:48] [116686] accumulated_eval_time=387.478, accumulated_logging_time=1.34033, accumulated_submission_time=31989.1, global_step=116686, preemption_count=0, score=31989.1, test/accuracy=0.6231000423431396, test/loss=1.854595422744751, test/num_examples=10000, total_duration=32378.9, train/accuracy=0.9321587681770325, train/loss=0.2351154088973999, validation/accuracy=0.7486199736595154, validation/loss=1.0621709823608398, validation/num_examples=50000 +I0831 17:46:12.849211 139540727588608 logging_writer.py:48] [116700] global_step=116700, grad_norm=4.850865840911865, loss=0.9246940016746521 +I0831 17:46:39.939847 139540719195904 logging_writer.py:48] [116800] global_step=116800, grad_norm=4.985005855560303, loss=0.9696827530860901 +I0831 17:47:07.034228 139540727588608 logging_writer.py:48] [116900] global_step=116900, grad_norm=5.280363082885742, loss=1.03873610496521 +I0831 17:47:34.186861 139540719195904 logging_writer.py:48] [117000] global_step=117000, grad_norm=4.726691246032715, loss=1.0884373188018799 +I0831 17:48:01.284756 139540727588608 logging_writer.py:48] [117100] global_step=117100, grad_norm=4.6572160720825195, loss=0.945054292678833 +I0831 17:48:28.373140 139540719195904 logging_writer.py:48] [117200] global_step=117200, grad_norm=4.849338054656982, loss=0.947242259979248 +I0831 17:48:55.542097 139540727588608 logging_writer.py:48] [117300] global_step=117300, grad_norm=4.7362236976623535, loss=0.9848617315292358 +I0831 17:49:22.631756 139540719195904 logging_writer.py:48] [117400] global_step=117400, grad_norm=5.037086009979248, loss=1.0426063537597656 +I0831 17:49:49.736171 139540727588608 logging_writer.py:48] [117500] global_step=117500, grad_norm=4.576347351074219, loss=0.9888850450515747 +I0831 17:50:16.926761 139540719195904 logging_writer.py:48] [117600] global_step=117600, grad_norm=5.159462928771973, loss=1.0366747379302979 +I0831 17:50:44.010081 139540727588608 logging_writer.py:48] [117700] global_step=117700, grad_norm=4.843582630157471, loss=1.020768404006958 +I0831 17:51:11.298532 139540719195904 logging_writer.py:48] [117800] global_step=117800, grad_norm=4.790757179260254, loss=1.042706847190857 +I0831 17:51:38.461899 139540727588608 logging_writer.py:48] [117900] global_step=117900, grad_norm=5.024764537811279, loss=1.0566422939300537 +I0831 17:52:05.578604 139540719195904 logging_writer.py:48] [118000] global_step=118000, grad_norm=5.075265407562256, loss=1.0940912961959839 +I0831 17:52:32.668577 139540727588608 logging_writer.py:48] [118100] global_step=118100, grad_norm=4.832976818084717, loss=0.8915734887123108 +I0831 17:52:59.838496 139540719195904 logging_writer.py:48] [118200] global_step=118200, grad_norm=4.826441764831543, loss=0.9139772057533264 +I0831 17:53:26.970192 139540727588608 logging_writer.py:48] [118300] global_step=118300, grad_norm=4.5998735427856445, loss=1.0039470195770264 +I0831 17:53:54.064903 139540719195904 logging_writer.py:48] [118400] global_step=118400, grad_norm=4.6290388107299805, loss=0.9524574875831604 +I0831 17:54:21.221982 139540727588608 logging_writer.py:48] [118500] global_step=118500, grad_norm=4.790989398956299, loss=0.9743832349777222 +I0831 17:54:48.317688 139540719195904 logging_writer.py:48] [118600] global_step=118600, grad_norm=4.774258613586426, loss=1.0503861904144287 +I0831 17:55:15.405466 139540727588608 logging_writer.py:48] [118700] global_step=118700, grad_norm=5.069082260131836, loss=1.0420067310333252 +I0831 17:55:42.587847 139540719195904 logging_writer.py:48] [118800] global_step=118800, grad_norm=5.208921432495117, loss=1.0600954294204712 +I0831 17:56:09.905094 139540727588608 logging_writer.py:48] [118900] global_step=118900, grad_norm=5.088920593261719, loss=1.0037319660186768 +I0831 17:56:36.992608 139540719195904 logging_writer.py:48] [119000] global_step=119000, grad_norm=4.895788192749023, loss=1.05488121509552 +I0831 17:57:04.169406 139540727588608 logging_writer.py:48] [119100] global_step=119100, grad_norm=5.0212507247924805, loss=1.1313481330871582 +I0831 17:57:31.268151 139540719195904 logging_writer.py:48] [119200] global_step=119200, grad_norm=5.0524582862854, loss=1.0196853876113892 +I0831 17:57:58.359970 139540727588608 logging_writer.py:48] [119300] global_step=119300, grad_norm=4.855984210968018, loss=1.0321329832077026 +I0831 17:58:25.536695 139540719195904 logging_writer.py:48] [119400] global_step=119400, grad_norm=4.882188320159912, loss=0.9793161153793335 +I0831 17:58:52.627118 139540727588608 logging_writer.py:48] [119500] global_step=119500, grad_norm=4.682689666748047, loss=0.9560534954071045 +I0831 17:59:19.724191 139540719195904 logging_writer.py:48] [119600] global_step=119600, grad_norm=5.328454494476318, loss=1.133991003036499 +I0831 17:59:46.865564 139540727588608 logging_writer.py:48] [119700] global_step=119700, grad_norm=4.931460857391357, loss=1.0062448978424072 +I0831 18:00:13.950495 139540719195904 logging_writer.py:48] [119800] global_step=119800, grad_norm=4.81692361831665, loss=0.9797189831733704 +I0831 18:00:41.275080 139540727588608 logging_writer.py:48] [119900] global_step=119900, grad_norm=5.17497444152832, loss=1.063366413116455 +I0831 18:01:08.438009 139540719195904 logging_writer.py:48] [120000] global_step=120000, grad_norm=4.767455577850342, loss=0.9464429616928101 +I0831 18:01:35.520903 139540727588608 logging_writer.py:48] [120100] global_step=120100, grad_norm=4.982605934143066, loss=1.0156686305999756 +I0831 18:02:02.623434 139540719195904 logging_writer.py:48] [120200] global_step=120200, grad_norm=4.888095855712891, loss=1.015663504600525 +I0831 18:02:29.767616 139540727588608 logging_writer.py:48] [120300] global_step=120300, grad_norm=4.875648021697998, loss=1.0243327617645264 +I0831 18:02:56.840688 139540719195904 logging_writer.py:48] [120400] global_step=120400, grad_norm=5.199045181274414, loss=1.0804758071899414 +I0831 18:03:23.935565 139540727588608 logging_writer.py:48] [120500] global_step=120500, grad_norm=5.099085330963135, loss=1.0528099536895752 +I0831 18:03:51.080424 139540719195904 logging_writer.py:48] [120600] global_step=120600, grad_norm=5.07235860824585, loss=1.1073801517486572 +I0831 18:04:18.183529 139540727588608 logging_writer.py:48] [120700] global_step=120700, grad_norm=4.947928428649902, loss=1.002941370010376 +I0831 18:04:45.289692 139540719195904 logging_writer.py:48] [120800] global_step=120800, grad_norm=5.007067680358887, loss=1.0194284915924072 +I0831 18:05:12.457620 139540727588608 logging_writer.py:48] [120900] global_step=120900, grad_norm=5.157216548919678, loss=1.0549672842025757 +I0831 18:05:39.796056 139540719195904 logging_writer.py:48] [121000] global_step=121000, grad_norm=4.975782871246338, loss=0.989166259765625 +I0831 18:06:06.883098 139540727588608 logging_writer.py:48] [121100] global_step=121100, grad_norm=4.987039089202881, loss=0.9979760050773621 +I0831 18:06:34.032985 139540719195904 logging_writer.py:48] [121200] global_step=121200, grad_norm=4.70930814743042, loss=0.9539833664894104 +I0831 18:07:01.105231 139540727588608 logging_writer.py:48] [121300] global_step=121300, grad_norm=5.046921253204346, loss=1.0125489234924316 +I0831 18:07:28.198012 139540719195904 logging_writer.py:48] [121400] global_step=121400, grad_norm=4.899130344390869, loss=0.9760012626647949 +I0831 18:07:55.348753 139540727588608 logging_writer.py:48] [121500] global_step=121500, grad_norm=5.0322065353393555, loss=0.9974268674850464 +I0831 18:08:22.453607 139540719195904 logging_writer.py:48] [121600] global_step=121600, grad_norm=5.067147731781006, loss=0.9801316261291504 +I0831 18:08:49.546783 139540727588608 logging_writer.py:48] [121700] global_step=121700, grad_norm=4.673852443695068, loss=0.9502684473991394 +I0831 18:09:16.704337 139540719195904 logging_writer.py:48] [121800] global_step=121800, grad_norm=4.734701633453369, loss=0.9925256967544556 +I0831 18:09:43.806273 139540727588608 logging_writer.py:48] [121900] global_step=121900, grad_norm=4.765255451202393, loss=0.9965633153915405 +I0831 18:10:10.892127 139540719195904 logging_writer.py:48] [122000] global_step=122000, grad_norm=5.245431900024414, loss=1.074205756187439 +I0831 18:10:38.303929 139540727588608 logging_writer.py:48] [122100] global_step=122100, grad_norm=5.11958646774292, loss=0.9919639825820923 +I0831 18:11:05.412484 139540719195904 logging_writer.py:48] [122200] global_step=122200, grad_norm=5.021030902862549, loss=1.0741068124771118 +I0831 18:11:32.493514 139540727588608 logging_writer.py:48] [122300] global_step=122300, grad_norm=4.888718128204346, loss=1.0632461309432983 +I0831 18:11:59.650493 139540719195904 logging_writer.py:48] [122400] global_step=122400, grad_norm=5.155271053314209, loss=1.043282151222229 +I0831 18:12:26.755922 139540727588608 logging_writer.py:48] [122500] global_step=122500, grad_norm=4.75963020324707, loss=0.9929556846618652 +I0831 18:12:53.864790 139540719195904 logging_writer.py:48] [122600] global_step=122600, grad_norm=4.835896968841553, loss=0.9625310301780701 +I0831 18:13:21.010526 139540727588608 logging_writer.py:48] [122700] global_step=122700, grad_norm=4.894546031951904, loss=0.957472026348114 +I0831 18:13:48.105910 139540719195904 logging_writer.py:48] [122800] global_step=122800, grad_norm=4.942352294921875, loss=1.0315544605255127 +I0831 18:14:15.254324 139540727588608 logging_writer.py:48] [122900] global_step=122900, grad_norm=4.83389949798584, loss=0.9488270878791809 +I0831 18:14:42.419125 139540719195904 logging_writer.py:48] [123000] global_step=123000, grad_norm=5.066993236541748, loss=1.0502017736434937 +I0831 18:15:09.732498 139540727588608 logging_writer.py:48] [123100] global_step=123100, grad_norm=4.823647975921631, loss=1.057628870010376 +I0831 18:15:36.833563 139540719195904 logging_writer.py:48] [123200] global_step=123200, grad_norm=4.729123115539551, loss=0.9157568216323853 +I0831 18:16:04.032921 139540727588608 logging_writer.py:48] [123300] global_step=123300, grad_norm=5.0395917892456055, loss=1.076278567314148 +I0831 18:16:31.125625 139540719195904 logging_writer.py:48] [123400] global_step=123400, grad_norm=4.518982410430908, loss=0.8206816911697388 +I0831 18:16:58.222921 139540727588608 logging_writer.py:48] [123500] global_step=123500, grad_norm=4.942296028137207, loss=1.0297702550888062 +I0831 18:17:25.401809 139540719195904 logging_writer.py:48] [123600] global_step=123600, grad_norm=4.629934787750244, loss=0.9277737140655518 +I0831 18:17:52.520854 139540727588608 logging_writer.py:48] [123700] global_step=123700, grad_norm=4.996158599853516, loss=1.0691512823104858 +I0831 18:18:19.632146 139540719195904 logging_writer.py:48] [123800] global_step=123800, grad_norm=4.748520374298096, loss=0.9453434348106384 +I0831 18:18:46.784220 139540727588608 logging_writer.py:48] [123900] global_step=123900, grad_norm=4.903236389160156, loss=0.9822508096694946 +I0831 18:19:13.859519 139540719195904 logging_writer.py:48] [124000] global_step=124000, grad_norm=5.01711893081665, loss=1.0271432399749756 +I0831 18:19:24.556110 139757377230016 spec.py:333] Evaluating on the training split. +I0831 18:19:31.481472 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 18:19:41.544848 139757377230016 spec.py:363] Evaluating on the test split. +I0831 18:19:42.435027 139757377230016 submission_runner.py:516] Time since start: 34392.77s, Step: 124041, {'train/accuracy': Array(0.9368423, dtype=float32), 'train/loss': Array(0.22361873, dtype=float32), 'validation/accuracy': Array(0.75012, dtype=float32), 'validation/loss': Array(1.0643718, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62140006, dtype=float32), 'test/loss': Array(1.8547386, dtype=float32), 'test/num_examples': 10000, 'score': 33985.024151563644, 'total_duration': 34392.77033543587, 'accumulated_submission_time': 33985.024151563644, 'accumulated_eval_time': 405.35461807250977, 'accumulated_logging_time': 1.4072437286376953} +I0831 18:19:42.486810 139540727588608 logging_writer.py:48] [124041] accumulated_eval_time=405.355, accumulated_logging_time=1.40724, accumulated_submission_time=33985, global_step=124041, preemption_count=0, score=33985, test/accuracy=0.6214000582695007, test/loss=1.8547385931015015, test/num_examples=10000, total_duration=34392.8, train/accuracy=0.9368423223495483, train/loss=0.22361873090267181, validation/accuracy=0.7501199841499329, validation/loss=1.0643718242645264, validation/num_examples=50000 +I0831 18:19:58.852967 139540719195904 logging_writer.py:48] [124100] global_step=124100, grad_norm=4.959739685058594, loss=0.9920551180839539 +I0831 18:20:26.205404 139540727588608 logging_writer.py:48] [124200] global_step=124200, grad_norm=4.724878311157227, loss=0.9513610601425171 +I0831 18:20:53.304197 139540719195904 logging_writer.py:48] [124300] global_step=124300, grad_norm=5.189523220062256, loss=1.1107758283615112 +I0831 18:21:20.388117 139540727588608 logging_writer.py:48] [124400] global_step=124400, grad_norm=4.62212610244751, loss=0.9888362884521484 +I0831 18:21:47.553910 139540719195904 logging_writer.py:48] [124500] global_step=124500, grad_norm=5.255875110626221, loss=0.9745398759841919 +I0831 18:22:14.651595 139540727588608 logging_writer.py:48] [124600] global_step=124600, grad_norm=4.630988121032715, loss=0.9354742765426636 +I0831 18:22:41.748955 139540719195904 logging_writer.py:48] [124700] global_step=124700, grad_norm=5.13581657409668, loss=0.9921715259552002 +I0831 18:23:08.876364 139540727588608 logging_writer.py:48] [124800] global_step=124800, grad_norm=4.983302593231201, loss=1.0238392353057861 +I0831 18:23:35.980696 139540719195904 logging_writer.py:48] [124900] global_step=124900, grad_norm=4.693772315979004, loss=0.9547627568244934 +I0831 18:24:03.062595 139540727588608 logging_writer.py:48] [125000] global_step=125000, grad_norm=4.958782196044922, loss=1.0629838705062866 +I0831 18:24:30.216333 139540719195904 logging_writer.py:48] [125100] global_step=125100, grad_norm=5.282406330108643, loss=1.0854395627975464 +I0831 18:24:57.537094 139540727588608 logging_writer.py:48] [125200] global_step=125200, grad_norm=4.66309928894043, loss=0.9215268492698669 +I0831 18:25:24.655093 139540719195904 logging_writer.py:48] [125300] global_step=125300, grad_norm=4.64963436126709, loss=0.955488920211792 +I0831 18:25:51.800070 139540727588608 logging_writer.py:48] [125400] global_step=125400, grad_norm=5.214349746704102, loss=0.9384632110595703 +I0831 18:26:18.899262 139540719195904 logging_writer.py:48] [125500] global_step=125500, grad_norm=4.932323932647705, loss=1.0501761436462402 +I0831 18:26:46.003131 139540727588608 logging_writer.py:48] [125600] global_step=125600, grad_norm=4.680225849151611, loss=1.0079078674316406 +I0831 18:27:13.174013 139540719195904 logging_writer.py:48] [125700] global_step=125700, grad_norm=4.795058727264404, loss=0.9400352835655212 +I0831 18:27:40.285763 139540727588608 logging_writer.py:48] [125800] global_step=125800, grad_norm=4.933342456817627, loss=1.0056415796279907 +I0831 18:28:07.367240 139540719195904 logging_writer.py:48] [125900] global_step=125900, grad_norm=4.708018779754639, loss=0.9355145692825317 +I0831 18:28:34.529704 139540727588608 logging_writer.py:48] [126000] global_step=126000, grad_norm=4.999928951263428, loss=0.9737023115158081 +I0831 18:29:01.598627 139540719195904 logging_writer.py:48] [126100] global_step=126100, grad_norm=4.926107883453369, loss=0.9490303993225098 +I0831 18:29:28.684587 139540727588608 logging_writer.py:48] [126200] global_step=126200, grad_norm=4.814653396606445, loss=0.9340658187866211 +I0831 18:29:56.054307 139540719195904 logging_writer.py:48] [126300] global_step=126300, grad_norm=4.775744915008545, loss=0.9669443368911743 +I0831 18:30:23.160404 139540727588608 logging_writer.py:48] [126400] global_step=126400, grad_norm=4.821078777313232, loss=0.975071370601654 +I0831 18:30:50.266683 139540719195904 logging_writer.py:48] [126500] global_step=126500, grad_norm=5.473040580749512, loss=1.1141996383666992 +I0831 18:31:17.432614 139540727588608 logging_writer.py:48] [126600] global_step=126600, grad_norm=5.071509838104248, loss=0.9918809533119202 +I0831 18:31:44.572504 139540719195904 logging_writer.py:48] [126700] global_step=126700, grad_norm=4.734167575836182, loss=0.9224956035614014 +I0831 18:32:11.658519 139540727588608 logging_writer.py:48] [126800] global_step=126800, grad_norm=5.002106189727783, loss=1.0041762590408325 +I0831 18:32:38.832156 139540719195904 logging_writer.py:48] [126900] global_step=126900, grad_norm=4.5831475257873535, loss=0.9289351105690002 +I0831 18:33:05.946062 139540727588608 logging_writer.py:48] [127000] global_step=127000, grad_norm=4.907226085662842, loss=1.0300579071044922 +I0831 18:33:33.056085 139540719195904 logging_writer.py:48] [127100] global_step=127100, grad_norm=5.230595588684082, loss=1.0330252647399902 +I0831 18:34:00.195689 139540727588608 logging_writer.py:48] [127200] global_step=127200, grad_norm=5.226077079772949, loss=1.003427267074585 +I0831 18:34:27.530808 139540719195904 logging_writer.py:48] [127300] global_step=127300, grad_norm=5.129274368286133, loss=0.9600069522857666 +I0831 18:34:54.643059 139540727588608 logging_writer.py:48] [127400] global_step=127400, grad_norm=4.5660810470581055, loss=0.9041241407394409 +I0831 18:35:21.826250 139540719195904 logging_writer.py:48] [127500] global_step=127500, grad_norm=4.850088596343994, loss=0.9477722644805908 +I0831 18:35:48.942492 139540727588608 logging_writer.py:48] [127600] global_step=127600, grad_norm=5.099522590637207, loss=1.016257405281067 +I0831 18:36:16.048957 139540719195904 logging_writer.py:48] [127700] global_step=127700, grad_norm=5.220772743225098, loss=1.0599383115768433 +I0831 18:36:43.204951 139540727588608 logging_writer.py:48] [127800] global_step=127800, grad_norm=5.22505521774292, loss=1.0208985805511475 +I0831 18:37:10.313564 139540719195904 logging_writer.py:48] [127900] global_step=127900, grad_norm=4.915736198425293, loss=0.993891179561615 +I0831 18:37:37.414138 139540727588608 logging_writer.py:48] [128000] global_step=128000, grad_norm=5.037959575653076, loss=1.0021891593933105 +I0831 18:38:04.561049 139540719195904 logging_writer.py:48] [128100] global_step=128100, grad_norm=4.761622905731201, loss=0.9119603633880615 +I0831 18:38:31.651180 139540727588608 logging_writer.py:48] [128200] global_step=128200, grad_norm=5.160921096801758, loss=0.985394299030304 +I0831 18:38:58.731973 139540719195904 logging_writer.py:48] [128300] global_step=128300, grad_norm=4.720265865325928, loss=0.9820288419723511 +I0831 18:39:26.119498 139540727588608 logging_writer.py:48] [128400] global_step=128400, grad_norm=4.894281387329102, loss=0.9295991659164429 +I0831 18:39:53.246657 139540719195904 logging_writer.py:48] [128500] global_step=128500, grad_norm=4.836514472961426, loss=0.9874048233032227 +I0831 18:40:20.358597 139540727588608 logging_writer.py:48] [128600] global_step=128600, grad_norm=4.784384250640869, loss=0.9857382774353027 +I0831 18:40:47.542778 139540719195904 logging_writer.py:48] [128700] global_step=128700, grad_norm=4.663294792175293, loss=0.9599646329879761 +I0831 18:41:14.685284 139540727588608 logging_writer.py:48] [128800] global_step=128800, grad_norm=4.87637186050415, loss=1.0406973361968994 +I0831 18:41:41.804425 139540719195904 logging_writer.py:48] [128900] global_step=128900, grad_norm=4.832684516906738, loss=0.989304780960083 +I0831 18:42:08.976248 139540727588608 logging_writer.py:48] [129000] global_step=129000, grad_norm=5.087740421295166, loss=0.9998127818107605 +I0831 18:42:36.070572 139540719195904 logging_writer.py:48] [129100] global_step=129100, grad_norm=4.804134845733643, loss=0.9764799475669861 +I0831 18:43:03.177804 139540727588608 logging_writer.py:48] [129200] global_step=129200, grad_norm=4.770850658416748, loss=0.9617881774902344 +I0831 18:43:30.369155 139540719195904 logging_writer.py:48] [129300] global_step=129300, grad_norm=5.032071113586426, loss=0.9827959537506104 +I0831 18:43:57.707504 139540727588608 logging_writer.py:48] [129400] global_step=129400, grad_norm=5.333802223205566, loss=1.0095701217651367 +I0831 18:44:24.780480 139540719195904 logging_writer.py:48] [129500] global_step=129500, grad_norm=4.880114555358887, loss=1.017694354057312 +I0831 18:44:51.943581 139540727588608 logging_writer.py:48] [129600] global_step=129600, grad_norm=5.064136028289795, loss=1.0611350536346436 +I0831 18:45:19.075496 139540719195904 logging_writer.py:48] [129700] global_step=129700, grad_norm=5.05799674987793, loss=1.0527386665344238 +I0831 18:45:46.199852 139540727588608 logging_writer.py:48] [129800] global_step=129800, grad_norm=5.037579536437988, loss=1.0226256847381592 +I0831 18:46:13.390495 139540719195904 logging_writer.py:48] [129900] global_step=129900, grad_norm=4.5822272300720215, loss=0.9281449317932129 +I0831 18:46:40.517341 139540727588608 logging_writer.py:48] [130000] global_step=130000, grad_norm=4.903039455413818, loss=0.9968441128730774 +I0831 18:47:07.597542 139540719195904 logging_writer.py:48] [130100] global_step=130100, grad_norm=4.8166608810424805, loss=1.0385651588439941 +I0831 18:47:34.753129 139540727588608 logging_writer.py:48] [130200] global_step=130200, grad_norm=5.072717666625977, loss=1.0414743423461914 +I0831 18:48:01.832816 139540719195904 logging_writer.py:48] [130300] global_step=130300, grad_norm=5.161195278167725, loss=1.0580780506134033 +I0831 18:48:28.928406 139540727588608 logging_writer.py:48] [130400] global_step=130400, grad_norm=4.966507434844971, loss=1.0647692680358887 +I0831 18:48:56.284064 139540719195904 logging_writer.py:48] [130500] global_step=130500, grad_norm=4.949012279510498, loss=0.9918736815452576 +I0831 18:49:23.392076 139540727588608 logging_writer.py:48] [130600] global_step=130600, grad_norm=4.729351043701172, loss=0.9863735437393188 +I0831 18:49:50.484762 139540719195904 logging_writer.py:48] [130700] global_step=130700, grad_norm=4.916469573974609, loss=0.9084323644638062 +I0831 18:50:17.637459 139540727588608 logging_writer.py:48] [130800] global_step=130800, grad_norm=4.961255073547363, loss=0.9821269512176514 +I0831 18:50:44.720630 139540719195904 logging_writer.py:48] [130900] global_step=130900, grad_norm=4.86811637878418, loss=0.8685524463653564 +I0831 18:51:11.793047 139540727588608 logging_writer.py:48] [131000] global_step=131000, grad_norm=4.792481899261475, loss=0.9898805618286133 +I0831 18:51:38.955761 139540719195904 logging_writer.py:48] [131100] global_step=131100, grad_norm=5.00923490524292, loss=0.9619134664535522 +I0831 18:52:06.041609 139540727588608 logging_writer.py:48] [131200] global_step=131200, grad_norm=4.644381523132324, loss=0.9350121021270752 +I0831 18:52:33.159012 139540719195904 logging_writer.py:48] [131300] global_step=131300, grad_norm=4.699960231781006, loss=0.8987125754356384 +I0831 18:52:58.519907 139757377230016 spec.py:333] Evaluating on the training split. +I0831 18:53:05.527402 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 18:53:15.763388 139757377230016 spec.py:363] Evaluating on the test split. +I0831 18:53:16.634721 139757377230016 submission_runner.py:516] Time since start: 36406.97s, Step: 131395, {'train/accuracy': Array(0.93939334, dtype=float32), 'train/loss': Array(0.21290366, dtype=float32), 'validation/accuracy': Array(0.75088, dtype=float32), 'validation/loss': Array(1.0653294, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62280005, dtype=float32), 'test/loss': Array(1.852586, dtype=float32), 'test/num_examples': 10000, 'score': 35980.980189561844, 'total_duration': 36406.970715522766, 'accumulated_submission_time': 35980.980189561844, 'accumulated_eval_time': 423.46765208244324, 'accumulated_logging_time': 1.4830548763275146} +I0831 18:53:16.693917 139540727588608 logging_writer.py:48] [131395] accumulated_eval_time=423.468, accumulated_logging_time=1.48305, accumulated_submission_time=35981, global_step=131395, preemption_count=0, score=35981, test/accuracy=0.6228000521659851, test/loss=1.852586030960083, test/num_examples=10000, total_duration=36407, train/accuracy=0.9393933415412903, train/loss=0.21290366351604462, validation/accuracy=0.7508800029754639, validation/loss=1.0653294324874878, validation/num_examples=50000 +I0831 18:53:18.486128 139540719195904 logging_writer.py:48] [131400] global_step=131400, grad_norm=5.174779415130615, loss=1.0998032093048096 +I0831 18:53:45.624301 139540727588608 logging_writer.py:48] [131500] global_step=131500, grad_norm=4.881472587585449, loss=0.9606100916862488 +I0831 18:54:12.938481 139540719195904 logging_writer.py:48] [131600] global_step=131600, grad_norm=5.015367031097412, loss=0.9480982422828674 +I0831 18:54:40.081138 139540727588608 logging_writer.py:48] [131700] global_step=131700, grad_norm=4.756710052490234, loss=0.9282273054122925 +I0831 18:55:07.175794 139540719195904 logging_writer.py:48] [131800] global_step=131800, grad_norm=4.550253391265869, loss=0.9029451608657837 +I0831 18:55:34.288561 139540727588608 logging_writer.py:48] [131900] global_step=131900, grad_norm=4.693728446960449, loss=0.9061131477355957 +I0831 18:56:01.437904 139540719195904 logging_writer.py:48] [132000] global_step=132000, grad_norm=5.1859331130981445, loss=1.1230512857437134 +I0831 18:56:28.556552 139540727588608 logging_writer.py:48] [132100] global_step=132100, grad_norm=4.817812919616699, loss=0.9043564796447754 +I0831 18:56:55.640741 139540719195904 logging_writer.py:48] [132200] global_step=132200, grad_norm=4.908921241760254, loss=0.9540932178497314 +I0831 18:57:22.794031 139540727588608 logging_writer.py:48] [132300] global_step=132300, grad_norm=4.8843817710876465, loss=0.9712030291557312 +I0831 18:57:49.869401 139540719195904 logging_writer.py:48] [132400] global_step=132400, grad_norm=4.748700141906738, loss=0.9776036739349365 +I0831 18:58:16.947039 139540727588608 logging_writer.py:48] [132500] global_step=132500, grad_norm=4.926920413970947, loss=0.9731858968734741 +I0831 18:58:44.307026 139540719195904 logging_writer.py:48] [132600] global_step=132600, grad_norm=4.861461162567139, loss=1.0152194499969482 +I0831 18:59:11.427424 139540727588608 logging_writer.py:48] [132700] global_step=132700, grad_norm=5.075224876403809, loss=1.0098729133605957 +I0831 18:59:38.520269 139540719195904 logging_writer.py:48] [132800] global_step=132800, grad_norm=4.839643955230713, loss=0.9747335910797119 +I0831 19:00:05.681965 139540727588608 logging_writer.py:48] [132900] global_step=132900, grad_norm=4.951384544372559, loss=0.9948177933692932 +I0831 19:00:32.799757 139540719195904 logging_writer.py:48] [133000] global_step=133000, grad_norm=5.130736827850342, loss=1.0611969232559204 +I0831 19:00:59.899385 139540727588608 logging_writer.py:48] [133100] global_step=133100, grad_norm=4.978974342346191, loss=1.0231566429138184 +I0831 19:01:27.055519 139540719195904 logging_writer.py:48] [133200] global_step=133200, grad_norm=5.0649800300598145, loss=1.0556695461273193 +I0831 19:01:54.156955 139540727588608 logging_writer.py:48] [133300] global_step=133300, grad_norm=4.9380292892456055, loss=0.9623417854309082 +I0831 19:02:21.291224 139540719195904 logging_writer.py:48] [133400] global_step=133400, grad_norm=4.953096389770508, loss=0.9824788570404053 +I0831 19:02:48.489248 139540727588608 logging_writer.py:48] [133500] global_step=133500, grad_norm=4.55886697769165, loss=0.9080998301506042 +I0831 19:03:15.590631 139540719195904 logging_writer.py:48] [133600] global_step=133600, grad_norm=4.626733303070068, loss=0.9675574898719788 +I0831 19:03:42.919795 139540727588608 logging_writer.py:48] [133700] global_step=133700, grad_norm=4.66933536529541, loss=0.9381409287452698 +I0831 19:04:10.087916 139540719195904 logging_writer.py:48] [133800] global_step=133800, grad_norm=5.014467716217041, loss=0.9839687347412109 +I0831 19:04:37.186795 139540727588608 logging_writer.py:48] [133900] global_step=133900, grad_norm=5.344640731811523, loss=0.9836798906326294 +I0831 19:05:04.281347 139540719195904 logging_writer.py:48] [134000] global_step=134000, grad_norm=4.737255096435547, loss=0.9720773100852966 +I0831 19:05:31.418731 139540727588608 logging_writer.py:48] [134100] global_step=134100, grad_norm=4.7850847244262695, loss=0.9970805644989014 +I0831 19:05:58.518790 139540719195904 logging_writer.py:48] [134200] global_step=134200, grad_norm=4.786212921142578, loss=1.0162544250488281 +I0831 19:06:25.631519 139540727588608 logging_writer.py:48] [134300] global_step=134300, grad_norm=4.86754846572876, loss=1.0083870887756348 +I0831 19:06:52.808140 139540719195904 logging_writer.py:48] [134400] global_step=134400, grad_norm=4.8439764976501465, loss=0.9804175496101379 +I0831 19:07:19.874026 139540727588608 logging_writer.py:48] [134500] global_step=134500, grad_norm=5.219333171844482, loss=1.0082950592041016 +I0831 19:07:46.948734 139540719195904 logging_writer.py:48] [134600] global_step=134600, grad_norm=4.9980902671813965, loss=0.9767093658447266 +I0831 19:08:14.134377 139540727588608 logging_writer.py:48] [134700] global_step=134700, grad_norm=4.984531879425049, loss=1.000212550163269 +I0831 19:08:41.458330 139540719195904 logging_writer.py:48] [134800] global_step=134800, grad_norm=5.057774543762207, loss=0.970267653465271 +I0831 19:09:08.540624 139540727588608 logging_writer.py:48] [134900] global_step=134900, grad_norm=4.818347454071045, loss=0.9613227844238281 +I0831 19:09:35.721169 139540719195904 logging_writer.py:48] [135000] global_step=135000, grad_norm=4.886783599853516, loss=0.898292064666748 +I0831 19:10:02.799037 139540727588608 logging_writer.py:48] [135100] global_step=135100, grad_norm=4.782504558563232, loss=1.0202122926712036 +I0831 19:10:29.924744 139540719195904 logging_writer.py:48] [135200] global_step=135200, grad_norm=5.052027225494385, loss=1.0283598899841309 +I0831 19:10:57.066369 139540727588608 logging_writer.py:48] [135300] global_step=135300, grad_norm=4.920061111450195, loss=0.9737001061439514 +I0831 19:11:24.159535 139540719195904 logging_writer.py:48] [135400] global_step=135400, grad_norm=5.008421897888184, loss=0.9688220024108887 +I0831 19:11:51.256793 139540727588608 logging_writer.py:48] [135500] global_step=135500, grad_norm=4.738471508026123, loss=0.8799198269844055 +I0831 19:12:18.404840 139540719195904 logging_writer.py:48] [135600] global_step=135600, grad_norm=4.7763285636901855, loss=0.9693702459335327 +I0831 19:12:45.513953 139540727588608 logging_writer.py:48] [135700] global_step=135700, grad_norm=5.001764297485352, loss=0.9568836092948914 +I0831 19:13:12.626114 139540719195904 logging_writer.py:48] [135800] global_step=135800, grad_norm=4.95102596282959, loss=1.0468212366104126 +I0831 19:13:39.971271 139540727588608 logging_writer.py:48] [135900] global_step=135900, grad_norm=4.995884418487549, loss=0.9872132539749146 +I0831 19:14:07.048359 139540719195904 logging_writer.py:48] [136000] global_step=136000, grad_norm=5.041899681091309, loss=0.9912666082382202 +I0831 19:14:34.137422 139540727588608 logging_writer.py:48] [136100] global_step=136100, grad_norm=4.697211265563965, loss=1.004833698272705 +I0831 19:15:01.288827 139540719195904 logging_writer.py:48] [136200] global_step=136200, grad_norm=4.846872329711914, loss=0.9596787095069885 +I0831 19:15:28.364338 139540727588608 logging_writer.py:48] [136300] global_step=136300, grad_norm=4.814752101898193, loss=0.9751400947570801 +I0831 19:15:55.473401 139540719195904 logging_writer.py:48] [136400] global_step=136400, grad_norm=4.85992431640625, loss=0.9683844447135925 +I0831 19:16:22.614467 139540727588608 logging_writer.py:48] [136500] global_step=136500, grad_norm=4.775761604309082, loss=0.9988193511962891 +I0831 19:16:49.693044 139540719195904 logging_writer.py:48] [136600] global_step=136600, grad_norm=4.654741287231445, loss=0.9687706828117371 +I0831 19:17:16.767042 139540727588608 logging_writer.py:48] [136700] global_step=136700, grad_norm=5.1533966064453125, loss=0.952431321144104 +I0831 19:17:43.946865 139540719195904 logging_writer.py:48] [136800] global_step=136800, grad_norm=4.5290021896362305, loss=0.9011249542236328 +I0831 19:18:11.261119 139540727588608 logging_writer.py:48] [136900] global_step=136900, grad_norm=4.809020519256592, loss=0.9874517917633057 +I0831 19:18:38.366147 139540719195904 logging_writer.py:48] [137000] global_step=137000, grad_norm=4.746774196624756, loss=0.9756983518600464 +I0831 19:19:05.520724 139540727588608 logging_writer.py:48] [137100] global_step=137100, grad_norm=4.814104080200195, loss=0.9646883010864258 +I0831 19:19:32.608534 139540719195904 logging_writer.py:48] [137200] global_step=137200, grad_norm=4.543582916259766, loss=0.8482443690299988 +I0831 19:19:59.718627 139540727588608 logging_writer.py:48] [137300] global_step=137300, grad_norm=4.92318868637085, loss=1.0219700336456299 +I0831 19:20:26.892715 139540719195904 logging_writer.py:48] [137400] global_step=137400, grad_norm=4.854905128479004, loss=0.9811145663261414 +I0831 19:20:54.021012 139540727588608 logging_writer.py:48] [137500] global_step=137500, grad_norm=4.905314922332764, loss=0.9971307516098022 +I0831 19:21:21.113704 139540719195904 logging_writer.py:48] [137600] global_step=137600, grad_norm=5.09190559387207, loss=1.030572533607483 +I0831 19:21:48.306022 139540727588608 logging_writer.py:48] [137700] global_step=137700, grad_norm=4.949118137359619, loss=0.9926077723503113 +I0831 19:22:15.409836 139540719195904 logging_writer.py:48] [137800] global_step=137800, grad_norm=5.166230201721191, loss=1.0907676219940186 +I0831 19:22:42.511585 139540727588608 logging_writer.py:48] [137900] global_step=137900, grad_norm=4.692147254943848, loss=0.9595528841018677 +I0831 19:23:09.903119 139540719195904 logging_writer.py:48] [138000] global_step=138000, grad_norm=4.794547080993652, loss=0.945045530796051 +I0831 19:23:37.001465 139540727588608 logging_writer.py:48] [138100] global_step=138100, grad_norm=4.727849006652832, loss=0.9876723289489746 +I0831 19:24:04.097178 139540719195904 logging_writer.py:48] [138200] global_step=138200, grad_norm=4.701259613037109, loss=0.9147742390632629 +I0831 19:24:31.258566 139540727588608 logging_writer.py:48] [138300] global_step=138300, grad_norm=5.004670143127441, loss=0.9913433790206909 +I0831 19:24:58.365516 139540719195904 logging_writer.py:48] [138400] global_step=138400, grad_norm=4.869284152984619, loss=0.9866592884063721 +I0831 19:25:25.499127 139540727588608 logging_writer.py:48] [138500] global_step=138500, grad_norm=4.787542819976807, loss=0.9775620698928833 +I0831 19:25:52.675925 139540719195904 logging_writer.py:48] [138600] global_step=138600, grad_norm=4.576642036437988, loss=0.8719696998596191 +I0831 19:26:19.802289 139540727588608 logging_writer.py:48] [138700] global_step=138700, grad_norm=4.788989543914795, loss=0.959772527217865 +I0831 19:26:32.697615 139757377230016 spec.py:333] Evaluating on the training split. +I0831 19:26:39.562992 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 19:26:50.504928 139757377230016 spec.py:363] Evaluating on the test split. +I0831 19:26:51.384441 139757377230016 submission_runner.py:516] Time since start: 38421.72s, Step: 138749, {'train/accuracy': Array(0.93939334, dtype=float32), 'train/loss': Array(0.21290293, dtype=float32), 'validation/accuracy': Array(0.75162, dtype=float32), 'validation/loss': Array(1.0644616, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6233, dtype=float32), 'test/loss': Array(1.8521956, dtype=float32), 'test/num_examples': 10000, 'score': 37976.918315172195, 'total_duration': 38421.72024726868, 'accumulated_submission_time': 37976.918315172195, 'accumulated_eval_time': 442.1525115966797, 'accumulated_logging_time': 1.554464340209961} +I0831 19:26:51.432613 139540719195904 logging_writer.py:48] [138749] accumulated_eval_time=442.153, accumulated_logging_time=1.55446, accumulated_submission_time=37976.9, global_step=138749, preemption_count=0, score=37976.9, test/accuracy=0.6233000159263611, test/loss=1.8521956205368042, test/num_examples=10000, total_duration=38421.7, train/accuracy=0.9393933415412903, train/loss=0.21290293335914612, validation/accuracy=0.7516199946403503, validation/loss=1.064461588859558, validation/num_examples=50000 +I0831 19:27:05.640232 139540727588608 logging_writer.py:48] [138800] global_step=138800, grad_norm=4.871703624725342, loss=0.9902591109275818 +I0831 19:27:32.801669 139540719195904 logging_writer.py:48] [138900] global_step=138900, grad_norm=4.864016532897949, loss=0.9534199237823486 +I0831 19:27:59.860907 139540727588608 logging_writer.py:48] [139000] global_step=139000, grad_norm=4.943868160247803, loss=1.0138020515441895 +I0831 19:28:27.203249 139540719195904 logging_writer.py:48] [139100] global_step=139100, grad_norm=4.882791996002197, loss=0.9475374221801758 +I0831 19:28:54.365959 139540727588608 logging_writer.py:48] [139200] global_step=139200, grad_norm=5.117842197418213, loss=0.895937442779541 +I0831 19:29:21.461155 139540719195904 logging_writer.py:48] [139300] global_step=139300, grad_norm=4.844820499420166, loss=0.9967927932739258 +I0831 19:29:48.566958 139540727588608 logging_writer.py:48] [139400] global_step=139400, grad_norm=5.00095272064209, loss=0.9310852289199829 +I0831 19:30:15.733570 139540719195904 logging_writer.py:48] [139500] global_step=139500, grad_norm=5.344364166259766, loss=1.000067949295044 +I0831 19:30:42.862852 139540727588608 logging_writer.py:48] [139600] global_step=139600, grad_norm=4.895310401916504, loss=1.0049301385879517 +I0831 19:31:09.993925 139540719195904 logging_writer.py:48] [139700] global_step=139700, grad_norm=4.81271505355835, loss=0.9203932285308838 +I0831 19:31:37.190110 139540727588608 logging_writer.py:48] [139800] global_step=139800, grad_norm=4.862547397613525, loss=0.9668060541152954 +I0831 19:32:04.279247 139540719195904 logging_writer.py:48] [139900] global_step=139900, grad_norm=4.919333457946777, loss=0.9554110169410706 +I0831 19:32:31.379251 139540727588608 logging_writer.py:48] [140000] global_step=140000, grad_norm=5.143237113952637, loss=0.9793355464935303 +I0831 19:32:58.753179 139540719195904 logging_writer.py:48] [140100] global_step=140100, grad_norm=4.6883864402771, loss=0.8688985109329224 +I0831 19:33:25.850000 139540727588608 logging_writer.py:48] [140200] global_step=140200, grad_norm=4.884824275970459, loss=0.9146053791046143 +I0831 19:33:52.942326 139540719195904 logging_writer.py:48] [140300] global_step=140300, grad_norm=4.7034478187561035, loss=0.9397494196891785 +I0831 19:34:20.127197 139540727588608 logging_writer.py:48] [140400] global_step=140400, grad_norm=4.777933120727539, loss=0.9515289664268494 +I0831 19:34:47.238719 139540719195904 logging_writer.py:48] [140500] global_step=140500, grad_norm=4.760321140289307, loss=0.8818340301513672 +I0831 19:35:14.339283 139540727588608 logging_writer.py:48] [140600] global_step=140600, grad_norm=4.550999164581299, loss=0.8968168497085571 +I0831 19:35:41.517937 139540719195904 logging_writer.py:48] [140700] global_step=140700, grad_norm=4.791899681091309, loss=0.8812588453292847 +I0831 19:36:08.595788 139540727588608 logging_writer.py:48] [140800] global_step=140800, grad_norm=5.185793399810791, loss=1.0263419151306152 +I0831 19:36:35.680443 139540719195904 logging_writer.py:48] [140900] global_step=140900, grad_norm=4.8069610595703125, loss=0.9864439964294434 +I0831 19:37:02.832250 139540727588608 logging_writer.py:48] [141000] global_step=141000, grad_norm=4.9310784339904785, loss=1.022621989250183 +I0831 19:37:29.946282 139540719195904 logging_writer.py:48] [141100] global_step=141100, grad_norm=4.714712619781494, loss=0.8815237283706665 +I0831 19:37:57.249381 139540727588608 logging_writer.py:48] [141200] global_step=141200, grad_norm=5.120866775512695, loss=1.0595366954803467 +I0831 19:38:24.359040 139540719195904 logging_writer.py:48] [141300] global_step=141300, grad_norm=4.798666477203369, loss=0.9772976636886597 +I0831 19:38:51.426273 139540727588608 logging_writer.py:48] [141400] global_step=141400, grad_norm=5.009456157684326, loss=1.000659704208374 +I0831 19:39:18.521596 139540719195904 logging_writer.py:48] [141500] global_step=141500, grad_norm=4.891257286071777, loss=0.9247495532035828 +I0831 19:39:45.727790 139540727588608 logging_writer.py:48] [141600] global_step=141600, grad_norm=5.019777774810791, loss=1.0210984945297241 +I0831 19:40:12.843512 139540719195904 logging_writer.py:48] [141700] global_step=141700, grad_norm=4.889074325561523, loss=0.9506536722183228 +I0831 19:40:39.940578 139540727588608 logging_writer.py:48] [141800] global_step=141800, grad_norm=4.792128562927246, loss=0.9047776460647583 +I0831 19:41:07.105039 139540719195904 logging_writer.py:48] [141900] global_step=141900, grad_norm=5.19428014755249, loss=1.0638078451156616 +I0831 19:41:34.182198 139540727588608 logging_writer.py:48] [142000] global_step=142000, grad_norm=5.102656364440918, loss=1.09727144241333 +I0831 19:42:01.301512 139540719195904 logging_writer.py:48] [142100] global_step=142100, grad_norm=4.784899711608887, loss=0.8872174024581909 +I0831 19:42:28.466230 139540727588608 logging_writer.py:48] [142200] global_step=142200, grad_norm=4.78377103805542, loss=1.0330266952514648 +I0831 19:42:55.802825 139540719195904 logging_writer.py:48] [142300] global_step=142300, grad_norm=5.148273468017578, loss=1.036495566368103 +I0831 19:43:22.916399 139540727588608 logging_writer.py:48] [142400] global_step=142400, grad_norm=4.943055629730225, loss=0.9457277655601501 +I0831 19:43:50.087111 139540719195904 logging_writer.py:48] [142500] global_step=142500, grad_norm=4.858746528625488, loss=1.0003998279571533 +I0831 19:44:17.172626 139540727588608 logging_writer.py:48] [142600] global_step=142600, grad_norm=4.9623494148254395, loss=1.055019497871399 +I0831 19:44:44.286793 139540719195904 logging_writer.py:48] [142700] global_step=142700, grad_norm=4.758121013641357, loss=0.9100338816642761 +I0831 19:45:11.457955 139540727588608 logging_writer.py:48] [142800] global_step=142800, grad_norm=4.94889497756958, loss=1.0123060941696167 +I0831 19:45:38.522715 139540719195904 logging_writer.py:48] [142900] global_step=142900, grad_norm=4.899933338165283, loss=0.9546157717704773 +I0831 19:46:05.611622 139540727588608 logging_writer.py:48] [143000] global_step=143000, grad_norm=4.883424282073975, loss=0.9496175050735474 +I0831 19:46:32.750824 139540719195904 logging_writer.py:48] [143100] global_step=143100, grad_norm=4.79231071472168, loss=0.9300960898399353 +I0831 19:46:59.831888 139540727588608 logging_writer.py:48] [143200] global_step=143200, grad_norm=4.912576675415039, loss=0.9672505259513855 +I0831 19:47:27.136380 139540719195904 logging_writer.py:48] [143300] global_step=143300, grad_norm=4.873079776763916, loss=0.9097369313240051 +I0831 19:47:54.315596 139540727588608 logging_writer.py:48] [143400] global_step=143400, grad_norm=4.949477195739746, loss=0.9759334325790405 +I0831 19:48:21.384182 139540719195904 logging_writer.py:48] [143500] global_step=143500, grad_norm=4.49763822555542, loss=0.9046100974082947 +I0831 19:48:48.475877 139540727588608 logging_writer.py:48] [143600] global_step=143600, grad_norm=5.130477428436279, loss=1.0155138969421387 +I0831 19:49:15.644484 139540719195904 logging_writer.py:48] [143700] global_step=143700, grad_norm=5.003990173339844, loss=0.9180908203125 +I0831 19:49:42.761430 139540727588608 logging_writer.py:48] [143800] global_step=143800, grad_norm=4.996431350708008, loss=0.9707123041152954 +I0831 19:50:09.837382 139540719195904 logging_writer.py:48] [143900] global_step=143900, grad_norm=4.718486785888672, loss=0.9922770261764526 +I0831 19:50:37.002488 139540727588608 logging_writer.py:48] [144000] global_step=144000, grad_norm=5.0917744636535645, loss=0.9691386222839355 +I0831 19:51:04.097808 139540719195904 logging_writer.py:48] [144100] global_step=144100, grad_norm=4.956173896789551, loss=1.018226981163025 +I0831 19:51:31.171204 139540727588608 logging_writer.py:48] [144200] global_step=144200, grad_norm=5.032227993011475, loss=1.0013444423675537 +I0831 19:51:58.322177 139540719195904 logging_writer.py:48] [144300] global_step=144300, grad_norm=4.616175174713135, loss=0.918409526348114 +I0831 19:52:25.644212 139540727588608 logging_writer.py:48] [144400] global_step=144400, grad_norm=4.9026641845703125, loss=0.9930658340454102 +I0831 19:52:52.753218 139540719195904 logging_writer.py:48] [144500] global_step=144500, grad_norm=4.835047245025635, loss=0.919486939907074 +I0831 19:53:19.934508 139540727588608 logging_writer.py:48] [144600] global_step=144600, grad_norm=4.9404616355896, loss=1.0046404600143433 +I0831 19:53:47.043345 139540719195904 logging_writer.py:48] [144700] global_step=144700, grad_norm=4.536959171295166, loss=0.8735368847846985 +I0831 19:54:14.151695 139540727588608 logging_writer.py:48] [144800] global_step=144800, grad_norm=5.001751899719238, loss=1.0408048629760742 +I0831 19:54:41.290979 139540719195904 logging_writer.py:48] [144900] global_step=144900, grad_norm=4.998241424560547, loss=0.9106251001358032 +I0831 19:55:08.386312 139540727588608 logging_writer.py:48] [145000] global_step=145000, grad_norm=4.887907981872559, loss=0.9948021173477173 +I0831 19:55:35.502564 139540719195904 logging_writer.py:48] [145100] global_step=145100, grad_norm=4.806793689727783, loss=1.0001885890960693 +I0831 19:56:02.672939 139540727588608 logging_writer.py:48] [145200] global_step=145200, grad_norm=5.214353561401367, loss=1.0018324851989746 +I0831 19:56:29.795421 139540719195904 logging_writer.py:48] [145300] global_step=145300, grad_norm=4.89080286026001, loss=0.9896283745765686 +I0831 19:56:56.890844 139540727588608 logging_writer.py:48] [145400] global_step=145400, grad_norm=4.898375034332275, loss=1.0072888135910034 +I0831 19:57:24.289321 139540719195904 logging_writer.py:48] [145500] global_step=145500, grad_norm=4.890360355377197, loss=0.956022322177887 +I0831 19:57:51.367664 139540727588608 logging_writer.py:48] [145600] global_step=145600, grad_norm=4.879941463470459, loss=1.0323952436447144 +I0831 19:58:18.475900 139540719195904 logging_writer.py:48] [145700] global_step=145700, grad_norm=5.052237510681152, loss=0.99542635679245 +I0831 19:58:45.654925 139540727588608 logging_writer.py:48] [145800] global_step=145800, grad_norm=5.049266815185547, loss=1.0674571990966797 +I0831 19:59:12.741155 139540719195904 logging_writer.py:48] [145900] global_step=145900, grad_norm=4.9487152099609375, loss=1.0172759294509888 +I0831 19:59:39.841383 139540727588608 logging_writer.py:48] [146000] global_step=146000, grad_norm=5.156197547912598, loss=1.098247766494751 +I0831 20:00:06.997264 139540719195904 logging_writer.py:48] [146100] global_step=146100, grad_norm=4.981577396392822, loss=0.9525746703147888 +I0831 20:00:07.392436 139757377230016 spec.py:333] Evaluating on the training split. +I0831 20:00:14.464185 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 20:00:25.451734 139757377230016 spec.py:363] Evaluating on the test split. +I0831 20:00:26.337083 139757377230016 submission_runner.py:516] Time since start: 40436.67s, Step: 146103, {'train/accuracy': Array(0.9429807, dtype=float32), 'train/loss': Array(0.20439605, dtype=float32), 'validation/accuracy': Array(0.7515, dtype=float32), 'validation/loss': Array(1.066416, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62600005, dtype=float32), 'test/loss': Array(1.8532897, dtype=float32), 'test/num_examples': 10000, 'score': 39972.81428909302, 'total_duration': 40436.67267847061, 'accumulated_submission_time': 39972.81428909302, 'accumulated_eval_time': 461.0949785709381, 'accumulated_logging_time': 1.613271951675415} +I0831 20:00:26.390035 139540727588608 logging_writer.py:48] [146103] accumulated_eval_time=461.095, accumulated_logging_time=1.61327, accumulated_submission_time=39972.8, global_step=146103, preemption_count=0, score=39972.8, test/accuracy=0.6260000467300415, test/loss=1.8532897233963013, test/num_examples=10000, total_duration=40436.7, train/accuracy=0.9429807066917419, train/loss=0.204396054148674, validation/accuracy=0.7515000104904175, validation/loss=1.0664160251617432, validation/num_examples=50000 +I0831 20:00:53.070495 139540719195904 logging_writer.py:48] [146200] global_step=146200, grad_norm=4.803895473480225, loss=0.9562443494796753 +I0831 20:01:20.130908 139540727588608 logging_writer.py:48] [146300] global_step=146300, grad_norm=4.913459300994873, loss=0.9702866077423096 +I0831 20:01:47.255327 139540719195904 logging_writer.py:48] [146400] global_step=146400, grad_norm=5.055189609527588, loss=0.9564161896705627 +I0831 20:02:14.557028 139540727588608 logging_writer.py:48] [146500] global_step=146500, grad_norm=5.276646614074707, loss=0.9162465333938599 +I0831 20:02:41.632702 139540719195904 logging_writer.py:48] [146600] global_step=146600, grad_norm=5.106797218322754, loss=1.01668119430542 +I0831 20:03:08.797596 139540727588608 logging_writer.py:48] [146700] global_step=146700, grad_norm=4.756960391998291, loss=0.8747119903564453 +I0831 20:03:35.898046 139540719195904 logging_writer.py:48] [146800] global_step=146800, grad_norm=5.217245578765869, loss=1.0309994220733643 +I0831 20:04:03.006198 139540727588608 logging_writer.py:48] [146900] global_step=146900, grad_norm=5.080924987792969, loss=0.9244135618209839 +I0831 20:04:30.158472 139540719195904 logging_writer.py:48] [147000] global_step=147000, grad_norm=5.021328449249268, loss=1.0281835794448853 +I0831 20:04:57.236829 139540727588608 logging_writer.py:48] [147100] global_step=147100, grad_norm=5.005340099334717, loss=0.9713712930679321 +I0831 20:05:24.345026 139540719195904 logging_writer.py:48] [147200] global_step=147200, grad_norm=4.924888610839844, loss=0.9444887638092041 +I0831 20:05:51.496099 139540727588608 logging_writer.py:48] [147300] global_step=147300, grad_norm=4.726351261138916, loss=0.9253100156784058 +I0831 20:06:18.576854 139540719195904 logging_writer.py:48] [147400] global_step=147400, grad_norm=4.750677585601807, loss=0.9350607991218567 +I0831 20:06:45.655931 139540727588608 logging_writer.py:48] [147500] global_step=147500, grad_norm=4.881363391876221, loss=0.9277376532554626 +I0831 20:07:13.025550 139540719195904 logging_writer.py:48] [147600] global_step=147600, grad_norm=5.0200347900390625, loss=1.0242422819137573 +I0831 20:07:40.125041 139540727588608 logging_writer.py:48] [147700] global_step=147700, grad_norm=4.979521751403809, loss=0.9865714907646179 +I0831 20:08:07.254504 139540719195904 logging_writer.py:48] [147800] global_step=147800, grad_norm=4.878149509429932, loss=0.8695657253265381 +I0831 20:08:34.430884 139540727588608 logging_writer.py:48] [147900] global_step=147900, grad_norm=5.0644378662109375, loss=0.9221640825271606 +I0831 20:09:01.540736 139540719195904 logging_writer.py:48] [148000] global_step=148000, grad_norm=5.00282621383667, loss=0.956488847732544 +I0831 20:09:28.636377 139540727588608 logging_writer.py:48] [148100] global_step=148100, grad_norm=4.818978309631348, loss=0.9855359196662903 +I0831 20:09:55.776511 139540719195904 logging_writer.py:48] [148200] global_step=148200, grad_norm=5.116714000701904, loss=0.9998196363449097 +I0831 20:10:22.881272 139540727588608 logging_writer.py:48] [148300] global_step=148300, grad_norm=5.023601531982422, loss=0.9741630554199219 +I0831 20:10:49.956039 139540719195904 logging_writer.py:48] [148400] global_step=148400, grad_norm=5.208728790283203, loss=0.9709794521331787 +I0831 20:11:17.110448 139540727588608 logging_writer.py:48] [148500] global_step=148500, grad_norm=4.755453109741211, loss=0.9202561378479004 +I0831 20:11:44.433638 139540719195904 logging_writer.py:48] [148600] global_step=148600, grad_norm=4.8876118659973145, loss=0.9601861238479614 +I0831 20:12:11.539349 139540727588608 logging_writer.py:48] [148700] global_step=148700, grad_norm=4.7784857749938965, loss=0.9150691628456116 +I0831 20:12:38.683444 139540719195904 logging_writer.py:48] [148800] global_step=148800, grad_norm=4.942794322967529, loss=0.9174060821533203 +I0831 20:13:05.791623 139540727588608 logging_writer.py:48] [148900] global_step=148900, grad_norm=4.999746322631836, loss=0.9616286754608154 +I0831 20:13:32.873371 139540719195904 logging_writer.py:48] [149000] global_step=149000, grad_norm=4.60487174987793, loss=0.8965438604354858 +I0831 20:14:00.037812 139540727588608 logging_writer.py:48] [149100] global_step=149100, grad_norm=5.144637107849121, loss=0.9518135786056519 +I0831 20:14:27.125368 139540719195904 logging_writer.py:48] [149200] global_step=149200, grad_norm=4.869401931762695, loss=0.8913272619247437 +I0831 20:14:54.236735 139540727588608 logging_writer.py:48] [149300] global_step=149300, grad_norm=5.179067134857178, loss=1.0208237171173096 +I0831 20:15:21.399432 139540719195904 logging_writer.py:48] [149400] global_step=149400, grad_norm=5.070855617523193, loss=0.9608543515205383 +I0831 20:15:48.483337 139540727588608 logging_writer.py:48] [149500] global_step=149500, grad_norm=5.029384613037109, loss=0.989356517791748 +I0831 20:16:15.581731 139540719195904 logging_writer.py:48] [149600] global_step=149600, grad_norm=4.937863826751709, loss=0.9190988540649414 +I0831 20:16:42.925513 139540727588608 logging_writer.py:48] [149700] global_step=149700, grad_norm=4.742585182189941, loss=0.908256471157074 +I0831 20:17:10.048130 139540719195904 logging_writer.py:48] [149800] global_step=149800, grad_norm=5.142239093780518, loss=0.996779203414917 +I0831 20:17:37.155332 139540727588608 logging_writer.py:48] [149900] global_step=149900, grad_norm=5.354884624481201, loss=0.9873426556587219 +I0831 20:18:04.290891 139540719195904 logging_writer.py:48] [150000] global_step=150000, grad_norm=4.6390814781188965, loss=0.9458861351013184 +I0831 20:18:31.373616 139540727588608 logging_writer.py:48] [150100] global_step=150100, grad_norm=4.672126770019531, loss=0.9393327832221985 +I0831 20:18:58.460784 139540719195904 logging_writer.py:48] [150200] global_step=150200, grad_norm=4.948758125305176, loss=0.9837696552276611 +I0831 20:19:25.657743 139540727588608 logging_writer.py:48] [150300] global_step=150300, grad_norm=4.881545543670654, loss=0.955140233039856 +I0831 20:19:52.781147 139540719195904 logging_writer.py:48] [150400] global_step=150400, grad_norm=5.178322792053223, loss=1.1292624473571777 +I0831 20:20:19.857145 139540727588608 logging_writer.py:48] [150500] global_step=150500, grad_norm=4.8780999183654785, loss=1.000759243965149 +I0831 20:20:47.039058 139540719195904 logging_writer.py:48] [150600] global_step=150600, grad_norm=5.027082443237305, loss=0.9273658394813538 +I0831 20:21:14.330285 139540727588608 logging_writer.py:48] [150700] global_step=150700, grad_norm=4.808567047119141, loss=0.970081090927124 +I0831 20:21:41.433551 139540719195904 logging_writer.py:48] [150800] global_step=150800, grad_norm=4.821190357208252, loss=0.9482805728912354 +I0831 20:22:08.599484 139540727588608 logging_writer.py:48] [150900] global_step=150900, grad_norm=5.027789115905762, loss=1.0106550455093384 +I0831 20:22:35.719373 139540719195904 logging_writer.py:48] [151000] global_step=151000, grad_norm=5.02651834487915, loss=0.9429337382316589 +I0831 20:23:02.809004 139540727588608 logging_writer.py:48] [151100] global_step=151100, grad_norm=5.158106803894043, loss=0.8859448432922363 +I0831 20:23:29.949820 139540719195904 logging_writer.py:48] [151200] global_step=151200, grad_norm=4.905797004699707, loss=0.9547634720802307 +I0831 20:23:57.019485 139540727588608 logging_writer.py:48] [151300] global_step=151300, grad_norm=4.807074069976807, loss=0.920608401298523 +I0831 20:24:24.102129 139540719195904 logging_writer.py:48] [151400] global_step=151400, grad_norm=4.82431173324585, loss=0.9498525857925415 +I0831 20:24:51.242847 139540727588608 logging_writer.py:48] [151500] global_step=151500, grad_norm=4.917196750640869, loss=0.9390757083892822 +I0831 20:25:18.318909 139540719195904 logging_writer.py:48] [151600] global_step=151600, grad_norm=5.097172260284424, loss=0.9732065200805664 +I0831 20:25:45.448368 139540727588608 logging_writer.py:48] [151700] global_step=151700, grad_norm=4.972909450531006, loss=0.9788637161254883 +I0831 20:26:12.829339 139540719195904 logging_writer.py:48] [151800] global_step=151800, grad_norm=4.862555503845215, loss=0.9667122960090637 +I0831 20:26:39.916192 139540727588608 logging_writer.py:48] [151900] global_step=151900, grad_norm=5.491297245025635, loss=1.0652008056640625 +I0831 20:27:07.021927 139540719195904 logging_writer.py:48] [152000] global_step=152000, grad_norm=4.955933094024658, loss=0.9750006794929504 +I0831 20:27:34.191848 139540727588608 logging_writer.py:48] [152100] global_step=152100, grad_norm=4.498730182647705, loss=0.9273554682731628 +I0831 20:28:01.316348 139540719195904 logging_writer.py:48] [152200] global_step=152200, grad_norm=4.753042697906494, loss=0.9542335867881775 +I0831 20:28:28.401802 139540727588608 logging_writer.py:48] [152300] global_step=152300, grad_norm=5.609224319458008, loss=0.9455620646476746 +I0831 20:28:55.562267 139540719195904 logging_writer.py:48] [152400] global_step=152400, grad_norm=5.08140230178833, loss=0.9581665992736816 +I0831 20:29:22.660192 139540727588608 logging_writer.py:48] [152500] global_step=152500, grad_norm=4.9004316329956055, loss=1.0175868272781372 +I0831 20:29:49.789518 139540719195904 logging_writer.py:48] [152600] global_step=152600, grad_norm=5.101755142211914, loss=1.0552104711532593 +I0831 20:30:16.977444 139540727588608 logging_writer.py:48] [152700] global_step=152700, grad_norm=5.044900417327881, loss=0.9416693449020386 +I0831 20:30:44.302823 139540719195904 logging_writer.py:48] [152800] global_step=152800, grad_norm=5.0514421463012695, loss=0.9287148118019104 +I0831 20:31:11.413744 139540727588608 logging_writer.py:48] [152900] global_step=152900, grad_norm=4.612090110778809, loss=0.874138593673706 +I0831 20:31:38.552332 139540719195904 logging_writer.py:48] [153000] global_step=153000, grad_norm=4.922503471374512, loss=0.9936745762825012 +I0831 20:32:05.663253 139540727588608 logging_writer.py:48] [153100] global_step=153100, grad_norm=4.998672008514404, loss=1.0102415084838867 +I0831 20:32:32.726550 139540719195904 logging_writer.py:48] [153200] global_step=153200, grad_norm=4.843101501464844, loss=0.8244028687477112 +I0831 20:32:59.888169 139540727588608 logging_writer.py:48] [153300] global_step=153300, grad_norm=5.432653903961182, loss=1.0285844802856445 +I0831 20:33:26.990855 139540719195904 logging_writer.py:48] [153400] global_step=153400, grad_norm=4.879928112030029, loss=0.9396483898162842 +I0831 20:33:42.525733 139757377230016 spec.py:333] Evaluating on the training split. +I0831 20:33:49.525431 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 20:34:00.156702 139757377230016 spec.py:363] Evaluating on the test split. +I0831 20:34:01.038476 139757377230016 submission_runner.py:516] Time since start: 42451.37s, Step: 153459, {'train/accuracy': Array(0.9436384, dtype=float32), 'train/loss': Array(0.19484542, dtype=float32), 'validation/accuracy': Array(0.7524, dtype=float32), 'validation/loss': Array(1.0662144, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6261, dtype=float32), 'test/loss': Array(1.8529103, dtype=float32), 'test/num_examples': 10000, 'score': 41968.87982940674, 'total_duration': 42451.374479055405, 'accumulated_submission_time': 41968.87982940674, 'accumulated_eval_time': 479.6059491634369, 'accumulated_logging_time': 1.682976484298706} +I0831 20:34:01.099385 139540727588608 logging_writer.py:48] [153459] accumulated_eval_time=479.606, accumulated_logging_time=1.68298, accumulated_submission_time=41968.9, global_step=153459, preemption_count=0, score=41968.9, test/accuracy=0.6261000037193298, test/loss=1.8529102802276611, test/num_examples=10000, total_duration=42451.4, train/accuracy=0.9436383843421936, train/loss=0.19484542310237885, validation/accuracy=0.7523999810218811, validation/loss=1.0662144422531128, validation/num_examples=50000 +I0831 20:34:12.586485 139540719195904 logging_writer.py:48] [153500] global_step=153500, grad_norm=5.084113597869873, loss=1.0130479335784912 +I0831 20:34:39.736325 139540727588608 logging_writer.py:48] [153600] global_step=153600, grad_norm=5.080477714538574, loss=1.0084259510040283 +I0831 20:35:06.812600 139540719195904 logging_writer.py:48] [153700] global_step=153700, grad_norm=5.0858964920043945, loss=0.9927721619606018 +I0831 20:35:33.915843 139540727588608 logging_writer.py:48] [153800] global_step=153800, grad_norm=4.949362754821777, loss=0.9818704724311829 +I0831 20:36:01.275110 139540719195904 logging_writer.py:48] [153900] global_step=153900, grad_norm=4.85859489440918, loss=0.9233124256134033 +I0831 20:36:28.360273 139540727588608 logging_writer.py:48] [154000] global_step=154000, grad_norm=4.657138824462891, loss=0.8708588480949402 +I0831 20:36:55.433013 139540719195904 logging_writer.py:48] [154100] global_step=154100, grad_norm=5.059042453765869, loss=1.0081186294555664 +I0831 20:37:22.581084 139540727588608 logging_writer.py:48] [154200] global_step=154200, grad_norm=5.427408695220947, loss=1.0356488227844238 +I0831 20:37:52.826363 139540719195904 logging_writer.py:48] [154300] global_step=154300, grad_norm=4.798146724700928, loss=0.9370341897010803 +I0831 20:38:19.881365 139540727588608 logging_writer.py:48] [154400] global_step=154400, grad_norm=4.671112537384033, loss=0.8765661716461182 +I0831 20:38:47.016004 139540719195904 logging_writer.py:48] [154500] global_step=154500, grad_norm=4.852830410003662, loss=0.9968103170394897 +I0831 20:39:14.117785 139540727588608 logging_writer.py:48] [154600] global_step=154600, grad_norm=4.988564968109131, loss=0.9660347700119019 +I0831 20:39:41.232426 139540719195904 logging_writer.py:48] [154700] global_step=154700, grad_norm=4.607135772705078, loss=0.8240559101104736 +I0831 20:40:08.404094 139540727588608 logging_writer.py:48] [154800] global_step=154800, grad_norm=4.772922515869141, loss=0.886858344078064 +I0831 20:40:35.726820 139540719195904 logging_writer.py:48] [154900] global_step=154900, grad_norm=4.715194225311279, loss=0.9271547794342041 +I0831 20:41:02.792451 139540727588608 logging_writer.py:48] [155000] global_step=155000, grad_norm=5.412055492401123, loss=0.9547130465507507 +I0831 20:41:29.982198 139540719195904 logging_writer.py:48] [155100] global_step=155100, grad_norm=5.200058937072754, loss=0.9818491339683533 +I0831 20:41:57.098812 139540727588608 logging_writer.py:48] [155200] global_step=155200, grad_norm=4.868923664093018, loss=0.9547141790390015 +I0831 20:42:24.204064 139540719195904 logging_writer.py:48] [155300] global_step=155300, grad_norm=5.250529766082764, loss=1.0281522274017334 +I0831 20:42:51.384639 139540727588608 logging_writer.py:48] [155400] global_step=155400, grad_norm=5.072220802307129, loss=1.065803050994873 +I0831 20:43:18.450259 139540719195904 logging_writer.py:48] [155500] global_step=155500, grad_norm=5.155465126037598, loss=1.0150226354599 +I0831 20:43:45.533814 139540727588608 logging_writer.py:48] [155600] global_step=155600, grad_norm=5.0285539627075195, loss=1.0055058002471924 +I0831 20:44:12.692747 139540719195904 logging_writer.py:48] [155700] global_step=155700, grad_norm=4.728820323944092, loss=0.9416544437408447 +I0831 20:44:39.806541 139540727588608 logging_writer.py:48] [155800] global_step=155800, grad_norm=5.025786876678467, loss=0.9812809228897095 +I0831 20:45:06.929914 139540719195904 logging_writer.py:48] [155900] global_step=155900, grad_norm=4.715885639190674, loss=0.8785211443901062 +I0831 20:45:34.324510 139540727588608 logging_writer.py:48] [156000] global_step=156000, grad_norm=5.601651668548584, loss=1.0576088428497314 +I0831 20:46:01.391088 139540719195904 logging_writer.py:48] [156100] global_step=156100, grad_norm=4.879785537719727, loss=0.9525125622749329 +I0831 20:46:28.477564 139540727588608 logging_writer.py:48] [156200] global_step=156200, grad_norm=4.730650424957275, loss=0.9632541537284851 +I0831 20:46:55.660117 139540719195904 logging_writer.py:48] [156300] global_step=156300, grad_norm=5.4593424797058105, loss=1.0482864379882812 +I0831 20:47:22.728219 139540727588608 logging_writer.py:48] [156400] global_step=156400, grad_norm=5.184152126312256, loss=0.9838722944259644 +I0831 20:47:49.826949 139540719195904 logging_writer.py:48] [156500] global_step=156500, grad_norm=5.061868667602539, loss=0.9094336628913879 +I0831 20:48:16.967075 139540727588608 logging_writer.py:48] [156600] global_step=156600, grad_norm=5.1330366134643555, loss=0.9722123146057129 +I0831 20:48:44.065597 139540719195904 logging_writer.py:48] [156700] global_step=156700, grad_norm=5.047150135040283, loss=0.868276834487915 +I0831 20:49:11.150758 139540727588608 logging_writer.py:48] [156800] global_step=156800, grad_norm=4.841760158538818, loss=0.9090273380279541 +I0831 20:49:38.299825 139540719195904 logging_writer.py:48] [156900] global_step=156900, grad_norm=5.305118560791016, loss=0.992932915687561 +I0831 20:50:05.580863 139540727588608 logging_writer.py:48] [157000] global_step=157000, grad_norm=4.81522274017334, loss=0.9100302457809448 +I0831 20:50:32.660198 139540719195904 logging_writer.py:48] [157100] global_step=157100, grad_norm=4.793705463409424, loss=0.9707159996032715 +I0831 20:50:59.806876 139540727588608 logging_writer.py:48] [157200] global_step=157200, grad_norm=4.846596717834473, loss=0.9258076548576355 +I0831 20:51:26.867761 139540719195904 logging_writer.py:48] [157300] global_step=157300, grad_norm=4.937160015106201, loss=0.9036328792572021 +I0831 20:51:53.955304 139540727588608 logging_writer.py:48] [157400] global_step=157400, grad_norm=5.080944061279297, loss=1.0339336395263672 +I0831 20:52:21.114437 139540719195904 logging_writer.py:48] [157500] global_step=157500, grad_norm=5.1489338874816895, loss=0.9846539497375488 +I0831 20:52:48.175624 139540727588608 logging_writer.py:48] [157600] global_step=157600, grad_norm=5.130618572235107, loss=0.9960851669311523 +I0831 20:53:15.255642 139540719195904 logging_writer.py:48] [157700] global_step=157700, grad_norm=5.270749568939209, loss=1.0861318111419678 +I0831 20:53:42.459716 139540727588608 logging_writer.py:48] [157800] global_step=157800, grad_norm=5.177969455718994, loss=1.0285476446151733 +I0831 20:54:09.550011 139540719195904 logging_writer.py:48] [157900] global_step=157900, grad_norm=4.862735271453857, loss=0.9337354898452759 +I0831 20:54:36.668863 139540727588608 logging_writer.py:48] [158000] global_step=158000, grad_norm=5.149179935455322, loss=0.9658536911010742 +I0831 20:55:04.021741 139540719195904 logging_writer.py:48] [158100] global_step=158100, grad_norm=5.340179443359375, loss=1.0032298564910889 +I0831 20:55:31.110728 139540727588608 logging_writer.py:48] [158200] global_step=158200, grad_norm=5.184858322143555, loss=0.9616742134094238 +I0831 20:55:58.208556 139540719195904 logging_writer.py:48] [158300] global_step=158300, grad_norm=4.977093696594238, loss=0.9402259588241577 +I0831 20:56:25.353538 139540727588608 logging_writer.py:48] [158400] global_step=158400, grad_norm=5.110679626464844, loss=1.015777349472046 +I0831 20:56:52.466166 139540719195904 logging_writer.py:48] [158500] global_step=158500, grad_norm=4.662341594696045, loss=0.8328948616981506 +I0831 20:57:19.582021 139540727588608 logging_writer.py:48] [158600] global_step=158600, grad_norm=5.0572190284729, loss=1.030425786972046 +I0831 20:57:46.727137 139540719195904 logging_writer.py:48] [158700] global_step=158700, grad_norm=4.930996894836426, loss=0.960013747215271 +I0831 20:58:13.826636 139540727588608 logging_writer.py:48] [158800] global_step=158800, grad_norm=5.361905097961426, loss=1.0094714164733887 +I0831 20:58:40.925853 139540719195904 logging_writer.py:48] [158900] global_step=158900, grad_norm=5.031557559967041, loss=0.9156042337417603 +I0831 20:59:08.082754 139540727588608 logging_writer.py:48] [159000] global_step=159000, grad_norm=5.041658878326416, loss=0.973647952079773 +I0831 20:59:35.157984 139540719195904 logging_writer.py:48] [159100] global_step=159100, grad_norm=4.904654026031494, loss=0.906226396560669 +I0831 21:00:02.469564 139540727588608 logging_writer.py:48] [159200] global_step=159200, grad_norm=5.0043792724609375, loss=0.9613056182861328 +I0831 21:00:29.762285 139540719195904 logging_writer.py:48] [159300] global_step=159300, grad_norm=4.8167805671691895, loss=0.9794607162475586 +I0831 21:00:56.865185 139540727588608 logging_writer.py:48] [159400] global_step=159400, grad_norm=5.0650763511657715, loss=0.9407323598861694 +I0831 21:01:23.955609 139540719195904 logging_writer.py:48] [159500] global_step=159500, grad_norm=5.127701282501221, loss=0.9812685251235962 +I0831 21:01:51.076891 139540727588608 logging_writer.py:48] [159600] global_step=159600, grad_norm=5.019761562347412, loss=0.9198150634765625 +I0831 21:02:18.181086 139540719195904 logging_writer.py:48] [159700] global_step=159700, grad_norm=4.790411472320557, loss=0.9666284322738647 +I0831 21:02:45.283429 139540727588608 logging_writer.py:48] [159800] global_step=159800, grad_norm=5.233885765075684, loss=0.9517183303833008 +I0831 21:03:12.447749 139540719195904 logging_writer.py:48] [159900] global_step=159900, grad_norm=5.068629264831543, loss=0.9952079653739929 +I0831 21:03:39.544128 139540727588608 logging_writer.py:48] [160000] global_step=160000, grad_norm=4.860971450805664, loss=0.9443888664245605 +I0831 21:04:06.651104 139540719195904 logging_writer.py:48] [160100] global_step=160100, grad_norm=4.965539455413818, loss=0.9381277561187744 +I0831 21:04:34.006537 139540727588608 logging_writer.py:48] [160200] global_step=160200, grad_norm=5.004106044769287, loss=0.9026545286178589 +I0831 21:05:01.079860 139540719195904 logging_writer.py:48] [160300] global_step=160300, grad_norm=5.2449564933776855, loss=1.0446202754974365 +I0831 21:05:28.215903 139540727588608 logging_writer.py:48] [160400] global_step=160400, grad_norm=4.820967674255371, loss=0.8940549492835999 +I0831 21:05:55.367661 139540719195904 logging_writer.py:48] [160500] global_step=160500, grad_norm=4.867816925048828, loss=0.9113267660140991 +I0831 21:06:22.483599 139540727588608 logging_writer.py:48] [160600] global_step=160600, grad_norm=5.097231864929199, loss=0.9306005239486694 +I0831 21:06:49.608131 139540719195904 logging_writer.py:48] [160700] global_step=160700, grad_norm=4.905991077423096, loss=0.9172671437263489 +I0831 21:07:16.761424 139540727588608 logging_writer.py:48] [160800] global_step=160800, grad_norm=5.203260898590088, loss=0.9387592673301697 +I0831 21:07:17.165450 139757377230016 spec.py:333] Evaluating on the training split. +I0831 21:07:24.243979 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 21:07:35.883810 139757377230016 spec.py:363] Evaluating on the test split. +I0831 21:07:36.765338 139757377230016 submission_runner.py:516] Time since start: 44467.10s, Step: 160803, {'train/accuracy': Array(0.94782364, dtype=float32), 'train/loss': Array(0.18040912, dtype=float32), 'validation/accuracy': Array(0.75266, dtype=float32), 'validation/loss': Array(1.0689249, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62670004, dtype=float32), 'test/loss': Array(1.859921, dtype=float32), 'test/num_examples': 10000, 'score': 43964.87986493111, 'total_duration': 44467.10048651695, 'accumulated_submission_time': 43964.87986493111, 'accumulated_eval_time': 499.20321011543274, 'accumulated_logging_time': 1.7564749717712402} +I0831 21:07:36.828720 139540719195904 logging_writer.py:48] [160803] accumulated_eval_time=499.203, accumulated_logging_time=1.75647, accumulated_submission_time=43964.9, global_step=160803, preemption_count=0, score=43964.9, test/accuracy=0.6267000436782837, test/loss=1.8599209785461426, test/num_examples=10000, total_duration=44467.1, train/accuracy=0.9478236436843872, train/loss=0.18040911853313446, validation/accuracy=0.7526599764823914, validation/loss=1.068924903869629, validation/num_examples=50000 +I0831 21:08:03.449329 139540727588608 logging_writer.py:48] [160900] global_step=160900, grad_norm=5.037517070770264, loss=1.014204978942871 +I0831 21:08:43.931758 139540719195904 logging_writer.py:48] [161000] global_step=161000, grad_norm=4.8889546394348145, loss=0.931659460067749 +I0831 21:09:29.322592 139540727588608 logging_writer.py:48] [161100] global_step=161100, grad_norm=5.163882255554199, loss=0.9386011362075806 +I0831 21:10:13.597917 139540719195904 logging_writer.py:48] [161200] global_step=161200, grad_norm=4.96826696395874, loss=0.9062102437019348 +I0831 21:10:42.245127 139540727588608 logging_writer.py:48] [161300] global_step=161300, grad_norm=4.9312639236450195, loss=0.9848003387451172 +I0831 21:11:09.413370 139540719195904 logging_writer.py:48] [161400] global_step=161400, grad_norm=4.672180652618408, loss=0.9350361824035645 +I0831 21:11:37.001859 139540727588608 logging_writer.py:48] [161500] global_step=161500, grad_norm=4.6825995445251465, loss=0.8460814952850342 +I0831 21:12:04.087852 139540719195904 logging_writer.py:48] [161600] global_step=161600, grad_norm=4.860040187835693, loss=0.8850826025009155 +I0831 21:12:31.208346 139540727588608 logging_writer.py:48] [161700] global_step=161700, grad_norm=5.053499221801758, loss=0.9557758569717407 +I0831 21:12:58.308220 139540719195904 logging_writer.py:48] [161800] global_step=161800, grad_norm=4.920664310455322, loss=0.907818615436554 +I0831 21:13:25.426695 139540727588608 logging_writer.py:48] [161900] global_step=161900, grad_norm=5.013975143432617, loss=0.9702956676483154 +I0831 21:13:52.587983 139540719195904 logging_writer.py:48] [162000] global_step=162000, grad_norm=5.0365519523620605, loss=1.014796257019043 +I0831 21:14:19.693897 139540727588608 logging_writer.py:48] [162100] global_step=162100, grad_norm=5.030987739562988, loss=0.9494566917419434 +I0831 21:14:46.794259 139540719195904 logging_writer.py:48] [162200] global_step=162200, grad_norm=4.881530284881592, loss=0.9614452123641968 +I0831 21:15:13.923695 139540727588608 logging_writer.py:48] [162300] global_step=162300, grad_norm=5.180329322814941, loss=0.9883328676223755 +I0831 21:15:41.308850 139540719195904 logging_writer.py:48] [162400] global_step=162400, grad_norm=4.720633029937744, loss=0.9268121719360352 +I0831 21:16:08.377823 139540727588608 logging_writer.py:48] [162500] global_step=162500, grad_norm=5.08119010925293, loss=0.9725327491760254 +I0831 21:16:35.536183 139540719195904 logging_writer.py:48] [162600] global_step=162600, grad_norm=4.925673961639404, loss=0.916546642780304 +I0831 21:17:02.643654 139540727588608 logging_writer.py:48] [162700] global_step=162700, grad_norm=5.053299427032471, loss=0.9621924161911011 +I0831 21:17:29.722517 139540719195904 logging_writer.py:48] [162800] global_step=162800, grad_norm=5.242842674255371, loss=1.0133157968521118 +I0831 21:17:56.902817 139540727588608 logging_writer.py:48] [162900] global_step=162900, grad_norm=5.092436790466309, loss=0.9443873167037964 +I0831 21:18:24.003329 139540719195904 logging_writer.py:48] [163000] global_step=163000, grad_norm=4.980007171630859, loss=0.946797788143158 +I0831 21:18:51.079943 139540727588608 logging_writer.py:48] [163100] global_step=163100, grad_norm=5.32796573638916, loss=1.0417016744613647 +I0831 21:19:18.220264 139540719195904 logging_writer.py:48] [163200] global_step=163200, grad_norm=4.902069091796875, loss=0.9210683703422546 +I0831 21:19:45.285390 139540727588608 logging_writer.py:48] [163300] global_step=163300, grad_norm=4.991593837738037, loss=0.9212214946746826 +I0831 21:20:12.618872 139540719195904 logging_writer.py:48] [163400] global_step=163400, grad_norm=5.0289812088012695, loss=0.9454931020736694 +I0831 21:20:39.791423 139540727588608 logging_writer.py:48] [163500] global_step=163500, grad_norm=5.080045700073242, loss=0.9385354518890381 +I0831 21:21:06.886953 139540719195904 logging_writer.py:48] [163600] global_step=163600, grad_norm=4.892621040344238, loss=0.9650776982307434 +I0831 21:21:33.983392 139540727588608 logging_writer.py:48] [163700] global_step=163700, grad_norm=4.910308837890625, loss=0.9428973197937012 +I0831 21:22:01.148880 139540719195904 logging_writer.py:48] [163800] global_step=163800, grad_norm=5.152501583099365, loss=0.9682492017745972 +I0831 21:22:28.242682 139540727588608 logging_writer.py:48] [163900] global_step=163900, grad_norm=5.202320575714111, loss=0.9867275953292847 +I0831 21:22:55.372308 139540719195904 logging_writer.py:48] [164000] global_step=164000, grad_norm=5.076254844665527, loss=0.9135780930519104 +I0831 21:23:22.526411 139540727588608 logging_writer.py:48] [164100] global_step=164100, grad_norm=4.870944023132324, loss=0.9574759006500244 +I0831 21:23:49.607619 139540719195904 logging_writer.py:48] [164200] global_step=164200, grad_norm=5.016909599304199, loss=0.9958412051200867 +I0831 21:24:16.716997 139540727588608 logging_writer.py:48] [164300] global_step=164300, grad_norm=4.9966511726379395, loss=0.9120484590530396 +I0831 21:24:43.850626 139540719195904 logging_writer.py:48] [164400] global_step=164400, grad_norm=4.668238639831543, loss=0.8537229895591736 +I0831 21:25:11.162338 139540727588608 logging_writer.py:48] [164500] global_step=164500, grad_norm=5.56754207611084, loss=1.0777037143707275 +I0831 21:25:38.289786 139540719195904 logging_writer.py:48] [164600] global_step=164600, grad_norm=5.247561454772949, loss=1.041256308555603 +I0831 21:26:05.433890 139540727588608 logging_writer.py:48] [164700] global_step=164700, grad_norm=4.908855438232422, loss=1.0104761123657227 +I0831 21:26:32.524930 139540719195904 logging_writer.py:48] [164800] global_step=164800, grad_norm=5.144124507904053, loss=0.9566861391067505 +I0831 21:26:59.615675 139540727588608 logging_writer.py:48] [164900] global_step=164900, grad_norm=5.10759973526001, loss=0.8996327519416809 +I0831 21:27:26.772604 139540719195904 logging_writer.py:48] [165000] global_step=165000, grad_norm=5.097186088562012, loss=1.001872181892395 +I0831 21:27:53.887424 139540727588608 logging_writer.py:48] [165100] global_step=165100, grad_norm=5.031259536743164, loss=0.9734193086624146 +I0831 21:28:20.982799 139540719195904 logging_writer.py:48] [165200] global_step=165200, grad_norm=5.001436710357666, loss=0.8973974585533142 +I0831 21:28:48.128183 139540727588608 logging_writer.py:48] [165300] global_step=165300, grad_norm=5.159297466278076, loss=0.9280201196670532 +I0831 21:29:15.218154 139540719195904 logging_writer.py:48] [165400] global_step=165400, grad_norm=5.284389019012451, loss=1.0294299125671387 +I0831 21:29:42.510778 139540727588608 logging_writer.py:48] [165500] global_step=165500, grad_norm=4.799660682678223, loss=0.8889108896255493 +I0831 21:30:09.674383 139540719195904 logging_writer.py:48] [165600] global_step=165600, grad_norm=5.063992500305176, loss=0.9494656324386597 +I0831 21:30:36.811352 139540727588608 logging_writer.py:48] [165700] global_step=165700, grad_norm=4.989663124084473, loss=0.9808051586151123 +I0831 21:31:03.888327 139540719195904 logging_writer.py:48] [165800] global_step=165800, grad_norm=5.275172710418701, loss=0.9803785085678101 +I0831 21:31:31.058774 139540727588608 logging_writer.py:48] [165900] global_step=165900, grad_norm=5.366086959838867, loss=1.0274399518966675 +I0831 21:31:58.171455 139540719195904 logging_writer.py:48] [166000] global_step=166000, grad_norm=4.808278560638428, loss=0.9155480861663818 +I0831 21:32:25.269322 139540727588608 logging_writer.py:48] [166100] global_step=166100, grad_norm=5.000518798828125, loss=0.9537750482559204 +I0831 21:32:52.429776 139540719195904 logging_writer.py:48] [166200] global_step=166200, grad_norm=5.039129257202148, loss=0.9233691096305847 +I0831 21:33:19.509341 139540727588608 logging_writer.py:48] [166300] global_step=166300, grad_norm=5.154554843902588, loss=0.9348160028457642 +I0831 21:33:46.624729 139540719195904 logging_writer.py:48] [166400] global_step=166400, grad_norm=5.177662372589111, loss=0.9203295707702637 +I0831 21:34:13.783189 139540727588608 logging_writer.py:48] [166500] global_step=166500, grad_norm=5.161464691162109, loss=0.9939601421356201 +I0831 21:34:41.080877 139540719195904 logging_writer.py:48] [166600] global_step=166600, grad_norm=5.1986260414123535, loss=0.9785328507423401 +I0831 21:35:08.144358 139540727588608 logging_writer.py:48] [166700] global_step=166700, grad_norm=5.176137924194336, loss=1.071574330329895 +I0831 21:35:35.298801 139540719195904 logging_writer.py:48] [166800] global_step=166800, grad_norm=5.161008834838867, loss=0.9538118839263916 +I0831 21:36:02.376659 139540727588608 logging_writer.py:48] [166900] global_step=166900, grad_norm=5.189355850219727, loss=0.9111672639846802 +I0831 21:36:29.473583 139540719195904 logging_writer.py:48] [167000] global_step=167000, grad_norm=5.27495002746582, loss=0.958970308303833 +I0831 21:36:56.667683 139540727588608 logging_writer.py:48] [167100] global_step=167100, grad_norm=5.406421184539795, loss=0.8850550055503845 +I0831 21:37:23.729930 139540719195904 logging_writer.py:48] [167200] global_step=167200, grad_norm=5.041898250579834, loss=1.0291266441345215 +I0831 21:37:50.800925 139540727588608 logging_writer.py:48] [167300] global_step=167300, grad_norm=5.264313697814941, loss=0.9286333322525024 +I0831 21:38:17.959343 139540719195904 logging_writer.py:48] [167400] global_step=167400, grad_norm=5.024592399597168, loss=0.972585916519165 +I0831 21:38:45.041955 139540727588608 logging_writer.py:48] [167500] global_step=167500, grad_norm=4.800239086151123, loss=0.9462224245071411 +I0831 21:39:12.321219 139540719195904 logging_writer.py:48] [167600] global_step=167600, grad_norm=4.9373908042907715, loss=0.9882769584655762 +I0831 21:39:39.484157 139540727588608 logging_writer.py:48] [167700] global_step=167700, grad_norm=4.894540309906006, loss=0.9004441499710083 +I0831 21:40:06.565552 139540719195904 logging_writer.py:48] [167800] global_step=167800, grad_norm=5.14202880859375, loss=0.8912181854248047 +I0831 21:40:33.670602 139540727588608 logging_writer.py:48] [167900] global_step=167900, grad_norm=4.999546051025391, loss=0.9719611406326294 +I0831 21:40:52.792378 139757377230016 spec.py:333] Evaluating on the training split. +I0831 21:40:59.773777 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 21:41:45.298582 139757377230016 spec.py:363] Evaluating on the test split. +I0831 21:41:46.194240 139757377230016 submission_runner.py:516] Time since start: 46516.53s, Step: 167972, {'train/accuracy': Array(0.9487404, dtype=float32), 'train/loss': Array(0.18130454, dtype=float32), 'validation/accuracy': Array(0.75286, dtype=float32), 'validation/loss': Array(1.0714159, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6251, dtype=float32), 'test/loss': Array(1.8615223, dtype=float32), 'test/num_examples': 10000, 'score': 45960.7723968029, 'total_duration': 46516.528638362885, 'accumulated_submission_time': 45960.7723968029, 'accumulated_eval_time': 552.6016957759857, 'accumulated_logging_time': 1.8376588821411133} +I0831 21:41:46.302539 139540719195904 logging_writer.py:48] [167972] accumulated_eval_time=552.602, accumulated_logging_time=1.83766, accumulated_submission_time=45960.8, global_step=167972, preemption_count=0, score=45960.8, test/accuracy=0.6251000165939331, test/loss=1.8615223169326782, test/num_examples=10000, total_duration=46516.5, train/accuracy=0.9487404227256775, train/loss=0.18130454421043396, validation/accuracy=0.7528600096702576, validation/loss=1.071415901184082, validation/num_examples=50000 +I0831 21:41:54.253643 139540727588608 logging_writer.py:48] [168000] global_step=168000, grad_norm=4.9667205810546875, loss=0.9056800603866577 +I0831 21:42:21.321823 139540719195904 logging_writer.py:48] [168100] global_step=168100, grad_norm=4.865715026855469, loss=0.9361110925674438 +I0831 21:42:48.368130 139540727588608 logging_writer.py:48] [168200] global_step=168200, grad_norm=5.346195697784424, loss=1.05850350856781 +I0831 21:43:15.512323 139540719195904 logging_writer.py:48] [168300] global_step=168300, grad_norm=5.15791130065918, loss=0.9996544122695923 +I0831 21:43:42.586421 139540727588608 logging_writer.py:48] [168400] global_step=168400, grad_norm=4.634342193603516, loss=0.8739444613456726 +I0831 21:44:10.676086 139540719195904 logging_writer.py:48] [168500] global_step=168500, grad_norm=5.0527119636535645, loss=0.9680424928665161 +I0831 21:44:56.845057 139540727588608 logging_writer.py:48] [168600] global_step=168600, grad_norm=5.020331382751465, loss=0.8990182280540466 +I0831 21:45:41.645688 139540719195904 logging_writer.py:48] [168700] global_step=168700, grad_norm=4.813370704650879, loss=0.895018458366394 +I0831 21:46:28.153590 139540727588608 logging_writer.py:48] [168800] global_step=168800, grad_norm=4.833280563354492, loss=0.9462729692459106 +I0831 21:47:14.936361 139540719195904 logging_writer.py:48] [168900] global_step=168900, grad_norm=5.263225555419922, loss=1.031261682510376 +I0831 21:47:59.934350 139540727588608 logging_writer.py:48] [169000] global_step=169000, grad_norm=4.633800983428955, loss=0.8590571284294128 +I0831 21:48:44.317436 139540719195904 logging_writer.py:48] [169100] global_step=169100, grad_norm=4.557015419006348, loss=0.8894205689430237 +I0831 21:49:27.558932 139540727588608 logging_writer.py:48] [169200] global_step=169200, grad_norm=5.018107891082764, loss=0.8986049890518188 +I0831 21:50:12.379090 139540719195904 logging_writer.py:48] [169300] global_step=169300, grad_norm=4.9812726974487305, loss=0.969394326210022 +I0831 21:50:59.510345 139540727588608 logging_writer.py:48] [169400] global_step=169400, grad_norm=5.124046325683594, loss=0.8985844850540161 +I0831 21:51:46.350718 139540719195904 logging_writer.py:48] [169500] global_step=169500, grad_norm=4.803107738494873, loss=0.9926440119743347 +I0831 21:52:33.924570 139540727588608 logging_writer.py:48] [169600] global_step=169600, grad_norm=5.2980217933654785, loss=0.9705309271812439 +I0831 21:53:20.843827 139540719195904 logging_writer.py:48] [169700] global_step=169700, grad_norm=5.074955463409424, loss=0.9472571611404419 +I0831 21:54:06.002731 139540727588608 logging_writer.py:48] [169800] global_step=169800, grad_norm=4.864984512329102, loss=0.9267332553863525 +I0831 21:54:49.419153 139540719195904 logging_writer.py:48] [169900] global_step=169900, grad_norm=4.832164764404297, loss=1.0206891298294067 +I0831 21:55:35.359116 139540727588608 logging_writer.py:48] [170000] global_step=170000, grad_norm=4.950588703155518, loss=0.9472445249557495 +I0831 21:56:20.729043 139540719195904 logging_writer.py:48] [170100] global_step=170100, grad_norm=4.86026668548584, loss=0.9538179636001587 +I0831 21:57:06.518306 139540727588608 logging_writer.py:48] [170200] global_step=170200, grad_norm=4.9346418380737305, loss=0.9659864902496338 +I0831 21:57:52.314643 139540719195904 logging_writer.py:48] [170300] global_step=170300, grad_norm=4.84830904006958, loss=0.9136912226676941 +I0831 21:58:39.017995 139540727588608 logging_writer.py:48] [170400] global_step=170400, grad_norm=4.830488681793213, loss=0.9700785279273987 +I0831 21:59:24.591482 139540719195904 logging_writer.py:48] [170500] global_step=170500, grad_norm=5.040051460266113, loss=0.9494431018829346 +I0831 22:00:08.023066 139540727588608 logging_writer.py:48] [170600] global_step=170600, grad_norm=4.852638244628906, loss=0.9527171850204468 +I0831 22:00:51.541023 139540719195904 logging_writer.py:48] [170700] global_step=170700, grad_norm=4.900259017944336, loss=0.9513531923294067 +I0831 22:01:33.187415 139540727588608 logging_writer.py:48] [170800] global_step=170800, grad_norm=5.142589092254639, loss=0.9712234735488892 +I0831 22:02:15.164900 139540719195904 logging_writer.py:48] [170900] global_step=170900, grad_norm=5.310712814331055, loss=0.9729113578796387 +I0831 22:02:58.022552 139540727588608 logging_writer.py:48] [171000] global_step=171000, grad_norm=5.082427978515625, loss=0.9081253409385681 +I0831 22:03:41.093884 139540719195904 logging_writer.py:48] [171100] global_step=171100, grad_norm=4.916240215301514, loss=0.9392939805984497 +I0831 22:04:24.265120 139540727588608 logging_writer.py:48] [171200] global_step=171200, grad_norm=4.984756946563721, loss=0.8829730153083801 +I0831 22:05:08.579635 139540719195904 logging_writer.py:48] [171300] global_step=171300, grad_norm=5.198332786560059, loss=0.9520210027694702 +I0831 22:05:54.140745 139540727588608 logging_writer.py:48] [171400] global_step=171400, grad_norm=4.943979263305664, loss=0.9582770466804504 +I0831 22:06:40.127190 139540719195904 logging_writer.py:48] [171500] global_step=171500, grad_norm=5.0407185554504395, loss=0.903640866279602 +I0831 22:07:25.626827 139540727588608 logging_writer.py:48] [171600] global_step=171600, grad_norm=4.895448684692383, loss=0.8646007776260376 +I0831 22:08:08.425002 139540719195904 logging_writer.py:48] [171700] global_step=171700, grad_norm=4.983148574829102, loss=0.9848845601081848 +I0831 22:08:51.615622 139540727588608 logging_writer.py:48] [171800] global_step=171800, grad_norm=4.924423694610596, loss=0.915143609046936 +I0831 22:09:36.255075 139540719195904 logging_writer.py:48] [171900] global_step=171900, grad_norm=5.048596382141113, loss=0.8485208749771118 +I0831 22:10:19.654952 139540727588608 logging_writer.py:48] [172000] global_step=172000, grad_norm=5.562159538269043, loss=0.9896900653839111 +I0831 22:11:02.590715 139540719195904 logging_writer.py:48] [172100] global_step=172100, grad_norm=5.010772705078125, loss=0.9415399432182312 +I0831 22:11:46.228965 139540727588608 logging_writer.py:48] [172200] global_step=172200, grad_norm=4.956217288970947, loss=0.8833256959915161 +I0831 22:12:29.287091 139540719195904 logging_writer.py:48] [172300] global_step=172300, grad_norm=5.030797958374023, loss=0.9704371690750122 +I0831 22:13:11.383102 139540727588608 logging_writer.py:48] [172400] global_step=172400, grad_norm=5.118746757507324, loss=0.9697637557983398 +I0831 22:13:54.187218 139540719195904 logging_writer.py:48] [172500] global_step=172500, grad_norm=5.206345558166504, loss=0.9677541255950928 +I0831 22:14:36.710945 139540727588608 logging_writer.py:48] [172600] global_step=172600, grad_norm=5.237006664276123, loss=0.9463266134262085 +I0831 22:15:02.426499 139757377230016 spec.py:333] Evaluating on the training split. +I0831 22:15:11.627892 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 22:15:23.803988 139757377230016 spec.py:363] Evaluating on the test split. +I0831 22:15:24.843034 139757377230016 submission_runner.py:516] Time since start: 48535.03s, Step: 172658, {'train/accuracy': Array(0.94868064, dtype=float32), 'train/loss': Array(0.17847335, dtype=float32), 'validation/accuracy': Array(0.75325996, dtype=float32), 'validation/loss': Array(1.0750405, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.625, dtype=float32), 'test/loss': Array(1.8682417, dtype=float32), 'test/num_examples': 10000, 'score': 47956.8501329422, 'total_duration': 48535.025040864944, 'accumulated_submission_time': 47956.8501329422, 'accumulated_eval_time': 574.862462759018, 'accumulated_logging_time': 1.9537787437438965} +I0831 22:15:25.239940 139540719195904 logging_writer.py:48] [172658] accumulated_eval_time=574.862, accumulated_logging_time=1.95378, accumulated_submission_time=47956.9, global_step=172658, preemption_count=0, score=47956.9, test/accuracy=0.625, test/loss=1.8682416677474976, test/num_examples=10000, total_duration=48535, train/accuracy=0.9486806392669678, train/loss=0.1784733533859253, validation/accuracy=0.7532599568367004, validation/loss=1.0750404596328735, validation/num_examples=50000 +I0831 22:15:39.207039 139540727588608 logging_writer.py:48] [172700] global_step=172700, grad_norm=4.9408440589904785, loss=0.8701879382133484 +I0831 22:16:23.420623 139540719195904 logging_writer.py:48] [172800] global_step=172800, grad_norm=5.219758033752441, loss=0.9864237904548645 +I0831 22:17:07.385926 139540727588608 logging_writer.py:48] [172900] global_step=172900, grad_norm=5.172638416290283, loss=0.9566007256507874 +I0831 22:17:50.324565 139540719195904 logging_writer.py:48] [173000] global_step=173000, grad_norm=4.9370574951171875, loss=0.9513857960700989 +I0831 22:18:34.628737 139540727588608 logging_writer.py:48] [173100] global_step=173100, grad_norm=5.188773155212402, loss=1.0096964836120605 +I0831 22:19:19.675599 139540719195904 logging_writer.py:48] [173200] global_step=173200, grad_norm=5.501114368438721, loss=0.9356459379196167 +I0831 22:20:01.595043 139540727588608 logging_writer.py:48] [173300] global_step=173300, grad_norm=5.11478853225708, loss=1.0001646280288696 +I0831 22:20:43.691665 139540719195904 logging_writer.py:48] [173400] global_step=173400, grad_norm=4.910259246826172, loss=0.9074640274047852 +I0831 22:21:27.294452 139540727588608 logging_writer.py:48] [173500] global_step=173500, grad_norm=5.332563400268555, loss=1.0502313375473022 +I0831 22:22:11.368495 139540719195904 logging_writer.py:48] [173600] global_step=173600, grad_norm=4.83388614654541, loss=0.918712854385376 +I0831 22:22:55.879478 139540727588608 logging_writer.py:48] [173700] global_step=173700, grad_norm=5.046889781951904, loss=0.9854961633682251 +I0831 22:23:39.244450 139540719195904 logging_writer.py:48] [173800] global_step=173800, grad_norm=4.963953018188477, loss=0.9539850354194641 +I0831 22:24:22.017191 139540727588608 logging_writer.py:48] [173900] global_step=173900, grad_norm=5.049905300140381, loss=0.8951940536499023 +I0831 22:25:05.553579 139540719195904 logging_writer.py:48] [174000] global_step=174000, grad_norm=5.104269027709961, loss=0.9532504677772522 +I0831 22:25:49.479288 139540727588608 logging_writer.py:48] [174100] global_step=174100, grad_norm=5.2159504890441895, loss=0.9204903244972229 +I0831 22:26:33.289925 139540719195904 logging_writer.py:48] [174200] global_step=174200, grad_norm=5.189445495605469, loss=0.956203281879425 +I0831 22:27:17.985832 139540727588608 logging_writer.py:48] [174300] global_step=174300, grad_norm=5.115742206573486, loss=0.996508777141571 +I0831 22:28:02.399137 139540719195904 logging_writer.py:48] [174400] global_step=174400, grad_norm=4.756511688232422, loss=0.9150068759918213 +I0831 22:28:44.536584 139540727588608 logging_writer.py:48] [174500] global_step=174500, grad_norm=5.12375020980835, loss=0.9469826817512512 +I0831 22:29:26.532136 139540719195904 logging_writer.py:48] [174600] global_step=174600, grad_norm=4.859577655792236, loss=0.8403810858726501 +I0831 22:30:09.335955 139540727588608 logging_writer.py:48] [174700] global_step=174700, grad_norm=5.153999328613281, loss=0.9412992596626282 +I0831 22:30:51.863125 139540719195904 logging_writer.py:48] [174800] global_step=174800, grad_norm=5.2303466796875, loss=0.9529769420623779 +I0831 22:31:35.723417 139540727588608 logging_writer.py:48] [174900] global_step=174900, grad_norm=4.8271403312683105, loss=0.8860395550727844 +I0831 22:32:19.169567 139540719195904 logging_writer.py:48] [175000] global_step=175000, grad_norm=4.649547100067139, loss=0.877926766872406 +I0831 22:33:03.866589 139540727588608 logging_writer.py:48] [175100] global_step=175100, grad_norm=4.99348258972168, loss=0.9112836718559265 +I0831 22:33:48.428804 139540719195904 logging_writer.py:48] [175200] global_step=175200, grad_norm=4.849855422973633, loss=0.860894501209259 +I0831 22:34:32.848647 139540727588608 logging_writer.py:48] [175300] global_step=175300, grad_norm=5.3136796951293945, loss=0.9717639684677124 +I0831 22:35:17.172759 139540719195904 logging_writer.py:48] [175400] global_step=175400, grad_norm=4.770131587982178, loss=0.8488084673881531 +I0831 22:36:00.938620 139540727588608 logging_writer.py:48] [175500] global_step=175500, grad_norm=5.07946252822876, loss=0.9661241769790649 +I0831 22:36:44.013165 139540719195904 logging_writer.py:48] [175600] global_step=175600, grad_norm=4.938067436218262, loss=0.8713827729225159 +I0831 22:37:26.827827 139540727588608 logging_writer.py:48] [175700] global_step=175700, grad_norm=5.037351608276367, loss=0.89052814245224 +I0831 22:38:15.321449 139540719195904 logging_writer.py:48] [175800] global_step=175800, grad_norm=5.253012180328369, loss=0.9890425801277161 +I0831 22:39:02.658846 139540727588608 logging_writer.py:48] [175900] global_step=175900, grad_norm=5.051004886627197, loss=0.8687795400619507 +I0831 22:39:49.262048 139540719195904 logging_writer.py:48] [176000] global_step=176000, grad_norm=4.952812194824219, loss=0.9004449844360352 +I0831 22:40:37.024426 139540727588608 logging_writer.py:48] [176100] global_step=176100, grad_norm=4.972234725952148, loss=0.922723650932312 +I0831 22:41:29.449959 139540719195904 logging_writer.py:48] [176200] global_step=176200, grad_norm=4.981405258178711, loss=0.9230483770370483 +I0831 22:42:22.373697 139540727588608 logging_writer.py:48] [176300] global_step=176300, grad_norm=5.3722734451293945, loss=0.9524778127670288 +I0831 22:43:14.325342 139540719195904 logging_writer.py:48] [176400] global_step=176400, grad_norm=5.10913610458374, loss=0.9017881155014038 +I0831 22:44:05.960836 139540727588608 logging_writer.py:48] [176500] global_step=176500, grad_norm=4.788115978240967, loss=0.8571447730064392 +I0831 22:44:57.280666 139540719195904 logging_writer.py:48] [176600] global_step=176600, grad_norm=5.1439433097839355, loss=0.9455742835998535 +I0831 22:45:47.660436 139540727588608 logging_writer.py:48] [176700] global_step=176700, grad_norm=5.3004302978515625, loss=0.9719034433364868 +I0831 22:46:35.206265 139540719195904 logging_writer.py:48] [176800] global_step=176800, grad_norm=4.894039630889893, loss=0.8853162527084351 +I0831 22:47:24.656785 139540727588608 logging_writer.py:48] [176900] global_step=176900, grad_norm=5.391083240509033, loss=0.9754825830459595 +I0831 22:48:11.102854 139540719195904 logging_writer.py:48] [177000] global_step=177000, grad_norm=4.900680065155029, loss=0.9589104652404785 +I0831 22:48:40.942126 139757377230016 spec.py:333] Evaluating on the training split. +I0831 22:48:49.969277 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 22:49:02.287639 139757377230016 spec.py:363] Evaluating on the test split. +I0831 22:49:03.344027 139757377230016 submission_runner.py:516] Time since start: 50553.52s, Step: 177063, {'train/accuracy': Array(0.9486607, dtype=float32), 'train/loss': Array(0.17960994, dtype=float32), 'validation/accuracy': Array(0.7531, dtype=float32), 'validation/loss': Array(1.074318, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62600005, dtype=float32), 'test/loss': Array(1.8686966, dtype=float32), 'test/num_examples': 10000, 'score': 49952.47198367119, 'total_duration': 50553.51904273033, 'accumulated_submission_time': 49952.47198367119, 'accumulated_eval_time': 597.1016075611115, 'accumulated_logging_time': 2.3890323638916016} +I0831 22:49:03.744482 139540727588608 logging_writer.py:48] [177063] accumulated_eval_time=597.102, accumulated_logging_time=2.38903, accumulated_submission_time=49952.5, global_step=177063, preemption_count=0, score=49952.5, test/accuracy=0.6260000467300415, test/loss=1.8686965703964233, test/num_examples=10000, total_duration=50553.5, train/accuracy=0.948660671710968, train/loss=0.17960993945598602, validation/accuracy=0.7530999779701233, validation/loss=1.0743180513381958, validation/num_examples=50000 +I0831 22:49:15.552106 139540719195904 logging_writer.py:48] [177100] global_step=177100, grad_norm=5.0831379890441895, loss=0.8729996681213379 +I0831 22:50:04.003920 139540727588608 logging_writer.py:48] [177200] global_step=177200, grad_norm=5.064516544342041, loss=0.9893490672111511 +I0831 22:50:50.799000 139540719195904 logging_writer.py:48] [177300] global_step=177300, grad_norm=5.158038139343262, loss=0.9798438549041748 +I0831 22:51:36.344662 139540727588608 logging_writer.py:48] [177400] global_step=177400, grad_norm=4.817814826965332, loss=0.8882734775543213 +I0831 22:52:23.379956 139540719195904 logging_writer.py:48] [177500] global_step=177500, grad_norm=5.0585246086120605, loss=0.9602863788604736 +I0831 22:53:08.817725 139540727588608 logging_writer.py:48] [177600] global_step=177600, grad_norm=5.078489303588867, loss=1.0060944557189941 +I0831 22:53:55.515661 139540719195904 logging_writer.py:48] [177700] global_step=177700, grad_norm=5.125731945037842, loss=0.9528493285179138 +I0831 22:54:41.524550 139540727588608 logging_writer.py:48] [177800] global_step=177800, grad_norm=5.323104381561279, loss=0.9729431867599487 +I0831 22:55:26.592607 139540719195904 logging_writer.py:48] [177900] global_step=177900, grad_norm=5.591817855834961, loss=1.1219244003295898 +I0831 22:56:09.626986 139540727588608 logging_writer.py:48] [178000] global_step=178000, grad_norm=5.024587631225586, loss=0.947258710861206 +I0831 22:56:53.604487 139540719195904 logging_writer.py:48] [178100] global_step=178100, grad_norm=5.032771587371826, loss=0.926495373249054 +I0831 22:57:37.575162 139540727588608 logging_writer.py:48] [178200] global_step=178200, grad_norm=4.8738508224487305, loss=0.8688822388648987 +I0831 22:58:19.548547 139540719195904 logging_writer.py:48] [178300] global_step=178300, grad_norm=5.066472053527832, loss=0.9636494517326355 +I0831 22:59:01.154452 139540727588608 logging_writer.py:48] [178400] global_step=178400, grad_norm=4.828902244567871, loss=0.8653801679611206 +I0831 22:59:43.291198 139540719195904 logging_writer.py:48] [178500] global_step=178500, grad_norm=5.3049726486206055, loss=0.9175033569335938 +I0831 23:00:26.946721 139540727588608 logging_writer.py:48] [178600] global_step=178600, grad_norm=5.098486423492432, loss=0.9334414601325989 +I0831 23:01:10.268186 139540719195904 logging_writer.py:48] [178700] global_step=178700, grad_norm=5.105940818786621, loss=0.9509732723236084 +I0831 23:01:52.325585 139540727588608 logging_writer.py:48] [178800] global_step=178800, grad_norm=4.982105255126953, loss=0.9802786111831665 +I0831 23:02:36.266330 139540719195904 logging_writer.py:48] [178900] global_step=178900, grad_norm=5.475265979766846, loss=1.0458036661148071 +I0831 23:03:19.389785 139540727588608 logging_writer.py:48] [179000] global_step=179000, grad_norm=5.247165203094482, loss=0.9071815013885498 +I0831 23:04:04.285653 139540719195904 logging_writer.py:48] [179100] global_step=179100, grad_norm=5.049513339996338, loss=0.9646730422973633 +I0831 23:04:45.924298 139540727588608 logging_writer.py:48] [179200] global_step=179200, grad_norm=4.897809982299805, loss=0.8885094523429871 +I0831 23:05:30.429682 139540719195904 logging_writer.py:48] [179300] global_step=179300, grad_norm=4.870821475982666, loss=0.9540774822235107 +I0831 23:06:15.405400 139540727588608 logging_writer.py:48] [179400] global_step=179400, grad_norm=5.36308479309082, loss=1.0525953769683838 +I0831 23:06:59.745378 139540719195904 logging_writer.py:48] [179500] global_step=179500, grad_norm=4.866727828979492, loss=0.8450534343719482 +I0831 23:07:43.892472 139540727588608 logging_writer.py:48] [179600] global_step=179600, grad_norm=5.078181266784668, loss=0.9708400368690491 +I0831 23:08:28.085355 139540719195904 logging_writer.py:48] [179700] global_step=179700, grad_norm=5.070951461791992, loss=0.8997238874435425 +I0831 23:09:10.949957 139540727588608 logging_writer.py:48] [179800] global_step=179800, grad_norm=5.224216461181641, loss=0.98350590467453 +I0831 23:09:52.120793 139540719195904 logging_writer.py:48] [179900] global_step=179900, grad_norm=4.855395793914795, loss=0.9756488800048828 +I0831 23:10:35.185401 139540727588608 logging_writer.py:48] [180000] global_step=180000, grad_norm=4.952844142913818, loss=0.8926187753677368 +I0831 23:11:17.184155 139540719195904 logging_writer.py:48] [180100] global_step=180100, grad_norm=5.116673946380615, loss=0.8991693258285522 +I0831 23:12:02.787490 139540727588608 logging_writer.py:48] [180200] global_step=180200, grad_norm=5.096442699432373, loss=0.9224395751953125 +I0831 23:12:46.362941 139540719195904 logging_writer.py:48] [180300] global_step=180300, grad_norm=4.917977809906006, loss=0.920380175113678 +I0831 23:13:30.046230 139540727588608 logging_writer.py:48] [180400] global_step=180400, grad_norm=5.192338943481445, loss=0.945069432258606 +I0831 23:14:13.167749 139540719195904 logging_writer.py:48] [180500] global_step=180500, grad_norm=4.633719444274902, loss=0.8726009130477905 +I0831 23:14:57.148041 139540727588608 logging_writer.py:48] [180600] global_step=180600, grad_norm=5.154436111450195, loss=0.9764580726623535 +I0831 23:15:41.952921 139540719195904 logging_writer.py:48] [180700] global_step=180700, grad_norm=5.0797834396362305, loss=0.8806819915771484 +I0831 23:16:26.124856 139540727588608 logging_writer.py:48] [180800] global_step=180800, grad_norm=5.229189872741699, loss=0.9755882620811462 +I0831 23:17:09.560934 139540719195904 logging_writer.py:48] [180900] global_step=180900, grad_norm=4.8038434982299805, loss=0.8953627943992615 +I0831 23:17:53.677773 139540727588608 logging_writer.py:48] [181000] global_step=181000, grad_norm=4.83842134475708, loss=0.8455545902252197 +I0831 23:18:37.012009 139540719195904 logging_writer.py:48] [181100] global_step=181100, grad_norm=5.175807952880859, loss=0.9235392808914185 +I0831 23:19:20.869152 139540727588608 logging_writer.py:48] [181200] global_step=181200, grad_norm=4.86659574508667, loss=0.8268725872039795 +I0831 23:20:03.684798 139540719195904 logging_writer.py:48] [181300] global_step=181300, grad_norm=4.944614887237549, loss=0.9143164157867432 +I0831 23:20:49.242789 139540727588608 logging_writer.py:48] [181400] global_step=181400, grad_norm=5.105677604675293, loss=0.9457405805587769 +I0831 23:21:34.934571 139540719195904 logging_writer.py:48] [181500] global_step=181500, grad_norm=4.892510414123535, loss=0.9019402265548706 +I0831 23:22:19.241430 139757377230016 spec.py:333] Evaluating on the training split. +I0831 23:22:28.328105 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 23:22:58.092314 139757377230016 spec.py:363] Evaluating on the test split. +I0831 23:22:59.179601 139757377230016 submission_runner.py:516] Time since start: 52589.32s, Step: 181600, {'train/accuracy': Array(0.9502152, dtype=float32), 'train/loss': Array(0.17258252, dtype=float32), 'validation/accuracy': Array(0.75378, dtype=float32), 'validation/loss': Array(1.0762789, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62560004, dtype=float32), 'test/loss': Array(1.8692299, dtype=float32), 'test/num_examples': 10000, 'score': 51947.90822029114, 'total_duration': 52589.320518255234, 'accumulated_submission_time': 51947.90822029114, 'accumulated_eval_time': 636.8429207801819, 'accumulated_logging_time': 2.812356948852539} +I0831 23:22:59.650064 139540727588608 logging_writer.py:48] [181600] accumulated_eval_time=636.843, accumulated_logging_time=2.81236, accumulated_submission_time=51947.9, global_step=181600, preemption_count=0, score=51947.9, test/accuracy=0.6256000399589539, test/loss=1.8692299127578735, test/num_examples=10000, total_duration=52589.3, train/accuracy=0.950215220451355, train/loss=0.17258252203464508, validation/accuracy=0.7537800073623657, validation/loss=1.0762789249420166, validation/num_examples=50000 +I0831 23:22:59.961926 139540719195904 logging_writer.py:48] [181600] global_step=181600, grad_norm=5.3622846603393555, loss=0.9922167658805847 +I0831 23:23:41.043959 139540727588608 logging_writer.py:48] [181700] global_step=181700, grad_norm=4.872025966644287, loss=0.9277018308639526 +I0831 23:24:28.741724 139540719195904 logging_writer.py:48] [181800] global_step=181800, grad_norm=5.053481101989746, loss=0.8754605054855347 +I0831 23:25:17.453391 139540727588608 logging_writer.py:48] [181900] global_step=181900, grad_norm=5.119720935821533, loss=0.9404197335243225 +I0831 23:26:05.551848 139540719195904 logging_writer.py:48] [182000] global_step=182000, grad_norm=4.961014270782471, loss=0.9360740780830383 +I0831 23:26:54.531404 139540727588608 logging_writer.py:48] [182100] global_step=182100, grad_norm=5.327876091003418, loss=1.0008430480957031 +I0831 23:27:43.267859 139540719195904 logging_writer.py:48] [182200] global_step=182200, grad_norm=4.836642742156982, loss=0.8613523840904236 +I0831 23:28:29.235894 139540727588608 logging_writer.py:48] [182300] global_step=182300, grad_norm=4.949620246887207, loss=0.9395437240600586 +I0831 23:29:15.248088 139540719195904 logging_writer.py:48] [182400] global_step=182400, grad_norm=5.1615824699401855, loss=0.9757777452468872 +I0831 23:30:03.448201 139540727588608 logging_writer.py:48] [182500] global_step=182500, grad_norm=4.949434757232666, loss=0.9301069974899292 +I0831 23:30:49.678436 139540719195904 logging_writer.py:48] [182600] global_step=182600, grad_norm=5.127139091491699, loss=0.945297122001648 +I0831 23:31:36.585882 139540727588608 logging_writer.py:48] [182700] global_step=182700, grad_norm=4.730690956115723, loss=0.8639365434646606 +I0831 23:32:23.365235 139540719195904 logging_writer.py:48] [182800] global_step=182800, grad_norm=5.299928188323975, loss=1.0132122039794922 +I0831 23:33:08.729767 139540727588608 logging_writer.py:48] [182900] global_step=182900, grad_norm=5.237121105194092, loss=0.9050325155258179 +I0831 23:33:52.941699 139540719195904 logging_writer.py:48] [183000] global_step=183000, grad_norm=5.451279163360596, loss=0.9576146602630615 +I0831 23:34:36.415539 139540727588608 logging_writer.py:48] [183100] global_step=183100, grad_norm=4.809786319732666, loss=0.8077584505081177 +I0831 23:35:21.548435 139540719195904 logging_writer.py:48] [183200] global_step=183200, grad_norm=4.993740558624268, loss=0.9438604712486267 +I0831 23:36:06.331710 139540727588608 logging_writer.py:48] [183300] global_step=183300, grad_norm=5.27506685256958, loss=0.9869638681411743 +I0831 23:36:49.812130 139540719195904 logging_writer.py:48] [183400] global_step=183400, grad_norm=5.384955406188965, loss=1.045977234840393 +I0831 23:37:34.656703 139540727588608 logging_writer.py:48] [183500] global_step=183500, grad_norm=5.10580587387085, loss=0.9955980777740479 +I0831 23:38:19.319051 139540719195904 logging_writer.py:48] [183600] global_step=183600, grad_norm=5.012576103210449, loss=0.9461967945098877 +I0831 23:39:03.887245 139540727588608 logging_writer.py:48] [183700] global_step=183700, grad_norm=5.064459800720215, loss=0.8691977262496948 +I0831 23:39:48.328100 139540719195904 logging_writer.py:48] [183800] global_step=183800, grad_norm=5.154045581817627, loss=0.9867819547653198 +I0831 23:40:30.876438 139540727588608 logging_writer.py:48] [183900] global_step=183900, grad_norm=5.3183159828186035, loss=1.0325847864151 +I0831 23:41:18.062930 139540719195904 logging_writer.py:48] [184000] global_step=184000, grad_norm=5.2383527755737305, loss=0.960383415222168 +I0831 23:42:03.118319 139540727588608 logging_writer.py:48] [184100] global_step=184100, grad_norm=5.278375625610352, loss=0.9775615930557251 +I0831 23:42:44.710427 139540719195904 logging_writer.py:48] [184200] global_step=184200, grad_norm=5.267536640167236, loss=1.0023176670074463 +I0831 23:43:26.273338 139540727588608 logging_writer.py:48] [184300] global_step=184300, grad_norm=5.462645530700684, loss=0.9463536739349365 +I0831 23:44:09.094367 139540719195904 logging_writer.py:48] [184400] global_step=184400, grad_norm=5.309229850769043, loss=0.9429004192352295 +I0831 23:44:51.181596 139540727588608 logging_writer.py:48] [184500] global_step=184500, grad_norm=4.915128231048584, loss=0.9007394909858704 +I0831 23:45:34.542451 139540719195904 logging_writer.py:48] [184600] global_step=184600, grad_norm=4.882896423339844, loss=0.865105152130127 +I0831 23:46:17.528820 139540727588608 logging_writer.py:48] [184700] global_step=184700, grad_norm=4.902183532714844, loss=0.8582038879394531 +I0831 23:47:01.344441 139540719195904 logging_writer.py:48] [184800] global_step=184800, grad_norm=5.1856465339660645, loss=0.9449483752250671 +I0831 23:47:44.238795 139540727588608 logging_writer.py:48] [184900] global_step=184900, grad_norm=4.925085544586182, loss=0.9033776521682739 +I0831 23:48:28.912796 139540719195904 logging_writer.py:48] [185000] global_step=185000, grad_norm=4.9524736404418945, loss=0.8486440181732178 +I0831 23:49:11.468621 139540727588608 logging_writer.py:48] [185100] global_step=185100, grad_norm=5.449380397796631, loss=1.021689534187317 +I0831 23:49:55.227444 139540719195904 logging_writer.py:48] [185200] global_step=185200, grad_norm=5.133976459503174, loss=0.9937183260917664 +I0831 23:50:39.787231 139540727588608 logging_writer.py:48] [185300] global_step=185300, grad_norm=4.688454627990723, loss=0.8530359268188477 +I0831 23:51:22.716017 139540719195904 logging_writer.py:48] [185400] global_step=185400, grad_norm=5.026909828186035, loss=0.8916329145431519 +I0831 23:52:04.859393 139540727588608 logging_writer.py:48] [185500] global_step=185500, grad_norm=4.869600296020508, loss=0.8848799467086792 +I0831 23:52:48.451534 139540719195904 logging_writer.py:48] [185600] global_step=185600, grad_norm=5.621570587158203, loss=0.9721260070800781 +I0831 23:53:32.050875 139540727588608 logging_writer.py:48] [185700] global_step=185700, grad_norm=5.180081367492676, loss=0.8848180770874023 +I0831 23:54:15.562083 139540719195904 logging_writer.py:48] [185800] global_step=185800, grad_norm=5.286177158355713, loss=0.961733341217041 +I0831 23:55:00.546586 139540727588608 logging_writer.py:48] [185900] global_step=185900, grad_norm=5.363322734832764, loss=0.994147539138794 +I0831 23:55:44.397250 139540719195904 logging_writer.py:48] [186000] global_step=186000, grad_norm=4.957035064697266, loss=0.8803666830062866 +I0831 23:56:15.247966 139757377230016 spec.py:333] Evaluating on the training split. +I0831 23:56:24.128071 139757377230016 spec.py:346] Evaluating on the validation split. +I0831 23:56:56.036513 139757377230016 spec.py:363] Evaluating on the test split. +I0831 23:56:57.067949 139757377230016 submission_runner.py:516] Time since start: 54627.27s, Step: 186073, {'train/accuracy': Array(0.9526666, dtype=float32), 'train/loss': Array(0.16909547, dtype=float32), 'validation/accuracy': Array(0.7543, dtype=float32), 'validation/loss': Array(1.0771687, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62570006, dtype=float32), 'test/loss': Array(1.869183, dtype=float32), 'test/num_examples': 10000, 'score': 53943.446397304535, 'total_duration': 54627.26826930046, 'accumulated_submission_time': 53943.446397304535, 'accumulated_eval_time': 678.5254654884338, 'accumulated_logging_time': 3.305555582046509} +I0831 23:56:57.547911 139540727588608 logging_writer.py:48] [186073] accumulated_eval_time=678.525, accumulated_logging_time=3.30556, accumulated_submission_time=53943.4, global_step=186073, preemption_count=0, score=53943.4, test/accuracy=0.625700056552887, test/loss=1.8691829442977905, test/num_examples=10000, total_duration=54627.3, train/accuracy=0.9526665806770325, train/loss=0.1690954715013504, validation/accuracy=0.7542999982833862, validation/loss=1.0771687030792236, validation/num_examples=50000 +I0831 23:57:05.184700 139540719195904 logging_writer.py:48] [186100] global_step=186100, grad_norm=4.806110858917236, loss=0.9289065599441528 +I0831 23:57:48.291831 139540727588608 logging_writer.py:48] [186200] global_step=186200, grad_norm=5.501133441925049, loss=0.9749774932861328 +I0831 23:58:33.189990 139540719195904 logging_writer.py:48] [186300] global_step=186300, grad_norm=4.9684977531433105, loss=0.9467901587486267 +I0831 23:59:17.345969 139540727588608 logging_writer.py:48] [186400] global_step=186400, grad_norm=5.216026782989502, loss=0.8990527987480164 +I0901 00:00:02.040801 139540719195904 logging_writer.py:48] [186500] global_step=186500, grad_norm=5.426215171813965, loss=1.0138753652572632 +I0901 00:00:46.720521 139540727588608 logging_writer.py:48] [186600] global_step=186600, grad_norm=5.139348030090332, loss=0.9900546073913574 +I0901 00:01:29.358174 139540719195904 logging_writer.py:48] [186700] global_step=186700, grad_norm=4.891494274139404, loss=0.9578847289085388 +I0901 00:02:12.236941 139540727588608 logging_writer.py:48] [186800] global_step=186800, grad_norm=4.9220733642578125, loss=0.8251274228096008 +I0901 00:02:56.707437 139540719195904 logging_writer.py:48] [186900] global_step=186900, grad_norm=5.279429912567139, loss=0.9530755281448364 +I0901 00:03:40.131458 139540727588608 logging_writer.py:48] [187000] global_step=187000, grad_norm=4.818136692047119, loss=0.8895902633666992 +I0901 00:04:23.860688 139540719195904 logging_writer.py:48] [187100] global_step=187100, grad_norm=5.097135543823242, loss=0.9391963481903076 +I0901 00:05:09.072693 139540727588608 logging_writer.py:48] [187200] global_step=187200, grad_norm=5.17567253112793, loss=0.9920602440834045 +I0901 00:05:50.844568 139540719195904 logging_writer.py:48] [187300] global_step=187300, grad_norm=4.809934139251709, loss=0.9135873317718506 +I0901 00:06:33.995251 139540727588608 logging_writer.py:48] [187400] global_step=187400, grad_norm=5.164144992828369, loss=0.9258674383163452 +I0901 00:07:18.489343 139540719195904 logging_writer.py:48] [187500] global_step=187500, grad_norm=5.077635765075684, loss=0.9095540642738342 +I0901 00:08:02.640619 139540727588608 logging_writer.py:48] [187600] global_step=187600, grad_norm=5.175909996032715, loss=0.9390788674354553 +I0901 00:08:48.927052 139540719195904 logging_writer.py:48] [187700] global_step=187700, grad_norm=5.173511505126953, loss=1.0058075189590454 +I0901 00:09:33.440837 139540727588608 logging_writer.py:48] [187800] global_step=187800, grad_norm=5.48660135269165, loss=0.9529780745506287 +I0901 00:10:18.843537 139540719195904 logging_writer.py:48] [187900] global_step=187900, grad_norm=5.313019752502441, loss=0.9315965175628662 +I0901 00:11:02.663884 139540727588608 logging_writer.py:48] [188000] global_step=188000, grad_norm=5.065371990203857, loss=0.9610649943351746 +I0901 00:11:45.652235 139540719195904 logging_writer.py:48] [188100] global_step=188100, grad_norm=5.40680456161499, loss=1.0355916023254395 +I0901 00:12:29.517950 139540727588608 logging_writer.py:48] [188200] global_step=188200, grad_norm=4.739466190338135, loss=0.8545449376106262 +I0901 00:13:12.551077 139540719195904 logging_writer.py:48] [188300] global_step=188300, grad_norm=5.367391586303711, loss=0.9507479667663574 +I0901 00:13:53.890249 139540727588608 logging_writer.py:48] [188400] global_step=188400, grad_norm=5.1487932205200195, loss=0.9922887682914734 +I0901 00:14:36.360261 139540719195904 logging_writer.py:48] [188500] global_step=188500, grad_norm=5.186774253845215, loss=1.005993127822876 +I0901 00:15:18.441021 139540727588608 logging_writer.py:48] [188600] global_step=188600, grad_norm=5.488479137420654, loss=1.0157907009124756 +I0901 00:16:01.847439 139540719195904 logging_writer.py:48] [188700] global_step=188700, grad_norm=5.251500606536865, loss=0.917496383190155 +I0901 00:16:44.558152 139540727588608 logging_writer.py:48] [188800] global_step=188800, grad_norm=5.016390800476074, loss=0.8924532532691956 +I0901 00:17:27.823493 139540719195904 logging_writer.py:48] [188900] global_step=188900, grad_norm=5.1955413818359375, loss=0.9385728240013123 +I0901 00:18:11.886467 139540727588608 logging_writer.py:48] [189000] global_step=189000, grad_norm=5.2858099937438965, loss=1.006028175354004 +I0901 00:18:55.100528 139540719195904 logging_writer.py:48] [189100] global_step=189100, grad_norm=5.346923828125, loss=0.957676887512207 +I0901 00:19:38.267696 139540727588608 logging_writer.py:48] [189200] global_step=189200, grad_norm=5.20168924331665, loss=0.9643958210945129 +I0901 00:20:22.104085 139540719195904 logging_writer.py:48] [189300] global_step=189300, grad_norm=5.363740921020508, loss=0.9793390035629272 +I0901 00:21:06.036294 139540727588608 logging_writer.py:48] [189400] global_step=189400, grad_norm=4.697900295257568, loss=0.8371133208274841 +I0901 00:21:49.615208 139540719195904 logging_writer.py:48] [189500] global_step=189500, grad_norm=5.039329528808594, loss=0.9119653701782227 +I0901 00:22:33.186929 139540727588608 logging_writer.py:48] [189600] global_step=189600, grad_norm=5.306957721710205, loss=0.9534136056900024 +I0901 00:23:15.922843 139540719195904 logging_writer.py:48] [189700] global_step=189700, grad_norm=5.466585636138916, loss=0.9276847243309021 +I0901 00:23:59.356027 139540727588608 logging_writer.py:48] [189800] global_step=189800, grad_norm=5.132716655731201, loss=0.9183045625686646 +I0901 00:24:42.017797 139540719195904 logging_writer.py:48] [189900] global_step=189900, grad_norm=5.483950614929199, loss=1.0313572883605957 +I0901 00:25:25.392404 139540727588608 logging_writer.py:48] [190000] global_step=190000, grad_norm=5.038303375244141, loss=0.8895503282546997 +I0901 00:26:09.663610 139540719195904 logging_writer.py:48] [190100] global_step=190100, grad_norm=5.241400718688965, loss=0.9587568044662476 +I0901 00:26:55.835422 139540727588608 logging_writer.py:48] [190200] global_step=190200, grad_norm=5.140751838684082, loss=0.9453198909759521 +I0901 00:27:40.768271 139540719195904 logging_writer.py:48] [190300] global_step=190300, grad_norm=5.034204959869385, loss=0.8800327777862549 +I0901 00:28:24.671232 139540727588608 logging_writer.py:48] [190400] global_step=190400, grad_norm=5.27098274230957, loss=0.9857954978942871 +I0901 00:29:08.447867 139540719195904 logging_writer.py:48] [190500] global_step=190500, grad_norm=5.048198699951172, loss=0.898148775100708 +I0901 00:29:52.244532 139540727588608 logging_writer.py:48] [190600] global_step=190600, grad_norm=4.833571434020996, loss=0.8375513553619385 +I0901 00:30:13.154329 139757377230016 spec.py:333] Evaluating on the training split. +I0901 00:30:22.384076 139757377230016 spec.py:346] Evaluating on the validation split. +I0901 00:30:31.785629 139757377230016 spec.py:363] Evaluating on the test split. +I0901 00:30:32.869454 139757377230016 submission_runner.py:516] Time since start: 56643.01s, Step: 190646, {'train/accuracy': Array(0.9512117, dtype=float32), 'train/loss': Array(0.16966121, dtype=float32), 'validation/accuracy': Array(0.75439996, dtype=float32), 'validation/loss': Array(1.0768428, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.628, dtype=float32), 'test/loss': Array(1.866245, dtype=float32), 'test/num_examples': 10000, 'score': 55938.98239707947, 'total_duration': 56643.01434922218, 'accumulated_submission_time': 55938.98239707947, 'accumulated_eval_time': 698.0477120876312, 'accumulated_logging_time': 3.8167998790740967} +I0901 00:30:33.347449 139540719195904 logging_writer.py:48] [190646] accumulated_eval_time=698.048, accumulated_logging_time=3.8168, accumulated_submission_time=55939, global_step=190646, preemption_count=0, score=55939, test/accuracy=0.628000020980835, test/loss=1.8662450313568115, test/num_examples=10000, total_duration=56643, train/accuracy=0.95121169090271, train/loss=0.16966120898723602, validation/accuracy=0.7543999552726746, validation/loss=1.0768427848815918, validation/num_examples=50000 +I0901 00:30:53.192149 139540727588608 logging_writer.py:48] [190700] global_step=190700, grad_norm=5.127620220184326, loss=0.9293397665023804 +I0901 00:31:37.116081 139540719195904 logging_writer.py:48] [190800] global_step=190800, grad_norm=5.568716526031494, loss=0.9613124132156372 +I0901 00:32:20.834402 139540727588608 logging_writer.py:48] [190900] global_step=190900, grad_norm=5.225732326507568, loss=0.898208737373352 +I0901 00:33:04.831553 139540719195904 logging_writer.py:48] [191000] global_step=191000, grad_norm=5.011116027832031, loss=0.9347034692764282 +I0901 00:33:49.766974 139540727588608 logging_writer.py:48] [191100] global_step=191100, grad_norm=5.098219871520996, loss=0.9408119916915894 +I0901 00:34:34.935952 139540719195904 logging_writer.py:48] [191200] global_step=191200, grad_norm=5.050279140472412, loss=0.9092843532562256 +I0901 00:35:18.456201 139540727588608 logging_writer.py:48] [191300] global_step=191300, grad_norm=5.256370544433594, loss=0.984431803226471 +I0901 00:36:00.799781 139540719195904 logging_writer.py:48] [191400] global_step=191400, grad_norm=4.946031093597412, loss=0.9383278489112854 +I0901 00:36:45.877397 139540727588608 logging_writer.py:48] [191500] global_step=191500, grad_norm=5.294809818267822, loss=0.9330874681472778 +I0901 00:37:29.115872 139540719195904 logging_writer.py:48] [191600] global_step=191600, grad_norm=5.085925102233887, loss=0.9546743035316467 +I0901 00:38:15.116929 139540727588608 logging_writer.py:48] [191700] global_step=191700, grad_norm=5.306346416473389, loss=0.9143280982971191 +I0901 00:39:03.442137 139540719195904 logging_writer.py:48] [191800] global_step=191800, grad_norm=4.930082321166992, loss=0.9501845240592957 +I0901 00:39:51.570995 139540727588608 logging_writer.py:48] [191900] global_step=191900, grad_norm=5.524637222290039, loss=0.9391636252403259 +I0901 00:40:38.090644 139540719195904 logging_writer.py:48] [192000] global_step=192000, grad_norm=4.993208408355713, loss=0.9396570920944214 +I0901 00:41:25.083390 139540727588608 logging_writer.py:48] [192100] global_step=192100, grad_norm=5.270130157470703, loss=0.9398738145828247 +I0901 00:42:12.218339 139540719195904 logging_writer.py:48] [192200] global_step=192200, grad_norm=4.768538475036621, loss=0.8741495609283447 +I0901 00:42:58.064351 139540727588608 logging_writer.py:48] [192300] global_step=192300, grad_norm=5.068260669708252, loss=0.8781179785728455 +I0901 00:43:44.882850 139540719195904 logging_writer.py:48] [192400] global_step=192400, grad_norm=5.169559955596924, loss=0.9626019597053528 +I0901 00:44:32.926654 139540727588608 logging_writer.py:48] [192500] global_step=192500, grad_norm=5.09379768371582, loss=0.9447351694107056 +I0901 00:45:22.745352 139540719195904 logging_writer.py:48] [192600] global_step=192600, grad_norm=4.940635681152344, loss=0.828417956829071 +I0901 00:46:12.294629 139540727588608 logging_writer.py:48] [192700] global_step=192700, grad_norm=5.164623737335205, loss=0.9356380105018616 +I0901 00:47:02.840484 139540719195904 logging_writer.py:48] [192800] global_step=192800, grad_norm=5.236565589904785, loss=0.9732014536857605 +I0901 00:47:53.835051 139540727588608 logging_writer.py:48] [192900] global_step=192900, grad_norm=5.301437854766846, loss=0.8929015398025513 +I0901 00:48:43.059973 139540719195904 logging_writer.py:48] [193000] global_step=193000, grad_norm=5.084157466888428, loss=0.8698699474334717 +I0901 00:49:31.368182 139540727588608 logging_writer.py:48] [193100] global_step=193100, grad_norm=5.157122611999512, loss=0.9262804985046387 +I0901 00:50:21.990073 139540719195904 logging_writer.py:48] [193200] global_step=193200, grad_norm=4.918476104736328, loss=0.93666672706604 +I0901 00:51:12.021564 139540727588608 logging_writer.py:48] [193300] global_step=193300, grad_norm=5.151683807373047, loss=0.9808118343353271 +I0901 00:52:03.223154 139540719195904 logging_writer.py:48] [193400] global_step=193400, grad_norm=5.007637977600098, loss=0.8987858891487122 +I0901 00:52:53.921513 139540727588608 logging_writer.py:48] [193500] global_step=193500, grad_norm=4.923615455627441, loss=0.876431405544281 +I0901 00:53:43.720736 139540719195904 logging_writer.py:48] [193600] global_step=193600, grad_norm=4.978453636169434, loss=0.9031649827957153 +I0901 00:54:35.751241 139540727588608 logging_writer.py:48] [193700] global_step=193700, grad_norm=4.623263835906982, loss=0.8179969787597656 +I0901 00:55:27.975906 139540719195904 logging_writer.py:48] [193800] global_step=193800, grad_norm=5.20158052444458, loss=0.9650880694389343 +I0901 00:56:19.006330 139540727588608 logging_writer.py:48] [193900] global_step=193900, grad_norm=5.186981678009033, loss=0.8980474472045898 +I0901 00:57:12.231930 139540719195904 logging_writer.py:48] [194000] global_step=194000, grad_norm=5.279059886932373, loss=0.8967748880386353 +I0901 00:58:06.581253 139540727588608 logging_writer.py:48] [194100] global_step=194100, grad_norm=5.283215045928955, loss=0.8775998950004578 +I0901 00:58:59.284957 139540719195904 logging_writer.py:48] [194200] global_step=194200, grad_norm=5.1260552406311035, loss=0.9229586124420166 +I0901 00:59:51.694808 139540727588608 logging_writer.py:48] [194300] global_step=194300, grad_norm=5.099612236022949, loss=0.8545802235603333 +I0901 01:00:45.259344 139540719195904 logging_writer.py:48] [194400] global_step=194400, grad_norm=5.280905723571777, loss=0.9352467060089111 +I0901 01:01:39.665056 139540727588608 logging_writer.py:48] [194500] global_step=194500, grad_norm=5.153232574462891, loss=0.9404498934745789 +I0901 01:02:33.415260 139540719195904 logging_writer.py:48] [194600] global_step=194600, grad_norm=5.155026435852051, loss=0.8865392208099365 +I0901 01:03:25.168556 139540727588608 logging_writer.py:48] [194700] global_step=194700, grad_norm=5.160574436187744, loss=0.9380514025688171 +I0901 01:03:49.011809 139757377230016 spec.py:333] Evaluating on the training split. +I0901 01:03:58.145688 139757377230016 spec.py:346] Evaluating on the validation split. +I0901 01:04:30.979604 139757377230016 spec.py:363] Evaluating on the test split. +I0901 01:04:32.055564 139757377230016 submission_runner.py:516] Time since start: 58682.21s, Step: 194747, {'train/accuracy': Array(0.95336413, dtype=float32), 'train/loss': Array(0.16278976, dtype=float32), 'validation/accuracy': Array(0.75426, dtype=float32), 'validation/loss': Array(1.0784762, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62960005, dtype=float32), 'test/loss': Array(1.8696264, dtype=float32), 'test/num_examples': 10000, 'score': 57934.57347178459, 'total_duration': 58682.21131968498, 'accumulated_submission_time': 57934.57347178459, 'accumulated_eval_time': 740.9094491004944, 'accumulated_logging_time': 4.328880071640015} +I0901 01:04:32.653004 139540719195904 logging_writer.py:48] [194747] accumulated_eval_time=740.909, accumulated_logging_time=4.32888, accumulated_submission_time=57934.6, global_step=194747, preemption_count=0, score=57934.6, test/accuracy=0.6296000480651855, test/loss=1.8696264028549194, test/num_examples=10000, total_duration=58682.2, train/accuracy=0.9533641338348389, train/loss=0.16278976202011108, validation/accuracy=0.7542600035667419, validation/loss=1.0784761905670166, validation/num_examples=50000 +I0901 01:04:54.550012 139540727588608 logging_writer.py:48] [194800] global_step=194800, grad_norm=4.959506511688232, loss=0.9924591779708862 +I0901 01:05:48.115687 139540719195904 logging_writer.py:48] [194900] global_step=194900, grad_norm=5.291391372680664, loss=1.0125889778137207 +I0901 01:06:43.162376 139540727588608 logging_writer.py:48] [195000] global_step=195000, grad_norm=5.354170799255371, loss=0.9368957877159119 +I0901 01:07:36.485148 139540719195904 logging_writer.py:48] [195100] global_step=195100, grad_norm=5.322921276092529, loss=0.9842922687530518 +I0901 01:08:31.994592 139540727588608 logging_writer.py:48] [195200] global_step=195200, grad_norm=5.432538032531738, loss=0.9020943641662598 +I0901 01:09:25.850913 139540719195904 logging_writer.py:48] [195300] global_step=195300, grad_norm=5.183285236358643, loss=0.8722373843193054 +I0901 01:10:19.327868 139540727588608 logging_writer.py:48] [195400] global_step=195400, grad_norm=5.262555122375488, loss=0.9559160470962524 +I0901 01:11:12.986704 139540719195904 logging_writer.py:48] [195500] global_step=195500, grad_norm=5.0275654792785645, loss=0.8501360416412354 +I0901 01:12:06.012057 139540727588608 logging_writer.py:48] [195600] global_step=195600, grad_norm=5.041245460510254, loss=0.9120882153511047 +I0901 01:13:00.018879 139540719195904 logging_writer.py:48] [195700] global_step=195700, grad_norm=5.08549165725708, loss=0.9187383055686951 +I0901 01:13:52.292545 139540727588608 logging_writer.py:48] [195800] global_step=195800, grad_norm=4.833141803741455, loss=0.878483772277832 +I0901 01:14:43.669051 139540719195904 logging_writer.py:48] [195900] global_step=195900, grad_norm=4.926247596740723, loss=0.8609577417373657 +I0901 01:15:36.804445 139540727588608 logging_writer.py:48] [196000] global_step=196000, grad_norm=4.961813926696777, loss=0.9240524768829346 +I0901 01:16:27.139338 139540719195904 logging_writer.py:48] [196100] global_step=196100, grad_norm=5.05678653717041, loss=0.8589613437652588 +I0901 01:17:19.230715 139540727588608 logging_writer.py:48] [196200] global_step=196200, grad_norm=4.812471866607666, loss=0.8642634153366089 +I0901 01:18:10.854696 139540719195904 logging_writer.py:48] [196300] global_step=196300, grad_norm=5.187776565551758, loss=0.9355684518814087 +I0901 01:19:02.238332 139540727588608 logging_writer.py:48] [196400] global_step=196400, grad_norm=5.144962787628174, loss=0.907798707485199 +I0901 01:19:55.625406 139540719195904 logging_writer.py:48] [196500] global_step=196500, grad_norm=4.840472221374512, loss=0.8996868133544922 +I0901 01:20:48.726710 139540727588608 logging_writer.py:48] [196600] global_step=196600, grad_norm=5.287326812744141, loss=0.9462029933929443 +I0901 01:21:40.978081 139540719195904 logging_writer.py:48] [196700] global_step=196700, grad_norm=5.30174446105957, loss=0.9596551656723022 +I0901 01:22:31.809644 139540727588608 logging_writer.py:48] [196800] global_step=196800, grad_norm=5.1461076736450195, loss=0.8835150003433228 +I0901 01:23:23.241791 139540719195904 logging_writer.py:48] [196900] global_step=196900, grad_norm=5.3053364753723145, loss=0.9215296506881714 +I0901 01:24:15.340554 139540727588608 logging_writer.py:48] [197000] global_step=197000, grad_norm=5.017614841461182, loss=0.9195653796195984 +I0901 01:25:05.879506 139540719195904 logging_writer.py:48] [197100] global_step=197100, grad_norm=5.062988758087158, loss=0.8928728103637695 +I0901 01:26:14.057518 139540727588608 logging_writer.py:48] [197200] global_step=197200, grad_norm=5.061221599578857, loss=0.8692669868469238 +I0901 01:27:12.021266 139540719195904 logging_writer.py:48] [197300] global_step=197300, grad_norm=5.258443832397461, loss=0.9018350839614868 +I0901 01:28:01.495664 139540727588608 logging_writer.py:48] [197400] global_step=197400, grad_norm=5.124149799346924, loss=0.912934422492981 +I0901 01:28:54.797966 139540719195904 logging_writer.py:48] [197500] global_step=197500, grad_norm=5.206815242767334, loss=0.9306174516677856 +I0901 01:29:46.528371 139540727588608 logging_writer.py:48] [197600] global_step=197600, grad_norm=5.147039890289307, loss=0.9926903247833252 +I0901 01:30:40.201736 139540719195904 logging_writer.py:48] [197700] global_step=197700, grad_norm=4.9493279457092285, loss=0.8642858266830444 +I0901 01:31:33.473342 139540727588608 logging_writer.py:48] [197800] global_step=197800, grad_norm=5.06521463394165, loss=0.9507614374160767 +I0901 01:32:27.248517 139540719195904 logging_writer.py:48] [197900] global_step=197900, grad_norm=5.192773818969727, loss=0.9224342703819275 +I0901 01:33:18.744683 139540727588608 logging_writer.py:48] [198000] global_step=198000, grad_norm=5.211413860321045, loss=0.8895381689071655 +I0901 01:34:09.962068 139540719195904 logging_writer.py:48] [198100] global_step=198100, grad_norm=5.280054092407227, loss=0.8785617351531982 +I0901 01:35:03.547225 139540727588608 logging_writer.py:48] [198200] global_step=198200, grad_norm=4.875807285308838, loss=0.8343443274497986 +I0901 01:35:54.596146 139540719195904 logging_writer.py:48] [198300] global_step=198300, grad_norm=5.231433868408203, loss=0.8687576055526733 +I0901 01:36:45.800036 139540727588608 logging_writer.py:48] [198400] global_step=198400, grad_norm=5.215671539306641, loss=0.9282453060150146 +I0901 01:37:36.418729 139540719195904 logging_writer.py:48] [198500] global_step=198500, grad_norm=5.314788341522217, loss=0.9560633897781372 +I0901 01:37:48.365244 139757377230016 spec.py:333] Evaluating on the training split. +I0901 01:37:57.573513 139757377230016 spec.py:346] Evaluating on the validation split. +I0901 01:38:24.670840 139757377230016 spec.py:363] Evaluating on the test split. +I0901 01:38:25.724967 139757377230016 submission_runner.py:516] Time since start: 60715.90s, Step: 198524, {'train/accuracy': Array(0.95336413, dtype=float32), 'train/loss': Array(0.16411076, dtype=float32), 'validation/accuracy': Array(0.75376, dtype=float32), 'validation/loss': Array(1.0798633, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62740004, dtype=float32), 'test/loss': Array(1.8716532, dtype=float32), 'test/num_examples': 10000, 'score': 59930.2302134037, 'total_duration': 60715.902230262756, 'accumulated_submission_time': 59930.2302134037, 'accumulated_eval_time': 778.1086664199829, 'accumulated_logging_time': 4.950573444366455} +I0901 01:38:26.174921 139540727588608 logging_writer.py:48] [198524] accumulated_eval_time=778.109, accumulated_logging_time=4.95057, accumulated_submission_time=59930.2, global_step=198524, preemption_count=0, score=59930.2, test/accuracy=0.6274000406265259, test/loss=1.8716531991958618, test/num_examples=10000, total_duration=60715.9, train/accuracy=0.9533641338348389, train/loss=0.16411076486110687, validation/accuracy=0.7537599802017212, validation/loss=1.0798633098602295, validation/num_examples=50000 +I0901 01:38:59.338351 139540719195904 logging_writer.py:48] [198600] global_step=198600, grad_norm=5.133052825927734, loss=0.9287529587745667 +I0901 01:39:50.431074 139540727588608 logging_writer.py:48] [198700] global_step=198700, grad_norm=5.262856483459473, loss=0.902496874332428 +I0901 01:40:42.186250 139540719195904 logging_writer.py:48] [198800] global_step=198800, grad_norm=5.088845252990723, loss=0.9036923050880432 +I0901 01:41:33.720593 139540727588608 logging_writer.py:48] [198900] global_step=198900, grad_norm=5.255681991577148, loss=0.8942970037460327 +I0901 01:42:28.918137 139540719195904 logging_writer.py:48] [199000] global_step=199000, grad_norm=5.022680759429932, loss=0.8936321139335632 +I0901 01:43:21.813764 139540727588608 logging_writer.py:48] [199100] global_step=199100, grad_norm=5.19335412979126, loss=0.953155517578125 +I0901 01:44:14.575769 139540719195904 logging_writer.py:48] [199200] global_step=199200, grad_norm=5.160800933837891, loss=0.9213219881057739 +I0901 01:45:07.971538 139540727588608 logging_writer.py:48] [199300] global_step=199300, grad_norm=5.307106971740723, loss=0.9299092292785645 +I0901 01:46:01.329692 139540719195904 logging_writer.py:48] [199400] global_step=199400, grad_norm=5.286625385284424, loss=1.0115433931350708 +I0901 01:46:51.549911 139540727588608 logging_writer.py:48] [199500] global_step=199500, grad_norm=5.439139366149902, loss=0.9880576729774475 +I0901 01:47:43.748644 139540719195904 logging_writer.py:48] [199600] global_step=199600, grad_norm=4.751157760620117, loss=0.8068914413452148 +I0901 01:48:36.818563 139540727588608 logging_writer.py:48] [199700] global_step=199700, grad_norm=4.882022857666016, loss=0.8971050977706909 +I0901 01:49:29.364041 139540719195904 logging_writer.py:48] [199800] global_step=199800, grad_norm=5.326032638549805, loss=0.9369748830795288 +I0901 01:50:20.689597 139540727588608 logging_writer.py:48] [199900] global_step=199900, grad_norm=5.408930778503418, loss=0.9897217750549316 +I0901 01:51:14.305603 139540719195904 logging_writer.py:48] [200000] global_step=200000, grad_norm=4.966436862945557, loss=0.8214758634567261 +I0901 01:52:07.603240 139540727588608 logging_writer.py:48] [200100] global_step=200100, grad_norm=5.4767866134643555, loss=0.9113569259643555 +I0901 01:53:02.158096 139540719195904 logging_writer.py:48] [200200] global_step=200200, grad_norm=5.208791732788086, loss=0.9281485676765442 +I0901 01:53:56.759009 139540727588608 logging_writer.py:48] [200300] global_step=200300, grad_norm=5.209992408752441, loss=0.8563293218612671 +I0901 01:54:49.896495 139540719195904 logging_writer.py:48] [200400] global_step=200400, grad_norm=5.000430583953857, loss=0.9254909753799438 +I0901 01:55:41.001586 139540727588608 logging_writer.py:48] [200500] global_step=200500, grad_norm=5.087251663208008, loss=0.9056901931762695 +I0901 01:56:30.793543 139540719195904 logging_writer.py:48] [200600] global_step=200600, grad_norm=5.142383575439453, loss=0.9256012439727783 +I0901 01:57:24.575695 139540727588608 logging_writer.py:48] [200700] global_step=200700, grad_norm=5.265192031860352, loss=0.9384474158287048 +I0901 01:58:15.305890 139540719195904 logging_writer.py:48] [200800] global_step=200800, grad_norm=5.086089134216309, loss=0.8874866962432861 +I0901 01:59:06.779576 139540727588608 logging_writer.py:48] [200900] global_step=200900, grad_norm=5.225386142730713, loss=0.908908486366272 +I0901 01:59:59.529716 139540719195904 logging_writer.py:48] [201000] global_step=201000, grad_norm=5.085196018218994, loss=0.8600899577140808 +I0901 02:00:49.514177 139540727588608 logging_writer.py:48] [201100] global_step=201100, grad_norm=5.9253411293029785, loss=1.0941433906555176 +I0901 02:01:40.845766 139540719195904 logging_writer.py:48] [201200] global_step=201200, grad_norm=5.199377059936523, loss=0.8707232475280762 +I0901 02:02:32.892002 139540727588608 logging_writer.py:48] [201300] global_step=201300, grad_norm=5.146526336669922, loss=0.9521707892417908 +I0901 02:03:25.738438 139540719195904 logging_writer.py:48] [201400] global_step=201400, grad_norm=5.149429798126221, loss=0.9310717582702637 +I0901 02:04:19.676191 139540727588608 logging_writer.py:48] [201500] global_step=201500, grad_norm=5.063365936279297, loss=0.84711754322052 +I0901 02:05:13.356904 139540719195904 logging_writer.py:48] [201600] global_step=201600, grad_norm=5.153683185577393, loss=0.8930842876434326 +I0901 02:06:06.167235 139540727588608 logging_writer.py:48] [201700] global_step=201700, grad_norm=4.991532325744629, loss=0.8199548721313477 +I0901 02:06:58.438301 139540719195904 logging_writer.py:48] [201800] global_step=201800, grad_norm=5.295660018920898, loss=0.9175885319709778 +I0901 02:07:50.219673 139540727588608 logging_writer.py:48] [201900] global_step=201900, grad_norm=5.51363468170166, loss=0.9488140940666199 +I0901 02:08:40.001881 139540719195904 logging_writer.py:48] [202000] global_step=202000, grad_norm=5.1116557121276855, loss=0.9149681329727173 +I0901 02:09:31.057461 139540727588608 logging_writer.py:48] [202100] global_step=202100, grad_norm=5.13820219039917, loss=0.9242303967475891 +I0901 02:10:22.303673 139540719195904 logging_writer.py:48] [202200] global_step=202200, grad_norm=5.155074119567871, loss=0.9845565557479858 +I0901 02:11:12.746314 139540727588608 logging_writer.py:48] [202300] global_step=202300, grad_norm=5.196496963500977, loss=0.908519983291626 +I0901 02:11:42.108112 139757377230016 spec.py:333] Evaluating on the training split. +I0901 02:11:51.094230 139757377230016 spec.py:346] Evaluating on the validation split. +I0901 02:12:11.203205 139757377230016 spec.py:363] Evaluating on the test split. +I0901 02:12:12.292357 139757377230016 submission_runner.py:516] Time since start: 62742.43s, Step: 202362, {'train/accuracy': Array(0.952826, dtype=float32), 'train/loss': Array(0.16415659, dtype=float32), 'validation/accuracy': Array(0.75398, dtype=float32), 'validation/loss': Array(1.0797056, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62560004, dtype=float32), 'test/loss': Array(1.869789, dtype=float32), 'test/num_examples': 10000, 'score': 61926.093482255936, 'total_duration': 62742.43178105354, 'accumulated_submission_time': 61926.093482255936, 'accumulated_eval_time': 808.0945663452148, 'accumulated_logging_time': 5.439246416091919} +I0901 02:12:12.739629 139540719195904 logging_writer.py:48] [202362] accumulated_eval_time=808.095, accumulated_logging_time=5.43925, accumulated_submission_time=61926.1, global_step=202362, preemption_count=0, score=61926.1, test/accuracy=0.6256000399589539, test/loss=1.8697890043258667, test/num_examples=10000, total_duration=62742.4, train/accuracy=0.9528260231018066, train/loss=0.16415658593177795, validation/accuracy=0.7539799809455872, validation/loss=1.0797055959701538, validation/num_examples=50000 +I0901 02:12:26.132725 139540727588608 logging_writer.py:48] [202400] global_step=202400, grad_norm=5.12214469909668, loss=0.8791223764419556 +I0901 02:13:17.690087 139540719195904 logging_writer.py:48] [202500] global_step=202500, grad_norm=5.035763740539551, loss=0.8733915090560913 +I0901 02:14:08.524564 139540727588608 logging_writer.py:48] [202600] global_step=202600, grad_norm=5.007563591003418, loss=0.8624310493469238 +I0901 02:14:59.777591 139540719195904 logging_writer.py:48] [202700] global_step=202700, grad_norm=4.734543800354004, loss=0.8612897992134094 +I0901 02:15:51.884085 139540727588608 logging_writer.py:48] [202800] global_step=202800, grad_norm=4.8520188331604, loss=0.8549655079841614 +I0901 02:16:41.487578 139540719195904 logging_writer.py:48] [202900] global_step=202900, grad_norm=5.375417232513428, loss=0.9270315170288086 +I0901 02:17:31.073872 139540727588608 logging_writer.py:48] [203000] global_step=203000, grad_norm=5.422729015350342, loss=0.9670564532279968 +I0901 02:18:20.610375 139540719195904 logging_writer.py:48] [203100] global_step=203100, grad_norm=5.046518325805664, loss=0.9126133918762207 +I0901 02:19:10.610336 139540727588608 logging_writer.py:48] [203200] global_step=203200, grad_norm=4.9898834228515625, loss=0.8408831357955933 +I0901 02:19:57.997908 139540719195904 logging_writer.py:48] [203300] global_step=203300, grad_norm=4.943655014038086, loss=0.8968619108200073 +I0901 02:20:45.293951 139540727588608 logging_writer.py:48] [203400] global_step=203400, grad_norm=4.793455600738525, loss=0.7936607003211975 +I0901 02:21:33.041471 139540719195904 logging_writer.py:48] [203500] global_step=203500, grad_norm=5.137630939483643, loss=0.834844708442688 +I0901 02:22:19.616274 139540727588608 logging_writer.py:48] [203600] global_step=203600, grad_norm=5.381656169891357, loss=0.8954939842224121 +I0901 02:23:07.365447 139540719195904 logging_writer.py:48] [203700] global_step=203700, grad_norm=4.9634904861450195, loss=0.9135277271270752 +I0901 02:23:54.935192 139540727588608 logging_writer.py:48] [203800] global_step=203800, grad_norm=4.922464847564697, loss=0.8915773034095764 +I0901 02:24:42.733330 139540719195904 logging_writer.py:48] [203900] global_step=203900, grad_norm=4.956044673919678, loss=0.813348114490509 +I0901 02:25:31.700823 139540727588608 logging_writer.py:48] [204000] global_step=204000, grad_norm=5.029631614685059, loss=0.8770581483840942 +I0901 02:26:19.640208 139540719195904 logging_writer.py:48] [204100] global_step=204100, grad_norm=4.774729251861572, loss=0.8761779069900513 +I0901 02:27:08.143625 139540727588608 logging_writer.py:48] [204200] global_step=204200, grad_norm=4.886242389678955, loss=0.8930196166038513 +I0901 02:27:56.464912 139540719195904 logging_writer.py:48] [204300] global_step=204300, grad_norm=5.239152908325195, loss=0.9082820415496826 +I0901 02:28:47.838387 139540727588608 logging_writer.py:48] [204400] global_step=204400, grad_norm=5.526372909545898, loss=0.9887380599975586 +I0901 02:29:39.622215 139540719195904 logging_writer.py:48] [204500] global_step=204500, grad_norm=5.08910608291626, loss=0.9347708225250244 +I0901 02:30:27.467578 139540727588608 logging_writer.py:48] [204600] global_step=204600, grad_norm=5.0705742835998535, loss=0.9702394008636475 +I0901 02:31:15.044509 139540719195904 logging_writer.py:48] [204700] global_step=204700, grad_norm=5.058753490447998, loss=0.9157432317733765 +I0901 02:32:02.479687 139540727588608 logging_writer.py:48] [204800] global_step=204800, grad_norm=5.588639259338379, loss=0.9256415367126465 +I0901 02:32:50.411152 139540719195904 logging_writer.py:48] [204900] global_step=204900, grad_norm=5.469030857086182, loss=0.9574636816978455 +I0901 02:33:39.015216 139540727588608 logging_writer.py:48] [205000] global_step=205000, grad_norm=5.065751552581787, loss=0.9036242961883545 +I0901 02:34:26.734565 139540719195904 logging_writer.py:48] [205100] global_step=205100, grad_norm=5.214702129364014, loss=0.96802818775177 +I0901 02:35:15.472491 139540727588608 logging_writer.py:48] [205200] global_step=205200, grad_norm=5.1035003662109375, loss=0.9582258462905884 +I0901 02:36:02.307738 139540719195904 logging_writer.py:48] [205300] global_step=205300, grad_norm=5.022372722625732, loss=0.8898489475250244 +I0901 02:36:48.118750 139540727588608 logging_writer.py:48] [205400] global_step=205400, grad_norm=4.985965251922607, loss=0.947791337966919 +I0901 02:37:31.313429 139540719195904 logging_writer.py:48] [205500] global_step=205500, grad_norm=5.104970932006836, loss=0.8947805166244507 +I0901 02:38:17.952832 139540727588608 logging_writer.py:48] [205600] global_step=205600, grad_norm=5.2114667892456055, loss=0.8741760849952698 +I0901 02:39:10.297241 139540719195904 logging_writer.py:48] [205700] global_step=205700, grad_norm=5.073029041290283, loss=0.9650536775588989 +I0901 02:40:01.567557 139540727588608 logging_writer.py:48] [205800] global_step=205800, grad_norm=4.941241264343262, loss=0.8655140399932861 +I0901 02:40:52.775372 139540719195904 logging_writer.py:48] [205900] global_step=205900, grad_norm=4.8720526695251465, loss=0.8863571286201477 +I0901 02:41:45.264009 139540727588608 logging_writer.py:48] [206000] global_step=206000, grad_norm=5.331924915313721, loss=0.9609878063201904 +I0901 02:42:39.048340 139540719195904 logging_writer.py:48] [206100] global_step=206100, grad_norm=5.245893478393555, loss=0.9630399942398071 +I0901 02:43:31.700101 139540727588608 logging_writer.py:48] [206200] global_step=206200, grad_norm=5.190505504608154, loss=0.8632317781448364 +I0901 02:44:23.332803 139540719195904 logging_writer.py:48] [206300] global_step=206300, grad_norm=5.197038173675537, loss=0.9479948878288269 +I0901 02:45:17.257870 139540727588608 logging_writer.py:48] [206400] global_step=206400, grad_norm=5.357416152954102, loss=0.8937592506408691 +I0901 02:45:29.543693 139757377230016 spec.py:333] Evaluating on the training split. +I0901 02:45:38.633274 139757377230016 spec.py:346] Evaluating on the validation split. +I0901 02:46:11.627383 139757377230016 spec.py:363] Evaluating on the test split. +I0901 02:46:12.786642 139757377230016 submission_runner.py:516] Time since start: 64782.86s, Step: 206423, {'train/accuracy': Array(0.9546197, dtype=float32), 'train/loss': Array(0.15765059, dtype=float32), 'validation/accuracy': Array(0.75356, dtype=float32), 'validation/loss': Array(1.0794744, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62600005, dtype=float32), 'test/loss': Array(1.8705279, dtype=float32), 'test/num_examples': 10000, 'score': 63922.77934002876, 'total_duration': 64782.85824751854, 'accumulated_submission_time': 63922.77934002876, 'accumulated_eval_time': 851.0713455677032, 'accumulated_logging_time': 5.971511363983154} +I0901 02:46:13.301820 139540719195904 logging_writer.py:48] [206423] accumulated_eval_time=851.071, accumulated_logging_time=5.97151, accumulated_submission_time=63922.8, global_step=206423, preemption_count=0, score=63922.8, test/accuracy=0.6260000467300415, test/loss=1.8705278635025024, test/num_examples=10000, total_duration=64782.9, train/accuracy=0.9546197056770325, train/loss=0.15765058994293213, validation/accuracy=0.7535600066184998, validation/loss=1.0794744491577148, validation/num_examples=50000 +I0901 02:46:49.111222 139540727588608 logging_writer.py:48] [206500] global_step=206500, grad_norm=5.202617168426514, loss=0.9641246795654297 +I0901 02:47:41.411192 139540719195904 logging_writer.py:48] [206600] global_step=206600, grad_norm=5.029752254486084, loss=0.8831148147583008 +I0901 02:48:35.248923 139540727588608 logging_writer.py:48] [206700] global_step=206700, grad_norm=5.016756057739258, loss=0.9199930429458618 +I0901 02:49:27.255084 139540719195904 logging_writer.py:48] [206800] global_step=206800, grad_norm=5.523675918579102, loss=1.0044432878494263 +I0901 02:50:23.465807 139540727588608 logging_writer.py:48] [206900] global_step=206900, grad_norm=5.0952887535095215, loss=0.866051435470581 +I0901 02:51:20.166116 139540719195904 logging_writer.py:48] [207000] global_step=207000, grad_norm=5.325405597686768, loss=0.9375600218772888 +I0901 02:52:16.543594 139540727588608 logging_writer.py:48] [207100] global_step=207100, grad_norm=5.393423080444336, loss=0.9074574112892151 +I0901 02:53:14.668585 139540719195904 logging_writer.py:48] [207200] global_step=207200, grad_norm=5.411454677581787, loss=0.9517563581466675 +I0901 02:54:09.398904 139540727588608 logging_writer.py:48] [207300] global_step=207300, grad_norm=5.204461574554443, loss=0.904299259185791 +I0901 02:55:04.280501 139540719195904 logging_writer.py:48] [207400] global_step=207400, grad_norm=5.037196159362793, loss=0.8675855398178101 +I0901 02:56:00.617881 139540727588608 logging_writer.py:48] [207500] global_step=207500, grad_norm=4.9890313148498535, loss=0.8920211791992188 +I0901 02:56:55.937215 139540719195904 logging_writer.py:48] [207600] global_step=207600, grad_norm=5.3708295822143555, loss=1.002864122390747 +I0901 02:57:50.592459 139540727588608 logging_writer.py:48] [207700] global_step=207700, grad_norm=5.073511600494385, loss=0.8668696880340576 +I0901 02:58:42.661882 139540719195904 logging_writer.py:48] [207800] global_step=207800, grad_norm=5.0301361083984375, loss=0.8774216175079346 +I0901 02:59:36.194826 139540727588608 logging_writer.py:48] [207900] global_step=207900, grad_norm=5.3437981605529785, loss=0.9863861799240112 +I0901 03:00:29.255694 139540719195904 logging_writer.py:48] [208000] global_step=208000, grad_norm=5.057969093322754, loss=0.894850492477417 +I0901 03:01:20.520169 139540727588608 logging_writer.py:48] [208100] global_step=208100, grad_norm=5.219923496246338, loss=0.9331371784210205 +I0901 03:02:13.425932 139540719195904 logging_writer.py:48] [208200] global_step=208200, grad_norm=5.331205368041992, loss=0.8658391237258911 +I0901 03:03:03.582855 139540727588608 logging_writer.py:48] [208300] global_step=208300, grad_norm=5.390745162963867, loss=0.9374880194664001 +I0901 03:03:54.161593 139540719195904 logging_writer.py:48] [208400] global_step=208400, grad_norm=4.948375225067139, loss=0.8729747533798218 +I0901 03:04:44.700158 139540727588608 logging_writer.py:48] [208500] global_step=208500, grad_norm=5.174788951873779, loss=0.8826329708099365 +I0901 03:05:36.012768 139540719195904 logging_writer.py:48] [208600] global_step=208600, grad_norm=5.1738600730896, loss=0.9117546081542969 +I0901 03:06:27.590649 139540727588608 logging_writer.py:48] [208700] global_step=208700, grad_norm=5.097403526306152, loss=0.825337290763855 +I0901 03:07:20.106319 139540719195904 logging_writer.py:48] [208800] global_step=208800, grad_norm=5.044439315795898, loss=0.9232882261276245 +I0901 03:08:11.350457 139540727588608 logging_writer.py:48] [208900] global_step=208900, grad_norm=5.175922393798828, loss=0.9277604222297668 +I0901 03:09:04.764065 139540719195904 logging_writer.py:48] [209000] global_step=209000, grad_norm=5.394362449645996, loss=0.9403526782989502 +I0901 03:09:57.073690 139540727588608 logging_writer.py:48] [209100] global_step=209100, grad_norm=5.283134937286377, loss=0.907479465007782 +I0901 03:10:49.620942 139540719195904 logging_writer.py:48] [209200] global_step=209200, grad_norm=5.316646575927734, loss=0.9177913665771484 +I0901 03:11:42.524641 139540727588608 logging_writer.py:48] [209300] global_step=209300, grad_norm=5.263253688812256, loss=0.9701597690582275 +I0901 03:12:34.581985 139540719195904 logging_writer.py:48] [209400] global_step=209400, grad_norm=5.150289535522461, loss=0.8630836606025696 +I0901 03:13:26.007793 139540727588608 logging_writer.py:48] [209500] global_step=209500, grad_norm=5.359416961669922, loss=0.9340381622314453 +I0901 03:14:17.037712 139540719195904 logging_writer.py:48] [209600] global_step=209600, grad_norm=5.081257343292236, loss=0.8694079518318176 +I0901 03:15:06.503660 139540727588608 logging_writer.py:48] [209700] global_step=209700, grad_norm=5.250353813171387, loss=0.9725810885429382 +I0901 03:15:56.844247 139540719195904 logging_writer.py:48] [209800] global_step=209800, grad_norm=4.9578728675842285, loss=0.8607099056243896 +I0901 03:16:48.316752 139540727588608 logging_writer.py:48] [209900] global_step=209900, grad_norm=5.009621620178223, loss=0.9333705306053162 +I0901 03:17:41.839717 139540719195904 logging_writer.py:48] [210000] global_step=210000, grad_norm=5.106468200683594, loss=0.9120402336120605 +I0901 03:18:34.520944 139540727588608 logging_writer.py:48] [210100] global_step=210100, grad_norm=5.289283275604248, loss=0.9164851903915405 +I0901 03:19:27.704746 139540719195904 logging_writer.py:48] [210200] global_step=210200, grad_norm=4.926700115203857, loss=0.8902142643928528 +I0901 03:19:28.824325 139757377230016 spec.py:333] Evaluating on the training split. +I0901 03:19:37.872931 139757377230016 spec.py:346] Evaluating on the validation split. +I0901 03:20:04.810120 139757377230016 spec.py:363] Evaluating on the test split. +I0901 03:20:05.867012 139757377230016 submission_runner.py:516] Time since start: 66816.04s, Step: 210203, {'train/accuracy': Array(0.9555166, dtype=float32), 'train/loss': Array(0.15610996, dtype=float32), 'validation/accuracy': Array(0.75497997, dtype=float32), 'validation/loss': Array(1.0815759, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62740004, dtype=float32), 'test/loss': Array(1.8746012, dtype=float32), 'test/num_examples': 10000, 'score': 65918.245708704, 'total_duration': 66816.03911733627, 'accumulated_submission_time': 65918.245708704, 'accumulated_eval_time': 887.9483625888824, 'accumulated_logging_time': 6.512280464172363} +I0901 03:20:06.284305 139540727588608 logging_writer.py:48] [210203] accumulated_eval_time=887.948, accumulated_logging_time=6.51228, accumulated_submission_time=65918.2, global_step=210203, preemption_count=0, score=65918.2, test/accuracy=0.6274000406265259, test/loss=1.8746012449264526, test/num_examples=10000, total_duration=66816, train/accuracy=0.9555165767669678, train/loss=0.15610995888710022, validation/accuracy=0.7549799680709839, validation/loss=1.081575870513916, validation/num_examples=50000 +I0901 03:20:50.026540 139540719195904 logging_writer.py:48] [210300] global_step=210300, grad_norm=5.099100589752197, loss=0.9033666849136353 +I0901 03:21:41.471609 139540727588608 logging_writer.py:48] [210400] global_step=210400, grad_norm=4.975542068481445, loss=0.8824818134307861 +I0901 03:22:34.061374 139540719195904 logging_writer.py:48] [210500] global_step=210500, grad_norm=4.950562953948975, loss=0.8531773090362549 +I0901 03:23:24.450270 139540727588608 logging_writer.py:48] [210600] global_step=210600, grad_norm=5.150694847106934, loss=0.9228275418281555 +I0901 03:24:15.651481 139540719195904 logging_writer.py:48] [210700] global_step=210700, grad_norm=5.025702953338623, loss=0.8377450704574585 +I0901 03:25:06.200978 139540727588608 logging_writer.py:48] [210800] global_step=210800, grad_norm=5.244544982910156, loss=0.9158619046211243 +I0901 03:25:57.274631 139540719195904 logging_writer.py:48] [210900] global_step=210900, grad_norm=4.929859638214111, loss=0.8668630719184875 +I0901 03:26:48.221316 139540727588608 logging_writer.py:48] [211000] global_step=211000, grad_norm=5.329727649688721, loss=0.9470930099487305 +I0901 03:27:39.206716 139540719195904 logging_writer.py:48] [211100] global_step=211100, grad_norm=5.25602912902832, loss=0.9449210166931152 +I0901 03:28:30.584103 139540727588608 logging_writer.py:48] [211200] global_step=211200, grad_norm=5.1287431716918945, loss=0.8842076659202576 +I0901 03:29:20.120266 139540719195904 logging_writer.py:48] [211300] global_step=211300, grad_norm=4.756224632263184, loss=0.8511444330215454 +I0901 03:30:10.911954 139540727588608 logging_writer.py:48] [211400] global_step=211400, grad_norm=4.828818321228027, loss=0.8599194288253784 +I0901 03:31:04.398240 139540719195904 logging_writer.py:48] [211500] global_step=211500, grad_norm=5.161365985870361, loss=0.9285765290260315 +I0901 03:31:54.495291 139540727588608 logging_writer.py:48] [211600] global_step=211600, grad_norm=5.312283039093018, loss=0.9032130837440491 +I0901 03:32:45.092459 139540719195904 logging_writer.py:48] [211700] global_step=211700, grad_norm=5.247129440307617, loss=0.9516128301620483 +I0901 03:33:34.874748 139540727588608 logging_writer.py:48] [211800] global_step=211800, grad_norm=4.607057571411133, loss=0.7894221544265747 +I0901 03:34:26.392297 139540719195904 logging_writer.py:48] [211900] global_step=211900, grad_norm=4.869307518005371, loss=0.8642390370368958 +I0901 03:35:17.363977 139540727588608 logging_writer.py:48] [212000] global_step=212000, grad_norm=5.008450984954834, loss=0.9228736162185669 +I0901 03:36:08.463917 139540719195904 logging_writer.py:48] [212100] global_step=212100, grad_norm=5.093581199645996, loss=0.9539191722869873 +I0901 03:36:59.480361 139540727588608 logging_writer.py:48] [212200] global_step=212200, grad_norm=5.203867435455322, loss=0.8432280421257019 +I0901 03:37:49.047150 139540719195904 logging_writer.py:48] [212300] global_step=212300, grad_norm=4.9379987716674805, loss=0.8450958728790283 +I0901 03:38:38.140240 139540727588608 logging_writer.py:48] [212400] global_step=212400, grad_norm=5.025061130523682, loss=0.8752448558807373 +I0901 03:39:29.007641 139540719195904 logging_writer.py:48] [212500] global_step=212500, grad_norm=5.153323650360107, loss=0.9192014932632446 +I0901 03:40:18.735784 139540727588608 logging_writer.py:48] [212600] global_step=212600, grad_norm=5.235291481018066, loss=0.8920343518257141 +I0901 03:41:10.573628 139540719195904 logging_writer.py:48] [212700] global_step=212700, grad_norm=4.923084259033203, loss=0.9090394973754883 +I0901 03:42:01.747575 139540727588608 logging_writer.py:48] [212800] global_step=212800, grad_norm=5.1715545654296875, loss=0.9721049666404724 +I0901 03:42:54.378849 139540719195904 logging_writer.py:48] [212900] global_step=212900, grad_norm=5.61404275894165, loss=1.0218350887298584 +I0901 03:43:43.873031 139540727588608 logging_writer.py:48] [213000] global_step=213000, grad_norm=5.3892364501953125, loss=0.9390974044799805 +I0901 03:44:32.140819 139540719195904 logging_writer.py:48] [213100] global_step=213100, grad_norm=5.290483474731445, loss=0.9249717593193054 +I0901 03:45:24.781072 139540727588608 logging_writer.py:48] [213200] global_step=213200, grad_norm=5.246114730834961, loss=0.9054976105690002 +I0901 03:46:13.726725 139540719195904 logging_writer.py:48] [213300] global_step=213300, grad_norm=5.045130729675293, loss=0.9203429222106934 +I0901 03:47:01.882310 139540727588608 logging_writer.py:48] [213400] global_step=213400, grad_norm=5.1387810707092285, loss=0.9174758195877075 +I0901 03:47:52.096464 139540719195904 logging_writer.py:48] [213500] global_step=213500, grad_norm=5.151663303375244, loss=0.8971946239471436 +I0901 03:48:42.630128 139540727588608 logging_writer.py:48] [213600] global_step=213600, grad_norm=5.194803714752197, loss=0.8475130200386047 +I0901 03:49:31.474825 139540719195904 logging_writer.py:48] [213700] global_step=213700, grad_norm=5.599254608154297, loss=0.9677801132202148 +I0901 03:50:21.417097 139540727588608 logging_writer.py:48] [213800] global_step=213800, grad_norm=5.40997314453125, loss=1.0112439393997192 +I0901 03:51:14.807210 139540719195904 logging_writer.py:48] [213900] global_step=213900, grad_norm=5.297396183013916, loss=0.943534255027771 +I0901 03:52:06.112943 139540727588608 logging_writer.py:48] [214000] global_step=214000, grad_norm=5.11482048034668, loss=0.8433535695075989 +I0901 03:52:56.101496 139540719195904 logging_writer.py:48] [214100] global_step=214100, grad_norm=5.192660331726074, loss=0.94236159324646 +I0901 03:53:22.202761 139757377230016 spec.py:333] Evaluating on the training split. +I0901 03:53:31.496147 139757377230016 spec.py:346] Evaluating on the validation split. +I0901 03:54:00.051017 139757377230016 spec.py:363] Evaluating on the test split. +I0901 03:54:01.180869 139757377230016 submission_runner.py:516] Time since start: 68851.28s, Step: 214153, {'train/accuracy': Array(0.95475924, dtype=float32), 'train/loss': Array(0.15814532, dtype=float32), 'validation/accuracy': Array(0.755, dtype=float32), 'validation/loss': Array(1.0817531, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6273, dtype=float32), 'test/loss': Array(1.8760587, dtype=float32), 'test/num_examples': 10000, 'score': 67914.09925222397, 'total_duration': 68851.28239226341, 'accumulated_submission_time': 67914.09925222397, 'accumulated_eval_time': 926.6902208328247, 'accumulated_logging_time': 6.9625163078308105} +I0901 03:54:01.683430 139540727588608 logging_writer.py:48] [214153] accumulated_eval_time=926.69, accumulated_logging_time=6.96252, accumulated_submission_time=67914.1, global_step=214153, preemption_count=0, score=67914.1, test/accuracy=0.6273000240325928, test/loss=1.8760586977005005, test/num_examples=10000, total_duration=68851.3, train/accuracy=0.9547592401504517, train/loss=0.15814532339572906, validation/accuracy=0.7549999952316284, validation/loss=1.081753134727478, validation/num_examples=50000 +I0901 03:54:19.172137 139540719195904 logging_writer.py:48] [214200] global_step=214200, grad_norm=5.291746616363525, loss=0.9927453994750977 +I0901 03:55:08.421885 139540727588608 logging_writer.py:48] [214300] global_step=214300, grad_norm=5.602275371551514, loss=1.0039091110229492 +I0901 03:55:59.023420 139540719195904 logging_writer.py:48] [214400] global_step=214400, grad_norm=5.197666645050049, loss=0.9371480345726013 +I0901 03:56:49.525973 139540727588608 logging_writer.py:48] [214500] global_step=214500, grad_norm=5.17739725112915, loss=0.8985241651535034 +I0901 03:57:38.286812 139540719195904 logging_writer.py:48] [214600] global_step=214600, grad_norm=5.16904354095459, loss=0.8873833417892456 +I0901 03:58:26.659506 139540727588608 logging_writer.py:48] [214700] global_step=214700, grad_norm=4.8706159591674805, loss=0.9196385145187378 +I0901 03:59:16.584637 139540719195904 logging_writer.py:48] [214800] global_step=214800, grad_norm=5.286158561706543, loss=0.948022723197937 +I0901 04:00:05.574454 139540727588608 logging_writer.py:48] [214900] global_step=214900, grad_norm=5.207430362701416, loss=0.9499319791793823 +I0901 04:00:55.754025 139540719195904 logging_writer.py:48] [215000] global_step=215000, grad_norm=4.990926742553711, loss=0.8281792402267456 +I0901 04:01:45.724997 139540727588608 logging_writer.py:48] [215100] global_step=215100, grad_norm=5.149868488311768, loss=0.9357812404632568 +I0901 04:02:36.136605 139540719195904 logging_writer.py:48] [215200] global_step=215200, grad_norm=5.248409748077393, loss=0.9625259637832642 +I0901 04:03:25.966657 139540727588608 logging_writer.py:48] [215300] global_step=215300, grad_norm=5.1366496086120605, loss=0.8769013285636902 +I0901 04:04:15.892117 139540719195904 logging_writer.py:48] [215400] global_step=215400, grad_norm=4.964018821716309, loss=1.0062419176101685 +I0901 04:05:04.340402 139540727588608 logging_writer.py:48] [215500] global_step=215500, grad_norm=5.2173566818237305, loss=0.9428892731666565 +I0901 04:05:53.880709 139540719195904 logging_writer.py:48] [215600] global_step=215600, grad_norm=4.984160900115967, loss=0.9021590948104858 +I0901 04:06:45.272846 139540727588608 logging_writer.py:48] [215700] global_step=215700, grad_norm=4.788780689239502, loss=0.835458517074585 +I0901 04:07:33.804284 139540719195904 logging_writer.py:48] [215800] global_step=215800, grad_norm=5.345922470092773, loss=0.9986333250999451 +I0901 04:08:22.976335 139540727588608 logging_writer.py:48] [215900] global_step=215900, grad_norm=5.213534832000732, loss=0.9682459831237793 +I0901 04:09:11.083518 139540719195904 logging_writer.py:48] [216000] global_step=216000, grad_norm=5.151462554931641, loss=0.8738299608230591 +I0901 04:09:58.583831 139540727588608 logging_writer.py:48] [216100] global_step=216100, grad_norm=5.086813449859619, loss=0.9186927080154419 +I0901 04:10:45.645851 139540719195904 logging_writer.py:48] [216200] global_step=216200, grad_norm=4.848623752593994, loss=0.8761664628982544 +I0901 04:11:33.411433 139540727588608 logging_writer.py:48] [216300] global_step=216300, grad_norm=5.241093158721924, loss=0.9542791247367859 +I0901 04:12:21.104605 139540719195904 logging_writer.py:48] [216400] global_step=216400, grad_norm=5.07101583480835, loss=0.9185527563095093 +I0901 04:13:11.579288 139540727588608 logging_writer.py:48] [216500] global_step=216500, grad_norm=4.984597682952881, loss=0.8720291256904602 +I0901 04:13:59.456171 139540719195904 logging_writer.py:48] [216600] global_step=216600, grad_norm=5.314583778381348, loss=0.9751672744750977 +I0901 04:14:49.419238 139540727588608 logging_writer.py:48] [216700] global_step=216700, grad_norm=5.466006755828857, loss=0.9345518946647644 +I0901 04:15:38.077586 139540719195904 logging_writer.py:48] [216800] global_step=216800, grad_norm=4.942784786224365, loss=0.8965688347816467 +I0901 04:16:28.309942 139540727588608 logging_writer.py:48] [216900] global_step=216900, grad_norm=5.509568691253662, loss=0.9572007656097412 +I0901 04:17:18.677536 139540719195904 logging_writer.py:48] [217000] global_step=217000, grad_norm=5.091160774230957, loss=0.8252366781234741 +I0901 04:18:08.240264 139540727588608 logging_writer.py:48] [217100] global_step=217100, grad_norm=5.339501857757568, loss=0.9927040338516235 +I0901 04:18:57.472813 139540719195904 logging_writer.py:48] [217200] global_step=217200, grad_norm=5.141698837280273, loss=0.8244807720184326 +I0901 04:19:44.286229 139540727588608 logging_writer.py:48] [217300] global_step=217300, grad_norm=5.125891208648682, loss=0.8821707963943481 +I0901 04:20:31.100716 139540719195904 logging_writer.py:48] [217400] global_step=217400, grad_norm=5.215120315551758, loss=0.9937551021575928 +I0901 04:21:19.136990 139540727588608 logging_writer.py:48] [217500] global_step=217500, grad_norm=5.033257007598877, loss=0.9200013875961304 +I0901 04:22:05.212984 139540719195904 logging_writer.py:48] [217600] global_step=217600, grad_norm=5.248668670654297, loss=0.9175803661346436 +I0901 04:22:53.942009 139540727588608 logging_writer.py:48] [217700] global_step=217700, grad_norm=5.173306941986084, loss=0.9010351896286011 +I0901 04:23:42.524721 139540719195904 logging_writer.py:48] [217800] global_step=217800, grad_norm=5.019108772277832, loss=0.8735446929931641 +I0901 04:24:30.690829 139540727588608 logging_writer.py:48] [217900] global_step=217900, grad_norm=5.224678039550781, loss=0.9725149869918823 +I0901 04:25:17.534569 139540719195904 logging_writer.py:48] [218000] global_step=218000, grad_norm=4.983287334442139, loss=0.8336320519447327 +I0901 04:26:04.204361 139540727588608 logging_writer.py:48] [218100] global_step=218100, grad_norm=5.2645697593688965, loss=0.9488639235496521 +I0901 04:26:52.205862 139540719195904 logging_writer.py:48] [218200] global_step=218200, grad_norm=5.136441230773926, loss=0.9232617020606995 +I0901 04:27:17.212783 139757377230016 spec.py:333] Evaluating on the training split. +I0901 04:27:26.374823 139757377230016 spec.py:346] Evaluating on the validation split. +I0901 04:27:55.670266 139757377230016 spec.py:363] Evaluating on the test split. +I0901 04:27:56.726056 139757377230016 submission_runner.py:516] Time since start: 70886.90s, Step: 218254, {'train/accuracy': Array(0.9552176, dtype=float32), 'train/loss': Array(0.1577096, dtype=float32), 'validation/accuracy': Array(0.75531995, dtype=float32), 'validation/loss': Array(1.0797553, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6268, dtype=float32), 'test/loss': Array(1.8753902, dtype=float32), 'test/num_examples': 10000, 'score': 69909.51558828354, 'total_duration': 70886.89947223663, 'accumulated_submission_time': 69909.51558828354, 'accumulated_eval_time': 966.0391397476196, 'accumulated_logging_time': 7.543886661529541} +I0901 04:27:57.180701 139540727588608 logging_writer.py:48] [218254] accumulated_eval_time=966.039, accumulated_logging_time=7.54389, accumulated_submission_time=69909.5, global_step=218254, preemption_count=0, score=69909.5, test/accuracy=0.626800000667572, test/loss=1.8753901720046997, test/num_examples=10000, total_duration=70886.9, train/accuracy=0.9552175998687744, train/loss=0.15770959854125977, validation/accuracy=0.7553199529647827, validation/loss=1.0797553062438965, validation/num_examples=50000 +I0901 04:28:13.955512 139540719195904 logging_writer.py:48] [218300] global_step=218300, grad_norm=5.42689847946167, loss=0.9464742541313171 +I0901 04:29:00.325412 139540727588608 logging_writer.py:48] [218400] global_step=218400, grad_norm=5.34384822845459, loss=0.9770093560218811 +I0901 04:29:46.040627 139540719195904 logging_writer.py:48] [218500] global_step=218500, grad_norm=4.801495552062988, loss=0.8362554907798767 +I0901 04:30:33.299871 139540727588608 logging_writer.py:48] [218600] global_step=218600, grad_norm=4.985342979431152, loss=0.8454725742340088 +I0901 04:31:21.228797 139540719195904 logging_writer.py:48] [218700] global_step=218700, grad_norm=5.232139587402344, loss=0.9690248966217041 +I0901 04:32:08.005173 139540727588608 logging_writer.py:48] [218800] global_step=218800, grad_norm=5.334331512451172, loss=0.9308943748474121 +I0901 04:32:55.718551 139540719195904 logging_writer.py:48] [218900] global_step=218900, grad_norm=4.996484756469727, loss=0.8513368368148804 +I0901 04:33:44.281551 139540727588608 logging_writer.py:48] [219000] global_step=219000, grad_norm=4.8881378173828125, loss=0.8675321340560913 +I0901 04:34:31.863298 139540719195904 logging_writer.py:48] [219100] global_step=219100, grad_norm=4.959918022155762, loss=0.871385931968689 +I0901 04:35:18.080663 139540727588608 logging_writer.py:48] [219200] global_step=219200, grad_norm=5.261718273162842, loss=0.9448763132095337 +I0901 04:36:04.777388 139540719195904 logging_writer.py:48] [219300] global_step=219300, grad_norm=5.069844722747803, loss=0.9205812215805054 +I0901 04:36:53.694897 139540727588608 logging_writer.py:48] [219400] global_step=219400, grad_norm=5.372315406799316, loss=0.9106943011283875 +I0901 04:37:40.487114 139540719195904 logging_writer.py:48] [219500] global_step=219500, grad_norm=4.93790864944458, loss=0.834879994392395 +I0901 04:38:26.593497 139540727588608 logging_writer.py:48] [219600] global_step=219600, grad_norm=4.809520244598389, loss=0.8052026033401489 +I0901 04:39:12.096269 139540719195904 logging_writer.py:48] [219700] global_step=219700, grad_norm=5.252195835113525, loss=0.9424095153808594 +I0901 04:39:56.615804 139540727588608 logging_writer.py:48] [219800] global_step=219800, grad_norm=5.2140045166015625, loss=0.9356333017349243 +I0901 04:40:42.652881 139540719195904 logging_writer.py:48] [219900] global_step=219900, grad_norm=4.8710479736328125, loss=0.8536181449890137 +I0901 04:41:29.078588 139540727588608 logging_writer.py:48] [220000] global_step=220000, grad_norm=5.391149520874023, loss=0.9055112600326538 +I0901 04:42:16.538896 139540719195904 logging_writer.py:48] [220100] global_step=220100, grad_norm=5.365960597991943, loss=0.9856511354446411 +I0901 04:43:04.664084 139540727588608 logging_writer.py:48] [220200] global_step=220200, grad_norm=5.106783390045166, loss=0.9476877450942993 +I0901 04:43:51.158096 139540719195904 logging_writer.py:48] [220300] global_step=220300, grad_norm=5.1661787033081055, loss=0.8866358995437622 +I0901 04:44:36.429673 139540727588608 logging_writer.py:48] [220400] global_step=220400, grad_norm=5.129538536071777, loss=0.816411018371582 +I0901 04:45:21.383845 139540719195904 logging_writer.py:48] [220500] global_step=220500, grad_norm=5.283844947814941, loss=0.9678705930709839 +I0901 04:46:04.471638 139540727588608 logging_writer.py:48] [220600] global_step=220600, grad_norm=5.321647644042969, loss=0.941784679889679 +I0901 04:46:49.431219 139540719195904 logging_writer.py:48] [220700] global_step=220700, grad_norm=5.2305521965026855, loss=0.9197605848312378 +I0901 04:47:34.157876 139540727588608 logging_writer.py:48] [220800] global_step=220800, grad_norm=4.89666748046875, loss=0.8537352681159973 +I0901 04:48:17.805288 139540719195904 logging_writer.py:48] [220900] global_step=220900, grad_norm=5.252561092376709, loss=0.9207080602645874 +I0901 04:49:01.668576 139540727588608 logging_writer.py:48] [221000] global_step=221000, grad_norm=4.995283126831055, loss=0.9387847781181335 +I0901 04:49:44.299897 139540719195904 logging_writer.py:48] [221100] global_step=221100, grad_norm=5.125156879425049, loss=0.8710323572158813 +I0901 04:50:26.202198 139540727588608 logging_writer.py:48] [221200] global_step=221200, grad_norm=5.203972816467285, loss=0.9590159058570862 +I0901 04:51:08.172515 139540719195904 logging_writer.py:48] [221300] global_step=221300, grad_norm=5.342173099517822, loss=0.9684510231018066 +I0901 04:51:51.469592 139540727588608 logging_writer.py:48] [221400] global_step=221400, grad_norm=5.202019214630127, loss=0.9197342395782471 +I0901 04:52:36.700766 139540719195904 logging_writer.py:48] [221500] global_step=221500, grad_norm=5.204736232757568, loss=0.9113171100616455 +I0901 04:53:18.803191 139540727588608 logging_writer.py:48] [221600] global_step=221600, grad_norm=5.098732948303223, loss=0.8688445687294006 +I0901 04:54:00.978747 139540719195904 logging_writer.py:48] [221700] global_step=221700, grad_norm=5.2412614822387695, loss=0.9510133862495422 +I0901 04:54:43.595762 139540727588608 logging_writer.py:48] [221800] global_step=221800, grad_norm=5.0428385734558105, loss=0.799647867679596 +I0901 04:55:24.396027 139540719195904 logging_writer.py:48] [221900] global_step=221900, grad_norm=4.9967546463012695, loss=0.8620136976242065 +I0901 04:56:05.639816 139540727588608 logging_writer.py:48] [222000] global_step=222000, grad_norm=5.202873229980469, loss=0.9019264578819275 +I0901 04:56:46.259544 139540719195904 logging_writer.py:48] [222100] global_step=222100, grad_norm=5.173943042755127, loss=0.9671158790588379 +I0901 04:57:27.778879 139540727588608 logging_writer.py:48] [222200] global_step=222200, grad_norm=5.098195552825928, loss=0.883571982383728 +I0901 04:58:08.724101 139540719195904 logging_writer.py:48] [222300] global_step=222300, grad_norm=5.321022987365723, loss=0.9466314315795898 +I0901 04:58:50.390620 139540727588608 logging_writer.py:48] [222400] global_step=222400, grad_norm=5.031403064727783, loss=0.8883345127105713 +I0901 04:59:30.867172 139540719195904 logging_writer.py:48] [222500] global_step=222500, grad_norm=5.149908065795898, loss=0.9178547859191895 +I0901 05:00:13.515396 139540727588608 logging_writer.py:48] [222600] global_step=222600, grad_norm=5.175168037414551, loss=0.9811258912086487 +I0901 05:00:56.591343 139540719195904 logging_writer.py:48] [222700] global_step=222700, grad_norm=5.267349720001221, loss=0.9873952269554138 +I0901 05:01:12.899752 139757377230016 spec.py:333] Evaluating on the training split. +I0901 05:01:22.184171 139757377230016 spec.py:346] Evaluating on the validation split. +I0901 05:02:12.578248 139757377230016 spec.py:363] Evaluating on the test split. +I0901 05:02:13.599207 139757377230016 submission_runner.py:516] Time since start: 72943.81s, Step: 222739, {'train/accuracy': Array(0.9566725, dtype=float32), 'train/loss': Array(0.15274695, dtype=float32), 'validation/accuracy': Array(0.75558, dtype=float32), 'validation/loss': Array(1.0793678, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62960005, dtype=float32), 'test/loss': Array(1.8752167, dtype=float32), 'test/num_examples': 10000, 'score': 71905.1672899723, 'total_duration': 72943.8052880764, 'accumulated_submission_time': 71905.1672899723, 'accumulated_eval_time': 1026.6069040298462, 'accumulated_logging_time': 8.02806806564331} +I0901 05:02:14.125226 139540727588608 logging_writer.py:48] [222739] accumulated_eval_time=1026.61, accumulated_logging_time=8.02807, accumulated_submission_time=71905.2, global_step=222739, preemption_count=0, score=71905.2, test/accuracy=0.6296000480651855, test/loss=1.8752167224884033, test/num_examples=10000, total_duration=72943.8, train/accuracy=0.9566724896430969, train/loss=0.15274694561958313, validation/accuracy=0.7555800080299377, validation/loss=1.079367756843567, validation/num_examples=50000 +I0901 05:02:34.984693 139540719195904 logging_writer.py:48] [222800] global_step=222800, grad_norm=5.141963005065918, loss=0.9152886867523193 +I0901 05:03:16.640876 139540727588608 logging_writer.py:48] [222900] global_step=222900, grad_norm=5.461738586425781, loss=0.9359047412872314 +I0901 05:03:58.744085 139540719195904 logging_writer.py:48] [223000] global_step=223000, grad_norm=5.523144721984863, loss=1.034278392791748 +I0901 05:04:40.755290 139540727588608 logging_writer.py:48] [223100] global_step=223100, grad_norm=5.295791149139404, loss=0.8724557757377625 +I0901 05:05:22.878063 139540719195904 logging_writer.py:48] [223200] global_step=223200, grad_norm=5.23906946182251, loss=0.9803253412246704 +I0901 05:06:03.806083 139540727588608 logging_writer.py:48] [223300] global_step=223300, grad_norm=5.258750915527344, loss=0.9197937250137329 +I0901 05:06:45.026330 139540719195904 logging_writer.py:48] [223400] global_step=223400, grad_norm=5.413966178894043, loss=0.846173882484436 +I0901 05:07:27.469680 139540727588608 logging_writer.py:48] [223500] global_step=223500, grad_norm=5.157536506652832, loss=0.8890392184257507 +I0901 05:08:09.190706 139540719195904 logging_writer.py:48] [223600] global_step=223600, grad_norm=5.050731658935547, loss=0.7874510884284973 +I0901 05:08:50.839274 139540727588608 logging_writer.py:48] [223700] global_step=223700, grad_norm=5.335750579833984, loss=0.927159309387207 +I0901 05:09:32.522250 139540719195904 logging_writer.py:48] [223800] global_step=223800, grad_norm=5.232258319854736, loss=0.8846101760864258 +I0901 05:10:14.466700 139540727588608 logging_writer.py:48] [223900] global_step=223900, grad_norm=4.881683826446533, loss=0.8007798194885254 +I0901 05:10:59.278696 139540719195904 logging_writer.py:48] [224000] global_step=224000, grad_norm=5.07036018371582, loss=0.931265115737915 +I0901 05:11:42.418146 139540727588608 logging_writer.py:48] [224100] global_step=224100, grad_norm=5.323392391204834, loss=0.9066078662872314 +I0901 05:12:24.869518 139540719195904 logging_writer.py:48] [224200] global_step=224200, grad_norm=5.26594352722168, loss=0.869032621383667 +I0901 05:13:06.657959 139540727588608 logging_writer.py:48] [224300] global_step=224300, grad_norm=4.887566566467285, loss=0.8749510049819946 +I0901 05:13:48.369122 139540719195904 logging_writer.py:48] [224400] global_step=224400, grad_norm=5.492791652679443, loss=0.9542075395584106 +I0901 05:14:30.849414 139540727588608 logging_writer.py:48] [224500] global_step=224500, grad_norm=5.074965953826904, loss=0.9694768190383911 +I0901 05:15:12.305689 139540719195904 logging_writer.py:48] [224600] global_step=224600, grad_norm=5.061837196350098, loss=0.8681414723396301 +I0901 05:15:54.460112 139540727588608 logging_writer.py:48] [224700] global_step=224700, grad_norm=4.826691627502441, loss=0.9221981763839722 +I0901 05:16:32.383496 139540719195904 logging_writer.py:48] [224800] global_step=224800, grad_norm=5.093710422515869, loss=0.8912603259086609 +I0901 05:17:06.371908 139540727588608 logging_writer.py:48] [224900] global_step=224900, grad_norm=5.322624683380127, loss=0.9293590784072876 +I0901 05:17:41.398054 139540719195904 logging_writer.py:48] [225000] global_step=225000, grad_norm=5.097079277038574, loss=0.9115729331970215 +I0901 05:18:15.227447 139540727588608 logging_writer.py:48] [225100] global_step=225100, grad_norm=5.352234840393066, loss=0.8734457492828369 +I0901 05:18:50.434811 139540719195904 logging_writer.py:48] [225200] global_step=225200, grad_norm=5.176937103271484, loss=0.9769225120544434 +I0901 05:19:24.149456 139540727588608 logging_writer.py:48] [225300] global_step=225300, grad_norm=5.121140956878662, loss=0.8887336254119873 +I0901 05:19:57.973713 139540719195904 logging_writer.py:48] [225400] global_step=225400, grad_norm=5.026456356048584, loss=0.8365431427955627 +I0901 05:20:31.703753 139540727588608 logging_writer.py:48] [225500] global_step=225500, grad_norm=5.142284393310547, loss=0.8570677042007446 +I0901 05:21:06.565064 139540719195904 logging_writer.py:48] [225600] global_step=225600, grad_norm=4.6825151443481445, loss=0.8323724865913391 +I0901 05:21:40.298228 139540727588608 logging_writer.py:48] [225700] global_step=225700, grad_norm=5.272064208984375, loss=0.9946851134300232 +I0901 05:22:12.447326 139540719195904 logging_writer.py:48] [225800] global_step=225800, grad_norm=5.2402496337890625, loss=0.9533406496047974 +I0901 05:22:44.054078 139540727588608 logging_writer.py:48] [225900] global_step=225900, grad_norm=4.856729984283447, loss=0.8267614245414734 +I0901 05:23:15.439943 139540719195904 logging_writer.py:48] [226000] global_step=226000, grad_norm=5.149362564086914, loss=0.8489689826965332 +I0901 05:23:47.917967 139540727588608 logging_writer.py:48] [226100] global_step=226100, grad_norm=5.282172679901123, loss=0.89263916015625 +I0901 05:24:20.391207 139540719195904 logging_writer.py:48] [226200] global_step=226200, grad_norm=4.9914984703063965, loss=0.9247154593467712 +I0901 05:24:53.100260 139540727588608 logging_writer.py:48] [226300] global_step=226300, grad_norm=4.962143421173096, loss=0.879338264465332 +I0901 05:25:26.001520 139540719195904 logging_writer.py:48] [226400] global_step=226400, grad_norm=5.329699516296387, loss=0.9033814072608948 +I0901 05:25:59.477441 139540727588608 logging_writer.py:48] [226500] global_step=226500, grad_norm=4.72149133682251, loss=0.822277843952179 +I0901 05:26:32.268574 139540719195904 logging_writer.py:48] [226600] global_step=226600, grad_norm=5.4829182624816895, loss=0.9126917123794556 +I0901 05:27:05.346985 139540727588608 logging_writer.py:48] [226700] global_step=226700, grad_norm=5.220470428466797, loss=0.9230283498764038 +I0901 05:27:38.532274 139540719195904 logging_writer.py:48] [226800] global_step=226800, grad_norm=5.133101463317871, loss=0.8413693904876709 +I0901 05:28:10.902464 139540727588608 logging_writer.py:48] [226900] global_step=226900, grad_norm=5.26683235168457, loss=0.9465717077255249 +I0901 05:28:43.773143 139540719195904 logging_writer.py:48] [227000] global_step=227000, grad_norm=5.121135711669922, loss=0.901813268661499 +I0901 05:29:16.428620 139540727588608 logging_writer.py:48] [227100] global_step=227100, grad_norm=4.928009033203125, loss=0.8093453645706177 +I0901 05:29:49.447078 139540719195904 logging_writer.py:48] [227200] global_step=227200, grad_norm=5.209933280944824, loss=0.9842454791069031 +I0901 05:30:22.494699 139540727588608 logging_writer.py:48] [227300] global_step=227300, grad_norm=5.702541828155518, loss=0.9609793424606323 +I0901 05:30:55.302270 139540719195904 logging_writer.py:48] [227400] global_step=227400, grad_norm=5.278942108154297, loss=0.896742582321167 +I0901 05:31:28.526643 139540727588608 logging_writer.py:48] [227500] global_step=227500, grad_norm=5.1370768547058105, loss=0.8735411167144775 +I0901 05:32:01.555264 139540719195904 logging_writer.py:48] [227600] global_step=227600, grad_norm=5.042890548706055, loss=0.930799126625061 +I0901 05:32:37.010277 139540727588608 logging_writer.py:48] [227700] global_step=227700, grad_norm=4.975520610809326, loss=0.8814067244529724 +I0901 05:33:12.783234 139540719195904 logging_writer.py:48] [227800] global_step=227800, grad_norm=5.240810394287109, loss=0.9263319373130798 +I0901 05:33:49.168888 139540727588608 logging_writer.py:48] [227900] global_step=227900, grad_norm=4.923221588134766, loss=0.854640007019043 +I0901 05:34:32.558911 139540719195904 logging_writer.py:48] [228000] global_step=228000, grad_norm=5.212977409362793, loss=0.9212753772735596 +I0901 05:35:15.227548 139540727588608 logging_writer.py:48] [228100] global_step=228100, grad_norm=4.910259246826172, loss=0.8728088140487671 +I0901 05:35:29.492725 139757377230016 spec.py:333] Evaluating on the training split. +I0901 05:35:38.593877 139757377230016 spec.py:346] Evaluating on the validation split. +I0901 05:36:04.713608 139757377230016 spec.py:363] Evaluating on the test split. +I0901 05:36:05.755409 139757377230016 submission_runner.py:516] Time since start: 74975.95s, Step: 228137, {'train/accuracy': Array(0.9572704, dtype=float32), 'train/loss': Array(0.15046822, dtype=float32), 'validation/accuracy': Array(0.75606, dtype=float32), 'validation/loss': Array(1.0797607, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62750006, dtype=float32), 'test/loss': Array(1.8712056, dtype=float32), 'test/num_examples': 10000, 'score': 73900.43143868446, 'total_duration': 74975.94643807411, 'accumulated_submission_time': 73900.43143868446, 'accumulated_eval_time': 1062.7228429317474, 'accumulated_logging_time': 8.600809574127197} +I0901 05:36:06.175903 139540719195904 logging_writer.py:48] [228137] accumulated_eval_time=1062.72, accumulated_logging_time=8.60081, accumulated_submission_time=73900.4, global_step=228137, preemption_count=0, score=73900.4, test/accuracy=0.627500057220459, test/loss=1.8712055683135986, test/num_examples=10000, total_duration=74975.9, train/accuracy=0.9572703838348389, train/loss=0.15046821534633636, validation/accuracy=0.756060004234314, validation/loss=1.0797606706619263, validation/num_examples=50000 +I0901 05:36:23.537736 139540727588608 logging_writer.py:48] [228200] global_step=228200, grad_norm=5.109195709228516, loss=0.847291111946106 +I0901 05:36:54.547669 139540719195904 logging_writer.py:48] [228300] global_step=228300, grad_norm=5.160841464996338, loss=0.8259518146514893 +I0901 05:37:28.083806 139540727588608 logging_writer.py:48] [228400] global_step=228400, grad_norm=5.538060188293457, loss=0.8842587471008301 +I0901 05:38:02.150082 139540719195904 logging_writer.py:48] [228500] global_step=228500, grad_norm=5.0425004959106445, loss=0.8998209238052368 +I0901 05:38:36.329100 139540727588608 logging_writer.py:48] [228600] global_step=228600, grad_norm=5.172792911529541, loss=0.8756396770477295 +I0901 05:39:09.262463 139540719195904 logging_writer.py:48] [228700] global_step=228700, grad_norm=5.025864601135254, loss=0.9205307364463806 +I0901 05:39:42.273328 139540727588608 logging_writer.py:48] [228800] global_step=228800, grad_norm=5.226393222808838, loss=0.8977769613265991 +I0901 05:40:16.196356 139540719195904 logging_writer.py:48] [228900] global_step=228900, grad_norm=5.263499736785889, loss=0.9219485521316528 +I0901 05:40:49.987523 139540727588608 logging_writer.py:48] [229000] global_step=229000, grad_norm=5.470120906829834, loss=1.0005717277526855 +I0901 05:41:22.865045 139540719195904 logging_writer.py:48] [229100] global_step=229100, grad_norm=5.219578266143799, loss=0.9156084060668945 +I0901 05:41:55.827646 139540727588608 logging_writer.py:48] [229200] global_step=229200, grad_norm=5.283169269561768, loss=0.8812978267669678 +I0901 05:42:28.773093 139540719195904 logging_writer.py:48] [229300] global_step=229300, grad_norm=5.058711051940918, loss=0.8824647665023804 +I0901 05:43:01.690493 139540727588608 logging_writer.py:48] [229400] global_step=229400, grad_norm=5.29608678817749, loss=0.9477274417877197 +I0901 05:43:34.977672 139540719195904 logging_writer.py:48] [229500] global_step=229500, grad_norm=4.923664569854736, loss=0.8533449172973633 +I0901 05:44:08.052167 139540727588608 logging_writer.py:48] [229600] global_step=229600, grad_norm=5.516074180603027, loss=0.9222011566162109 +I0901 05:44:41.203690 139540719195904 logging_writer.py:48] [229700] global_step=229700, grad_norm=5.270562648773193, loss=0.9427632093429565 +I0901 05:45:14.284592 139540727588608 logging_writer.py:48] [229800] global_step=229800, grad_norm=5.442511558532715, loss=0.887548565864563 +I0901 05:45:47.096210 139540719195904 logging_writer.py:48] [229900] global_step=229900, grad_norm=5.295755863189697, loss=0.9227164387702942 +I0901 05:46:20.525373 139540727588608 logging_writer.py:48] [230000] global_step=230000, grad_norm=5.127828598022461, loss=0.8649423122406006 +I0901 05:46:52.973694 139540719195904 logging_writer.py:48] [230100] global_step=230100, grad_norm=5.302691459655762, loss=0.8674868941307068 +I0901 05:47:26.785141 139540727588608 logging_writer.py:48] [230200] global_step=230200, grad_norm=5.28776216506958, loss=0.9079924821853638 +I0901 05:47:59.202728 139540719195904 logging_writer.py:48] [230300] global_step=230300, grad_norm=5.022533893585205, loss=0.8880480527877808 +I0901 05:48:32.246781 139540727588608 logging_writer.py:48] [230400] global_step=230400, grad_norm=5.0455403327941895, loss=0.8570905923843384 +I0901 05:49:04.677040 139540719195904 logging_writer.py:48] [230500] global_step=230500, grad_norm=5.372920513153076, loss=0.9184428453445435 +I0901 05:49:37.219638 139540727588608 logging_writer.py:48] [230600] global_step=230600, grad_norm=5.587920665740967, loss=0.9265112280845642 +I0901 05:50:09.771788 139540719195904 logging_writer.py:48] [230700] global_step=230700, grad_norm=5.236055374145508, loss=1.0084688663482666 +I0901 05:50:42.637523 139540727588608 logging_writer.py:48] [230800] global_step=230800, grad_norm=5.731800079345703, loss=0.9543076753616333 +I0901 05:51:14.937855 139540719195904 logging_writer.py:48] [230900] global_step=230900, grad_norm=4.967644214630127, loss=0.8564640879631042 +I0901 05:51:47.473142 139540727588608 logging_writer.py:48] [231000] global_step=231000, grad_norm=5.014541149139404, loss=0.8100097179412842 +I0901 05:52:19.966750 139540719195904 logging_writer.py:48] [231100] global_step=231100, grad_norm=5.310692310333252, loss=0.9450331926345825 +I0901 05:52:52.293031 139540727588608 logging_writer.py:48] [231200] global_step=231200, grad_norm=4.872323513031006, loss=0.8590764999389648 +I0901 05:53:25.377665 139540719195904 logging_writer.py:48] [231300] global_step=231300, grad_norm=4.994529724121094, loss=0.8368402123451233 +I0901 05:53:58.077226 139540727588608 logging_writer.py:48] [231400] global_step=231400, grad_norm=5.3622727394104, loss=0.9152543544769287 +I0901 05:54:31.302560 139540719195904 logging_writer.py:48] [231500] global_step=231500, grad_norm=5.396509647369385, loss=0.9337939620018005 +I0901 05:55:03.598842 139540727588608 logging_writer.py:48] [231600] global_step=231600, grad_norm=5.255952835083008, loss=0.8997480273246765 +I0901 05:55:36.456600 139540719195904 logging_writer.py:48] [231700] global_step=231700, grad_norm=4.880581855773926, loss=0.8657580614089966 +I0901 05:56:08.911623 139540727588608 logging_writer.py:48] [231800] global_step=231800, grad_norm=5.106797218322754, loss=0.8131548762321472 +I0901 05:56:41.464051 139540719195904 logging_writer.py:48] [231900] global_step=231900, grad_norm=5.183785915374756, loss=0.8894537687301636 +I0901 05:57:14.224547 139540727588608 logging_writer.py:48] [232000] global_step=232000, grad_norm=5.110013484954834, loss=0.8886953592300415 +I0901 05:57:46.540298 139540719195904 logging_writer.py:48] [232100] global_step=232100, grad_norm=5.21185302734375, loss=1.0014300346374512 +I0901 05:58:19.236147 139540727588608 logging_writer.py:48] [232200] global_step=232200, grad_norm=5.4166669845581055, loss=0.9251216650009155 +I0901 05:58:51.956782 139540719195904 logging_writer.py:48] [232300] global_step=232300, grad_norm=5.301342487335205, loss=0.8434670567512512 +I0901 05:59:24.676029 139540727588608 logging_writer.py:48] [232400] global_step=232400, grad_norm=5.119884967803955, loss=0.8767211437225342 +I0901 05:59:57.329455 139540719195904 logging_writer.py:48] [232500] global_step=232500, grad_norm=5.257816791534424, loss=0.8869935274124146 +I0901 06:00:29.763236 139540727588608 logging_writer.py:48] [232600] global_step=232600, grad_norm=5.246023178100586, loss=0.9517991542816162 +I0901 06:01:02.830261 139540719195904 logging_writer.py:48] [232700] global_step=232700, grad_norm=5.04565954208374, loss=0.8970433473587036 +I0901 06:01:35.693655 139540727588608 logging_writer.py:48] [232800] global_step=232800, grad_norm=5.205224514007568, loss=0.9111096858978271 +I0901 06:02:08.522520 139540719195904 logging_writer.py:48] [232900] global_step=232900, grad_norm=5.366690635681152, loss=0.8691924214363098 +I0901 06:02:41.539203 139540727588608 logging_writer.py:48] [233000] global_step=233000, grad_norm=5.2992329597473145, loss=0.8730559349060059 +I0901 06:03:14.028834 139540719195904 logging_writer.py:48] [233100] global_step=233100, grad_norm=5.417654037475586, loss=0.9946203231811523 +I0901 06:03:46.895428 139540727588608 logging_writer.py:48] [233200] global_step=233200, grad_norm=5.149653434753418, loss=0.9320715069770813 +I0901 06:04:19.554609 139540719195904 logging_writer.py:48] [233300] global_step=233300, grad_norm=5.062996864318848, loss=0.8981925249099731 +I0901 06:04:52.845618 139540719195904 logging_writer.py:48] [233400] global_step=233400, grad_norm=5.344503402709961, loss=0.9438152313232422 +I0901 06:05:24.722344 139540727588608 logging_writer.py:48] [233500] global_step=233500, grad_norm=5.376211643218994, loss=0.9524419903755188 +I0901 06:05:56.966510 139540719195904 logging_writer.py:48] [233600] global_step=233600, grad_norm=5.225225448608398, loss=0.8655856847763062 +I0901 06:06:30.136746 139540727588608 logging_writer.py:48] [233700] global_step=233700, grad_norm=5.041961193084717, loss=0.9015315771102905 +I0901 06:07:05.262347 139540719195904 logging_writer.py:48] [233800] global_step=233800, grad_norm=5.2122483253479, loss=0.8341056704521179 +I0901 06:07:40.827224 139540727588608 logging_writer.py:48] [233900] global_step=233900, grad_norm=5.126073837280273, loss=0.8901228904724121 +I0901 06:08:23.564607 139540719195904 logging_writer.py:48] [234000] global_step=234000, grad_norm=5.703407287597656, loss=1.023517370223999 +I0901 06:08:57.494192 139540727588608 logging_writer.py:48] [234100] global_step=234100, grad_norm=4.944810390472412, loss=0.8790634870529175 +I0901 06:09:21.753865 139540719195904 logging_writer.py:48] [234169] global_step=234169, preemption_count=0, score=75895.8 +I0901 06:09:22.575893 139757377230016 submission_runner.py:857] Final imagenet_resnet score: 75895.78674530983 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/eval_measurements.csv b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/eval_measurements.csv index b3ffbb3c5..0494767ad 100644 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/eval_measurements.csv +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/eval_measurements.csv @@ -1,39 +1,39 @@ accumulated_eval_time,accumulated_logging_time,accumulated_submission_time,global_step,preemption_count,score,test/accuracy,test/loss,test/num_examples,total_duration,train/accuracy,train/loss,validation/accuracy,validation/loss,validation/num_examples -83.23402738571167,0.0,62.0377836227417,1,0,62.0377836227417,0.0013000001,6.910612,10000,145.27190852165222,0.0010762117,6.9101987,0.0011,6.910505,50000 -118.17159128189088,0.0278799533843994,2058.049063682556,4518,0,2058.049063682556,0.029500002,6.248613,10000,2176.2921516895294,0.039580677,5.9223266,0.036059998,5.994765,50000 -183.0349597930908,0.0581820011138916,4054.1717386245728,9801,0,4054.1717386245728,0.007900001,7.6627584,10000,4237.352629899979,0.009745695,7.5413218,0.01064,7.568091,50000 -315.471892118454,0.1032261848449707,6050.326933860779,14506,0,6050.326933860779,0.0042000003,8.537536,10000,6366.033954143524,0.005859375,8.495525,0.0052799997,8.498083,50000 -405.4710731506348,0.136307954788208,8046.917476177216,19882,0,8046.917476177216,0.0032000002,8.626876,10000,8452.700548648834,0.0034877232,8.590018,0.00326,8.605198,50000 -484.79101037979126,0.1682186126708984,10042.881111860275,25270,0,10042.881111860275,0.0025000002,8.380475,10000,10528.060062170029,0.0020926339,8.377973,0.00232,8.369723,50000 -546.4171376228333,0.2008829116821289,12038.886108875276,30022,0,12038.886108875276,0.0019,8.119603,10000,12585.776431322098,0.0023317921,8.121338,0.00188,8.11584,50000 -613.6692273616791,0.2397420406341552,14034.850373506546,34981,0,14034.850373506546,0.002,8.15415,10000,14649.07592201233,0.0018534757,8.153365,0.00166,8.154742,50000 -694.5466666221619,0.2730488777160644,16031.259686946869,39688,0,16031.259686946869,0.0018000001,8.218177,10000,16726.439479112625,0.0020328443,8.241737,0.00196,8.220464,50000 -770.4708964824677,0.3067083358764648,18027.405270814896,44428,0,18027.405270814896,0.0015,8.245352,10000,18798.586426973343,0.001434949,8.262915,0.0018399999,8.247698,50000 -869.9464857578278,0.3443114757537842,20023.873562812805,49135,0,20023.873562812805,0.0013000001,8.255711,10000,20894.611929893494,0.0016342474,8.264104,0.0017599999,8.257346,50000 -950.4290370941162,0.3771963119506836,22019.93824338913,53896,0,22019.93824338913,0.0012,8.235645,10000,22971.235975027084,0.0017139668,8.253426,0.00174,8.235537,50000 -1019.7672593593596,0.4138531684875488,24016.297024965286,58654,0,24016.297024965286,0.0014000001,8.181261,10000,25037.01320934296,0.001694037,8.188425,0.0016999999,8.179385,50000 -1088.661203622818,0.4541494846343994,26012.28999519348,64006,0,26012.28999519348,0.0016000001,8.0927105,10000,27101.98451256752,0.0017936862,8.101958,0.00168,8.089027,50000 -1165.955510854721,0.5219497680664062,28008.3186814785,69149,0,28008.3186814785,0.0016000001,8.025281,10000,29175.41971540451,0.0015943877,8.021358,0.00172,8.019661,50000 -1237.9831821918488,0.5660214424133301,30004.49159431457,74083,0,30004.49159431457,0.0016000001,7.9134455,10000,31243.70892095565,0.0017338967,7.922397,0.0017799999,7.906208,50000 -1304.693614244461,0.6050024032592773,32000.55635213852,78998,0,32000.55635213852,0.0018000001,7.8329372,10000,33306.56731748581,0.0020129145,7.8298244,0.0019999999,7.8241224,50000 -1390.706758737564,0.6433432102203369,33997.315504550934,84001,0,33997.315504550934,0.0019,7.752259,10000,35389.42226982117,0.001953125,7.742617,0.00208,7.741793,50000 -1450.920217514038,0.6834356784820557,35993.23620557785,89270,0,35993.23620557785,0.0019,7.7178745,10000,37445.64071536064,0.0020527742,7.701437,0.00222,7.7050667,50000 -1505.3919687271118,0.7256448268890381,37989.50494408608,94561,0,37989.50494408608,0.0019,7.670427,10000,39496.46735692024,0.001813616,7.658918,0.0026399998,7.6551237,50000 -1589.862429857254,0.773186445236206,39985.52833747864,99468,0,39985.52833747864,0.0018000001,7.6621127,10000,41577.052481651306,0.0016342474,7.64568,0.00204,7.6454577,50000 -1611.5242595672607,0.8123252391815186,41981.95154380798,104689,0,41981.95154380798,0.0018000001,7.65344,10000,43595.22164225578,0.0016143175,7.6252575,0.0021199998,7.635452,50000 -1691.907369852066,0.8714404106140137,43978.46745610237,108937,0,43978.46745610237,0.0018000001,7.649656,10000,45672.22894263268,0.0017538265,7.6281743,0.00204,7.6303515,50000 -1816.6612305641167,1.4030964374542236,45974.4246840477,111895,0,45974.4246840477,0.0018000001,7.648998,10000,47793.9038836956,0.0021324935,7.6150737,0.00194,7.6296186,50000 -1847.798466205597,1.7638790607452393,47970.05360388756,114967,0,47970.05360388756,0.0017000001,7.6456165,10000,49821.3028011322,0.002072704,7.616824,0.00188,7.625837,50000 -1885.494007825852,2.183001518249512,49965.64317059517,117934,0,49965.64317059517,0.0019,7.6490345,10000,51855.24300432205,0.0018534757,7.623401,0.00182,7.628771,50000 -1914.1544938087463,2.561258316040039,51961.32361125946,120883,0,51961.32361125946,0.0021000002,7.644687,10000,53880.19545173645,0.0013153699,7.611876,0.0017799999,7.623189,50000 -1949.544005155564,2.9746978282928467,53956.820291996,123836,0,53956.820291996,0.0019,7.6381392,10000,55911.70957875252,0.0015744579,7.5977,0.00174,7.615778,50000 -1985.0605597496035,3.459739923477173,55952.328662633896,126748,0,55952.328662633896,0.0019,7.6265955,10000,57943.46052527428,0.002192283,7.5961895,0.0017799999,7.6027384,50000 -2012.879814863205,3.8696322441101074,57949.51488518715,129664,0,57949.51488518715,0.0019,7.61915,10000,59969.13365912437,0.0017338967,7.580449,0.00182,7.594372,50000 -2054.632348537445,4.288118839263916,59946.0051176548,132610,0,59946.0051176548,0.0018000001,7.605457,10000,62008.08043074608,0.001813616,7.5625515,0.0018,7.579831,50000 -2096.377536058426,4.6587910652160645,61943.662088871,135448,0,61943.662088871,0.0019,7.5847716,10000,64048.056831121445,0.0017737563,7.552586,0.00204,7.5581985,50000 -2123.342755794525,5.124191522598267,63938.85355734825,138357,0,63938.85355734825,0.002,7.5671916,10000,66071.15640091896,0.0021324935,7.5125346,0.0020599999,7.53964,50000 -2156.486003637314,5.550399541854858,65934.31674814224,141279,0,65934.31674814224,0.0019,7.5543923,10000,68100.40691971779,0.0019730548,7.5101995,0.00222,7.5270987,50000 -2195.9613015651703,6.055431127548218,67929.82773637772,144193,0,67929.82773637772,0.0019,7.5415044,10000,70136.11622142792,0.002351722,7.513111,0.00256,7.514347,50000 -2251.2275352478027,6.4613258838653564,69926.05036187172,147094,0,69926.05036187172,0.0021000002,7.5356674,10000,72188.20590472221,0.0018734055,7.492316,0.00244,7.5076427,50000 -2284.97248506546,6.899955034255981,71922.94839406013,149980,0,71922.94839406013,0.0018000001,7.536077,10000,74219.54028224945,0.002471301,7.497335,0.0025,7.5091887,50000 -2318.299463748932,7.3489696979522705,73918.37983560562,152933,0,73918.37983560562,0.002,7.536695,10000,76248.95145559311,0.0026705994,7.502706,0.00278,7.5088005,50000 +64.69926476478577,0.0,53.80131030082703,1,0,53.80131030082703,0.0013000001,6.910612,10000,118.50065970420836,0.0010762117,6.9101987,0.0011,6.910505,50000 +86.44905090332031,0.0397703647613525,2049.826003551483,6395,0,2049.826003551483,0.43960002,2.5965824,10000,2136.365298271179,0.6256776,1.5689651,0.56868,1.8507421,50000 +108.49990367889404,0.1349964141845703,4045.7496836185455,13740,0,4045.7496836185455,0.5343,2.1114912,10000,4154.4920818805695,0.75713485,0.9526457,0.66587996,1.383565,50000 +131.3270001411438,0.2016239166259765,6041.659068584442,21092,0,6041.659068584442,0.5665,1.966288,10000,6173.354122161865,0.80626196,0.7468492,0.69604,1.2423426,50000 +152.86800980567932,0.2448570728302002,8037.608729124069,28443,0,8037.608729124069,0.58580005,1.8870754,10000,8190.942818403244,0.82924104,0.63326895,0.71349996,1.1654677,50000 +174.9927954673767,0.3151042461395263,10033.702872037888,35796,0,10033.702872037888,0.6003,1.8511142,10000,10209.286883592606,0.8547313,0.5380873,0.72433996,1.1205891,50000 +196.35537314414967,0.4192686080932617,12029.607981204988,43147,0,12029.607981204988,0.6056,1.8409466,10000,12226.716282129288,0.871313,0.46765828,0.72818,1.1009811,50000 +217.0973782539368,0.5025744438171387,14025.61386179924,50502,0,14025.61386179924,0.61230004,1.8371395,10000,14243.602019548416,0.8854432,0.41532198,0.7328,1.0901387,50000 +237.31938290596008,0.5800423622131348,16021.685701847076,57855,0,16021.685701847076,0.609,1.840022,10000,16260.028512001038,0.89253825,0.3850234,0.73662,1.0837384,50000 +257.79344487190247,0.6483263969421387,18017.576412677765,65208,0,18017.576412677765,0.6104,1.8365499,10000,18276.52331066132,0.90124756,0.35447967,0.73748,1.0785924,50000 +276.8868010044098,0.7893733978271484,20013.38549757004,72561,0,20013.38549757004,0.6108,1.8346378,10000,20291.634605646133,0.90824294,0.32824045,0.74074,1.0726177,50000 +295.5284707546234,0.8865344524383545,22009.43017864228,79915,0,22009.43017864228,0.61490005,1.8260653,10000,22306.47315478325,0.9136041,0.3116816,0.74292,1.0638095,50000 +313.69797587394714,0.9892408847808838,24005.31016421318,87269,0,24005.31016421318,0.6208,1.8269247,10000,24320.700825214382,0.9151985,0.29835552,0.74513996,1.0610034,50000 +332.67340207099915,1.0685644149780271,26001.21587538719,94622,0,26001.21587538719,0.62310004,1.8336954,10000,26335.716311454773,0.92352915,0.27080315,0.74688,1.0591964,50000 +351.16429591178894,1.2089686393737793,27997.165731191635,101975,0,27997.165731191635,0.62320006,1.8342712,10000,28350.35700964928,0.92490435,0.26681402,0.74736,1.0602218,50000 +369.42036747932434,1.274552583694458,29993.1374783516,109331,0,29993.1374783516,0.6223,1.8452394,10000,30364.70661520958,0.9286312,0.24832252,0.7486,1.0602936,50000 +387.4781656265259,1.3403329849243164,31989.12622189521,116686,0,31989.12622189521,0.62310004,1.8545954,10000,32378.873791217804,0.93215877,0.23511541,0.74862,1.062171,50000 +405.35461807250977,1.407243728637695,33985.024151563644,124041,0,33985.024151563644,0.62140006,1.8547386,10000,34392.77033543587,0.9368423,0.22361873,0.75012,1.0643718,50000 +423.4676520824432,1.4830548763275146,35980.980189561844,131395,0,35980.980189561844,0.62280005,1.852586,10000,36406.970715522766,0.93939334,0.21290366,0.75088,1.0653294,50000 +442.1525115966797,1.554464340209961,37976.9183151722,138749,0,37976.9183151722,0.6233,1.8521956,10000,38421.72024726868,0.93939334,0.21290293,0.75162,1.0644616,50000 +461.0949785709381,1.613271951675415,39972.81428909302,146103,0,39972.81428909302,0.62600005,1.8532897,10000,40436.67267847061,0.9429807,0.20439605,0.7515,1.066416,50000 +479.6059491634369,1.682976484298706,41968.87982940674,153459,0,41968.87982940674,0.6261,1.8529103,10000,42451.374479055405,0.9436384,0.19484542,0.7524,1.0662144,50000 +499.2032101154328,1.7564749717712402,43964.87986493111,160803,0,43964.87986493111,0.62670004,1.859921,10000,44467.10048651695,0.94782364,0.18040912,0.75266,1.0689249,50000 +552.6016957759857,1.8376588821411133,45960.7723968029,167972,0,45960.7723968029,0.6251,1.8615223,10000,46516.528638362885,0.9487404,0.18130454,0.75286,1.0714159,50000 +574.862462759018,1.9537787437438965,47956.8501329422,172658,0,47956.8501329422,0.625,1.8682417,10000,48535.025040864944,0.94868064,0.17847335,0.75325996,1.0750405,50000 +597.1016075611115,2.3890323638916016,49952.47198367119,177063,0,49952.47198367119,0.62600005,1.8686966,10000,50553.51904273033,0.9486607,0.17960994,0.7531,1.074318,50000 +636.8429207801819,2.812356948852539,51947.90822029114,181600,0,51947.90822029114,0.62560004,1.8692299,10000,52589.320518255234,0.9502152,0.17258252,0.75378,1.0762789,50000 +678.5254654884338,3.305555582046509,53943.446397304535,186073,0,53943.446397304535,0.62570006,1.869183,10000,54627.26826930046,0.9526666,0.16909547,0.7543,1.0771687,50000 +698.0477120876312,3.8167998790740967,55938.98239707947,190646,0,55938.98239707947,0.628,1.866245,10000,56643.01434922218,0.9512117,0.16966121,0.75439996,1.0768428,50000 +740.9094491004944,4.328880071640015,57934.57347178459,194747,0,57934.57347178459,0.62960005,1.8696264,10000,58682.21131968498,0.95336413,0.16278976,0.75426,1.0784762,50000 +778.1086664199829,4.950573444366455,59930.2302134037,198524,0,59930.2302134037,0.62740004,1.8716532,10000,60715.90223026276,0.95336413,0.16411076,0.75376,1.0798633,50000 +808.0945663452148,5.439246416091919,61926.093482255936,202362,0,61926.093482255936,0.62560004,1.869789,10000,62742.43178105354,0.952826,0.16415659,0.75398,1.0797056,50000 +851.0713455677032,5.971511363983154,63922.77934002876,206423,0,63922.77934002876,0.62600005,1.8705279,10000,64782.85824751854,0.9546197,0.15765059,0.75356,1.0794744,50000 +887.9483625888824,6.512280464172363,65918.245708704,210203,0,65918.245708704,0.62740004,1.8746012,10000,66816.03911733627,0.9555166,0.15610996,0.75497997,1.0815759,50000 +926.6902208328248,6.9625163078308105,67914.09925222397,214153,0,67914.09925222397,0.6273,1.8760587,10000,68851.28239226341,0.95475924,0.15814532,0.755,1.0817531,50000 +966.0391397476196,7.543886661529541,69909.51558828354,218254,0,69909.51558828354,0.6268,1.8753902,10000,70886.89947223663,0.9552176,0.1577096,0.75531995,1.0797553,50000 +1026.6069040298462,8.02806806564331,71905.1672899723,222739,0,71905.1672899723,0.62960005,1.8752167,10000,72943.8052880764,0.9566725,0.15274695,0.75558,1.0793678,50000 +1062.7228429317474,8.600809574127197,73900.43143868446,228137,0,73900.43143868446,0.62750006,1.8712056,10000,74975.94643807411,0.9572704,0.15046822,0.75606,1.0797607,50000 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/events.out.tfevents.1778614855.ae408ec2d842.13.0.v2 b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/events.out.tfevents.1778614855.ae408ec2d842.13.0.v2 deleted file mode 100644 index 95c7f2cf5af84b234f804e5428e09c500192c3c2..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 360787 zcmbTGwOZwCu`EMb&P90=1R2l@>hU|uevX={qRH?-Q5}@jmE>K1?LhxsK1Vyb4}-tLbKh^x>qFrw%q`YqNCIdc?p6Y)h z#GgkNwy9uGeFs?44|B|NB}G;ZSw_UYYsfgdwr*WGAYQ&h%J~kJefQCEF}{DNzP~@n z3DX1LDjGc^8DhvN+$V~At=j#7$La4L}-gB;|a_nfwWuaYb8u z1VGkoBN<{y7DCL76gB~5VH4d72~x>{oU~#(GWn356F?reBjtR^BEJZPgrzSV36KCS zk|BnCMaaz3XC zshdkO#E>2c$?bLiC_uaxw`eFqqUy*(l6NwYekQ(!0J*n}l=C5Nzb!_HxpVAJfIRai z8DhvNgmj)AeGVW>s;yf}5Wl*zkXA)!a3gNiU;POn_0>o@A9B2>7G@lO8TkVs?>mwV zF~kBPHY%s~14ON)p@sx0<3J)Gr{YEov2ZX4h~67g&WH3q)(kWBLmSTm$k~r1LkwAt zkn>$jwE!~b^UH1$Bu7EEBP08#;YK7s^0@~P`9UTFZ(w>y_A*q(D5YnY=+zWs-*CrW4h}?FB zxXs#=1dy0yw_y?_L`fDh_C^YB#9{OHK>+bON6Pt-J2Rp%<8gBX9e{MdLNbI9@4W~K zp4+bpKz0s3Ya&5Db08ttPvJ&Po=|!TAa%S*IUnMwI~FqvwzT&LNW)VkLkN*Of)J(7 zds+cxP^&{`5~Qe}Y)2}lp2Uqfo%m)oKz2?i<$TDB_zjrRqu)J0fY@v%8A6EnX@ux{ zH1-0>$k;cQ5+p-e7P9R_GHyiY<5dFz;(mgZ^C8iTx?o1HMdiBzQgnu72qAJ=2+3Y& zb{!xCl3zJUkp1;#A#;qFj^xa>nGKM@JEWWsnboE(Lb7aQJpeMdl4J-W-uVbQ*W<Zw(mBuHLE*^bQeVUAjtETiK9()9u<=R?;0+=GxjUzgediM>iPgb=x}2)Ui*j!Ub` zQYBo1B&oRuV0z`Qk$q+&eYK?%)_?bKAE{Bc` z3i8`8LBblzLf&0V#EqD1H+}{{oOY0MK4hIud(1czaiAC=A$v)N5aO+fkalCHhXbTu zgP#c!q?!YH{pu)ggt4mYdVt(*PRjX^ud!n=W8m3t(*ZJ2on#0h84VG#(EiwWfYcqa zE>(gQH5Gs?9F4+hA08(vfO<_8s5U;uyAUpL*IUmyAwF5%x>$>*tQtiU zB(SM0Bxq6`Zp4$Xsh6Q636`Xs4|#Z|DP~Nn={6Q1_3cT95R##VkOW^ZJZf35Tv94Q zDmaiqXA+s$8A!KOq`cDA4<4iJy5Q9Dl=`!V=EkKka$NrEYDa~ae zZc}1#BPuLb?W}v1R+Fj6hekAI`;t}3-vx|NRSc^ z#9Hq#Zo~`g6Pp0Ceh?|=Lv}RZhnMlKw7TGV#CSN#5JC)05b`q4Xb3=R96Zonf@HOl z?MPHUa~`q3b+8#g<`k21KBTnL1V4=Ub-AygBT-Ec{@DZ}#Cr-t>K8mb0FXJYMh=i5 zajj(`dnYj+IX~GvC=qwXwt#;+r8PLkN+xKuAR9b4`GJs(Hdhf>d%K zc`cdqh*q-j;n`$4G_{;%$$RhTlik1IUsZZ!9H9T3cC2 zmk$RJk{-C#9Uys1q?`}YI5QA48s(?pQEP5vk|Bg-xFFR_2=N&U2Abn?%ay}#_d^~1M>=ZW*I#Oo=$q+*1mLp_wjdc7zvU+Zqrv&j+mxXu_ z--jC!W4k{VASKqMoDbRHr;i!kL$?kDh`l4p5JJ?w5n`QmIBxeFk)7WNG0c2@;|q3&~ir zA3sg+_op=gvaN!Y^C5X3j4&him z6rZJ#EZK3r>?k>0kF#5kdjae5a5|~~i+R2H+6oxt?p|zV{Ch*3`y&m zvXGJ~%%h>nC6;HQBi}!gaz12UM@59x+0q$Tsp3zPA%w_nM@ZX;H!o$RivB`QTw(QA)lLu+yRK&cakB5nD0hNt8)c-qxRLL+)4?OrzP8w zcpYXRsjM@&BXp$h98%7Qs0T1lE*Fm*G6f*{mLx+6@!pG&2aC@ZLPr|x{qbFbBz2L6 zIM>>ZpXQghH_8Cw6HChZkgPX$n6de4tu6p*bDU%dAyEeq;(MYHzhKAgGO7FY*B$3D zZCS|2`^?U6P`wkcx!WEh<$TE3=6wd)L zQ@`%Ojp)3+L?ye0GJ$F*hhlC$s9#Of}x!VvRt-VQx5MpowA!i1+Q2jxA;dfdAt706cL1_Dd5LvKl^7f~4rmLO!nu$BpoJ z>b(viuN+7@A5tr(HfFr6V~k@8wU&_#Atd7*LT;X_{|+Eig0`DUkO)0lNN(PC+z2nF zOuYKd3nS%x$S$*&c=a1GtkVMM$k`~8A%qlXBE<8Wa|3|vXpp@?f_V3og{(>0g^2#7V=~R(~)@= z=PLl>*pQU-A>RG%u_Jz4^Y#Ped~=c^gm~v8q*oJboZGAab55uPspLQk)R`0FrYdbK zfE-eo`=^`_aZL|FNbv2Pj{w=GLNbJqsJjR$xO#CmbfmLgP?Q8I=quZicbyqX7hUgm z0NJvXl=C4PJKYhoeYtldfDBkkGK7$fLWFdw9%lm({a$wuOOUjFvXIK#%s%qIvvnpw zCg_uLKIFmmDBMTpIGETs0yZTM;S$~%ek z7VF!Pay}#{Y(8dqb$Wr1mnLhH3?W4BF+$AaZt4L<+jVl91o0an3t9ezxkNd(vF+MWREzrQ?Jg5(U8?a0kl%s0fUm3#3U zqE9?2=R@QUZbXQ2%E6V;k-aBKh7e-@5+Mugh7SeEo9q4VNsz=rvXCZ@%s#R@f5LHC zTB-L)IUiDg9^(uh8-$62jkc5$n~PaU(v&ZN!Ou`NyQ3 z54m%wJ!Z@}{{yFuzm<^;A*A>tLSDS~vV@M<>l^-%AQ?ksAtfW2rPZwG%#X0NE{q`M zd`QYG=3!A{)svynkx?cjLkN-kiV*D=%jy6mCwY6_l3%w>_YakYjQPSq`l!9fi$>3L zq?`{~wQLi1L?dSr{?;xfn`8(f>faG^;#r&R0BO8GS4D#O8^}V6qM4;t8qvH9bi_8E zl=C6SRSx4N%9%Scc+2$o36dd($l=>;sx+pr2T0$ga;+ptIR}#YgMk?PL_G$GwHGPp zLyFcKV@C!qYpDfGD|H>o5JD1bAjJIs)ChpYbhFTqAbFfORONYZ!Hrn>Ai@S9=lhUy zKBS#dcg%QvauAM9>kJ_oLWn^vgxJU(y#|mKM%%kdkfdRQR@{jB&HLd@lD-Kk z=R;z*^~a2^_xIvm6uD_6LkP*MjS!1n-#Y-rEV87J1PL213rW=o!i_kqr92KgQhA7! z^C2_*nD-iNU1fwfAD$c`8A6D89fUX>T!dG@txxwEDnY6_kcNwyjvR8jw2&t%x5T^%qgPLmiBd+$z0*0<0rFu`e@h7xI8qjpe<>7~#r@FR_@lu#BS<+P zV&Tu6i<(^gP!k|u#*qvmM7IG#4tCRt1IV0yzV;HNf&-am!raQay(!!PI?~3Fl=C5R zyO~!nh4uBvL3Y*ABtrbHLv(IMGK3J*MhG$5{0$*B)9QOmkd)D~kc0uudF0GQYy8c2%ZsF(59y&c z8keWtsdGJHX(?PI8A3>LV}$5zw!$G@|86q^BuK;$ z)l`HGN!@!1APKigh7cmx3?T^}mc9o_ZnMIj5~PF!DUWBCR?mn>_%@an?MXQwvhQ9u zglrmMj{^f^x{wSZB)B<3DgtgzM#zJf2P8<=IN6T4FJ`WOTULhRD4=pFDd$7n!j%nS z@^`RL@gV5PgEu5Y2vKi|5Z}{JjbYw)rr|%lX4~O~h&f`gt zVE)@@(mQO<^xbVn!2Eah6cX}x^WRpJyW=K(30YJZ=D%I%khH&=|GuX49=GeTWm^rH z|K4UvlK!Xp@51IEQepnPu^mYi&wtl&tbxXY)9vurHg{Y|j~UPKd^ z|5o=T8UJ$r`}+vx`Y`|9+lOSy&VQ%RUx{DzyT7eE4ePN+0IC0r`R|*H{4nVHmsxFL z{(ET<3Hqn=-!hw*uCv7w=D#ySN$TIve}7Quhp#z&ez)Kq%zyhwlDvO3|7~FEfWLZ3 zK46>)$Az;8NTl$%pvfE;mXzHug8A>QaU|v6&3~tV8~y-}l?S~VoRbbQapPr=l>yh7 z%S?~3Ckx^1Rpw91`Nzs58mn=IZ?>^Xg3shzFv$>pCW)=^SZN%z*c~8>eOKp5kYE#8 zNQbt}*~_ATo+Utz3?=1!$m1a9SFk>4LkDSU^5)j)x2%?Sb`KxknPCQ zQ0Bs@vUPegKpY;Day}&b)MLCY*fQbA5`auDB^g49c{_x-pRRZT9a(kg$SVnwW-1H0 z7{WXn%35H8cM579CFOib@}{o{sa-R>BS7k$A{jzRg*rl7R(Y-kNRymbRT3m>qAX;_ zHs%#`x>F-?aAeYLQqG5rlVc7Tj&tj@07&NtBtr=C?tqZabxiaD^5)1~xo5v_o%>Ca zg_xK#S9m8Jo29_g+I5PQ^C5967x1Xna^2vU06Bk-WC$TDng}VouZEA8LKa6UNsuxQ zB)1Opy;NaIN9@QoS5nT0+|Nis$jA2j_;$CL6(mCliRy@u@A+-hp(F2KRyUC#Ig@2O zGRKY?Q;^pgasfKhu8NfNAy3`v;5B!2XfA%8dsIy_gb>}%2nih2<}^Uer`xrYAc<3C zA^H25oxQKkgb@JQXimxnknzlWlGd%%#8+_n+K>z(Btr`!dNs^70HQPRd>084GF29` zun)7dFUgOY1CT{Fq?`{KdXJeBo1VN8-*3^>iDU>NrrHSce|{nt)u4@39^5NEF@Zo zxkQ=ZEQha7Ipjgg`4C%OW=!GH)I@xid~7wz5JG}=5mK+tqEvurd@7qOLHuXRLQY08 zN3Gh;=WK+IoRWL-r<@O2P^gDXtCq!xAJ7py1(G3zsP{z3>kf+90MRj=y-0$Tb08)< z%$dFU?KAlEn~7scIUh3l%?E@i+I1WP5UWWfLkLOig^)crhc5@n*8VxJ5+u(|wj;+> znT{N5tXcvc(Hlm}`H+1s%o`B%-pp7F5dSeGLkKbGgODgQOH+UpKGye=AW5@iA;Djn zp9S338g?BZZC;UbKBP|y^IN;y3(x8UWW@)PA%x`hMTp^%ZuoTHzSXt>2@*D27Sh3u zxpJ+Z)E)=fRWwOCA2KwPd9zQGW%HW@k4v<1kCizGt`@g9Vbe(|qk0dm^-Y_tT)oGS~_-pZUu zZl0_A4m#p&O3L|=z+ICO(pzWY8i1^yK{A98mB9!Zd;INjfYeEykR(B(=gC5HtQp9Q z4VMc6QhI@u^C9yi$05XORs#O5mGxIih7b}p1R?ed_S6N)gjyMAB}kyTEM#A~KW@aE z({Gjn2~xpgBm1WNY-dE3mYTJCY0`B*PFPwZ2clzgiG~!1k5|$+wX0NKQTGJTh#S zE8a!1JVMI(kO>Ao5c00HttmhjCzA{z#B?}9hIX4!1RcruY5GWlq|BFv+*r@-Bg=N) zGlq`z3MA!x$O6R?2q|0_)d3)}TSm<_ATA?seUKpDoIgF5u#;I@tvfHl z*PY%sCFOjG-fd=XZ{#qQW&pW3gJcLHa-$HktHEa#fP~l>|BxUh9LS5I%+7wMyip=7 zt?tgGoDWg$+8H~tqI%PCfZX;V8A3?#XoMX9vN;bR?W}hyJpXlrF3U=GY3=M2fE)36 zu{R!mcRG=BKIF}vWtef=Pt^+`W8F!H5TZU7A&;N+$G5xH`BYm~g2XM9g_I_&!Ho#- z`C}S%WT%|`pK?A#&8`(@yfuD@^A=eOBtr;E9ET8#PhST>N3>(y)Feo-wJfArvvs%; zE3?1Z17uhdDd$6+4lvhi{ogugK}SZXkqjZkzyu-R_QW>>NK^f5oh3*m2U32GdEZ`x zzyutw@*Y9T`H=C)4DohQ`{oWIHnd7PF63 zd>%0jI$}M7l=C60YcW49@Z8fDe{0uZJjoD3%qJqmF)X<)bfl$Q?qCU$W-ANvv1TA` z%9m6D#LtVA3m`+88{u(%c6^4V)oUHe5JD;@A;k8QHU2#!y<;Y$B}mjFSx9bu=DyZ| zA@OnmshLK~`H<~{PT_twwt@R%fb_^D8A6En6of<+&e{zf+4b?{WC`N8SQcWsn>iu& z>Z!F8AfZ1=IUkbZu>c|aqWgsd#IA1gpG^=#RHh;1VN-v6K9awB^gIbt#(`8TF-yz8 z)5&6hl&>e{d`Op@%v}_Zg?HNmmrZgQENm)k|BiX&P2%3+TODPqNP>SRe~hi%R&q^nJ-qo zy!3EqPt_yke2Ch)bX;0Z{ZHTzYgY{*8A3>g8A2v~i+T)@l5Xy+BuI#ZEX3z2(~%YG zcdo$Fs^6EC^C69AI3T34)wENvw7v}{8A6EZY=ndynK}X>7hNyuLo%)uBV=8t=lJTS!)r-~5K=q`A)yL4<^p7Eo32|WNRgv#N9tT=Ix^jH zMGJtu=}5}?kZZwP@yzbDRv&*OQKUmMgb=5B2x;`?U~7OlXKvpmK{A|VA)ZOhr0{a% z`r809IhU04A@v3_zmXUf@Bm*xJM9k15JKcE5YnsrtoZXj0CHO!~%rr*T-Dhu_&_6G?^;5l5*!vlJPGmDd$eQ@)Ra17nqYQ*-6UC#%FK<(!;k`8mz|=)};O~ zCMk_8E@99y&-vjnNoiqEg8u0wrOf8Ns~5QvCMlb`lGMMSq_j@kg_D%^vy~%Zl2UaA z$@^!Mls30;lCmu0$qhJnzVjuK!s9|Tb6gmI^oTl4QUUD5BR zbcpeHmOWNJT+dt>d2XzG5mtEl8dA%|Lq|+vnA^c3#E%8{{+)oqV z0dprl{JsP!=RgvgFbC|mY5t=DvN4&I^C9Vdrs4HOn^y7<0b-C&GK7#sD}-E~QJn}K z8NTUSi3G`Wk?n|$J##&A@V2i3K*GaFIUkas$&4wC8F>rew|6y)WC$S!)(ClgY;_}m zbhz03jRZ+@m4!6D&D^_*eXfpg9BFc$l=C6eijH7M)b9BxLq|s6B^g3Uo()2pe{6;~ zYA3E;_(_6)`i=W^zIX_9U&e7#VmKPl%!9<_UlkjuF)vjMU*m}Cec6^jva$1N}% zApT=7D@u@JH`$I9_GcgtZu9Wxyg^e*IUnLTs54#-ZJC~pw*_a-AsIr5w>?5MmaW(U zkSAt6R3%8}GFiyD$;?r!`S32=VQJ-Cl5#%eM`cs&$hPRmILdS1j${ZSDvk(gdZ68B zfW(ZlZ6iUV-DM&5e=v~Dowj)E+-(vm=R?LgE<(sax6?QhYH3C?gpepFgbW@Z`3xYZ zVvl!}Ab}pTkU5Q*o&Dsfv7MkJbx)FVK4g-rBSI!@Y>z*%?tGSH2qC)82-$woZ3RFc zUa8qbf>dxId2N|}>wKB7b2FoE;vMH0BT~+XT-fD;9m(%-v@bx;jw2aDi0M*m)PC@f z$r8kSr7XnQota6>>UgmSK&*{PIUjNW?q@<^32&2AejmzLkJ1>M996bEoQ*dI%S_> zEkUwY$#z6xDg&|IeG&h5O|505oDVq>yBIq%HdkvmKt8S_8A6EqN`#E8>bnsj@_Lm{ z5+u$`781V94>#h#pkjQNSl>aUoDZp&YlRtSOG5C^%QhTNGK7%CRR~$Vu&5kv8rU7v zUoJs{y=%xy%*QZ$yNk*TRk-)mJdLX7Qz}d^A>~0$%RT*m|52OLvMVzcdu37~LW{aL zS{7@JN(W2RC!2gFOC`^;r3Pbpq3h@i7Ry&uJ>OzhGYKv4N1fULmiZs35TT`vu_P@q zegu{Qd1nJ9OTlVRUv_V1R#@uDsXf55@E%pqw|ra8OfGwcK9~)bGmohdp(SxO_GOy4 z{s*u;3#-~DS<-wsma>h^Ro6|2hgZPjxtFTvTUH%$#=dBGh?xl%?N};AXmR&Ji)O{; z$zZYgI(fHbiSp%Glp8a*RJ>g#)CNm-ttq6QZ?WpkTtXh0|M?17G?l3kp~b)#Ek6#9 zvjU5&vTw9x@$=(YESfNfC$NMyqw4vV%z4bC%>C;;_ku;a9Tg(9DEOhJ+;;I$ zuqZ`lAC)X+JWHkzW0|x;Q3Wig4pH@di_M04xW?MtJ#+~yT@tAfp(T&8oO*fA2P{1Y zv^XtUa@KJAa?p&KpeQyQ9RQZtVN^ZeQvIBH{l4zFtU+MeHHHciTEf;~U+l84+Jj~F z^3j=+C2=js(xQ?%js@gT(gKUFA63t{=rl9KzI+It;S82XfmDdlV!jqFp@Rb2g5`0A z_ch59vW{aZaAVfk#R#W)U}^T1s^?ptDKW2D)O|7|4lJ)~t|l8Hv}mnEOB0P^Rj^cj zEW0CFKJzR-cNxp#YYNU_QLFKS)blMbA2Pq8=vTTD_xqz|o% zB}%yIR*B`jR2vnrwv7lH!kjnQ?zyXRwq9 zn7oiI837#2)->k(@M=4UsbI0$O4ai%bFVm|WkJEwWU%bnMTH10-T`PaZMv}wSbWqH z-%FPL>p7O2%b31+8dYV0#kw<9&$nDK^+C(FyGJ^LMYjhPBD9QJkCu%~o0x;;<Xnde9PgQ8rYXB=UN_wzBrtxLWGtA#*#nQUI8q@CSDCB zOI{$SFUPwvw010TeWI7SY~NZ^?b{hkIWcRTdxn@z;d}O6(Y3gZbHkZZ;^grd4JPL zTe4L1EO*W@*JE4y7d!!r^8>1$Z)q~i8uu524&k`JJb6Zi2rZu(i<5TN6tH+FyY-SR z#X+3DL|tIc`-QFU+k&O*IjWv-XozYT8Z7hIQ1yJvj3&(WSh{My5m;7jq(X!i|6sIa*uPv2mXncnjU`L; zW{&0M5azs}u;fukurxYM)$=WFpV?qvW~6@a1D4}QsSu&XbTe9-K5iNWmg>v_6C_LE z7LKK6A=8)3rfqM5MdvzI&$k5VDx>9x^~LjGF}zEK2rVtPpyj)1=zOp=YV9~fvQ+Ra zNxc}$xpS@YJF=@LRnNBst0ID~yE>(Z?PEG}cI5TV5>1T90lW#0hH1dV>Kk|kmr$6|Mr z>C2+!mV3bx*o~^^TePPgL(8QjBS(W}e{U*8Xz9NVEo+W0NkU6*_)5v*y`5uGXv-YO z=HxoP0L$XLR6XDF#5W!-)pfSTgGK%k6(Y3AZAXh?)COm;bhW#_RC3pwNa#EMED1Eqf z6#9~2P1W-)g)L5?O0UH#MtOF=lNFN18E)7-$!F{i=e zww|izTU75b*Ooq7zR}PZ^DR_}(2^LAeHr()tOpEZFFx7jmSl<9S>vxJ%|~ys#F1y^ zdTK4<;b~b0l_VH7Pj*tougiI@495)p{kU(Um79g)wJa*-?`F+U9hiz6m219cHO$K% z&!gi0YS_Faf!VV5DJPo2Oz_FORMh`8ZN6r^fdNcaM?az>#pC8Jk|!ebm~y}oSW#U{ zsUT@Z$;_J{sjGw^ecZF~frEd4^yZ`Al`Fn^O@;i+f%85?bF%e*hwbw9eW7^0k_wZZ zIG?`k8xD%QXhki7g;~Fv3iykW^J|T3;}Xvw(S7R3-||9fAJ>-Vpft z$C_hP>%viDRSPOpc$Ap36F0>#MI*_57n}kPq|jMT=EON5JQ@=Pnf@ z{9N3b&t(P;KivP4C2EZ=4_-vBLVuEuYGsyb9X-?G$dIG)vV%`cdMrEMQ7 zL})RHz#eOyN62-9$NMkJ1TOPqDSXKwAdcNge=?wfpj_ht$4=mRb zs1Tt=D+(iL%RmCUzU@8g|#(mb0IzdcI}cbLNi8(F!L#j&1!( zg$OO)d(l!AXMxj%OPoFQB+LGN980w#(-*git2Du4vVyATTOLL*6GZ_D%bGx6TKiEU zLd&RqXsPjO^C4Jc&US?ZC5!)lj-{>!^KG_8_dRIo=tkA^EkWt$aG{PizGVuQ_baIo zp+#jsT1wvdT7t!_M?E9SQqHr)c`$1%{PM>f=*z?YR6XDFs(xcU@7v8Rivmle0Tm*& z6fl; q2!`8LaBykyBc!0AgsGv-cK<2n7p!7|v2s^?pbeY;{`(siC`f~Au^6(Y1m z9l*Y9HhGOt;-^_gO_MB12RW9ZJD4?A75K6iSSszQdcI|N6l7yTqlbPUJxVL5XgYg9MpI#@pcpz8S+!!8Yn!My|Pu}^k@0@IiAm!~F!MbDF}=Udu-S&I8h!0_46!IJ7jg$OMz zV$rhISopbL*dEI%X_CeJ2*(n2g1P1|S=Sk##*K5J>iL!@<;;sU7wXJg4i+^xDnw|JJA#&j zdOkQEWj{UnqGT!ISw7Ea?kP>zc&iU?N)*qe04HC8e8*?O=nYfOa*E$YY6 zazFRYNU*GZp7K(%RPrnrY?$Q{JVDnFEU({E^?XZR_HWS4+&pC>SQdVwLWGtw#&UOa zULsi1@4WmVSqhGG`r+#@^YYjIM5y@$UZ{C0&35n2>Z zqGj4Xm(gHR>DaiXWGUlWJXDzLWcB*r#(>432UX9v6m_#hOTUlt*P$;n`cWZ5OCDp< z_Hy$EON&k;J4lwCWKLi7qE_N&C|8du1j~3`s-ABN$#Tb#l$kx+fn|1oDnw`rOUAw| zAFGb{#$)q5yGoYC6pqDvJM)Y$w%LR*u(T_r>iL#j*YS9MxpR9+Q?LwuLWKw|<|$}- zvHMXI=*!|pdA%h|$SICxxCgWEx9Fhs6fE^yQ}uky?&XHqmt~{#Pk_Zqg9;H^v`(RA zlwIB>u-tJ~9U@si^DOUfF_z}52jDY{gR7`|zNILU`Qk9^=gMASab8P>2rU(i&@w6wEwjoF zxItfZ&nPdDEdFOWmgW7KWAce4&ADJXcZsU!Tk4Nun!6=oWguAQU8h2X7L_w-$?5-Q z1z6(xEwz&@hK|Mk#rFO<1+b*Rqr*>OzGGEjH9pj~4fsbq`^U zwHQ7iUa~}A;8FtL})2yEUxYSKf-`> z^<}w2$#U}IUryd{(Nx6g?`8p>_}0jP8&r~D@?J-o8Ht^ol=>Pb@AurJQvPo8Ub9{c z+^C<&vOdD({m{o$++R)JFL)`BS2)k&RdZqTKD&&H`kyB6m$iw050m#M6;z~n^1j|( z4P^H5X&D83`kqfzkhG#?ChzT=F(3U&hptvIc_00Q3i+3l_hy=!%V6?8TfPQaSF)4$ zYhSg&oospZx&T<1rIxi(l$|a6PUzPdtRy)_=Jy?#vr0V$=<%Z1L z(o?tl$U%=!zNbQj7N<+tkt(pwDiA>mcei4{{%~J@BKd{i+2{sqGHEP72L?Uq6n4=L#TSb<<@&k zTw~8RHOK_ZgAr7S&?1+Gmf3;1-@&qJ$Zf?}zn(6X@GP!}nZHAOUp`R{EQ?1`^?Zw~ z8FO{Mx-lcf@ZR9bZxPe9NRQ z%t&JN!Y{kQa`inGBDBo7f|kJ@_NjrTMdEfX$r5~(V|ja%IgTAQS6KyXY{x08o^RQE zTMaL;3L<;zfMwKqDnw{ezls)Pk2Uz7#h~b_o|2`KXVF{8d~2#_5Sa;<<437_zGYPH zSlsuOCsfr0%Z^i2h|p5TSn9v8u7bYo@6&m(WGTqu^kt4cV=?^LbuCyX45jM%mXMPd z(2|_q0pA}pU?deHv?S(WU+nTLamq7u-t^IuCG8r=GHM_5MRfPN5%}lV&TUjR4-5n2?kqh;Zvb2xq5GpLn~WGUlW9(H8*m%+Efx`1Wz4yvASIitD< z*Vz0yO>v^=>K-aYXvt$NxBINV3w`N()!SLJO)NxZ?aT-wS^u9!?25CE2j$y7bx;_-nw zCI=6jlmV7o=~Rf&VtxZH{%hlK-+vbRY^`JoxyiBIsbcPy)|z%V0W8h-QuTaG`v=U+ z8x9A2>kF3Z7%D_)(YlEi%UnmCDAEa?xK*-z=2gZgi~k*t#n_HHza)f~zXi*w7^1)v* zuan0dm>dUd%;OjpBDAR7L5ubFj1jQEBv>b3mMrBwOKMkUf2sUwgx|%6E}`oAmYY$` zuhXZzII;pP^E{{!p{0Pa3|pDd7%bP!I)x}-5f#ptLs-AC|@}2oW z4piqofHl^BC>0{KWE5gw(!wr{2TNIFbNTXL_s25tb1aIlnHMqG*7q?2%c?e1J>RnS zZ4CCsZRG>}m0Gwa6(Y3w-$%>go-z3A^t3uD4JAu-5yukTmO1Z#``V*ESPrOA^?b{X zT}RQftcIB`^u?hi6(Y2l7NI3q{@MrV%d02#T1l3`2ONuQDD%DY@)#duu-q=E>iL#e zDQ0N#9N4uxSUfAK5TT{T1GMa}6R;mFgMJ!mN|p+qMJ1j&zYO!gSPGVx3L{88-(r4a zGFn>fuY3=d7!@i+Xenka)l(gJf+hb}WOvDu|B%y{o6*bgco;Ecsegh|uEn2rWZR7vL|5CK>fKk}MI$9E;;&W{rJ09exTd72~OTzNPIG zX8dJYgP*IxQhOQ|BDC}`MvKOpsW`c^zvUVe$>RN(W9b{k+`EW>F#&&db6bO|=Uak4 zUc>!mhw+^?U~%b6g$ONjkI_13*&Zy9nR6#JrM zbL9*4#bXu~BD5qv!@evoJDLlYZV{oOk|pgq$MR?lb8V^c#w-LZR}N71e9L`{1Uy%M z2u{aeL)phuAwrA$bF`FZ^wk1Ok3)C%NS3HljwL*t`GsoIx#$?MyiBC(`4)pc%p0l3 z<($L6+jA_13K3ciO3`Ad_wE>2roHPOCt3W;IF=x5=6dX&;R&21j{i#4^DRT3Gk>p7 zVcxiKSYz#L^&uM}v?!FJrFr;`xnOC%D$;DJzC1lh)$=W(&nj?# z88Wm0&o8@=P$5D~9%GTOJhl`p)~i~dlPozeIDK*1#_aoUgS%V?%fMq)J>PP^JP7xf ziql2-w>p=UP$5D~*bD5-kdzGkwXNOIHdiG};!BQY+zDoRJgXSo7A#*zQT2RFiURW; z`PzYgKVXejm{K7^i}_2mm^JTE3kCyg*K#b9ESLUc`TxN{XA4`09u^B0EOoY6;PKyg zxPHrHhC|c)^iGC%MQhckk^~d0a_?$ji~8JYvIQnqyEdg#{%&G*>Y|SW;e9tf?Z$S1 ziPa@)RNP-pta=tQTef*e$5@zH-J(fF{ZA9CdPjG-!o+HwZd9arV%1LZ3o<)ZCgJ4U zqh3^yw4!7tR-e6KfTlL>3}9mQ+(0ViUrwxc{=VibOsr-YQem;E+WL?YCTSnT-qfDrze>$-$vw@pRPL6?z)wk2A*uS4xJr=HD1Z(x6 zUe_ZqvD$Dh75I-PRt=KIShcXDV*cI4YW$RQ z{oqWHc)94Abi@gH#W{NFx-fU-XPusO5RRU9H>i64(X)>`^A`{XTsk@oj-DF#s1V`j zqV)=wR+z6vDOlWJH>;2=pLrIwJIvL_=1aHmuLcx`t^NAfUF?k49Lo?J=G*N4mcs_Y8e1Pv)$=X6E0>_f-RH#(V>;?tW`BSdKoR>iL!lV_xGLYi97a23XQn zjL1d^Ed`9l{7yjzSoUAq)LpXVRdD)pa~@-PX56GDSakBJdcGySEAx)k7Y~vQ!E&~M z3K3eODzGovu^K60IoADIf60>cj$`ppV(yrfo#>PZmZANrdcI|@%{lB#=iL%Z9Ohk~T3*xI!W#3`p+bZf z(<-#g+~C;}EH!!`aFZ;7A32u7hYq;Zx9mvADbIm}sd~O8`Sbz|v50$(?~MFtM1=?~ zEk2@U@p==ylXa;5*VU4xf@kSJiMh61FK4?4`jV$j)$=Vqmoay-=cT!sL0?Yvq(X$2 zV#YG#S@BxvOG(onfs!Tv6Q?gm%FJfyg$3X*h-Rx(^?b{9|9jY%uKv+@7u!jT3K3eO zKVe@CX5Zcpmf2bU+a*iNXO3lzB6B_F)TWja^yR?~s-AD@JhD6fg6PzJdjsf;-#scs zXmR?CmVKgmZBcab+Vt)QGBOY?evQNR@}1mm!H6LU6l$E zTI9Z>#bZM$PT<|VJpY7bDdAZbWigX0)lVZ{!5WLbLDlmuuZ<62UsU35Z3D~6dsK+f zlEqj?y!OT8*pkCZ>5?Vu8>cTH>oDiab2Xovg5}#Ws-AC|9iM@gut+QX#odlnDnw`r z{)T;#J2w@tljHqgWlNU0YK~=E7W3daILs5@vD#$|RnND~xy<~@)E)I&+=jl?45vbb zmKoJ(d9GIn-?et^%*cGn68xQG8Pt+FCU0DN0^dhI=rUE$x0D@WZcN=cn6V5j4|1sx zp+)^WT1pOHYy|th!QQP8B}*mG(xl-cTpr^}x_*N-rVvNf^DPaUIbq1#!oheOENMSEmQKOU?I7>R4e*zcYh$Q-z9l4-xe%&5{ww}Hr54Ai z5TV8WCt4zkvf98J>)O=ryJU&-t||N0fI+^@`Nh`azB2SBsV!B{w-g$h;Q2-SR(HIo z6yAvn5n9ycM#61#-#qzoU{QV+TldYcdwqVZIhJ@M=9+)#hDrFV&YU+?J>Rlkbt|qh zhf97qvAX9Y6(Y3wF_zE)5je5x^{rN8$x_C%xQ=8j4H^s_0)43^Uz^nPEiG&hq2ytUbYU=P^~!w^-^K z-*;fS_JRr#TE^ADHFopnU3IXuZN00rWJ&bpSbDx^P7PjbeC5EhSCgveTQ*Huf%{8> zfx#-Uq;{i1gqB3c5+A$F04zTmRO(5V5I>IPC5z`y2hYNuxwEnOzQcT%zc;9vhj7t zd01PG*}slsY14uEj@%);ElzuMPN(YmmM0hA zqD4JE(hd63@-h`7wCrarGjufY-RfSS)|yKee}9hUeHC+^+;`y@oWM)0v5wU9E%(DB zaDSO@(k>M&3lymkp+%uK_N9G|%g3QFE#;otN|th-W!5)lTD{=d&iSy$HY!r}e9N1s zd(dKM`vQ++#~M)~LW><^DQ%-41eWnVdbvoJyZ}yLUZya&=o>V=guhgY9zfOeE%uen zgtyA$B!B43UPCHGXvt$NUA49Gyx&T9$qLDmw4P&0eZt&_vHens6L_lzQ1yJvL6zFL zzg+NnnF5voLn=gQ=`D|IY;NLx4d_d+3%P3~OV|dErP7u;CTHKjxfRw}t!q?0-|`|P z3fEZa>o8-m9LuLdgqASIa=-ZWSFkjh+b&45RP!vRJDKzThTKhf-!j#js^?o4XE9$K zj6FxiL!d_jaS@ z>YUM~u)mCRqC$igt-9D3)$iGO5BBJU5lNCIFoC5{ZW`F6CbrpZ5CO3$x=Ua{~nTLJx zxaNr0{5iv@5TRv?0EKWXL4aMeXIS9O8L9#+gTy^aHIB3OY?{6+b!i(++R)K-tu6!?1$sI6EJ;y z-~$!)KTY4>+t6@5OyA0XqawxAw~mc#;rn8p95&+}q@^{7kuFLrN@n_Y;W7r;Wwdn~ zOy7=Cq(c7X^zFLt*DPWBcCHE)COds=J$4q}Al$V7P$VqOw#})4znH!?DY3+)56a2K zFnwEJjY|5b)3-7k_{Hg&A57nN=}5)?{q*haMJ;jqHZ{(~45n{?bfp6S(e!P_s--x6 zyYpGh5jf)c_M$R{M~MT>QNsSf?)fl%t38m4`FGQ|`#-O44QJ?mFV{YmjyNg*IpWBC zGq0S@oS`l5d*gIen&vuE&p&#ew`E2Xnq~LIJHN4JREY3%Nnt*hdOBAZgJt`+DQ_f8 z#1@Xlq=>nqjqG;Z70zl^kyJh3a^V7V3rYE0RVeg$>p?0+XlbK}XM*_)Zs5B;>OVQ~ zS+aO<lYgqfhAx&RnND~9?U$6-`#QiAm~fCNGe2VDPb&MmRBqTOVne- z`jRCpgwvPjo0xBaPqs&%1WUbFR6XA^=-5H*%Z(i-$za*_feH~?Mk(PMd-8f3-es5) z6W?62#BJkP)+RCE%tt0Zk_XGvr&K-P;xHi_E#^Dwb^*(kmI-7dgqAqQ;@jpX-pTrG zcc+773Es}JL_KESQ?f)&c?4LJ+^BlKr97zwElqbQF9M697ZoD3sMN#0OkX>z9jvkS za=p7tmP(#w)?()OWQG|t_kyKXJ*u8>8FYOLTHb!v!uMggs8S(9iz{PM=>H6V5#bTH zuD@g{2<7yps+1WtZ`?2dC-A~gQuTaGL=N-KJZjU0@6Z?dbSgw>DPSxOQ}XaHg%y8l zK3uY-g>fvQip;4&@6zcF(3hxYEl54zQg0mdtE>}|MmXVpSeFVBTJ)81jdkl|eFN5* z!d6Wa$r81LV|o0YIq%Oq-c1242P3F@zGcGE&A7%aLrgC~Uq&6ELWGtm#uBQvxKq-X;G67D^^7#PU z_6OB%TM{gNJxGYslE^H!eL`_^<=tWvCt(Q-kXYO&vs2u~NjJa3_EKUssh3;6_a1?3 z%yaAbXs~!}AR$VNUL{;(*Qag5|A_d!cKK>y@e7n#axSuMfAX+=eAn8|O{89KnRSyL z_cs|{umwxXHWH$=_%TaXLF5ovW8;sfZ4{OQnMG?myBpVf*VKd1m&*~PUT&%5c>??L z%=uS2uzZUpAxcXDvutk>gA=PwqAU3eOU5oqU)tvBamcbIT!j^LBvc!xU04w`kvDcdwfs zsf8lurH@i zoeP4#6g?#&O3ToyxW-1jKY{OB>whjLO<2PBNi1dK*?!EXT%Gw~Y50-U%Pq~8u`BX6 zMV3RslGq~l_lHng!kH!H?&gosmqF9(J{K1MV2P#Gd3HWlV`eh`b8Ef>saIGYu&L;m zHx`r!%Nr*WqO@pN!@l@AKgZ*K)TyPp!cru&+;d=knRl?#bLfj}G^v+cnuhJeHRf{m z*l@6%zCuEjmgUT{vqQvS=u3%vXTAzc=6*?E&dp=Lk)N&GHWe(VpOAXF<@DaYXt`6z zdH`6OWs?x4C6if(^*nwDEMp8ymBEz>RX62LOLC#jcP3@)(om+Y18 z8-r!J83|EZ`c}s^c79kmKHV|){8&w42{|aS_{Fp9W9u)}z)9lGS4q9x5;L4l;apfU z5#Qf`JehSBanAFQHH!AGLHP(K+ zeHZ9UpM@kuY0<2KeVMsr_zke!U)!s(u;j@sPRH4Av3s5GIDlo$1yV1!SpM9CmUDfY z;=4SxK9CTl#erGI{PeX3i%W%DZGWLt`ySXAKeM2b zV41F4>-UFHTDsK4H8%H>>mabWSO@nOmVhG?%dF1q_|nAtR|n{e$!tFm#V2OUxc#yETgi0(KSD9thMO{1}tMZ!E%Plijvd6hk$ui9Nn-w6J8!EIS&qbN;cRT8Yq?!RJW5+%oPf`=MR8Zj+% z5eZRR+Gyb#^PX?@68f_9`K3j|;(c6VSy`K%Cs(kr!%45oxujlhnbL+$5-$(BfY)KN z^GS%(;>|46^`_%*wl^7tCRe)w@IL<(8#yL$NO{M-E7VzO;Wx zLX?(#W@(zVvJCX4Qgow3zzcaRJNeS|mhi ziDH)BK~V?5;t{(fNLYeSN-Ui^vfuleemgEfU!oF7z1;G;4ExjV(JkV-gJst(5~8%| z)W*Ji__1RPSSBC;bXZtiPf08<%-Hr)!_dtKEWOT=dby=XQ8_#x8}j(qH?ZW#k`SfE zm02pJcW4KeQu+f=3rn`l61;H^E|2%MobZY6o83vh-16fI+m9XX`nC*Ms+*D!r6rqL zrrIUqxuuWUxfo$d443rfb_l!ox-N8Hf9Ok{VWeJe8CZqA$9MUnJFUSoc?=0rS}f|| z8oN;XTORafO65-1g(d8?#NvFKS<J$rzXX(7iN9Y~=#UVlC5%}r z+NRZlVY(-Ex4#k=_cMPx<>_Qp3MUPtI&K>WQ=WFsNRnd8Q}=O6+;p1w_vr*vo_pGo zlz*D?v~nzlA8OLMDWzb_bEqMS`@1R6QtQ|!+mPnq4O5;PJxJ95Y0C3heY4dt<$1U- ziBwN{e$Rc2!>p^ahT|`^T(9uODA@O-g>L@8axshkb3#v zGd$=l9vZ72a>t3HU)@QF@_VVlz8A+UVR&}_DX~w!uoTEF*Y(&h^IOUN@m~V<96V)d-2@-4k|pzV67CU$DlG-68dI z%hPu34%XG_qZ@)H`5_5WT1@KUw%0cF2tHF%xzD)j!V-K=VmWKeuKZpY(5VerMp=-0 zxut$=6Wm@}?<-#)ECnM-h|&_wEZc9@35IPy(fMk9VR1h%vGjh$EH_QsJ_AesI#MsU zOwjd5%lWrutAXX-CK95wl&z0_DR^X#SAIX~)YKK09GRtI7CQ^9wC|KHSjM&@^>WMn z3+$@gj|xe6#iV^l5~8%&F-!9~4XQ(5UZ*bWC@d+FlD=3pV3sPUyPgA!{$o-vxA^6= zcSgQ_n&5_zciWw2OQdedE4!lNXX+2lCDbI5Hq+V`$z10-A zmoYth;uKD!ZX`r$@n;s(+!A=Tba#Wz!-S_>v2a>^ZdladcTUu@OMN8Ke zQ?`R;?F15{v`lV@YwYQRPG4b-)th8BLs;TsB$k*|b`G=nVDJO5_|G8qa!Wz4Q)ua+ zzXYF%(03pqN=qEGEPI{M4J>Kr_stiUkXVVO`aJf2vh|+bPJyLP9I2OE60WgJ;oqN+ zz<&$6d4q%~ExJ0`m%PhCufY=1x!MY0@w_Op*!$YySNd7w3%;l1r3tB*TXLGSD~#2b zw8N)v4w#b=rNxt3`ft-p2g{BAmg|HiPi8q@kqw-UN?*4D`ciWQsh3;A${xorWbgjR zCt-W3F^+^NEqTl`r*uBP_quq|2_Ioey(H;N@73)1a~ww3ulY}TO^%`Sewv8J!2`pKwUY!sYmn#yBogwQ>VYQJ3U>WH_>gATgY3vH)rixeY zfhA!x2~k>HnC0o^Sp1I&vr*Hdge6O62`R%YWeQ%frwAsJdbwrKglV|#cR1Jt-+kL+ z8VONavY16LXn+UwWpLk1SA`}1s-!Q^uCXifPm9K!fxegwCiQa5Am6QM$@E%{|KQfi zl7uKN=1p*oZJBz!Iatm|cD^Mnq1Pmq04;Wy+uX~2HCQ@dA@y>L&H9aKiR}1E8`fB- znqux;PldIJ7~oBdu=FSk@k+lH1ch538I5`LJ3 zC@oc*Vqd0I$isJfby`s2rLg47EPX$-bIaS2Rq>rGX)8#*+%hI*6IwiM&f_~)?cGR- z(z1YAR;|6h8T!)d#nL=sNlTFQrABjh4%21pM@v{^KJQ4q-11^(JX&fl3dg6RCVn9y zN=q8EOuo4gpHl1l{cEAHL|vCy%!1gI=dZhtSzuYHQSiH7Zn3P`6D?~OrE9?&>!(RV zl$P$zaE%?le$Np4l3vcT)W^TBj|JV3Se{*D=P;KNww?sbjq0RcZpq!uHoAoU`Ced2 zs!KwYmLO&se?hl4SPZIOsw^z7i4x0;pKPDJ$i7;8u&k~_>g5)@Wo)7-cbknLSZo`U z5T!+ZYlSnuMQ5q~)4M@G*qJ8@e_N7MOZJ%I`xpgNY zN{e0#Tw^y{-^TyA`}KC2k+AqBODq##vHjSa?>F&!vfgJ>FSj(k*c@;3mS}8Q7A!?Y zBt&WPV-}x}JMlVyH|=kIg{44dX?Kck`})xaqo6MdjYz%RlC94!Oto0((*`W-Tapl^ zrGQynK1Zd&8XGrh*brgKNRjj5TK457zii9XF zvCQ&lg+(%0rVZ+8D=fjcC6@BxGw>^Y`t1}>tkxe*>gATJHoGunz|G+KV0ko=geWad zTVh{yJ2e;w7NdJ%_QK+RM`CeFWB0{g1k{R$zHEL%>gATRZtVVxN23<_zn|}4kr1WD zomm=be60i)Elu4e!jdDim^WkBEnl5kRvj#-Ym$1orFniN_NDaTZaUDHu=*rKX~|(0 z{o7md%|`{7T-FFnN~)wUZ~fS@@{y@g23Q8(B=vI3)D^*K$;~rL0E^CD5~8#WZG~%W z@|*2=O{v4U4;zIg{I0~3(uW-@+ds9A152(gsh3+464(gul7hnz!D2s$geWcH%yR8$ zg&MHNT6~`4FD(A|B$gW8*{nj;rMoY|GV>>?ms`AguB0Xt&0` z49IoHeX@=2t-ZohB(sb<$&QtMI(P5{OOuYIUT#@kgDun>-Pa5OOV6$(L}^*hES*|8 zctT$ex%NLQESdKueW{wye!ui6zX0Fl(%6gC%Pj|u3~_sjnmO?~SiWx|AxcXovuyE6 z#sBTmi;g=hEOBWPOWEb@_kMt8$Q|g*+pDBrZn>r5jg~X*#^Cu_d=d#!TKcxZH8!sD zLYyQ{@~nSJSVGbzmJ*X#U#eBufWKdcY$f$_%kAN8ql>Mu@FMi(MIZ@LT0)q`qki&s zSYx}AHzf*-=L3mlYk77KvvyurJ?KlrbW$(3`1%jWzU*=An+BGkrzAva(QJ!-srV%m z{}1-c#`5=tB~NAv+QP2Gj45i0_xN?OXnCz;B_)o2?pnd=Zif~>i_mQe@v`8m10MNjylHp+|uQWr+?SW_nzM# zu`BZFVaIXa-%yu?D8H9f_Ps28Wo7`DZNuGv2}?wV#4&4SENgeN;kiqi^_##x=#dGj zms?I&yN8x%Z9`&Vz1ljG5Tzx6S!}#ogATO zvnQ}GwO@Y2D_KSJNQlxB&n#!Roy7OYtoGb#AS|Jo63dT~?8>k1wsW1qvgj(Qms>7I zWTNH3*Qgxmi%Ag)QCiyQ;Tl^O!QL0UW4TUGVex({u{cd-cS>fC>zV_W`Rho%+~QDw zGg@x9va1A^giRzwY4K*3B`c0kgEi**(Wk$#P4+0f5z!oV`1C8?KNx&_;! zC8^L0ua>s%Kthxjs}8uvg8K~{1ATG4TWylCL_L>S+#0egCQlAOnFp4@D%F42%PmR! z+3}_K*fyE4#*)oRh|&_pEai1Kb^yzrX?tf1OHh`?lHQtKEscJB0iRUamZFL5 zy*`>%8utW?uM-JTT68*MUwS{tI185hg;|cm;`%~j={uTj`@?IS6o_jISTuw=_D6IxBjkD=w!1OHuXTfKna^>Rzt zRT~WXR_4Ge=*y2GBt&V+W|p%?Yg&V4?K0o>!jkw>(wD5Q?CMX+AYXhBR_nE-UT*P- zU=w(+7wbfTW&B1GqO@4(;~IV7+_z7#fN5JA>eeaa8WW1L2WkDJ{?l*O6kH5wIW|MlkrDFNh zXqnXO#}(*Hgd+)2S~8erL{Q!+utaCs#t2L78;NDM@nZZKDIbmT4p!tlQZKixuFxMt zI-hTV|3z@-3kgwLObl_2wVqoH&y#~5HNP$_!EYs&s!!RlX~%UnyTKZ}s@39mz1(6L z!rtW>H1EY{=*yv@Bt&TmW|qaaM@mCq7JYw~DlG0f5=&h^Rp4xF`4}uqjh_Fmms@sK zjK;pW*KdaJynix_geWa#J7Hhu%K;#25Lt%Ie?<^OektVm`lQJwvMNq%*35?Druk$SmBYYBUw zU*Px5Sz!4PK|++4B4)`OG9?}?-+d-j7naOCNnd)eU{7|HEZG3x8n9p|sh3;IN3px4 z^U`z8z|v$N2~k=mcfmEbKK4!|SQah{Y9K6eA0(E+$;_gaThIymGUPa^ms`GWU{Bj# zn&*2QELWb95TzxKSzI1zUIa^abk~-`67o@EY3IStE%yeUwg=1PYouOoX%WM&^}Xu3 zKMgF#$s|N+(KW)pnA*kSbd*ou&HBRP`AK4#IEdYNuHUfLFtBtPOzP#9O>IZxe(c22 z4ZFb7$&!R9EuPG>|IRa zfh0s}8QB%rSU_Z825c``?XyP+OT=f1#i%|z=db_055B|uoDr#)TYeU?Cn6@;^!);T zS>KCxn)O(d0f=}GdIY#Q`7K?bcPwxA09a@ae zkPxM%UN`K^kfAA2&=;S|S7r!{%U6kI`5E?m-|$&>JJ|Ls)+Y6G%aK7L_=UKYURDHs z+1ZGMC@n6`5?x@~7c6Hk)O8e=ESV)tk8S&h$NjtsmOk;MUT$gHbue1Kwi$@~v3@Be zL}|%lmH|m4CxFH6+WHm35?>(c%es#2oWDt@LvO&MbBol=Esr{|_ zaAwPgV42w3CPY}$zDxQN@sX{u8I^L;5^X^0<(ALsY@^dz(-fbEitSE9l$JDR+4X55 zzR4xbEBd6cL={OaN1L*H;tT8y3So^o_>+3M6M``{ewt|($d`+*I19@ zYw&&KpW6++AS^*YB$j?rYnBC|~UaA`FR z(>($CRh776ZnPiCX|u zo{dM4kpDR4xz62c3QT#{A4|giobrsB8-TN!FVnKMVPWc8lYqaO@-*tP50mU??->nK zp53RDq<=f*`NszwZEiggraV37lGuNq^4z?93{H6lUD#6}raY@GB7y&E%G1Ap7*2T> zYW+F`d+$jrNTzZx@t*A^Ryd5;gDKDQt|aEaO?hf>oe&CpPe-%l;-CMzS{mk3^3N&H zI@j4rPu_BiEwI+}hmv~v-t${HyDHbB`#5~Qe}feXQGPFaCU{g!Dt!+y9{x1WnJ93TEbEw zvo!KymR;U%6To7ZMC#?1#@6gAL;rV|8o_!UevgDGEd|Un^-3SS{(x z(w1xrr_4#G5n!2qp47`NoxhyGzMQIWupcZ#FOd+XWqdDOW9K`k;1o`=lHXeiOY9nn zWnN=8s1bka?PahG&?NP8OURKQXgPlG6TWLLvpNY;T4I^yqpx=uZ2O(v2N?=Wu&cxp z)t#-eHWy~&JH590l6tx2!A{nf-4A=?oA#IQAR$Ui)85#ZBi5sQ!Sd(pYM;3T8eiT%YP*yN{c(QXgZhF0L!hhPx}i?j?8k^ zlTCW1RXlwg*4S8EQZKjEXfY7KkkoOu`2H9RI})O_iP_LjC_Ii{;OQdm;lBz+lL zp8ZHL8=KPuEczWuz1)(rnVp?iUJ`_NeYP2q5T#{kA6#Rl-%lz8OP=1QiNX@TPGU)_ z#cm8F-SEa|NsDzQ^>WMSdgrk(4_;)vg8kUXUL-_m31^nC=j!1dtUZx3n4gT%7=6ubIk zmXwQAIQKe*|E`x?&NgG)%P;!?Jcr4ePC}HHzNWaw-uAWV36>*GZ~6#Jh=;^7UXNY< zS$Z-6kNdf8Nxj@+dyh>>Y+rW!F7)NOAqi1hLYSp{2N#@H*NyBNAS|Aq5=&tgJHE^) zF@HR4FV5FUz1;G(j2~|MWg|Ph2TNiy2~k=!`(a-!HJr_%FNq%(9uSs1nI$`y^`-Kl zUq``W@}1PnEt=na(emP`-*&K!FS+RVhfrD^m}P?3jPGEX5|t1pEU6nMeK|IjP2e?& z{eV}pu3RDYa*M<3u4sw*(0nRbB5slpr6rYFdc5g95G>a!HjWgQ2rr4{_G`B7e=BN< z$4Y%IQZKi-jbI_;pYDkR%clk;L}}?_hHLD{%sSa%v7c>wMOXqhNi2bj*!J?^ZXsSR zy|sYU%PlU^Y+UT&pFle*tVS znQo+BZkgkI7(;GeowOS)AzmazX(?uoeOb~vJP0hsEUP^bmModY<%SJ@jC$L;Z3WAh ziKJd`33d&{khLAX3&C<{8VONaY?`6#_oh+PMiz+$x_lHng;+ZAdwJ$!^*>9QidtnLnl~_*IW4|j~%x{X5_v22H zdbwqHS!?WzThZxc=!@Gc5~8%U>5psdVC0r!u*SZaW_=SD@68fRw}b4JJOh^{euloB z9zyEnmVCESXqoV>EZ&XVJeq_kE#Ayh%z99FuuLx1P~*#A*ZT5hmVyRsAYitmDL!#m zxiYDjTh81#gO&r=@8J6|3TlxMr6r$P<__M5<*v4{L~W5+M$}=`w@)m;;d`$y z%qI16OSjAHj_<+&bC!U`%aMdAEm6#3TV?WFSY!2^t2Pmqpsf;1i+r{p`}Wk-8TxW` z9jTXFaxK{J7xQVl_!M+IZxW)k=nTZZ)QqqC1}q-)Cbkt8*KHDu?^kwid27ciTw`&| zNWI+BPA3D`*t3wD^`I}?*N_mU#g$o_+n^BUFIhFqyEY6)sh|-eHER{puF9g=u$J3-< zZs~es3R;dk#5jWGd^8DBS}ZJZjm1P&z-t%Z6ZHoPOW1aaB}11@tIvwBzZWb!N*VmF zms>V`V+X0i8gI{mrLz$UQCh;7rBm<8tDrB2Edoaii{B23C9Qx>MwY3xCLAoqJCJ(0 z#d_ry?8~0obK8UEQx_7VwA2`ceR(zg$Rn_X40~WLECn*lk0$J%#WAN-^}*719I2OE zz8jXt>&Y#juR92qTpJRiv^X(K?yceL!Lp@7T{~gP*eU5tod|Z1Z@N{Zi(n~Nlhn&C zW7Z^LUk-elmPw z&~nSTjV<)WBZPz~EoFybU%C|QoCB7J85_0YMITzr=YD~l6tu%^VeOp z%(1$&5iHv8NQly6$1ELd`%VXo`-dC*ge7ISq%TL0EX9xUIifB8M@0MYq+V`$bnFC% zjN5O15-ge}6MugQr6q+~hFvXw1}y6enjI6C@F0oB@)NuFy2B|HrwM12Z2r4mZm~CN zhx_Cm{W~-POH2h4qO=$c#WmLLM8ZC>co?jT5ElPE63a%th4__rw%Ak(wij(D^vc>QQ zS~ldp-U^n@9wbC*DPoqD#^vz;lp-3}yD2Q0dnJ7tpvAWR4V!%MEl%azl6twtF`Qk4 zz16ENJ~@2OfP^S5lZW9Ndm9sn+so-WOYRFx+&+n=ReN?Frq`kFNwCHqFCz7FOQj?> zvAS&iM7+nB=uASCmN;f9y*mZ}PpRwmb5DgOBv@kEaF`umE|>0?2NvB_QZKjkeSa3$ zSUcUg6R^hCWsnf1MRz#%Wz`9<(=bf8+&r<^*T1gAxbH9dcT=A24`$+|;jpl7^}CN z$C|;EXZ|b_^?#c3%rE)v8%%k=olhdwQ=Y^72O@Ka!MP)_qQ)&HL1IPyG39A+g#k?u zt?da@o^4ip6N|SP1pccj&&-Aw zamur+{>BlompFHrWGeR(?rbm7_f^C|nDX>GL1O;fl;`r)RR(Z`ZnMI$xY*-({%?=- z$Gbd#-eJ%B6c!sk4%T{(8>yG?J$F`TyP4S$fj8lMdG19*l;4Xd`(EBpn~DDr`dZUn zQ&{q3mg`5^6~=_m$9sY$s1~W0TaN8ym%^Xi4!RGP4Gl<$(vrt4aanP5pfC9c&({)` z)B}>fJgd*noNDDZ#CH=8IYH{>mRZ-v;!({wGIAbRT+fjZrDfy@JgRLvsb>S0mQ8=^ z2us94i6!>>BK#Of^7i2r&i1yXUT&E*F#^xd8|{s343?RlNQlxB!7Kwk_l*Ec`&olq z3rj$V#G+uVzmm60HEFG4U5T&KwNbHO2zPppbVw|(u zP*_|JNi4;J*%iiRZyw+~ijI9I^>RzVvleI>=iB5cSX%ugAxeu2v&@)Ou?JWNj(^!x zSh8f6p4M!8nN)7xKG^mLO(*qo%Wdt0Xc@EU`6I9_ol8QLmMmtOJa-_z3un;7J_Cd$ z{;;Gk<#(~Z6dM;h6f6(2NWI*0CVV(r%I{Ao2Ns815~8%2kHR&UP`7nI=!<==l_P~E z^oYdru^Zd=Z*Q20&&2Q4A@y=g*he<)kz;R$R~aVhk`SdOlv$E>dg2}5fWbE=3X6BB z#NzOSeT?DB!A8&*4J%SFw>VzjihXgr;eHSLvV1ZLQCg~4Vqc7U+v7ChE#D?Hg(Y8R z(KBUt77HIw{05e}fuvq;>2~7@T6S%2fp>f@_LC5$WdXC?sXk{ltg!$O8%JSDJ1Xf* z!Z^0Q^ax*t?|659&zlxTD&^>WLv3rEmWd-x>0OX}5=geWcDN8=hR-PIeXqsE=A<}NHj$0e32 zjppOWXg&5RPI$XVk$Smhl0UPY?LH6BEjPuH5TzxES)xl+tp;mskls*VVQ~$USj?-j zZQtf=%2U{09vG2&xus(CNc=+j9iBB1*4VFJBt&V^u)@AL>VNA3mV?cF1B4}8X8GhZ z6+gzXc|{MwVxmXt<(7@D4q(WEQO!-jGO`N^QCem&OVw?K0buc|`1*jbB%YA;rAg{^ z{20j{N0tCfb}dpbx3s+Mfgy$M+Vuv@wuU4`X-Q<3)z&$u!IECy@RYEGos?LbePUPS zgO44z1Iwryq+V_*fA9xdTuX#62FvbxBt&V^8-r`i_HGLPThLd(RnfxYcS>RjaALp3 zF5mYZ0G3L-Nxj@+csm#^{#Uwg0!!nABt&WPW0p>xs@Owc5{#b33rm5_;;?|7!yGx{ zfKvs-`;dCM<$ALOwCrt@;tiHN14)R|Qot-a7K@L8WmmY)9bw4`m-OX8B0E-|9_ttb zeVGzN>gATepAl$zYv1+*ST0^8Axg{mvAD+OL_DqnmMVS=9|=qBX^F-8B|BDjH7IKa zmgRw@UT)dcj9vM?=Ij{>mJ|C)h|&_vEcc&An1MxeQSvKc2|gpSteD5H$lER6y$LMG zTu8m#5@Vi%eOWf-3;qwu9d{CXFM1&pm&5k5F6MA^VJVVXjBVLl-Y6PCH*= zzC_+8^>RzVG7IdB_PQ4M57a9kkPxLMlUdwzL;2L2VK-rki;`IOcVfrN!<%(tvn0G79&Bt&V^oP>R8_Py#FuxKY0jun zKY!QDExW25!oDokn}+Afv!{^|r6rYFwECvwu~Juen}e`K#7ZpRt=Mm|T~6*bz%qRj zsh3;Q64(ypMqDJm%^_kl2~k?QOvW|V`(p=u_pMW3&1J$8a8Y8ZHkeIDe!c0t4J`SC zNxj@+a@7j^a>U>czMJr(B?(bl0+{9f(&l*WBJHlzT48ayB(XeO$c~ju);#e8%g`sJ zUT%4Cl#TG7HO(Cd+sn(BBt&T`W{rJmxHaTCZ2OaY#%vOnESaTEbGF8QIA+CyCC!}F z%Pkk)?ZdvzIx`9Xv+K!F5~8%&GE0Y|>vg~~Be~uVVTr#i=}SEawogt;am6R?o1~L^ zxuw|_eFYMtw6vLmYpm3?S1qA0-9m037Z&d;63geG?9ecg5*C zV<&NYX?;Ow8CWK+CLu~oKC^ssUoao~GW^rdE5ed?RnnIoVbkzqJnrW33@pw?q+V_b zUFC%#Jy$NkcM^MP1pNLGN{f{Zt}&At74hHaOBz2;7M7@M5{s8NJGXq@r5;`%yKF}4 z(L+qO?RYOP0GM{+F6gi|!AEB`98Esq=}=D!hm(hEEWU=|Jk` z7Rz?S(NfI4$8oTLi;gY!#V+AuFs!j>SfpC8{Dj(k$So1 zdWW@WvFH>y2P|dYkPxNCm06a!9QOuGuK_RL2}`!jQtUQ6m1$vw z04+JO?xn$UcN_^(TC$nNqR-`SFc>(Xw{fX&e_d1Bb^UKARvpVf#;M7+vrQYq#OgH* zlBAeeyMMH^sh(IhbUBOPeR`uO4PiyyT10}xiuz+>b^c2ReA#Pl zB2287a3&%DabmU6xg`ZKvAW-tg#9_O>K_z|bEac5j@N*N8R|g-{$^se%#O>LG;92e z4=}O%z=tIL+lkdbKH#j1JBPu<>e=li_MazK>wa*?iPe71+T4eU)$Ktf@Lx@=dS~6i ziB?Mi*=@PRaj$t^s4F! zOU6w}U&;nf!H;2gA`S0g?e0qI<(9}ucBkY>ZZ`HMt~Uu$S~8eri&MkLV0p1)fRV7o zCP^%JJF*|5mE#xuf;DzDnAFQHvpQ_VFXUt25#L~qxrLGtrNv}AuCXsqj^c!O?H)UN z3rlda#M0G;{Q|60BH%4p0{fAAxuxqa_TZy+J5BtDQ1ihgL}>|TmNDZV+=agQoXH<3 zEbb{1i$+&=$M@x>R}NrV8B6NrmIM9RMt9QrDo#fBi6zIBE|CyrDpVZ4OCuuKES`2358ms@@6(=MtS2mp| zEdF;SmLa#<)zZ1ss?CEnR&710ms?7W>y2y7Y(-ixuta&25T(VRS<0T6kI(o{jdNTo zEJZTQfh2YgQ@U#O7_gK{BK301_F4&Oxi{zuKI41&E(uXuikRhAzmtt%d)d%6WsR_8 zrb_x!qZR8*xBNFG<}JdyU#RQfMtRK2~k?&m}N%6qde$K{YBPWg(c*k z#4>gtyBlZRyf5C3tLsAQ1}i_MB}A;OX;v()&> z4%vnuhX#UWHp7XLqu;?2EMhZ(ry2Ns^7TYHW zpVo>4%ZMJNUT)dZ)eXOpH>2j^6GWD#Bt&V6V3wG)mDj=YWNuoVumn7iSf^Kp=T+uRA@>g5)f^K74deN&Beu-tbeAxcXYvzYqt#5-6W8^vY{OMHf;FIC5}-!D5m-thrT zMk`V;x7gQn#WhyJ@y#8uT zQZKi(el!C6Vr~D{5G)zfNr=)C#4J~)HNtlkZTgwoNLXB-Ni46{F^lKSEBJ(JV%fXD z>*banLC4TCVCj3DjLfV;LX;MbdDxc&PBA}Wjg73M-&$C*WtP-_Y|^WG_h_8L+0lg5 z%Pmb3`=MofTs531ifm0nl$IIHa=F673t&m_Gsi$!5}!-@QtLf?QpKQZiZ*O7l?IV| zxux>`cWAk|cAzP&vE8Fch|-eCEZ=88$F~LyZV}f*Si-U-mKrVCaldBj;sUTt>p<$| zmesy~Xc@fi@Jq0m8IcgBMQ=W?F$en?yf*IfxSpA?_`Q%=oDQ>d%aiMWo&rnRY*H_` z^fQS^%L$kJO~LXskAx^Ke$0{`*}FNcu}xm1hYCx9%o5s#-Aj7tc6_kl6OlJQc~mj_$e@n!JFW3#}r^bo0+TU;g# zMN93zTKEstt4@#*rDeP$uCXU4yE#K&O!gJG5ti6j5{v0|b{)q5?kfB*f|^H3z1))G zH3WaZbp83TCiF%3Gzn2!Vwq)J`#Jc#QZuoeov;LFODwy+obW5vtGDL~^kw2*QZKid zxt+(5UJd5r?|th>Bt&Uxx&ZsqZ$n#r67pF?&xOL`{#s(Ow`KSEs=plZ04ymjNxj@s zz0?u3WbJ*l1p4x@0|`-D+?l1tz3F&;thLrlXJN^aS+e#{!;kS`sSUm@C`^ac%Prb` zns8yxFFaQ2v?L))OAfPGMVIn`?WN)`9d}_#c_Zn|kOOQ2&vlVkB=luLxkkV1<(7gn zY~aD*tEUe1rGHftqO=TMh-)l8eb)i7Xq24mBP`)>C6>cwndMaIJA68Qizca;TWVhL z#^w&bmRTGuvucnKr6rtM+Kunt2P|95#_tdo{~U>>lO8)(uDN^`uUl@fO6uj73)R@Y z*Mg;|kAcObHVIK$v=?Dto^M==zs0Jp|Fut8iewg-p6pmTbI)PCj~wAk>gAT`<43VC zrF%cXW994}Bt&Uh&Mf(n$ErhLn(0^^6_(6gNnfUEur>BPG}sHa7u}YmUT!h1#HMfc zPrKk-HoobR5TzxPSzgUe#{VMFG7UK+EOGB7maX;JedH5oLw|t9elV$*TQ-gh!M-?5 zjB|oDW@Sl2l$O4WagEuT&BJ45%?96NgeBy?#Bx2Jo%5F{`wk~o({xF_+%olAAX?Ju zpXd)3e?1bSw1hB=y~k6$K4$W$TY|87=1DA(eb{x&JMn!VLtj3+dNKg(dZaq%S91Ey1t!Xaytu&#q@0w|>{lEmiW+=kPxM%%Mx5;@zXz=z#989aq&B03HT(jc31Ak@6RoVll2hyIS7C9F@~u+CeY{=U7s(=wR6H%TW+ipl%kd)TeZwBn85!{q(crzGW{ChzC#XW)lwG5Ww0 zn7kjHL*o8!@;=0leX``MADvdyw^T^3z>C4cTa?^ zLqo&)cNfKq`eX9`!ZHTjJmcCun7rTHhlKpc$$OiIK5!rj`*ZT%ugeXbm+!Q- z_h49YMBc%qmHLCtVe-D~7?SjFC-49Gfc4vdkA%tlr`9C)pC|8I)^x|o z`_w)@sW5r(J%a@PtI2zFw``oeZ|-(*1?=%!Igm`{UZOPHOO#p={s<=TZ5NZ6|2BEw zvO|Mcu=l)~u(GDi*}!3Ux>k6yuz5-hJ+|B z-pmr$!fqT`j`RraA}skbOYS7L)P2$-R)ghaKT%tnlGlYaFE&0r{ zqKEw~u;hJh)mK>33M73AYR#@PlsP;S-?tQ6qV?~3x#dvsL9{H7aKbBD>E%g?(qgq7 z*VyNS&0WCqA|i0Guta^6SlY$1NiWaF_lrYc>RcxEa!aAr7PO42c{U6BVs?XsC@oRU zl5)RoJXoHOFdHi@L4^{_tETq&G4>sKfKvs=6G*+>a;mp2hFC1lI|r86Q%Q)@qO$_~ zQc?ePC9u>Qvcg7KT)#^!Npr06V?3^mKCX&TRv-S!;qEDuHt)r4o@Z_N{cJA zq?TNczsy%B+_n>zY?X%Od%gH??L}>|QmP3WN z(!o-1S@e2g@%t&UbpOuwV^^l@;Jcz{ttItxOJ{v;yc?HvFmwu7+%}RBrKQG7>`Pgn zZ#YRD<)wJLgFpPvKEjli^#x%xn z9SWA#cDcudC3uy@vOSOO$2P|<#&;CWTS)5VmRasz*q2M+UI&Ba#R?LlvDv4g#F+c&*gxi47q2ayn^#g$p&o=(KR z*lt;KNmz1Z7Kb5hjWuo)vKuUwc9VL!|vWHU=AJFNh)=vp_w zD=guz5=-jXW%w}$#aQG2kXXzn^>WL}Rhbxa;DdG!SZX_x5T&KgD%|#WkDJ#YEWeJ8 zdn_#eYb6$|xon?2tVGsG=u7TdQZKhS4KhKCeVR{Gu;j#$5T(U}S-hrg!Y8}FOgx({ zEJZTQkqoxRuD$Gl|3W)$2C0`@0%F#o<=_}oyz8UyKthxjXJ*-dY69-Z7G{6{BrKV3 zlD=H2%K9?Exye`P%j*xMUT(Q%xE=Rn^nWDe$ zE`+#CEGNsbNw0&m(zKv2daFsj+_Gb6h28@ylrLAljH8XUo%MXjkxN%?0ZYXVBt&Vc zu^PAi%@M|qU~w_~Sw&bp*Gnvh){F2feb8|}{y)rwVWeJexwLH)hBSTl;VxJr#*h%D zr603Q$uq^LZjOW-)e)9FnZsuEr>%LuYBPgAT}yV->)lPTa^L19y@bVOqr_r7ja`S)GM+dAEC&Kfz1-5YEW1A!V80OG8F_j?2~k>1n5DhV ztF_RV-q~jc3QLyEa<39Q?uWKa$M;_M|48cPmg;$WxKNF47G%R3%l<|}l$QC-VqI&} zHL$pUs5Dww;=Lq&Y1f`@FW-uqw}rl(DE;hTPA-TkLNJ{tCFIjFRR9o5T)fFvuw=R zv=aI<%;L~2Ve$5sS+vdOCXx`PrK~G%`zOC_ zb_GjIjW3SEk}tD3)MeZL`)c(Tfkmqlsh3-Z{76NM#%_0fg6K|75~8$pWtP;}C+dOa zC3aB?A)^L&xMV_vS%Nums^IH-HMhACS`P?FYQ7}h|*%mEYlk%<8eQ8 z?+$liiSm_LvKO;M!;I4=1HobyO6uhnvog(bduegbgoG$9Da_JkOY{U-W804Q z*eNWoeiBO!Gj=;Wv(>5c&=XB6X{dFvBt&T`u@<-e?YjpZ0*mpM zCBedyEwdDauw&)6hha%z8MK?!%Pm{i>_$u7y2g0iuXB)uC@luevUbbV6tD#A-#98P ziCZLn38~1o{imL*@hND{9;9Aw(f;`yE%utmEn$sKHzgrT%QR-$GT#Z0FK5p*JS!|= zTP2n`qu91T`tE0E=u5d+QZKjE-O8>fw@MA_2bSm8NQlzn&n#bd=RAPESl1dGD=dE7 zB$l21+3%Me?O`RrvPy^4%PsAuvwiZSORp<}rI#)VQChAuOTPZ`JzyE%ZFhpO6v!-9 z%CcjncK0C(V9_-t^>Ry#3b%2MIn}cZ2TS)sBt&T`VwMWi7vS;bfyU?C!jj=H=}XTg z?Ak@^PxD@Yr9>rCFSlfzWXJur`P=c0A$c`Ph|^ zReKA*!ENR|QZKhuueK5`kt26i0n0BZ5~8%kF-w)CrX#?z|3QrcVabtMs<^YhoA?NNWI*$vm3KCKI8KPEMMM|5T)e{vpjPR#`lqzc1f%JTl#Qoh= zwEIx@$$nm390gO+`}dKk|I<`-LG7w5VJbTK2#Nd;Q_-W&&Ci4_z3nLyBv#ZPQ_&lz zGvK^e`&PqL^tJOOXv1WZ^lzu4|M-AY&-rzSspykwB=(=DqWv!T;8b*cBj>3w747ka1pce3 zXqR23dcqxn&Gp;iRCL)_BvZAQaH+HokE*+)>sE!SXx(=t=D$rvPkC?a1$)o9@gX(D z9w+60dz?SsN1i{F?LCdx-SdUDo;sD(%lDqEQV!vfpk?$(yi0mw4hd0xFU54(_tG|B z_X${h-`ML2OL(Bfa%MD}ByP}UBVH{XmQU*C7RU3e(c-$q!w8N9CPgGfX&K8bm+qV$ z4wjsUAuWZ)f0x8EWgVN4&}&x_&(1r(AoX&~`q5i)_usbpa-5JTnM*>H7Efkr7Nn1N zNmrfFHV~E~nPo}%3j9jvg#N^TZ_GMR>gAS*gZnYWw`wv@6Z%~uAxg^yW+^}VWM%P4AFSpd2$3{lVzMh9y8QP?h5TzxLSsF~4gYUx0 zJXddku*3yPEcX4`BjEW{#+89>f8AwLFSne`j>W!MnhZ68?WN@n5~8#;cE^3<<0(cr zp)YTLZ675pA$ug2{4K06Rt?7Bop_saJAT*8Ez48bC8T@hYK{j>|EeTJX&K2Z%{#l| zw0ic#f!4y}xmRK-Ux!^WNv;ri1p3nI1*w-?0(5h*FRzE68UvP=xguzXIHXLo4>XPOQb&uQCcFH#n`9YMX&_a ztguj6Quj&vlJk>I;1!iK!3l}44M@G*60zhc_N7a=36@}~*Mfv7EpM3R!-4SbU@=Xd z>?|x1!4k{TY_`;Ms~^X!49gyodbwr%hiJ6q9O*k4ESfJ!h|*GTJ#PCy+?zB3OY;$1 z)(K0%eu>2`lJzBi?YtSV#R!9T=vr0X&q|gl;<8>5~8#WW|lVF%@=?rcUrc$ zu(%wMSWIrRLqnGnXYnbu&JLtrZi!sYrf_W9&S?&PdFDhylol6eDRzBOd$1(8?YTo( zvSb#E5H{`cz5Hue=u2}WQZKi3J-h?g*wd6H^}tf5HwjT%PBKec)yB2KVmftgu&~4* zl=P*a3EPitou_FImSS^Bz1-qrVThKRg`;YLrR8E0qO@c&%WJ)9_(a4syM$xH5*i}0 zyt&4<{Tqh&@I6>x7m|9pW&U#37dyK;_%F1P&Ll)>skH&OeJlGXkB8vd`s&rc*oX)$M(nFGyVgQdFe{ENbpFS8sz z&!!1atjK!}ma82}z1&iB|4v+E73;J~hBfBam4qlQOPR%`c+GBLNho&py0D}jmh@#q z8@3;7QKG|6u=tcG^>T~36+5@o+`2Fl2VkeB39;HX|WQOBD~?_Px?8;_u4l2Y0>_7T2Q^OZmp^ zPDzi{KGVUH9!BcrmVL!vqh(pMd-!e-#|RRlwDe|{He>JOJBmE-Rrn|@*)oekG`oYP zb>9ppit1#NdbuUrc^_K(%w04U*4X1$Bt&Uhz$|H{0&toz+|s;ASQ3v(`r;eFcBhy7 zXydfH;bc-Tx0Lc6ftIE=E0Vx6YX%8XS`ILaN9~R{wf17!hSER(x*{KTTw-xgXA?!y zZA#CEHCEn@)XOdR^$hSjOk`l5FIcW`Bq2&m8nayNxTgU2V}-p-R23G#Fp0%~6zj{~ z5$Qd^5@Cnb~sL~-U#$*CM+2zBz?(R$SmO{UGaYv zCLAR7a?1r@Z`@u|<_6-O_)cLYL}{^SmJSQ;e?eaYzW;15EU_mgmLsd#bxTkCoA`d^ zd8K~+u9sUDEjxvlVbc!8!S*t}A_-Aif|zA`*+w`W_2xv+uEG+0N@59U$)4z5{Hb6p ztg+dVq+V``3S__cgQ{Il1dDSV2~k>ZGfPIRU(=y4zCC?Rg~dHwVhQTb?)W--+^q-} z-TtIrZmIl>O?YowFbkhlF&{=klopMRxb3fW^nM2x%^RNw3rmj7^56`+&VR^uNac@k);%U)ytgxhfX3$xU+YHtRX&6cEI zZn^!1ZFFs#*2bv<--#qdX_>(+Zk1=^e$2sdsg1CNpOIJ+HP{;KQ1fUiSW4_B^>Rzq zQpvdOAJ&-_3VnHakc22LJDBBB-Pn%M7yTbucEaL+R${4U&dx2r_J4ryR=-vH#qWB# z<>&rDw0K`1gzs5&szgGRmPBS5zW-G-uuT1>vq)HqWR~pf?1_kVC!el{zWnb!zQ60` zmW+PvSb3}VKs@LF-+O$2hbS#SnZ*;n5xi_&Kq!0Zw?&^Clrm%T;Ev+tz;`SZ*bxga}KX%+ljI+xEK-Tyh%LSoJ%kUT$f1 zkX=)1{HU!y^u_Zb2~k=Km}TOJhj>jX{AkTn!jc*#>5Kbn_S8+pnLK==JLW8@ms`?0 zvxVBA`|0gqSrl$KVTaNECcq=VZ_$)y9Lge4+cVsVON*DWjkC_D;%adjp2a*L%7 z`{B^wxHVWumrf)^Y4Ku~#Um!-)9GdMO5G9`ml%m})85k}68h4hkkrd9>#j{g zOCNhL{4eNx#Y_MG5K2odvrOIM=?2^WF3lwmg(XX7c~F-fE00vphz5&IKTD#=tYU8U=uqu_*n(uf7>oTPwM5Cw>S3S zHKk#{UgABApi3k~X|ZIMBhl?}pIrNSk3wPbz9g}X(O~y1+V}a@AJ*8aTv9K$*n5Xz zUkYoj#^X!1FC;{1ac7n*4{ub1SHB)O{iepRzpnK?|DWY=-bwtn&vP7fz0iB>5SYI8 zjv+~k>02|`AIO^gqWNc-zI||wr2NzL?eoE}@k2e@(y|py-$o^qxWAje4f$1spUiBJ zu{BKJR=Q83{!i1l5!z{KFnv4!F^T*S)3;8WOe(>OD)y2Di52z7^zFy4?5ltEsQC(* zzI~cQLjL3QZMBm*4`BMXQa%a$bNaSu#v7cO)v;OP2MhCIAqn`K>D$>_`IvN}W0UDH zecQA6+uuw4-%j8D@d2xRy?PX;Z_kz^vHv`MyC`M{PT$_I)yD&-Z@W|>f&XgyHvVN5 zoW31U+VDH^>WL}YwXPFL}VoXD^0g9Bt&Ux;Dcuy$3Aw&D_IryZ!r*- zpequ~@B!@1DaNATda#5aBlU92*t2EuY~w(7u7rDZ6y%;;JLCs&-xB^wKi z>s5&*rZl?~U-iQLe6UQbPU__rx6|yD#UZ~AJ`+Eu9tlxe{vT)e{nzvN{}0^EUfFx^ zot;hg-kT&NJ9|q)NRkkeq#~kHlnPOjNW=0G4~ZZp;%Fw_u6o!ZS8JvI=IEgE4l^l*G#|bAvmhW!$e@=1`ZW z$4C;TrH&=e{W@h5arb(w``4wyV)a;JDQU+n<{dYc28)ItiI-c}lwt2$`&HHi-?>s4 zK$0jeqnRbY$>14Kmv`I8uM?JhndMApwo`I6CBqRc-S?4rx#h)yBe;$Yum0*A)Mf4w zl0<3Q!7Mjy_lJUISliRvg(WUTQWtv@c9)_2b60P$Z0bhh<(BU$>~`3(ai+IsOwYiJM8h+_G@yA*{?96&;qD|X=`zbGO|~01;K3C9ns)0qBwlWD%qxL&|7p3gZeSVufh18{HZjYs zxOVvLb*thv+=V3~Tv8X?9A=qS$H^1wva&phms>hb*1@{0F?x!3vDavlBudL=W?6pb zy$j6!Uejjz3XA7ci6y5kJBO)KP)-9Zdlry*xn+eOJMKH#zOaM3tY1o!C@rzfa_g3% zKUh+d(t?D=IznRk8p1lgVy=$g4VIM|BwlX${fMo}gS#!O1D5uAB#F{e(F*7O??c`= z#`gEw`%G90WtJAVSzRoHBR_$~eh7(|TM7oSf!Zzb0KV79aWqMyvVF|gl$N#3Vy9oO8(8LK zZb}iBz~>T6-D~VgmB3Bu3!p9*-$}gO64=xP=l&LZ(?($NEWWwuM<^}!%o4Jy{3Wnl zon7!nST4ViSf=@~dlzOpd$xn+kv@r+TbliLLd&Pm&oaSsZ6!&fv_vvX*BR|bfn|B^ z?S;Z(7Adjpf5UcscR4?L2$pY6NW9!K{hTXW3ZfR`SDWqIk|aut#tEGJdh;JG0ZYP| zj~d1H{O>Mymdw&Lp6xE2%pQi{bUQwc#LF#Dnmk2|kJ0rYsEfxul0<3g%`94_KYs$t z^BzO13QOcmNnM_=XHD}d!w+|cx(pvk;^mh9mDrQRuY2kg155Q0B#F|poLNpksD_>Q znU9Xv6&CL(iN$>*YdHI`{zxHMwmc*8a!Y-WB&^G$#$g^{S^tV8QCe)6<#qV7CbR_X|%jn_kY1;;dbz@qD_GvO{y0cj zf?rAMvaubz-XGp=w+D=|&Rt2o++sa`F4jd`!+thcHufP&loowvxtClIcYHVd=!_N? zmsp9VOaMFXSFnAJ$NgD>BwlXGNMKLgnf`Pf2bMZvB#F{;l3Cim@G*nBJbZj~im+I` zmRMpMu=~le@sY2gE<0Y3c)8_S_HnFB?9it8eV8S$NfMcF~2z%PnQ|zoW(2BGex0;_pe4C@u4uWtH}-E?@}{dc0Lw?BXSs z79MoL|9DTN-Rrl z*?hSi+VcTeeB0eCikDkzwPPnjqwYMI0Clm~B}tT)2h4ISsl{lpv}kq8Tv$?NmdEYb znyjtuhMmOAJCk_1Wy#WcIL3ao8HM*P%k?5jl$IQ3>3;St-seBors8Q~2}_XFWqn*jTKudU^dU$XAtTsuaw6wCux&QQTo1IXXM_>0{6c)F45{uhe zb`CQ>a3sE8d0rujms=uhvbQ+t^j%gP##qBr`-^^r(lV1-#yq)w5-eTkK6eoot3-*V za3`BDhYP;nhq?qTCh>C1@5&KamrcjCe}TndHA$kh9AXx2o1h(FS?5&tmayc@EGKN) zeg0V&x;_R=OE(fPw?uVeU7laxZ|e${IB$|fY4Ky0`P=&8w+A<0(7h)taqlH{iCDw# zjeFNSg2$J3S4q6wvQ&2t)}@Q{(%S7lZdCUT%puWOHB3{cC%$j7}p-l$Iu^aqc&4{{Z*IlMH%C3X8)Bi6x8Q>uXu- z7oI0?-9qB!7R`a|yKr9rJBI^HEfbPNX_?F{7HRIjFkeoa40tOnrpXe^NHcbP$=Rq~ z7wY2cMdIa_MW+|x7>oW`t`ICkK9M9!iwU#nwZDbmQ@V7r-X~#6msxa9uzTZuo<*Jq z%XK>vFSmT1bQUdE%Rb_Kak)&AC@r^`WqYwZv!E_3cWllPmWUKdUACNMby+p|RRCD( z?I-baOLfaWxR+FMPS=A_m!sw+iPDnFEPE0=hk>O@MD8zP@l2Ii+FxPk{5pQ!dx52S zEQyy}Jk!`sHkUJv@NINM6G;-KrQR8w`@?gK?}8@B7H2QiDgM`e%ab4fw)?hTTFJif zn#~*I7rle-+XoILlcM`J{YU{0H$+o8Y8hC}yls{oQx-F^Gxa~d`o8$K`X4kIc%gd`Fp z>W}W*m6sUMEW~FzblHm;)M@SW<&RF)ntUHh&ccHiFaW`SRj%lbibDp!dyY?Zj})7KEXZ?j9-7p)Tit^0P4>6g{8 zdLGHh{QtSE|F`1&@$TDm>)BbLrCUKNjP;TEBwoIHrn<1JS?wP+Ho%dfX0fcIAEEqT z#Av6)ys3uTt1yKFZu^Ue82XbB|oa?7Lz=D1`&w7F3YEb(C^iPEx* zS*|`PJPwwNa~#?WOTs5fU5@Tzb;)T}2j5XNc`=EXTQ;oUgO*~SEf#@g@hXx;X>ntg zAC})Hg5|4q<(|S4m?p7g+q35E{_p+p-5%rYNxa|FttmV?I)6qd{B63hF3?2df(4Br7T#$ra3c)2Bh3u_ZMns15EV1-X1NtBivXK^Xk zw0hzOmeSp?juaNN&k~Dyy`4BboK@(vkOzPZpLenPuO7W=U}`f!*p8LP@;b(tGq3wDgH_#`pPcctMgVEnAqybKtE) zsLO*|gXaoMWQL?J`r}v!UYPbSy!#X1gv84&ZwqVU+4bsI^dw8SyXmG9l& zf#v(2%+Zfy==%rgW#GAEC4iVwUXh zS9XC#^NsZ$VM&r%-fOb?a>c4Y-uWGoOycDhmz~~d(fd#u=gZm*l0<1)$1LC9@3H~Q zjhC5+geCZ^q%NBtv**Z7U)6pFb;*BA;^mfkuh^QL>AZR>)MaBTNusnkFw5wG8hT)P zW!2V7SX{CtmVx~JWF3BAsR5SSwj^F|F>B1e4qRc>P%E&wJCY|ITB0B3hV|!{OG#aQ8Z{3iI-cJmt1hd@K?E4ffHj^@+2#ZO+ z#8TFR)g{R?0N=INb}osRTdc~nfm(IXEj;ee)+b4nmJ+r&_q~2+r-7wxah+&kNtIce zJYkm}S5nu2=&#z-kl0<1)!YrQ(jBp)$9T)IXSlkLEmczH% z&f@f;QTkxfN+$7g%g56RSeLR_G<9Je+mb<&C@t2^(%<+n-iI0UAvs%EtbRx==00q` zwB7tE4lM8Mq!h)=EpD0Y-i40e#K~YeFrFk)T0)qm?jw^IFkh}`&iN%Q`7(?47`C%m zDyaKMu;k>Ec)4YkAG>g{%Y*$SUT)Ex&rTkmHW-9E@$u#)iPF;PJg#HSx9xlmb#c<0Ttis=e@ZNN z2JHUW{th?qfkkH}iI-d2G(Lp$<#h3tkx-Xs8%PqRMVDFhN{;aX%bmizjfBPFm&7tZ zj2$Zjw?Tqei?jJWs*c`v1FFFvBSN=GPiV{w!&iiTVg5Z z&aTDWUsk}U)Mi&9@p8*UXLg?4#{5h?)TMMCl0<0)n}-S{xzMQ!g6_^%#zH` z$1KizumFTk;Ka_KkKl2u#au8<2&$*T_;JDmi^39b^PAbV5vN|ou#m3$t=6?vrEEN z=|6u!T?S4k@p4OpG7WG|&YPBn$I2P=NfM>Sn^_+DF2pX+^<%7UgeCHTq%Pix>=bv= z=RRp*xigK#%Pl$OilZept=cOXV}bKX5~bx6vv?h;vkdAo>`tbWuy~tFEIau1%a^*> zzk=m{5Q&#tjxS(OcleA6$Lsx$Pe~G`rGY)p{rHjNYl7vQ%P==#u{kKQL|kV}P?^!6 zb;06si^R(<1zE>&?oWRmiTC;}?vf-*%Xns4c*`UJEVbfK-WC?4LlVoPOm>ajqtx(| zU@=S~@p4OGz6Dx7>W|qDV{Bh0NusnEF-yOOjY@%K-kbCQVM&r%;ybWo<^GGdRlt&& zOycDh-OgrcX=3mR?_GS%AW4*#8_e>xyB40q48FB6Tv&n+OX{-HjnyUPi0?YEX!?_Q zx#fE{yQfsy?1(1RWkfJZqO>G2OUS%a_;km)Z{ab*;$kMTG{4S{l@&kqo&}a~#h(_% z%PqCmvV)XH%q-j!cP&SfC@r-v;@sD5{u$rkHsoaI_rhXvL}F=LkJaU%Rvxa&^X*8y z+;aa{1sr4Pz8zaZT@qbL5~XD%vmB{9VJ*xT-`iI|3rmj7aAr@&ZY*YM!&I>7hLI#n%Qj|dRrv&Vs|P=+pD!#i=90Q}+QKf?bEkfI4wlK$BwlWL zVa@i$hg_V1--mhqmLySHT$shbMrsz+<=mI;B})Evf6VWw#PYNv8)L7tmSSDX{UGsj zi(R-2j2axiw&EbU3ntu-I8hEWQ2McEcFW7c-zP9Z!&W zxy9@f>$$o-*?%~UvBBp^5~Zc81J3;w!H@4l^J%x0Gn)#_h5uRp=AFdFiWS4I==$fz z9)_;yvl~e!MOXB`fNxl$(q-4}g|29iog~XYbw%4uEQ>#?q3_-X&=sA!kEHv%uIPt; z?3cB>b9(`FMK?P_QvIL0qC*2WBtutp$rB`{x+{8Rx$oE&UGQqkRTxnbHYAZ4QGawr z@6pW1zyABVA9tWD+W8_$@*lgR&zu;&8M>mYT_wr>?24ZBq!bR!0T=G(!N4?kCkg(h zD|%)B@|dY}l@iI&72UyyWcs&V(SQ6v!`WpDpex!XkfikBso>9gb`mQdT5=T3tiFeUz0Tdttj)&Z_wTN zc4EabIVM>>tC+CW)1Y|ozA)A|{UGu3)zirH3y$@2QA7DVB;FSN2-W{mYzSLDJ$WGw4E_QDMmh^8VUT)bjZwFdJ>pItk8V@XXzvxFOEgP66{OUG4 zII2AltTj+r!Yn0qxl)PeWq$4jEwFTtCP|bQCuVtnvor3* z9}YAdDJ*WsC6<9(**m>j&hCYGOKqQ$c)6u?EPKW`y!%|dYBG%>NtBitX1N#J+zaZ` zy?xvyVX?B3ST;9dV{G&LqeozjIdmuSa*OQ(_Rh#$V>f(0b5C!s2jBVu{dWJ6MI?`*?vR_A!Z((;^HI(c-=gfTYR zNZ&|UOsyrBTX)#itoA~aSg`axMB?R^)%+!zTf0W#mBt85l0<1K<%Dzp{QL=%z!Gk5 zcTiZ;WtL(Y%(7nBaT{2+Sd)0UWz@iHIL1zt@wf(c8D&S3C@npirFSipW?-q`@u#J* zM4Xn?C2cyh_&gpR0+v&~NW9z<)0N#VE%SWcNwDl3NRlWm%b3M-R(0HsYkjbfjj(v0 zkyyr0VcqJLcY0)iCEk+6%Pk9yU9m3Hs@F^hOP8}GiPCb0S?1KTw1&EzGC%AnEY@cw zmZU|jE)!S03X$M<9u<>nTk6liy}!9rA3cfHaL}f0G9ow?uH1?5!&zh600Sj^5#Eb+zI zPD$<2g}CeUEri6&E#{k8XXMD&weTIQ)nAY#N=p#4*p6OO1uPrakA5R8Su)GXAMCvg zXO6KignI2D@p4P;Aa)MZV2Jigm@lLDkt9k>KC_rF8--n-hYK8%g(cEXQkQ)-*?F=~ z?vT4M#&kPwYy++QDWIr zk_`{v?jwprUFK+9EsB?0qK2_no<&JX7ojd~Dw8Bii$Ajr@*NNZmPV#!st8Mx%u=@& z`x%vDyz%-a{uqguTi&GX#^cNJ4m}Qm#r`x&qO^QvmWQ4pGr#o}*0HEn<+g(5r!h&Qw6t)>xqocQt94+hSi--#u(&u%EN}O- zCBSpzvC^`rBaV~BZbB9vczJQ$CjXP z$8TDK#dQFQms`Hw43@{U&&?ASlPeO7$$U2VE9KU) zg1VUKlX$tM-qXiemooQf;k}DND@hWi#h6*{CG77FmLP){ON1p=W~noh?O7!LTCN3_ zv+GH`+_LMp6#r4-u&a`~OipLr>h^AZa>4TI zDT$X`=GYxUORL)V@EW;(3`wH2Br}Vieywg$7xxv0TZP5VMPh0CjjhR1N9qm)i|$4e zFSl5>JjyK9OX7QcZFiC+N=u!qIQR3soFc)p*xA=uSgfu|ES?c;&!V77)()`zzCq&U zme%?6(bBWtCQBG&-+f3DrDZg;{P4Yq-ykZsF8i>s zeQ4=gqX54@mgq{7C@l%hlJRiHaIl?&Sg!vNJnIutKG+l7+`^07DfMs%2Q+Hu8b(dJIzp?8V!|jh7!5F*zmc+{~Iu|FS zC3D8A-(WFJAxV@L9cEdvpuP)O42rGy5tekB>zYuP9t%jk+|v6cyM9SJ5H}kv zNlQo)rKRFEocji)%jrX1!sCtJ3QM8Pvf7fJCmWQRv=S_qsU%)*XBiR$Gb#>d~6RN|!NW9!qNtbQn=vQuV9xS1E zNfM=HEwdc4AA9&G;cKm%SvG8dcfx<;ty25irIoT_N#uOQ`!CJiZ)`JB|C5&)i88 zr6rPC=HHFQrw6mIxt1#Z*F9J>Z;9n*Wj6QEx7pMOEZ4nAyxbDljP3C)@Q8Q}b-86xOA_sY?T}#0HZj zN=t8MiR_s+5E^NN4lijgEVpj|ZHIT%0S(qn-LQE8baiygRJRL)MClzfM7i z_pfs#%RhB^uh>})f7HSQ`c0t2+u4z%`@0VB)T8W|S({`Jg%0nuYb4eGslz+P#~~0p zypP=^DgQ%odZr)Ao@HRU)wj^|TpHC(U|5b;#LvAPR@P1qGs3)uvYcohrkw&426gj?tQlcjcZO)oCSGoXG#JIDfqN`XK)@#foc}@h(Hj(7UWd>5C0rm$l>f_e_P^Y&&>Y``^~n2L7h&=CkyyMMva8vkkZJ{RRMUJw;^mgC zk-c%rY|`qRH`KUT2uY%}*f2{<-sJ6Yw(+D{ing%W_)07$CE48{&l)NCelo`nBwlVw z?#;TQ?H~1O0+u`7NfMWuq4SWbv;-YPFB;P*)YaDH<5U`W%IjT zXsJ{G3Esu-v701OTC|wu;X~u4VA+~FXtuBf-;vZM-j3bHe%~n*y9qVEl6bkL`zCf< z-zy~M2-Kx`0ZF2?=yS`jH0;8Oth{-Vu(;fnSk|>}m{_jaNw$r6rhI>^q#j50-bkD{K;$9GS&oB&*AupFQ4yrSKq$ms^U5^g~PSDm9bA zQp1uYQCbR^MK2)?$Jp4!*}H`$#$QsG{A{)+cXsF$3YG!sBwlWLyfF|h!>5MFfn|0M zNusoLaKm+M$w+4psLQowJ_m)x@4m$Hd^8(l<-T9T4sWe>BwlX$5yr;Y)paKZfaTd% zl0<2l&n$zzjunDs-^~G5!eaM8Vkwo!c729K^vj02{Jclv<(5g#1F)5*5B#F}E$1Kh> zt+8AEi(R{y!V>pLQkPrS?707Hf5=O)G`mXT<(87a&Y|V{1%I^E@E}Q)mJDWj@T~`a z>&D;8I9^!%gC&;mU2J!u+0vm+!1C)HiI-azEc}3$kc}TApe{L%B#F|}rv^jc_6;^mg+%`MQPGjx9@STatMBua}3v)p}pG8!z7 z7v>iTOS;VBt--Fv%9~o_PJG%G5-+!8e7}p9uH{?db*1GEl0<2_#VnP4=IDaOWa06W z8h`CBM1)G};?R;EUuxT(*au^*_zDs)x4gbv0e4AvcKsa(4<_+)i<_YhS`wcH;PuOq2$Do;*~KhL-KW0?i(!B7Ho}q+ zE~$&>CwAO#u=&$wuuL97;^mghiyxt-y;G|}P?u{HND`&Rjag=8^#2JKry=FFgeCB) z%yO0OEH3UGjpxaC$C7xtW!l?qXlbxmdmLEorjjH|OCqzBZ+9scEVXm`_7#@P5fV%P z8|)k=HF4(>s7t3`BwlW*cKkJ3YNWi%0n6{w{fd5s(o(|{=l++$HLro?QTBcvVKI9q zv1GPlmZhgU;Er#zUnE{`iCub)Su)DIfMu=5fubLwvf>Mf*=jz%lenTAiI-be7&OKt5$D|Soy6y3bS-PSKtd{tmc&|y29f9LSm^D z#-6q^{v57-Kj5ND`$bj#(aOp6d>l3N4)tgvBOOV)0z6 zhkt3ODivJ8vb{5jms{Kqup3kB{D02?%fntIiPBQ}CeHl_jT>UO`uJN3>x9MVrNpxC z4LiO(GhDI-ELQ#`UT&#Tf<4)#(P=QACu=<>NtBjB%u;7luM<$0d#$_g5SApFCEbG^ z8roghi(Q^0yOMahC1fyrPyDa56RtvCDr%D?O3ONCY42L&4Oo5}o9`2r;3!F5cK%>% za_PbGd!a7(%}Bi5lIA`G2kP7T*`>i!<^)Niv^X$Jm#q=5V5yyX-&|N+q9qpF+U#lD zQineu1IzAw5-+#>3SlSPU2kWX0?W5z4~l+-(h|iiJEzvX2bOcYD_IMRMU2D}x{+Bv zooJ5lSef(77wZ67xz@m+vm@oPXZA zh^k=miXicFi>uygoG-Ob)yD2y#~6}CX<5lE){{C8gSu?a$?y;szgUSS(wg13oYW|< z4p{pBCh>AhWJC6(igRbnN>G;pWipF?gwkTmEc$1<)CNm+Tb(2aDGkl0<1K;e~U5gyFn?VDWi<{i(2|$}EBW8+S)PK6?t5 z`yWWW+|slpyUCVyyy^$AG|3=Il$P$yVq;Ke4rGY@)EZ#Y-&1v)Ey-`^JC)V0jr% z;^meGwb?or-K)O?Sl*<~EBX;ii#4-M)=x|UOY>aUG-0uNE3v%)v>5-=$NHCy!E*CA ziI-b;zh!g3_1^O>z~WlQt>{N6Eg{UZr_>Z1upC`oHdk2kWtN$Z*_!-l%y{fp58OiH z<(A8Duj3!0n>TJQSd5HF5~Zb(S(=Zr!@i5fM+W^8mbe5-U0S_k*J2Ag?89$E4L?QV z<(7**&!eU3@>0HFDRG`8QCd2A<2q(wFc9yOqsv#C;PvXSU$XFYjR%3Z*v%9 zm%EdAxy3_+o%1i8Xi*+45BrfMN{cSD#LV}&3S+GJ#q6rW;*cn@RGr2y2`d{O!MD*> zb0P6^%lgdeSeFlVd*MB$1)d~{(qhRhPQJn2p(!}=TKyKn^6r0@zj+_|>sz1ko!E`; zWz>Og^%XuOlcHPwbdiN> zzy4U#&y~=vZvKfR`H$V|3uc_U4BhJQzK~>pcB}7MTmu&j%NM!}VPK~HCJFwgTYbyO zdYGwewKb!lTislvLD3NZx83T0{J^$D4OT<9`nW12?LT*`AMW6X-Rd8jR=x<`>Nzz@ z!hhAR9v9FDyVa|N_rRJzrI1mn+&vLFM2`&~`gXqWWKo1+xETcR;^z zu+;Hv*;QE5WtO$C*>g)F9+w=!^5qVRms_e28H>B5zpRtLfTd;-NuspmF-w=iy)D2} z^5k}HVTnkR)WxnH`%Amdw!rr-Z9h-q<(3tt&9S(>a^4kyCG8SPqO`QRjZ=Q*j9Up{ zak+YHsIYi`kXY8OWF2_9Gtc7=meEZTFSmI5vAUG^dG-S=7w(WGO3Q3!F?peL2`pMy zG{y^yb+W|LeH^>em>&~08pc?j2Z@(ky4{J!x{RALwIo=4d`S|e#f({ITEG7W7N44W zGlZp3X1QLAojENrvMURgv_>ReZW&vF?Or!*I1qQ@OSUFSl$N{9(j}%5-u+oJCRtZl z5>g~}3Cm+USb5zGr-EhbG!iejSQQ51+`o7${WR2N**ua&X~|-i)d6uqP?zO)bCwHB zV5-F8@QdAHG}sW@A1qN1v&B{1JUAjc}-zh9+ zpCp#GCD@wWGRv_G)Ma5jiI-d4SC2z!INKk|-_vnZ&lhmcP zFMEHC)4YI@U}=>=;^mf?n$55-CtvM!0!zz}B#F}UiCGM)%)sv33%TB>g~dBvVu|0+ z?#jIhJL(6P!@o(q+%nVmAX@BuUc~p5EGTok=tn3m4SaF#f9g;PpCV}3pwmTRvH2{q zB;>HJwaCl?ct_r~C5e|?9(=ommfVo`_~q8njwFfFGM-tEoj1ju#ZTotuL+A$hQu;_ z47+Jt$1oC~QcI{#;^mfu9|odD_e2GJ8ft6J&*F!vuPlX$ts?+3fX7*SXe->>YIL6Rsfwfu1I7koal0W3OB`(lK};)}!*vtT<8 z^`^szw*t!(R}wF`>|UXX?^ZuEZnO{7rNE0MQCdbaONoqR?2ypf@I6sja%7g=ne13O zrp?eZV2RmD;^mfJ8WZsk8Pb27GgzGWkt9mXHfAa4k%8|gJ9v6py0FB2mDDBHi=D$9 z9ycDpwdZb6;^mgQ+2hbMDEcry33>JkNusp4FpJ*y3&mlKO`LckPgwl2C6=f$?A$W$ zjSW8489$4}%PoP6_M^q@<2rm7PQXHvL}`gTJHITZ&R3Xog`6Os@}o5Z@FWN1JotPdsPKtG0Bx!O5bN4 ztHW;X9s`zc4@kV+5_y^}hW)SE;(eI9AtZ^?GK5)PPh5|0KzKaUwWhG7$}FFJ*bY{? zq|2|sQhyPNms`C0v8U6wJje}%x&L}ONusoDV3zPKi_=h-%~h*55tgt#NnI|qV|DrL z*m*KoUe_h@a!b5@IF2#9(cf{5&1puGC@oIRvgWBBc1ET&*wRi|+`dUHkrUbV{;lz8 zAHd?#ip0w;NBXYE>-`1&S5AStZ0(;W-&>kwAeAr&8gp(LtQG)h}IPrhaVD4=V$EP@|cz< zKB;1_OXB4g=cBLKACeu5o%eT_kt9mXb7qSZ4{PtnWbt;c5XS#G1mnw z+Sf_E++sV14b<1`($0cq#VwLVY3a!<=AW+OntXR(Ya?Na_$jH&p;2sK%rL|q-{ug1 zl*G#|M|ZrzF?Rm>qvl}oIz^HwEz6k2eRWCfTp5vc>5#B^{*qX{%dq4A&dZVKp)Pwq zlX$u1aHcz2?Cy?x2^P&fl0<1a!z?Gw`ptv+@}XSW6T)KsTVh$=h0T|!iN8*O#jHAs zms`GC+(ygAfzJ!UGPgcSqO^oDOC_r>t-&&U`zbqNDU?|<&oj&X2BX`6<*5mYms{FA zJ&u;0JN?4IQu7E&qO|;CmMS5MI$(MDsPa`|Nifp*^Q~t|$Jo8T9=0`Y!Ls8SiI-dU zrLcwl@|Q38gsMd>NusoL@yB&+Qs4IYwC%w`OK%EG;2w#kR5NzzvH9kVwP1-EMdIa_ z7FCB}UFMJ2dmrkeGnph&TJ)I3?p;(*s7w6|H|`3{Wn+mYWjQ+^)Aiqld$0Geka)S} z%+J1PG2MLh8Cc4AkR(cr6|xq+|25}HqvC@n$E(%bX15!7Yh)1}XaB}--*)R)Z{i?v}3!BY7ZiI-chhg+g0 zb6WZ?u!JX&BuYy@vpDy-ZU>g;7oy^XCDKGv7p>pyiMu-s*TsOPnhS}STlU(s3+*=> zj^6~!a8Hs%X=!&K=YD)tza?P#r8h1`SiJX1ESA>nxUZSF%^EC?R+4zRCDz>@>+(>~ z0Iw^5tS3p7mbuKLWz=L1SeAFW^hH=~_Dd{{@3OhyTdOsGyW@Tz5-+#ZKEm!N_ewr` z8|w0EFiE1cm@|t_r#_XzvaEF7LSZpFAhE=6WXH+{_S<}*F5Rn;c)8_40NWP}__4Aj zSR87TBua}vvv{q)fW1B8eU56B`|Dm`lFV|*itX_Yn{-+eEKP=zc)8{ACmkGP5#^FL zf#t<$l0<3w$}Fj67k7cWY#EkPRak;eC3T75r|ZoMpW~kR*a;+FZW(7e3N3w&rs6*G zgc&4>($eAq&i$j>Yc-+CF~Vz52VpTgsPT8*w*$IIVfSsbfyG-x_ifO9l1b5hYv7!N ztTjQdrJ?(Fe+bF)Pu;is+)HEk?Vh}+PoVp@!gG@D@49bKU1Gm%+a$}I(0yz5ilq8K zb>B`uU)d45Z>zl{Db?M#KR&17fB&=HGJL)+F_k0|BkGUt+cI5p@vk?$@~JX(-v(!r zB>%Dd)-BH~4Z3eF@<_5jyKfJ4ZGZzarrxZIFffgOk_3O#eOt+(8D_dZ*sTwC-;e1A-z?Id2ldfMFZ!?E7$NM&pIUlR6`B+CC~I{ROuowM@cNMKX9cpqW0I3%$| z-DXGV10789j(peVBwlW@tj^wlnXx4CrOkRQ)bC5^*|pij`wa45tba8#m|H7 zENbqYn-53mXOBs|+){n?NvwQqO|xhOJDav$G|f2_p-6V5_4Emm(q*b z?n1=IhelxeUDlu|UT#@DlD#A#K-+dISejQONtBj!X7TRxqYqe8W1deF7C$qI<{k1<*%EGFg>i%VB_ zWqxr)>@2W2B$9Z!u<9azl1y*Ct=RGH<>FgEws ztZIqR#nvpIR1`0_G>W=_mRnsXKY+UAl_g1(7EflWJWVqMEIl9h-!3d+M~^q} zTtMRGmXnp-FiE#EXHUTxOI<>eC@l%hQh%L+2GnJA#l{}O;$S7QoH)piFChy`;oZ^~ zHA%eOa_)pVT25zI?gb zveKQsBw+ugv9rMPbTmn#v|MJE?C^WIQ!>hL_$y)YJSDL>gs|hj`OzTUDLL?v#LF#( z=O5x2tLRYyy9#QAlO#$@EVCR5(HaDGY1Y6#QCO_4C6-#gY#r;g_zS)_F0Untms~LoE}ET45~ZbL5YByV?*#0^Szj$WO;`$LmNNU<@kRH-vL7(U8r3K9a?8ut zN6?b}wJUyEuSRo{L}?koEZLXz@##V5{GYkPl5kp5mzTBKF6o}}ecprRcK2UJ@p6mt zhGS^yZ(m*u>eAbQBvD${GRsJt=kF*gmww|ESgN!p@p8+& z!R%NW-ha(OsLS@wB#F`z$t+C<*Y<(A|6%)aO<^&!kyzpnvd;VSU$^uCOTs=9FSn#M zV>`ausfTugrII;GqO@o{!nyB%BM0j;VBV+t!jdJk zL_}4GGkE=CvxX#5T6#0fs^-a^p)SqVwrD9Vk>@0J@$1iy`~GpKv2*3?RuV6_?2mkj zbN{Hb$#AGkZDW!|X<5!JNwcet0?Ww46`h2|+g4)vn$E6YHrc;!0(E&=dSy|(+>+3L zBU*wkdEwK8ag|9DrNxF>?Cai`43<~!uX_uN&3TEXgez;%Jgj{z7c6ajNxa;W_nfU` zdz+QTr*3)$kR(b=II~n)6`BZ^X*GuF2#b-O#FAQ%-G|wHWHojY*KADU<(6lqk78Xm zM3f7LF}A2RNusnA3&y#>D!6wRSn9`GjT4q6nI*%EeIw${$K{q#m;Co6UT#^I@Bu9c zGkV?t%Z7B4L}}4tmcEg>_}!k3bAHVbmf#DLy8K?q>QX^tExu(#Z!w9NTMQoBqUGw_ ztubKPw2CBATJ)Kv{lO-ep)RleN9qZSi@n667s)y#3XTmM0(CJAA@Op{kj;0kbUB?Q9R0F73PT7M2(XNnLh$vg^u~6+@SS#kxC*ms`qP2BKw|_Y-`2@L_+F zL}}^p7}v4HxA7KW@pHd^NLc(FC6>fJtS*sPQeJ{Z&zHo@E%&p#;Qg_+-3H^+wu1sl z5~XE6vy3*qYz&s@$a2;FjK;q?=qf^-iuEl9zI~Ze=UXdh9 ziv_dzPk4{t?g(47+E!RhE=ep+kF)C+Bi$_A7t3Eo;^h|i^)qmc?XbR&d#|51k|avY z17^99*$JPBDADoKC1FXGS(epjV{Ad>w1xRxy50|A+%^O(Z%;)?iomuC@ne6 z(kJ>iezR+5YDITp3A-$*OQID!E&Q?J3f>!!_(J047Rv^{Xem`~56=B--$@dsrBw*d z{ZHmw@P4xC;vqi5;^r)|6r0J8mHk&al!3Ywt|swv%Z)eeTLhzyrQ>;W+07)0(lV1- zri}Z(3FiK?bgKYivAQC$>|Dg^a^?Clybn`9kHpI@M;EaRx|nJ)_;kn7UnGgra)?>p z?p*a4EOTe3hY3r*%+hKgn=iJL{P7-az(Nu)w>)YSiDRtz;H!9Dx!!;zQCj?%rF7{z zVPGk}vSF04#9futrHnDVH=f`yyfTcjZ<|QG+!C*Cf|g-ktMT2P= z@vSEczX*%zb%|w26gwX~G1eK^p#I%;E4^_h5{bHz4tH%a#6@(Xzs6M`svgT81Qv(sGMgrWgNh4VLBA z?MhYn>mF=`tE4W6KC}0_sc{#H~@9| z^|E?xVX=0XSb{X!JtgO_jqvSSDQ`)<++u7q94*VoXm13|vlNmhQji zSspvQ@3~a*g%0m)1|;3zb$Fkj#(vqdX&RfL!@KM{lIs7|;k_Zfoi%iL_t-{K{)Z0l z@h?NKz)Y{bmn0G+>W>cZAE^wu`rVwj(BVDxFiG+sJG{s5N=t+e?+M3AvOhb#bAC3* zNw&Z=q67@g@n=bbzv=MqGo~G8@;ee%2p!%B>`A77+u{Ak58M@9X$^FEZ@Ws;{&R=- zqqUE*!`pC1XYBAE?LiX$yAJQ13E1J?Vaqgphecywl2f%x7;RyzL_(+4@1euHUI0n+ z-#WZc_6_OtU>)iz4yz)$>i&iMY(nia1#sR?pq> zB#G*ODYlFKFB+b16~L0&H(yIw5^hM21T)$)%X9bk_`W6IVI*E|(Z9iNRfJhroDY`N zaU_Y-;>IkY4Qu`Yi)|a%{=yRIA+bbFV!PMwgC~`Sx{T8z@p4Q4V)mYr)2^fO9jmRh zNfM1Hw2felD_?99o3bUI!&8Jm-Mxn)N?wiCaoVd__?%d(Rs ziPEx#S(>)^T^=mkj@Df!ERnY)b$K;@1OBC5rftOc`7Jt4;^mfy4(x7eQs7{G@4~GM zB#F{;g<1SYU5f&XZtyfiVe$5oSk`r9-BD|%n^uQ0_Td?cms=Y5WnDPC%?IKc?Uz?1 ziP93sETNOSV&}@5s>ikoi;cI$a-b`_YD$d0`volZ>XCT4W!_6SEbid8rxIX{RcuC* zC@q!4aqhPqc&`=IMdz){K4CGsEwNPH&5kcRVU=%z`yU<>=JITrYhh82d5wsLO#>?M@3zu#coJ6H2f{c9%;x@qHL`Hz_Mbv=0#y~@s(K0uVZ%@Tg+S;33Zt@fyB!#jV;;ju!D+7t_F3v``X7}SYqx->asSP)n(Gd_qZmPO(yYj%bl4PXemB?SqjYk51&aArDY|v zl<2T}C0L@aoevQfzq=C4u}t{%iGDqV&X5cl!<1?{d)~Z;1@#s+mLvmuOo3NzHEK>`1<4~Vw@KYb^GIMe%aW_eodLqMLi?tB@p0i#4;1 zAAGhWSbSe**AW)00EuOEDm%CA_1K_~XL(-p=fiPECWEaT^m%78KU z_(P?k!r~Alv6#IXzevgK_6l9Wkxy8hr9rxF@t6C1~(l4JRQCcjSWoLQTRc5jE6*t%-7a9saUw~SmNw6E?)%+69u0Nzdt7<5 zuy{U}SZ=;ymOf5b+Jfari7G|$a!b^5)~#-~d8aX0HuWS)l$P1dk{9WPZ>G56<7Of( z)*%wh-Rb-@@?4nVb1mr*#?%fPe>A_#f(`VwoY9Hb#XG6iTw-x+yO&+)3T1cpg0 z7WLS*Si}5oL13x3lElj`;Rfs-t9Kk9r-G%^29iW+Y4!r=e$NSx%b_l=+bwPi%jG8$ z%j}x$e)6)DQ)0lfdLN0GTTBPACn6k7SL6FHj~^jPl$NQ?vUt-LJWn3lf9pMAF$lAouGZ$3KkaKU3?$&y*@ zwAmPId$<7}EAzrhyxj8cY%2aCjdp!-1xxQ}l0<3oW|nh0>u16k>$_@qw6H`zmDI(! z9lL&cJf`b3s7vG-5-+!`8hZsT-?pE)0v7LJl0<3w#4Lv#tFH#jhl;r$gvC2TV%c(< z-Mbi6@o6AfCeT#D*biu!sxlg zVmXYR!)PpDi`U30&Lm!LSv<-Z$Jm|FOnjGT?p=~ZX)$7!hK)Sp!4iAfsG6`O$t*K@ zb9Ub}%`B*k{S6W?xBMt)iIy=J*5dyDW*?G7X}Q5H#`>0fp*g5(_`nXra`469cCB4X z_<&t&abq@|hpx4NXp%|MwN|_yYn|HoFzx_!t=YUKS^lYO?T`18-Vb)W1zl_ADJ0$B zb*)9WXTR)ue)L}GT04_LQvIL0*81FkR|C4%X6KQV|DkK`_m&wgVMN{eNfL<>^=H@G zV+LH(^tL;6t?ev5sAwqs$F8+ulUlBZt~H}uB-x)`Yg z{0&L?ue#Q}Yc0U8wao*6dc*2HG>PO?trA9oY?Zj2UVS-qt@TJFY5rT+TG@T)uoruD zx=T;7;spP1#rfmCq`jQkRnx9|Yj46>cWQO5C|Szs~RN#f;}{IWOE^09DMG+6ZZkt9k> zt(Q1W#_yYe7K15?Glj+CrOZ-_U1`j#cN@Rvu<$U6ms^bOtK-?mhb5nefu+%Xl0<14 z$t-2;8m2>Cj{ll&AS^jDi^E8L{7dJL@xu2!Z>m3|C|+){s8k!1yvWYY150>!l0<3Q z#w??YO`i{z{KSh}g(W6RQkTi*d+=x6uz!Mg8QOj(@p8-LP*&VGXYalMONm^PL}_th zmUf|UQoyo2fAJw<@r#yN;_mK5%bJIcV`08D>$apQUT)d&l|6qs_T~f!u)JPQk|-_l z%rbJwX1v2V%e3IEu-L^&EQxDaw}^yP6%WZ#EY!t~w-MZW)xp-tDpe_PKRX7oBk=iPAEJ zS^S1NW9Q0m*MvL5k}9*5G+{efWgiXI1IuO~5-+!e`LoN&rcpZh<)fnyND`%G1G7}r zC^rY{a@u%%xUhu9O6t`^-sFSnGxaS9j1@-=6D1m9wi);^meD9_+Zk_1D_9VA;}!BvD%WG0V#z zMh#$$oe$BjQu(jD3~_HHb=mWgS)L56{Rrx^IG)7IEptOyYr@ywG3~%IKAj{{T2?d5 zrb`*v?cq53SQBCKkC#}6e_>0|o+s~c2W#O}5-+!mY5f_;nECgnJ-~8sE=i)a*fGn1 zM#r&Rz2l9UU4+Hqt;ABTJzK|)eZPZGs z+;^+{<`7s;jC7qOEa@^!&l>DK@z;-@#rMa2*+b&xmNi}5qGkTUZ8gBs$BQIUT6!`| z%!UQHgEipzdp%)^cqge#?oxIRGda;Y73vcBgv84&4NjfLO%;<}Y2QCiM0 z%O~Ac6~VHl<)VGUV*OrX+0MV*y5M-kMzB;_MB?R^F`rq7#Lpi?FM?&!DUw8K31gN5 zmz+$n49n6xAuNS5i&iMxeCjjN5Z~)lzx|@3c)4YRt~<_`+ddBlfMwMzl0<3w#Voat ze+~wVS;yL!ge4(KQkNQj2KY0IH`2#WuQj7cyxekWffXjv_cX!x`I#&xNtBi@uW%jP zd-&Zfu+;CY;VmqIA0!snt?b-lJ-MHbY4^Wrx=SUKzMUPox zCa=Nn+v-E>JQkMA$r8)dNM@NdZ+I`T99~7@4O+>*a`7S_dk>sjn5n$VXdQCix?;@r== zcW^dX?g!;ps`A%4zxPLpW#2JYm%Je>H-TlyGZHVij7e#Mb@^^sIT6O#-dK`EX_?C` zR)&kSVAG?6*U16Ga_ZCH?#7M3Ujbj&Q1NX1HrS2Z{eWar?8doX&&CqXDHu2acH?@6 zk}UtU8+UrdFWe$LyTkth?8e1BC+YrfH!i9r+pG$0V|NX9lxc*KZ(_uF* z=N(D;A9mx8m^H#TtOR}}iNuKdV>j+pAOm)u_(m6Y;|j7!lK;3H*QT)LFW8N1)5f}J zT>ZHlx9Vg!oMg@|=4imcwCZxYD8b+C#-+GxW2V^`hQEZ}xUs!RrhmH|_s0(m8diEE z?8faIK+^v6Zd}`%?{PP7^|U3nup8HAI7#@gcH^#Cd*E)|jCGB=!7A~39LcF#C5#TU zRbtZJR5RF(vztoN{I}h>WcSE&uzFTsGIyX@acus#;{352cVR2L%TTdl$>Xqk-ds!K z<*R4CdF;}wW+~e>@W1$NAxTvKOEGiyzto-6-VrRFiYJX17NazYC8jUyVO;4NI~Xh` zl}NnYQgtP}(&!PqY%W;r>XIZ%i$Al>ujsA^7L&z#3xp*}W@%Bt`eioEY_=IJ7cP)^ zxn;@ZcX=(8qm#=ZLyZVA9d;LHoVR88^v6S{@XS4-7f0}{i$}SQw zx0J8Io+UNQx0?cWd9j})QCg-mOSNmYa5t`1ooPpf#Uevu`PPbU3ij?d(FH6Ktw_Aw z(rhzZKytqJ#yeT{v`7-A#gthlRA1Q*>f*j)`*~r>ky*;mWOsfy=QbDzmRe3EUT!&m zdJmoj-trq#AL_FDHA$kh_%O@mUggGu#p!cD7h#FXl+-18F56k$*s(T_F^|^$i{j;$ z4c{uE<@k<6cm)_dpCnOQ(wQadTL|7U3BIw{Pgwl2B$mv&Y)#(LWe$FcQtu9lms{44 z9E|H&t@8sXL0x)1BuSK(MsYaz>zc&lRa3^tyf9(0`y#QFZOXbm!fVGJgfaFxo5ag4 z3lR;<*0G0{YNfM>SlUd62-h$8hP3`IPQ&_^XC3T6)XUpjoed9Y&m*jyYUTzs2rG;~U zi2Lv5V2K++k|-_7%yKlZ^)*<>biOsHQuVJZbGICcW%dR3Ws0;)8?hI=X7Pqa@d``E z6dYs!kF)!Z>-qct$6rZGB73i_WM^gXy+gj7hP zq0;!iJU*A-?|Hx7JI8qU|NVY>T)Urh9^?LaoadXY=uKd$phJ>KEmdCO+;3dNr6pJ% zXSA%NSUfTXmb$apI+i_bq&n1PTd`KEc(LVfIQtI6ZKirBurw`8l1MEBnPpu>;ykdN z$W3giSh7WyS4r$C_MS#-@VnSinn|j7vBgH)5$keocoGqQl0<5WXO?{yzEzJy%tLtPxrNfN20+-scs-8DyS0Lz%N zsS6Z~bB@3=CY0TarMJjD0Cnj+h{TI6jaIQYy|ny;Q^1mGN0LY_y_uzS?_>Cegn6Gw z>lI6i$l`W}-M_frbyyFUjM*e!Z0VN7Hq@~?&>r_&I=+J>ky@58%YsFg@1QPL=eKTG zEDv)9b;&-+)@1wa)C92X`bOf#mc$9|@LbvE$P#=DXL=)d)j>!te#}zd{o?|#eEd58 zkYWk`DX^3&$F42k45X)Wc4oUFvcQuCa4ZVYI(^l4F`3{@0MN} znglBruRMX};B+=TYP2og2I^wmk;IEFZyU44FxmeS{)$u2UL=XsQtA!P{hP1XxIkS# zn?1dzSn@=c^hxYfV=M2(reJAmLE^<0%?2r0ml_T1HJ~oNXOJXPOIK!@&}%R5f!AsN zt|y8mK3`B5|7q-8*?dOry{3M;$Zm+-;UCGO_E40^O<-U4j!7 z%k=_*r67gP7vG7qo`Ge3jSyA5*mAAX2eeF6>$el?vd@qtky;KiOI&c&MzA=2~Tc zUGs0YD^c`$a_$)ROsUD5p##Cv>n4d8TXZ@Z;hOAz&~FmdB|M5Gky^CkaqbUZlbHaP zm6cK|DwZ^n|Wu$10k z*g&y7StqEAVJCJC)8lMfDX2@IrzBo%3Gxd>%SoNp8BiDZQfdwV|AUZPW-^OW=gzC4 z8TFv)!VXHBo$E{dZm0b=&q`pIG&?qU6Ld+F+L26>E~(bVbY$%~G+`WcNrStREPv{f zn$FC^F6p_a{r#a!ItX2jdS-izcC+L!T4I!!ir!Hx)pp)LvCGBQPQp&re zo1$MMbM};7`7okZPauhu5mls1n&8KP2~lg_L6@{(DoOH>UD6kCN)3T7=}UK8)wn9! zB`s~)2^WkrRn22zU_RNmNR{9>UDBa@dt#<2`*CfcOWOJ{$@I5f(jo^q@MTCKbVHC3Ve7#4hRfO|P~?m$Yv%N%&V?(z`d`Vwd#d>3bz%mB_q6a>`Z-I~TS}Y3Y{kt5<<#O(97lwHPzY<1$rhgT?i$+j_-v+d)tlGh24G;rGJ+2UtvUNW9pR zv1}rav6k;|w*!l#`WDqeNG(&Dr9h+MH?UZh((zC%0geJo-f%X?3N9JoJHM{|NW9qc zW8nn+ZT0GI%PWELjpeZ)JDpmvb+*hPt%U$|5mkB@nXy5 z=o2``^o*zDj_!>HlO$40F0%xN6gLCQP@}516ibY=pe~#0vH5cGON+T+xiyW%i!C$D z^Cyh8s-K6tjPWK(q?V@da2?a!@+=rEn(xa#S1jQh1(vOK*^^)64%HmN5?x5*#g?x} z-LWp$7aZLO7N4r+R0kooOk$Q3?zXtQV~x{hNs7g1lfV+#irvlrRO^A?Uxr;K@nTED zdG?}xK-Ue+!IFB9B#~OSGRv66tGMIt)2>-R6pOu!z%uD8+kNZ%$Z?OLF71|)c(G;3 z4EA36kC;?zu%!EwBvQ*&W?5g$b{UMZ!XC>@l>7CRAz5T8m&3jzf10=<6fC#bka)4h z;w5`xqW@}B3$R=}N0LY_8O&1KY5E(mY&_#vS+U&PEU3%J@@(#>Hm-r^{ly6+UTo=H zwh7Lc9n-pafknHDKP zudZTUW{>JK1}rCrlO$5hCT6LSZ2JK$2DwK46-&0rQo)X$_dm7TkLQ=p<(sJD#g=0M{E zGSZx+v4k|a_~odlfw?HAsk0hTq3rmR#fzU~4`iQ?>O>41a;>|SSVCh=m6ySgD- z*7v`PzZy`{pCpl5MlwrFGmA%HDShb3X2s&XO<*aU&AwN*9@F9o)aB|c5-+yo=dfE*srCWq{u6_VxPMH)#f1+Qi`Nc;Wl9LU<}bC!ZV*_$*^zj$WzAbw+!2lJ z)`4Z*Mv_Eo8O$uDQi8EAopTPpQ!IHROO!gh9@}-x61zSViXBqLi!J^qeX%YRYE&2t z7JprmL~2>XEU7cp@49RXHRE) zg5`W0l0<3=WR`^4x|P8)sd-4T^1rV6ukRFCJifAP81*K@&w*vuG!iei=qLBax^&!} zRvRp?dq@(gPL3~1m~oAtS^VmbMr<^LKu3oT~MZDC<$wRoY0)$*Ptwbje1 z8EB45`;OhMnM1x^fX-sKMI@7?vzYWf9ZOW$Wr`Ve7R^?XEPv`OmUT_VgYxXGwhKCo zGwn&b-*pxfYO%)KI>$#A&{@oLA*ueS&fSm;!iZ|R zha^%)RME~N|NgRqrspE)EEYdVlKf+5@%Z|SAEC1tHTR2ZTovss9&XkPCt0(g3#KqI zhcr!9CHPHevFnzBm}#o^PIKri`WTZ;f7@9sa)9wo8`Xr);+~Eq?Vme~4^Dl?&f@)n zmCs;ju?I=`SDnSdXLWHp1=#;@G}rm7%lfGK{48x6WclhcAh+ zdZsU^)m>R}(nPDLk16Xeyxcdh3yk&Gfh1nMda5_6glA}XmW$gdb z<9tipvC2KodAMSEvRiN_@C;-RIup+vDgcYC-bPit*m6UwGuAlrX7>wVdC`a@ky`wj z#i;rje1|dRWu%2-3Ed;GOpa&YlvEhyfjcBPCXjfsrF-UfwCugU-w-Uz|Y2Io_1X$`hk|a_~=_Fjf z>aWwYgZa{{dItx^Vz*adIkJO|F^|W4AAm)-7l{{JOxv$P%fqfOR)VGfY?4H3>CP;z zVjk}UOUG@dI~7ZU$P&fhE8kw-0KZ#uyG`Q7mW#{SJ6St*_4UBgDUT$PS{5>kSzZo~ zu^s*I9#t&2_X+A^^`2ePu1@eUgu1kLBJpC&RgExycRno~g! z>SC=#(Vcd_n5sS{!NW>cHeBwlPupQny%vWIgOe6PHHRgy$%dB!X)%hhTC zmP4Pk?kSe7J_3tJH}){S*q01^-@jG;5~_Hy<)P^#tc#`hTHGDwlsQQvwP+{f+}CO| z8^>6p`TaP>k|DALe`Ze^UfSy9z!e*x{ z)WvhVajIg8IUuM@%^dcy!>Ax(6Ihx@l6bL2Pn~@ezu@W-+@;o0t+nbPq?XytG9uvG zZm@W5iYrho;RgklS5w$>I&=OqyaV*TMdHPl#{TtizEnx6^9<^uS3r_TE&G^d#?XW1B*dj5+{&&vE@;%R%o%m_qiHaayF18Qp-bT8Krvx zpUQ<9kF2X$?0p56hYi^Gm+@AEU&=6k13qG4EK;2)0-rbS~Nc5+@JmZ zD}GO=y=bVhVo4TRjCI+X>@)QGGO#RLP2$Ct_{WyGjancOXfmmNv|CI%l0d z%$FMlb9*V4dxr&eF&n|2{0^y6>oQo5YtK`~i!J(xd!ePn($Nku#?19e5~;ToG9K>|;p0*b+X3U5^b_zj7DqQcClN>L8?+G-k0&UHt-@9+9z) zS}K+^fxqp>t+olkj?i;0pET&kIes9SB;B~x9_d)3Mq-MIHdKjJ}Y zwVF5yx^a&)NxI*4h8}d|uI7?tVyK$p( z2I4|~^y=`VFfe!bZmLS~n{J%(vXPicZ&_g<=*C$OBANcS8&~82CwP_~2;I1VQ6%l3 zyK&!pf5mRx?139bK{sxYIZ60e-8jF`^>8}noO`APtHcEhl2f)y*i}uzf8qQkUUi`x zXKO>!{98BfyXPVN)+g@VbYo@3x&GgZQ{+3p$GzB<)01z<@lGQ@g2ang&)3~8@c+_6 zyXpz}zXaYUNo4;QH3Rm4+0%As2{;pMz2wkKvG|=9Se&id)kc#;ovOpURnH>vV#|(U z>~ylRneig9I2$>t4nk@f%`7%=S0;mHcF)Aoie+<&Jrma?Yn&MTG#k;O=jeX}^J{+%^osj`a1i!EDAAHp#< zTI&-2-ck>5l0<3=VwR|h@!i2<6|&&2V!3@zP?s;cZ0_HF-W0pnXVOT#*y44Qb+1dR zr+YwM4%T0&ItZyHfmyO`jPduDRxDf;t5^a;1(rAE*i!~4jay+*m#@c3yx20jSpn8% zRoP>>gJ?iJNg}mWO2xT9%c&Utp5L)nlfEdHt>*=ni(YIU8+-ZG8?bcjOyb3s6Ybfz zKF)VK;wsMjIoljHu;JrLu6@rkF8@JhSvxIOTz>bFSd--9glSx zIQSgCTiPLoB#~NHFw2LrO^1PHf31Uhm3}>ChzS$arO`^Zj``0u!QaAZ(3!-GEpHNr zprw3=K6tLYI)@~YS^}7*;@p2>9&`jOhS9d&6pQ^ufn~cs>&DH`%=7?DWGsmnTc)3l!R~dl z!gdp(E_w+hiPX}QSthQ}Dh`&+^smDeOR~sf{Fc4%KXzS79AkNEX{va!<=XMXIL0z& z-P8d~K`D|%YFWfASB8A)0hWEA?prFBdzS=txns*LXXXyU?=X(nk5a{pEuG>zU|mLc z)xx*d7MPPHQp*u;dD0ZWi;Z##U8Gn7FAFR+YOq^Er$rb2VT`5wlX$Tu>*s2;oO+p$ z>sbC3l0<5WVU}$R4q&ImD$UYSv3NuXEHh$RH|}~bt2toF&LQz)i~l@!L)WE>t|yGK zsv1_RgOFNErs3Rg{6Vh~jIqWIrg|!tY>_2(7CXOutUY%mSmsYA@nXxiFYMZKs@lBU zV0q+0l1MF`nWbP-({^Cl@pS)D#S(W#P?zQo8a%oen}(oV$0NfCOG$tZ`E%K zmTt9uR0koo%w?7U^%uBvcw?>Op^7Eqs=#v1kXawX0i5FX{II!aGX}Af$ z?+-K~Nu(AZW_f$qc@flQox!}@ipBSuz|u6GS?W1IRR_z|aU@=B=~n74)@4h&VGPuz z(-x9MYKdl+MS6jmUDg)th}*EGZ((`T+K}dg73_FQG1V3`xAia=9_C$(z>2Vb|xF zF-an|v}2ZiU$R}HF2`n!$Wts2ZwTtrFN&RCJbs4Zn=8$2NxaxHX<>7;)GM~l1?o~| zB}pQ+*f7h``EBQb#WQemsmi~u`Gao?EM97Cj4iZz)gJ0HD~!a8Eu%m7MvGoRr5vzC zT_s7Rmfg%UVxj@=>*H_Tw5DS5x+SnIddl9AXwzqWL$Czfl6bKtU>o};zQ57!CScjO zh9r?%?lFtH#`S^Fz{#9mx~XCb|IhN9@BB(miozzI+36NTp|jX)0m&rkEKaDy`gdhq z_B?>jVu|G>%bz-nXY~{CpvGORc?LR*{&pnY?>dW%g4vOc>JvN_I*VRTB-Q`aS@ap4 z^Bg*hQ?`(l^3GzB@BHR$a!P~|RevW*q>QK{oy9}5SZ8t0rQk!*Sv22AlKf+5@$2fL zrJ=KUqeD5>xGLINjK0qLfmNE+sSX3vWL0{#9qu)}kw3AJ>tr8Y*I^2WK;#yyl=HEJt?gz7)$dUo_?SF}l&7hA(v-?W&- zi&xLh0~+C(;FfO3_V9l>vzjE4{$C2%|K(Ak>mIO7EzxIyVu`;kI1@DEzfqJt_!d6J z?opP+i!Cqu_QD!>x$CYAmf=-N5~-!-cbq0Imj@jKON)^e%@xb_I|56@Yj#CDyFjZa z)TO%qwGFE!oVnq?AEPuynCJxmU3~xhJSg`MPYrELkze2I`Wvlf;WH zl}1#+x}4e60ly{HjwDH>7Nc~W``@+OJ_pOmwP#N$meBhGi}x)X{FmC*zJ}lUmfb|+ z#g>LW`(Toe&zG);y7b#dl1MERm}NoL6G>pXbZ}3EV)2d=STeS-JB|9^JK8~A7M~#T zVvBCjF|5m-doQblWmO1CBDJ_N%L9#wLtyFu&L*gX(fOuDeSU%pz&PGIR?WriwV zY&mIt3F|V!sqbU3tQ|y>NG%b}(#5XoSg@?_y8XRkNf23DHf8h0Y=`YFu-xiL;>DJl ziCfTe?R>S%V3|3LB#~OaGfVij$je~4=;D*ESZ+TQ)WvZCyEDIOaSnHQYZFG|#g#@rO>N5T8h>D73YqY?UGoE)z)6MF@7*mfW@nTE& z+2ha>p?As~#@HF%k*b4`TAZ0>p5rzAEzf@ARyR~E86rzP3w90Dx49bL6SkO3;>8xD zMy+ujd#Sc*6IdQQk|a_~7_;2{(BucyMdMrV_KGFuv7j!E@32mMWAEI?P#1T@KB{=J z#j;IPtc%^x#pz&aHHjpVT2h$hTx9l0uuQsmrJrI6e2sN%(z zV)mo3E)Kg)@z)$2x{@SP%W!5{rF&UF;lVFKNDCAGFfLar%zaU7-P0mNxayS zP=kGkVQo-+C)A~c7fB+utY?;QN2~3GG1l%>?bV7US!6ll#F`vQO*4mr<(PXsc(f)7lSv|0~L$M3xOqV0^8ZuV53EfHH> zur9qHnAZSHnVlqw)Dq7uCI^GrcjTjtGZl+(oWQcS2mAfY9JMX@7S3H~5-+x_F>8v~ zW6m!R;J2h^r$`d1rQ8pk`-Q_wn?qe5*woRg`sw9KepuK|XU8#j6@>!F2zF?a?8(>}CNmEbqsxFu$mxbzH(>^Kp+aVZfb z)8BUEiX7mqQbEn38>fGlr2TU@F0F1hcH?&M8F2@?aS;zm!oTXqS{}(^@e|cy9qc@xhe2!`NRxH791eW}z z?8$F_vu3B^Ofd8di5FYG<*h->?&;g`y^NtoI;w+^T3#}X<&Z~h!E*EPsWFPhD_&sf zna7?mK0IQg29^)eBwlQB@9l_|b`P~KL0$aQNfN20R5nieXX{_&g5~J;JZr_0C$fy{ z$L?l79Be)UEOz@yyx1~mQ5ah6%B;sPfsrz8*myo?gvBbX>)aA@q zcD1oMZA=VU`g|wxVvE@lH@ph`Ht&2*sLO;xl0<5m&n)wklCr>3*t3O`V!8fKU`cAg zo|fu8p8N$YP3=g$*fJ;59_x}hXE$~i>Ypb`q?UusGCAF6I#?QxIlNP`_`Mfc>R8$0 zVW=0LuLhPZ3lcB34BE90leD{g`aaa9-b|82YI(veW2&CPosmq3-||x|n-c^U+rjJ( z@TaEvN3g6ON8-g6s}pROry=)_oCk}Y1xX^cXyxGCk9eBg4C*pskY1Q#NfTMDHQD=^ zL(_M?081}35-+yYJy;#}(g(Q($Oqk`=adUhZdvCeONX7ExgP<^_-AkO6^mW6z+z~`o-j^p z=-Uk}8F?gLZ0Y9`hZg^S8+Sup9+Zt&9fZ`PmWy-$zV+KFP?vQcQKhT>ddiR>vg~@y z&Xq}pWANMS!UrT?Y+1E%FpjYiKOSH=ZnxG0)j>!t#>_He)BOWrQER7PN3q=gD5%Ta z4Xk?|xvg|C)WvZ=i5FWegLk9lZp%UV9fsa9l0<5m$}A2+KdOPHlSx!7#S-vIV0pcX z&HV#c&tPY9V^0z(@rCrKe>|THDbz`bxiAfRE<%u)<{&M{N z`ai%DJA}lGE#-sG;22vSoq)SLo;M{)q?V>XaUHwqm5&yeN=b_qOZXRoC1E~0zZ`r# z8=w4o3?uPkOUW${ur7)7JGX_pjIkq0q?Sp{a%I!nGBCzwRvz!DSbS0imi+N-O^&*} z(iQ4*aTJLcTar#*!@Be~7*PT&UK>dgsbwp(wEwz2J6Tz}f+e>v2QcJ@;ocn~w!k=4cPL4&Jc=~iR4 zDqd{q(}lJ4^uAc)j=TNalO$40IJ11nJ&)g#T177^P%Lrjg1VgaX6KjALmlv~=$UaO zUTmqa8G`er>KTKjP?yCCB#G3L$}F9O3=5zx+qbIeSO0a*ACV!jSbb-YP>yWMLrY;K zi5FWc8tUVG$?ceoyVTyyCrPB1I{7&F^V~gSpy~0v+$;mdVwd^bZrtZy@9{IV#tU77 zpc`jYgJhC)<4SJ-jwQ-UioFKixZCwemOpjliu?wi$Hddkpc{9q2}$?6Zd{e(>3C#? zH{WYQH!j4Or23z_aqjO*eTQz`=#C`iKXl{9n~$CYGyP}}l1Ld*MY?en9x~v*Yny*S zH?HXblH?z|ab-G9Y5?81+*PwxD}y$n|4zGDqs~oC0S8$Ic{)^#3xF{a>0qi@gD7g6AVNIw}_D zEP>_VN7li5`|yAhSdLF7@nXy2)2xFPIAz#auuQQfNu(BgX4%;|^fg#U?Qj^RSW-k5 z-xloM(ujCkA+?-kmgstehJd9+aM}dL^6-bC zF6#=}w9$#&hr4bbzDVN5mPGAiIL3UI{KPRf=LSh4wInmk9^akXVA-+$=p4looGq}F zbzxT<2G2e4-Jgk@Nxaz7XGW!-19eO4lq_LuWii`gq3t#EEqE2!&VwY8TB;S`Qamzo z@Mf^2)?2Vzv3TVOEDg+T@n5>h{?;BCW4p(bc(LV-7yFQ_&bDHB2e^MbNg}lj=9Y(6 zxRY3zV}Ccrk|(khKgaGg%Dk(v5iGBQNxazdrY) zcG*qZr&!{11$7DR!5W-bYiHmUZQ>3RFSfK*yM@K=eoM<2=KdQ$l0<3=WR{bw^h3Z> z;*;?y#d7_pz*6ZH`}l!#SuO*>5@kx_#g;YUFR(5~W;q|BE(wk#iPTb|5a+(rx_}R0 zDYrlAkz(1LFR+C7Wlv1LRgA>WVz*-?UTpENe+n%-whj&j%d68QiPX}US^C!PichhP ziVb+LSkgq6=Ch~cVGKTWu^-gsMHdn;wwP7ufk~cLp0gM%?dOstQp+-CS+eKYN~nu# z={xC)BwlQZuXGtLi?`pzU7nuSAxWebe`fK#Y3K=-RXrDL z*7$X29$F}{Y|Uin{WIlJME&^$uFC-_x*E18*~SY%~BFCwuC2{U|sf2tQ`zvY|sgkL~1Fa)*mh(4!<6c zop{Hsr5h?1yLAG~XASlgdqFv)&S1&fL*m7jKJx<6(zeb1D6sVRB}t@~I?S>-Y*`r8 zrC9XNc8VoIWN9>xjj<7VJF~#jL(fYUFSfX?O+d?}8WY-qrFMUkL~7~4ET`=UZvad4 zEsgsrmfPzEbo(6Qr?Kfi!EI|cVJystqaF<<%rfKiPU1xEEO8&Z-u(V?6Y5? zSTaPGG549pcUg6OLt=>^i5FWmj*mx6sR7-wd%Y=;B#~P7GE0}Xo34Om#f?@jiY3ND zP?z=F*tv3|<1XAqu;M!sFSgw7!#+xTwfpu{7-LWLja3IBwVY)Z4X(m8uF*2UO;hZ|V#&L>HvmIurd6P2?J>QZk;#0kaX z<0P;&8q1!;XlwSxxqoRti5FY6sSXfB#~N@nWg0@-{WApQTzKP#bWO) zusqji^X1KnTg{*@5fezf*z(~X+pFMvRPrOR#BCr+q!x8`JnuJupo6~~H>YsxL&cIT zvZNU>OL)KTpTP1#*IgAawnPnPzpY-=btmp2;Zlhtky@%Ti&xt6D5%T0SeLho<=#d? zU8WhabEV(S0NibRWrfYEc(Fx)E898zMcNM>V?l#R5~-y%v%J^pfqO?)J|Fc>u>@`s zSUUQ%xgV-M(i`g1<12|5Tb}u_=P(KCl>L8?+!OY_DGj250rEC3L8a01i z^Lw}mET0>)d*tjeW84qpYzq=Ew(K?bz;(>w)Y;EqSztnvNG;Qtzx zk}a~lyuilTl08-hP?xZ55-+w~N@mx|My`!r!SY${yXqjMmNm@sxccb;Xy8=LeEW(pj9X#yX25T2_4soyBEeNR~f!7IVLR#)Inr zEYuM?i}4vG-S0Y!xvuQUx_n)32A#!ec_h{U)LAq*doK_=i|>lnR{itkoyDJ=*}eDU zmoTCXbx9&+L>1{QCQW9Y#WQtQRDsT7qjDt4KXw*(RlB$mI*VBo>ZoW%JB#W@tRI-< zn-C8JbDL*fRf6Ai7Au-8#7A1Ai+2r%&f7*8;uPsDhK^!a8!N|a$HQ1(dxXS`SI^)P>`X9VzGh$e zzl6n+B(ndD+D`U=@g82{DOip*uijI!M7RnpZY$XtdP$Gjj$oP7nZ%1N-5hpdjjdXa zc?*`!14t67C6HMnHOGX5WlOtrqZNy(BDLIMmX;bV@Eyk8U7f8Ji}MzNWql<3bn?ND&-e}2f&L_3Y_Vv-+AYPlHZz90 zOdC#;NG%hjtpt|E z2_%Ws@{?KWChXURF&443tFvNxxK&UW)4}Y0|JQf?u(KF@l*EfI?>}G1`QpE#3|<8` zc|(#&EfqBIt||8In2%uDz50!pVhMH^SfabK&SJZhG5Fq}VHk-QTi(tZjh3ieej!ko z>ggnj)Y6PuZXNGA3oQ4yA2_C1ytWA}Vb|Cg)7z@`1jg8|{Uly&F>ttn7OS?Ac-K_* z7)c_v^ktUV^q8?=>Gh#jm}1EjS$Z8~?=O$3I#(Yo+I4EH;>8yKR5r$@gkQoPMKs5f zBvQ*{X7PNVkKJp%BUX`$C4Rf0E-@|GTQ~_{@)N*v`W1;6Tk4+agmeGp;DJ54fdKvwLL3a!`xRt zN|Hz|naomd%MtvowfGi6Z4^tWm%y_5D7zjj-OqVHSblUN@nXw__~AHTO3(AeT?Fe5 zAW5W_(#7z+-(&x=RZy4N#cld17VljGOWJJq6#M@1r&Gb=T!zGpEgDHraEujn-lq+9 z@vTgfNG*++W%$LC*djnaU#VDb?-A5xhZD0@{27ba$&0scEPVz|#(S}92Mt^mOMth)^2w6DAyKL0f(WQfW-N&pTVnIrFH>|sZ-H;F)GJp{ zbr4dE3$yrzyqF7h@vwQmSFvo}E3oXV&*sY*oizM5?ut8!7hA^d--&e@x65-I)FuBC zNg}oQGRw!iQDeX|xySlbiX}s2=^w@RTq*RMdmHLfdkBdaTXHkmThZ0tw#K(Td`(Fb zspTTG+_=9U-zsoaFCL*-V)hB@a>A8ekLjKm`WP%x?Mb}YGW0#WKy|(Nv=-FGbOuQx zwLD`MjhDq%KwZ|=pC7GQ!uJa-qf^=QvCEAtW`M=Yg2an0Z)5i07#q?0jSW}|>_`%+ z|M#uFQ3b|z&%&m29tQPrO{ktydHCT-y3&At`JF* zNG%15Cv3MQ)Z3oM0buu=KI*%Vz89G>7Mv+XC4%VHM-*C9ytd(^Z zI#|udlPrJgU^$#Fj@xIRs5G%5bg-a|f%WRx#{gRcK^A9XePC4v>U@)xmmHfxX{m;?i{mtavZ|NKWZ0 z(U+|fEw3IlgAUf>QzXs5b+9^D7;phrPy4mC8YwHzz5iC6A|0&KZP}I6icNbf!B`)? znZ%1%&(<~cahW-H;WYm4LMtzlMEZXj#r`j|LhfgP9Sh0Bc3M`x3u_swxtJssQ zG$#@-w#?l;11$@Sb-*1(wjLr$q!tHeF}^tb9$21)jhdubvPG64=kHANfN2$BeN77KZ4(sd}uuWkYaK66IcemWABwGTAx=3%ZfH6 zUTi5C%ib;hd2R#lxVyhINg}mq6vsOa=e2%qp)OyZT{)vzQbd-lL+nO$TlHD^ZJbLy zi5FWY>#*=@Rz6cK!TthEfjc|D{M^v*5m@?! zlX$UZ&knX29=&*AKGbE*HC%3wX9{9VR70qP?yXf!8&z*-I>P+2!e zDW}DL3#Y{M7Zaf_FLX!}sl|&~#<}D-0n0m&)M|?5`U!z$x+Qzcpm+6oOBiEjr%Aln zvUlqVtczy1uyC+6c}8vBvMO7Exd-=v8W#I?J=&Z?_9+adRky9vyELRCv~4x2zA*rn#7AO zXKS!~u?}%2o57;)M3P7?&6%a{x({!lE>FiDTBBIJg9H|}8mx)FF+!s#&xSjE0(Qi1(xN5*tO-P%0C)`<(3hN7h4vFdEpof zHND*yEYFQe5~<}nv+R2{5x>DoC}Z+ov1EuW3r*RYJiW~tL$G95CGld5`p5@p`OzdA z_m0vxAW5W_SIkm!#nnepmxwRMnTjRmoS-gMT-mv@T7*MsupBDhS`{z0Jki*OmQM}8 zUxT`=sX>xREm_PG{Ju#7SQaEz)2{pLnm;^LU^!KWjj_k;s)m3iD~-g9EtRiYp=Frm zzzblhl|zz9EoDmJdEeW*y*4yG#*R&{tXMMtv;1GvW1+>2xh*WLtQIe{uv*@;q_%oF zH3Q9|S9G!Q)%IMK^3aVtV?i=Wx^bth(~uS87(4>Hao#gXmOpjl^wmq?L0zxdeHC=$ z^cIkGzw5?b=*_xuCYfC$p&M6hIZ5?Db>pI6R1bh|+%h|oQr?Z5SFRW`&zd>u!HCLt zB8ij{Riqm?uOaKkeH}Hf6?EfXx{)OR*p1tB++YTDX z49s31KB*G?rW@y>yACsz^)2QG-8i?;B-7t^iQTxk zA$>YRH||6(N%&XYxG6b@a5{|*|FR!eiS-(vRsTiVD$%_rTO~H%o?Z&Naqc=K&A)Zy zj(4h-1gqzhXXR@vD~`{3!RlG@5PL6Us@HUU7kl;>5-(moZ{I(S|Cho`o;vV<`KeJs zbr7=ui<%+(zkE8S*%~a-@vT}Z7W*)PCAc^H4LrZ_l(#T%yM89|VoQsAW@woo>`)FY zyV6M#sigS*m!krS8eIc&D+-aN=#S zWKASVq!x2#@z|4?36_~N%9$#bdlv+CX%fWh5@flb5ar#=Sc(KKzdlHVZ9#u|SgQZ(d zl0<6R%q$kZfqTKSQP*{;V)3{ruviXa-34oZjb2cf9=%Dt*fQ?C500_CFWP;;@^mFh zBDEZ5mSekhHiE_Yqo%WB$rf3>i?gQ;m)>7?28(|E;;MMD#d70Qtjl4OXZm1C98Z!+ zEtie1155U3l0<5G z&MZl%Ct@dl$(0HIiY4N*z>>O|tz#YaT|A*KpYuq(*mB?CJk}-E${6>=AJ=B0>L8?+ zG-mlxClYsgYG~T$f@1NF5LhbhVwU2!-Eky>=L z@lL})TMvJCVR_dMk&4Cniomiyjdic9+K1szFA4ogyx5|Bw-#EAv?IEJWz7hZL~1c$ zmI?_wI>Ow)xBN}4Vo4EMI$E*r!aS4nZ=f#u8%Vs^GVBxE5pm0wHMrBuO&5|xYU#u* zx3+}h?<^)e2Yps753dU9vT6#u9$Qx`Ef*|PdI?3oBvNegI`Y>ysir@>*})kvhw|{WUwqdL*m7j*Pk2Xe2HG%64x=?3nYou z;>axdD_@ob%l^c-wG>O9$TGp1jj@;>uhYRYCXK|4E%tpPur673*QUbUFP=@3NG*ilo?_=)IpMWs-YpU@wzw{Ck7I1y*oWF++5Lbdky;)xi-V;B zc6<-pM2%D|ezydcpQqTHk>Nu$8^ah|w1dQpE!{M?VqI!aT9^-(PFG11spS*1+DH@J%6C3Y1{LOU@?DAl1ME> znI*Vn9-d!L=VkdT7Vk)bV9{k@A1-mw^7kL3>{@nTEL=E3+&Es-FxuWSA{|5<+1jhi%79XmqX@7NZHZd_sUE~-qDZk*dF){Se^&$1VE z<38$l{lw`~U9 zxWxJ-rMw$g8&7h{@|M9PRN(v913@f!|HgP}DxLpQEk8s?kui8`t=j9rn;H_muX5ff=0CO_ks`-MBUy4wz|O#FTZ=jVqH%GW~5guE+s8 zm$kNnZrp$@lJ?KtxLdchup8H6-_WtpjZ@1f3ID1a7w>%&CTO+$s)M&yukfD=8~ZhG_L%@PXZFXqC-<4rc5;&cNmqojQ9X&wRH7%}}w#JQSP>D%WOb=n@$gxGUPJ zog`jt$E83$bA6kN?i4932we)9}l2sSrZxc=__G-Lhv41SG3}f$NyT6}`yFKjh zK;p%gpCj4RpXH6e;`d}B^GFh@#e!MB7C7UtZLc~zezsys7FjH-u_swK4((_Gb?JYF z#EUI~YuN8Dw5t$+@0BOsAxWf`70j|AulG@yFK&tMs};+=CxW_oOk;C@gM)1VSQ^yX zrivF^_V&Ao^Cf8LqGB+{u8bu~q?YZ>QZTgxzW1lqKgvz91U?m5s$FE?ll9E+u>>rW zSCe?LC1?|Sd2ns#%edP^ffq?4wFEFrpP2hKpe|j!?(J7B9?t}pw)5E2pGmX&<8Md| zI7s5fmJ;C}SQpnd8*mrFn*Jn-)N+Ga3aW)_f~BxtM4)2H7FkYrW@GHC?=GDCUrk87 z*z(QY7cJ`>AD9PYYEce5%;@ciR5ho%POWbooU9z6C&)P0^9D?&@{aq3- zwpf|4U)z51G98~}U5_S7q?RAd(%r$hA=G8c_(KmBOGJ#oa(6N7`aG*2GX>_$h?yi_ zY$-L!8tY{|ZSWwKQfH&ublfL0w|YPx_`
    qocZRd4KfX3`6QM4b z4N1J%Qt*PkSN?55yYf(%;T=g5sih~g)ERjmyFTuZCTP_Eb!VO;vUu)f@BDU7I6oaM z7H>$r*fQi6+fC?=MpP}ZjQT{9NG;=;WlN3ceW5M`d>)opEDvJ^b+Ig9osuyj9qNJQ zyg!K-TNYJhzqwMi`?ynJaR?$wq?Sd@(#OksGFURpu4<@Qg5w01_pR9V*rC_H1HkgM zD~T6doV%9BYsfG&K(XYBEKyc$zL+jOa~A4y*NMc7 zEvCn};F`R)?|ba}ct?;VQp;s#nb~+h{@zmU=H}*#CH}RbE}t}+rSE=i{Jo{C!$`c? zQsQ+FwCKJH!Z%l1ts+UJmKbKyn6a=l)J5-&=S;DKM9ShN7R;pVMu+%R@l1MG}m}T)g^ZQ__f7W=PVo4KOb~R+*`urGQ4!@6i`jEtn zEk^b1(UK5a&jBoBKaeC+OJ`)i5FWI zwevyChk8e=gXKoCHL8Pt5@476Mz@nTCr zN<%zXt{t-C9@J&zSCT|(ablLIGoM?4#pCC-4~oU^gTP`P&7Q;b)fkB9mlG|otK!9$ zh|28E71L4s@jJf})+CA4;=?Rf$yxX&@yJq3vlL5$$dcZG-B-TaQ1d3#MYB4I7h6tN z=#67+M1yUHP?wDjNfN2$JhN!jsk{f89-12U^&9-U=D(8o+iu*r+SRb}Wt&yk0lIO^ zbV(*jH_m1j>v?6BxH|~Caec~@EPv|81>9(a2lXIyRC(yet*B1Y{jMA5yqF!?y${=S zpc`jakEHsax^a(cKZ%8IT&@vG`48Q=XK$-ihY@wvm?Tn0RFQ6+`e@dTYf;&MFLdK7 zb|gvuu^YE?t4SZ|#(BIT$%=O4J_Xog56#qI$q^WsYFiudfjMuH#q7yb|NCTf(TsU> z|G$qm%j7v@rVhQgXhJv6#*<|F+iqNu19a7$*ciHTXZMk`f9}Rvbko6Zoc{3-2cR2Q z>j+8sSKYWOH(7V%>$7h7RB+P?l2f)ybbrWJi446~JD?laHkhROw{F~^4cFGenV_S4 z>i-q?_J1o*k)0|vKC?4H$pSk(L-*K2;>D}ypo{FvY1fKv`1@quhe#6H|3xjD{a?CX z*%_!!U7Wh49H6Gk(G?y7^3T0S#N z`R)U7*Uj)A?OQ9Bt;qt*!x!wz$uoGX4y-|^mXUa|Wn&|DDOq~dx*u>>+kcWIky`p^tWmDXH-E%yN7h8_@NXIjDi_FzEz>;H4l1MEz znI+w68@?moBec7zVu|@EsLQE3%u-v;x;WJ3!#5Hyw)DPJ0>@aB0yDfbf7_sk>L8?+ zcFfXeuSbRPUEN=DK)%mHsa6_=%vm)_g%l+ataf~hKq>t|~wx36mNG&$Z za&*c3vS3;Fal%H$VxJj4&6^Ad5%>hfBy(Ed$vYT0M zMOMy-y10G|@>eYPz6k14>jAsdn5*8z8pc@cPZBS-v?;qC>vGC76pZK)2!bB1;wuDS%=lwDX%MXC%;Z%}DYPrWO zYo@ivohrvi_q(T9JiZDng%4QA*EQ(A57cF19ElfORvWNyef00D;nSaK2_%Ws@_|__ z#%{xRvCj_4c%@jfMV4hr?3DfS%WK@LVANO=FSfL(JGnt73^u6VxU39IH#OzQatRE)QOkc(J9&DJ?u#`f3Ii2TMWeLe)V?EtN~- zHB9He199&6UDHOr;jcUMh;IVR$Omjq-gL$kyVv)2lX$V^-RLt|mkDJX90yDBYb1%( z(vn%CTkO+?y0|_JE~{94zY8p-Q`mQYd zEMePAECq{`bNT*?e93AZuTU-GWnV<&!ziK8lBY%w0x2`$OdZ3e&?i?VN9Sd8f|^hi=@R1d>wT zjWhJoL1sbUosB+ns`L-1Kyk0S2aFnO3R+o)Q{cpq)s`D?_!%pk$ADi zc;*qb1boQHF6nX2r>cXHTD~(&`2aiIt*(=8O*6$3V5d`bz6^N5?q+jJXncV&cEpav zi!G%qv!|tz4JW>X`C_!0B#~Nl_3=)l_Vneop)QWcmd#KsTh|FJtDdpBZ}_3zV6ZfN zPvXUvWi4H?E*lL!@VnULO3PITA+^+JmNm^nwt=M}_55jiZQwVsdv(ytxY;`hqC zQ%StoGT(~rR@dxYL)?3PZ5Bx)wTxkwoRXt^KwZKP4E9zm;Tr^&cU71r^}?0OP?yLT zBwlRE@5z4AJ#bkUyfgn+Ziwn2q?UQiQoZ(bd_%%fquwdS;$ttcteV2kFQ+Tt*#nl5 zrAWNkGH&&3EUvR|n=3H)FI6B(q!wppDLdx!3#d!=jGhsS#oj?+F$!a!v|7g$;M3A> z1teZ<*|BXbTCBzm*#eeN#e1m^LTWj{EaSsNOu^FinQJn>;;tYq~`Yr)d+CrKi;gfYwO@3-(9tRCNbzEdps90he5oW$NLxZOOY6Ih!3 zB=KU)unINsdTgyxzX@QOQmmKiAf%SZ%#vYq6u*t@+V#wL#S-Wwu=IV(ECGgj_F(C_ zg2an0VLjOH@b#X291NBl$4L^YC52f$_Eo|i-90Xq&@%jWXYS!Fuyo8|^Tl&pt!ZF6 znL*;kmVKApalX7#&xiqwdlPNdK}apd%i?+eYe7f+{g}vSODijuY>{Q~LiWUD*36jg zU|ASX;>DI`+j^m;%C-u)tMi^@l0<5$#VqUn>#l^kfA3LUL&XxeQBaqMiR}DxxXOgK zVA=PT#EUImuKA+H%Am?-sLRcY2C9RQTG}(qzQ|{Xz*4%|WfR2`u}NUDyu~cq6B=Ix zi&HxiFSab6HUcdj8}8f-mQmeE5~*c4vy^cw8w?h&i|q$07GD>E#e6oqPG0MCzARW4 zW|Mfa<;ZCErk7Q~dc5YpS=B^!5K_wwW=XZ`gl~F%o;PTsVsYLquz093%eUeu+`%$% zHHjBnEY|J8x?FPVWDa%7cP2@smi5e{SxWCU)FrCJr+JDcMP!Ls&N^774W}Ii%h<&v zUTpc4*99#DIbH*>oU)Get~vV^Ub#cYd(wh!YxvE$oq9!Vm#1T)LvgVFn;EDJ}!R#q^LGLlmpe`G(k|a{geP*$n7MKc_S9U-RX6FS)=`6+382> z#TMUK_TRU|ua3uma;c6ZAySJ4v*<4L$2V2pJ-zvXV%fb;VA(r?9rrJ{sl^G_hEh^5 zwhSw>!ZlXs;`Q&LFD^Gph}7c9EG_F*;Co7%4_Ww9v4pJ`Sgvhl$I2w5yN1vgOIK1a zw)8ubjFy{M0`Yw8tS1SPS|&2f^aI(|F!WG$t@}mif6n=DzPEhjO_kv<+TcVeVXoIU zn8xkyMUo`bxUP@bH16S)`XgW(7h_FQ{%snkeYG8aDD&q-R>Cw+-c=<<)hfE&4OUdq7!stcs1K%bp{8sa zmsD)!1=F})<4MS8PUHFv`r87gasL~vp z5jc%|a-nBen8pRZAW5G(jr-sO-X9$HB~0Twz9X^!K8;&*zA;YY%FkESgK1n*y?Sc# zCr#s~pJDIj-ER02-)7Wamt;!!67$(!;=7#YKfpBZk{*fqylI?`&0teFs&%d&)I{0i zJQeLdpGB}M^2V1#U%^^m+0Is7FW!3|_;wF&%z4JPXW&TC-JOI;zn4YqdpSB(&j&0T zLmISGETxfxBf(N1cD8Y5m*)(yTv|rz#TJV-?0$b&j~I8bOgv0Nq?QfL(xSDM3s@Fq z6j>^k%nbrd_t)&+pL<(2xPYZaKB*U5Cb#g$t8$HI_BaBT!b>DXYDr}lGym>*|FS`G zzP)0J-6*hxWwYlnEQhN6z%sd%)Qc_a{rcb<8yl#-04x)3k`Sro6tgr4vc>=Q*w8u4 zO|dMA5?F$kuBSF()TcATnM z%0-rwnKSS!&EAoQXSAk0NWIuHQ)dE(94-pLw{NZPPeP=YSIqJxWakUmUhdB65};TL zqXm7L_>`TUFU}g33w-w;(rNa#8ECx%)E>bVH^lWU6mNu%8B(PN6A|X;sS7vd`Pn!gN854LqRXnmvvBhESH)uI%?c@a($0`ycwajCdTWQYoz|wa2))K{%6)Uh5n6Lq- zN!-AJVEH4B)Qc@GulwTm5_hE<-zIeb2nmr|)-X%@t;5^EGN;Dwj$(KTa=|*FgJRhqFR&Q6u{Eaq zxb6Y4*w~YLvBfob4O$FVyp9D+Ulj?FT67I?jYSpc4FSu+4J~^smdI@ai>W(1hj|zi z^bRatdJj_9i!JXQ*s=2ZIZq$3jGRS6q?QiMQm4kN5-bjFI@u|fzeJX^W$Y^UYWM0m zu=uG+z1UK`hi!Dz&y5HH%N8FJBDH+WEQ2)?Gr(fm(s-m|xtSp7%hB=dxWD~&03Kf! zcOdm*OKJmldu6Nir>0=>?nXkS7B^;DQ5|^)EIs0rd=*R1c7Y|IzunQxQo|qm(lMFT zi!GI#eQ^%8K&L7y9ux<_1A3|zb$t*P$xp?PFme<(jie<%4fyH4t`(61%(7RsHmx?K* zUToP_m4tnX-2Njzf7x+136WY7nB~op+9S}H&^{U)6-$-Ka>binkxwXEj&~AIvm^Cl z%TarF4x^v73I7#l)Nm3awH#xXnz?O5z+xYLGEuRV>=N|l&<=KQB0gh(yb%wk>r))g#af94b_mgwCA%j`YO5^r%Ful1!A<*Mt&me9em z*cZoXX^CLbxxmIlr6xSun->;PEYBYIy~EN?}Y`;XcASVgfmK1-UsiPVcN7Ax5S z_nOwdbzq6xMna?(6K0vF*(nt)cLEkYQY>Xjg1)#nWryq_+kAJ}UQR}ndayv7Ag6Sbk1pzbmy<>&^$u?L<;9wseYOcNEpV z)b9XT;*OCJsl|m^bb9@02m_okQ`{SN{pa_~{r8rSOcuYXZ;nGet<&XkFj>rz2R&}m@M`RB5@y^EE)}FpKMq7H7l4bdaNK( z|7o(Q>i^Io8V;8c2j5qOcrMjAc3DWS`8HF8{F| z_7a=yNv3QsG0=nUC5EMMH-^b#FJ}_-d6UJ5Gt8U8ksvnYwZ5{)*}X@w_x!q;{c5r= zK8I8EKEINB@!r#7HXBiN3R#8EKjyt5A+qnKjvxD8qP^#K0n3PI2Rka3uoQvCF_}%U zYE}CJVY`iLm9DNATQn9jOKm@Y1L*M_2NEK+1T)KdyM1P0nN`1}uVQ&BvRJK}g-dPp zgP*W3c?P6jY2s_hWi`X%1B^rF#W^ng1($K4yS{ zn=AC?!bVarwpi$~pDQ+{@%S7@i((QYwPY~MfT4fj|12H}bMRIynW-X69@}2p9IeGG zSv4C;z1XrUncc(q$h^tV(3jMB5+b#nW0u)2wi98E-F~-frecZRC$QWKXJ@prHg*%B zFRgV$)%9Xa!AVD4W7>Yn^T6U_NJ6BRhs=^6KN5eF!W0g=vDmbiBbVv!Wm3e-%7% zBlTj7@1{mL!7^H#jnBot8%IK`P!TQZKex zWtrm|yVNcr0M^)y$s|N-@nsgrR(G<&l9X`mvSN8IvaE||_sEw`c` zz8socI}9vMjTfswgw(Q+SwcJSz`NEmR84fu{`qAde@I|ivx#l{<^Fnjg|VkQsTW&f znz9q2%$^hQxg{+h5+bz}GRv`1xrMO3I2aomE0$%41(ru4Y#KKwJTndYa#cm@#TMPE z?5$`6-tH)Xz8nl7AyUg7W|^d4=na-y6DM=UQX#VB@1KcFJxgPk6IiUgNWIw7r$Ji` zDRNA}^D*BIBt&X?%`B@|cdQSVy?=iDonkqYDd>yKsrx6K}TJ)RaIZRpbsKsEpcYDwX#j^j1z!GiAmin|fEqn$`x02M0EwSDs z(K7$9yS8Al`yUCBTFjVbdi_@^U@;xqZIWV%JSwm>S;+Ru{eQW?0Q%C^kJO7Tg?T-3 zKX&)xIX|$Z29pq}#g19VLA4PY*J#~;+=s>FI`>6-!Q*!164e9bfhwHo|AGTbGi0v89eNdqZ{h`dj$6 z?ckdvL~4m(mda~>_+QZt@7+&QEIW=1EQRsxcV&L}r7vNPIn;Zgt`}P>TKC1i)ciEH z2KrL36$z1Al9*-2xV`xQms9Gu%2F&VP6#Ycy6_d-=HSe{J(U z?vtMzk`SpSk6BDl*|dS}#YC^oImJ>Xvb45i*T+mok2wJr7bj9Lww&4TfPL8+`?eYM zW%3ddBDMU+EPn6*d;lzYdEM_SmXd5iUxF>!BwoA0SNFm4;yI}oTgKEGg_c*>=ca(= zH+^&UhmcyHGfRfK{xGntjnRFsSkiL@7Rv{08rL~70`FG0xJ2s3mLtDAp=J83`H#R7 zuJeWZLr5)O7~yfhsh7nqm^6IdA^VMDx%=Mo;iTcL9h(o#YU_d59Q!)ussm)7V~G*_ Y9ES>>m%=r2RUnD^yldpiZogRmAC)C`v;Y7A diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/events.out.tfevents.1788165989.f782bac2bd13.13.0.v2 b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/events.out.tfevents.1788165989.f782bac2bd13.13.0.v2 new file mode 100644 index 0000000000000000000000000000000000000000..8c5ca393247d8cab4146d490c4f3370bc9238c6b GIT binary patch literal 520738 zcmbTn>%Kno``lxWc^tDra{u?=<9lXrl;z~i9*;fv#Jp|8 z>Ao(Wa}0(paGNu~U$1dmDw^}%=J@&e&h+*PFqq-v>t^6T-_zT`?XQ3La|-bFobTqF zq^+2t|0k{Lc-AoG4;oua{z}v9q|#AE(c8z*Z-`2L`KEH3iizgW%&Z%zD61&>&2aH{ z@%6Lsm^`oNpW5B4R=rg|um7i3M@36T+1=O0)oG58@9cjE95}2$7y{&y?msj76QFun z!(O>O2R9RT|Cz>o@7)GGm6nugi=@|ISHhyuKggv-g;20 z@1L-BN+utbyIlG^OkYJ$r7;yoZ`D&xW~+i<-_u|`#ssEoXY1=trigNU`zatD(^i`UW2rkO3-Ecr% zcr|NQk3Uhfr?$6IzE}Krl%|S?N?pIdcA$R`pLM+_4uNv{$>TC$80NnFyZrx#q0jt( z4Q1DV4Z^?L|L+`oxeCnLyxgT?*q_aKSK^VVT=enp98(oz74^Rc@@#)^m-%k4PJaH= zXM6hj{VUIDzUOT5#>KsBerxcbP+PYJF3R#h{toS>VyM!B4*l1^dqZ5knJaJ`Z%$mc zc*vi)MW5#!RkoCG`0rh=kBX5>YdX%`$KBm?jyoj&RSUn~ogojs`us_Jcx`Hla%%m* z6U8$@5Kbl&+9_4``xDZ1YuQU>pT>WObXV!7qBhsp&29GFzkYtF89x4V=Ks41GMm$E ze0;pYpQLNIC;m{LF?5T(&0lkhtL&_}Vaa2h74xg-yo7N*V)WnD{CRQh-X2KFyJlDm zzekG+e@BRak4$XSt_@Sx!ibJNRN9azGUa6%pH^feW5$k!Z(u_7NR-O?j$C_3bnKaX z@`Ul9A9C!(@DIxNDN=?QVl@;-$CM*>$pEQt)4nZ%Bq+#2tQTBI$kAqxZU7`zODg9> ze&mHBWYm>qcLB0nU&;_ewjm_wRQDeM@jf#~hd?6g$U^kyq$1?_>$d{{Qa4>H=R6)t>0c7_Bshkh7e;9@t zV~#I$0f_B0DMJkDfsnT1t5tgwbKbCSV|B9-$YJ_q+=MuBGL zRDjeQDP@QuE(kd?#7+qyZWFG}A&^oI#B~U>5#?Wd_5#TJX;L{K@~z1RglIf*G62Yp znNo%rvKS$&D(^f2i19-GB?OXHU$!HW%lF_$+5#AdxDvkZtZ;a3kv7TOSXQrCp?QKIH0^6_`Z4Ui=lnr$bLY7S(>Tn6Gg-hMMc?qx{jd`M%>aD-Hb2b4ocLb9X`Aw=#3 zLQ=Z*HU&sT!Sme&QqVxQBT+4wjhJCwdmSOKrE)%`t+ypYl%^J50Z8NbQic!`bO9mq z>hA{uq}u3UEPL2O`kc0tJIUjOzJCZnG%TbHAw(_{A@8>- zJ_AVI@LBN$65CJ~vS}N$5g!6WPXQz)QYz;|);q@GkFoiz#sGi>M@tz(NKhU^c5Ur5 z5g=LFeNGWbXd_ujJ%zQn5%B|Ws{rJFzEsYKY_elcb4t4?wE@VoLMcNCkt;$-r@p-x z0A!3xO$vdOb0GN@D-mL+Kl~s-rZ347I;}rlI8zE%~A#y(ua$-!eCUiu( z<*w@la#~Fm;+xKF#5wC(-T>LqQ7Y#{#%7(xALIBM%Z>o)*+t3_Ld@z6gVXrI7Y(}u zB&naqm zJMsg8gtw4|q|RX=Q5|PZ1jvgxshkfvGMhQ5W-S^uA0U$xr3@h?O&1{}9=`1bkWXEj zeI<}e4n+5vBX(rKz-T>y%`b!x?NKk)-eBPV)4j|X! z7po9RbZc42yOWb}BaV)l_zEDePfO)|$PHblBd&*^E(ge@R4GFUNi#=?$F20w0C{LK zqzQopwULE*wHkwvs@|3U0BL6~mGdE$(@iiVWL$tdK=Q4m3?W3$3L#cYW^4dR@E+CH z1X9d_RDTJvyr zJ6XtnC3D<}`ol+<0>sf+D(6Fn6**u=yW{H*0p#&ODMJX6b3sUhCXJ5+UU)}x!kk<`@EeRy1qbx*O zkr@`#9~dtJNM&=WoDbR6r#}PHio*lq>2^|v5F)o2A=~FpbOOl9bq-?)Bv?ZhGGLG~ zZp6o2&su<_L`vm+NXJy>n&wVOwmEdfJzB~TLNtO9a#$h43m{{XwI&cqDF-sOJp;*) z@Gt;~W13XXhfE&O4u6ce{EzkkG02oMgb=e;23ggCE8$eEvQ@f789<*DfelBg*QDU8&^jYu2q zhiAX{zf0wONXwFjn31<{RSrO&$S3{12|`HFT7*2*7YV zSv@%wI?`{FRL+OgRky*6@@Y8;x$Gik2q7`y2FNHquY^Sc&q#N@VV zuK`k+DwXpgGt>rR#>opl8=)g}GNcS4ByA%?ip=%50>n17VE};?XvucuNp%<82$NfN zi~uralT^-!9Gb)I?DErAM*!q~l$0Ta6mCYygXY)N05Y*6KafDuv}GZY!@DA6P*ll# zfT(pcSh7h8$10jWPyWtJGi`U)P5lCnkS%`(a25v-(>)vC~ zk;B|i9&wox^fP_AhGK3JbXoQ$wnAr(BBLCHFGl7(IAYr#VBE;dRegQyY ze@o?jh@od|%t-k(7}m%K4BP0ZN$huJ7GX05P?eGK7$rg9tGR_ca2D&&(Z%2qa2R7BXmid)$a|UEbmK z$Qd80oDcCfXJ)?6<6I{K#5h385JJ-85b`r^;bnkyy6t?7Kx#OU6PuZ-Gbn7*7U)P& zo>b0xn@z33Xs)`T}~57VK>>1mRg$`C>{P9bE6OYk~?3^6}={8a- z=R=HcGv{j`lqNa=L~gv4A%vKnLCCpP9r3=F*~k@{1QKo_3u%$3hQFps^q~U)(X*4v z`H**2%(3X`fMJOMQJW%V2qDfX2#MJnb_qJ7eS1_kfmCuJZ$cPIFJBX9fYi&A%K4C0 z*O}9ty4r0s=`fl!fRoWjb=G;tMVoiZPbT`H*L! z%+NZP({(RE9t@H)gpjl=2s!vJ`7l5hExPcEK%#rfLWZwrE>n}s4%ui9e8{$) z%EFNW{FA)f+kCu*BZ zL4uWBkMCz<$WLeWQG@<0Tp(-gvdpEB|Fo2=&ekTFrnWsb3s$C>jg->3p(N))f=GH+Hi2;N-j^DW67Mp9D`DTjeKTq8+0=>yniBUoGk~E72SXBVzgfBcU)aXkM2h zh4VuG4gC3TzI3e*OZ=uc3H^Is_}`XwU3oeo80N~pfnWbGDyWz2Txq|OIWbyMvUUZG z@W-d5a{gSoaWr!!)yq>~6MmC}sZxgUH<8Q4-(*KAY zc>4PvKLL`gE0yyh_Kldy(`PyxS|PxRB<3-&6uH8 zx8dMR(2;FsQaK+|=QnefsF52R3y`H&Qic!`^9UiA=D+R-9T{Vh+?qi0jb%IX(x4u8 zMDdIxuI!pHS}Ny5;@@<`jEgS)qXD8gLCO$9(w-pXs3U=-n#e+|hb!Vn znB-o>cU!$DN#%TqhD8s|u+R>~N3j2TnDO@}2qA^f5E2!!Jr+8mp)y*BKw?Z~A-9y6 zC79+qs*Pc2wQMGp^C7M}oe>i5XFt*fHXUQrXPW1^^@($x#7&k zvBRda9su#ZDwXpg!S|V4i{rx9ECoo}Eh$3?aV|s1^5xIx0OU&BKyv~~>@N#Ru$RZ7 zWnW99VQB-$!K0r3+2N&Y5G_y=Rl@rDrE>Eaz7AqL#HL)O$!?CHiJN72g^dD*D^=;fY-ZQ z1LVwDshkf9T{0YZ_5llj<8fr~L@7fE(Wpg;=L4(9&=C)%?z0Fa)JztlasD@Uq?_4~ zH2@i=CYAFc9_yIX_%ZD>ZJ;But)vVwL=L}Z)2Xh{8i33hTW21Dlye{}Z+=BcQO##R zfb?>Z%K4B)1?DZ9^hy3*0P@30$`C>l_mLIDsaG^00PP3e4#40DRZth+`Ls6 zfaqFE<$Orv+05`fKG}8?K(fb38A6Cz9fZu@ovQ*Jaeu#M8G)Q0B0IFS5C6iBbX#Ip z4UpyEq;ftaNUb~WBU8VB)don-FDXL^$x%eefZ|bj7o}*IV+esn4V8u5^$wSUo$Kdy4kwTr4kRdtfrQvD!iBZf z&!lob#MYX5BVYc9(=O=9mDf^+5K^gxkk@n06v5EiGq`*+ffQQEc4Wy{Is7$O>{7!w zDNk8R<$OpTPv&}Lh=lbun zPnlHChpZmPJXp`K@9+~K?w_O#A;ho|LYgK=+Ur~d^z(zolIP5@CK zDV6gf^YoYtE91xgt^*`yoRlGiq%}s!{f28l0K~rP(P;w7vy$z|@k_4|qBd(29@*n+ zrE)$by2)$InA*MY6hIEt&HsB7gb=VCq0*SVkh1?v(oNJZE7fk|4 z^((2I4~e_1f{<$ws(A0W(;Jd9lwNi!ZR$rE5&j0K{dk?>hpi;y~VbF=xNM&v(}aNZ=x= zoDUhL#JmjO>a;$dqKsG}We6cT?Xe>>6Xk{hWWtqEp9my>tZYYqnJ^tO+<0OSK!)Cu z%J~qTPRy%32F||=VQ6LLN*O|kb4P@1$cn|Y->}V1zY|ERjVxqEYB3J2^1F+aeS6oZ~R|Rb%V#rLi~q5$Bw+tU5Tq! zTuY>KJ|tl>b7ZgTH4oq0`&1!i2q9`(2&w-fG65iS&jvIkkWvn$!vbb#^*-Ho2MjH% zG^v~qdD4NouVq*NJDSRPtvB-+VBO3xSI&@z1XZWeT8 z#T2QW4>7pS3@sIf*Z5S|)J@6|Leg{*vczR)4s_(@!t7225;;K@axMNbLIRJhYz~lu z6;e4LGGi_CK1;u0v3AgrcrgFatxe!A0YZMQic#x*bO1a>rcW*Z!L;ldJsr~y=+IGm_I>?g2{;}=!jN@ zRL+MuuqTb58m<}wkR4T0h7dBhJ3^AvzIedUiao4jOdx3!Wg+hyy+eqOM^z9&#-5bQ z`H;n}m;++7=|M37DM*nrgb=wN2W~ zQ7dL%m74eddpqbzb%vB7glP0aNY&I^?*KBi&PHnjDd#{U)tS2}&$bugz278Nshkga zmCEewAG{|y1H`nslp%y98X=^ux$RVdxbL*LC6Js+vK_J9l8+sE{M%Ietf~AnFgL3?U@P7$E^MwMzk#pzwAYfkaJ_g=qF;&b3CY-q9C2@~eSV&WBu3 zXO=og4&B}nAgh{58A6D&DMH%bPZ$D_(LGMO5l9UO;_l6yqI3@0Fc~28Or>%@WV!(} zYyG;_SREj0W>SU_QrQzR}xgaq|RNR9sSh0qcEv*vyTk}*vdVt;_?NWQ~x{6ItTWvQGGxif^B zesjuveE~A+hLj+Imtpks6E7vjJ~a%3>{G)E|v2kO4ZEH zuB&1a1(0K-r3@h?W)MP@EAvi3M~+>27(^i9&a#lMC6)= z6ODm?(l0x6KY;{s{w|Ol`&i-Cg&%WaXpPR7%K4B7&zYyX zkw1Hl2gsQhQic#BXN8c)Wy%i#GJpH#BLq^+fh^W!P8u%-6sG|sYM)fjhh!gOUMDQL z`VJwQN2ClPBw_?YG#@U(Ki`tq$T5*XGTmf{R?&+GIJ9o5TH|Saht^U#AL6!67k75Q zd0%cpM^<)}GK3I~kqBvLq|ydDvZIMsGJzz_l!bhBzKW2S_lofAkg;v0az5m|I%R)3-<{_ldgM@`JwBj|Tay}%Z2XoE- zc&P14=!mYKlp%zejX}tUOM3W8@sY)6G6;oN1u@AgWAo(7$9r<;hSqVGyh`lRx z+;;*%4wOn6LP(GuLL7Ik4FX8y(~=hi66_@lnYE8OvZu~n`~)ENUrXhD$YdSnlNl#1 z&f>SxHZxS;R=Sst%gj$&N%aVS5B9JU^*-8K4 z59Urm!1&kgVOOnpQ7Y#{=JaOXw`iXJ1e?1uL&^|B3>^`&WxT~~faE{8`kg=$XUjt7 z-DJ)uChQ%%7a-@JOXYlsLVNbz)>>=4b#721We6c@lMu2$ec5V&n0juIv-|5cg~&Ow zklfmvIJBBwb*Tc#c3-KS4{4;ryl>&Cb4m$@meXP>LkO{%f{+HI*5e&#duux-0;%Re zv~J`eWY^`v_$FoSc&VHZiF9Vp@Is6<@J}WNPL?u+kiw}5Ic3^z8gyiKNt;FlQs5)o zk+4s7{A2Q&pA>Q(^`P%y|K=e$d3?W3r6(J`aA9aO}L?5*3N+9JN z$dX!SAK73bHxeN8j!WfyNK1R>qiD6O-0`}!)mbS+2uXB9$bs8=rvRes+Srgla{OdF z@_gQP99jYED)5gX_n$45^C541n1>njAAGn69l0}K$`C@#+!5lSS26$~Hsc=}6UgcL zvXIg@3`F~#{VM3l1@ z97us0^LR;p^}>tL5ue>sIUk~z$9yuw?@{~j07*Y6We6daUI;OE+#Cpy@q>DdB#^=d zvK@JF_BM88$Y<+50O@EcmGdF?-}+!jegvMg1;~=IQic!`G#er2J5_Q4lD_}DErDbN z$U;sHV{Um&Y&8KN_^BpH<$Q?CS>|rq**C9n`OUT@DMJWR^FhengtdDCl8~9^Kp=4o zWg(v$GtbFBbnAsr_fOB{46O;hq;ft)c{wv{xtgYiz|d;ZPs$KN41E#ewN)Nhc3rz^JCi^v zIgnxI%sz6+_nR_66z)sqe8|u5%$n^jwymB4r1XiDA%vv)A;h9v#7O8!Qib+x0?At} z+mWqfu3$%A8XUquzi{fSRL+O2tjD}}@FS3v zC9)8OM$C=c(G4yvfR0=|ES2*iUtE|k>DHgMrv-GR%?T+(2q|2Ekm4huSpZS=PgzVL z(Mx3^8@4m2@jtAq+XCcJpj6I>tU1K&BR!wb_yUl+tECJfWbQ(Q3~aRa0zejcFJD0* zL4mT6M%$Uw_&T?zSpsC9zf{hLxZ3r@fCWN&J-vqSTU;9`We6b=OAzAPxziGWc;z+ONFbTZ zWIGbGH3N6{!|m>80z~<_RL+My)n`soW*_>2FX?W1EoBHH8i5E=z1$3+TyFGzyp=!_ zmdiri2QiOeZI@P70wjKkRL+O=31s#W#o^hDU}&{kC1nU9iOUdD?`famaPzsQcT6;a zLCV##s`lj@jqw?BeREW@`5rmcs zKlNE)x$o|EfLN+{me0Mj(W27Q@f%ol=27*0OEdpzXz}}zJPIsMi>MHxrIfK4pDNV> z%kR3qjuK1$3Qk}8%}7DZqetonU>Vq$s^?oC*D`0pThiy<0L!hGREW@$xB~m)WOFD5 zEPX$IOC*-ml^l!r2j=R1l0)(pu#Eaa)$=WNLIz-8-rn$t0?SH;YH1^cmW3i9&1lCP@ zOe~RWI2Nzg=h32gacTlsDjHGsd`sye<_+Ez4$YjvvZN&yBDCnPL5ps;k!Qj3#&=mE zu~hRcCu%REXS(ArgU;HNQTmqJZtEdp6rIN8UIzQDKEKg@yydjo? zP)=Wr?=$BxgEg1-15WC-Vl@Y-i`SV0m_$s^?pD*Nn!#JQ-d10W9IEREW?Lv=%KXCrdhmrTW_ADq@KZ z<5(_tV(x#bYmXlSmaJ^5o^NSTJsB;ZH+i7N{t*=-v{;3qWyh^1i@_3f_Q+3S30=pr zL{4MQ_!~`(Edk5P%TztzVmOGg^xf8d3-qPe4Jt%vQCo)=t-kd>fTjKdAB72jJpwN0 zSt?4>ao=~|JnIWsbeB{0e2c$<3--m{$Q>=`S5qNEOFm->I`d#XSPu5=t4u68;hetI zGfG3t(g=;SV0qq;s^?o~Ol8htLXLS~g}%%mLWKw|G2z&kAFT=es2N`R%o?Mbja{noJ?v1CMWEK0S> zxWD`glphBc9eE9@o^Q#V;DwfJ4ny$xl3bq(5n4hc(6ZLkWeix7RX_A3mbgtE%edFf zwU|xPhX-H@XhGHUEep?tqD6o9j)~A0o%U3S&|kSWrS)dCv>Li04J<{A0|ygJCC_qj6LU9w zRfJ_{u$Tr@^?b|l-Rscu$-4_4Upj(LUZnlchB#couH&@wj?ElFp?hJa<-?$YtZ61|mUanw1BeYrmI z+9R+WJ4Dsa1rH*9@zRY}R#i>ce610tDS!~5zi|vqK zp9q$9VN^Ze@*#CIT6(?iGz=`8H&Y=(i`+J}Y>ZER41Kw_aiI&b6!R=!FEh(}QdUpj z3zi>IR6XBfIP?fwoGd!}gXKsJ6(Y1`GL~k|zMlijz!_#9#FDw4)0g*o7jTTNuhEzV zmWJL`J>Rm|cL!P~^gDGLEdAzFAwo;UcI->=*sD#!@@--RA7V*};#i)4Vl11~r&obR z!-cBnTl{@bpykR;TMw|D@}NS57KbRb6kqS902bd?_x*__Vh6|au@Q5h-^gZ~5?DHE zQ1yJv$C(LenepAn4lKuYsSu$>V+UGp_;=k4miOCtE+v*Kp5?7PGmnirIc+~!O!ras z0?Q?)FEOS|Z-V7T92Fw8lrk2b*nw}r;-BEQidgb@a{7|kgL&p^u)AhBSn|$L^?b{s zHJ7k2iAsm|gT?S76(Y1G?!>-W7_46k7ROemVZ@TUi(_%sX09v4?6V$%B~^o}=UYr% zZlPtPw###{IO$R$Ld(KkXbD-7JrpeM6%{rSOH4G!^39kz?oU0Z`3x*t4XJv*<UW=CDM0E=;*JKKpRcsIw=bvJW-aZ$Xw3M>k$R6XCKdi^z8 z^k-gC1YlmmYWuVWnkmTJ;YMVvusvk?(^3_6LJA8QLU+ZzGZ(%0b1V3 z*-ikTjVFDQ!{f_`2&$fMkt=(S7Q6Nv zKZ0fb4k|=wiQ0pGsiUtK50=Kev=fLWaWBVmYX>urEiisj1AU2(rRw>X#49Cex!+~} zAh6gaP$5E#^Io(xd%j>3SmyM6cZygdV>p(#5zM*e#IKHhz%sT!RnNEVjd_oj_gBVk z220rxDnw|}jX}%BJ*G)u+3t7x9I;gMEK4kyF(y~J?k89-FQw}FmJQb`&@#U)@&H)O zf~gRprINAut-ZDzECKo}(}<;DAEz%Vnap#sE2jpJ0Es-ACI)};z9ZATu9154OI zDnw{W*@t~;J5oCXEUro;ZV*e_evTz@EOSpOf5|x?uuM{;>iL#NIsd~QU>W{i;`CQ}C1T5;7 zREW@`7K@hE2hD9@k@bS`tA)f;{4dM@t=RQ-@to7cWyTDDUzZt+hN#q+Zz`v$*u#Ac zUX|^Aya7MMv2!bxBv?Yevgcq-Ign7RG5F8xSGH~oT$ELIQz`#x4f(Miy)aIB%;id0 z2OAMf#r@MF@)mXXV${59`6&29)7RpusQ<53A~Q}oZ}xn+ zh`BPibv*hJ#(Gf>RnMP2Ro=D6*|Y4sQ#Jf9e)&|0@OQ~)ewRCAbDo36w#u}WSaJ?> zjs&?1%o*+0XLg=o=}=46^DRa?ZP5~GoKOpv?sfZ08zHpB9K<7mQox7SV6hk{|AAOe zAL3Zf+b~akYIi@|1eUPLR6XC4^^LLgH;l(8CF(P%5TV865L&ETb&3H?bgvs=8u zIDIMZ%zU~)C)lwySY9uu>iHJOfnCsIA(!n7mRW145TPY44*ODLx!Vl-qIXnBg;+9< za4btFF;B1x+of#=%kU(so^QE#tS4HAyVlJGOWt`ZL}&>;f|id?c8*{%x2#YjmbjxF z%aK^-t%5^qrs5lTRyI^U-?E{V30fxGJn;m}wTV=S&|-5GEhFQLYr%5YB&8LxgvWC% z#q!M6?BcWK__q3cGpe3%xntc6Etej??FW`gBd8FerFA@79^J|B50(L*%R3NDCC{R& z!0h|>NxL#(jD7Q^>iHIrD@JG;b0gRkEXj+h5TT`zvCND=6 zUshZh)DA39dsFp%iE8T?qtT; z#s%%$KwnCOsd~O;wbO94{NA@j5Bic2PK5|9awpId>p$ceSPGJtm=Q}c&+>j1a~J!i zM~6dT$xfr{`Icb;%>Gi_Pwy01_GVHcLQ5uNSyO3x8!WwU4Yeee%tTIKel=y@@*I)m zI0-Dn?^E@B%cP6e*q327MYtlO@(C3pv_vFgUuHPoz*Ph)vy{gWOTtNx<>F7~?$6>6 zM^eG^&6ld@TMjFjqoqmb3LIlA7E>WYi^EB@>}gs40Q%BDEYFTuB2IBEH9<^Y8a;8x z^JLX6R6XCaqk05dw(MQg9>!SmPAWub(Kv;cPmRunf~CfI*JNU;;#rO#XDo>$G}}X8 z?hK~t`4*)GqtW8xzW*3l4p~wmLQ5%Qx!G#nE3lMInLeFZ@{>4ynK_*~z7$OK`3#oJ zZd5(ra`en7w45HVv=}ULv#AiFB{2#6a`WNpPhjzh?(Rt}si!%XmPO1wrqe?Y?`<7B zNY(Q#b=poxOKso5mSAx{MuiA13s0kEo72I3upGRgFqc?j&TuTDzZgs7{rSu}%r&Z> zZ~3sx7A>EyAN~%Oo_DAap~dVBTH;+w@OppExXcB_5`30p`JBOAzhrznas$Sg>tw2) zZ)wud1}!}lay7w{KZ6PpTI!!gOZJmMyhd(acXJ@Il=3X93CvpsaZ~okfaPODs-AE0 zKfvrSLF#+8z|yuQ6(Y3cFqZp`LL0#tyYgycFtKDMbNaGOl{p{F*uVQ2Se}(r^?b{# zRm>P`)Yd2uEC;Ko5TPY18T+zpSTr7A+TZB7j#v^?IF{Ba%u2|BKF34AvNngR=UcjN zV&>%ilQ&HVi$XpXBD6TCpk+*7lQ8Iu*2A(*#1eUqWAS;!>@P!mY&!`Sv-=el? zJnk>&HAgoE%gYifL}<}HhnC!?V{^cwvNtJ;SgLuJvPsPG<(^t5uI`A9qw4vV+E8YU z`9HjiE4qynsSu&1lCcClNWr&ngIyNwC6z8srH)$=W3?}uVv z+9$Q&0exBNON9t6Dd({-gHsRTlX(4)7Kex>?E=SgrpYCIg!8G+T@|oA{Z7^ME!9Tj z&@ylTw%%abD?d!y2%#nD0$LPK$5_G`dvQSJ7_r2rax9tqnENoY5n8NL(emJRzbvp!U6-3gETI=UmWWv9nZ=bMjqwfd6OmLs-}2@2 zFzics;fyk{)Z0yk2rX(C(emAQExt{-%m2uEVkzfY`kY`sraeE`(i{4su$-#rTb67$ zL(7|05h`F=vziJKTJjl-R^69qY5L6X3bEu|;`HU^8s`3(dh0pe!P4&yRnNCE~|_0gLqqDnw|BxrBZB+RgAgjIlQJ`ezc$>B}5TLTl!_vU{uRXt7>Q)$=W0_5;wO zuzUL&uoSPLLWCBN%V-%Ikc=zt4h*cHO)ODq97|Mv+zOn>>RD3te2d?7=Kk2p z;OqCm@^lOpBD5H$p`~DNV@2qT&f&Wch^2;S`PPq_lehM*GY!U=aUxaEw>%4B##qsr zqm^K3kV1tBE!B)=UfUD6>ZY-7WC5`hUg7j5Z~PHFzNDXz!gH84J*j%WWyw1|9AhOZ z`Ypg>*_R3tTGFmyU*`2HGlaf;JK^$*STfQ%mT$k9m0fPjR(F8DXq}_#`IgFl%>A*% zp}UH~GV=-*BD932qh+$@L};jK)#*d`p`{255QG zzYO0b4s)bJgqGGBXgP63e;HVk%+kIQOC`^8y8*NB_dPWY-}JgMf~x0RzHMQi@!9T6 z!qpuUY^e~TrI4|d&+v5ui$>0xU&NAkjnfy?uFUyZ=J*|Vp)Xw?Q}uky)S*mYD!kS4 zO=7zuDnw|ByM}#f*s?c1dmZaAweG~fo+GDR=UBd4FvtBd>4$NQ<+)Jxd`si~%sjSF zKNqi&BRr`Pp=IuMw8ZOAYz}>Cqo}J&EYUYO7M=1`qdN1*uAj#O^GEalVjP{o;kO?Hg5J0upHPz)$=VW4Ykl> zAE>?pEIGTV5TQjb6D?U`cW@f*Mq9(Te=Qp-XRIm zQcVHNY7;6%Xvt(Op%)_W!x(#-woHRqGH-JF@@fDxC+j6C;Hn$NTU0&Yvi}3~&Xv2% zP<&UxDVGWnS|V;@U*^@*Kuhq$;a!L&;TFg8bT)GzW>M!g3!pE0g;YJ?a;q+5`EjEZ z&n-8+p+bZfhg)blk#gV}^kqz!CI-Y3ahqc~r_ESqf75>s7C%F(o^KgpppN^?%ZGF0 zVT@UtQXxW%#%;8;U7!#GmaD1-y@{oYXYn1)Smrc)WCWI&7F0dok`>kxEm6np+`!VL z0~I2)lrk2r{m*`b#d6|-{=|}>#pz2;bt3Nj>ng*4fknxgs^?og+?jXbx~^<-4lI}4 zsSu$hF$??R@1c&@$S&y~Lx?5y4#%=#IdcxP(fXYhSccxF>iL!@*)6aypL=J01B=3a zDnw{mcn2+A4=u$nASkah8$m2FcR7}*LCi;63lDX{RnT=Jsd~O8D6uJ8jJ!hZ!4kZS z3K3e&?xH2S;-V^yvFR6;Y=|W|n`0U0#a!>dc>D!dsTo|S>iL%UXBbQNEhBuElyQd& z5nAeJqh&_dg)hL;dg~o~VkzZWB5Ig(%j561<2@zyJybp4a__bp_T{6t$!6$_+d(Qs zXvtwLcizO}mnb!oqNWl{Rt~2xnP-!5-~YI|8-7)@ODa{*w`^;piWa}kDi^?V{VEkA zv_$1#U#zrt;62zkMNY27l9_&#p`E&rSPJfQ`jRq|v8YTMGYtBYZ$Q=aEp-CwphaQnP+V>6W=w?$Eh+c0 zFWC;J_(k0>PgX4_mb5&MWzFT|xbKfh--2UIub8UmTZ~sRtJFNal!roJ_PnD)gqEN@ zv@|=@?>zJ+Z1C9C#1i{}W103j0WH;Yca}q6zFSfCd`r}2X7%9BWy?;2#nF}u5n8Ms zpk--7E4)T-GPl)wVhMf7vBdQ`g_f1M6GOqWXc1Mni1R2>-&mIg^wh|m)A2>asu z!V9k}i>;RKBbL+o97``{<{ZXm_JE^cS!YYt^DU!Be@4spSGm8y(sdFQBD8qqqotwB zH*@HV?bo4k#1i$GW0}&GIk&9*(jp%$I=`uUzD2w2D_SPc8yN%^CnXDMBZL;i$7tEM z^5z$?^z&CeK`b>qi$OUvCtsQ9j9=E%T29sTE!|qZMTQoM`V_f3K}9s!F@EfpfPq&>mD?De^T z?~pvu+L=l$8BaNu%8`4}a>q0ruV21&r|S8Zl{<5>FZ-kH@ts~}V=6>w34Mx|W!su> zg}&r=nt7F2;tDvHxqfkIIW(p=8}^qRJ*u8>v58~87;?ek5`G!nsuvX^wAd7&rA3Z@ zDOie5>)#@l@Mj#0!x!dVg~i2-af}6AqU!k;(?K7xFIg6O_-;?!bt*(?Y5fc>PMhcB z`PiDY&pE_W$+LXg!yGFErrY>IUs7&S^?b|b-p|mY6yIwb>@PX@s1Tv0kg-gwV`Bsh zn(MsPdrU05p8w-Yd#w?6INF+DduaiGZKbw~N)oKJFKe%fDVDF(Hp5DLy;>^eKdrR? zaM&2*I=y*64OZI!R_rG6{%NJXs=-!_GI_Gp99G&lP^F^&zgF7!Gn>^OR@&caMn#HO z+W-360=I4Ti?=UeL}{v1L1aY9th7&D_Yp%1hCS~BEA3M>sgVD1rTvdOMhURe-m5DW zCcD!9h5k=mAHUbuVJ!?yM?)&$A6DA0`}P}?`kr#w2rKQUnovpq?Mi!@4YV7kk^(F3 zT?bRK|9Pdo@$=TW(%!0YULvftcOOm#{#PsQb1p2#mG)5^Z>@q^qV5k}P3@$9^5E2^Gv>G38HEvgrHG=aWc>_CMGEvAKNNm2NOPq36ug?%KJpdyYX<27@3 zezI=vTjkt;&W>~A+WLSIHdvZ*GPVxFaC_z9fW z$9)_#2P{T~R6XCKu5%wPBiD4r?<+reLxl(}nT*B5X>A->2HUs!O)Qx&IejUw$2`Fr zSa^FbSTvlddcNiJlx(zE)j5D~NMz5XLWGuxm)IAbF1h`{ayZA?aJ>XoVu>i`SQfls zEP)yyaYaONW2&BSnY;Q1_QmLP)LPj0kF=sfgcglrv{d+xI}DasyZOzCrHW@sZ_m6{ z(CqY%p3s-9sZ>4RQr(}K$1a;{YyeAu8xwRDLW|iOw2bl^;S7Bl5wzQYSb|GA7OOaB zPM&!>2UnA@Az(O6<7{^X|st65nAe(qGgxMXm7B*X*Z(}v6S*GI_H@? zzmEO(7(!njXjAok%k@^bu`lhnKE-|CzB?5nwB#_B(+{uU(}lqqx&w$M>n*1*Id=}> z{<3!Aa=gnh=nhrSx8&trL`xm(1@6!n^#@dl&=U0)`!djP4t|j;dUVwgVo5CHSZ?M- zp(V)jP6_m7o;_91w}#io@H)d<~%tm z))~JQy;Fy(=UcuXNk+>(-4kA5u{EGVgqBLiGRPzXuf>cGj+{s=1?8N+lsPa@78l(9 ziK__OAE)a1meXF$Jmw_tjn~_{XQ>dOC8Zqul5O1(S9BlU(_|X4q*ZV%BlY5Oe>u@% zv<{51i*Zyv-!dZmB=#lM{5F31s3egJ5n6&O&=R*QXC{oXgBD7e&M@y~Bg~?!eLuAm6(Y3cGnT`duL{7jp=7`!V#%rG^d);3^S1he&WgCw z-uweq&$rC#dBXYu=|zRzU>T%Lg$OMkAJG!Hz^gse$3yya9lVSzk7Y#iV6{03_qdebl0(X-!i{y^Lk>b;aSel-G_Z~Y`@N>%sd~O;Qp7>*i_ZXSGw93ax?7};5L&7k%jYsb9Aj38r$rJ=;b%@?T0CN|#inE| zZv~d;iBvt`qLzFdEv|OOhG1EkLWKw|X`it#Wxf3HP2yg!baoL-Mis|0!k9US2~kVJ zzeO;&oT}$rUaw;Q29eE+&)#5BsG>rIme4A+yzMmB5B8TeHWmAbCGHEy5;T)JhnZSu zdJyy_(2=U=TZ}B2Vl=<0#(nqH0!z#$JpY0cV!j?8i2_3ZRq z=!?})s-AC|-jdn(T}qqcIm`zID`_Kymbh=&m#W`C@b8d}ucL91SW>Dv7Taj%{@C}B z-gtj3S(U2iTLvZ{#l9#Ux{?WfS<{>f5nAR}qs8QQ6~1}zHKOz?u|$98SQILlzjLK` zXg+=+bjvuZo^RPVfI05xz1`UWEX^lTAwrAkceFg2J5ve9Sonoww}>U^2gfp~gz3xK zTOr?IjMcqH)$=Vz2M%FhTrviSgC#7B3K3f5exPN{mHjQiqW&--mspBoL0>*AiL%T(ahga-CWi)8TzvIEEOWOXvkT@AJ~vf8MqkwvT4T+C1MF)!Liu& zWS;L&@%l6m`Vx77s^?q2KV$mh^R6%MFV68)h|m(uSmt>dYJ)|tc2gr_Ddkz(-DfP_ z6cX|4jW_FVmg@PIE;|q77^{=I?jiJLeM2flXenhZUbiar!Lq2tz6G<>pCO^eg7x zg}X^g7WC!v2dbWLDOkt+J=o(Hj5NT~DD45>r5h2rW5`9~SOeFYUFwDhivV{FIo56WQiJW}gKEK%z?mJSD)dlwhl zyj~1rY{Mz4o^R>u7>$-zikJ1FFU`(VAwo+OV{zVC2fsw=XmfQYvDEM^gYGl)n1152 zX<*sxNY(Q#n#;q{a^SQvuITRTLWKw|HH;{9v5@Lzlz_EnOF>|uJ&5(1@ zmkCX&dcGy=>U!+U%mXi9f<>n-6(Y3gDq&wLe>TA{Qw;6+VFj^-Z{%2(bzsKWYrRlB zPyW?~s^?pl6*K4jN8cK@2FvXpREW?L&RDdzhRZ=;96Fp2C6-E_<)IO?(mu593cN_XM^E!&$i=gF%=_FI6(rw0`xv{W*d)1jK(!E)PU(?(*+i{SJ{&5L;-=Ct!?ygydo zm#XJmety}2V=TAmE5378XE7BbwAj_hG3GpWJ04%w{dU+!EGe5fmJe%~`}|>J!|)4k z-pYQZQunsMTKL&M!zSOEwAwo+EW0}_072i>`?cRAeu|#j?Sb|id)2rccDu`m4=^}7sXEI#S&0b&W-!m;R`V4l6M9Z?_u zy2;g4s-AD@p~G03`=6Nt7RL-KL}&?OEN7R#z`lGrmw1#|ig^}8OXgnR`eu z=|k1?EiGzRVPBeRnBh7;?}1c^(4y7=`!fA`*HGxot_k<95la=%qBn>+?$`ZdjPEDA zJ5lv~i{E4B9Ok7?bNnk!_1vitp=CZ}Nq#?ZC@calSe~9sEP2}%{^{?)CuC2;(YF3Z z1+GrNyOv54tWIy0ri>{j*X!xP>hz}1sg(b;I^AJTUyR%SEMXt4PQUY(iua|) zG3xv}`x~%2-RmP2_5Zax-OKRuP*|OQ|2q{aUY$PlSR357&G$Dpggre${;JePGNNQw zr;q(ujv-0aV{gFf^i9fC$p5%Hy{x#iJgiRdtwx2(u1?Rn-VPr?HI03>1_tJUR#d=0 ztWMVn>4-_!`h!SoRMLODI$dT1Jz`dWgw^TZ+Ena+UY)M{%K%rWUyG|*53AF^ zbfW_QtJUca6ZYZi^pTye8o}(XX+&iTXNf#!me}~Dq6Ai_TlS-3{xNJj3cu;eQW{qbRk4{oIQ;7rNgOu{_N?dx*d-MX}zob;CJaTjtUX}F8R#w za{FZ?d$3&p`7WPWQlmJQ;d0EGQ&wXgTy48qv0SR>Tk`KSXB$g*)>Q|~a8)WqXc?f2 zN42ub#4})V7=QHzvBd1)SWKgtzRXCSum}1wXbM%&x7hDwp88x|pEwFE!8523p(Tc~ zSk${-4=fr(!%B!HcqhkVSMLB$g;me|!b8l=3X&BAJzt&yL^4yK+AQsd~P}Y7lc5Te(RvuIfA+LWKw|9*pJV zRxNd~)LiROMJ!pnIDN57XYOKWAH8r3`Z7G5s^?o?b`Hh9{C;pV82Xa;kO~o6vKY&d zytq=Z6wE03K`e>U97|9Nb7kJTtU-6MG@MP<^DT3JF_spZ&+u&z|M^si&|=sK$5{7H z0r(x?VR`ZLlm2?!BXT##(xUGU++RM2r*8sF#51a%Zwb54oShFF89p5PV*Hv45n3V{ zi{0t#8PJ#7i~jYArJ836vuExyOgX-_9at1^Q1yJvwS~;SU$k+nJy^W&QXxW1HDifh zRf>N@b!Aci#>7&vhtrpKip*UGug{Nb!4gtN)$=WnzA$&O?*_Rp1k0pPREW?rr7@1N zy!tH%L0_J)S8PcvX?r=Aeh$o4lT*W;J;3sQ8&%J@^lr@TFIEQ&wZYPKFBKxRq%oEe zEzV2;OAo`GcEl1J!?DP6^nL@}sGzd{hg-R$T zA%rA^5JCtcL?MI_k`R5*+20?m*Sgof&pDqzV68{b*522!uf6uWC|faEfAJZ1ca$1K z;$_Rv>rQxoIlOjLUZ~6di6n{ElEy64_lz3~76XsgO_e2Xw@??`E9^Q}_E4}BjIjc~ zBwn_>ezguQ)hB(l1xr!@Nusq3Dvo0;;n{h-j-B)Q-C9|~0tCy4?d%!mp7q%LU@;9O z@v@~@4|W|Jymf#-Sb{@H60IeSS%Tf3V0Xv+DVIAbi&vmvIkcEv#~M2(<5!;vb4a{w zSy7kWUoyWX+yaaHB9cUFDOUpPve?WLuVYu|Zt1Qp*^(ur0qcY;ebN_q6y07*;$_R` zw~KL%)tl;%J6BvCNfNDPIkVKcV1(~4pKFilr!2{Pgt{2iXI~bJe2?7+W6Y~DiI**o zz1Yr`qg9H!LS0U_CP}oGWM*mO_&OQp{?ef}hAK;RkYHJ8#X2Du25uSymdO?*Ubg(~ z>wtCXeli~GVmO&3(OSBd#4%Q3h!1wEoJmR>r7XdF1xsfK_BGCSdSmQTTa!xSWy`?= zj%fM1=pml2H6WcN(OQC;rKjQ6o=}%cW?`1f;u`;vIsGNm!Q zjvcIg+YQFpiy*BC3B?2tn`!1WREa{SEbzOGfpIUlfI#}`zC-Jf+yaRjY z|9!|BugT9XND{5ZhFQWN4}1ugdgeoBD@*)7p)M~6vupCoz$cDiu^mt1WlOCAD{+jK z9}~D5>QZqkNussHGs~Ch<-@_!`C$14$`Y|(utaa&hVNos%(pcI%a(8wFI)09Sc;Y( zOO96n%lxw>iPqAlG>)->Gn@;9#UM}eGG*~SAXw5%u)6FrX1hE;7?OC|QmHOmADa`> zsS{Y7OOPa5i!ZYTOq*HT@r*aZ{F)#;*>N4>G zNussHFiYv^8F)<&FP?9kvVvRiX>jPys@&!x$hM;0y|YAjY$%% zr9oL7Vz0&0Li$hQs@2Mn- z*5bx2H8xLd5A$X9F^@oH$&f5_ZMWhz`CfsRZ=f!zktANWq{S`6x^!)4GXyNtqe&92 zC4*TWS8JFXEL$us4=77Qs8E;GGHjQp;p@e|V7WY<#LJdt?(CUAs>`=GU}-d$B+*(% zmcudDtZm2_u(WE|@~E;z9uX{S{MZ_-!|<}W4%4zLiI**nd$DJj`Dteiz;dMzNusqx zGD{=ZB6#ZA)yug~DvRGy!4jOsc1Yye?cD|HviBB=mo0Uk&creH#rQ{Iu*`fwl4vc} z%41!|?U{|gLlPW(?VPeW9TP0KP1)K-&CujLV5xMJ#LJfDig;}5SyClSr;*I^)X@xoBjSB85-(efcCmHK0hMg<)U)G*ND{3jm05hO zLw3M?8QW~s4P}W96Y6sGG#g`nbQb;PNuss%tAJxH z;7RmUurys)GZ#6c-gXk?rfYd5q{pAp)SRfND{53R7I@I zCf7H38lBgVBhQp2OR~6>WADg*`~C`sx&-7s`X^qtq~Ekf%f6M9@%^RCUnGgvvY1(Z zUwMT)?^oBhPgR!06GB}gSF-i7H7V{>!IILL#LJfNgXf@S=)=5tP0nadl4vc7%re!? zEePgIzh^x@Doa#^U@29DeY1F&)8h_UO1&fTvSniwdwg;(SW5G+E5%estxgFTvvfh{Vel$H_}@j9pFrfSr(yo|7b6i$P_qOUI%0 z$AYEPp+kAc{QEm{nq;wP$WA~oFnjt4EL~2Jc-ayk$nKSs=Op5FEdM!@L~F5TmJR;r z%Ymi!^96>=5_ejt%c1Y?cujtM;88)alzT_wWy|O??0L*CbLKM`V~f6!Bw9-xvz&Rd zZXj5uSLjk)S;8U(i$@{$9eM5iHC4c3TZ6>QmOG6W;21MZ8Hnp+l?f@RWd_P*c$`Ck0(K^t!pFIxsxVROGmkBuo%mjeDIiPqx9Ea`uB z#`{aBWjCrQOSWXGb&lODV=ooK`O<$iiI*+MH?y_Axp!KW1j|QPl0<9CW|sZ7;ptG9 z5evO*DNFKMp)Qp&*>&vf!aK!ajM>|fc-i8xi#_v)m~}Y|mTqfF60OCuDvq&)`*nJN zrS8&+4U{GNoM0(+#sla6!JRenJMz9&Bwn^yj+l@0<>ZIWx547#LXv1L(abVt-S#42 zsTEqMxv~UD36@7i*!%viJJU-;T@p`{c-b;3&JHaPAJ-TRmKov2*+5$)(G5lb+!yiqO}aGj$^F*)Eon_4EQ*F zfU-nf6f7M=*%(U*I8_|#V)~uL%a%I6^U&hh*JK~m<$f;nKYxVQ62UAjT$bR8R4I?D z3|AK4Xu)#UnRP?$?$LKGSh}W=c-bOqP5sFi&@ggG#F#;e;t~r zEGd#Dc^Z2jJ6`b^c6J#QIQ%DGw&b74)-4yiw($i^P!WoZgfPo=hi|wxKD%UrMatq9 zBUsK3XW#cN#+kZ;rE6spFIyV7W9$6w=hvAY-H#;dy!d0Oz z6%Mhz)_a!3;faSPYe~Fp*;}1`uk7(99#7fG-bj*YEeXuB(x*ADjj!l3-9=dpf2-a!(`(9bsY&-5x?pT;4(OUeNWoe7OPB6y2Pkr@L7N=Oj z(qt9;Ug_CoRdbm84=$2;*|OruLcAvD9r^Jf)Mfp3l0<9CR~zdx@#fr#VDW!=c9*iG zN*3RrY`!$I|KI|akmn>`w%j<&_V#=V491g>KEEYNw3cbia&3*L9at_JtqoF^*c(D! z9K6_d?8Ss{E5Y)(G>Mlj`RlRw{l~K}iZSnYEnNwk(GCOF1!nH%B$v5?E4Eo%Z93 zjK|TIT4vo}@H_DRE|N@|PWx1wkJuXg^2+sn&}sknI?3{%I_-1MXor7PtHCeVL#O@u z6q4?rI_O1Xy3@YN^mbLTLc^Faa z3@iVsq8d?u>$I;C_zshJw~Q_Uo%V}LkR<=#+fMtx{egM14=sgG`wC4++W*{X|6_DF z?6m)Vxk7E|v|rPjB>b;B?Jf7;$LVzZei7Ww_^>m{sl7_Lu&c!A2C)mF)4oAZlIDNw zw9jXJ61zLD_}{#tt~gQuy5jt;)4q0V_T;qZP_@S})^F}7@$%Kvpd)(&+~++LPax=Y zlqAvqU!vImN6BqOnl3HPF0q;c%d%$RyQ-RI>>yD9TP?xzQ zND{3jj9Ff$e_RP;EWu}0rm}e57c5IYvsJmlEvw*ltV|$@mn{cou=(<@pQpmFajy@M zBw9Pv%nHJk;KcEUc1>>Qj@byAA@Do z43b1^NoJN;rVDq2r9k8(LuH9h5G>IXSY3jDWR(L;qe~=Swp6~wR1b&*)Bc6;yRtg%a*?)r{LS{E)UaFV9B;2NwgND zhFBLr_Z_(Vc7}E1s>+fsSuA{9@U@|SrQFxSvSa~?mo3)ehiHb+Ovc>QerR`)9D+*hJ!GORy{3vHI1cj1g}=H6Uce4hxGH67XWSd&qe44^K<+LL(MGW`kLA+gPCXDP6Z?@E$rExydM zvFgkkP?r_AMz&KH`=^5C`Fplnx_3(b8esX9K;mUfgI8<^-q}gx@H#dsnSJRQY6c=O3c#r>M5Kr)i;rN*%DBYJzIWADPsY3iQP()Xf2bOK}V5m!x5+q)>Sf6C)Urya++8r#x}%ZlY4NzjyRsWeqP1i&%f*K6@ND&S zrwUC{mW1a*UA`Y=_ZQn)i}369gNr0yw%l@>fMe`?iMz+3F7DS!60K!qQygPOT@u$p zUCN$(I$c>JUkH{ePuO)VcbMS}7-RJ>ka*eR=*V_@1vRxU4wgq(NfNCkl3BVvoO2Z{ zxnAtEQ5L@x!E)7z?TlR1_6Dv6zf2?Xvc;=@#Dbq zeZZVW%Hs4=up|VqFD3JXD&r}EYetfI*;1n>tIMeU?{H^iQ%jOWYjI+hCC&}7gJ^g_ zi&i*XuJj z7hRMk^tE7dKf&JjZ(1!Z3YKR}NxW=XGuRsE%b=@iouMu@R*@uHODMBk{q2Pvs>f=& zxGRfis$jAAWAnx7!Ww)>KGl%K%a)@%+1iEGd((Dc+0=?8(OOEiz`EGaHbqOvuEV{R zB}=j#UCzc>ip4zqJtdQVBwn^G8_U+m24#-HF17N*ND{4OF|#;*WhWraFIv%0SrXp} zb;(%A>eBL>lR6(wV1ZVF;=MZBYYn78543)Spwe) z7SkM7myEb_c)C`?WD+l1jx1#N%JWM@@b{F`tVt5BC6HN0c<#n`u^2DAFlBL06D;%F zvR%=w^IlAXy7b>o;$=&@g>0?Qrq=`9h4XPANusqFw8FX!@OJzLW9;j_?x&R{O|lHi zX1hGMAAOB;KX(#|mo1lnvN1L%_rVyrj#Yn6l4vc~%(9{VqrG66`Jm_pWr=$))Fq=7 zdmj61vq^cV%kg9qFIyfpW4odYT%L))4b?Z5B+**pn8oJnN!+ozW7qu{WeNKrSSkgv zXZ~V8oBjosBf%tIwygO+4d;Geck{VWmjj_BiPqAhHIA{o%V&>)x(rC)aZ6deJ_?qt z^VwRT^`gjmVDbG<;$_Rg_iQ)e{6gVP!O|%A&Od*I*5bu1^$&J=2Nt)rR`-=9Te39q zVs+{8%&8Yx=DLx1+2UOT)_~B}tRJ#{ zHm&b?egjL|OcF0!9#tEUmhjef&Olu(=945^i)9-eV}8rpwuQPFumAi)S)xA)mQ5Ym zI!q;pS$MLZr3ZGzAV`s`BqtiKMR)nGnnPQ z%MRS-ndw8~WlK(8Hpap|Eb;e~%LI@lT1%a_SQocxMe#SfOFppwq%5vq1j}fC>e;yA z{VZXOWhaq%*|OY-?N&GHxC^_}FTEy7v=&!pdC|682^eFEU;1PzOS)tk{)4TJFSy_A z3|Ks}NxW<+GM7E`-bFSEKlb0>kt4nemU7M5 zeg8>?Pk8ECv5q8OwoDFSr)=bmDZCZx64Q+&(OM#y#U<+3Q?Rs+KVYORzL|pMY2}T$ zo*Zk|wmXcm#!pDRY*{~XE?$%EPi;N`mWULRL~E(k9_!-#=p*jFO)9^rq_Wt56D(P_ z?Bt_cKg#1y;wWPhFI%c)u}(-gPsf#DnNyD>(OT@8rT^%uc$$Ohj}8@-B}KBF7{TWL z?7Uq+L0yXF?f55NwnX}}=dp(mV*Ect1VT`rQpSQZQ#AFF|**Ti6 z!7i*;3ZE@279sJnrLr~aw0{y&p%qvzlqN~EmR=ojj4i3O&l4<7JKr)QZGENusqB>xgx!ez>*`SdO0`(^6S7Buj_0>;!~b zHm`Br(&##gmo5E$*falMW$h+|W#%1{L~F5SmOR1!IL2}g)#{)u2|tCp^xMhy!xT=s z^%UlQwFnX~TMF^hPs}R5Z2^{p=SUK*C4pH!1mrA*y4-E{+e}#^vjxi#6Sk(bJu_?> zShlq%@v^1kQMS(S`1J%{$C`E}Nwk&@op6jDoYK_?EN25R_Er|ZUxFpi7mg)XUjYtKOcZG zw!AQjmo4uMSl3O&h{4zmRlhVzqP0w8mitRbWP!!Qt=u?eiOmt}axXP9E0P_byn%UXdhPizl;mYj_FwkXY@SJ5O1%B+Hbp>}0*)Bc9{Q z6tjX!ylk0!p4BBM_C9tFJA{%XT1ytQmh|a&J=_md zdM=5VEncVC_e!h5cDN3ccL_`K%yGw3aAlY5l!&Kd4KwtC!a(OQ4frS^tTBd$pK7@ixqt zpgSa9wq%WFmU7LT#epUNW0FK`scD9F*>`-jKUiuw`fX4a=hcG6=ofn)JM#8Rb+FWP zA@Q=Mks({>pZ>(n8!T7dNfNEanOTZHvuX{NfgP|lnX5N2aqIMOB%Br@$d4l|g3A?O3zjQ8ND{3jj9I4KU0Vw*|2hF7SXsQB1xv_0_FlQSnBz^b+<#8uWy|Lt zY;Am>rPp+@)OtsfXf5TsVO_i(ie3VXbxo(k%91TvewSw_99ZS8m^5j@ya|(*m>C-6 zE|JSP-?Wg4_>>v7-sdiK+IQGRGHE*POYe!t)?kOAI|rfDK7TOD@}D~GW4zkqA9cw* zFdaJWJB5;T|I}%}e;w;6$@y^0NlPWydPB&EL7{^+?%$c!ub z8h>Y^a|}tOM%3Rr?I)SAPWv98rw@Tn`^=jp$^Y1C|7CE;YS3wa5TH&}qN@Imz_jcG~~#4}8(GUM6(fe|t;P{^w5n z59LN-r+v)e9TlO|e$FS7@W1M`j|k3%10(mFYxv|g?>ot}(f(h8+5ctm;S`--DH05-(fg+Ok!-q`^CI zwe?yzVv}4UbZw@#(sCy_T*Lk z>N9HuNussHGfV2#aW}zI^rUHuvP5hYEbp(fr@)m5S5Jj8_V_%Bmo3qbth4K8w9giJ zatgRgl4vb$dg2%x`RZy4SmxI&_)b}THwhNwuI%fz#hXbMVCmPK#LE`1jckWR$bvrj zJ0y?VktAA+FSGbLhL;43;mMnyl*QgnutYs!s|?fa>f3|m$axYkTjDpe-Gu2DS?qV< zuaP8LORip6mxb34S%YPcpI4T$q)3(+ceV>>%z)B(nnRU+Bwn_3v1hwH&-`eg1Y_(* zC`qEVOlFn}$s6%}vWfoVbCe}!vrw1so7r{j*@6%gs7td0Bwn`UdBQp(rk*eQ11vGe zND{3jhFL0asF(qDd3U8|fpP!d?GfTGSbX@G#UJkd+k<8IZ4xhA9$jK@v$=Pc!`-)o zACe?mOM~7x#sX)!;!fgK0iTSN#mz&o4EexTOn!Yy$G0Yr5E3t2icV*%KgT1I?Vv6# z!$=aX#f@2>PWpwX((EX9vZS(PNS2h>ZunSsaonPHP?u}vNxW?NUWJ_-H`U_q6tGxU zCrPxH3}$&3vJ5*_@(;ADpezZVLS4?Za7N3o+P;Uu@_Yn|mn~~ou=%oK{sU96M2{y) zw3d;5aEyI3D2`{Vm+~A@U0EWx2$r-Mc1`x%^Rfh3Ht!(uvZY}#TV*Jg(+$s99kquf z(OM#zW&he6_}jMQ8dj~NEPh)BONAxu*)l8B3{MOh){4Z-mZLq`v*p$I%kbIKs1r$| zwN&eib&1%Pie1pT+P!L|EKb`5OZ8@K^~WcpIPNO&F(dJ^rHu<)#r|^S(-0VAgZq#q zT8k62?5yd#1g^;mUqV|dOR8jP)Q#O=>Vyx%ZYYmz5-(dmk7uU_oC~RV8!Q9!9r^P| zXf3JCQhH0?FtB+3T+~5XV!edA`1N34Nmtf-e+cHw(|n8n#LE`r5zhE!@ng#xxbpkP zh$PWk`t`#x_9Zk1yKc6%G3%-0eU!!1N3cYOvlaPo4<6xfb+(#M;$_R1&uphx;N0SP>Y2|n zl0<7M)gSB9aq}nKvAQJp-9gHdC0fR^FTUL@#^X*eTSF2rTgta&W31+>q1dTnT7o3e zS{5^lo$LEiP?zY=-XoMHal25L;8^zlGPd?++*M%Mo5ahOoMY@u$^E(=f?@9a4kAgk zmPBS5xVJg}7WByrlgBAbl&@eZnay^sZR))UPYK+0oy5x)gI?@;?8Bv|xbwbdJV~Op zm=3@(w)I&J+_^HnO8Y6w61YRK_||0imtnqnX2NxBoe7DTEhGA}uFlgfd(MZtbZ<k%NE-MY>fSNp=}CS#(gG9w3Zfw zaE$HEuGbVSjdE(NRTeLQ!P0CrJGpgiiKF+xa>AhIpLp3)p*dSqT7SsL1}rm+ktAA+ z7qj$CDAEQjwrOuSC`-0vIWeB?@V1?ht29_nWs-Q=QYDMM?`J+-d3xOE$Aa zzqs-NEEVq`@>G`O-9lY@db8cPyQ}5L-M3|uNW5&xv}BzsYd8EX1D3t7ND{5Zaxjju z(1iS*!E$KllI_Y89UxdT=CE%TyH|C`j&8rbBwn^uoXvh$;kk2U1Xu{AxkJ%S~|l3kORS=gKf%hN+7Ubf__$i5{_X;S7FSo(yMBwC9rx744B zzooXbX3}A0NtY~jC$n7zNga;A2a8D~5-(e#f3Ro%@hKJDz;dJ|Nusr+Gs~CSc{V~_ zBJu}?D@%NkP?ulb*g8zX0t@gt*{KMLmo1;u*qTy}am(>!y@O>)60K#}P#j~%LG5vO zRDZ`=XOtykuVCpkly!8!*pj#n>SF9d;$_RavTW`z?OHhm>Js5jl4vav%o6>&KX$6v zIk$~g7T;jO^3$5_fk@nXI|D3lHk{*$b2}Jg zBW-eDR~Gwyf+hJYdrp2*WcEU^JXu2GWy|sn?0KxZU$;h3m*oy5iPmDzED3Q_PJ^X) zsjIh@B}K9f_G9bGn=Wr?0+wc#NxW>?cZ9vaEKlr$&&fVDNfNCkg<0I{-uVcYWsf`( zlqF`rP?w3r+5KhQr`rd?vhW3omn~U-?3usV)Z3m=m)LhCiPq9 zBxMOXAXrYe*^F!BovvTO?=N?=NW5&>afr=*(*fbwrM4^A=s$mi))K-j^ZYNY0L#|# zRbMHKTZmxUd6_*Y&u{zt1=OX@G!idcx~<%VYkf1mKU)KJd1OP9Xf4Ieu`XXPuEjN_ zLNni`DNBZAF)YunV}bTX9Kf>Pmc+}JDwEl><*@;uj(}yra*{-Av1OKvzV6rsz4T|) z7iCE}DAZ-4AKR;7nXBXqusGy9|0iCy_)cceV+OHa9l=u5h$PWk5}3vKS%r}>#)i*Y z_d{7C4+)lhli4@EK92o8!18nfiI**QDK0q1nh#!%?=Le~kR)15hY>i&0{fL)4wh<` z!*Y-R_qSKS!-6IHHG6;QbtbJASoTaO@v>!j1@?VE*=mjt)FszEl0<9qW0p>1`sW2p zt<4n;l*K7jutZE|>z0#qZ_NSAXCD$TTS|3d&tvgs4kN)bIDjP4TJnv=y0jS@whSz} z%RDQhEUA*kZ5Vq;9_;6fJBi(^l6cv&W+Xc`VEM~>IA7M8kR)2mG-hf4@B((*x|Bay zT3KR`NOcL>g!h+=PQxlgUG^H2c-hjd58FxHvOvOWs7uXyB#G7%%Pg@$Pewpp>Rq+1 zq%5IF1jIu@g@X2)aY!WlJGvHuv|O4BrVB!^b3v*3x7YjxqD-_gEKe)Alu# z#q*e8DbtmG<9k$dBJS&}(TBv#mVmXaF3}~1PXbHUV3I^@@nn_>UDqcINXf0XH;?nwTL9oQHjBTPUiD5!r z`lPXE%QvAj^1yt#wU)%omW)B{o5jb2pRg{q+(;6w#bPv$F^ly>%YkKPVfWU`5_Md# zJet79*o*paaj!z-G9+HMwCT_GC_epkb2-$dSXGilYl&i(`S%*&Ubqt`(>p0kV7OpO zoWkl-zi3xesLO*UBwn_B9K+U=SGm^1xnH?0NussX9D{YSov|VU>awqRlkUpmd_u6a zdB?t28nu7o4;JH!Bwn^GXwJUxyAB+VCxd@BCP}mwXJ#p}t{495jD6wEzRHp&Syr83 z-(S8aR(K8;mog+?wrt$NPNvw_>S_Xvv6WRw60IeTStegUz8&gPYw@`u$`ThL)Wzr) zyI1Dj7W)z`GaX30Yx!Tl|j#p{${DYBJa z$5NZz%YwSp8cpJ5%fbe;@tMD5!U{b7nL~AKG7VBbGCB8dYDtVQgsw~-(#m<;L z^Pe!OWCoVp{v=+ue70lv%H{PJ%!az04<<>pmgUUy!>`9*U@0&;d6u#ypBC!Uj=%3$ zUDy!!TwSv!@v^1Lb~eTun?~jZi@^etL~BWAmPY2MaDD7tuOM4xiH;O3ZLHa|<(xB7 z!=Nr+gGsz>`BK6T$Jp<_O{~H4b`(jXwR9VYW9-L(KGhttFUQcJ&OweSLGU_Ha}d*Rz5p*_zG$(334r zLtT2@A@QA*9g3_|mUPK-VJKUNX>#uK zNwC!ELE>f0g2`;YT%T-x4(gJB07;^?*f5Jjw+QSU9-Ex7Nm=6033bW$lZ`Q}hX;y- zCF~7}mn{XCvS-WBPdeb*_|#7%iPjR&ESc{n^P!PIb30hB3A) zg~ZF2`JU@=?ib!vzz^!uG>s(DT5?Umx?D@Q$F=c@&foScONwN9xQ2bNO!v3*2Fui5 zBwn`6O=I^LyAqf1`-{mSl0<8n%q%O8B;zmVh1#AuqAW2Ng}US3_cAjK&N?CILW%-Bh^hqlw<4W1+PQ$uE zcY2mJ$)xE{zgqngcBi-Su4xS2>9ZD)EdQxHy}CyS{G)DNTQ?NC)9b7t>HeuZ{nH6{ z#?#th!>Z7oe$)rWCsLc1*_ks48d>rQ{L znsuj_9%k?tbf+)&B}x9r?sV6v8dIP<-7|nB`+Il#S@WgXQ{JY1jl(c7@9!rG{-Ha) zaI?YY5#L~`o+PsushxF+4Wn{o$hy@ zB>bW|`sXichr5 z|1!FxED4u|y1d%Mo&ujwG{rBzGfI(ojm3mLZS0?N6IU6^S0YKYmXVWijGY{8fxo4e z$1yHJSt73p7E24(rM7KCY+0zwQWp|0Ti(`ZPl3UYt1pAPM0$`UT1zCeOyB%;9*nWh zt+pm9i(iajDVU3`Vjmpi(HtxbI+J+W;@yTkG!!gTdy*ttOSQ>Zml55cV^^p7 z4$D`{;&fH8G(O38tTui7=`vV)1(A5!GIJT*9d$Ut+W;)z4w58Vixac-^xlYFYU2ym zeXlI3lBLrTc1?~BdvO#j{fm)!*%H%oE52zwIG26nLJR8PX^SvLx zC`;@$p)Q67Y>bU+oU#ZkD?&-UY_Tw8J65y%WR?X>+X#|GYw0%y$C!!5KP#LJfWS?t+zfwl1n7-Jp2ktAA+=~NtJ`AV(D^T-ST4z8pu zfwu%pR(|#jbGz>-+=WxB;J!ccvL&S#yYD~E+))=S&P7NPttF6IzVEp38^+k~3G-?w zi*uY{F?qnAVTuH_!PU~)Gf2E_85+x;llQ$SSR5>s=aD2@i@`Lk%l?TQ@Jsxuz&7=i zB~7vvOJOVW&6f@70rSQEAc>bPwq@9SX;C)03Ro5&CrPvxYi4;mw>>`de~SOzL|Ni) z3w7x-dL6!A9@{ozC|F)RA@Q=s^*MV^o|5-P5LiaOBuTWEIA&>(m0kkI*vw+rS}RM~ z9l_!{jMe4Uskit%R_M>=dWtYe(J!OS91=UbbxWXJ6u{EKNQQmYoww60OB@29B|O4)qMdGIChz5M_zJCs^uz zW$*iY!zzb@rP^2$FI!%mW%tShby{}-%h4$$iPjR$EEX$m)`R74;^9%s5`15<6#2=% zE)25&#LJd>)z~}oo(G9D!BXxhNusr+Gt0B_X{TU}WmhacTUp{0gt{!4%idr1Z8EL{ z7VBChUbZ|Q&c@inDTCUBWnx2;L~9vljbkjP#2kDco6$enR#_q*3YIPl+55iXqM!v} z>1j^lWy^wZ8}MG)Vdwh$U@16`B+*(Tn8oer54=}?tP-?LS$rP}mR6apL$#tyv3g)B zSck;RmL`tu^phH9{xiTbzcERowN#pgb+I|#BOAup?SgY0mBs$CU>UTV?YwW5Ql>Fj zOp23u*RKUg}HAn~$g z;zM>#Zg_5P99SNgCrPxH1ZH{Jc{1)=i+Y!FL|Gz}1xvAe?0KwxxkE`{x%idD%a-#| z%<{eS!hK*l@{1(VS~}R^7;{KY$KO)(KY2PrS^S;}mg7Cx80%hm7QQ0~ULf(ZrAB*p z9UDI{0(agoxJHs_Eq=^W&+t}T7-J*nJDgJ%r{{v@S;sZFc44_?FYYAnvx3CSmizYX zIeGih(`TVBRacWFT1&pUSeJ$6L%)M%&dR=*l_gcOq#Ln!vDe?H7lbi(c`b>TEfGW9 z@&5Ak&~@yDY_^Fc(ORZ4%ZAt&d%$9HruYqIiG3l|CD#=;#yXrShu>aDUMBIfWvf5i zv3l~;N9?-kbdw~}T4I^y+FzUUfaR}l5AP~VXo_IjUXacG7WLA)KwTE~C-JhSmjQd` zAG&b31&pze!$=aXrO7-TV}70Y8G)rz_kc&r;`vgrB%EYxuoDbM>;=o&{v=+uyneyD z2OD3C_zIS7!$}gY#gkbUkG~rPmMeXyJyVt}$@1eTdrpp;WYq#J#fOu4*)r@nvus=I zFcd5~79@$*lEo~CtOnHrOX(2}Q$ z04$zUNfNEaVm^+sdxtKS2g@{rf*+M7>a}3$UWM)O?qFq~4VL8MBwn^0@$f{;<$_hP zi{N#6l0<8XVwRd69?t;F(h4`eDobFhVA(&9Svo%dfF~dX)FAP)<#$K+jnCqlAO3cF z_xdD>)>6|J>r!asudiUa9OaR%EY5EPi|ewrc&{vMzOftB<=s^hFI#f5+3E8Zd*W)q z7_+%el4vc?%(8#({4}uK%r!RegnxfWPLnKQA#APBtZV!gu=o@r@v@~!dsde=F7XAx zGO;*GqP3(k%cql%@S1!$e@#PWiF+&5rRZC>SE2p%d-%RzvnYv|En$OvaK2Pd>4MM6 zd&-a`TFamXIL4-j$D2c4q87Xn=f$p)R#U zND{3jnOP1j3d(`H1l|3sfwDw@5G?nHuy3y&1CzdjJdoNxW>?`;pC;VOF*KLS4Q)kR)1* z(PFGiWQ6%xuoUTUW2!9alI7@Pwg!9c_?Kr;mk4tbFI(&`u(gZpP6Kh@MW=BjiPmDn zEZc5REDn})JzMuwmiSLXT?V#cW6WVp6rK+JVi1X!E&W!qXUl54L-3kxIg%vNTH=|d z%12M!J38yx@BYdX@ma8#&Sqn5^RJ|eP?ryHNxW<+=fuWXi#mC6O=;3+l0<81vjoT3 z<-HZ~9XX_Z%rIr~{UTW03bOC}l~+8t2A0fjBwn`c+`193$+q|M;67Ndek6(3;>#?F z@9#On7`xPZ;}~VJ&k!u%-m>}9zv;Y0usk#+@v>z}A!doY*Big_)#y!$y&g^zf3E95PL>a8>N1ovn@EI%_tw|ECC5BmwL^j_C zmVf=y${b}0$rLO>huIiA?D`n@!G7vX;$_SC&uomPHyF_tEVG7^Bw9;@WjMyHhVRDx zV_oWoEL0Y^Z-S-bOx6imzGe#UQ<~9?#LJfW86G(IU(DH`4wj$oND{5Zjain?Z1oiC zGWztQ70Qw!Sz?c{@5oNCBJh1bJDSAHmb^3A^H`u^Q5Uc{#F8XhO9rz%k95aho+#U= zuamMQWC?Ytl*8_o2Isd1z;!Hr2#J?1D_q%n@|Q2{7eE6>lO$Tp$mKZ3JU2f&0(DuL zQ`ALSBEJiky!^M*!wrh#zKih&#s0+07P~a|%s)M>H}35TFGiAREs@NU_jiptV2KKQ zv{_mFeh8MZ+-&V)WYrqjn=t1LiI*+rKU{IXj4X0=BaE?+mq-$=rP>OtOXZG!cuL^* z-+o@o;`CFnWL9^=_x(}#Jnt8ezAAM zeugMZ=r6%?B$++Kbh=avPjY!Rmc+}JIjgtfd^yp4xee5%jTK3vwS+QDPCwIa(9j(H z!Q-g1MEuM051saPBdl=d?A`uz8Fbno|4uS#I_)i19>-4mgTpRmK&O3V9?w62&wuK) zUvZsv+MB=bTNgU*`;Ol7C*41F+7B+uI!YGx*=7Nq_Wn~ys{d1`y{~278_;Qgel|&| z@3e2&wG#gCPtA1f1+!ztB9cgrsK0gE7ca#+?H~UbxC1)vC)trC|6`~9yjW9OBWAiV+AkeC?Q3o$nf}{O`@j8xL+V)0 zfKK~1yGYvq+-dLj(jGhQUtY;q13K-y1(AgRRi}MKRxKPDp&{?_?A^TwNlxumB7$8d z?5_K_hfe#yj*~S1Tc>@aK2PyX;@$63Bh(eg^S5yI?6QS*sD@SRaT~_^jtmknUp?!U zU~ho?bC1U-CzouJMEid!Wrufyz2gHM;0`_g`O$OAk|kM+>|%GdkX46qr&phMBwn_J z=V5oX7So>N*Xuc7ND{4OF|+Kn-;Aq2Rr@TvtSpH+LS4@OW;=;PPu^Gxbx9gR;$@3x zhwXS*3;z*O9`0(p#*idhOCqzJe7c}0)a5~7{~O8@WuLEJb;GKL`K;&6pKoSpkh?@K z<9uiPu{ZN>3ys6T^6nss*H~V%FIY8J49o;e&2W-LYpG_BV=OpqRerEEyj?b4Spppd z%ZXBKS3yKv*->Dr`iR8KmWTP+7^|@DLp!i^eNK{SEkl^)Zu;dOU^yO;^hjBp9R*AO z8f?ByzL8iAEDH{jc-i7snpyI0jP(G^%i|=8*5br0#=(xe!SZ5u&@*L8lPnv1vp0>` zCwt>>L0?@&;$_RG?^|%b%zSu?34&M$3=N66PdW7Pn)o43lR0 z;VO3UFA^_X?)bB(^NSS>FM*|c{#Jkf2(6`p1J31nfHUzNpcwP3kDkXas% z?ts5VaH}tgmo4t~ym5>r)hsX(EGLJOBw9;9W^w4Aog3;>wbzwwWyzK-7Vhl6?_JHf zI*c*X!6aU`XR(d>)DzNs?$SE19LcTbpfQxjb`Y-iiO-?UB4js7uEZZ0E|S zy)n2dci)!8%a+9BY=yC$!NV?4mt)IG60IeaS+34M_X;dMRt_~(mgu#D<@pVEe>r<~ zE1nkQvX;clmb}H-4sWY|7jc!L^Cpr+YkAHr&5NwTx|~g{T3lIzodrwy4px^(rxv8b z82k8=#LJe~&g?o?-!l>K`^(>xBw9-;N1XfhV?I}by4=|ErL3~Jt`jV_W^AX|#X1e} zOMLuJ5-(e-uHB7uzgp{#_!~s+_mU)9OAlr-y?dtxSn_(GuA(gIlI3S_HpU#+^x6h> zx#dsdWy{JaHurDOaKqD2w(TQHw3fxpa{KGlDPT!k>r_iw;@1mx`COY>f(rhK0?Xdb zBwn_>ujPexDRwddcY4`+lO$TperB1w*cEq1R;%5&fwDxn2o|q;Y>e3#E{ea|b-fXZ zmo0UF1)`$v=nhz6o`?JBY-~ z7CRTVBL8F6x+P#~HHswBT8gZ~xnJeit4d(`GB2^Mve<7BEbXk=b!=cl3)~@*xs1fi zmI}GpDt3tFw_#wpwTdLsT1=Vc@R3|N|e%0fIva=9gmmn}ut zvghQcy&nz)OYdnUiPkcYS*qra?F5!RC8qaOmY9t~T}oA8&oE|@xh}#O3#~}vWy>NX z_Ky7Yj0JXe)~Z30Xf1)vGPd_;JT2%)x8?(sC1jIe8TW-*@;>P636=-_NxW=XI+VR5 z_n6ukPo+6IoFvg&?lVhj%I;E7m({nw4^tL5H^Fi?m34PK+&T#F`(GcCc-gYGydPf2 z7KCQuiHD)jNfNEazzOI6qH%|JfW@lBtue}yAz7yOVLR_9!dadv-k z4szNB^X0QKNuss1W0v-9AO8kRm!jSil_g=bP?t9WZ0@f(=;IG#taUjOFI!smVU}Y4 zcky{FqAE$EwOBLDWml(9VCmL$!gOVcbQdgMR;(@;`aW9@7WWP$UbfihWAmj<>b)OO zmzmv260Kz?vpBX|G9D~<2bkC>i=T&J=^V#)NCX9%b^^=x5hPx=l$yB{=gXne+i_Px z9ZQl#Yl&l)rHAg~?~Qky^m(DOIC%;d_vdVo{ zNwgL(X6dzj(J-)_2rK5UES}p0%l+%tV<+` z)^d$m5+`=S(;Tv2BzP%HmSp*|m|e##hqO~+y zgV(XH@5_q&-# zPfM^&aUC6_EMYqYOXUcA+_DgTV#furwA@PKWs4i{gj{^#OK-3w?;uIEmeI`etMawB zU^!RV_?EJG?G!9ApV(;*9=Ug8m)hAfBwn`Ud}eDGw`z95--cRQg(T5hT$!be%SwFa z?|uKxJ!Q$3EYDW6wLTlKwI!i0!;?t7Y#GvgFJ6;x{;C@bb;)@}l4vbwnZ?mF%^K>m zx=UE1vLx>k>f-p2J&zd|nSBf_1^SYB+0tn@yYHtQx_~?Ie20)ET1z^!n3Qa936>Ea zOJ68Uw4Y$v9?Rx_?o{I3YpWV0mRD`NNH*a0M7wrope&oEKVpPPZDwmC_nwX9*5gj`nmj%<`C`J1w& zOP1FW?3+dLpRa~MU5fri;$_ROhyWa8TY9^=gXL%ml0<8XV3yqNO6-EVv~dXhtt{~Y zLS5{f*?h6d$-omr1J{#y*F@*qjHmbc6jKOxU{uq-Y-C;y~>e@Bi8 z6f86EvSfQ^t)WU{U6ZX=T*1@l<0g?w_FvM?3PnKUQuPiB&qXEud7MCY;k+b?k|0fJ8XwBwtWLhqO}}lmQ(Bdx&WWy|F^2?-V~b)wkb zo`xQs9Kmudj>OBBji1?d%xhyR{*Gds1d>E+DYG8u{-U3G@sy2eX_Xo&i`zcIa=j$G zR|YH$!0-EAzLI#^QYw(WBiH{cV+GVDJ)0!aT6!_d^gT%?Fvh}OrngX*49T+Fhpk(- z$o1wS)Mdj55-(dkT$m+n#s@q#V00!)qO~k#mNA##;BP_4+lRMTmW2I6U2>0O9o-q( z%ld=Gcs+@iEjg9hd*!kvck)189&IK`w3ZNN`EkbZ5!9vhFb6Ybi98@!x}0UFYFaGUbdu!vFq5o=ku{c)$tHXqO}xr!MXpn!h`-$mmAfK4^kGVgMy`OQC63(VcqfM z*5CC=ylg4U*Ty^M+FcszvbGsXqP3VY%ZXbBv9l{J@UgkFq)L{32ibdN|ASozLR}sm zB=NH4{ZY0yey)F|>0mi_oFvg&Y?L{- zM&f0QX)9*AH8OHOSj^s#Bw9-lvuwLpG7T)*ZKqFGme9k3Wx-7LUK!;wI}qwpv@401 zEjQ=0eM&R8B;XiJ?n9DjEeXuhZRhDwuxwn^YNoPyh6{g$`W-{uxu>NzKJ`n4LJao0`*C}YNPalh!Cp;4#56Ja!qbC!fnRD+erdE0TEG^5p~jzW*fI z8+#KP)*wl=mOIQ+xY5KUuy{SLyFpo;!vss7+idNkSdlPOuslj3@vAsHB@G9`>&ikv-oj%c@WYTn}8y5}6?)2*| zPrirl^p*QbmjBe9Uc`s4X+ph|Aryoxy$^PD*erdEX_LOgPtndv6W~)?^;2*lvyRY4enJOOYnIF2-$A2K1 z{@d>Kzx{zeC(U<5ce-6BN&BC>(^t)1kKO5EcWNW%ZBJKeHWAFS!wCWo<0 z&LZEaKhs-xm9URsSBWjv-@8C}`tH9-n*Xgkeaelh$KX!z;%CW&>WbraLb!Tvddcnt z78!oH^1D7?sXy`Z)w54Gw(^_OqAae;wKXD1bpMxJvF!iyX4G-qecSQ=!(+;lEm<}W zXHT@t#-!oy+k;6YUbZ|c!`?J}Z{%wUceN$2NfNE)H?#cJ;hP=QrR4pcr<5f*La0k? zGq#fT)WZyabGS+piI*)^z8%23TCYk;xMTIwE0RQOX|fTYoG!(U#GUuEHcdOPEYT+g z%b|*_+t&MBt1fUoUF=2TWy_H;_T*#`@BllA^7xS?TFXRc>D=NBc6ZEf+A&61f=>yS zo154w_L`(?eZk_`j>OBBs4;Ae{qEB;8^%~EGm=DW@njZD!_IhSWc#7HZz+rGX~AM{ z!_Fiwv?f3PrpiDI5-(e(JYa9Lh1Z|%4wl`MND{5(GP86|kG%->^3DmMmtmZWy`|EIr@2Cn`&Pq)-?4{_L&k zZ1#8j>XY!9#LJfTZP;pQV6#v>n{d?+l0<8%w+ZL|g{X-0U`ena`9fJD&Ip#6C+z;R zef{!wFkkwuC-Jf+-JiW_6sl1b-!yVIlO$S;1+xV0i`xR04iAmrDvR$~!D2pwt(N)@ zEtn51^`4V>+0wKG8)NmH?eJ{%jc-U2tz{##gay>W-BD%hy-8OV`*VWDaysvZn$X+~ zET+v#ylg4Co2{65jL)hFb3a#ml0<8XVwO|>hU=g%$KD_MrYtFvCH*`5dcAVTg}Y#x z_>IKN7IWuAc(3&M*ZH|%DU?H!Xf2GlJ#U2K4Pwz@Nz0x znl2^rvSrUwHpcn}ZNRhDT^&dgt)-?L&iyH!Ph@~)-;+-HC;xl5N62}>(xWxIS6a+5 zi2}=|S0rAxd@8>W>oUZv*IuYg(~l&H)?&^q$*VjbfMv{$;zr8ic0sUw@m`E|Iq{`G z?!c>apTx_SqXumE?YvsWmVssVQ<6k$ab}jozSVK}?Z@khC6pyYvTUx$b`=<|If3{6 zXYnLnwtU*euE}9pzwr0Riaa4nw3buM^8Vxi{7M@2I-tC=BwQ5gQtk!2@B6p?)dt4c zv2+qITeiGs_sWaI%)UWg#%7TuT1y(UIhl0<76#4OwF7A*qH>cv?NmBs0@VCj(Lh;#qo;aoVzjyjNd z*>bEg`%+T*hGAZ)i}8ArL~C(mmKX~w{DO6INK^}DNtGaqjn-Tn&G#v-o-vFI##fvyKSk zg)gST7z^D@l4vclG|4h2gVp71>2KLkm+qfPylffJg?(9ERWu&2$=$w_ zBwEWtW;t@ZQ(>@di{8ChS>kRAb(z%nT>2fb%4JPrj#c(*Q%h2TF_%2pz6iK4B7BBwn@z{Oh2nK7Sb&GAuZM(gf=X^X6yd zjlK-?rTBc3L~F5OmQ&`HdVwYGwu_IlMBfoChWFXCWsP37MuVlsV-hc0sw`n+tWfI( z_!|)+&q)%kWjC`-9&KD5ER8Qt@mH4MyMkq7Yu4SdICozBJ*B*{Bwn^OFlNvE)stEp zL0xvllO$S8JhM31mER1OD1%19%HkR?SbS~Se3`yKe{HDCr{W}Dwsh&qo|BLEevIG6 zHkBtyw3Y%}aPCKz*oBrFy}uq(mUPKdwI%yr*{!R|Q?LXMA@Q;$t{s~%P19}1LS1@} zCP}oGHp~*b@-eQ#-hFufxU$6G6Y5g;8e8Y@@zfz6EXSXdc-i9Pz@9DJ*Y9Tm7Mr&u ziPkcMS(eslhU@$pL)JwqOT>M_lAgitm2HL|s|1#E79?J_yo+Qz@B*I{SPYg8lSmS+ z#g|zgp0n_V`O?yG;6-KeeIQuUve;VR;MX6rz|y89iI*)w7g$|Bom%!2Ebl6iBwEW& zW@%cc8UEId-I|Kml*K+luoPRz>N2;)N(Zn6ZzA!s#itj0hN-+HYdcuBdXXerORlXr z_cxUrXbW|j`YAb1SyCj+&hqS;-*@yfZ?Kg4N#bS8lhbUiFY!*zQDC{9=h>e>LThQp zEbWW;YXg=dje;I1OUy%|E;UZEJtQ}tna%)9qt+x|wiF!7x(7`YpPGPWV<(bCYnjX} zbxg+Nj#VeexlffPu(M zjIr|jiliw^hGeN#l-*yZ_qkjFECr8}c-b<^f_18_j1Dsf%ej*ziPrLiS#sAbJ_jt{ zUM752mV`v1E^l|TuA9AYRy+gC=?^4cwp?D%?)ybYnCAn_oJ^8LYiY3U|8aKbVLf%- z)6QQA&-zny#YufwJao-HJ1;hp}H9MUS4DQf8V`&W-^v?J94(CWo`%j z+k={eKs;{=(;(;is_N$)N0!22Ac@rCOe__)c819E`FHoa!eW!fSY{0*XBA>@X7)pM z$ye&}C!V*YJRMP)yE7Mpy1t`@RX zYk+v((z|UCey%LFXo0^BzE%e$ky^@$<^1vUH<9J$#;E4Pl9l9v`3b( z9Uz{!guW+V*GlQQEeegX@TrYnXQrS34t% z^EeRCTc!*ldwqGkJ{2QN|7jqJ)G~)yO!LC=Gjhn4@!G=Tn#))s^~e|tXwoGPS=lYUAH|-)Um7FCyomkey+wVjc!|x!Tx7c?e z_m|%lS!YpQa%(jE^G8T6k;I~06@$-O55HE_Ls-(^GIcqsMBcqt>9~;s_X)Cf6cLqtMmTAN?<)AIT@4vZZJVIFP@)^sWujHEiKrz@D)ullIi03UK z4ahawI@xa`vaAdPNu(AVV#)D5^9a=?`a|9E!cxRpyxNfazJ{HJ4zheG1@XM4);&^} z;R^nHkmYa{NFue|AeK=%AJUO!-_Ivgg(c}dQH2BvQ+6Vo6jxiGSf>N{!EJg(bU?sY{GA*-!Qv(+j^#np6+O^OnOC$Qs73JREmH zk8J{yNG*ZHQaR`W-cwq8GSpI7;yyB#m-%Gw8}68n`;vN|2l2c`FP2!mEeFPBrMV8$U zK@zE@`58R-w^~1Hk1R(+X6zLfhhoO!HHDnFENNo%0a=2hKs;~pHX(E0y*l*(vUI-( zl1MG1h-IFlqc^gsP0_FumJ-gA|AJh{EWXvmdwsccK|F8SUO=Arouc!fAj|b7Ac@qn zjaaUlbh?Ht3-{$X2}|-PrY;%}$b1Q%I3f(y<>X-y&s+Qx$d~mR8vMl1l}0WgiPUn1 zSo)Qp+<+|UDOZmROZaETqI{HGlV>jNf#1Ij$^!Ac2T7zB6?Z)M z1KXvlqlRYJ+l?*=%hmr_{-)EuB4-u8q8%$Mx!4`yxN!Bul#Nxq&+EEh|8h|FP5ljZMjt`gR&u>ggGGLSx<79>nuk&$CZS2hkA6Q;X66%iW_OiR}NShB5hn*?e-sH)NT& z^P0b~Y^}Xprw6&K4bxA@r`1Q*2T7zBFJkH7(+i)%xz^ht zQds=T7)wF`dE;B5rG`(gI1K{vyyfXna-yhL<~Q7>wrMy>BDK6E7UzJCebE?GZ%{i{ zSe(ll%e-@Bbw230i3h5St15`+EsjG;_u#p$Ynr3^qNxdzNG(cd@oFQtzH={R(OLQW zp|F&5mU~6yjqkHrHF2lPmJJ}Dx7;WtW9(c?82(*_$=g5@sbwIs+^Y8uzrlLb=X{c| zq*X9=d8JNHu8i1u@Fl9t@WCLSw;Y^9b{QrPo#ckV1xciqBx2b-zJ(&P=ydy$CoDFV zj73G8ba&|1_wh!S3Hl(Ow|Ms@^JQb}n9pd8Jsu2_NG-KI@Z7&KxV8hb6n=>NC@ck> zI$`-N`)oyJ5!fC zi^~im`iPHtoZL~57!@z zWk(p9FJ?Phl_1Of01(ex#yllECZ01lM<7eHFpxxQNg$R%earA~M6~FXuP7{)oWsf{~_qxyp+QcE?lG|aNVze94dPDDdtNv~$=vTZQA zzYOcw2zS~KZ3N*^RY@#qKv_F`_5%dd-@v zjVvCEK|F8SMJ?gmRPp^~of$|XwZsui^JbwQ*fKbyv#=C#mMi%c~#|&s(~;B`5C(@AW#5Ea^8v5~-!LC!YIChit|ni}~g?`oa>jm$6I= zBF`_8w@(Hj%Z1$_p0|8&LGCZp8jrn*ENTZp5~;<6SR(o^z+Vac^`-A%VR75XSa$v< z>tv7b3XPEEW)+C%Ei;#rHNX1ZzNW~sy;i?Je}vTHKrDTJ-@tE5=2`t1E-clY#o;)y z{JwGwTU>2HJa6gUm8|(&J-BlOSvop`BvQ+5Vp*rxd@LGc{ny5f6PE1#OkGZ9?#1&( zeYyS}RF?*uK|F66(~aCKJI+z8fh?D;KoY5?gjgbPhvNJGM&H9zg(c30v3xZnr`D!L z_1cOo{enR}Z^`~m_QzUp*=LI^TO&XcsYUxdp8HRabeVwaV(?snoY z-o(GDvhp*C=Pgf!G?_d<>FEE#buCqWJ{x!hV@W$r#+bhKo-$-H z&H(Ye<$fHw?`s_ydJtLmy#q<4mev>W+>f}QgTLZ*xzXqy!s332u^c-`#@JVPGu&FHvTQOyH`LGsl}RD5|Z`=Aj|Pa$p?ickF)gaNajn|>}-6mv}gh1 zdCSjOvZrKVdlLU<*XK4MiPREIEMrHntw5IYbay9Vi9gKLrQTce&f>>!v$?1)W$!>d zZ+T!s=F3)x?Z#+~c@%>rQcE7OD787|f-DLTmL3-tKL^I5;7Q&ktv&Pxf4#AO5{Tz5 z1Ki1%j=$gR;DIa=sUV5eqIMC_{UIAV!#XNFufP5sQwl7oIPneVhY@CGrSkG1DUV%IG$Ga7VYfA&8e)I+MDLGwzF@ zkynlYNu-ucVtMGl^aZNRx1f1p!s6-7SQ=-LYw{umU;Mk)6K{cd-qIwRynEe#@tG5< zOF|4tBDFO3!gGJo$p}{&odt2iQovbKYmvS2FB>McLl!$*5YJn*Bgq%3COfsX;6T-2Ife7hSj?u8=StP$J9zKnxjBgEEi0AD^Gp9<-MS&m_RS!P z)RIOlPPPt5kR?@h$SYxSb!9AX`;)b$t(yiuLz1fn;(1H06J(v-yKTh{WJ&1=l1MGe zm+;(gaLW*nu??N;WeZCsXL%S+=6*eoR`|={&b>f9Z&^N#tdlkRy>vmABpJ6oYN-zmT}|u2cx;ar~!!QE#70ub7hS4j1kD<&>SR@TGkPZ zlJ;Btj_+X|&thSTKEYT)5~;$*2R}E5^VY}<%cj9I3KoY4%;WD24YH{&Z z$m08|vgXSFzDF+NEVm|-d*znckGKoEEEmM{mYEI68fKd3)-|Xu{fa>nsYRbyj_8dF zLY6M?W0Zs?=@e6!A7jWnzGa%Z6=?3aF9GqqW#V|U*O#y=7Qf@W{S73MTFi*0Tg$KS zk)?k9Lyd$b-Hh0xZ{JYi_%|JYFnR=hBliiAEbwzba&;UuKmea)Iw5KQj z?eu=iQ`LmU?F?htJ)NwR50uY(jq38HDTwDSL+ID%(~WEXL>65QkVI-pBo@Cz<@oC- zqc>}43QIL-IX{7{#};qMn1U?*c7u4{^4*K<^Y=Kr`4qAg*@7fe%P(S4YiRC@##qws zw>rX-?atJtRsregb~V3w5?L(lKs;}0QJcI=YW%|of0;tn86=TfdV1sQSev+g_<6r5 zI!sqs;?6Rbm;Piu7VH&z6j@$A1o6COZ|jPI4@_TvT! zi^Dm_GHo3By4L)oT5XV}%mu{rmRYuBZ8`kH+o8y^{tQSWwcIC`=_-!ps4i9w8xIkd z63)`<9NC9adE7PvS@x@gc-~U$7}@Kyue}N1Uq-h9Nu-uaVwt>mH-7&Tm-uR=uq1mj zb(!>*T*tIx6Uvb#I|an^mQ{A-dEb1>Ec}H~&kT@6YSHznuHu{o-X>R}MiTdow7`^)&o1G^)Oz6(eqwKx%rTOXBe$ntPg4-;XrzQ|an zm5}?(#H)7r>}csO5YJomH<0U?LB`x`$dYOcl1MGl#8TBPvJ_cHs8%c%mORd)bDNG!x07){Skk)~Oi_a-Nxjd=} zS(J=GJZ~un%55rOQa8DaZ7Q-7G=Aqcz08`J^$35zWXlePM=fecMNr>$EtvIf7hM< z^$Dq3(>HbtQFpqhDoFKz>P}zv_~!-Got~=+Qp&s2ql>ig|9$;&^E#lJ-o67!Bu3O< zy3?ECvz|K$&?Y#W(_y3jhyc;4E){l3j+~;|AiD(;W{G&s&1_le=2#z1nS& zrT=A+L~02kmO$T*__v_12buT^OX5|gE=DWJ_0cQ;HU7nvO)f325lXXhSMpT#D zTR{@3WgM}zc{dRMUf;|f(NV%u$yu(SBXd7V)4vn4+#CSndCR?3bjM_}Azsm%8G|HJ zixshac`yn8M)%HJ_Hn|J?$6XE->Y1M$4&Z4lYXGG5uF0M(^kB1j^& z_!G;!4(5euj5S&`^^ved2QZfAU&xdBtp%fRBg9XE-YSwjAf)6*~NY{MtuRYtZfJ4d5d#>GREq(c#lu3mv#n8q?V?B zc<$G0tvDam<*;3Dny}afF_v5%@@DbQj)$L-<>65f&s%~z6N~?mZ(UGbmYxDhq?QrH z5;4qT60)>&@y!yJBF=K$knEPu%{`NgEJNKvJa1WVNjkd<+c}O!mTMP55~*biv8?*i z=qIvRMy<~mmZV^&E=7uDzMMO<0C(MJ-v{x$rM?$=>yvO;qcgIkJO)Xm79V2yH9yb_ zS;B@H77I(rb;hD*N_Kw+B`)uQ>XJGh#Pb%H8e~1z@#61}$TDR*NFueoCYGt+mHm&td*2uD$Ojz+@x0~I4)S(kz>)R%w{DbPsBTLA@nYzLf z9>G{nIgss`tkStnkY(iv5YJoYWRfw~qwpC%eXBeHB#~O&iN&?1O(wD=H`M4QEbg}$ zi%&YazvxcZ#J@pg+ycb&mV>8=rB1)+_$_IJwjhbr@`PC4d75277Kgn#{e{IklCdb& zAZyEGVU|l!T`cTDJZ~wlOYZw`dWYb--}@LyBDK^Az;j=zdHG#rnb`cQk+9@(7PseQ z?&~EQ8zIXeO%TspzF3p51{6ha!(BJ8+JhugOK)On)$;v5WO?YcVWhCcM=^C7P(a35 zaC8>#w*BA+;(1HGePr$%jT|}yU6apyf+SMQGGb|JG!g%%eLzy*3Buxco3Us_lQqmc zEteQn7ynlvp0_M)Mt&pWccf+svTVr#Nu-t&#FF!&3V)CMR_BJ(g~j;}V^KIt>hf&f z0eot$Yi|(GTbA0BG3Fls3-9%PHULSamUv>B^WLNn8e?kqFXjqMIcM2DpX@0O344y$ zV=Y`jJa6fDnsnXtdbM&Ys!P>rkVI;!B9`ArcIzOEs@Cbn!jcxv)TPEzYrNN2w(V99 z8e_jYf_UCip-lFaoSV-Xjx1^2K@zD&FA!hHHVtf#PkC-rS-w(OBJVO5r<3Fv`Pf@s z{F}p}>L8xCe4Rw*{*os8cpqkQ8<0e5Sx7A3L-fs1UCI>n)(MMe3}gA0PC8Y(S>@QE zx+toEc;521C)vBO8d7!_)x}g5B#~O2iRJMRzjw$o{&DFhVX=v2EHkRedhAs5tc}QW z{w0X#EwwI^6W#@{KYT@&{+S?&)DlB1diPTzk)_V+sGY)6z*(kEBXzmbaz{S0Y_tUN zyyamGc@{fd>mWYuv3Dm(BDItg%gvBgCy>S4#b%$dB*rmy`FMp`LW?UdAxniji03VL z9}~+_hnEdeT@HJJBvMPqAUyYX&ApKCW6bozZZBbpj%O@)Ym;|LT|2r2Aj^C|5YJo2o+I!0UYhor zh%8^PgCtT*+h9ER<1&sKB1_M)qpk{z*F(lKSc}a4;eB%O7;7~R#PgQY8ssGLgybjt zkY%w6NFudNBbJ7Gqw#w&rNoB8!eW=eSgbyiG4|Zt<2b6zhhz}XTe{sLb*WkCVnQqtnoBvhBl^FcgsQI9A0mrH4lUn0xaWgv;v z(&{>%`@dXHgd$5f?VgW?#qBX;S?5lkkvC+#!C!N*?h4|0i(`9o9wu{V_b6mh>jRQV zEfa|4no2zWI&kZ-uP=n9nzN+&k}rhzG`xWK#@(YqJa2g%PWF^Gj?_PaEavebiPW;2 zSnSuBEh(Hf1hd(%M@sik=c zp8FqLSmL?wlB4rkSR9@)mas$QI`&pie=@4ek3ta7TZVlho%U0nGYm+l1bDNo7MV0`?&wt{1%dh&RF6~@`;*mwK5lAAnY$KM; zJB~)k^20T>T3C{wGj&`a@7sYHJge+gqf_UCC?FHGp_}MrHpHU3D1d>QCSBPbR z_JLc-QZe4D_NxEBM-G3%ST2{7cduqQ+*Tt?%b6gaxAgx(>hkRP_A#g~YZrhdQcF6q zB&{kaLzcw}W0Zx(J(;l-L~q1(S$FA9fZX%m9ea7K`dEYy>Fp0R>K#>^OgudvNyijk^H*J+CY#*YI#X4^QIYL%L!#e zJz;T9V=Vq-$#dnbsBnBv$+$U)=Ph@?lFqJXYMZRk7;|Y2l1ME|H}Kp)S@fa`e; zihYEooU`1IA-|ojxTZV)iqnGCAfC5WSdeFAZPN*dkfpZ;NFuciB$gqr(|nL+x5iTg zVM%+%)a7SqvWA(yZtQI|#^T$9c-|6fLY~F0j2&lJogjsfq34sd=Gg>4po2M6j^FN0ZF754`Q(o*p-Ira=m!rXkqbu z!&r*fkacqXW%CkHUCw_6@w{b8XVTT#Hs;|hWO@AyB#~N@h~?|0Gx%3Dnw4o!5*C|u z#)72NdoLr4rh?9&KSFA$6^7^j2m@t5RF_vSA7%(k z0cUZJBF|zWzbaoKORcjYp0|v0By0XhV@>uW%T6zlL~7|vEWKQ8@jDAWkKlR2l9<8N z#b7zP@9#0{fqQwPyg@u~nRGl5KO@gv`92I;yaPZIsbvMRj_PtORwYPxw20^ zw_V6`q#uaqEnmBnx(x4|vk+MZ4hBi2mSe93WpL|VeGdzZ-CM?TNROQDIg{J> z8nVnh0OEN|<6q>tvgZ79+(X;Y2_%tP;)o@{t;KQF(44b<@lj#%_>bjpI_=HYlHJ&; zIqjFDPW!I5Ad{ri{>Ei{+-bje?2|~;X}?x^$)CUHpE~Ui_8EewKxN%nchqU0*Bqq# zyH5K%F64CCAf=)ssMCJ5CP?*v>a<^aWsC*tv|rQ_q?C8sTkUUynfr#1z~7%z>j4sp z5%rf&d%GKjI7z=t#l@)8{`~-ua?FZ8D#pmo%VnE1M{}j&qAH{3Uff(f9|yJJ?k>=wAXci zaSe6amo5Ye|5c}b_hKFA!T`QJM2YgV^f zj_zvTwnd&0SDYg5>KXi&>|($AkU1Y+J$3ehc>e0Sp*FdCdev=%f8*}`A&^A+|MHdm zzl7^FHAa@u!!~DyB`J@&6PW!ZrwKKeY0N~HQdbbqTZ+b#-?VSMab7v9ahf|wBDHkB ziB}tqx-4FYEK_ff^AeVjcZ?;;gIF%TNen`k#ScI{Z+Y)XRprzauHvnL>l)M7#` z#vNMXCqV6(CRc^UEuXQtFCb5VrQ5G2BFpecAfC5eK1<$`4&MDz8CjB_gCtUm1F^KY zH>NeJOYLE=gM_7;vz&+}r*Gfqx#3ly#&{6VTapdPQ_D zw0$k_2uoZ6V;R_&y!A;ed$R>uS_gx8-ZK6=8Dpo1+v0Dny}JpLNG;kCc<#^W+Mp$} zR9-T=FDyPE7|V*Smg$iv@J^QQrIz=RMdcPqBDE9}OV1g{3s7Co3|^EY zEa63rrTGK0EDH&_SB=Kl;87r+xA@E<*W}!y(mu!%GZ7?_T3X-2bAO(B-Py>}%e6y+ zu(%g9ma1HG9Xqh47yeGX@>vkiTWsEwx7V+Wb&HUt#0w;mS|$-o*x1T;$TDM4*(YJK z{=`^LwIpvBbar3IJHN+AfOy`b@}8{6?$|89iY!|vfFx3jHLT7rq?_g6D~n(%Pe-fCfq|IF0o z>O-=&JQTkRf0br&0f^@jQM7SD3V;`Whr zK_5vThoART3PC(?S>KA>E9>06R*c5jv{H~nYH1vW=YGh{I}ymzvWB{zu-H^EmL0mA z@O8}kjM+j|m*(|AJa4JWJdW36FT#i6laZ$ygCtVRFk*>beq|K0tdGd6Ti=D_aBNoK-miNWvI(FdquTtz<0oMv(h{)U!VL z?RAfzAfC4@>qyr8R~!0=BTKl#;6Hza)S`SF&;6k188wily;b8$!cxguR=*=(q{@tl zoQmqQ+7`s~mfc&){iV0J$~9#9>;RHTEr!IRTxU)#WGSrkYKE|+e`o6QC5QY*MDDKf zdy(aX3W(<|FU!d_IpfB>O~`Us4J46T))7m~LzA8(%fV=mdBPH1#aNE4CHKnrCPVQ$ z+3z8U=Pf%I63g@5$@pG5>={TRwRjTC!;c?&B8##5a#LaP`oUOk`I6_C8k-v8eanGu zK|F6+tV+HX^rdFbMO2qtoj?+)!p0S^CcgGMd5YJnN z_>g<$S^GF=G{!PIf+SMQX<~8t`EVkdFDq|OwhWm= z*ZeQnoWWmm88H>a^Ong4WX*3?s~7${u=gC0L~7|7jjv-}o&|&;%X5qW^wIoxA0}=u zWAUyg&&aQ4O)Erod1VFSdCT1CiPU0BEZ?&h<9Fg3DZ9=Ii_bpB zGGi)P!$iy;wHR6YWP*6!VsMUh56%fJ*o-XZ`5=kZ;z}$x3Jt%Zx>V+lzAP*b`x#6A z3vx{!H2vy9WQqR>;(5!Eab%x=-NelokmXVtNFue|Cl()r{+7t1tI_b9u#|9?Z?WWF zIqp*mzK)G*58`=CsXsZZu(K)*pCta!6(o^bDv9M&wciwES^MF|bzw=iVd^rt?K(X7 z7dC8Ti|R7A6vXqE>!--Q^3G6`N>rDcRUnDfqI(z5{jBDn@Z8UTayCL(!fhE#Whe6d zV!S#=2U$$gKs;~p*CpK@GYf{}uQ-)wfh1DPd}5gpbLR-EOZ$3D?+T0i0mgE5FS(Aj zIlrJ4vS>5~@w}xpklbIi9(%>1x(sOrl1MF1#Imi974Gb^dDSIeSga2+7SB^;4RhtZ zCjQFC$6OH4TjFn$`LaOM5AVZ77lI^GOEj@)#(c+Xn9I5qPlP3pv$QrR?~*o;Iv$Ve zvZpVI=Pkzuk@GO8g6x)}x+ocfBvMN$v3zRR9iLboXmTq>SmN!Nx;VZeCyC=W8uvyP zgQXyzw8wMemhS?{493;I!Ge5*b&S3n`hf1%a)Bgg~C$KSso~n_sI85mwO@0u^k|u zw|v}5=KiQafwVUSv7D>_dsLq#b7JV%B*LzE|G08d4Wo#;gPJ zyu~~82;LjNtkzToS#mdnBvMNeu}o-r7Vq`F_P_p3SRx%5%l$MmU$%cZg-_ns3J3AL z#cdi{ColZ1i1%P!?|>vyi&iY2`{j*H@IFlAxjTLdi>D)J`A(iIzr1TY9Mz@1DTwDS zne@v?_cJVZp}Oo}1(HZDQ;22Y6#ao{jNLmnq|TcEKEK#FF_y7U$-CDS%U2zdrM)|d z=Pfbw$r|Q|>Ht4vNw^4-NG}W#tCKQovch1d(?Z^OX*pAWP#w5YJnd zP9bCLO7}W?$dVogl1MEf#F7;88h>HuUGnp$!jgD|sf+#zvgTi8;MNOSLNh=-Z~3*8 zys!MQe4Q$?G{^@@q?UYQF*{IzzZkM4)~%(m1UfSoqpz;`I(Glp>P5)%I}yb5mL4O? zKFo{#k8xezC4(eVON%%>_g`ntZiB|yyv3$%g~jzKW9jZd_F*zVO^-#EJ*z-GZ^<+w z`(tMo1>>`DGd6-GQp-4EvH3P0f3e-Vfk7u>spKp%O~`%UPhp)KvTRia@w}yb3$mY_ z)ZZF+Lp9U{Nu(AlVp%w$7=JaOr^>Hx!jgWBsf$}Davi(iqSFsqMyP{$-qJFHSZbe` z(H_m0KrN6&YVjwQ)P+ZeqA`|`6W>=@qFoqE;T*C~wt4UofBj_gM-b0j6k8HYyFNWv zAWO3{kVI<9CYDUI3;3%v!CxH*35%C2W2tq7T$3kGxP#YYhkt{3-Vz#1&M76Is`DIK zp4D~w^G8T6P4D5kKf7_BJF3goZPSMdi`{X?vUUTx@2`E*H5!ev;JqN8w`?#Z>txHW zZCW79%|jrG)G~rtx?gXy7Fm2RYK|3_BF=KT51B8eR^`6PGTINs^Op9(EJI-|Oj1%P4K^aavs|I}zJ?z8Zz`TI{iZ)ssf-oKcuWe!Dk zSyT@sky5JOd8$l7Ja3sXmz*CvZ#_s8S@Kqc zBvMNsEtmx2c}wATvW5xp_%;w(s%L>D zQj6jPJol&n@PCY&fOVSHIV3Ep|FQf{ce>qu(w+WfbHfPKoxZpo$Rz1be|5$Vcc*u{ zdT%o7PPgm^viwtbdi6BYovv$lZ9VEvAJGq_`@8P+xuIm2rn}*u&!{_H-w>qwKXs>< zj4m`s-Ra{-fRys?bY*+eo$fI*_#(Oj7mWvr#EAM!clwHlq&t0_W@vxZoxWu~Nb)~+ zr^j|my@|ThPj3au{@R^BGvYPwDW4S5I|U8QW_v(_zv)g7v&z7k4kte7gSyk#9RQjB zZFl-#{=j;_d*4Le=^vdy+JEj&S2`Xrd|Z9yddhVzmo8bm#KJOS#h?V#o!;UkNcgY1 z)2;5D#>w}7xC^>+CWz;+o@0DT*UjbXkt5O7bK63YME3tuV*vSonY+}q5Lt!= zo;WHj$!D26fw@1qt2NE;(FD!gUpGNKZwbsL-&f8a7WoTVobQ4pQp;*$3ElR*1X(U_ zndc@f;U0{oZ#8+dIMPZVcd8uS1LAp0{efinr^+z&JF*xb1WBY8cVY>9Z z5`UhlOW~az_&WByY#;s|l4Cj`p0`|5Cp+?4toSGP~`;mIG_=s-iJg_c@5?Ee)rW)p>_EDtXBA{S`yh$Xg%c;`52f|X$ zSsvz*HEY!78hcP(uKfVzU`+^Xe`ET`*(c;0d#o2-+o=S3VsmIuv15~(GUSOz4$$7}vfznk@iCH*Q>m!7-G z+v}Qh9*;woO`#y3x9q=3=Kd9xi1Em>GYTY;T0RlWtNe^AG+$y#&`FpTVC zKN~w|0J6Nz2JyUQVG5c1LmR)eN0zt(kVI-}`v}kdTA!}$MwW*g$F~#~uWO9O<|6r> zE3-Wn5y*0SCy3`Q#ZSq4td_#XVaU>VKS(0AOe2=>3BB@>CDpEJTVb*DXDr>Hkh*Ls z+xHz=%8r6~-r}1}-uPbKysZ#f98ZBHQi~0-*coPiK$aiJ(>e-E5od|0L*~ovE2Vhv z;=2ck=PgZ|kWR?ui@PKv%bUv}iPUm~Sjs-M!(R>PyYGBAVMz*L>T-G?S@W;3Jk9$=*evN)!BLiiOQV zJa0)FO`gTHpE*87mj10l5~*c3u{6{VxJSb3|Yp+yj?0RC7fm8Sn@2^a=~ES)wwni#PgO8 zeaLSRxsDB9jx1F#KoY5C8?h+3B;c<&&1~wwT3C{AFm<_XOip?oS-K}5jjjY@zzZng(W;w^cu ze9-fTEgECx`_m@w9i1%Q>RMg%hEY>#}OW{{?O}^22eKS;--oHUSZ%NG} z&tgq;qVOr46h(zUe}vRxOf0k88;wC@Of})5t+3>Ameyy<+`qK#lRdH+ya4gM#iBJi z9rdYbTP~_g@N1AnYOx>|yO}=%kmdBkBM!n6AHmdR?Qil$UA1rX@fWG=R)TomQW-_& zzJBU{BV-w~9wd=kyokji)D(9Pdq&T35f;B&jOFcF@?Pv$jn!jOUCtzcc;3>?lC1eh zudau`*xvd%NFueoB$g&V!MH=!Ok3}?usBCD7LUW^>+{7`{f{Hd;I1H^w`|`@-tk== z=GX|;<$iCFL~2obis$~;^^sf97)$Ue@f4PF&eFDMD_+B_ZL%A`f9Vno;(3ds1zE#% z*RjRVVs~zWBvQ*jVsTbK-2z!ueQtRROIj3Dm+Q^QGxENAHy@$8m<54&-s0VgT$2a2 z_ilpfqIVM{ky_RgOM`Yss>sqm%i3R9B5yO60oLTKLg}aAp2)JX0L1eaql4t+eV$s} z7G&vM0+L8A9>h}GEDOI&+HqxMsIYk6VJy$OlWVfO?t1)Ox%N7U=PgSV$$s+18P52t zG|3SliPVxrEI%e_HbZq;_oh*#u-HU17VT4HjJXc$hsW6FJP^-YKG!5;Y?a;JA*e1n zA3+kSrB)K2`=KE%@xJBOJIOJ^Qovb^KazczGwaQ_p)ocj9>nvOZSTo?EXCp=-nX2c z1d>QCeTijRcDo!@mpU0|9tun1U8XL&i%DH-)m?c7S=!};c-~TclFXOF#2@%KyTXb= z5~*bcvGj}HuYoM34VEPdOJEFR(bHXougP9kOWl#BL>0vImTv~+eWh~O0=(uo*91wV z7B^zq-`HUlvUI4cmntl-v5dugH(A5X58k~QS!^1Ec;52Xf;?B=HMGMW-8EZ)BvMNP zvGlt2(*s#{4f&EGER~$)%M`MP>C#Y54OwExfOy`Lf1JF3nd9H0Hk$kWr+_3E)zTs4h?LANeRO(f1gOe=d1nxkRrDe=+2D5{Tz5zt#~;Kg-6rBjQvVNFuc?CYByg z6SpBt?d!8jg~jVWV~Ki2p7+O?e>6m6%r_Fm^Oh|~$QV0*WjMZ9wu%Euq?TjEvTNtb zY-9=cZ1Y`M>>eXCtx^VAfC6h>rCcLPHgBtWbxDoNu-t%Vlg?=8h3UXj5ya> zSmF{H%i|I<#$FiQs*UR6J`u$8mXs`VfAL>C^gObt&IC!M7VQ^!?zg)=1)sHUc5`_T zVexs)SSCCp?_bJ3zr|-1$Hahm-ZG*lsf)#U6&GZ=_YfqJT4oSSY5%+5&~XLo_smCVmL397}JCHDb)+vzA;VCm6>m$ps5RgP_2_}~G&z`f9x_Fvuk7v>!HRH}14wqmWmOI_;mif-L{kX>aT`1Yc%#v&Y;*o%RlALAt-| zw7(ukcFu8_liOhCaXWkboujHykVuTE zzjWF^=~;x6?0uK^8g<$mMu8;%W2gPt&HhJGr~RNfknFFW_M@g1;{N!{234_WU~Yc| z68uf4{c3|^oXIX_jsfblKl%)0`nR3-fB6G@UM+5bI_-5{gS7wLY5z1d9CzC5yj(RD zb=tqo1_}RFr@ejqi@2tGo^SAXh3pGJPT5t$`VhHF4DNVO0d?9B{S4ClZ=LoLHY<|Q zonUd{5?gV_N&C+g=P$oS@XL|B^-(;Sfp;0!l!18u>RC}nR_6w(iDBqY@Z<+bBKv=- z;Ya>oPBcH^g)Dowba4=t$W+GC`zm?TXlNXE8{G-~_keiblIcY(|uF-eT=bI__*8AK*K*(N2&=Y8gf> zRrA(QLv;xV+U+SU1)ODOELoi|>9}$c8e<7T0vflHpA{?%GXmJO^2PRY5#&i4G#C2_vQ*u7T>ZSra6YT9i}q+`m(M z=Uil&u-^TauvBuE8cO7p=b7VAMj^}eZ6Kbvq$-oSZ{Ml$EM)QD1CmHBhQwl^@NzY> zn51lo5tj4}rY_E#$(q$Q?H1m}Hp~R^yd{>t^$9Z!!X4c&-+?4j%Q|A&6?uCQvaEVI z=%KJgXEGMocyd~Oj%KhTs>`A0AfC7AmXW$-m>gs~8V~-Aic!`Ccz#EX$qb{;!I)Nloiy5&DNZb&E>e6$i_D5j}$z?2a z*O2R2W&M<=s4fjRfOy_AU6p)eHF<{_p8K1(f+SMQX=16P_Sywmlxh@x5f-<%j3szF zS;Oqp`Q9B_>ed7Cyk+1z^7dMNTR$yS7t6*ViPVxvEVEB`Geeg2W!Ec(rJA!ud6Bvd zd48-ZvQ(Rcc;1qtN9N0-qi@e4%hk;wiPZ9oSe8DVdk^Lylt zZ|2yYJ5XJEw*~RMMR^waT9D6(G5AeV?_~bK|%22W% zTaaLbJ0f~q2JyV*sx#U7bz0nK1FB2NHIPJVaV3^l4GNr5T@0J%HWL_YDQLsq5XllQH4KoY6tKCxsj>8+0}yN(645|$Fqa;z!Yhnd?( zX9KdBoB{E?r zS0J9Z7!;G|{XREs@w3>H9FRn6aUz!d-(_vld>L^dv7fM5e`G9gEy%sn=Aa$^UAVcol1MGl#1gdmc{sATXCE^ZmORd4LcglKF!Dwgs!Ojz5YJmi zhLbl~dwWjKL1V1q3rHfhloCsp^TL8BBm~8D&&oCjOyKY$dW%8#PgPQ zeq`^$&g1!KWHA{Dl1MG>-{84lzT+9bS32x%KTcTuiWy7Q1#*A6(sq|4vK+Ao@w}ye z7gCqB_Z5c7GSCGiky>UEi^_w#fw`j~G zr`25!yW=moeG39fq?QO`xlvI2EUJrj`N4(4lJ=RYi<1huCNIo+`x9B-{si&7Wt9UN zWAnYkGEiNHDcb${BczrhVo{oqicb}!PMN%1SRzXp%f^}HUa8`#HW69MOh7zu$r?cJ z`!%P&#qVFNO+gZ=MJpZ8{paUa)kSp~*P+E)Ve$OJSiGK->saW`-Q-3y9|}OLWPazf-$< zn~|m7caTJC2_cpeV}0B`_}gWJwXh_XF?F%mBV+8_*)_kBW$0ND&s(~mAp0<8c_}-P zrP2!|ky`SJ<>}&Mtlxo-6(i2%;4g^W0 z7Jp*#wtI))SxjzkazR+4zcH39H}br%9IfybS$cK@@x0}IC-N+I>Dz^1WI5ayB#~OO ziRJo(%vY!`_5s?y!s1oQSa#MV-B3}RuX-ZOq;L?=TMTW;{+Q;kKKMQI>S&NeYH6B@ z=l-~$+xXpUmyZPj!eaNGvG~6y9e0M)Rq;A`O*x3?E&AI@T|VUQvO{&zss>4BDFNg!gGK3izqc@ ziFlRxOjxQpOQ|LKonF0URs0JF)wv*^xAf{o)?*&2wX9KH9v6ZnQi~C>m@A!Hjx2-L zA59aM?4L|s%66^8?~zXq&KZiw*ftQ)TY_&K#n1b(uW#co>}XhnBvQ);Vwt!3h5@qp z{+N{|EOEaW%WFNdhEWcz#{2va6=wg5=Pd@FWIZ;m&$MBvE>;ad5~<|^v9y?W?IW_x zuhTtWSbTmnmW*2DI`-}R?8c}reN{j_Z%GXy?_WaWOBImirz%JywImbE_mk&v2hrn4 zT&<#swqBDE-HiqplPOV1(8uAv~Fw|p5+?)x4S2IDVNWRC$!q?Q51qM~^>42?1U zI{T}HC3!DXm!u8k%M^;nt)?Q&1r-p_TW+S3`4TeyIsTHqry58iwX7zV60Px9k>yOn z_?qkg`|dS-A7e>hLG~^h78W)_mbNcIJa3u3lU$Rx-u~>1#@LJ3Ac@rCPAuaUo8tZC zT_>6-35)xF#-cW#JXb1piNa^DOguq6Z+Wntyx*@f&lrs?4ShfospSc=tTQ-p8r9|f z&(|u#Vr|1%PNc06C&s&BaC-dc1$edot;-~_WNG&yT@Z4`X*9EU( z0$e>+g(Z)(s68fY{v|5iH=;3C%LByomagr|HMvnteY|&3bO|JpT6zu5s2q4X;;X5u_J53oKRifjRZ-gmSx0Zr+>u?jj^G- zdg%y@-vP!lY!&$e^=zezv&ho)0Ep);hmMl-{Owzxn1?L)96=JP^J8_YKs;~xc9pD?17c_54ppC;6@UH+sU@CR((c@-gDhqz z9`+QLa?Wz90U2X27g*shQLZ@$;(5!tKIALs!%k1=hw9SD8zhlhs)$89{1~1u^R~P8 z7nU?TrY`$lkY}+0$sXg-7+bLq#PgP(wa6HI=)A@TS>7H3Nu(CNTznmSpmTl!s>|+y z^NfTg@(^R`b)T#)k0ojKM3&3LK|F5>%_G-j1^>=?uW$W$kVI-(NGzWE6Ytwc9ba;PQ5^kcEV6Xi0g^~9F~qVl>^uGv<(yU_(}bmf zvow7`_W9LqBMzdvT+{*ayv5=QxsGk|7>3XER&)hPq?U4Gsjsl^398G7E>?4eCDDPY zOUPU@_kE}SSdT0by+J&0nLn27DfvAqnvUwSc_2t4wRC)o=l;!}_wiZlIqOF)7M4Iq z#!@?)?8A&{qJe*PCUP5y=Ph@w$b01a2lDZk!X{aRBvQ*9V(EA)7O(j?E2^v%7FQ?6 zQXWFqmUi>qTu@zZH3IRxMW-uyU+M7Z;8|3c3F;t;)M8I8H^#T=jK)}AEvPy1(mAA5cNAa)nRqexmO5xcMN}|EW9u*6=a*s5{-o5~P%Or(5lBgPDC-EUZ8y zYLgX6Bu3OMc?LT*?_lvrV zyVDbHRXju8>02&?g#W5LUCH?>uIZwlkG#=ULf;?clwKv?ldHsq;Uhy(cY0w6Nb|pS zr{7(n?T4w6<^n( zx_Fy_BvQ*XVre?7Q17NrEn_wCh>0v1N;PdUme8r7L}po#HwaLoyW+c z+Xf_&T5O2r=E?c)$kJ!(G*@9M;w;TSli%r8zAs*8dNi03VZ^qXF`J8{ElD1X` zNu-tnVzHav{WY>|PRKnkEFrFpCBF^X`8_zl8ownCdjaBkONS$5j5S_SH38K{=M6|A zwX}MN=YCEL)&9s5anQ#{Slo^?mbRP7+qlf)g9}hy4mg8&-tx2uIptZVwl^PHM!A6` zQp*Hl`Tsb(6Sp3>?s5Fa$Pg-o%o&rJ6e1B)=9#EWDKcbCR0<)4A_+w*X_AnnSwaXQ zl(CTzLMrup&i-D%%UbKKea`cI{(!ZvUafoB_uhN$w~@kdJR|>N(5i4@spKq&x{))9 zN6a7f5Lp68fOy`rdJQ=<@{XTrEoA9C79^2cwh>Ehb;b3lE(*(rUKW&12kVI;UAeN{%6RMEK;d+hh!jf{7vFr&VGg;cR z^y86bWh#j0Eq+yGrR4GGG5ELC^6!EqQp;Onc^OL1S+p7OC|Ov7ycmo1MsmN*oKh8p zEV{!$Ja2hxLT2SA1Z%HEb@^lfl1MG8`S{+i-{=Y66+Jol#BE`5JH}X)Mw0I$Q)|Zy z$YOL6#PgO{x5-SFz1b8iWEpY|B#~MSi6y1<2JW<9+;PbRVJYD(eHxPU{62R1z7Sak z^#t*}<*+4LSzKY+7I&%?_5(?zmQBPG61f)trb^kHo>{_@>CM!ohu&g5^dA;(u7)h{ z--CGG5+6wJ7t^z8_*9yM-#`+nC74)D*L=pi*1VFxzZ90}ZC~(HRYrNG(bQ_}=gMzPSUk)EntiEG&*E7>j-;S;5jcdTAN56paV* zyro4)(xG}jpeA1RDV+k6NG&6YC8~!B-bsAfWa2kr$>l7)*N|Dc)6KuMM3#p$K|F7{ zMNdA;+4~mPW#IylL~2<}EGZSn4NzTdR;ZQ>OWH}ME=sm!heVsbEAW~3y6ZtaZ;44I zGrw!k_j!QsmuL1MiPYjxEQ2G47({%f~K$SKBRXGY$YOZVz!qq^j(f_UCi zJcvA#xBRe+K$hd`Ac@rSgjkY7cdkK}eOJ!b6Bc)0#xm#vc~{FE-YDoOkL(Ekh)CtJ5Y=)AweLXw;0nEte9uh zeUPQm1&~B)SwSqb-_FWG*O+=q19f4EJj`k217k|d`S)F`~nRyzHK$gpzAfC7AEGBaoGh0~U z(@)k807;~lT1EKYU%G0}2xR%N)}pVl6mpggQDhg6)yb{6yW`mw5YJoQ>5>)S4jbR> zMAw*-3rHfh3?vqpu^j`D<;l?Q+QO24mZ?i|Jn86uXp*}X)#dAG5YJn@vdEeD+Tlt) zk;Ts#B#~N{5{p&m#GA-+eeI{=!V(d{SWb35i0}Og23zs(kTkRh@x0~YcCtfapQZ)g zxpLDPB#~OYh(%@UivnaZvcED~SUdt5%jtCTe(C)D=0sE%n;jsYx9qw{R^mI2_MCz0 zQez)TBDLHm7VGatj!?2J^+)WW|HqZ|w3dCTHsWKOA5rW^hi+m;NH zNG-kJRm&k>e{bBv{h5t&a`?AbOIU1OQuK|F7%-cQEKUrR<*p}KhW z1xch9Gh(?~@D87LI@#vULSgX_W-K-}$-DA!u+Cj%(b@>&dCT@SWG$&z=I-Ula>4;5 zky;KBOH4#J15}rGRe?)|#q~U6F`yQc$D__7OM*hspLpIf&YHYmnw|A%fh@h0KoY4X zg;)}+7wkY5JDoKvg{7FY%xp_`tqr~JhP%_>j{@<$Wl~*YskclCpAzV31d>QCKZwQq zt#viBJkKArPFONRn7Y(AAzhu-{w z%SY#qcn4l1y&F!#;uXeNPREhcwc<+dVoRPOi03WIlgYPO#(=i?)PQpnK@zFOjaX)P zHmQlOvDZ)bx(Z7rXR)BaIc$G5yaB4q!&e}lx7f@eyQ9)p?Zqb^ZZ7~yq?R~h*?(f_ zZDff#J@tUFybkB;Vo#pQ6?zG6kfriGi03W&I^lMN`9?PVG6h?v+kkl9@}VAi@87x__7+*{Yye55mQKa^-akF#v@Noz zD}M74mY@rarG_q9k2_QgZZwJ(mgp$P zGUNrB8}E1XJ)Tdlz6auYOXv+UhMDocV?3&h-xH8TYEl1)@BMY}Hm*i>xv>06oUr&@ zVk`@0k^ANMgY~1(HFj$)i03U&>XEu=bkV?d*}VlMky<7aORse%vymmd^NAb6;&_>{ z6c&;1%A}HPypuSr7l`LADqYEY-zvY}2UM4mfgp+0;z%q_e@7`I%kSok(uF0Lv)ma? z){@TNd2$ie#h^Wi=Pd)=$vn)Q6Tk3t>{C~eL~4m7mUrI<;8UDz`)EB9mb5EOU3#`4 zCr}%ovnfZGRyrV_w}fsddwY6)Fg8PVQPT%Wq?TM_$sX_$pG3Ji^4l|E35jMbsqM)9 z(xG8aA+nsk2I6_kFWNVSCO;_;lEO7n4hUA)GZGZG|`TE-De%01=z$ntRS;R0bP_9wkIUi5vlb@{mfjc4Hw}T{7%Q|Aw*4l&T$J*W;*;H84sbv_kcpTagg)FA!Hk!hcoygS1!i>zrN36$i03U% z$KCNsVVYBiZ9$gC+8~M4Qs)c4_b+9vErW^0zbFImg}Zxij)|~j++gbBv7P(^^c90JS7eEK3gUUo%I;*p zB16~%2>WVBje;$Z)YuJG1~~@d5i07G6x$w zzYE@mrsBsQcF6qTrY<&{$-UoW_YQSr8QdAf^OoUmWc?+z-WR-P z;n@=;ky`qG#m_NaCx@}9p?S!-Q>%n!_pN{Iw9n66g`dvPE!2}xr~UP*Ad{riep2OP z+-YCmB*7bX+Pf|SS^lllKBwMD{6|gDv;K-Y?d?{AbpO<8KXW*FGV~j-SQ~ZPtF8yB z{!g9uJ~2OMqfYxAdyrDzX|JM6I_(b`^i@JvRIUq1B(A8xb=s>~l1}?>fmXXvr@f*Z zNb)~++8f`#qlG%{M;!*q{@!V?kXQ#_nD#}P7tn>N;tdk~L#O?X*$OyQo%Wt#sMCJD zAIS9IcG~~#541XVa}VmYPuk)8=g#@}PW#27w{fTa$paS~qE37B{UG6g)oJhEAqrpr zQ}i67(bN0L5s*{%lyDqPo)R&$KJG!C_Q5AWn*XiSzP|Y_Jfa;NcG*@uaZ>*C#Q9sN z{j{HChB5tIyX)v$SGfY>`KM>!%j9d;=ljJ?=>IY<9wd?dUusy8|4Z!6{kU^@=kR@Y z!V+|wu`G2Xs|y_(1ZX2m%dsGyw+!r0-q3;OBk(y`4iiBVsl|g>dhQ;A&yMPSaGH~_ zxZPna1zP*@rPfPnOKW6V=MCa{%i*SE7tWj-L;X-)zWaeBQcEha3|@T@|8CEYPin5h zQo>oxG{`k(_0k#tri$}V5YJoI7m{u$PwQ6rGzZ0MkVI-JBbL#dyADI6jiGMu_6tj9 zI#ZXmTgWQu6gS^#s4j<&f_UE2u`f9Rq3h6n*HB%QPk|&-OV4lkQEWDHH2!Vd{&&uK z3QP1|#-g`{j4>=0p2Rcq%Q8VcZ&B(;zBKAJm~s+ZvOyB5Wj?Vy38{_GihicB+DBM? z?lG3qDsqhtKQT>Pa(~rWE{*{1{a_1W@y})~YA4A&Oi|wE zGGuww55)79uARg2IC+42VJ5On*8xeS7UiG#-d|ojtOv3*OPkkLSX^@$OVJ1NjeK%e zC%oeOBLl?qmiyz$4vA@7RPp!AkY^x?)H0e_KE8{6fh?ZaG&F>zn6rd*Cntoiee^8_ z)x|gk#Pb%DZDg)5Vstx{F6YR}dP=`y7NGm3y%k6zwLB-5+xs_tK^BLz_Ctil>osF3$s+SGP9y3EA&X@g zi03WU^~k$&^YUT13;Nn6kVI-}REF<;pE*f*uJ3u6!3bfguv=WvDJ=PBk(Vme+5Xx=dV6?w7>8=UkHN)OJW-k&s*wr zCi9kFrzcH8_shx7Ac@p+npiqd*T*M>s*XH2O<027F_y4(%kcO95vNB<$l^L4#PgP! zZOOfVDECkbvY1Z+Nu-ucV!6Hjwk4{I#jka9g~ctGu?+MhGRdhnkXm ze@E$fRb<&{43bDKR>V?wgXK_Em(;;`ZGHq zpzf~=l1MH0h~;JJYJ5syy!TvtVR0;EEU{zB`z2)FeSGhioB{E?<>z3s%kzVW<5g6b zqvt^qsij6azV{CYk6(`J(zcd{v#{iHmSG#nshU%-tKm6V^+q6`w`eaU-<2KmGVy7r z-Bdvmsii-$l(ZO!*OHFjDB3M7X+=z3?v5g-X`Q zV}FoDYAGd_t@&{?kR|>^otwgvUc%JnXgQgKOoW4Gl=Ic_k+n8X7!V0J5XJ& z?*U1qmf6J8MMYg7)n&}pqnX0u@s+W-c#-#h(+`{R`hIB=i03Ua3S?|~+IGr(WLa|? zB#~P75X+z=1$dmid*R9#!cxXrymZJq^1d?~o~SNXmLQ(DY!4$RQHEDY~%A|65_n{>Iehz5)3bJOA}u0J2y#2l2cmzXy3H7o7+?i0ZPkEl47@ zd?J>;pHAZwbqCp3779y3DPu8wM81)C9dG4_EWcbpJa1Xnm^{bKRDY= zvHV=AyA4^UzjgZ|EUrHo%hEh@jk$GutBWiRKY)1N@_YcX^s9L?5Y^@KSCB+%aUqti z%^wax7Oxg_ehW)6XDRe1WBwmYI+49_?Lj)8G^6w3I?)J3ZwdGD7mnT~&5aAqRJRc)Mql!|9;C#{A&&K7J_)*(xMX?kBz#t7J!iQNPiPZ9zSgc-|PDhrxeV+9a zmY_<;Qofyh?-xF~wHR59hk$tAk~NdeDIE{EvL4kXVFXAbwWwC(d%tJ;%k8_B#;ITWZfI-^h2Gxv8MKQaYdP>s!&TZVzNRe+R_#meiAE4}8DyZg{VP^&^l(YUw~M zju*B2qHD}(jM02y$>l8dl$YZB<<-{Q`N(225ybNrm8E1KY>4WX!>BHaGe8olMTb~S z?7HcpCg6)ldMy){386p+ml__d$$hiiMrG0_=8N6?sV(gb8&aN!8(K9s5|{- zD#-G0-RU7ZBk{kwTR@j!)SZ6lK1laZ-RW26lM&7sZ~FSe0?^Z)@ zqVDv8KR~j-cc)jlHNu01Ly5l5=)x?h1PT73JN?&SC7h{IlgF;8JH4W|)}KrKzwJ){ z+aI_!VQ>!WPWNmC(*ApQdiJ$PxI6v(ieH0KcY34>NcdlMr^jVp$Jc+=lkZ{ZDPhnC z3QcU`D*%oz5)K-p3CtdiR}MUV=DQ-?Dxz(hAiWiEN#ToGsKy(?0rvG zuyTT@;1TV+mLQ(DC=4RA3|?o(tw1%7Y7de~E!M;m6W|trEDAk`trr$|7shh%GT9w< zEPm?_^bGplAH?&PcU#G5<9garH)QEO6eN*aT!z(c7nV}aGSQdJWUUOS zgJ&3r&IIwirED-+T`1RGe*;+_n1Uoy%L!uX@S|`qvb>ygXS=Xu?O^I+)s5`HGuk$- z6xC&)1&HS@p~ny6-Rc_-*l$FZP%Dr`YPn1-9y@GtcSrw&hjt4~>`ul~|0l6T+SS0j zZ>M^Lc;2EwJBRy57vVFBozH+IQpq<>Jok*B#~N-iDgcZA^tKSbZJ8m{CTdtav&{%zY%t3W(&F?~s%$tBrE zt5IE!ZvaW8mIz|`GT-_zve?F1-xU`B{fxytgzO|})n(^R5YJoU){xm!pYQs( zbGYAqkVI;^MJ#HgyWkxX2Mq^g3XAIj#$xiAtokJNubF`?p_4#7Z>jE0x;yrNSc+GD zy3PVgq?WhDQuDyatEeuW0)9OgmSWD5{)4PuJJ0Erh%EiXK|F6snnt>ylTIGS-(mrm zK@zE@f>@Nlw8LF$bpw*$2up@LQRu(%2cg#Zfi~VJg zL~2p3g~u?D+jZW6>XO!ZPoc0x9b_!^=aZS=i`stp)H4+=5YJm?)*<)Hy|!KODI2K+ zK@zD&i&(TBbh;sn{y7-lXBvQ*J zVp%jh9G~K3@iV6Oy1#aNqc-2zFZ zmSAGJ-ZE%EvJ7u*+)P;9jxd(B4~b>Tt8`mr8P)>C^A_WJ=keUdz&u0zjohdmNFuc) z6U)c_-PDl9xV5sHu#|9?CeO*YnC_VSc-1GRGl=IcUPsCIi?-z_JU=!@6C{yZvWaDA zlhCQilKAdfY%ms!L!oi03U&Mq;(5!>I;5-fN~8J4=o(9G1(HZDbBJZ{!h)x$E}JrA zM+!^Yai%W5ili79yb-wWYaTwh%O0giGw~TipmWLYm_acj3GDsq|>?M}1nziu`yiO}kCkl)E z3C6Nv5qZCaSUa6S7RwDFp0^l1BRh$6YB%hJESXzD5~;9G8SV$ za*~UwS;#SD(VPt8dCQ<+a=*0Hsy75#cFqP#q?Y=1@O!`0ipo+{mk+Z>uND>;KgP10 zo<{djsS!T?WRN+C=Pep*$b7Q4{_rNqa@i6jky<(u%NebD{>bveq`^jEDda5Cb;xr} z>%s9Ks4jJ8fq33hHJW_C_*jp{zd@9;03?xGbcsc=dgCl)ao?G-RanwbGj&-x${gSO znjgQ_N0x{KAfC4z@FQdXirsP3QC-x%KoY5C8nMI}C{`lN%bMOU!V+^D{b?N>< z3GY>iZvf(XOZZc=Cf@MonF?enRt8C=mKDSjIkgwwefy!(Y_G6*_%oJiKgb$i(xzm* zE;dQQ{7*b@S*Jv4=DLRIP!o zu?qu0Ja4&YL*~csXlo@ROY@;1iPUnESTc4E?}98pMpk(VOZHi&F12FFZuQ;EU8W$* z@9`j>w~T*D#$#QA%XcEn@+lyR)N+Mbmd`BxiYyj=l1~XsLI7h4NG3ap3tmRzU(qlg z0^)g#-YoKd(K9>U9$9imfFx2&2C*0$EE$C?$KSdJ2#bFpV_AHXjFYu)KEI6W;#~~l zdCUGuWXI}|CS8Uh%a~G-L~1D_mZIyc@rkl1ME}iN&(G0Iydj_wILJSSmToo`>YufE2o=v_*B9Jp{z_7T-d$_Ijem zNE=j__ai_Osl|aq;l zCq74(f+UbcYVjhL<6T|8AdBApjBH^^31uw)x5#^c*ooVRP+cBnfq35H@Q!>}8nlka zUC_?2K@zFu0)L%l@UiCrJF+BA z07;~l-o%prrExm4Y+vnEcl}@M$Uc#b<%a=z@7r{^fIIGfYzFbXWmXxP>pOm_FP`f= z=me5TEn|tLURR}u$g;_7Y$IWDyvSJY7Loa}B}-1?d6?15K|F8CEg}11l6CIm4%Jp` zKoY6Nj97F{j?_YyHDM|$!jj8b_O>DW$M!5=lZ@_{yelA{w`eUS-^g~4GVwhB=y;Gs zYS~OImID{KAWQS<&szygS`yp zS`HCQqa`PukY&|l-wwhOa*45wI!nIA?%U-}Lv=BI3F3K+a#!+QY0-Hw{@!n%50XeN z=ZU4@#_u=C@+5b2H(_zV%vhX0t-*C^KR5CbvZ#Fm@x0~J3NnW2?zJ1w!<_vNl1MEn z#Nw@%fak~h^c~q-SV}pIS{50P#ny|6M3$X8AfC7EF(P9a-SUZM=$V{95+spYa){-y zr&b}Vi{=!CLBf)Cg{jNg-d4CS_mj`z6RBe7f_UB%y_#HOnaw=#-1u+{kVIV+l(mb4uMUui*WbtwKROZyD8<%v;uqIPHb%k{$(;NG(n3;rD*d zPWxBLa(e60QNrSTm9a#5l56br)TVgfc<4?L&s+YpHh}l`^k03y3$iTV50XeN-HD}p ztDftSWvimucwupgVJzGA$(nfUlG%7o{I@cQ=PmmNkbB>H)s1LmF>MW!NGl&VcF?8^c?G92jY3lvJs>%%6+WxshS6!KoY6NmRKAV3iqIf zW}oaA3xwtOe=Ps6p?R(Oa+^Npix#b4Yrbgns0Q_GD%a>#=Wm_%jz63Dp-y}64a>ro zo&D!O^?&NL9~*SW4t3gBH3TW;o%ToF+vB@qbY-KD=!zPq3KEGc>TjL)@v5ZLzMHrG zP}FH}pbnD!kDd0OX$osmr+tskAlctL?LW*`#r^S4S-*YIh55QCNbnDx_R-B-;7nf@ zRMkhF_DlPLO#f}C{onq;Y2%x^qfYzvIw0-8ciQh%$-$lW@kI-JpicXRBSFIds?$EA zEEQiEj%ix4=!us#7UYyYB}O`sr$lL@dMngvziT2$^S^c4@6<|(LT|N?&6Zn;Cr(5x z^YlFBL`EC8?98&zwSKQPi07Z4)w$#iy{M=e-g&Rv2_%vIUuw9K|4Yo)t4ERL*U|yY zg~j7KW0|f(R(uujtKl#63#&jpZ&3&!mc-YCaEEH39Y`XzoF$fVZZUbtQrPe3Dq$(( zETdD%Zo*lrj`&MMcPNPGEt&(#*K8xBy?EyL$0(3QYKbG32X&06Bg^tucQ*)2b{tce zjDzIVfQxav-=S+vtuu({EssZ#8I#OWrfTTkbL|O|NG(r^#dZA)yram(<>*#nNr-1G z-RF~?UjE(7@ha)ieITBDrmJ(tq)ElLN>eBv%sf)1qCoq-_ zEi#j}d18J1)zmW(#PgOL^~wEm+cTjSvK$QqNu(CV`uI!3%*+{g+uO9hoYJ03k2)3*RNu-ue)H37hKvb974L=_gmSWDLTZ^p3 zyBrI%NB7Hj6%fx`>enS55z~&B_eYikZ9o#KWhk+X>SvB;<(f^o;w3B@Nlab#`4LO6 zi%YbS<=I>i&s!`4$i4qyWSBX!+_V5mq?YN#GB2aaDO8s?Q4XhsB`TS*R_vHjW z|8kBoi03W8kCA)d^Xk?bs4gF-fh1DPPGV_WcO*VFK>260P+_U$EDt=%d*9G>9DXL( zsiFKQp11V2CMzY@CPvGUWph1{L~1!jET&C2d!V{FnPy!Sme)6!y2M^0vkY4YsN&O4 zW-bNsyk)5;S-m#hy{8IU2H1ckQcE2MILl(CQ5i!D&lFC>*ttMwiMkciMLidZ= zOc2jo4C;|txwk57ex@5+EPD=sBvQ+JVsXFs9zT+-2z#z+kkl9 za&|kZ%jAp>c&FE_4IqisQd<$f_tRDk!zUmdA6Eaqu#|9?Wk<3lr2F# zZ`n7USi;*?TvW!?FmIUC{>or5Vz84n9bjH$aAzAT_ykmI-SHRbs zciR7m0!gHnW5n|2&MSN>O+(X5KZPajE>oAW^fw}oOx`gS)x~8gi03U!?vOFeJGEP> z=sDKM1|*SMB8jDScfI52-nR>~uM(D!dyHkzd2+v4DaSTL7OhAS&s#peB=46aW1J$9 zWq1roBDLHhmV2FJjFH7wX>#2Sf9>{gzt33mPm&Wuk`CU*C+wK3fq34c@sfP+{~A;j zfGll0fFx2&F0sr%8X1TzN-8ZI2}>zw>D`x%Ee)3@;_ksRbr8>6LidyVrOCTK`;o;- z10<1JeiKXf;PSc15^wfGSy-|%n7W*7PQEKk%iTYqYix@Zi03UEP006tit5V}WI4PR zB#~NLG{9pR%Srdtk!8gezgEH$`+%`5iY32kKd4~|9>Yv=0r9-0_YN|)Tre#`6G#!{g-X`5=kZa+p|Z z)vw(EU1M_+_YM`7h{ud&u`N02_{qwcq39YrdK<*^mLKED`{m-pn~hLiG#-K^QcDQ2 znCA`mM3xgvri~I7k0*>JtivKa2kRD{k9QR9wE^+GCBH4X#u~d#--#>}Hh?5j%MD_g z?zBP$S>AkVJ6>4IILi=S@_S0nnoq_Z5q-))JZ}k9C6;MTQt{4}>(wBM)bf&8)>ee? zL>4u>ZjlrvX=BP`Be$JUq0Ui@w~-Zi>!+|DvVA?*I2@RkVI-}+7Q3@71}=2 zN0yi8cbE!`|1-wYx-l8Um^G}KjV$F+AfC58w#`2~C87J4W*^GA*d-ekHyyf5>QkPHlCRC!j_zwh0q?S>{Qg~_%-Wgf3 zPT5*miaATgPjZ5r%`mIos4iJsK|F8SHHyr`9BMZBIQMV5_yKs;~pnM>Z49=$zqSEsf%NFuew6U!<0XeE4&P5`vs4apSZOP5|$FqGQX0%U(&8T z!(*6vM?gGpdDV-IEho;k|AZ`gCqNRZWf-xXbhN_VP+Mmgg$YaMTc$2G>XY^T+?>hL zs4lzBK|F7X=s@N!YW-Hk9o@f{gCtVR3}SiSzDo(Z#*7|CToRV(cZ|i`k#uyIU-7`- z$iK2dJa0*POHSFy3X8(mSadE(BDL5M%ShGuC}hbGTOTVdKDmtLWnI$Q6;e|x2Gu1v z8pQLKtlwllIW?hl9lFMrCV(VT%PwNEjO&C?4d^#&WU{b0<}sE{tw>$gJ@dow{V5Yb zJa2jMnRMLMYSz6QvY1W-Nu(BEVoA)-!M~rpdY-~#I5epmL$q<&be5Ni+y~wv%=YeONBg@+6AfC6> zu1`*K8R2_xGO`q?gCtVRLt;^zJgyD8U!Gm^cq%L*1&rmLI(d$*XnGBwfRGdi;(5!Z zD~Iv>rQo?eUQ61U3X(`IABg2iqKP50lucd!Qdryz8H>9Px%ZtX1oTP{VD z{bROGcRC?U7gLZ#YH3R>Nt@?SM3#!YiVwn)Rm9Y#i5IzF-Z>cIb+IO&K|F6UP$T#L z$6I8s@7xcNL~0pKETb1KuSS;9uadqBOYD2b($<%BLycN>-3-+wwh+YgmK(jvUInMC z?{A~JEGz*@q?Sp1lcU4rEmRF|M*!b7lt4lFsiD|bS&nGu2yNge69F_p$d5hNWcGsPyPs zWC?l;l1MGy#4^wC=?i2jAN9Jiu%v%v>e8E@sN47cy2Z%iW(?wai~a-hz3CZHEi1mi^pfiGJZDcLzbDxm3s(Fb_r9LH`mF$WuNUsCLzn(vmlKJ6nc311k?t}bLP=~6+PBxE^!4#e}8E!~MFQ1dfBF=XBakVI4i&*ZEl+6|Fq@evTuTRxp9WB#M-^KhsAz>^?})H0q}x@kn> z_x{70R>OtG^&4YZeVVL^AA9ize~V3Q2;zB**I{CbX>_e2x?i%DK@zFOf><^$9yb=< zFXMLg87(ZuoTb%Z@?E*{>1j1om(T7Xp0~W7N5+=hhLUfw9bO=b)Ut(GrnNkPe>Kyk zu{ccD_f1;Qs);P;%0WDDIaE#N#~SxH`-bYGTkFT4KSFBpAeOqHn&Y{P z*!S^MgeB@bWBL4otXVvZc!qy@qI)Tb=Pk{blkwPo`_}D{Sd_Y{59_qBWTLqH+ zz0*EJLml_W@17pp7+shXHh~2H&}ko1+7@RzcRqb7>a^E<4l@0>o%VnG1MjXa%R`;^ zjo*Q^|K4fe-7Oz?+VA^zwKwXtPbmTk|Eo^>NQZm)!f3ZWqA_|(SbqgMWlxC&`jn`= zTbheH?LSn3H2+(t{lJ26SI}E+pwDPi@x<}^!8|=}(#a@rLDqPD{-yR=5YIn7ov)Iw zrqTT;mZJZQZYW41`@hskB~Q%BU;bsueTiT|O zk<+>7Bk|dU3Qi!2)KW$)QFy3)nKixKF@<5P8YUxQVotIxPMRif{GQdGt+{zhChmT~2al@Am7tuBLNDIXCmbyA* zN70=9DtL6hZ4gKzwHOe~z;Tg1k!9zhUps`QgtOfKNjiuICbU_IEX$6Ac-|8Fjm((z z(^=wxuCb@5K@zEDKCyIvdZrv%@*EP}ge9|rsmp=@vh#jM@frLZs^fNpc-~TSgsjG0 zJR7kFS-!c0BvQ+IVo`84!n>o+58rcGSfVQ#%Z}ruE;%X*si-cVh9I7|Tn!_u*A9;B z+>mA5M36*k*-tF`R=@B}R*#XByoJT*H)FYdp1fZ+b!vux7w-Ew5YJoUHHl@ywHbI< zwDA;>L~02nmf@QA*U>fhx>pN7VR5WtEOl;^?+c@~@dHs^mJ9~*yd}bmyek)U?SgmV zqDFvV2zs zNu-vq#8Mgj4ga=nXH~09!V==B*t27U_6_Q+SY>Pb|DP78?YI(KZap=*jw}<}f_UE2 zaUhwMv;Ni}@Ai160g^~94V&Wk{@PWRGf-W=T+@#g7I!DcvamIIjw!zEf>*GH-T?8u zrGr1Y_s{e@jl0x(+yzOb77b!q^Sl>+@0+UCPZpL^&hqU5=}>ivTZ30g?{)<7yhU#| znK3zdK^xEfTK51+q?X~t5~Ec93)Q99;*49ulC_v_!_HGpoeFE6<30I-g37U zxyAwp{J^K5M63fzq?Vb)qFApb-c4w5{&B|JQWsS7sj$Kn|!}~S+fBDW|vhgi03WcT9W(a z!QD9*&~q#`1tgJLT#04drdxQWWXHqqFNMWr2V-e&OzxMm@iPihT|!@gc;2Esi0qJ< z=^Hr#Sz5gVNu(A(Vri)1?2GP~0L@r*V-635|8;$J_PZ+rF|^xKKs4gw`fOy_AJCR&tuLrj8 zh3XP<5G0XW9udp4B_{KbCEIZ8S7C|R#aPr2k+J2dOUF+ki^pUT&sz#Z$@fcYZblKZ z7|aGqq?Tf08LAb8=PovCj{PMp9!qRx=YAX^DKzxE%S;Dh;`jdS z5RgP_(I%GXZH?Zex>P+r-&j}@_Ar(XFNj4YEUz3{9=!zdyk(&#S&hr`RT+=2G0S|A zL~5B#EcJ&~okf;{q_wKT;=h-%7;BN=x>5BW9)v8X?tpmSQv92Y$JVsCzY1A~KLSal zmgU6qW6b)^$kL)xyN$58x-ph~XL7=h%Cru6Jf>BD$e(!LqCSu8@=V`51fNHq)dVDw zTDB9*)Ej<%k;TNNRwrR8<}6CVWMy%6R9}3Ccc~kQ=PfUN$tvl#8zC!DU5%o|ybw@vFVEE)Tlx+wHsjK|5<3xj>oH8!Xli03UD4@uX}$2v`M_u$#yAc@ox zMJ#qRt5+h+JL~6N9 zEKgP6)kT)0=jIF+7Ow-0WnVdYzqI(Cpo{8qJ`%+97TeKe%&*e6Hr{#PG6p1(TJniS z-*`J7^DnC3ez>qya+XboWT%%_-{Jyfaclpgppjp9kW3OGPd6esNnl^*6f4R2G3GQj3~09>bjM z+N=}0##T zj4UaNMze**?J#56nnb=UyZ^}hiYzh8AfC5q&m$-5?#wUyh%BvJgCtVRB4YW~uP^{v zc3)OrC@dwM#j6WB0ioF3awW3#NdWP@Wu7hhtQNse-gy=7Tx*eIabT6^;TqQ90-y~En&nGyw7$Yvb64R zWh*Q`M;ME;3z<{u7OjhagD6@T#PgO@gUGj-j{jO6WYHQ0l1ME#iRILnM)(`qs%3vW zVR1alSVH!?68i{XnKN~kVYKYln0 zOD<=5qegZ`_p>!wh3=R6HGcky=PgU<9PGv={T89RJg5hfNG-pJrJ2*C%g9pYm9$G( z(!7|uJX=QAELwO>v_h71r$Ib#$<85Tm{W>Pw;_vmFi0Y`D68Q2zKyMBC9*Uh>bYN7 zLXI((VYA4bl0q}r9mtY58pQLKv{z(r&y%%t@W~Wgj6o8qMUz-!{G7HRi*?pKPhoNQ zW-KmAL62eu}W9pJM7VdKnof2b^3x z1J%Vc9K`dM&^creiGsaXC9+gq21%rrZ^SaH#wdIum93M<9bt*^Wh_lK$Xb&2&<*&z zvc_Hz&s!AL$+}o*#BTgsoq>ly5~-z;Dt_<#9DkOFuCW_;=06k`4?o6|afFF@ zNT0zVp0_A@lX*+s+n<*pOItmVL~79^mU%f^_;jtRtL3kRCHpi}7suJ8=b^8i8a|0~ zS2Bp_EtjIh@pt8o*p{7Q_n$?{N=ziH@=Jruo{QVh=CLQyCHSpSoEbhxe zJa0Mnot%91bNO;S4^z1sB#~Nn6U&bBg_7I?|s|$C+{Q6J~I%{ zTRy6hHNH3P%<(T9Sy+N3Qp;&#Syo?lG_t(OXjLIB#hhhcQ!?fc|EP_-)Gj{*@w_Ev z+eLi8q+hy(Pt;xc1|*SMt`W=l0oU;9Co@!YYi|B)9XTU_sf(o%ndjel=QTdbyivUTa7KP^cy`MJLAsbnS2I;jBmP*c|*Nl9NRn!>W8Ci5MfOy{W zXcg%had6pMj_RUt6(o^b+7rvQ&&A=$qGhMjR#;x2W9p)xK;~gg=0w{fi@r06=Pg!e ziKS!F^~K0?bq`1)wG1JaZ8MbDBg^AM&pQiCN-$$_yGi;i%JX)3Bg;e|5YJomj*vC+ zZpIz(T9T4KNFudNA(k)W?eMzT-L+?W3QN#=#f+`H;(1F%Ir*+suDN<6 zvdlgQl1ME+#NutXatf-;qR(Y|!jc)v)Fmj0?6v-`p^bmxU|1Q5=Pf2#q4>QYe(1*- zWGSl#Nu-ub#8NBeX-#A?agHA&EYV?%CEbmDBQGfo!>4Pta|7|bWg$J8;`Z$#{7g{m1eT-Ra{GkvC0ItnPNyoxb2I$Rz1bKd@sg?oR(& zq?d)d(`^z#mVfI`AAf>$r+crlxr4gX$KC?z{;50t%4YJVc4Th*3#dDNRt8A*f9g)Z z_^UD=b*GQX0x9L)=`F{R?)2#jqwp%&uGb)uxT5~no!+w-=}x!aS2qoHr~fPfN&d&~ zbiLT0xu`ommJLvMdP427f9{-r?@o8W`x$qq@0gZ+9(SiV015xA?(|gar}!0Map+lJ z^u(K@406hz5((MlDKTM@YD3hW?%xWe`QN(J!*lJop{J*dZcTP@-oi07Z4O)ikHCc74TBhde)%LR}`_J65SK>jb|=A7_BmZLhKX9Ef zN-S@7`{DEa4vqrxyv6V^v218|st&RYGy+MamKrVamqt>cXCq{p-X>~+u;g-R*$Rj4jKG(bFW363Y0P0eTHPL*Z7KoY5?6|ponjZZ@sohZ8{!jg7@sY}j0vO{85 z>vxXGGVD5t=PkX>$xK%7nw#+sys{LKL~7|zEY<54gdmI4#VIR!St+=)>)HJmn}wy6v(#@u?)~B8=i(Vg zzh@wxw}ieUU$fd;PWX(XU2i}Vsbwp%4E=Fx5W2=P$DVZ*maHhIE*kFSthH%(eZ$c; zHgXDx=Pffr$>_YQW|tgPm+ILdiPUn0So&N_$1^5}e-CsOme@;-W%C=-ZM%5ob^M!< zDGG!B#Pb&GJ!EBZZl=p-WJy*6Nu-u=VsYMf?*OWczC+o5Ve!4pSp1io;n%<4`_U1o zE{eNBJa0+dOzsyKrD4aAWxqQ}BDJIui{(I_B4qj8Hqld9T&^&dAHB&g&%B~zN04Pv zHxSQT78#M}*r7|SP9e*{J|KzI@`hNhPp^KCEODoI`3OrPXHl9+RT{^z~OqWry8u&_j2Wh`A47h%g0y9&Gu$FvH> z^Om%0WThlBc(?$&mSSR^dgp%r|%3$7V8tAB80^whOyL{OJ++~bliO!U1LuJ zKs;}G^_Kh=L2>ug=E!0h29iiEV~C}qh9UkP#rO9juLw&SXHl^vD_H3s%R3@V?o$xY zTMqOg@0V^1Lh!vm;1x(BwJac(rAJb5=kS;x+v0>J`x;Z1jdRJ^^2w4Pcn6+wHHhae zZx^ZaH<@9q>tT&&erNOr@w{bL z9daVo@LHy8kfqKLkVI-ZM=W_gqw(*;>0iluC@ijVjAhsgGFLS&YQYk8zt|*zc;0e) z68RQ8sv3`f(>@{%B#~N@h^6L(pLiuc<-N}{VJYS;2E}BY{9CneU1U)y1@XM)&@a;6 zF}r3a{yin13Xnuy570z&gNkVI@Fnl{Zq3x z^+t8c4+lx4mJ!5~KX4O1$>m;-M}@GwPGag3vyiOB?>aRMpSrpyAH?&PYa_{SkM?&v zj6jyEPauiZGMiXN#Mekf*H~ctIkoKm+U=2&%vf@sl3kvT!-8;US6+7z&szp|BReD- z1m2BDmN|Vv5~*bkvBdgy!@r@r3osnz!gs>_2}@0$pV+YQDtviD9r=C_)7C`!zR?x6c(Q}#yk%kwQkTpV7w02OquC&d z)H0P=biNFYMV9`vdm0K$$Q{O#MStHiLic)iRF~c6AfC4F{nK)+kR`?hB#~M!6U(i9 z=f$Wlan@tag(dbLV;QGRPC&SRcv634sj(i!^Oj~Qs~c-w;{QpT2>#_6~5`k1SUffOy_guQ#z6d!&9rbvdyVB#~MQiKWl0%e7Hm zwyn#s5f+yW#Vm%E-IiPX}XSUxp$e}L*z zXuQZ?SkfObbum&T>y=0J8sgLE&1Qml-m>dBc_te-pM!Vb{xAheq?Q51vc#s+9a)~_ z_jDGPh=+`&?goXquQn>#E*_lw645YJl@-Uj3GSl-<8yO3q*K9EFeS@QpJb{~E{zTX4* zjgUmjmXTRT$;!wqvR9OqNU{~lmZoG?l0+FPR6;3)P!ypEr6hz=VC4yKsekyv3Ea@Lx9~YM3CyeDle-nIv>0509o*Vc33gUUo`<|p% zp;??0-j(J}B}gK*+##06)9nh8MRCU4Q^Mk&!C0zt$^FG{aLRmCm#_0cJa4(0MxK!^ zS1fcub(yjlB#~NP6H7wnxK_xr`skT+!cxdt*4`wm)KYz`A4QgFp&*{OT)s=j$@lHG z@pvpI0wj@I%8BLXH+@rNF{`oWlCY#^GIcT2Bkz*7#&p?>EMt{Hyu{)|cGsG7u`yoR zrP&xHky;w7<1x&8y=wSbOnH`moUlYcWh_coWVh3?-&gfSmKv5Ip0{kRL5{KMKL@@? z$C&L>kVId|>dJx}0qvcXUg&s#ozCiDD5UXJlZmXAgt ziPSQeSkmlok41H9)%kFmu;e^r>Y~w{-1nma&B(p0}uWAahDaol?t?<=z*N zL~7YXEK#m@`;aAOUBB1D67rm}Jb6sUFvT-w;rB1IoIpHpQEx(?_w(En@tFUmJ4hn6 zgb|CC;xW7ua`CXrJYm`Og0U>^L*7~J`FvOx)x~@ai03Wsnvya9woM=Jp}I7k0+L8A zH;Kjc;?qP_mz0``#llj;S*mRyyV8u9Gg}Q=GL=C*Z+R0)&X+yzPI#x|yv87j)RIjs zWvUPHx!>;0&ac9f_L8Z~S3B}7w(+wAUOgCf5ybPB`s2vFrESZ&-KZ`zuY)8~OBu1) zMocS5bqUfpt`wG-9LACwMBcsbe3puLZ+trf#PgP_O(*fQnD6O^smL;6JV+w7G-!qI z`+I#=U6G|*Tq}i5f4xTzc*R)0B$7F$CEp6&&@r~;4v6P1dcDZGUo~dGJF*x(1WBY8 zO=8)6_IP7t`TF6tlCU_wW-RM(k{*)AcjoOtmg8k0p0_mKP3A7%xy{7){T`JdiPSQT zSQIu^&q0=ogiuvs$>l5^yO3-0q1ncVk>z;`i03U)Z^@dG=JxlaQC)7QgCtVROkyc8 zHZ(_;DO(+z3ro@)rY`1<$r#2nwY(Bp{5?QCZwYEgu46-;ui-DWZS@67q!wFZ8E^Ti z1+t9S8`)M^!gCo*n+sdqmhkQp-+aIhpCG zhAhhq>US0vueXflP7c|zJ#K)mCbAfR0r9-Wd?Xpe*p0Kn?~*!~gCtT*FtK!h+M*@0 ztUj92U06yv%Z2eva9s|?#^HD3dqY7yZ_!F5=YDRVLxHF+9+4o4)N++rymhN(BFhz} zBfW(s^Bq%{C;HadVqHUhF{;ajMj)QIxY3=$+TF@*fGpqDK@zFu39;NWEB}lvCu=Sm zC@it>8B484vdx%WlcVOW#ye5^ya4gMrL`03wN^V8i0_rt-hm`ii(+eh-#1&b`Z`+BY&N5> zk+1~)&+<1b?Y)D^9!LicT~HK-GIU^0ECdPuW~F`0C~cf+gndOLw9;Pv3&`}ZSK8Ab7;KVmgjU*{Hq`rl z=KS+Y`Q%M@F3R|{Qi z9FFeLMJXVO)H0A*y6i5Tj4VH{Pn$0+$puVZx{fAywdL*;nxi|k&K(fXTik|_w{a`7 zQ?rq!`9qLIYMDwbb?5r`Lza4T7OoJMh!2d#s**fuls%qz7g;(i1o6D3UIlq-dfK*c z5VDl407;~l<;2otp50eu3BT)UCoDdNj3q~%bgpDH_Wh15rfDFaw-h9kHznT<*UmzH=6LbJA6?wB$2h(H-jx;(3dcGkF_VdwdJLU+jd# zAc@p+m{=A}^XiJ~qS|kmv#?};Wa^UaHv?aqHO4-Dhw8F;7>MUB&oYSR`MML2QC(_} z21%rri^QV*(+BT4WN$w>-sDSb~cg%jMH#RcCxb+zC_{`?VmRw~V+z?k{c|&$mPtqb(qb z)RIpubCTxbRnVUw?e!HF_Y%gEvXFEV`&Y}XjVw*GK|F71vX-3tdAg(Vs?H;CKoY5? zMjQO3aiOXk?z|tfX!aptDda5Utw^`J)zu0-lNE0R;(5zvTQd9erIQYRdp&&}NFueU z6H7A<^1X4#OPzv+CG`_im&{x;hN0YZL-u)!ou`o9iPREFEE-PE_zl*} zrLU8PCFCn(NjXMl<xuJ(lk)`W25YJlzqscY7ul;jdWcfTBB#~O)5X*-}J#Qn+epBl# zVJYD(Ceh^nV(NL(5Lvd)1@XMaJdwOz(0cs_f2ViQB9KICsUVg~{x?gIrR3#+SHhC^ zovDjCeH(Yo`Q-v+>5>oPdCQP8a=tV!$h(N@ax^1ki}>?NFue&BbKn|kDDWl{cPh3Vaeqz-P4I> z`OrBlWbs)Jl1MEZh{b8K2CmD~D9xIi|LQ78s$lB!q#ik6 zUVccei!45|AfC4@{Yb`^FT2h3LY5^7Ac@rCO)R6^mf?<~nb$wp6_)Ty#*%GF=Et-X zOD-UbUJ{7sEwP^DI`*mlV!Wa|{Q*cKwVWcBW&INIE`gfQV;TyJR~2J1Q6XJ8elbh& zcNJE>1M$3NmKQl+&iYQm9YtD2Ac@qHKrAlZQT+EfB#~NH6HD#BVYs8{{PwR#!V=-i zSoZIph3Cfizj#>()n%SHi03VO^U2(}Ym=f_WT|%$B#~Nn5la)DD%`m;!2a$8Ve#3@ zSbBZljxFK79dYMM*;WwGTPAinhr2vG)gEhtENgdzBvQ+9VhQi&+YOyB0m0tWgylPD zc{7TP`OmDohj$5Fa~j0+mfD5nxiX~heRX8%cpfB?TCNex7W;#%k>&jJ*|UWu+l{Hq z=;`Dc*}lixBj^}g^&Z6YmKjaReZRxw)lZS7st6>JS~7@b{pl2EWSJe&ZoaU@ZDTC+ z^T|7lxus(gk)P%RbXkT@t-O5~-zbJAB{w3Y(3;rIxWh(neU^w=!s~3}Tm#5-h5YJmW(+<4VeiKe2%PfD8L~3bIEE}7@PDgdw@^<}NVJYM+HF}ZN zP%X1`@ilp65s2q4jjNIQvGZ^1)kK#5UqKS7MW0wg^-J)#ZQoZLYA-CQJD9p;T(QE( zSa;QXQVB#~NH5KEw)r$4f|H@v??Sp0S|7PADhLUqO9 zDR`xQcnc8ETQ0UBy9NE)5QN`}yR`#Jq!u@7nVJ`ZEM4y$-6Jd&oF&ZD5+7p+G!9Kf zb*c9V#PgO%zGS6}kNUUGs4f{#K@zFu2(gsu8n;K5xrWR9ge7M;Qx~0ve z<$`$LQgMN-f;Jm-c?hyR{{WImEj8QY`@U}VikHYTOLb?Iu#|8XGYj%wtn`KO#L(A^OBTM7C zAfC4jQy}-sD4Vfi$Z}yJNFufLC6-1ZKk=Bq&O!BfVTti(EFUb$S7+iphyFsA3&%h_ zZ`n1BjA3S;Y4sLamV|*MQp-eQX}vN1Bs#{rMCRNUmVmvCrR8MO3paUsG2V%C^JNgv zTUvi6WB%^lJn%e!?;9YA)Ut?J97lO|N0#}GLhcKTqYq<=IY{1FtZv@7KdQ^BRv@0Y zEICWYFsoekaBojkN03BnaU_81HwoK-EMG^A%odiUeN0{2Ig$IzfRy9-+#fa(#PgO;v&kH6 zfvz(C7IeHRNFueIBbJ<@F%6JK)uiqlVF~wTEHk^1d6;d=$+%Y`*&M|4mcnDCE|<%6 zN>N?B7J?*F%ROT0+f*qCS(fyFoG&b1`x%RZHn}G6X*&Xc3Cwpmi03U8orxuOS2Mh; z=EyN1iPVxyEM|7~ijie)FTWCDDdj8%24vo{Tl*FGJ+h%bi03VRRLC(l;DH0aCa*LE zNu-uaV%c0n5BDc4EiwNlESY{xT^c)(x$(trw^pEYzrHPq=Pe78$$0E|^Tapk7<;+_ zB#~O0cEDqp8sERYF6Tv!+jqWDcJw@9{K1&#3NDl9{cdNTs-U`Dv;^_I#b6CN#;kAHKR|W4 zy9^|eS}cg=zT44Ibd2rZVy`AFd7NcIIq8L4*K5FMWa+;R#PgQ5cgWZ>tX`w@$nwSm zB#~M+63c-LD+eJ<=i3G?ge5tEsmn2a(yL%~Y|t5GNq7L_d5fn%8S_VJJ-C1@HW?s^ z)UuaYv>cs(Axr7W#u~yBafq?R{37pOH;hzsK$e(D5YJoGf04QIwI;3SAj_;vAc@ox zPArF_EUl2GvdPQN!s2t7u{gVs`Q*^(Li`>%s11ncEsc(nUWHFhJn;OOL1&OeYDpxP zz?k~yki~Rtu(q&#=PWs$$x8bg+XhZSb#a~w;(3eXPBP}dz9`KY)n&+RkVI;EK`e15 z?Te5lrpB^9!jgT2sf+C+a_(O*4Bvw+jg&z=Zy8-d=3&$)sN(l84;q0aQp-1D>9aVq zI$CVndd*%VVF^B}^mnVEPv(%TK&^gX-lA2|1Ga!nl2y>}P7c7UpbZVzUPh~+8@Pci z|FjC)<_cK_{qgH;Q?v?t)ozgP?^Z!;btI!`!)7!N9 zdXXU6pR1s6x%b9vmk&Sa;EfK!oG;|ESQ zzvYHjK?fv)wEw&c+PSmRh_UsQ>MGS)ZE3OG!gjS;amSlz6?DU7knmrvf?nMJ1D=Zi z*eU1{y5co@401}Z5_LP`5w%yl6}!+X=!~Zz&3{`3J+ssG7bL^^Olk*(oy96co4o5Sls|gq?RtkQs7_R1J$KRNHbGm zaX-db>e+9@hg$iS_fL?eOC*TrEp;1{S?r#RmY1WtB*%axQi}nxq}Y`PAdBs*95Z1l zCaTNCR1nWw9-bvnv!B<$#a#u4HN$`Z2&rW^u_X2=Y=n-nkE8U~3X9)y z#-h@hjLy$#^-n<-7h@35TRcvXC%~T#Q>LQ2ES?3DNG&IbWnau&Uu3zpe2J5=RB)D} zwxp{d)o=*jm$ZBhi03Ws>yUl05`!n4Lzav7Ac@p+omh&MG>#(6@}j!Cg(c?%QL7{K@`+e7KWZl-OL=~Ikg$Y=FqUo~X5;Hv zZ$C@CAG!G#5YJm$j3(Ey*`bcpQC;-QK@zD&sS_R#U9sqezXg41dE>LfvMZFas78_w z33ZEO_-Qs~4T$G0YLEBeYx1vEyK(pJZhMeKYUw~M>e_nvd$7SdZ!QT-31|7KM6P3b z4;$=6$5=r=i03WSeaLmp)7s%Ws!R9JAc@p6h**@OXDOp&EZ?GeqOhceF?AW(l{}fd z`_A=5mdIrwp11feBv0nnA!gffyBdzpGg(a7>*gBAVWsgRa@wW)#I)QlJ zQgM0Gi)IQ#dHkXk+v%Mh)Tc>n&aRvAr%#Vdlb9Nt2XF|7bq+*P2Z z1>$+jfMw))->sx2-jDo;E=VG^D0Ig6{e783yP&#Q-_&U*ETx=fDSdk#P;fLB)g>hb z#PgPk+sHAdUaY?b)#Xe&NFuegCKk7bzwmv(?Wn-+!jc)u)FpQ;c@|4dveZGAYm-4d zZ}H6|Z%T%@YlAyh!)JgbQi~q3jPlrd57i~fpmBd;i9N?yPQ52@;yrG@S%xfX<3T)c znc0cdMPul}>ZmT&rhz0<%OqlHp?$&=SGly#aUEUwXvCAtCWz747z(HL1KSAlrm(sDF; z-nX6lwh5|>LG8}Je}vTHL@au@hr1w)--+t8ge8x&7`-KyN_YLQ$WlG9#_xFEk~*3k zV-K`Xs31#LC`cl;93&RU>1O!cZ=C;nfv_Z>XX^5LC$YqDsfovy?q5JWZz;S_-YkB5 z=DHVIR4YIdsU?b7wtspw9@XWj`gj{*iMYU6S{@_k%ZVdqC8#dmoj^QqY34=VEcRBq zjrWOH>j9ETEy={vVBG9HWEmZLcZ0C_Tx2XU4arKCb9eLbt^o#7AfC6FeIjGaSBuWz z9YP;m21%rrx5QFeQh5qlDt0=#3d?uS@_r-P!R^xC26IteUi*M}-ZCtRj4g9tH1C6s zv3rL=5~-z%Se|ISz~3PWX=CImEZH$kUA)(j9T3JHXo2UHdTN1q-jXqe+$&qDG--=0 zHztE5QcE*UJceoPs;!IaQr1xQps>VUVk}#gNO#ojLtF8#G?!X}c;3>^g*+pBt8BDD z7K8R6iPWM?ELk1W@G7+)y*nKjmf*{b<-|pD9V_0J(G(qH^9F%<-g0dQ>A=e@KfDWB zI*tHIq?R$n;%N5*ue5*Hc6)@dxL;u`$ET1Eyi18e_&ZnmT|hi=c##pk9$V!zi#*^{b^2sg`&s$#kkZbboxp#jc%R3*CL~6N3 zEM=+=cuwirvW9uWQo&im`p?A2ShRf-?p7bN2E_B0mLBAM@#(3L=PuGVgCtVROJXs< z)5a9lWmK(#&%%;(ovF(-GqQW*$OHX!(J^+R0f^@^gCtT* zS7NdKR<|Lt^xS0JNLY5=U@YF&WXyl=Or3e?7|U}7@w{bx1*yw|)YwbNqO%($ky;Fi zWqkMTc&=}HL1t@VDd8+D4TvRS)^oB0!WR(FTW%$jdHyb`k5?kgqhBD2)M7>~rH)~^ z`!>^mmzJ=k-DK(#dyUK~-KphlkLq&&1c>J?^AD2oSlf#3y^-Z)1V|#atRa@p_P@f= zG1lO)c5h*cNnk8%4amIZGk=#%WQjZp;(3ebKC)v-@wKyqkVP*7B#~M?h^20-<^p~CX{f0n=b)=g|SQT0%c{EoUjdu|4qBwd~{@p`z+bE5yf&8W-su?xuZPhFnJ z(@B@-iOE{psLQkWPLS^Jx;#G~Bwe1XQr_iR zts3d_G`}(OBs!wj1%gE4i29?;b7%WvJXH8l^w|b=d0K>kB>%C?Gjv(%X4K_5D;6aC zv&%DUnm+E$JYdte2pyREp;PFAX=7)xeC~oZwib3vY^?vkP19=cAB;1(T&lGeb$K2+ z3o`xNF3&%H;0fmggHV^J@kNmKpSwJJUsSGrmn8F z@Ltt*6G2YdRl+fVTqOonwycA?JU#A#H2{u%tw9pm|5EKF`Crs#zdV5M1X&8tCJ2k;Eygm^ zjXX6Ct#cN4u6(Kw;(3dKBDoWIyqv0qEd82-BvMNpvG`{U8-XnL$(b{RC6}|@c|+D; z)_0g_fa58`d za%lysOVT_L&s*BnjlfTUD>iI#K^B+AAc@q{kyuO(w6c*!smDTBVJYP-R<#`Qb*yb{ z6Fe)Y-yg*DmYLyXbbcyP^)|9J90rm|ErW^0W}MzMWZ68g&_h@+;CmKRvijK z7VB~l&s)rb$sXR7PhTV=ORt)bfBy)n#fDg-cdW#FtG_MG2o{#3$&4jIm+T>75PA)N zPwBxo5YJnBoh5I42NXWQYp>T-&;0!(q!xE#xjUgL-edKQ)AzH&;+n!(HjW|pO3l1R z=TTk0?*Q?kX1fErHacYwn27m#{AnE(=Q@XPMKV+$%2}*qMPW z<10ZtZ;8$%W0)?xb@7*uzbK^s{t;5kWnwYC@bfOR_@8o06qe-sOkGxXBOTrbwIBUL zmZS?Hp0_;qBX!wn*0>ne#r8T#BDFjwmNe}vxI?1E%`#P3A|5c70g9w9x+ddsH{okf z5YJn>o04m?a_O%EWZAwSB#~MQiAA}d#dTDdl1lSu!s3(4ScV@b=YAKO?w<-fnAG8(FsKgCtUmGqHHsK5)a9gC?znrI52!ekUCg2i3DLAxlXRi03U{A4wO^ z{5N-NA&X)-NFufP6U)jU$(xWRX!^I#!jhWK)Mflv@?1IP&H>!%wQN3!=PlLdlIQ)p zwU56>7R{v~iPREJEMAuVejv-eg+)DuCHe_tv42e7`m}oW!5`IS&Pou^TZ-3_v1NM8 z8pDxAc|AxXwWJWs^z&WuN=WBlcLxcJUj}1&aD5&=U(Q=hAB!ws74Q9y=Pk|dkQGGv zW9oHAmJX^QiPZ9rSSAg1ZiVX7WY4!z!cxIm-i#p~tIPY`twffO2S7Y;3Ajt%#E1UO z!kxs|0znd~HC0%0GMT#UvoynXX*zxRNMx~T0OEOzLN9V1b6r-e z0@X!rB1j^&H1CGTFrxzxbwqWsUZQ6%ED28;i(4Xj8&{^|G8tLiY(YG439CzNLPWG`cqS6QR@lfdCP}ja^Ii(ssw-Q z#&7^gBDIVqmXdDMFQK}qYY(y$mR-*nOTkp~X7Qm%|L(}rL=VLCma-kBn{b1_72c!h z+hCAHYOy4i!G*){e&oYTt2qcu31|7=U>u(>`uCG>BTLsR5YJmqD3Iq$*QM(3(fOiS zyZY}RA+>BKmO|eOJU_Ox>G|!#l9tWX<>U{t67u2PHG`04Lt7BfTPk{z=gQ+o%klnX zDw-gP)UuCQt`2^mjE=F3X=(d}CFVI}dE1k8uIz8GgI7c8=YV+La=JNLZR@{&jvA_q zaz02RwVWZA0o%zpyXKAv3lx@s7mUSpF?s*ucCFevcYr60qvUr++ zBvQ+5Vwv2~bP1|UeD37a!s7Unv3Svaux5|zi7khVKs;|*X;03VP&=(|$ny9rNFuf5 z5X+IXQTfP{7&iQpu;g-ul1MGxh-Im& zVKY>hjkf!8gvIMMXL&|e4m(u4{s>uKZ3gkYrQR0ujJ(X`5$?e29s`m{EhC6!%!|nx z$nwtD{-dyza+Zu0WY>W8Dz>kj;uaEo+Ho+?~uf$YN<; zP}}LRd6?K-#u9d)jFU4DyW;mRp^HE~Z_zkJR;pZjU9%ywc-nv@Qi~_C7&w?bL6-Rw zQq+Xy=v&6pG?+XiADg6szjdQ99mMmNAO57nJL_WI{>bvv93+uiLWpHbL6-t#IoP3V zTVZj1$5;klA}fe`sutnrmwVbEp0~uigyXrs1?Dlh^L~6^kVI;UCzhC`BD}Jz#o1e3 zg(Z)(Xl)`ZA|?e~J%p}fUoV1q-ZJ+QsmnFxmL8}sw%0)tsU?e8=5*QF4>dX+h%N3X zEGOUpZ3o_pXXGlNHF^FQ)Pc7o5oD5d;GO!=4|m|TQ;uGWI`CShfGq#if!F08>A<_~ zr8fa};N_=*bbr@@H{6AE;7!x3r-M519y|l7{!blvuJfBzLmhZW-hhcCt72_*TC9e5o*wv0y|c&#)`ejitVcHnK$ zFvNXxnQnVs(1AJTD@gD+`z&tVFcN1{$-PsCI`FJ2K&F4&f%nG`yj-la1a;t**ZlNb z_|F}9PDV{|2cE;`*Yi*Z-f|_7@LzS{<-Ki=uXq}vimB)-u}Kx=lwBp_^2t@A%>j>P zr~_|jbCBl0b>Lmzpo+h;p|mWm-BJZDc;#RU28#?La(lv3W^m zOvY&F;E~h5%^-=?GM!lb@44fzYkgnYXN|DLePArnw@KHU`K;47kmc0|5YJmGoXPpp z_o6PIVI1xRl1MEpi6tTVDDGPG)_CP8EWw40rT8OxGJm;lX9}`3>I&j{%dcAGscBrm z4*XTk(w-oR)UusewtJZ2FYI`{v)Cmp?jIS;eKj&V56RT~jx2LaKs;}GX-eKMEZq__ z5}hyAzk?)F%Q0fv{XPhDL^d-%H2G znAcd4L~2PRmWOtI)+0-|UW-l(OLQ?~NzozuU}XjNz%`Yx(V zy%8XZ)bf#7M)Wyufh;k33u1)DuY|Fv+K^f7l{Xg~qPpz)1mby%?^p6>(Z^-PePkK& z10<1JYU|+pelvr(cgS*Q;F}x5Qo&iSM6Si>%hR+O`;jGV6o}_7J2c2OSt-2@p8feg z1tgJL+7e6iXFmIo<a@vDEr%jgK+&!)x%pQuiu|=PfCn$;|KWWfSqN-1!8M zL~2<|EK_#5Z$OseBlmn1mR+TcrNWln_Z=HPJBRABbRvl7EiWvIWl*Wg1XPzLrXY#b z;zBHU4=vb^ELFvgzY9wVXW8+EjQPKPDI16^m%oE}-m-WFc~f%eoT3i0Y^>Jg_m7ZT z0*Iw!*zpI*@_LqPP3OO6e$&dBy1dQaijT2_cRkx8i?J1m=Ped}&*JCpK6%@2BFlMe zkVI-ZPb@X70`b@9J2ZS+Usz(kG8W~*WJScrFrS~uGPDJV=PkGMNT*lh-SPMtS+5;P zBDLHnmNf=5@m`)K?rJTBCEy!lNi8M!mzBHgjzX3!ZxGL04(%iNeMOr`c)!?J2SF04 zOw0wsts#14*Qo zYCZ6MfAWwYis*c?@%PsemR!!VdFIW?j70c3d+4w6VM zEr?}J+n!&L<@LI(0m72>gQ<({6|(9k{?3PKs4hX?AfC4@I85G>n(w=Xzpk|{2qckO zdJ&86mi0H0Wpe+L5yBGwld&u}nSrlkj6I|35!=bWBK7u-p0-CGtdcHJ`V@+yk)H;8C&YrIKCBG`i})kq?QH5 z^00Z=SY&y&Nu-va#1a{6?u0CE)Asud zOR{6VKi|L9Jxo@1%`6DthwAch4v6P1jpmWnwr>`VpM)%ftw0i~rI9Wk!z}i@fyZMO z#;zxXCBliZsFjd?u;%9kyCTb&1t6Zc1Z$F2YI9?o&qtP+Ga!l7(w$g51E$tNb(!^K zY^1RGI5U=8^T~TLn=}4fkfqy85YJnzN0C(oR$2()1GPz+3C^RUg$QM-?QIT0)6sN0?a^s!P&B=qgG)dXspT25oK@&K4%MYBreBq?q;6yCQh0hQ zKKGXwOvCg1O>csD-lBPhj4eyUJnZ&RBvQ zlKJG|gt7ssE=_-cc-}H*An8``T4@o2>e8-y;qMlgufpuYHThFl{P85~*biyk* zeiMh2S@P!bkq)SvP%{=}l5`VpxTK4_2|u_VuA9*LCg~=O3!i0FDgZ zngbGvBkGTC!jA@|n{Y`zV_VcsX!af?`H$U%#~)prkGcunY83oFuKw&M%>OVJ_W-u< z*Ru&aFctGbg1_k|G+r?QXF6bVv_9%4v@8ah{%tqmA3xCN)Q_5|o6zSQNc+#-grEDi z#odG@Q+m8W-GmBNAmP92CJd?Ujo)r4&Qn!JS3HASAAXlXdX<<#t`eg@P0B{ygbDRP zn*Y{Kc-r*C8FclG2(A46lm6%?{7{{I!L52Pt^4R$KYI?u^HdQV0h&wowsjx2dU zKoY5C6{+#aVaxDWH4XRl7$+S=C$d+;DrGgi3;MY+AfC6}P9#r&!*hG#x1{Z-fFx2& z5wVQXvcp}T+j>pfCM*G7jHP5NIbWQPG&VxV*ff0*&s)Y?l4ER~aR}ZA>$3q!BDK`% zg`YG`^Gu(kbAL zq!tZgdGsjp8@AlZJ}fM`oMnqT>B1RMv;u!+BkU1~=PijAq%N~@%De~)dL0H1~GM3L>N!OZJw$n#snSTt#^Oi>z`|;Q=Z(QRB$Wj^#l1MGfh~=F_ zUEH;HaD7I+uz2|}mXZjvU+nbAL_CY_?hN92OU(nMYi-k#`x}s@+YXRKYH=l&sF^Nm zs4nUEby9?-l(XnXkgo-8up4(B)#Y{ui03U1zGSx4;NBv9>*Zd zaEn2i!jidT)6f(Cnq3F>(d~fw_Hjl9eAayE@I1~3m}Qql1D5G->tHcCE9e~ z4`FfjV=Q5zWRKPEXX?B}mY;J$Ja3tqLiQ=CHB0#zs!R4FkVI;!-W%Wdi)M7h`(Hk8 zY^UJ**KBDXXW93fbQ5kl9FM!A72AM#-r_NZtXH0)UxHT=Ebj!8NG&ai#rnhRd#Enu zZNgN9CHVkT7uT2M8M*HMXUeE9&9p!~Z@D?2JipA=I@21}rGqX=BDM4;mNp6Jrz1<5 zYg$WTi8#nuN(Ydc-$7blZphNv1H|)|+Kb2+cHHgXFGH5~z95OzGM-qXTAXWwEH%=9 zbP^UHf5vjng{&eNn(vACi>><-#PgO04au9uSHo=`k>$fXkVI;+A{OJtC%Vt#uA4siE40a7KoY5C3$ZksdeIiuI3;u*Q`FKG~LgDit|K|F8qJxazf-CC;mMaS6q{ve6e5%*UE@nA7;Lga}u^^tegnAOou{B2qBFoVPkVI;EO)ORk??RBJ>cE~w!cxdt28|%s ztbj z%hLtA$TH#?i03U4!^m6G{nLKwqPpn30ZF75BVwt3RR!-AUF*Arx3E-jmSIt3?dA+8{ z(tA0G=PeTslJmuN&=UMzg{;*eiPW-(SR7kK<72GD)`3yN5^|ie#15N@$1n%ocj9kD z4fO=^yd`WSnd{4%T6zW5W#WF2L~02mmS!t^YoTMTU$Ylig=N_wL5_MaaMOWH}ME?)Cy;kpD4T-zVj#la24 z^Oi?-eDR#pCe>!Ro3OqINFueA5zC6}8>=JB+s~(82un-|V`*SOI`31LJ~2g>$(10U zx7cQr=gNo%2fLzU%&GR7-#`XPwaD4mQ6fmkg>L&a(6r}sRZo;q} z(oML{zzOe7_;D0S^?&Ln+#RHrfVv4CCV`alZo*gEq?_=X`sGk`)e15NiNq20M>nD6 zaMDfKez?VP)Jjg*Q-R-O=$iZWcs(=gn#_N0Wo9Nqi(`lKSA1m?j{Tj?1H-qcP_5K8g&zz zRkQjn{#V_E13aeUyXy0}{rGL5eI1Zfc9j@*fm|iR(hAq3Zo)AQK$`#7O}MvHP5d;w zvS3^vam6X+uAYyQNeAB7%zOANfsMX^c>d}cdXKzWY|{S8TXgl*F9%6v|4X&2d zUvrphU}_>PM{vsd%uluqzK-g$YXL|iwG1Q{o7(qV zqq=mdJ$+9mh6SoLMZ}Urg?yP-lDBTI#&0q=Nlo*DPNF8YB@|Sb?e#TFHt@o|1D5h zvd=Sh>258{YI{1nt9E}AmqxW_P- z6Sc_LvPS0t>Byp<0pfYf>0zWUZw_}^jVzHbKoY5?Mn8PtPwhQm0pZfIofDcVEUA~6x@=fP z&i&*qt@2P^E-V7^yrn@Psf*fpYdp3*;RupQEq#b(Z%(B)vW&3S&lQ&F%Zz2hHPThE zBw=lJWQkq};(5#B-NZ80$Pa&!%EkdCky<7YOOp_N2V~KjZ&oZUepeXFT>7Jc~VPEl47@EF_knN0*O6mPI2j{1lc7&eF=Bblx``G#>Bi zRd*tY=Pd>bq@zf6;Zgi07sda5eEt9b5mJi-v5frc`3YGZN1m$X_SejBPApTG*@|T5 zck3bFKIj;`H3`J?mTk?*xxXOb<`{JD-qCQByjQ{AS0J9Z>{KAnFUMvtz;73n^Fb1+C6ZWv1}WpcJvI!rP#2bvYm8;hP4YJG zSohpUs4iinK|F6cA4*mcG}>_luT<$h86=Tf?h?!4C_}ur$J7lSIt$CL>x`xU0y0h> zSlbnMN4YzLc-|5oPiExfb*ym*UXPt1iPZ9jSVrZ1y^M~rmMu5v3QGxR393!rE=+S+ zjXNVZbO7bWd8W)Y~a>+Pgq_D)qGZyPwq^m$7 zRy__`wwi%>-f|?6++SL~wZUJXudx6mky>v%dQR&+ab%v2c=fRlFL~p8IiH&9QS8<9%fl4 zi03Ui!^!?+uUBPPqVuI|4oD)kY#EMB)5%kkG_Jl1n%>qumIZV2Le z%V0ZljA;(B?TPAgdK^e1wImRWwoZHe)#e?ox9t&@QqJ=AJn8VRqg-K)EMq%^c;1o_ zMC#()&U6*BOwj>Jq?YHz@}&4W?pU2>?HnL1nYWp`s7x@$uZOpL|C);|1J{9g-qPEI zbVpt6X{nAZvm8JYspTuNn0b%D*W`hB7o8NA*d)f{wU*4oB(=81dq)|kgLvMeYe05$ z*sU`cujnp(0g^~9Dg*F+-=dxwUTHtG>G*TPa`X;kDNrN(UKfpM7mw<)LIcF}mIIH- zb*xol#x+!z0a_r5)S^W!4&SEY?x^8i%VUMb^)6$%zM0G^?KT>sfGnQ9K|F66SwuP` z8`QmszjEG7A0&}lh7(I4AJuqtO}_u?$!%fD<196rl6^@N=KA4|Rf7*8p11THO2)~S z5m{NNF2-L#5~*btvD|vz&<0u0#NG+?0Wq{=t{GF?rdNZF3OGGkb@xDan`G@We+M07EfZQb5tiUo#?tsB*<;l?{){cMTz?4Sc}wOFQkUaqNn~%r7a)n$ zQgneri~B>ylCXq~$8 zr9H81>;J71HK0A3YT8v;F8t5(H=U6I&Sa)+T%qAy)ESv#1~N%HBTro+oskop`6r>y zNZkb>%RhBS&ekBEkvWnOL2TA^8XJq3` zqdTI`$gSrmI!dSJ4hkqU-$a3+;r9qmzPq{%pt>ECun z{_z8=H=kb%bw<{o3ex^_XJi8x9o!lD+*$t!>Wp-o2@?LR&dAk9cDSRdX8XSF(baqV zJdjg%m2lK1SBYJ-GM1vw$oY#vn*Y`r*k$&*I*#YXt+ zCl)I}Jb(3E*My8VKK04@h5na@Yd{j&|58n#{4bvrEsh|IU-Queg(dY7b0_d_PG(G2 z6i>l#;$u62c;51{7x_YH;j~!_$g-$ANFudNBbIx2UU?%++p&5hg(dniW2y0p+@ZT} z>AC`4gZ7N*_dA}q^d3M)w2hDK!uw!d9uJa8Eh~uS`gt?FPf4ftdXt64FO9KW9!>Tp zoKovp9b{?S9mMmNhO^1cZhg1o?QCJG;4BTEl9}IP z-EmEjWvx4i=Pidjl8&Mqd!q1O1v|Vz5~<|~u}sWYn2Ich)ngY5OHMjdm+V`l+hems z3u+*V)DlB1u_tZuOjfTMhgJzo!V|`FsU{hnpD|go z3|WTH2l2c`w+|VezkG2S?}ssADM%uy_eMnf+o-%b=tV+Hbu&hS}-eYxE zAc*HJ`%_3=x}S`QL3QaG3X(`IeTk*BZ+J^&$=jM8Dl9QsjAh(V@~rf9Qw81wPsa-)VwY>gKd$85%8zMEK-FSdJw&KL7C5YJoY zB#}MvoY%C)$CyPWNFud35=)d#jxMT;s_n=WVaeqz-u}emSm$jbvUJ=8;(1G6FuA`x zdT5=3jztokVI-ZM=WN()$y#HqP}acu!O&0EYp+c;m2&<-J|h(<=rDdJa6gQgItr# z`!7;Omi`k!5~<}LvFNROItbN8{ZRiBVexv&SemRL&o2(^ex{&f%=!d~=Ph+tkS?54 z3FGlI>eMqJiPVxyEJicamLbcm#rhS(Qp#DnW|944uH4Ya-5x_%fq34spc)w`zgRbY z2&zlX4IqisQb{a92D&?urPG76itc~S{AT7bb?Mz`8a~F3-!jDeF09T3@w_FX1sRVy zTYgGMmMu9TiPX|`5FW!g4Sn_!S#Ew&Ya%SMuNcdZt>js(UqB1IAI7Z;5YJlbde_UL@6`Uv8A%jZ4h z8Tpp&R(#(d^c5tLT1FGg6ThW+9wu_f8XaMAeZyFK?IYb$E5>RjAj^xrAfC5a94F&3 zqpKZD&@ndFA0&}lEQqD>Vcp^A7;ESA*lB#~P75{rV>wlY+g*o_WTg(c!GV_9&Aocq=vADAG^Cl?UUTeKFCaq_?; zWroNSyb~mmTEdBC^OC+A$Z~LV^?AbL^Nz7>Zc4s%Y&hoRcw{-f6~yxvzbrCu*>k1E zb7ZmF4U$MLiNxaaA;uES%+*y!B zYPn7<{;TKUubu9Ro?I*}IYmrerY$GqF<)a%yeF~dH4x8RbSlaDqO#Bh@8r_v7Dyts zWD?8DZISI!U5ZZ^R0vB#F=Lq;Pv*yVZ7lkP>S7ZE;(5#H_v9Fh{50@9s!M1bNFud- zB9?=_igAbc(9^S&w*U3KA5y|t29=Wi{IqXmw?~%Y7eG93`B_FRH8kAsV9QmIL~2nQ zg75o_zc$~6>Jn8lwwbW(`ovf|bR>1rcPbr&sRx$-~D-*iWXtB^M_4Px*6qwc7ag&>opJId@_XS{D*mf?nG zs5@#|Dai6q-BCT7k?yE#Q7ZT=wC~G7y1(m=a;~=nk8sS(>ZzdasH_@YfB#eer|zf) z*AFZ~-BFE{KuURc)Ozi9__x`NRB%B@l&dO8B#x**x}!c!D#l4(Oh4_7x}#b)2TA^8 zchtF7i>y$0)Z-o?*`M7}dneAvJx%AWf=khXIl5=J-+%Yte1%qTpcT$EzO~jU)Ezan zKgjfNyQBX2fhAGPT~T*b&0!$zKX*s<88`rUM_sb8t%bUy&KZG(|EfDm*KG%WI=E|R ziLVkl6F^ScRU%G@TqS-EXrYF>qdcd9H2r;`3SB*ICj@DWD^3Y_^}Mf2*7$Z^ z9*5V&zc2;y{MB<*G@1ELnWK@6uAY|WAc^dMsWyoGFW%Wbe<4enQyqO_N&C#)30wo5 z@F_R>#>qf*-p5#LVB=v!cL_`QH^$<6jC=th`^C$l z=os5H2E_B0hp)(NX@Fus{G>5#3P>Wgd?1!vX6O7+T})m&9uO9b`X%G+t;*;0qe)lUjbDQB_iMb`P<*)<2R zjf1tH6- zP!P{sbZ(RJSW)dzJd6D*3M7$Q788q2r?^wdqF3uqmaw>1FqYh0^0lDw3s3PL5-KVn zp11URNY4Ewu37g`UCuWJNu(AhVwpen@K9vA|Kk38Vaekx=EKNWoYs9$!Tb6sZ%tcWVc;$THOya%c#;I6f@MIfHH zY!4zcSs`bh#v#ik8<0e5NhX#VcTY}2b?FkF*Fac&eleDuDP(L}KhV+~o%^4zf_UE2 z`7OC7KbhMd?@jnU5hRgX-V#gg(Sv6p%Z0-_t%T(}XE7T{*7ybmu9=N2rwTwkZ%OJ& zp2bSnDCwiRRDA(Sq?RgTIaO+c?=N#>hIA2@Y)9p;?UdRoDJ-{Hz1l!YvASBd_6lE5 zk+s*U&72F6iRIkDC&tJUeFr3wTAB^RV;Ik-+rv>^Vhg_X6_z+B#-gM{ zW`1vMSZ;*-ck@w&X=YAOAa7Q-CrPy)S^o)=Ehg?UeO-j4-AAQ*qO1I=acc6 z?+d7C~-SlnG0OVDCsS#~mG z1F|F)fq33B!HbOfe-t&`fh_&Mfh1DPd}8s`j_-yn4XoCi2}>bonY@>DdR;UuR7IBF z)*zm@R97J5WTPR24kF9AwIGSqvWZx3mz}}iB3P2-vO-u=U75O!b|;-)x1xIEFY8%8 z1o6COTNxR{m>wT?r&OT?Kwf70iBIC>Zj`Q*AjxDwzp0}t^AkWA*;}l(x#e6?VBDI_+mJpBn zGmz!y$G%8pM5;rX$eFF-tRsr!|@CH3__G7VWu-+&}i%ROSrc;S_g zELH`VjtWZ^XDJy;RuFv~reBKAm%im7p0|uRLGH)e&r8Hz(KUX8BvQ*OVrexl^8&Jb zFFfTdEZKXRx|DSw^U2Bzmz9vEUO9;8EzaR&d@1dJ%dr^jV1T|ch)|{d#xoZfq35XZV7o7yF4smEwZFE1xciq#$)lgAF)g7&sK|F66*o&MmYM+BmQC(K4f+SMQIAZzVGFLoScDfUi zCM?CATv--g~iL6 zvE;5I_b@hVv+y(W3vCe3Tk=0t{XJBy_I>rVYfOy^#mrKr{#^J6tzjv)#2uluU zIUG&iS+qGi56_S3SAlrma!ZB07n^*iA|BOcLygwIe}vSMK`cuKxQC+#f$}N)y9i6w z|1AHdb7k|4)(WU|C2uy!B{;6}txMB=Gou{5Ty8`X& zrKocy%m}3VKXtBL8xy)8b*_||ft2#jm8sqxFf;1JaXhP^ zYY7sGBkGUNmAQvW=gQ*GR$Wo&it~1mLQ{?BwLw>jX1*Y&^eXX+TqTrts?S25D?0;0n*Y|h zGW@OK4Kxy5jD4yut~hZAnUNsoHW>+)Oy{Pg;U3rT0KQZ<)P@+&Oi5(eECz9DfLs zNG%A`Bk>$t`Q(a*x<17KiWG2hbvy(ouB#r^`yv3;xsY{UEZTv;5)>A+dsbv(g z9BWnc1+vV^8NO6l(hoCrX>ylX!Y=KcjV#F@K|F6sx=LmkYxK=EM|G)L36e-H^N6MN z$u8XK^+;orsjyr>!dMillRKw`nlbf|<;E}&&s)kh$X#HqZSOuJ%fzuDiPU07ES+|o z-iGQjp~T->SbSU=%k39r=J(Xjr+DpktM4G5w@mIzx^T{EkH))gXH`G_`$tGE4#d)I zx}pxUteyDNL0IgMG8TshWG3rg$E$da?~a)up0_+nAv5wmEk@(r>DTCiBvQ*MVlmp% zbt|$c44vyHEcu*e>{ilExUWhdcL8_M0`a`1t~#+?ske9sI>v5I0ZF8mC}R2gBC`Rq zOqll8S6C9=n7ZhHCVNM@Hu-?}T01fa#Pb$YWwP7$x39}j&Ks7FQ3(av_kcCkyPSVS%n=6YM}dZ*dq#W*OSfo4Oa(CC&{bky>hw!{dI9BOCD- zsmxaC#S2R%XK{N;x(XDl<$OmLA2SfoTQW__)9m;$GjV5RZ%dFwYH3F-Mh(^R+G`t4 z^)zA0JjT?;$dG)A(qqIy{AO{&WDw6=DpX0A=cEhM=cBr8o&}OfEklS!d(K(heOveJ zksM)(KF(M)wP)k=#o_ab&8RL@!$3T5QT{-lALb?uoQo{2qd*d=MTb~=O}dZ25_l{s z@3XM@doq^L8RV_c)JU14^^A?i}sty&37 zvNuzgZ$;$o!T`hm#^`*pwg>UNWv?Q+Pu`#Y?LDeXU1yL)Y6&NnK8w;uA!FP-Rd2!u2#d!_ z#xk&s)a62>o_KE$#Uc>TTZ|5n=lveOtMNX*<)t8r)RIdqQ(C^k&tf^HUB(Ja6=!L# zN!Bdtg(%`TB{4rhJa5^wjXYP{q?>1;W9)d%uD^eT)bfj1Mn;6=J-ys7q)Zo-nGJh$Mcp&gUOo`zl+CbAWQoOAc@q{k61M9yzt()D;3_X7M4I? z#`2;M>53k7ZNLw7?!T-H;(5!5tE4VVW-h@U-fJ6!BvQ*%VtI31^Ef*9dnzv7BrHyT zj3r0O3E#uGWUd;8EUT43Ja1X)bpx;7yRs$|Kfjna1xciq<-{_f<|e#~V9v0?cEVE3 zSt9eu+qk|19-ASHkrs&OE!s=TSn2V$3+^ZyI29z3TDB3(y24Lcw#ss2n| zEMrJlLAhfU{?hTFa1hU1=7*ANa_TD6RCK;HxDJv?Ek}uE(&#~rk)`OSnWwNso@OkO z24wD{ZLZ-rRF^ezAfC5$j3bt9E&Jo^SVl5PBDGu~7KaHY_mE}z(!xMt@d{upVVU#s z`C|IXF9lg#j6pnaX;GV;`+GeXTA;e@GzUqfmixpqEo<{~WSL$*FG5&s&oCDCQ)J#U z$VaOwvONC<;(5#0_oOa0eebIw%j?>Dzkh_(l1(fvEhgI{OJ~3RcZ4N}v$*7wJ&GCx z-NK#s`!0fb-r^fY_FEb{YxWRivAzb9NG)H9Wz(=%cz$fXnpd*0#04^SQOzU!E4$t9 zk%;QDt1pP>ElQio{Ftrdim}LYV<<=>wWv(MQdok|G8XH1q+7kp z;vn7&$EzQR=PhAoWKZH_?>zDQmzBdn5~-yZu^if?S`%633|DwBEG|Ke#b5-fOZz2n z@Vlg2<3K!b2{}wUBUhD2;jXnE(?AlbWjwKTGb`DMEYI%WsSuVj&a&t@x#v&q(xxXm z#$uB}Ja17mA?JS8vpuiTF=n3*l1MF!h$U<2t&7Ow;c!IJ`LAc>^m9yIa_f^hSPk>G z_}p)F7{v1y!wO%mRMJEpM2%Uz#vo? z{bnGZw}ek2_b_Ucdf>W5wgX9|mV?C7u4e0>s6jx-RHMDHeEXl}Z#q|uipbqeyA=n+ zQRhm3OOQ#@xf0ZZbgo#Q?yG@1R|2+!EdSKG^0Wrob7jQJbv;q%%Je-T-QRVtBn%*( zE1z#2C_~uP|EY7O&{6#~>Ri!13R22DS9H}#=Zfv!f^>95IeURb;)wdAb7k~g z(z&9Yeta+LTqFie#?p&#TFY68JTxoL| zB>Y#MD>E99w@#lfuE*aHSHBK&%B~W&C&*R8y3d<<)VWf08>IPfohw6~>*GGzD{c;* z#1+TqJahHz*d0c6 z{duymB!)0`v9TeRqBqatP+eXYfOy`r%+4Lpn0Pu||B5WmpFt9-rIc8lzjnfRfo(2U z&J~u>P{va871%Mz zYbq?2oJH{v*`+q2Qd(vd9%24*h_r>@oEr==PgZb$aQRc*B`i>aN#JBL~79^mLoR#cwe8fJQCsl+nJTNkfcbl#D6TUf%c zFqX1RavkeCBo%k8?sow3yybl=IbR+bD=k2lH7+2D)bfE?^3qnnLv>jadi|lWctkRm zHJW7Rclq-P5vVSsqCh-vd1_Dg^?Bl7Vvj6yZ-XRKORY(G+*iD{=Ml2_Du+B5mMYG2 zbchZ<_wBnT;;+$NZVuvkOQAcdi`K#=cosXUJxC(8v?rFXrN?R@OS6iiTw%$+%GBk} z9VPCt%TQuzKJ5kGH||A~f)ZhgjbbeQ zrxS~EX|1m480**;#PgP;@uZuu`u;%tESB2^B#~NX5sSK-3;u%Jq};l{geCAAWBKGo z&XYO7oj&s(~jA#dXQ-Z|=wELtBy5~*buvDkK7JrtcUvtr7d3rjI) zDQ-);2`3FoG(g8#OgxC^ElTO+?Si*n{&i$o@B}20TD*v*;*nE5WN|La>mn?vH<-F~ znNGSqOTq@>4)1y8AfC6BXOK6Gb?0rVK$cEFKoY6tGO@(1xz-L@&eYE8D=d*W8H?A_ zE%=)JWlUC0RF@BpKs;~hszog4Yw9dPma8p55~<}OvFxlr_Byh>co{oVSiEjAmTsox ztxwZYp?8oaDG$W+mYJ`~^NUxno=cHsSTRTzkmYIoiOIrZ8_ihKE67_P zU8DPpkmY3xi03Vxo{_Jd8}-Y>b(!@7B#~MaCgXA6(QO`HL6l>Be2%c>a2CBbWbddm z-&fZD`!{pO|kOJN6aIMDqzSTQ zrh<6hGU*Drx19Xysy3>NOD0Gnwb&4g)xl7_SHY%BwOxcIJ(j7ZPFw4xUgKm%UI?$BG1V4R!rQDEI$1~ zJZ~vlK#np0xdnLNg>l0{5~(GGSUfMSi$}+phN?|~u=w0#EY;%4JBwB;2jErmU)q6q z-ZC$iocoOe^6H?v_;v$Hq?R~h$u`#1K$b4uj$9HJ`}>ULwlcYn1@GzXjOx;=FNo(Y zW9N};^6h?k_mRbM7)T_xj`z@s;rQtnRA6tNU-V)G` zJR@HyXo1g{F55v8spT87e15En=f|E6>YpSmiE&I_rstD=7MI3f#e1z?9|Pif%Zohn z&O$BBP!ZK7XfjA5wKSQ6?_tjB?`w+cQdoV~b72X6z*w{u$yb{nyDVvjj6>XHyb#>&_l$G#$qt~H3~EuCi& zOC6hCuaKqO9wd=kEQv*T-!QzFXMm?pLt%+dWGqHWWFBVy&;9sa((|(*p0|{lk#BeC zOmM^Z$-gdwBvOkDu@pTxUmG={+*_mDR#?*hXZf4%sPd&`&3)PC#4V^hD$g8bl5|IP zH|vP^_3@gv)*f|7E!YaO{8M+-%z`oa5xz#(R$-_+%GM5~`@8O_q3WbNs$zeQYN$KP z+5x2cKXpfqeztcK>W(^j5TulMM>&oq-BIVw-nT|)$3%CKNE}grbVpgNBi&I&5v#VL z?x=U(AjyC1j!N#Y(+G7(jlTnu{n;HA7iNWfnufj}T!aqHsSoGS1C#BJ%4}eRGby(^ z6@1P#}C}$GJ6o}j%t<((*ARI)S=ipxI1df3#-PcJ8J(Mknmr1N1beT z3wHuIwDiS$SlASRoU*G#^igt^m}ZdI8+AurD+X!)TX$6RgJbbp()b~hx```}e-d-` z%uXgFbiVe=Q|MT)KNiIESI;zSG7=;n*&K&P0;kC!iR^z-I7|MQul`Lpp;0YwoxuQM zad^mBlADmbz`Zwhh9gVqW)RO?u9cFf*+{2fcwMa7c92ABxlJt9EmQE@h3Sox#|TRi zX9=<=`;@dxC_jKKeOiNf-ZFk4vFOis4nlS5-326(T3!%KsQ#LMs4lK+lcx(y@*}1$ z#}3)zVdFs)4-I6A+yvrzOIZw=k)PbSl?}4&+Xj+IEoH<~I7;g%vRwJ_a)GdfCo>kO z%jBu4&bo8>iyKO}*$79Cw zNs+9JJ-elWHL6Q-EQseVSJ#jyjp7$moKRhC5ge*p2k#jhvn^t$BvYb>(-C{RCBg^m*wNPPkddgUA7m%l>u@kOd zKy_&z1>$*2*8nnG`Z8wHYgCsXF(8T5;zulglit)vmK~~9H-x2_vpio!#+Rf7zYoar zpgoA^Eze$(CqS+6gg3}?NDU;BTCNj|*_jNy{!+J2K!UKOK4a>lUP#85R*x&1Ba3xo z5YJl%8IT#q9sbJAkY#BrkVI;EN-W2k-0?#D?wfQk#@$gfx9F4#i){vDDQZgIF5Ecy3V;2?%nro! zmN(DIb*#VRKD>%x(LRtwYUw~M^=h{&LUjpS(NV$WubHeI&eHS^Sx27Lzr`wajI}Za z@w~;mE}3EUx#C$9oiEQVKoY5C7_l^2@Cd*0tuS1mEG%))nY#F%A?N;l6>Yr!(!w0X z^Oljv$o*KYiz(U2GJP*dBDKsWmZ-EtS5RHPn7?Z&EWs}r%aC8BbLG+gWsgu@nrs8{ zyrs&P+>gyV;)b8a+S`L9Qp;Lm@tNo!g)A4gJ9ZNmmrTYo(VD!$db@cRUVmxr0OEN| zbA9rrgk#hzHcdldO0i_&}b0m4$oSw@zT)lixKoyViPn3aNf-tx+w z?0@M~o`uK#vTq=X)Z$GnWvniLaTu9MK`q`%wmVc;0e)CWz-P8yb@RFuGaX zKZq=aZ$T2NWdN~+9_wR{EM2sW4hT!=8^*GvjC2#8-gtB-I>vfU1M$42vLhMyzxCbI z4Oxn3gCtUmHnCh?H*YVp+zmeEDJ-sU8Ow@=q|5V}v&&9oIUWJxdCL(Kve%lag-0&3 z+`I{rNG&UgzZDDmjaCDS6|2II8($WLdrp#Pb&QM`YEFYU&8QuaBBB zNFudtCzeM)s_;Gkm1DM-g(WkGsY{2p*{gdCPoT^89k|d>DSNd>aIk zNGzxNG)%O z#jj{5uFJuyruo8B#90=9CeM}Uu6@AA*u>hZzvFq!lNn_0;!u+l>rh<`8-gTK%Xeb& z_@<2ay!XzqC>NIGe5Nj2tjT@yr?08~(J@x_2E_B0zDDHUG9`3OA7pv_9wd=knoh^} zFtGLA6~#%?wufGrRFM)&B#*U8pQLKM`Ovj?|44sE3!E41WBZp z-o$e5ikSkkJQ?<(p|E(oXDq7kNcZj8RnDuC<;pk^&s+AMAuH~7D?h_4hYhBIBvQ*n zVu@~gIu$ht)S7y@rLcJZ&+<2&E7!e9=Zeq6BTZ1}in1=qBvuTv;~~B>Y#ME5}zPt4GkW+S* zh^AKw)lR7cQ0GefsUXdN>s--RJJ}XpJr5+o8Zyb5Ec=plTWOVghoC@Ok ztLM{RWKpJ6Z0I-}ri0IIR$tV$Slxg4_kRG@tMjS^O28f5-C{ zofPuc$KzL7J5-nTbwCoSC6idLruNoDb?JG2)&^lo{lwJe#(VPC$2;Q_UVk|w0>tx{ zH8JEa(6sQi0kZtK0g^~9<-`)~))0Tkcft4u+l3{vn6ZRzB7No0EgIw9wr|b^@w}y@ zGnvV1zU{$mWO1Jll1MELX5c4{t7kGYQC(WjNpchxug{ESe(nZ*zOnvO zpmyXsw%Tj*3RIWLP9TZYGKN@sj5vh9e3W;;j=!+vaF*ja`nWC=U;OHZ>N4Xoi03Uo z`;$E+1{5~XL>7k(kVIT;nonK4PP^$4%;Z)gYNdCP^} z$ML=0rGe`mkmafaNFueE6HCSCQzy~+vfHiYO<@TxV=VoykT)fx%SJCpb!qMc;(1F! z5AvoY!8oV{S?UIYBvQ+MVwpQ)!AxW+Nj@1bEH34YM|+spST-R1_a_K^Bdsk#B`1y@ILB;pS8Exo=fh-vn9av<30JrN+7=_rvY?rM^8e^BB2hR8B;42b6~hL6db#et?S zxGwi6gCtVRCt~SvpDSgz3fAuK+XjKzE`xlcaa-10cG6s-gCyk*1$a&LLvu#O?B z%K;0JL~5xs6Oa4V_S)ffeg`+St9SUX8AkiBj74c6nZN3OzUB&K@wEZ*yye7C@^;~g z?ZGf)F;{K>`$tGE9f?J$)pgvhzTEo zIulv+2Y@6}i!QPFU)cW!SuC6G(-4->?~FyM5$TM)tY2pcvOH-E;(5#TByzq?tz&f! zS*~;iNu(ANVwrJ#FaA2Pp2wJx!s1%RSmJfaN|m7V$Lb-=P+t(wTWaT%d&_24xp;lQ zLJ&wIwd^LA8l5iSuIP<@LnaGLC1+W*kaR_FEejckEGLJ7c;2#eFBvQAp1Ruz)n&n0 zkVI-ZK`eJxE!c;SvF0iMbA=`I2UC~i2+|?3$}>L=S$r}aDlEx%OkFNMB+oBq8Gd+uf5#3W zp0}uHlk1pvf+z0qzNQ9}NG$`2rM9jM-b13m%k-46gzsc59-d@;x!Y?Eu1mWB5YJm8 zHjr+@Q}rr)qPko^50XeN(~0G!#vuG97n37f&I^mjF2)k}ip*PjKWlsuo%;sfAfC6p zTTjNyZEqBH=d!{bl73AKsNv~c{k!8mo5YJnljv)8R zU2GcNMV6iiKoY6NgIJ93KOTWBPIg&Oge7)2V_6?f_9!}edSM{4R5^lp-qNBs8DA0u z4&c4k&K(9xq?U`svQhUbUJcbF(cqP^1nyxh8`j(5v9hO6n`6jw|0#&)Ew=5*O33?_ z1@@>e<}X1KsU@CR42MLhpt>mEy;LMDPJ0>4=KW+o*|zbmeaMoV1mbzi1bZ^RoH=c= z9a*kD1xciqx5Uy!@8B(Dd7``btFRPvmYTWbxzc{yX#D)*y9mVd7O!CP?)BN*BlxSD zPAfqYsilfonqIGl_a;>Is8#F8U(frg4oqFVG)Y~a6%8&#bum5#;(5!&p5)v=kP;P) z>N4akNFueU&cgRFohmH%qh<&7w6)EIW#2yKzw5$jwU*3cbTzi#kGgPPyaJgdT{s!V zWdE4E+Q+)0E}X-8Aj>~>;ar$Rx^U*+-QHe+@=j~?Fh11pZ;VjgJqgDY@ z{hzvU+?_|7qAncepCF~Y3+K992Yl5^?mnO;I-;ysS`-{XBUq0X?xr|Gh)`rHt4`i@Lm6#^#9w#*e)D@t35bV zneX?ns0*j=S&-@9cH#W-1FsdIT#mYMtU^KBf9}HR9lRQM;mlQ85P-UH23-LO|5X>x zr^GD$)RtvuWH|G)PZu@!)M7;}k;ir(K$hBu zN3?|{&WWkZfnKENcUM}+N_4(ht^x79<)IO=RIHnd?*f;YgCtVRVPY}tS2++_baaB| z3rnywV=-Mvp3Hm9uV{oUn))D~w=~NnZ?JlvTX_oAWzS;mJ=6A=gN(wpZM;4>j4nYTLyP0*Rk;qYmXs|of}9ZwZsxj>MN}rRF~z-D>n#B z8E3iKhrIC}wx+lO)x~Gls^9UvWr80$_r2YQd_k5k3qTU7^kv1NFud#CzcbVo|Pj@X1R@@u-G4F zES}rQ+qgqNdgC|ofj>b!Z!ud;>XJQEr#?Dg^lEMY{UfB7vBc88T_3!MM6-ckLxd%t zv+PSG_b{KQjLATC=`jk#^A^nv@|INF=xse@*)$#`ky;FhCCW<)cSuyeRlP1OiAR{a z^jk`w%**u#BTGpUi03VCW#l@hzrPLMZQGz6B#~M+5=*&p1>Vv9vE`OHVF`6*EV-rR z9%jpz?YMJgqaBFnEu*)S>sYm0*7$sRxDO?I$s(cn)OszT#qsquWa%) zx&+TIzNjv7$3Z-Axj2e^S#MLa|8jJUjqwFZq?XgfqA+g+zUNOmwDGmDRB{%xfE{>z z>1I%*iY(5aAfC4*HYV$03u}hsnXCqWAc@p+lUQ0cGkJyT(%19m2Vu!{W9m|wMP_~r zZz|$0oQJ1DJa3uwl$uMk`lbPhl?*rBDK`j#pAwX*X7&MF*aoVn#RK7aE!BPld&@1 z`W60Cn28mL=Pe)9$UXnQkH3zhy8N^SNu-ue#NvIS8Q!a4shL4rVJYG)e$&VtX*;)| z=jhyDzXinembzESJ&dXL$0NwH+7={{T1F7d)(m?uR2R306MG0t@^Pjv6T6XP%)0wD zEo5;?1M$4&&IEFdRi8A!2D03K1(HZDbBJYP;|~pxB~|_1AYlpjWGvH1k#oOMi%>ki z6b=IMyd}ksjFq-lPjJ`T$x$GQ)Uu9PhQ`~eBFpV*FUJathZkdU@wdX`%f_`UZlk)) zp9|u7i%Mg%B4WVjp?Gyi#zK%pYS}|9Qz!1ldwN~_U@${isyIu6CV75IYi&LeS%!WA z@w{bk!38`A>f&CCyFC4Wf+SMQNn)wI`mzzKi&>-H2Evl<&D3SzZu0i}^8$Um4_16j z5YJnt^&BS<2(L=ua(@iRQvSMjOIYGH{z!C2bVCtsv`HAo9z$MnB~ zc;2#n2H9UZK`rq+vQ#NF{rw}PmdC`hM$a41_06&T8E6$-~%xG&vTVZkXVJy!(lDYA_g(`;Ve5qv* z;(5z}m*m{n2-S%}mdQ>aiPWMv2ao&Jw&mKNpY-twCM z$J5ASlmwDUEw;p>?QFCTSqjhF+!vM{&Y~Ga<}OYvw8=%5%t#Q=TdwUV?<`iWvYLl1 zHSd5VQp+)7`PO1uFJw{9K9wRYai^KO`1sI!%bKe%BTLiwAfC54dXdii&F1~^D(JB# zAc@oxMl602c9)Um-i+W^!V(<7SZ-F3uINv{hD9Jtjrky+x485r$5`o$r;|}#UM~ho zq?QC?nf-R?F=R>6Sym`4E@v1^ff1Qg>hV(xcdZ@P2l2e6xh1*he--v%2(qXz1xciq z9AcSbkpB!>2KH6{A}nQ`VpE?KW)%)08;&*x})w6y6B9$ zqxvO-l=AMVFBTo}6LE&_S|4;o9Zds?#1ZvJchmzV(jBFsxxyB8M@`NGN&aJZ)B)9o zgHU(W`AU%N&+e$(jt;n|DeYIu4s>9;%$QCOOtw2}<3>lEsZGGe;ix<6(;SfL-*!j+ z@dKkzH%&*~QHvLXwEx^4Ww&z!?vCp`0T)*YqOV0>qE^<1L2y}7vJT>sw{=a25F&e~)osQmR5 zcNB#!2l4#XQ|&D2Uk+~Zq!Nt;H%vhi+5e)TLH?KGUx}NMCE!!j&cfmo#8{@Z-HOlK zItC-Fq4PGX62$YChx7{weF{dYBFlCK)!#otYMD$dy0dCqA&cL|oqdGG{v2aTYfARQ z`F_e4cdpn+fq35X{4{xkWikCoU1Yg;8zhlhmJ*9W*iO8&tMydf;lh&7S#s8swF`AK zx=ulLsniAWyyfFwa_2Ors&go^#4Z3yq?WD3GGx66o@G$EHDjW%BnC5esi8vdXm=E> zAA~G3AAoq?61tAu(GIx2>L#-MN(M=!mLtS6^wFuh=op)1sW4kuLeDdn7Bk5XqBpEH zgOO#~E)dUK(!u z?jWAG_^kE9W7XcEaNL3SRRbiETC#{GchQXQs4gcvR$2&4W(ZT4ZcWJ3Y}j+(*XS6_ z=?&s}OTZN}i+yFjUQ1M$c0)iCspSi?xb7~+Yp`O+?Y0w^=upOD{FZcks5bxJ9a&VS zgLvNZYcY9h`ZfCQ17z7g7bKBd8qLE`8Y_ARc%r(D&+g|eEdCc6%iwclChN4T?mT2_+zj)Y5}ko+aLMM3z`LlVif-5XM->btbz9r_9^@9n~e} zB#7rN)!&m235z=3LCE422$D!GTEw#ImNA}X=sUIjX<;eiEMZY(mf`k!x>BTSaR;~JK5nEvP7tWc;51( zF**0wb%?{W40T(BBvOkrv0R=yyEd}q>{PoiEFPB`OVDoemb9t;#E!^PxF5vxmJS(Y zM*hhrmviVC3qJyqNG$=x@?vbm>d5lGR?rh+sp2e=d&pj%Z|9ERi!555K|F8qc}2QC z?)cegB1?I1kVI;^MJyU&SMgq+k2_Du5|-=;rY>%e$eMUf?W*(0a;i6o=PlayWUqqZ ztCoyMmb}3riPVx#EP+o;pQE~@dUSd(EU{M@i?1&E8r|Zs!+1^nDPIuJTS_mJYx3Tm z?e-u`&mfRQYWYkoC3Zvb^UKs@Ju8GIFp{wZd6H*jbMqMdRn2@~5YJog8k4o8{lnUm z&d4B;L~5z0hsXU*-ODbbx~L9|ujcmGY^l>##&Y-@*=KS4xKm%zG4^dNi03WyW|21~ zb%yNmM0MFX1tgJLIunawn~l-PGU!-qWnn4iEIJCa@c6Q&@qz$kQSk-wyroZ9GV`l! z@^~4tEDZulq?VDya_Yh-ynoEk)cGxhB{hnvi{U-e8F_Yk1n!XVe+A-siGh+ z$kJ3FB#~N7iDl)Euv~PEDgRv1M_9bBGZs?|a-V$9Y(QOfjD^%+_&c7rge)P?`wzw} z!22sNY6g->EqjS2f8KZr3B!ji*T_FN)$8KmNN6V+vV4v6P1-%prReZS-3?2U)&-1o6COVFsDI(EF&4zZ4en z1tgJLo)C-m^lBTBMNPYjfv^PMVl3ajlevp+Zoc@-;N9kdc-}Il8o4Gby{U^kimogI zNu-uSVmUu{U_)f-P~Ullu((7smc~0t*V>I+H*2H1>}U$&c}of1n{aF2wd>I_*0K#q zBDK_*kH>w@pxE)qGN@a9Ghr#?EJj6Se5t+CycM!c*8uUnMYWV%lP^zeHx6044FyT0 zmbS!_(9;F)QKU37%Su?%W0<;xZzSjbl{%mBetyoAKs;~xyqP?|^#AYUF zi*kYeIAn=)y>w7meC{xoVfBgPXXJkn~0pfYflkVg`xlQ+wjmVOG2_%tPjEE(+ z(5w(ylIxB>E-d!3jK$l6tWbSsv}+Qwls5tKyu~<@jQi~m^jm=}m90S%sl|?1ItRz6 zB1_NVx&FeE&snZs*2eeA!%bGkAWOe-AfC5;G9+DVjZL@XJ-i!C14*QoJ|Ig& zr8QTDCG;L+Ig>-W)}{vBazhpkRS?fxifzcb-?Zq(MRbf+wE;<_mPBI7NcxWV?@xZ> zaz|KP?=zNsS2Dg_Ykvptvp7B*#PgQV^h?LqFH&)L)Zl!OL~40QEcI`#4MlZvt@xZI zER~$)UO)1Td{$HaD5}dsg_7U#yk%_{avgISvE>J#*(v!>`Y;D<4hj1bnOD-dCS;p8al>q`duj! z7XNt0;!sJtZ#VZH(iU0XUjXsEMJbWoC-1FR8_&V^jRZ-gmMO#%ndy)BzwFjx=@(&f zNMJ0Dds*Z0CH+Y!ygI#qM-b0jz8xoXN{UX^YofZ8bO%YKmSx0Z{-F(C*|oN$YjyX( z=EsUS%cr|!-g1-eJ>2EF<2#7wEgcV#=gQPikL{3Up<=h+KSFA;CYBl%V=Pe9mSYQ1w!k0_rAw)Ei{^r*6WF>qs}@ zg@jK9sGD%jAdv3wx(Tm#C*6c&79Jdkx(TO^1gZW{-Gr^4yjqF62@A%7l=5!EExkM9 zv!nIZmUy3|v}qubIHLaOCTvxkbQ9Kn-Kzj~6L!-DN&aIu;n>g4F{qocmnlg0XE))| z^-j14(Chw1{Kdm+&mHN3$#xUwSsuWd?vpX@9WtujQ!j`^ zqgwJf5YJydD;&wFrqJgJ?v9!@4J48MFAA>Yf0=%?;ca9|?9!vFu!JWumQC*DE$P6Y z$q$f4zb}aAEiU)STGE9ln&KJxal=3ospUMe*u}>mM3%kLAsWKs@sP2MTyBrAK{3rP zS|UqG5{Tz5$8V7Jm)~=`FF>Q3N(M+GwcH~XN6Q}VktOnA-{Hbi#aYJbla8W^A$@Rn z_Mzh-p0~8OByUQtcG?kwESm#C5~<}Cv5fpsIvH7vr$>$#mh4ANU5;0fj@9Q=uj2JF zTjzmz-eM6zW*L@Rj2n$C6^lR;sil%woGWH5Lzbwr(iy@Mo6K1($*kNtqg;IMH~tCY zc}wUda_2NsYcbvrV|UH6-#5R>G}=a(;58*r0?-<>=fm z>IdR^i(@@On}V-n%{={jB8z!H z5YJmYPLe0U3$4uX%layy36VF}CzINFudZ5KGC5TQ^W$^7nMyB`lGtj3wEE%$Uq}Cfy!uTtGZ; zIcrL;V+PJj&dAcx10<1J4iL+iep9uPMdh2fv#@wQWi0u>$ZTnDk`kW%30FMvJD#`1 z3?^?$t=-)46X5##Ac@p+hFJDSRV+c4^`9Pl2#f7A#F|CKcOS2fi>d_iyk*50 z(&edfG!?H`-m38A_m7ZTqKPFnv(5r!2@m%46P6s#vZ*=wT2SBltAfz^GVBV7=Pe`j z$@%g$ZB-RI#?+!g5~(GFSUSw`!QCFF(L*i>OI#XLm!d7?O-Yo(_eaQL@c_j07O$OT ztgLI)sSsJ#KL$ypmJ(u_dTH7iRF}h}gCd0`IGwS?SCRX%v4wux$l|IE;(1HSDssL= zoIL&*S>EV@BvMQL1$f;5wm}cSDQO+m{Eo1=WH6SJiDaH6IeJ(fWHFBg@w}xpHUxLx zmk-SfN0#piAc@q{g;;*n^L~ix;=DB?Nm$A_i?0S*snX=r0{oVAf<1`mE$vs4bN|t} zUJa4O-3cU-T1FAelTHyck>%65Ch5YG{+y}HZcTE&=sfk^fa+p$8N~CJ!{5lbf9T^# zd>wPS0g^~9^N6L9*Trhc;?(HG8)3Qrg0b{bCF_;-=Vn$R%fdDwp0{XiB45-U`RZ0j zRF|6?Ac@psMlAjNcd|wnuhT7xgvBS5vH0~MU!$8o`ZV5iq#)DbUa zITi(yNG%S;5@w9uP*4gaCepDBa7!c1}KF=oO%iEJ(b&zF# z0!SjYoFWzj&1u7tCHAQP4`IpYEa?kLSHYE93V8pR%o8AoJ19LD)#buf zkVI-pAr^y(3I$Y`<)1wp2utWI#^TYOJnvf{o8yd*v6bIJJa5@@mOP6c8XAE2zx-0& z>-Ue4T8fCpWAHotCf-)9UUOk_&1Ni0_T;&8@bps~ktM%Bi03WKv&k_wz-PHXs!NR# zAc@pcb0Hr0UtIaz64hm^$@ccbQps5^s*%;UJ=^GaL6)Z$AfC5eODB7-Xy0CqpZB-f zfFx2&J7Td2?ua|Qr`GwXCM=n+nYzRqEx^yn7N?EcBa4a(i03UH&SVZY-dJZCs>{Jv zAc@p6gjjxPtipTVx0>zWS6HIoFqTEl$ZDuFEACZCmR}1%Ja6flb`;-_y*lEH_l)#j z4w6VMI>d5WecujrjJ@dBWSFq{zhx{Te%oucrRH4_&s#c#lIMNxMSbU^W6UoR zB#~Nn63gE2ad^L_yn~8Ugr$hH_yv%Cbk`gt zky<>7#dy4bV^o(}F_}8TlKhUTOPzUS74-MhpFbnZl6Vl$TU_sxbN{-E67Ia;k^+)Q zE#bs6VA7nQ$Wkxk+k9aO&t)tM_T=6&Klu?}o!;#Yi03T_Ym<5YbLzGc=otH00Fp>8 zNyMV|vuh@@RCI2+R9HOn7>imIx#zcbSII|pna~@=^Oglsq-!m}%_|OBx(xzJq?TM_ zIWzr*53-!vHF1rwRB;xaHe_$Yd$rHwedD~GKs;}G{eY}cy=0)#23b0~f+SMQFJhUz z=gewk=``4OgRo@hGj*vpo~*dDZd8HaS#%u@;(1H16S?Pi>yxz(S*lM0Nu-t*i|{>+ zQtl!Aj6CE_mX)x?7BH41+u3-0@!y(+S9B+Zf_UCy(afEUl@=GmkmXJ!NFufLBbMa@ zPuD?pakOZ>Q&S8(F_pq=ObC#;*diWSSe>cGkS^DUKc-~^DMaGu_njy=P<$goI-#v|+2 z%O$0#F4@w_ExAGseZYHzt0Sq66iNu-ve#FBW|33siX zJNnsISRy|#mXR+>CvoO-hf~O+?gQd^%Y6DQw)jU&EoAWz1WBZp3&fJX;@lckm(q8w z&k2jyN5-PnZzev*65F+vHCNE@Vg|HN{~cqxlb$`w1c)Fi$~U? zOTuFNiLvbZLOLUdgn#abEH>{!Ja4&og>-nYQpvGFby-#ll1MGt#B$<{XH8^zd*|de zVaeevH>;EHDFqvTYk(|TNg$rL46}~Fb!l{MGyXQzw^WcsYWYemJyaXJAxn#`DR+b= zu9&Gykqg;dz1|(|b;y#Q0^)f~ix1@7*SWY6?@Ma>5+spYR1ERBAFbas2U%<$bxahN z;LnU@1>G<9M1I32$Z{ba#PgO<%g8l({fbC@52Kw8l1MGRh{f#Fm=I)X=)EFUSX@dN z%k#z(q|9xZb4JhKoY5C5wRE?^maydaUVV)S6I?ZnYuLZ zLgvQn-d>1za7&*6;(1GCIJt*eVmlJ|_MD#ul1MF^iRJmHM7$%__xDpj3Cr~|#uEB~ ztb!i#-VE<}__Ykg^OidM$@nt*L?xc<^QZzzq?Uuk5;o!_K3^_ob^azSKIM!hJe~Ah zX}Jxi03U4$z*)#+itWH8ei^w0!gHnKw{ak(qR)i#zxMrOU0f>eZ>HqR^>Ea6__&Sa*4DG3vZeTn;ixI`8$Gw!)qFJG}R& zqRxBkwIItsb>4S#CY|?Fa%X-;o%avSLAt-|ygx?w@l`J~+=x2wO}2tm|EJFT@9j1v zpw9b8b|9s^^FG~=?BjdY`1wF|L@n9}5{V<~kIws`U7zsnXymQ=pHSz$stZW+A3N`7 zoM};rI`8j11IhmEydN^p1^3H0u9kEW9hlWsoauqdcHVzAJB%}_4R6pNb>8=C0W$sD z&ig-p;6NXrAk=x^wgX7}&z<+3((G~P{nkL$E~xYVTsM&LUv=KU_o|CK@4I$thi5s$ zdV`#@tHhXia+O#W|NcJeyuUaQr1@{1_fZz5y7=np^j1||aq_vVr%FEQSY7uv)*Bt` z$;UxFfAw_OPTt0OhknHCl?V8PB(ncS;W_zVa@trqAxnswUt3{G{KAX`t{cdV{Ei3d zYmg3kl4 znPS6X5YJoUhg`%@O$SS6`5=p}H%KD2C@scM8vEDm!ZVEJ*Z1`i7T2$gCCGrxGT0rt zgZIG-z6;`cOV5Sm?tJpG8P>>RlmwDUEnSIa@S7|=`;*v1b%?N3a+abVf$NG*EAlKL)w1UmO^Jr7M1mgw({B_fqPnU8p`j91+h z`Ga`g;;l)ZG;X~eAC2mABN!x+TGkWG?6ZgPN|lpAN;<;gU&UCS)F3mz)oLo&M0I(2 z3B>ajcT+OUaQ^6&*T`ac4J46T_7Tgmqt+);T})@}m@h02KN!n(6l-q-N z-twRpImYUA*^bA`tp`97spT56j110dh%8E*%8i62`6pKwFEY!}L!DCeL*~LsmLbd zOUmK)smM}z79^2cJ`hVTlQO# zx_r+1j^AESG5|@WmRd{jxc_a`6Z{n?@2H!0!cxUqk~fhVlfKpF;cN1o`yigToIOCE z%*}_o<2NN+o`57$OM7D3RkjY#GU#kjbQG3sJEks6X!os+efyQ@7}MJd;(5!3U1TNX z!$zI-(J_{82a-rFLy1K>Z}nYtjGgE_{jjja?qn<<=aYRFPefYRK$d#FK|F8S7)ZvK zUEyUa$Wk>BB#~NX5zEN3^DfAuF!%6rVF}#DSe6YY-Gm2|X5*DA4VQv=-qN54S#29J ztqY#{ODwWTvbp9ZVM*Q1)a7IovWlQQAfz#>%Q1Tp&szdQNL|({nl43k z8SV;_NG+F%#rT~iwkR$LyCy7=dl*Z9bFvz0P~;cU2AhL~3bC zEKldv!+YM3RiBY5EOGmox}<54H(2=>)Nz;R!}Bm1^Fx23W9)$@NFufL zCzjw@?)V!KaVZ;egeBOKvFuc}!`I}~%N+Wkx(vw%@w_G0nLNMPD{QWZELsI1iPSQU zSc>1D#H$FR-v<^6i_3n_GM#i1pRX|6hb$M$Ks;~BH@<@Fl0Vh`7P8!{0!gHn6~uBn z_%MDO=Xw5mnXr^`mJwsgx!>=&P=_w>ZmT$HiLNH(%zC>lV4Yx^aEK2eFRCQ7B^z~SyO$Zx*jRGnTLq`B z=vmEI6&Cx0jHSm2vg&5z#@twB`SK3L^On9Tq(j2KwUs-v94!G!q?Xsj(xAo`yt2#0 zYIj><$>%I>W{^3hcU!A#A&a3si03Ug+{svZ>i*bH$ge61B?0od|c0Ig0-OLWe^OixO$JpC- zxxIxY^blitJD0q(7=0#nDXPm7OAyanV*8S3RQ z?aEjj?T97ONB;z}w6_QGyk*30@*X*;cm-Z*zuyTYky>1c<;6+GAIM@8;yX`R{Esr0 z%s%9uMMlcB@5quG3F3K+?S0Z6RXeWTCsdbKF(8T5a+X+bsH{DTES^J>4TZ(Qjj`NI zC38xOst*g1<=Z$A&s*lrB|B0@ELN(6EH|fvBvQ+5V!3kFpg*z{&ik=SSc*7{#RsyA zK+k5;TVzpR3gUUoF@JJD*5}eve`F~#0!gHn7sPU>Q*FFUVDF^vYaUZNu-uCVp$Qj^ggPKWxu(W!V>PmSgNa#bN_n` zndc850pfW}-7#dW+_7K+-n}to97rOyC@;t3e)RwP%8=zn<@z1M;&F_zv^h)m@3+{l zz7f?$L2<+Hc-~_2iaaBKXlSzx)g`VzNFud#BbISCWAWU$?x@Rqg{6wK`1+G~;$6F) z$KN3dnh4@~Oa0AcPHD{d9|w^oY6eImwTvc~uHH*EQC$}8DLN=D*~gi>Y||twy9|?_ zT}5?Kbpi3bWxyeFO>Xd|Cf@Vj!~-OeTILf=kBkppk!9!pjvm4i>&aMF?jrNa&3BH$ z*W|~pAfC5;yhiT%ca>G*zKe}sAc@qnfmoWIYJ3%0-Ucl55tcwN#^Pa1x^F|qknh86 zx(woZ%d8{hnrs~N<`_E0YTX1$q!veF85gq#KfjoLwhI&%CvV2G{v=sxAK^5A7dpm@ zzkzt(VzP*gmEqCj?2$!5ap&(JA+`7s%XXc1V^Li;K933&mSWB_%8RUsh;0_Ff-KwL zfq33h@qoO4iL$f6d#iW&0Fp>8*NNr!kQ=X&rNaGfq_Cu(VCquxiCo8mA4NqWOVlY4 z&s$V#kllim4eQ|@4>tsXBvQ*$VwpME7w;6dXV~BvVTnA+ShBszoKpP`b#Xt;ZY>bc zThwcjbHCaADA*1%B_>qPk=^07;~l4#c9dVT2JnU&cIl$Pkts&T=oF%w2r(U5n?&x4s4O zyycc28Ta>uygPy{mkL1=sbv_kyxTqk@8+QNPUnrV#Q8FHDcwtUayg~Z|01f(>W?6v zx7>&)!$)Gd)^*^R|Ng{6$M%&15D$EI%|Hyc@o+z0Wz0+kLu_c>+uLAky^Zog@>oW#w^~2w! l{&a5R?*y{nG3EmK9adXM&q8YydRzi&{@WUb;2iDM{|`)mCDZ@_ literal 0 HcmV?d00001 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/flags_0.json b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/flags_0.json index d54d7dc93..2608bef10 100644 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/flags_0.json +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/flags_0.json @@ -27,7 +27,7 @@ "xml_output_file": "", "experimental_orbax_use_distributed_process_id": false, "experimental_orbax_use_distributed_barrier": false, - "submission_path": "submissions_algorithms/submissions/self_tuning/schedule_free_adamw_jax_v2/submission.py", + "submission_path": "submissions_algorithms/submissions/self_tuning/schedule_free_adamw_jax_v2_bn_fix/submission.py", "workload": "imagenet_resnet", "tuning_ruleset": "self", "tuning_search_space": null, @@ -38,7 +38,7 @@ "framework": "jax", "torch_compile": true, "experiment_dir": "/experiment_runs", - "experiment_name": "submissions_a100/schedule_free_adamw_jax_v2/study_0", + "experiment_name": "submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_0", "save_checkpoints": false, "save_intermediate_checkpoints": true, "resume_last_run": null, @@ -59,6 +59,7 @@ "helpshort": false, "helpfull": false, "helpxml": false, + "only_check_flags": false, "chex_n_cpu_devices": 1, "chex_assert_multiple_cpu_devices": false, "chex_skip_pmap_variant_if_single_device": true, diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/measurements.csv b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/measurements.csv index 726ff2440..c91806365 100644 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/measurements.csv +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/measurements.csv @@ -1,1598 +1,2382 @@ global_step,grad_norm,loss,train/accuracy,train/loss,validation/accuracy,validation/loss,validation/num_examples,test/accuracy,test/loss,test/num_examples,score,total_duration,accumulated_submission_time,accumulated_eval_time,accumulated_logging_time,preemption_count -0,0.280216,6.9083138,,,,,,,,,,,,,, -1,,,0.0010762117,6.9101987,0.0011,6.910505,50000.0,0.0013000001,6.910612,10000.0,62.0377836227417,145.27190852165222,62.0377836227417,83.23402738571167,0.0,0.0 -100,0.9382999,6.75622,,,,,,,,,,,,,, -200,1.7159138,6.4328136,,,,,,,,,,,,,, -300,1.9524416,6.2667093,,,,,,,,,,,,,, -400,3.2473595,6.1781263,,,,,,,,,,,,,, -500,4.891164,6.262416,,,,,,,,,,,,,, -600,2.3864083,6.0814447,,,,,,,,,,,,,, -700,2.753381,5.9718924,,,,,,,,,,,,,, -800,3.687202,5.987758,,,,,,,,,,,,,, -900,2.6280944,5.933118,,,,,,,,,,,,,, -1000,2.3546267,5.840041,,,,,,,,,,,,,, -1100,3.599295,5.88187,,,,,,,,,,,,,, -1200,3.9748979,5.886465,,,,,,,,,,,,,, -1300,5.010844,5.7755485,,,,,,,,,,,,,, -1400,3.174834,5.806589,,,,,,,,,,,,,, -1500,4.3649273,5.6979904,,,,,,,,,,,,,, -1600,2.6471348,5.748484,,,,,,,,,,,,,, -1700,2.4704807,5.695845,,,,,,,,,,,,,, -1800,1.7465972,5.610014,,,,,,,,,,,,,, -1900,6.677021,5.88741,,,,,,,,,,,,,, -2000,5.0624423,5.7971625,,,,,,,,,,,,,, -2100,3.3601139,5.8100057,,,,,,,,,,,,,, -2200,2.5220048,5.5251393,,,,,,,,,,,,,, -2300,3.1667466,5.5600495,,,,,,,,,,,,,, -2400,3.2634444,5.671502,,,,,,,,,,,,,, -2500,2.538338,5.54154,,,,,,,,,,,,,, -2600,3.0445223,5.448047,,,,,,,,,,,,,, -2700,2.3397136,5.385078,,,,,,,,,,,,,, -2800,2.6597435,5.375867,,,,,,,,,,,,,, -2900,2.9278173,5.3871202,,,,,,,,,,,,,, -3000,3.3318179,5.32281,,,,,,,,,,,,,, -3100,3.5130124,5.5266876,,,,,,,,,,,,,, -3200,3.021631,5.3545885,,,,,,,,,,,,,, -3300,1.2741576,5.2872806,,,,,,,,,,,,,, -3400,3.2336977,5.463455,,,,,,,,,,,,,, -3500,2.3230338,5.291337,,,,,,,,,,,,,, -3600,2.6491072,5.3013506,,,,,,,,,,,,,, -3700,3.639723,5.378151,,,,,,,,,,,,,, -3800,3.0273232,5.2039704,,,,,,,,,,,,,, -3900,2.917794,5.1859326,,,,,,,,,,,,,, -4000,3.2125683,5.2224617,,,,,,,,,,,,,, -4100,2.9455898,5.1019897,,,,,,,,,,,,,, -4200,3.4075716,5.2506337,,,,,,,,,,,,,, -4300,2.1821988,5.0772753,,,,,,,,,,,,,, -4400,1.9292452,5.1112666,,,,,,,,,,,,,, -4500,3.0773811,5.1388855,,,,,,,,,,,,,, -4518,,,0.039580677,5.9223266,0.036059998,5.994765,50000.0,0.029500002,6.248613,10000.0,2058.049063682556,2176.2921516895294,2058.049063682556,118.17159128189088,0.0278799533843994,0.0 -4600,3.0757809,5.022922,,,,,,,,,,,,,, -4700,2.1589346,5.15825,,,,,,,,,,,,,, -4800,2.142906,5.0416126,,,,,,,,,,,,,, -4900,1.516194,5.013075,,,,,,,,,,,,,, -5000,3.939658,5.104778,,,,,,,,,,,,,, -5100,4.2181535,5.1436095,,,,,,,,,,,,,, -5200,3.3677444,5.0368257,,,,,,,,,,,,,, -5300,3.702074,4.9691706,,,,,,,,,,,,,, -5400,2.8452134,5.0268507,,,,,,,,,,,,,, -5500,4.9707375,4.811808,,,,,,,,,,,,,, -5600,3.4141655,4.9023495,,,,,,,,,,,,,, -5700,2.4545686,4.8978815,,,,,,,,,,,,,, -5800,5.332715,4.974868,,,,,,,,,,,,,, -5900,3.2696788,4.857068,,,,,,,,,,,,,, -6000,2.5486977,4.8113956,,,,,,,,,,,,,, -6100,4.340679,4.9837155,,,,,,,,,,,,,, -6200,2.0728767,4.827755,,,,,,,,,,,,,, -6300,1.8308903,4.712989,,,,,,,,,,,,,, -6400,3.7294922,4.9287744,,,,,,,,,,,,,, -6500,3.4071636,4.7604427,,,,,,,,,,,,,, -6600,2.9674847,4.7801595,,,,,,,,,,,,,, -6700,1.8668141,4.697668,,,,,,,,,,,,,, -6800,3.7065735,4.7670755,,,,,,,,,,,,,, -6900,4.6483707,4.7852592,,,,,,,,,,,,,, -7000,3.8844688,4.7861257,,,,,,,,,,,,,, -7100,1.2563435,4.7136173,,,,,,,,,,,,,, -7200,3.540576,4.6784687,,,,,,,,,,,,,, -7300,3.2213483,4.840146,,,,,,,,,,,,,, -7400,3.2844977,4.8261228,,,,,,,,,,,,,, -7500,4.4265137,4.890481,,,,,,,,,,,,,, -7600,3.5370514,4.6698093,,,,,,,,,,,,,, -7700,3.928209,4.8132963,,,,,,,,,,,,,, -7800,4.1096754,4.658948,,,,,,,,,,,,,, -7900,2.525869,4.547046,,,,,,,,,,,,,, -8000,5.1850243,4.8210316,,,,,,,,,,,,,, -8100,4.4520864,4.842931,,,,,,,,,,,,,, -8200,1.3208221,4.5518627,,,,,,,,,,,,,, -8300,3.1019812,4.5623493,,,,,,,,,,,,,, -8400,2.7229078,4.578224,,,,,,,,,,,,,, -8500,2.884135,4.6114874,,,,,,,,,,,,,, -8600,2.6330776,4.473111,,,,,,,,,,,,,, -8700,3.8220706,4.6068187,,,,,,,,,,,,,, -8800,3.309848,4.6149096,,,,,,,,,,,,,, -8900,4.8462615,4.680875,,,,,,,,,,,,,, -9000,3.0733392,4.596997,,,,,,,,,,,,,, -9100,4.547898,4.743021,,,,,,,,,,,,,, -9200,3.2441206,4.6345625,,,,,,,,,,,,,, -9300,4.2056413,4.600101,,,,,,,,,,,,,, -9400,3.1424062,4.475791,,,,,,,,,,,,,, -9500,3.134442,4.515594,,,,,,,,,,,,,, -9600,3.334294,4.450712,,,,,,,,,,,,,, -9700,2.7007835,4.4899807,,,,,,,,,,,,,, -9800,8.78771,4.5756187,,,,,,,,,,,,,, -9801,,,0.009745695,7.5413218,0.01064,7.568091,50000.0,0.007900001,7.6627584,10000.0,4054.1717386245728,4237.352629899979,4054.1717386245728,183.0349597930908,0.0581820011138916,0.0 -9900,4.035606,4.5563397,,,,,,,,,,,,,, -10000,2.6494067,4.4724283,,,,,,,,,,,,,, -10100,4.734405,4.5196056,,,,,,,,,,,,,, -10200,1.7634783,4.4559884,,,,,,,,,,,,,, -10300,1.9574803,4.4244103,,,,,,,,,,,,,, -10400,4.5312862,4.3619804,,,,,,,,,,,,,, -10500,6.095785,4.574256,,,,,,,,,,,,,, -10600,5.773542,4.5945807,,,,,,,,,,,,,, -10700,8.404025,4.700728,,,,,,,,,,,,,, -10800,3.0360038,4.4262486,,,,,,,,,,,,,, -10900,2.9663846,4.3502684,,,,,,,,,,,,,, -11000,8.538077,4.614937,,,,,,,,,,,,,, -11100,2.58923,4.356302,,,,,,,,,,,,,, -11200,4.8081527,4.4986863,,,,,,,,,,,,,, -11300,3.9248297,4.4430733,,,,,,,,,,,,,, -11400,2.7789543,4.425868,,,,,,,,,,,,,, -11500,7.6228423,4.5680304,,,,,,,,,,,,,, -11600,3.257058,4.408654,,,,,,,,,,,,,, -11700,3.370618,4.3430862,,,,,,,,,,,,,, -11800,5.211456,4.5456705,,,,,,,,,,,,,, -11900,2.5772142,4.376078,,,,,,,,,,,,,, -12000,4.9535356,4.5590544,,,,,,,,,,,,,, -12100,3.306491,4.4554644,,,,,,,,,,,,,, -12200,3.197167,4.4501705,,,,,,,,,,,,,, -12300,2.8215528,4.2788863,,,,,,,,,,,,,, -12400,3.856448,4.349657,,,,,,,,,,,,,, -12500,2.7383,4.31394,,,,,,,,,,,,,, -12600,3.078698,4.4544454,,,,,,,,,,,,,, -12700,5.4603825,4.4363747,,,,,,,,,,,,,, -12800,3.1067622,4.331504,,,,,,,,,,,,,, -12900,2.612004,4.3515515,,,,,,,,,,,,,, -13000,3.8724933,4.2877936,,,,,,,,,,,,,, -13100,2.4128487,4.1776004,,,,,,,,,,,,,, -13200,4.551445,4.221446,,,,,,,,,,,,,, -13300,12.10603,5.087637,,,,,,,,,,,,,, -13400,4.0941777,4.3070536,,,,,,,,,,,,,, -13500,4.601502,4.293885,,,,,,,,,,,,,, -13600,5.7885237,4.470435,,,,,,,,,,,,,, -13700,3.5543041,4.3319473,,,,,,,,,,,,,, -13800,3.2628272,4.282591,,,,,,,,,,,,,, -13900,3.066159,4.3300147,,,,,,,,,,,,,, -14000,5.9923234,4.254425,,,,,,,,,,,,,, -14100,3.7241254,4.340272,,,,,,,,,,,,,, -14200,7.0247498,4.561752,,,,,,,,,,,,,, -14300,18.445704,4.336795,,,,,,,,,,,,,, -14400,3.8160427,4.2095313,,,,,,,,,,,,,, -14500,4.3744287,4.188452,,,,,,,,,,,,,, -14506,,,0.005859375,8.495525,0.0052799997,8.498083,50000.0,0.0042000003,8.537536,10000.0,6050.326933860779,6366.033954143524,6050.326933860779,315.471892118454,0.1032261848449707,0.0 -14600,4.467907,4.2669516,,,,,,,,,,,,,, -14700,3.3470304,4.3462744,,,,,,,,,,,,,, -14800,4.7857275,4.3258553,,,,,,,,,,,,,, -14900,5.4956465,4.244761,,,,,,,,,,,,,, -15000,6.0854445,4.3275557,,,,,,,,,,,,,, -15100,7.0976405,4.310113,,,,,,,,,,,,,, -15200,10.45203,4.28819,,,,,,,,,,,,,, -15300,2.3430781,4.353918,,,,,,,,,,,,,, -15400,3.6630425,4.374762,,,,,,,,,,,,,, -15500,3.7355542,4.3146195,,,,,,,,,,,,,, -15600,4.433247,4.2850113,,,,,,,,,,,,,, -15700,3.806504,4.2035556,,,,,,,,,,,,,, -15800,3.134234,4.219532,,,,,,,,,,,,,, -15900,5.6245284,4.3095727,,,,,,,,,,,,,, -16000,5.1426725,4.389682,,,,,,,,,,,,,, -16100,4.930809,4.1630583,,,,,,,,,,,,,, -16200,3.6823552,4.16663,,,,,,,,,,,,,, -16300,3.5907009,4.1396,,,,,,,,,,,,,, -16400,6.19738,4.290329,,,,,,,,,,,,,, -16500,5.625287,4.2814417,,,,,,,,,,,,,, -16600,4.5783668,4.142776,,,,,,,,,,,,,, -16700,3.7919612,4.3298225,,,,,,,,,,,,,, -16800,3.7367795,4.252203,,,,,,,,,,,,,, -16900,3.255845,4.1587377,,,,,,,,,,,,,, -17000,3.4928837,4.1722126,,,,,,,,,,,,,, -17100,4.9969463,4.1048155,,,,,,,,,,,,,, -17200,3.4076035,4.1547966,,,,,,,,,,,,,, -17300,4.81383,4.1707573,,,,,,,,,,,,,, -17400,3.3729806,4.182557,,,,,,,,,,,,,, -17500,5.335816,4.2266583,,,,,,,,,,,,,, -17600,3.7623515,4.2105193,,,,,,,,,,,,,, -17700,4.0723586,4.2333765,,,,,,,,,,,,,, -17800,3.9344776,4.2579627,,,,,,,,,,,,,, -17900,2.5470374,4.08372,,,,,,,,,,,,,, -18000,8.017417,4.3492064,,,,,,,,,,,,,, -18100,5.6542134,4.268447,,,,,,,,,,,,,, -18200,4.561398,4.0348454,,,,,,,,,,,,,, -18300,3.614934,4.1169233,,,,,,,,,,,,,, -18400,5.3847723,4.18013,,,,,,,,,,,,,, -18500,4.8149834,4.08442,,,,,,,,,,,,,, -18600,3.6389842,4.19518,,,,,,,,,,,,,, -18700,2.1904624,4.0891123,,,,,,,,,,,,,, -18800,2.141636,4.0688114,,,,,,,,,,,,,, -18900,4.326212,4.1618667,,,,,,,,,,,,,, -19000,5.230624,4.149301,,,,,,,,,,,,,, -19100,5.573018,4.1819363,,,,,,,,,,,,,, -19200,5.422349,4.191166,,,,,,,,,,,,,, -19300,4.9523053,4.2868195,,,,,,,,,,,,,, -19400,4.7175326,4.1128936,,,,,,,,,,,,,, -19500,6.7898293,4.218955,,,,,,,,,,,,,, -19600,5.6592836,4.224696,,,,,,,,,,,,,, -19700,7.3466115,4.1840143,,,,,,,,,,,,,, -19800,4.838632,4.3253455,,,,,,,,,,,,,, -19882,,,0.0034877232,8.590018,0.00326,8.605198,50000.0,0.0032000002,8.626876,10000.0,8046.917476177216,8452.700548648834,8046.917476177216,405.4710731506348,0.136307954788208,0.0 -19900,8.406875,4.280713,,,,,,,,,,,,,, -20000,3.0530674,4.073484,,,,,,,,,,,,,, -20100,9.283489,4.2329326,,,,,,,,,,,,,, -20200,5.7091684,4.1942415,,,,,,,,,,,,,, -20300,6.52517,4.078738,,,,,,,,,,,,,, -20400,3.7838748,4.03094,,,,,,,,,,,,,, -20500,3.3891509,4.0229764,,,,,,,,,,,,,, -20600,4.95285,4.13231,,,,,,,,,,,,,, -20700,4.070734,3.974365,,,,,,,,,,,,,, -20800,5.4343624,4.1464214,,,,,,,,,,,,,, -20900,3.2595744,3.9693594,,,,,,,,,,,,,, -21000,6.0546336,4.160526,,,,,,,,,,,,,, -21100,9.128398,4.1398525,,,,,,,,,,,,,, -21200,2.6798873,4.096501,,,,,,,,,,,,,, -21300,7.498567,4.174957,,,,,,,,,,,,,, -21400,9.177841,4.2797837,,,,,,,,,,,,,, -21500,5.297593,4.1277456,,,,,,,,,,,,,, -21600,3.7307286,4.087367,,,,,,,,,,,,,, -21700,5.655477,4.0202885,,,,,,,,,,,,,, -21800,35.790287,4.1087584,,,,,,,,,,,,,, -21900,3.84333,4.0212502,,,,,,,,,,,,,, -22000,5.458952,4.042703,,,,,,,,,,,,,, -22100,2.7214105,3.9247923,,,,,,,,,,,,,, -22200,2.6311631,3.9167001,,,,,,,,,,,,,, -22300,3.7071095,4.0104403,,,,,,,,,,,,,, -22400,3.8738332,3.9798355,,,,,,,,,,,,,, -22500,8.183264,4.199581,,,,,,,,,,,,,, -22600,4.083929,4.101125,,,,,,,,,,,,,, -22700,6.538111,3.98141,,,,,,,,,,,,,, -22800,10.706647,4.1743784,,,,,,,,,,,,,, -22900,4.3562045,4.0752883,,,,,,,,,,,,,, -23000,7.3156056,3.9747617,,,,,,,,,,,,,, -23100,9.278941,4.0692983,,,,,,,,,,,,,, -23200,4.4624643,3.995635,,,,,,,,,,,,,, -23300,6.2670517,4.0707054,,,,,,,,,,,,,, -23400,3.690238,4.033261,,,,,,,,,,,,,, -23500,3.8872423,3.9629023,,,,,,,,,,,,,, -23600,12.841017,4.2371216,,,,,,,,,,,,,, -23700,3.0231092,3.9917762,,,,,,,,,,,,,, -23800,4.0215893,3.9973109,,,,,,,,,,,,,, -23900,11.097367,4.0281706,,,,,,,,,,,,,, -24000,5.877704,4.1196938,,,,,,,,,,,,,, -24100,6.811452,3.9925466,,,,,,,,,,,,,, -24200,4.1025453,3.956572,,,,,,,,,,,,,, -24300,12.08199,8.693407,,,,,,,,,,,,,, -24400,3.175768,4.020027,,,,,,,,,,,,,, -24500,7.4916625,4.0345826,,,,,,,,,,,,,, -24600,5.1023107,4.0331416,,,,,,,,,,,,,, -24700,4.7690663,4.057292,,,,,,,,,,,,,, -24800,5.653899,3.8886619,,,,,,,,,,,,,, -24900,7.1365256,4.245449,,,,,,,,,,,,,, -25000,4.1434855,4.041425,,,,,,,,,,,,,, -25100,5.510652,4.0795584,,,,,,,,,,,,,, -25200,5.802321,4.0171223,,,,,,,,,,,,,, -25270,,,0.0020926339,8.377973,0.00232,8.369723,50000.0,0.0025000002,8.380475,10000.0,10042.881111860275,10528.060062170029,10042.881111860275,484.79101037979126,0.1682186126708984,0.0 -25300,5.0090413,4.0756607,,,,,,,,,,,,,, -25400,4.783087,4.096098,,,,,,,,,,,,,, -25500,4.42902,3.8388517,,,,,,,,,,,,,, -25600,3.5773208,3.9284706,,,,,,,,,,,,,, -25700,5.6360416,3.8540926,,,,,,,,,,,,,, -25800,84.76252,4.006638,,,,,,,,,,,,,, -25900,4.2678614,3.9146504,,,,,,,,,,,,,, -26000,9.864656,4.1194987,,,,,,,,,,,,,, -26100,5.3442726,4.0437245,,,,,,,,,,,,,, -26200,6.1160374,4.0446,,,,,,,,,,,,,, -26300,6.744386,4.144932,,,,,,,,,,,,,, -26400,4.537485,4.096885,,,,,,,,,,,,,, -26500,4.2122874,3.9606616,,,,,,,,,,,,,, -26600,5.0336595,4.0358725,,,,,,,,,,,,,, -26700,3.8216372,3.9123623,,,,,,,,,,,,,, -26800,5.373471,3.9777694,,,,,,,,,,,,,, -26900,4.3394933,3.8900046,,,,,,,,,,,,,, -27000,7.179563,4.0710373,,,,,,,,,,,,,, -27100,5.7475667,4.146737,,,,,,,,,,,,,, -27200,3.6109614,3.8999686,,,,,,,,,,,,,, -27300,5.520955,3.9459789,,,,,,,,,,,,,, -27400,10.086233,4.044467,,,,,,,,,,,,,, -27500,5.7071314,4.1303377,,,,,,,,,,,,,, -27600,4.8181553,4.052229,,,,,,,,,,,,,, -27700,6.1694202,4.108068,,,,,,,,,,,,,, -27800,10.649848,4.1889877,,,,,,,,,,,,,, -27900,5.7529473,3.971198,,,,,,,,,,,,,, -28000,3.984212,3.9278066,,,,,,,,,,,,,, -28100,5.3417387,3.7957175,,,,,,,,,,,,,, -28200,5.5505147,3.8924904,,,,,,,,,,,,,, -28300,3.4175365,3.962369,,,,,,,,,,,,,, -28400,3.3096542,4.052376,,,,,,,,,,,,,, -28500,3.6232696,3.869679,,,,,,,,,,,,,, -28600,12.6551695,4.1518097,,,,,,,,,,,,,, -28700,4.246352,4.051933,,,,,,,,,,,,,, -28800,6.4363475,3.9160285,,,,,,,,,,,,,, -28900,3.4846153,3.9209905,,,,,,,,,,,,,, -29000,3.8023717,3.9441152,,,,,,,,,,,,,, -29100,4.8633313,3.8467498,,,,,,,,,,,,,, -29200,6.0180755,4.0778155,,,,,,,,,,,,,, -29300,3.9182436,3.9736567,,,,,,,,,,,,,, -29400,4.738921,4.0044084,,,,,,,,,,,,,, -29500,21.797781,3.8951948,,,,,,,,,,,,,, -29600,5.425934,4.1751223,,,,,,,,,,,,,, -29700,3.7484279,3.8656957,,,,,,,,,,,,,, -29800,7.528026,4.079552,,,,,,,,,,,,,, -29900,17.990604,4.0182433,,,,,,,,,,,,,, -30000,5.3135643,3.7454405,,,,,,,,,,,,,, -30022,,,0.0023317921,8.121338,0.00188,8.11584,50000.0,0.0019,8.119603,10000.0,12038.886108875276,12585.776431322098,12038.886108875276,546.4171376228333,0.2008829116821289,0.0 -30100,5.6709237,4.107811,,,,,,,,,,,,,, -30200,4.0549097,3.8724434,,,,,,,,,,,,,, -30300,3.2703683,3.8781092,,,,,,,,,,,,,, -30400,3.7733035,3.8501124,,,,,,,,,,,,,, -30500,3.84388,3.8641956,,,,,,,,,,,,,, -30600,9.286421,4.0280323,,,,,,,,,,,,,, -30700,7.415321,3.9021547,,,,,,,,,,,,,, -30800,2.674545,3.8964734,,,,,,,,,,,,,, -30900,6.04399,3.9797688,,,,,,,,,,,,,, -31000,3.9037101,3.9740214,,,,,,,,,,,,,, -31100,6.759665,3.9689639,,,,,,,,,,,,,, -31200,5.3561325,3.872446,,,,,,,,,,,,,, -31300,4.22913,3.950346,,,,,,,,,,,,,, -31400,5.4625034,3.923957,,,,,,,,,,,,,, -31500,3.5583322,3.7323518,,,,,,,,,,,,,, -31600,3.913525,3.8800182,,,,,,,,,,,,,, -31700,6.7979503,3.9069986,,,,,,,,,,,,,, -31800,43.569485,4.001134,,,,,,,,,,,,,, -31900,3.2232947,3.862296,,,,,,,,,,,,,, -32000,9.973514,4.068392,,,,,,,,,,,,,, -32100,5.164908,3.9653573,,,,,,,,,,,,,, -32200,6.021162,3.9379249,,,,,,,,,,,,,, -32300,4.3055863,3.8091025,,,,,,,,,,,,,, -32400,4.115223,3.90779,,,,,,,,,,,,,, -32500,3.751346,3.6571703,,,,,,,,,,,,,, -32600,4.704988,3.9540267,,,,,,,,,,,,,, -32700,3.8842516,3.7744231,,,,,,,,,,,,,, -32800,12.891077,4.0785193,,,,,,,,,,,,,, -32900,4.596135,3.9260666,,,,,,,,,,,,,, -33000,5.231579,3.9262362,,,,,,,,,,,,,, -33100,3.3310785,3.8254242,,,,,,,,,,,,,, -33200,15.770643,4.123315,,,,,,,,,,,,,, -33300,5.116667,3.8323798,,,,,,,,,,,,,, -33400,6.238741,4.1003957,,,,,,,,,,,,,, -33500,4.1911044,3.9873357,,,,,,,,,,,,,, -33600,8.030036,3.9302359,,,,,,,,,,,,,, -33700,29.971622,4.0041637,,,,,,,,,,,,,, -33800,4.893659,3.9212613,,,,,,,,,,,,,, -33900,5.0590734,4.155683,,,,,,,,,,,,,, -34000,6.436087,3.9019113,,,,,,,,,,,,,, -34100,8.254065,3.8732133,,,,,,,,,,,,,, -34200,5.0181775,3.9908419,,,,,,,,,,,,,, -34300,4.086965,3.8376045,,,,,,,,,,,,,, -34400,5.838453,3.9188938,,,,,,,,,,,,,, -34500,3.3457096,3.80754,,,,,,,,,,,,,, -34600,4.9952664,3.8173242,,,,,,,,,,,,,, -34700,7.9723835,3.9199867,,,,,,,,,,,,,, -34800,9.7428055,4.0741715,,,,,,,,,,,,,, -34900,3.4913144,3.8093839,,,,,,,,,,,,,, -34981,,,0.0018534757,8.153365,0.00166,8.154742,50000.0,0.002,8.15415,10000.0,14034.850373506546,14649.07592201233,14034.850373506546,613.6692273616791,0.2397420406341552,0.0 -35000,3.725205,3.8181734,,,,,,,,,,,,,, -35100,4.1176467,3.8415713,,,,,,,,,,,,,, -35200,3.2027967,3.7598162,,,,,,,,,,,,,, -35300,3.222495,3.9731207,,,,,,,,,,,,,, -35400,5.2092566,3.9930968,,,,,,,,,,,,,, -35500,6.290946,3.8703933,,,,,,,,,,,,,, -35600,7.147085,3.795021,,,,,,,,,,,,,, -35700,3.3251483,3.7409275,,,,,,,,,,,,,, -35800,5.8374767,3.8579075,,,,,,,,,,,,,, -35900,8.081065,4.0949206,,,,,,,,,,,,,, -36000,4.552894,3.8279955,,,,,,,,,,,,,, -36100,7.85722,3.9836643,,,,,,,,,,,,,, -36200,4.099224,3.9244785,,,,,,,,,,,,,, -36300,5.8328805,3.9188013,,,,,,,,,,,,,, -36400,3.3009632,3.7246418,,,,,,,,,,,,,, -36500,6.306325,3.824336,,,,,,,,,,,,,, -36600,5.5483522,4.0186877,,,,,,,,,,,,,, -36700,4.460649,3.90034,,,,,,,,,,,,,, -36800,4.213024,3.8723621,,,,,,,,,,,,,, -36900,3.9558437,3.7579007,,,,,,,,,,,,,, -37000,3.4334342,3.8330417,,,,,,,,,,,,,, -37100,4.3328915,3.8027668,,,,,,,,,,,,,, -37200,6.3554034,3.8227448,,,,,,,,,,,,,, -37300,3.4918911,3.902834,,,,,,,,,,,,,, -37400,4.307879,3.9332376,,,,,,,,,,,,,, -37500,5.1152515,3.897512,,,,,,,,,,,,,, -37600,5.0823617,3.741316,,,,,,,,,,,,,, -37700,4.148218,3.720445,,,,,,,,,,,,,, -37800,5.041351,3.9966173,,,,,,,,,,,,,, -37900,4.9701896,3.7904515,,,,,,,,,,,,,, -38000,4.687297,3.9169598,,,,,,,,,,,,,, -38100,10.095328,3.881401,,,,,,,,,,,,,, -38200,4.370095,3.7748427,,,,,,,,,,,,,, -38300,7.689476,3.8940072,,,,,,,,,,,,,, -38400,3.9452763,3.9027624,,,,,,,,,,,,,, -38500,4.0301948,3.8300538,,,,,,,,,,,,,, -38600,3.3662624,3.7661119,,,,,,,,,,,,,, -38700,4.8496184,3.842304,,,,,,,,,,,,,, -38800,4.880418,3.9234312,,,,,,,,,,,,,, -38900,5.4037447,3.776256,,,,,,,,,,,,,, -39000,5.294405,3.7983108,,,,,,,,,,,,,, -39100,5.6256614,3.8836744,,,,,,,,,,,,,, -39200,8.314431,3.8155646,,,,,,,,,,,,,, -39300,8.624082,3.770013,,,,,,,,,,,,,, -39400,4.9023423,3.71177,,,,,,,,,,,,,, -39500,5.408239,3.854054,,,,,,,,,,,,,, -39600,4.6616464,3.8548274,,,,,,,,,,,,,, -39688,,,0.0020328443,8.241737,0.00196,8.220464,50000.0,0.0018000001,8.218177,10000.0,16031.259686946869,16726.439479112625,16031.259686946869,694.5466666221619,0.2730488777160644,0.0 -39700,5.738367,3.773004,,,,,,,,,,,,,, -39800,5.091378,3.8936028,,,,,,,,,,,,,, -39900,3.8261428,3.8260934,,,,,,,,,,,,,, -40000,6.3494725,3.8886676,,,,,,,,,,,,,, -40100,2.5586007,3.7873971,,,,,,,,,,,,,, -40200,4.754927,3.7953649,,,,,,,,,,,,,, -40300,3.9491239,3.8576734,,,,,,,,,,,,,, -40400,5.3736057,3.8455992,,,,,,,,,,,,,, -40500,3.3648274,3.8079717,,,,,,,,,,,,,, -40600,3.0873358,3.8526113,,,,,,,,,,,,,, -40700,6.0015206,4.032993,,,,,,,,,,,,,, -40800,4.579732,3.7816277,,,,,,,,,,,,,, -40900,5.601803,3.90361,,,,,,,,,,,,,, -41000,3.840319,3.6631515,,,,,,,,,,,,,, -41100,5.629751,3.6964374,,,,,,,,,,,,,, -41200,5.4698105,3.9539602,,,,,,,,,,,,,, -41300,4.2248483,3.8048253,,,,,,,,,,,,,, -41400,3.8463855,3.8221073,,,,,,,,,,,,,, -41500,3.9706495,3.9117174,,,,,,,,,,,,,, -41600,4.6688204,3.8157606,,,,,,,,,,,,,, -41700,3.5994647,3.8129673,,,,,,,,,,,,,, -41800,4.705359,3.7153249,,,,,,,,,,,,,, -41900,4.423021,3.829566,,,,,,,,,,,,,, -42000,4.7451224,3.910371,,,,,,,,,,,,,, -42100,4.494183,3.7962198,,,,,,,,,,,,,, -42200,6.2629266,3.689824,,,,,,,,,,,,,, -42300,2.985189,3.7441576,,,,,,,,,,,,,, -42400,3.9152787,3.8090749,,,,,,,,,,,,,, -42500,4.015066,3.7186825,,,,,,,,,,,,,, -42600,3.1087058,3.7514994,,,,,,,,,,,,,, -42700,8.181146,3.8451028,,,,,,,,,,,,,, -42800,5.713146,3.8196056,,,,,,,,,,,,,, -42900,10.422153,3.7484965,,,,,,,,,,,,,, -43000,4.037586,3.7140613,,,,,,,,,,,,,, -43100,6.0442543,4.029402,,,,,,,,,,,,,, -43200,4.870251,3.7339015,,,,,,,,,,,,,, -43300,4.9496794,3.6359887,,,,,,,,,,,,,, -43400,4.7664447,3.888214,,,,,,,,,,,,,, -43500,5.8735533,3.8929005,,,,,,,,,,,,,, -43600,5.140307,3.8528392,,,,,,,,,,,,,, -43700,3.934191,3.816639,,,,,,,,,,,,,, -43800,4.757514,3.7692919,,,,,,,,,,,,,, -43900,3.8450017,3.7140095,,,,,,,,,,,,,, -44000,2.8316438,3.8117716,,,,,,,,,,,,,, -44100,4.500588,3.7888203,,,,,,,,,,,,,, -44200,3.77068,3.8026738,,,,,,,,,,,,,, -44300,3.2905622,3.719396,,,,,,,,,,,,,, -44400,4.76021,3.6966324,,,,,,,,,,,,,, -44428,,,0.001434949,8.262915,0.0018399999,8.247698,50000.0,0.0015,8.245352,10000.0,18027.405270814896,18798.586426973343,18027.405270814896,770.4708964824677,0.3067083358764648,0.0 -44500,4.4958463,3.7190838,,,,,,,,,,,,,, -44600,3.3361692,3.7332382,,,,,,,,,,,,,, -44700,6.689782,3.9469922,,,,,,,,,,,,,, -44800,3.378697,3.829051,,,,,,,,,,,,,, -44900,4.061113,3.7271245,,,,,,,,,,,,,, -45000,5.553673,3.8520184,,,,,,,,,,,,,, -45100,7.022553,3.7395623,,,,,,,,,,,,,, -45200,4.5973434,3.7561932,,,,,,,,,,,,,, -45300,5.6338344,3.81748,,,,,,,,,,,,,, -45400,3.5725238,3.7917628,,,,,,,,,,,,,, -45500,6.70774,3.705049,,,,,,,,,,,,,, -45600,4.136595,3.8933012,,,,,,,,,,,,,, -45700,4.269683,3.7553263,,,,,,,,,,,,,, -45800,3.559002,3.6119606,,,,,,,,,,,,,, -45900,3.9800136,3.8060846,,,,,,,,,,,,,, -46000,5.036452,3.7876766,,,,,,,,,,,,,, -46100,4.7088304,3.6907206,,,,,,,,,,,,,, -46200,3.013004,3.749632,,,,,,,,,,,,,, -46300,5.657477,3.8257113,,,,,,,,,,,,,, -46400,3.1413274,3.8631847,,,,,,,,,,,,,, -46500,4.333033,3.607487,,,,,,,,,,,,,, -46600,5.9079123,4.0293097,,,,,,,,,,,,,, -46700,5.030555,3.8685927,,,,,,,,,,,,,, -46800,6.7889166,3.7499352,,,,,,,,,,,,,, -46900,3.8781905,3.6512356,,,,,,,,,,,,,, -47000,5.0112987,3.8031,,,,,,,,,,,,,, -47100,4.3585315,3.7042851,,,,,,,,,,,,,, -47200,4.522954,3.6638951,,,,,,,,,,,,,, -47300,5.3607135,3.6689029,,,,,,,,,,,,,, -47400,4.054584,3.6959949,,,,,,,,,,,,,, -47500,3.3621557,3.770347,,,,,,,,,,,,,, -47600,3.6946125,3.7792609,,,,,,,,,,,,,, -47700,4.776403,4.095537,,,,,,,,,,,,,, -47800,6.280883,3.8142836,,,,,,,,,,,,,, -47900,4.5244503,3.6844974,,,,,,,,,,,,,, -48000,5.259481,3.792388,,,,,,,,,,,,,, -48100,3.2222986,3.7685966,,,,,,,,,,,,,, -48200,4.037337,3.862459,,,,,,,,,,,,,, -48300,4.1058645,3.724509,,,,,,,,,,,,,, -48400,4.117225,3.7693763,,,,,,,,,,,,,, -48500,5.883688,3.8842068,,,,,,,,,,,,,, -48600,3.872678,3.7335644,,,,,,,,,,,,,, -48700,4.875818,3.8303635,,,,,,,,,,,,,, -48800,4.4876447,3.7771163,,,,,,,,,,,,,, -48900,5.419809,3.6684613,,,,,,,,,,,,,, -49000,3.9664187,3.820809,,,,,,,,,,,,,, -49100,6.176051,3.9147296,,,,,,,,,,,,,, -49135,,,0.0016342474,8.264104,0.0017599999,8.257346,50000.0,0.0013000001,8.255711,10000.0,20023.873562812805,20894.611929893494,20023.873562812805,869.9464857578278,0.3443114757537842,0.0 -49200,5.3658414,3.8717103,,,,,,,,,,,,,, -49300,7.160287,3.8151584,,,,,,,,,,,,,, -49400,4.511857,3.821358,,,,,,,,,,,,,, -49500,5.5613055,3.8931565,,,,,,,,,,,,,, -49600,7.927204,3.9708521,,,,,,,,,,,,,, -49700,5.550275,3.816207,,,,,,,,,,,,,, -49800,13.793541,4.068351,,,,,,,,,,,,,, -49900,3.1742601,3.7585921,,,,,,,,,,,,,, -50000,6.1742935,3.9537544,,,,,,,,,,,,,, -50100,5.816035,3.688892,,,,,,,,,,,,,, -50200,6.294397,4.0017257,,,,,,,,,,,,,, -50300,5.034474,3.8405762,,,,,,,,,,,,,, -50400,4.528432,3.7026956,,,,,,,,,,,,,, -50500,5.817104,3.7765853,,,,,,,,,,,,,, -50600,5.0261707,3.6592016,,,,,,,,,,,,,, -50700,4.6259723,3.7639756,,,,,,,,,,,,,, -50800,4.592901,3.6707706,,,,,,,,,,,,,, -50900,2.920498,3.599956,,,,,,,,,,,,,, -51000,3.7707229,3.7798963,,,,,,,,,,,,,, -51100,3.6521912,3.7572396,,,,,,,,,,,,,, -51200,5.3303742,3.6340604,,,,,,,,,,,,,, -51300,4.9224415,3.5592957,,,,,,,,,,,,,, -51400,3.383927,3.7486959,,,,,,,,,,,,,, -51500,3.5161512,3.8142517,,,,,,,,,,,,,, -51600,3.336362,3.932082,,,,,,,,,,,,,, -51700,5.712367,4.0350304,,,,,,,,,,,,,, -51800,6.018396,3.6324112,,,,,,,,,,,,,, -51900,4.898832,3.7984817,,,,,,,,,,,,,, -52000,4.5093904,3.7492769,,,,,,,,,,,,,, -52100,6.878928,3.6713605,,,,,,,,,,,,,, -52200,7.803661,3.946499,,,,,,,,,,,,,, -52300,3.543654,3.7522478,,,,,,,,,,,,,, -52400,3.9371057,3.7056189,,,,,,,,,,,,,, -52500,4.671799,3.6870632,,,,,,,,,,,,,, -52600,8.388179,3.8928216,,,,,,,,,,,,,, -52700,6.4535336,3.8706117,,,,,,,,,,,,,, -52800,4.0045238,3.6131546,,,,,,,,,,,,,, -52900,6.0720224,3.6321619,,,,,,,,,,,,,, -53000,4.211054,3.7647564,,,,,,,,,,,,,, -53100,6.6874847,3.6929026,,,,,,,,,,,,,, -53200,5.5525956,3.9167283,,,,,,,,,,,,,, -53300,10.889021,3.983602,,,,,,,,,,,,,, -53400,4.9901485,3.8310883,,,,,,,,,,,,,, -53500,6.350291,3.6494613,,,,,,,,,,,,,, -53600,2.9172637,3.6889215,,,,,,,,,,,,,, -53700,5.431403,3.8153787,,,,,,,,,,,,,, -53800,5.747957,3.7604089,,,,,,,,,,,,,, -53896,,,0.0017139668,8.253426,0.00174,8.235537,50000.0,0.0012,8.235645,10000.0,22019.93824338913,22971.235975027084,22019.93824338913,950.4290370941162,0.3771963119506836,0.0 -53900,8.2547865,3.887939,,,,,,,,,,,,,, -54000,4.050254,3.7449822,,,,,,,,,,,,,, -54100,3.4184213,3.642449,,,,,,,,,,,,,, -54200,23.04608,3.8561249,,,,,,,,,,,,,, -54300,4.293613,3.6969109,,,,,,,,,,,,,, -54400,2.6232061,3.7131348,,,,,,,,,,,,,, -54500,5.22178,3.6465604,,,,,,,,,,,,,, -54600,3.2922797,3.712771,,,,,,,,,,,,,, -54700,3.7804375,3.8474088,,,,,,,,,,,,,, -54800,4.7113585,3.7810936,,,,,,,,,,,,,, -54900,6.9935584,3.8010979,,,,,,,,,,,,,, -55000,5.08542,3.7087271,,,,,,,,,,,,,, -55100,5.367065,3.7550082,,,,,,,,,,,,,, -55200,5.36235,3.7832875,,,,,,,,,,,,,, -55300,3.3240387,3.7667408,,,,,,,,,,,,,, -55400,5.192694,3.587033,,,,,,,,,,,,,, -55500,4.3675733,3.6223228,,,,,,,,,,,,,, -55600,7.9266405,3.9210904,,,,,,,,,,,,,, -55700,5.15066,3.7417364,,,,,,,,,,,,,, -55800,6.285863,3.8508568,,,,,,,,,,,,,, -55900,5.7810807,3.7156157,,,,,,,,,,,,,, -56000,5.723335,3.7640681,,,,,,,,,,,,,, -56100,6.50888,3.8194268,,,,,,,,,,,,,, -56200,2.6089087,3.8350096,,,,,,,,,,,,,, -56300,3.257226,3.6177363,,,,,,,,,,,,,, -56400,17.74145,4.509633,,,,,,,,,,,,,, -56500,4.025693,3.7712443,,,,,,,,,,,,,, -56600,3.49037,3.707316,,,,,,,,,,,,,, -56700,3.7792127,3.6467433,,,,,,,,,,,,,, -56800,3.1950445,3.666729,,,,,,,,,,,,,, -56900,4.512518,3.7533352,,,,,,,,,,,,,, -57000,9.118009,3.7812755,,,,,,,,,,,,,, -57100,4.3791356,3.7261348,,,,,,,,,,,,,, -57200,4.042726,3.850298,,,,,,,,,,,,,, -57300,5.2832785,3.8037422,,,,,,,,,,,,,, -57400,3.7623196,3.739806,,,,,,,,,,,,,, -57500,3.3871002,3.629765,,,,,,,,,,,,,, -57600,3.869986,3.64782,,,,,,,,,,,,,, -57700,5.1228933,3.7367048,,,,,,,,,,,,,, -57800,3.3173306,3.7255926,,,,,,,,,,,,,, -57900,3.6714208,3.729186,,,,,,,,,,,,,, -58000,10.119244,4.0623727,,,,,,,,,,,,,, -58100,5.583807,3.7240503,,,,,,,,,,,,,, -58200,4.453329,3.604529,,,,,,,,,,,,,, -58300,6.116827,3.790073,,,,,,,,,,,,,, -58400,3.2840781,3.7023377,,,,,,,,,,,,,, -58500,4.162328,3.7367325,,,,,,,,,,,,,, -58600,4.8851166,3.7628174,,,,,,,,,,,,,, -58654,,,0.001694037,8.188425,0.0016999999,8.179385,50000.0,0.0014000001,8.181261,10000.0,24016.297024965286,25037.01320934296,24016.297024965286,1019.7672593593596,0.4138531684875488,0.0 -58700,11.201581,3.8576274,,,,,,,,,,,,,, -58800,6.1667876,3.908225,,,,,,,,,,,,,, -58900,4.3647437,3.7510939,,,,,,,,,,,,,, -59000,2.83936,3.782693,,,,,,,,,,,,,, -59100,4.0229163,3.5408344,,,,,,,,,,,,,, -59200,4.412282,3.6593525,,,,,,,,,,,,,, -59300,3.9784074,3.7265763,,,,,,,,,,,,,, -59400,3.785562,3.831666,,,,,,,,,,,,,, -59500,3.1105008,3.6872025,,,,,,,,,,,,,, -59600,8.067553,3.7192867,,,,,,,,,,,,,, -59700,7.8072033,3.8594031,,,,,,,,,,,,,, -59800,3.0368054,3.7472112,,,,,,,,,,,,,, -59900,6.2840085,3.9188628,,,,,,,,,,,,,, -60000,5.4949694,3.7369432,,,,,,,,,,,,,, -60100,7.1960325,3.6609,,,,,,,,,,,,,, -60200,4.5695915,3.8068452,,,,,,,,,,,,,, -60300,4.449434,3.725252,,,,,,,,,,,,,, -60400,5.508068,3.7967577,,,,,,,,,,,,,, -60500,8.081078,3.9384332,,,,,,,,,,,,,, -60600,4.1946535,3.9076388,,,,,,,,,,,,,, -60700,4.542024,3.720483,,,,,,,,,,,,,, -60800,24.939323,3.8733072,,,,,,,,,,,,,, -60900,4.585266,3.7948785,,,,,,,,,,,,,, -61000,4.1284704,3.8336833,,,,,,,,,,,,,, -61100,3.1275475,3.5531368,,,,,,,,,,,,,, -61200,7.289842,3.7774649,,,,,,,,,,,,,, -61300,5.13781,3.61988,,,,,,,,,,,,,, -61400,5.7867785,3.747478,,,,,,,,,,,,,, -61500,3.387427,3.7106059,,,,,,,,,,,,,, -61600,3.310374,3.7420902,,,,,,,,,,,,,, -61700,5.4570603,3.7426698,,,,,,,,,,,,,, -61800,3.1062095,3.644948,,,,,,,,,,,,,, -61900,7.4637523,3.8270926,,,,,,,,,,,,,, -62000,4.6428843,3.8305101,,,,,,,,,,,,,, -62100,6.257846,3.8697796,,,,,,,,,,,,,, -62200,5.086574,3.753459,,,,,,,,,,,,,, -62300,3.2557068,3.7122712,,,,,,,,,,,,,, -62400,12.972366,3.8092206,,,,,,,,,,,,,, -62500,5.8288913,3.883705,,,,,,,,,,,,,, -62600,5.3443694,3.5373797,,,,,,,,,,,,,, -62700,4.4532585,3.603692,,,,,,,,,,,,,, -62800,3.8442364,3.6976507,,,,,,,,,,,,,, -62900,4.5232034,3.6880288,,,,,,,,,,,,,, -63000,3.969894,3.8623176,,,,,,,,,,,,,, -63100,5.4277024,3.7538757,,,,,,,,,,,,,, -63200,6.095492,3.6877685,,,,,,,,,,,,,, -63300,5.2611756,3.6736197,,,,,,,,,,,,,, -63400,4.93059,3.8247552,,,,,,,,,,,,,, -63500,6.38839,3.785384,,,,,,,,,,,,,, -63600,3.6646507,3.6697369,,,,,,,,,,,,,, -63700,6.811043,3.893231,,,,,,,,,,,,,, -63800,4.4329367,3.8672183,,,,,,,,,,,,,, -63900,3.641994,3.7438078,,,,,,,,,,,,,, -64000,4.290968,3.7717102,,,,,,,,,,,,,, -64006,,,0.0017936862,8.101958,0.00168,8.089027,50000.0,0.0016000001,8.0927105,10000.0,26012.28999519348,27101.98451256752,26012.28999519348,1088.661203622818,0.4541494846343994,0.0 -64100,4.03059,3.7392213,,,,,,,,,,,,,, -64200,3.826038,3.687376,,,,,,,,,,,,,, -64300,3.8314059,3.6277883,,,,,,,,,,,,,, -64400,5.7069654,3.8561158,,,,,,,,,,,,,, -64500,6.9179378,4.0117097,,,,,,,,,,,,,, -64600,5.23885,3.7526627,,,,,,,,,,,,,, -64700,4.6671114,3.9492297,,,,,,,,,,,,,, -64800,3.566571,3.7210023,,,,,,,,,,,,,, -64900,4.719355,3.522327,,,,,,,,,,,,,, -65000,3.8931165,3.8704774,,,,,,,,,,,,,, -65100,3.4531975,3.67401,,,,,,,,,,,,,, -65200,5.6988673,3.7660434,,,,,,,,,,,,,, -65300,9.512504,3.9528341,,,,,,,,,,,,,, -65400,3.464954,3.7395873,,,,,,,,,,,,,, -65500,2.9016626,3.737907,,,,,,,,,,,,,, -65600,3.5644042,3.6590564,,,,,,,,,,,,,, -65700,4.454152,3.68656,,,,,,,,,,,,,, -65800,18.038452,3.7878706,,,,,,,,,,,,,, -65900,4.6601286,3.6610985,,,,,,,,,,,,,, -66000,4.548684,3.6421475,,,,,,,,,,,,,, -66100,7.015832,3.9274187,,,,,,,,,,,,,, -66200,3.0047283,3.741171,,,,,,,,,,,,,, -66300,4.24214,3.6540935,,,,,,,,,,,,,, -66400,19.151278,3.7354596,,,,,,,,,,,,,, -66500,4.373426,3.7781403,,,,,,,,,,,,,, -66600,34.828712,3.7973447,,,,,,,,,,,,,, -66700,4.487275,3.849683,,,,,,,,,,,,,, -66800,6.110285,3.8515377,,,,,,,,,,,,,, -66900,3.8985333,3.7125442,,,,,,,,,,,,,, -67000,4.0154867,3.8387241,,,,,,,,,,,,,, -67100,3.7285028,3.6013346,,,,,,,,,,,,,, -67200,8.064487,3.8656154,,,,,,,,,,,,,, -67300,2.764653,3.6415305,,,,,,,,,,,,,, -67400,4.752501,3.8310046,,,,,,,,,,,,,, -67500,13.809174,4.1034575,,,,,,,,,,,,,, -67600,6.7908216,3.8764968,,,,,,,,,,,,,, -67700,6.6559215,3.8610497,,,,,,,,,,,,,, -67800,4.6885157,3.7705617,,,,,,,,,,,,,, -67900,10.376777,3.7260685,,,,,,,,,,,,,, -68000,3.272739,3.7654681,,,,,,,,,,,,,, -68100,6.9552503,3.8168383,,,,,,,,,,,,,, -68200,4.9848127,3.650938,,,,,,,,,,,,,, -68300,7.1247673,3.7664585,,,,,,,,,,,,,, -68400,3.5716617,3.723964,,,,,,,,,,,,,, -68500,4.700095,3.7076175,,,,,,,,,,,,,, -68600,6.5638614,3.859867,,,,,,,,,,,,,, -68700,6.6468334,3.8042753,,,,,,,,,,,,,, -68800,5.0861216,3.7159767,,,,,,,,,,,,,, -68900,4.621492,3.751352,,,,,,,,,,,,,, -69000,3.9854782,3.7047172,,,,,,,,,,,,,, -69100,3.988013,3.7583046,,,,,,,,,,,,,, -69149,,,0.0015943877,8.021358,0.00172,8.019661,50000.0,0.0016000001,8.025281,10000.0,28008.3186814785,29175.41971540451,28008.3186814785,1165.955510854721,0.5219497680664062,0.0 -69200,5.306702,3.754941,,,,,,,,,,,,,, -69300,4.08526,3.6303535,,,,,,,,,,,,,, -69400,18.792742,3.789513,,,,,,,,,,,,,, -69500,5.1767235,3.9641402,,,,,,,,,,,,,, -69600,3.6631536,3.7188802,,,,,,,,,,,,,, -69700,6.028405,3.9732318,,,,,,,,,,,,,, -69800,3.8252249,3.7279024,,,,,,,,,,,,,, -69900,7.379957,3.7672143,,,,,,,,,,,,,, -70000,4.9746027,3.7026181,,,,,,,,,,,,,, -70100,5.4211082,3.6874342,,,,,,,,,,,,,, -70200,5.935713,3.8266892,,,,,,,,,,,,,, -70300,3.9207327,3.6833205,,,,,,,,,,,,,, -70400,4.3597083,3.6687775,,,,,,,,,,,,,, -70500,4.8332796,3.854022,,,,,,,,,,,,,, -70600,2.8327942,3.691423,,,,,,,,,,,,,, -70700,4.827691,3.8247018,,,,,,,,,,,,,, -70800,3.6906624,3.6557286,,,,,,,,,,,,,, -70900,4.8639016,3.745768,,,,,,,,,,,,,, -71000,6.0107765,3.6589684,,,,,,,,,,,,,, -71100,3.5852568,3.7620132,,,,,,,,,,,,,, -71200,3.3434498,3.6127138,,,,,,,,,,,,,, -71300,7.28145,3.7989526,,,,,,,,,,,,,, -71400,5.1092415,3.870923,,,,,,,,,,,,,, -71500,3.305421,3.6970563,,,,,,,,,,,,,, -71600,11.261924,3.888277,,,,,,,,,,,,,, -71700,4.7061434,3.6927242,,,,,,,,,,,,,, -71800,4.7104845,3.724838,,,,,,,,,,,,,, -71900,4.517913,3.7259455,,,,,,,,,,,,,, -72000,5.608821,3.7892253,,,,,,,,,,,,,, -72100,5.4255047,3.5181646,,,,,,,,,,,,,, -72200,4.031135,3.7068253,,,,,,,,,,,,,, -72300,4.0281034,3.6361604,,,,,,,,,,,,,, -72400,3.7963147,3.6341317,,,,,,,,,,,,,, -72500,6.4693685,3.7034178,,,,,,,,,,,,,, -72600,4.9265127,3.8012738,,,,,,,,,,,,,, -72700,4.69615,3.6061344,,,,,,,,,,,,,, -72800,5.65134,3.8129778,,,,,,,,,,,,,, -72900,5.109966,3.6676354,,,,,,,,,,,,,, -73000,8.042091,3.9227042,,,,,,,,,,,,,, -73100,3.8468657,3.9130213,,,,,,,,,,,,,, -73200,4.864156,3.6076195,,,,,,,,,,,,,, -73300,3.9252834,3.836092,,,,,,,,,,,,,, -73400,3.7207236,3.6526165,,,,,,,,,,,,,, -73500,4.1928473,3.8666072,,,,,,,,,,,,,, -73600,3.4965532,3.704314,,,,,,,,,,,,,, -73700,5.111949,3.8142,,,,,,,,,,,,,, -73800,4.5913997,3.7854831,,,,,,,,,,,,,, -73900,20.22457,3.8525372,,,,,,,,,,,,,, -74000,5.5176306,3.7346404,,,,,,,,,,,,,, -74083,,,0.0017338967,7.922397,0.0017799999,7.906208,50000.0,0.0016000001,7.9134455,10000.0,30004.49159431457,31243.70892095565,30004.49159431457,1237.9831821918488,0.5660214424133301,0.0 -74100,3.6965876,3.8169959,,,,,,,,,,,,,, -74200,3.9310112,3.6546402,,,,,,,,,,,,,, -74300,5.118127,3.721345,,,,,,,,,,,,,, -74400,8.482264,3.7857862,,,,,,,,,,,,,, -74500,3.7454543,3.7504117,,,,,,,,,,,,,, -74600,5.7869554,3.666864,,,,,,,,,,,,,, -74700,4.5513244,3.9294753,,,,,,,,,,,,,, -74800,2.746604,3.7471697,,,,,,,,,,,,,, -74900,4.2196574,3.7441413,,,,,,,,,,,,,, -75000,17.69132,3.6279054,,,,,,,,,,,,,, -75100,3.8422835,3.7112203,,,,,,,,,,,,,, -75200,4.2522106,3.7691536,,,,,,,,,,,,,, -75300,5.1085796,3.7320158,,,,,,,,,,,,,, -75400,4.3232884,3.7169309,,,,,,,,,,,,,, -75500,8.185216,3.797991,,,,,,,,,,,,,, -75600,3.0675159,3.8118052,,,,,,,,,,,,,, -75700,4.543264,3.6588106,,,,,,,,,,,,,, -75800,3.8403318,3.8873868,,,,,,,,,,,,,, -75900,3.8007188,3.66755,,,,,,,,,,,,,, -76000,5.2328486,3.8331556,,,,,,,,,,,,,, -76100,4.129122,3.68244,,,,,,,,,,,,,, -76200,5.218644,3.8238728,,,,,,,,,,,,,, -76300,5.487956,3.9197707,,,,,,,,,,,,,, -76400,6.472754,3.8573077,,,,,,,,,,,,,, -76500,5.157866,3.585194,,,,,,,,,,,,,, -76600,7.1468635,4.0143666,,,,,,,,,,,,,, -76700,6.5523233,3.7112405,,,,,,,,,,,,,, -76800,5.857106,3.854035,,,,,,,,,,,,,, -76900,4.97733,3.7486687,,,,,,,,,,,,,, -77000,3.442297,3.6275914,,,,,,,,,,,,,, -77100,4.839957,3.6837785,,,,,,,,,,,,,, -77200,4.6235127,3.7830105,,,,,,,,,,,,,, -77300,5.9867287,3.80227,,,,,,,,,,,,,, -77400,2.097766,3.6201582,,,,,,,,,,,,,, -77500,7.8115273,3.8525956,,,,,,,,,,,,,, -77600,4.759082,3.7891784,,,,,,,,,,,,,, -77700,4.0051374,3.806543,,,,,,,,,,,,,, -77800,3.459436,3.6744864,,,,,,,,,,,,,, -77900,5.6197267,3.7515533,,,,,,,,,,,,,, -78000,4.0651646,3.7438774,,,,,,,,,,,,,, -78100,6.5018306,3.8045645,,,,,,,,,,,,,, -78200,4.1934614,3.775304,,,,,,,,,,,,,, -78300,3.427759,3.7099495,,,,,,,,,,,,,, -78400,5.60765,3.7451289,,,,,,,,,,,,,, -78500,18.030796,3.833187,,,,,,,,,,,,,, -78600,3.556237,3.680437,,,,,,,,,,,,,, -78700,4.853753,3.8841105,,,,,,,,,,,,,, -78800,11.598163,3.681364,,,,,,,,,,,,,, -78900,3.5925112,3.6197612,,,,,,,,,,,,,, -78998,,,0.0020129145,7.8298244,0.0019999999,7.8241224,50000.0,0.0018000001,7.8329372,10000.0,32000.55635213852,33306.56731748581,32000.55635213852,1304.693614244461,0.6050024032592773,0.0 -79000,4.0211067,3.7101815,,,,,,,,,,,,,, -79100,4.4491353,3.824151,,,,,,,,,,,,,, -79200,4.0453434,3.7108495,,,,,,,,,,,,,, -79300,4.000266,3.6974456,,,,,,,,,,,,,, -79400,3.6480021,3.6287162,,,,,,,,,,,,,, -79500,5.443443,3.9547458,,,,,,,,,,,,,, -79600,5.32156,3.9649365,,,,,,,,,,,,,, -79700,9.374236,3.7234735,,,,,,,,,,,,,, -79800,5.1313076,3.6569352,,,,,,,,,,,,,, -79900,3.6012301,3.6748557,,,,,,,,,,,,,, -80000,4.468842,3.5644252,,,,,,,,,,,,,, -80100,5.765341,3.7199028,,,,,,,,,,,,,, -80200,5.4640336,3.8716712,,,,,,,,,,,,,, -80300,3.8925505,3.7248075,,,,,,,,,,,,,, -80400,3.8439584,3.578797,,,,,,,,,,,,,, -80500,4.649687,3.8866136,,,,,,,,,,,,,, -80600,3.8016524,3.7550263,,,,,,,,,,,,,, -80700,3.3088977,3.6975558,,,,,,,,,,,,,, -80800,3.0811157,3.8028116,,,,,,,,,,,,,, -80900,3.7506456,3.696291,,,,,,,,,,,,,, -81000,4.108608,3.6531448,,,,,,,,,,,,,, -81100,4.126156,3.6517215,,,,,,,,,,,,,, -81200,7.018788,3.7769496,,,,,,,,,,,,,, -81300,4.0605087,3.6988864,,,,,,,,,,,,,, -81400,3.6833708,3.7389665,,,,,,,,,,,,,, -81500,3.8807416,3.650499,,,,,,,,,,,,,, -81600,4.3163033,3.7095718,,,,,,,,,,,,,, -81700,4.9549804,3.8093486,,,,,,,,,,,,,, -81800,3.854112,3.7295532,,,,,,,,,,,,,, -81900,2.5804994,3.6572938,,,,,,,,,,,,,, -82000,15.087785,3.639728,,,,,,,,,,,,,, -82100,8.313381,3.8191564,,,,,,,,,,,,,, -82200,4.265251,3.7999153,,,,,,,,,,,,,, -82300,5.5458493,3.7700853,,,,,,,,,,,,,, -82400,3.4135382,3.7258217,,,,,,,,,,,,,, -82500,3.7948663,3.6742144,,,,,,,,,,,,,, -82600,7.2945714,3.748433,,,,,,,,,,,,,, -82700,4.4718003,3.7154083,,,,,,,,,,,,,, -82800,9.751736,4.11737,,,,,,,,,,,,,, -82900,8.06493,3.8888378,,,,,,,,,,,,,, -83000,5.980417,3.7389822,,,,,,,,,,,,,, -83100,4.623166,3.7919197,,,,,,,,,,,,,, -83200,4.530367,3.702111,,,,,,,,,,,,,, -83300,4.380783,3.7214885,,,,,,,,,,,,,, -83400,4.154782,3.698268,,,,,,,,,,,,,, -83500,3.7099662,3.7450035,,,,,,,,,,,,,, -83600,2.187424,3.6890948,,,,,,,,,,,,,, -83700,3.9286623,3.7358787,,,,,,,,,,,,,, -83800,3.6030777,3.8510852,,,,,,,,,,,,,, -83900,5.400223,3.6882818,,,,,,,,,,,,,, -84000,4.0204525,3.761488,,,,,,,,,,,,,, -84001,,,0.001953125,7.742617,0.00208,7.741793,50000.0,0.0019,7.752259,10000.0,33997.315504550934,35389.42226982117,33997.315504550934,1390.706758737564,0.6433432102203369,0.0 -84100,7.5532837,3.7580602,,,,,,,,,,,,,, -84200,3.9432058,3.7173567,,,,,,,,,,,,,, -84300,5.370383,3.750895,,,,,,,,,,,,,, -84400,4.0891137,3.8355253,,,,,,,,,,,,,, -84500,5.105422,3.7051065,,,,,,,,,,,,,, -84600,3.1101549,3.6448324,,,,,,,,,,,,,, -84700,3.4435122,3.7893312,,,,,,,,,,,,,, -84800,6.47303,3.9054017,,,,,,,,,,,,,, -84900,4.874751,3.818531,,,,,,,,,,,,,, -85000,3.6587787,3.745954,,,,,,,,,,,,,, -85100,5.4139752,3.7489014,,,,,,,,,,,,,, -85200,4.5998344,3.73968,,,,,,,,,,,,,, -85300,5.4846654,3.5675845,,,,,,,,,,,,,, -85400,3.7795837,3.7404718,,,,,,,,,,,,,, -85500,3.9007351,3.7400858,,,,,,,,,,,,,, -85600,4.1185427,3.9590888,,,,,,,,,,,,,, -85700,5.2974415,3.9031549,,,,,,,,,,,,,, -85800,32.51614,3.9326203,,,,,,,,,,,,,, -85900,4.045722,3.718803,,,,,,,,,,,,,, -86000,4.867931,3.6247306,,,,,,,,,,,,,, -86100,3.8269339,3.7825634,,,,,,,,,,,,,, -86200,10.420803,3.6182625,,,,,,,,,,,,,, -86300,4.942003,3.83794,,,,,,,,,,,,,, -86400,3.6761374,3.8127117,,,,,,,,,,,,,, -86500,3.9985268,3.6732237,,,,,,,,,,,,,, -86600,4.8697395,3.7111259,,,,,,,,,,,,,, -86700,3.0467222,3.7269819,,,,,,,,,,,,,, -86800,4.3500204,3.6749876,,,,,,,,,,,,,, -86900,3.87247,3.8380852,,,,,,,,,,,,,, -87000,5.2048407,3.782299,,,,,,,,,,,,,, -87100,5.1716237,3.999425,,,,,,,,,,,,,, -87200,4.018753,3.7867396,,,,,,,,,,,,,, -87300,6.306531,3.6776438,,,,,,,,,,,,,, -87400,4.656188,3.7544003,,,,,,,,,,,,,, -87500,4.4258246,3.8112988,,,,,,,,,,,,,, -87600,4.473336,3.8432086,,,,,,,,,,,,,, -87700,3.483896,3.6533632,,,,,,,,,,,,,, -87800,3.2544804,3.786353,,,,,,,,,,,,,, -87900,2.7823796,3.723776,,,,,,,,,,,,,, -88000,4.171748,3.7442374,,,,,,,,,,,,,, -88100,5.7823634,3.726451,,,,,,,,,,,,,, -88200,3.291215,3.718237,,,,,,,,,,,,,, -88300,10.135679,3.8133903,,,,,,,,,,,,,, -88400,4.347157,3.6833625,,,,,,,,,,,,,, -88500,13.583102,3.8161712,,,,,,,,,,,,,, -88600,5.2135277,3.8343408,,,,,,,,,,,,,, -88700,5.157245,3.6808124,,,,,,,,,,,,,, -88800,3.9589875,3.6379862,,,,,,,,,,,,,, -88900,5.9786267,3.819407,,,,,,,,,,,,,, -89000,3.4327266,3.7041063,,,,,,,,,,,,,, -89100,4.097994,3.8290226,,,,,,,,,,,,,, -89200,4.7462044,3.814223,,,,,,,,,,,,,, -89270,,,0.0020527742,7.701437,0.00222,7.7050667,50000.0,0.0019,7.7178745,10000.0,35993.23620557785,37445.64071536064,35993.23620557785,1450.920217514038,0.6834356784820557,0.0 -89300,3.850599,3.6804526,,,,,,,,,,,,,, -89400,2.8794518,3.7205637,,,,,,,,,,,,,, -89500,4.0694456,3.7814646,,,,,,,,,,,,,, -89600,3.0679464,3.8672256,,,,,,,,,,,,,, -89700,3.9825113,3.9328063,,,,,,,,,,,,,, -89800,3.7572145,3.8066936,,,,,,,,,,,,,, -89900,2.7074008,3.7679105,,,,,,,,,,,,,, -90000,5.139957,3.7176876,,,,,,,,,,,,,, -90100,2.9498293,3.7700586,,,,,,,,,,,,,, -90200,6.139378,3.7961407,,,,,,,,,,,,,, -90300,5.150022,3.7076917,,,,,,,,,,,,,, -90400,2.9947093,3.6556406,,,,,,,,,,,,,, -90500,5.591922,3.8734381,,,,,,,,,,,,,, -90600,4.661605,3.7834394,,,,,,,,,,,,,, -90700,3.029676,3.6389909,,,,,,,,,,,,,, -90800,5.1798496,3.7577572,,,,,,,,,,,,,, -90900,9.363433,3.6611297,,,,,,,,,,,,,, -91000,4.752153,3.7161677,,,,,,,,,,,,,, -91100,4.9372573,3.7749457,,,,,,,,,,,,,, -91200,7.168136,3.6796145,,,,,,,,,,,,,, -91300,3.3876786,3.9478106,,,,,,,,,,,,,, -91400,4.504289,3.6985111,,,,,,,,,,,,,, -91500,5.968009,3.797285,,,,,,,,,,,,,, -91600,8.734096,3.7522974,,,,,,,,,,,,,, -91700,6.9926353,3.727034,,,,,,,,,,,,,, -91800,5.626188,3.8925586,,,,,,,,,,,,,, -91900,4.888878,3.803012,,,,,,,,,,,,,, -92000,6.231159,3.7174501,,,,,,,,,,,,,, -92100,3.919881,3.838605,,,,,,,,,,,,,, -92200,6.8617024,3.7765038,,,,,,,,,,,,,, -92300,5.418927,3.7680013,,,,,,,,,,,,,, -92400,4.013881,3.8469505,,,,,,,,,,,,,, -92500,2.4908092,3.5705185,,,,,,,,,,,,,, -92600,4.8931417,3.9030793,,,,,,,,,,,,,, -92700,3.648907,3.5375795,,,,,,,,,,,,,, -92800,4.8440037,3.716641,,,,,,,,,,,,,, -92900,4.258031,3.9172316,,,,,,,,,,,,,, -93000,3.3702343,3.7809746,,,,,,,,,,,,,, -93100,3.7311397,3.6108165,,,,,,,,,,,,,, -93200,4.093848,3.6165645,,,,,,,,,,,,,, -93300,5.4255333,3.7741704,,,,,,,,,,,,,, -93400,4.7932963,3.8074324,,,,,,,,,,,,,, -93500,3.4705706,3.665578,,,,,,,,,,,,,, -93600,3.1339188,3.7078183,,,,,,,,,,,,,, -93700,5.6766086,3.7883143,,,,,,,,,,,,,, -93800,3.687691,3.653184,,,,,,,,,,,,,, -93900,3.2860339,3.6221054,,,,,,,,,,,,,, -94000,4.784071,3.7334151,,,,,,,,,,,,,, -94100,3.138641,3.5924613,,,,,,,,,,,,,, -94200,5.464802,3.658163,,,,,,,,,,,,,, -94300,4.9754195,3.904367,,,,,,,,,,,,,, -94400,4.250818,3.7946713,,,,,,,,,,,,,, -94500,3.3583648,3.6458168,,,,,,,,,,,,,, -94561,,,0.001813616,7.658918,0.0026399998,7.6551237,50000.0,0.0019,7.670427,10000.0,37989.50494408608,39496.46735692024,37989.50494408608,1505.3919687271118,0.7256448268890381,0.0 -94600,3.7039092,3.8781142,,,,,,,,,,,,,, -94700,12.766131,3.8258934,,,,,,,,,,,,,, -94800,4.4381614,3.7057266,,,,,,,,,,,,,, -94900,12.79607,3.865994,,,,,,,,,,,,,, -95000,6.7418456,3.723462,,,,,,,,,,,,,, -95100,5.626379,3.6420093,,,,,,,,,,,,,, -95200,24.62333,3.8588133,,,,,,,,,,,,,, -95300,4.399186,3.781167,,,,,,,,,,,,,, -95400,3.988272,3.8645115,,,,,,,,,,,,,, -95500,3.225506,3.7032738,,,,,,,,,,,,,, -95600,4.8474717,3.8203042,,,,,,,,,,,,,, -95700,2.5323994,3.7457979,,,,,,,,,,,,,, -95800,4.304119,3.6176066,,,,,,,,,,,,,, -95900,5.248994,3.764946,,,,,,,,,,,,,, -96000,2.9983847,3.6920943,,,,,,,,,,,,,, -96100,4.023457,3.7714958,,,,,,,,,,,,,, -96200,4.869568,3.7292516,,,,,,,,,,,,,, -96300,4.9994793,3.7000172,,,,,,,,,,,,,, -96400,36.604977,3.9277248,,,,,,,,,,,,,, -96500,3.2503793,3.8179295,,,,,,,,,,,,,, -96600,3.2179625,3.7143464,,,,,,,,,,,,,, -96700,4.1996784,3.8640037,,,,,,,,,,,,,, -96800,5.025609,3.784147,,,,,,,,,,,,,, -96900,3.696889,3.839906,,,,,,,,,,,,,, -97000,3.9724557,3.6676583,,,,,,,,,,,,,, -97100,6.4573812,3.8307269,,,,,,,,,,,,,, -97200,15.202039,3.8486834,,,,,,,,,,,,,, -97300,3.295957,3.6481185,,,,,,,,,,,,,, -97400,5.11084,3.8010767,,,,,,,,,,,,,, -97500,3.7680202,3.7407365,,,,,,,,,,,,,, -97600,3.773195,3.6281195,,,,,,,,,,,,,, -97700,6.7987866,3.9346426,,,,,,,,,,,,,, -97800,3.5334237,3.7927132,,,,,,,,,,,,,, -97900,3.6296153,3.703312,,,,,,,,,,,,,, -98000,6.921455,3.8794885,,,,,,,,,,,,,, -98100,5.560491,3.828596,,,,,,,,,,,,,, -98200,5.7989435,3.784206,,,,,,,,,,,,,, -98300,4.6190376,3.7910116,,,,,,,,,,,,,, -98400,3.9360058,3.7589045,,,,,,,,,,,,,, -98500,3.0322351,3.6895761,,,,,,,,,,,,,, -98600,2.9373536,3.8033998,,,,,,,,,,,,,, -98700,3.3407788,3.7084126,,,,,,,,,,,,,, -98800,4.5877295,3.7536163,,,,,,,,,,,,,, -98900,3.666169,3.578076,,,,,,,,,,,,,, -99000,5.0842104,3.8843122,,,,,,,,,,,,,, -99100,4.985973,3.808479,,,,,,,,,,,,,, -99200,4.305146,3.788444,,,,,,,,,,,,,, -99300,3.0768301,3.618441,,,,,,,,,,,,,, -99400,3.1495545,3.6296272,,,,,,,,,,,,,, -99468,,,0.0016342474,7.64568,0.00204,7.6454577,50000.0,0.0018000001,7.6621127,10000.0,39985.52833747864,41577.052481651306,39985.52833747864,1589.862429857254,0.773186445236206,0.0 -99500,5.896803,3.9320316,,,,,,,,,,,,,, -99600,5.127389,3.7925925,,,,,,,,,,,,,, -99700,4.6965423,3.870345,,,,,,,,,,,,,, -99800,3.9106686,3.6467907,,,,,,,,,,,,,, -99900,3.7363012,3.7557092,,,,,,,,,,,,,, -100000,3.792368,3.6697068,,,,,,,,,,,,,, -100100,5.4569893,3.7061567,,,,,,,,,,,,,, -100200,5.0573874,3.8636231,,,,,,,,,,,,,, -100300,4.172719,3.87791,,,,,,,,,,,,,, -100400,4.745176,3.8258996,,,,,,,,,,,,,, -100500,6.5660534,3.7368302,,,,,,,,,,,,,, -100600,4.786131,3.7488484,,,,,,,,,,,,,, -100700,20.541248,3.7330308,,,,,,,,,,,,,, -100800,5.5016584,3.6011453,,,,,,,,,,,,,, -100900,4.4557633,3.7115223,,,,,,,,,,,,,, -101000,6.2710056,3.8584757,,,,,,,,,,,,,, -101100,4.170464,3.6536942,,,,,,,,,,,,,, -101200,3.901454,3.705207,,,,,,,,,,,,,, -101300,6.1757793,3.8873682,,,,,,,,,,,,,, -101400,7.3739944,3.6051688,,,,,,,,,,,,,, -101500,4.433085,3.7269304,,,,,,,,,,,,,, -101600,4.545472,3.6889546,,,,,,,,,,,,,, -101700,4.9912586,3.7542663,,,,,,,,,,,,,, -101800,4.445479,3.6531723,,,,,,,,,,,,,, -101900,3.3302562,3.8209147,,,,,,,,,,,,,, -102000,33.32331,3.7291842,,,,,,,,,,,,,, -102100,6.682193,3.644393,,,,,,,,,,,,,, -102200,3.8319016,3.748837,,,,,,,,,,,,,, -102300,4.994502,3.7763066,,,,,,,,,,,,,, -102400,4.46946,3.8610172,,,,,,,,,,,,,, -102500,7.1836433,3.9487262,,,,,,,,,,,,,, -102600,4.961598,3.8653562,,,,,,,,,,,,,, -102700,6.295808,3.8098378,,,,,,,,,,,,,, -102800,4.743649,3.9129732,,,,,,,,,,,,,, -102900,5.166383,3.8248997,,,,,,,,,,,,,, -103000,6.142827,3.714141,,,,,,,,,,,,,, -103100,11.05951,3.8204274,,,,,,,,,,,,,, -103200,3.927036,3.7877688,,,,,,,,,,,,,, -103300,3.3664362,3.75898,,,,,,,,,,,,,, -103400,4.3231983,3.8141727,,,,,,,,,,,,,, -103500,5.8472376,3.837212,,,,,,,,,,,,,, -103600,4.9020624,3.802246,,,,,,,,,,,,,, -103700,3.7248266,3.8341973,,,,,,,,,,,,,, -103800,6.6299777,3.8163543,,,,,,,,,,,,,, -103900,4.132186,3.8516862,,,,,,,,,,,,,, -104000,5.598032,3.6790378,,,,,,,,,,,,,, -104100,4.2840433,3.9085011,,,,,,,,,,,,,, -104200,3.2889755,3.7162023,,,,,,,,,,,,,, -104300,5.2575274,3.7812662,,,,,,,,,,,,,, -104400,6.402026,3.811129,,,,,,,,,,,,,, -104500,3.1554387,3.6744394,,,,,,,,,,,,,, -104600,3.1033025,3.7667708,,,,,,,,,,,,,, -104689,,,0.0016143175,7.6252575,0.0021199998,7.635452,50000.0,0.0018000001,7.65344,10000.0,41981.95154380798,43595.22164225578,41981.95154380798,1611.5242595672607,0.8123252391815186,0.0 -104700,3.2510293,3.7815018,,,,,,,,,,,,,, -104800,3.760888,3.7837732,,,,,,,,,,,,,, -104900,5.1061974,3.6924949,,,,,,,,,,,,,, -105000,17.579077,3.8624177,,,,,,,,,,,,,, -105100,3.435383,3.7376752,,,,,,,,,,,,,, -105200,4.442686,3.7771134,,,,,,,,,,,,,, -105300,3.9319317,3.8173125,,,,,,,,,,,,,, -105400,4.413467,3.8777738,,,,,,,,,,,,,, -105500,5.5260453,3.799898,,,,,,,,,,,,,, -105600,3.6940482,3.7890053,,,,,,,,,,,,,, -105700,3.2887032,3.7253997,,,,,,,,,,,,,, -105800,3.8688643,3.750141,,,,,,,,,,,,,, -105900,3.9151685,3.8507104,,,,,,,,,,,,,, -106000,3.6723917,3.869387,,,,,,,,,,,,,, -106100,3.0799263,3.711652,,,,,,,,,,,,,, -106200,4.5108323,3.8725243,,,,,,,,,,,,,, -106300,4.462537,3.639622,,,,,,,,,,,,,, -106400,7.3450537,3.9319954,,,,,,,,,,,,,, -106500,15.922429,3.8099992,,,,,,,,,,,,,, -106600,6.613267,3.9996538,,,,,,,,,,,,,, -106700,3.3610792,3.7452784,,,,,,,,,,,,,, -106800,5.3537564,3.6334987,,,,,,,,,,,,,, -106900,5.467445,3.7954278,,,,,,,,,,,,,, -107000,6.998638,3.913753,,,,,,,,,,,,,, -107100,5.831449,3.6220708,,,,,,,,,,,,,, -107200,11.557222,3.991922,,,,,,,,,,,,,, -107300,4.460024,3.89603,,,,,,,,,,,,,, -107400,4.135595,3.7365088,,,,,,,,,,,,,, -107500,5.4577084,3.6871662,,,,,,,,,,,,,, -107600,3.8390932,3.7695012,,,,,,,,,,,,,, -107700,4.4817915,3.779157,,,,,,,,,,,,,, -107800,4.019161,3.8200526,,,,,,,,,,,,,, -107900,2.7714932,3.7639713,,,,,,,,,,,,,, -108000,4.6806536,3.7370164,,,,,,,,,,,,,, -108100,3.0377078,3.9234962,,,,,,,,,,,,,, -108200,4.945133,3.8763032,,,,,,,,,,,,,, -108300,3.865592,3.7258415,,,,,,,,,,,,,, -108400,7.269028,3.9254532,,,,,,,,,,,,,, -108500,5.3218293,3.7838445,,,,,,,,,,,,,, -108600,10.783357,3.806242,,,,,,,,,,,,,, -108700,11.696349,3.6098814,,,,,,,,,,,,,, -108800,4.0714593,3.7698712,,,,,,,,,,,,,, -108900,4.5637183,3.7248135,,,,,,,,,,,,,, -108937,,,0.0017538265,7.6281743,0.00204,7.6303515,50000.0,0.0018000001,7.649656,10000.0,43978.46745610237,45672.22894263268,43978.46745610237,1691.907369852066,0.8714404106140137,0.0 -109000,5.376041,3.760328,,,,,,,,,,,,,, -109100,5.657345,3.8204675,,,,,,,,,,,,,, -109200,9.227175,3.875277,,,,,,,,,,,,,, -109300,2.3602028,3.5669336,,,,,,,,,,,,,, -109400,3.8874097,4.048623,,,,,,,,,,,,,, -109500,4.3420963,3.9647741,,,,,,,,,,,,,, -109600,4.491919,3.8505788,,,,,,,,,,,,,, -109700,4.9978333,3.9355803,,,,,,,,,,,,,, -109800,3.8086424,3.9143074,,,,,,,,,,,,,, -109900,4.3091617,3.6726217,,,,,,,,,,,,,, -110000,9.459061,3.784495,,,,,,,,,,,,,, -110100,4.7404833,3.9171824,,,,,,,,,,,,,, -110200,9.23133,3.8291678,,,,,,,,,,,,,, -110300,3.637351,3.842377,,,,,,,,,,,,,, -110400,7.4110465,3.8116012,,,,,,,,,,,,,, -110500,3.1321383,3.6745036,,,,,,,,,,,,,, -110600,5.3074417,3.582342,,,,,,,,,,,,,, -110700,2.6436422,3.6574836,,,,,,,,,,,,,, -110800,19.198784,3.92822,,,,,,,,,,,,,, -110900,3.5715544,3.7600322,,,,,,,,,,,,,, -111000,4.1539664,3.7033374,,,,,,,,,,,,,, -111100,5.543024,3.871777,,,,,,,,,,,,,, -111200,4.6366763,3.6521044,,,,,,,,,,,,,, -111300,4.0463314,3.788677,,,,,,,,,,,,,, -111400,4.1235437,3.7942095,,,,,,,,,,,,,, -111500,7.0936127,3.6910822,,,,,,,,,,,,,, -111600,18.019808,3.9077935,,,,,,,,,,,,,, -111700,3.39752,3.9913716,,,,,,,,,,,,,, -111800,10.306468,3.9151318,,,,,,,,,,,,,, -111895,,,0.0021324935,7.6150737,0.00194,7.6296186,50000.0,0.0018000001,7.648998,10000.0,45974.4246840477,47793.9038836956,45974.4246840477,1816.6612305641167,1.4030964374542236,0.0 -111900,3.9275575,3.8042986,,,,,,,,,,,,,, -112000,4.138191,3.7380276,,,,,,,,,,,,,, -112100,4.6038084,3.7808924,,,,,,,,,,,,,, -112200,4.3356366,3.9187803,,,,,,,,,,,,,, -112300,3.0787988,3.6989148,,,,,,,,,,,,,, -112400,3.8041463,3.8773782,,,,,,,,,,,,,, -112500,16.20001,3.8334985,,,,,,,,,,,,,, -112600,8.872913,3.84942,,,,,,,,,,,,,, -112700,3.020047,3.7359579,,,,,,,,,,,,,, -112800,7.824386,3.814516,,,,,,,,,,,,,, -112900,7.2096887,3.8773196,,,,,,,,,,,,,, -113000,6.990146,3.9106002,,,,,,,,,,,,,, -113100,11.177308,3.8120627,,,,,,,,,,,,,, -113200,3.3114169,3.947348,,,,,,,,,,,,,, -113300,4.9605455,3.7409136,,,,,,,,,,,,,, -113400,4.188526,3.8735268,,,,,,,,,,,,,, -113500,6.301297,3.8350315,,,,,,,,,,,,,, -113600,3.3167946,3.7566357,,,,,,,,,,,,,, -113700,2.9355977,3.8058612,,,,,,,,,,,,,, -113800,4.4839053,3.776876,,,,,,,,,,,,,, -113900,3.7172391,3.6984591,,,,,,,,,,,,,, -114000,11.39768,3.7443218,,,,,,,,,,,,,, -114100,4.159839,3.8136022,,,,,,,,,,,,,, -114200,3.752556,3.7094502,,,,,,,,,,,,,, -114300,7.3795185,4.0115905,,,,,,,,,,,,,, -114400,2.7422314,3.83433,,,,,,,,,,,,,, -114500,7.1590285,3.8269649,,,,,,,,,,,,,, -114600,5.699358,3.847859,,,,,,,,,,,,,, -114700,10.639413,3.8803086,,,,,,,,,,,,,, -114800,5.9916005,3.8578885,,,,,,,,,,,,,, -114900,5.5205927,3.7692356,,,,,,,,,,,,,, -114967,,,0.002072704,7.616824,0.00188,7.625837,50000.0,0.0017000001,7.6456165,10000.0,47970.05360388756,49821.3028011322,47970.05360388756,1847.798466205597,1.7638790607452393,0.0 -115000,4.4481263,3.8422027,,,,,,,,,,,,,, -115100,4.403424,3.8667254,,,,,,,,,,,,,, -115200,21.891872,3.9015439,,,,,,,,,,,,,, -115300,3.2971532,3.8190055,,,,,,,,,,,,,, -115400,3.4605074,3.7726648,,,,,,,,,,,,,, -115500,55.11227,3.895568,,,,,,,,,,,,,, -115600,2.9291873,3.711916,,,,,,,,,,,,,, -115700,5.43193,3.9283447,,,,,,,,,,,,,, -115800,4.47112,3.7677941,,,,,,,,,,,,,, -115900,4.5517497,3.7989235,,,,,,,,,,,,,, -116000,3.6519709,3.8780327,,,,,,,,,,,,,, -116100,4.106604,3.7333598,,,,,,,,,,,,,, -116200,5.7141576,3.968327,,,,,,,,,,,,,, -116300,2.672927,3.7668402,,,,,,,,,,,,,, -116400,3.8861983,3.9910412,,,,,,,,,,,,,, -116500,6.089529,3.6987512,,,,,,,,,,,,,, -116600,2.686511,3.7109423,,,,,,,,,,,,,, -116700,4.908961,3.8116314,,,,,,,,,,,,,, -116800,4.274679,3.8371181,,,,,,,,,,,,,, -116900,2.6250956,3.8440568,,,,,,,,,,,,,, -117000,4.186766,3.6785276,,,,,,,,,,,,,, -117100,3.3089893,3.8470113,,,,,,,,,,,,,, -117200,3.7952948,3.8503036,,,,,,,,,,,,,, -117300,2.5078144,3.7403162,,,,,,,,,,,,,, -117400,3.857513,3.8079453,,,,,,,,,,,,,, -117500,5.1463213,3.7657351,,,,,,,,,,,,,, -117600,14.902926,3.7564468,,,,,,,,,,,,,, -117700,3.076664,3.7177477,,,,,,,,,,,,,, -117800,3.445043,3.8301206,,,,,,,,,,,,,, -117900,5.425839,4.120443,,,,,,,,,,,,,, -117934,,,0.0018534757,7.623401,0.00182,7.628771,50000.0,0.0019,7.6490345,10000.0,49965.64317059517,51855.24300432205,49965.64317059517,1885.494007825852,2.183001518249512,0.0 -118000,9.385499,3.8958511,,,,,,,,,,,,,, -118100,2.9379132,3.651403,,,,,,,,,,,,,, -118200,4.6236877,3.9199445,,,,,,,,,,,,,, -118300,2.7990782,3.720789,,,,,,,,,,,,,, -118400,3.9721272,3.6497588,,,,,,,,,,,,,, -118500,5.5147386,3.7769046,,,,,,,,,,,,,, -118600,5.439056,3.8420036,,,,,,,,,,,,,, -118700,8.089831,3.9625082,,,,,,,,,,,,,, -118800,6.541725,3.8788805,,,,,,,,,,,,,, -118900,4.2585464,3.8112612,,,,,,,,,,,,,, -119000,4.3126283,3.8589547,,,,,,,,,,,,,, -119100,10.229043,3.974662,,,,,,,,,,,,,, -119200,11.070024,3.9006264,,,,,,,,,,,,,, -119300,11.379998,3.939808,,,,,,,,,,,,,, -119400,3.5441818,3.9538083,,,,,,,,,,,,,, -119500,3.0569496,3.7393396,,,,,,,,,,,,,, -119600,5.33516,3.9989693,,,,,,,,,,,,,, -119700,7.362302,3.8755703,,,,,,,,,,,,,, -119800,6.288874,3.8165514,,,,,,,,,,,,,, -119900,4.81928,3.9429932,,,,,,,,,,,,,, -120000,3.2867503,3.7361712,,,,,,,,,,,,,, -120100,5.2188926,3.807652,,,,,,,,,,,,,, -120200,3.5767217,3.8174608,,,,,,,,,,,,,, -120300,4.1150055,3.8690505,,,,,,,,,,,,,, -120400,3.8997452,3.7922268,,,,,,,,,,,,,, -120500,3.3161087,3.8541164,,,,,,,,,,,,,, -120600,4.229436,3.9411607,,,,,,,,,,,,,, -120700,3.6027894,3.8828125,,,,,,,,,,,,,, -120800,5.321989,3.9399219,,,,,,,,,,,,,, -120883,,,0.0013153699,7.611876,0.0017799999,7.623189,50000.0,0.0021000002,7.644687,10000.0,51961.32361125946,53880.19545173645,51961.32361125946,1914.1544938087463,2.561258316040039,0.0 -120900,3.5195842,3.8635147,,,,,,,,,,,,,, -121000,3.9423573,3.850008,,,,,,,,,,,,,, -121100,4.5462575,3.981271,,,,,,,,,,,,,, -121200,3.2338,3.7782588,,,,,,,,,,,,,, -121300,4.964594,3.8536155,,,,,,,,,,,,,, -121400,2.5844753,3.7594428,,,,,,,,,,,,,, -121500,3.5755427,3.7236176,,,,,,,,,,,,,, -121600,4.9097505,3.842507,,,,,,,,,,,,,, -121700,5.0955276,3.921885,,,,,,,,,,,,,, -121800,5.5242424,3.8590653,,,,,,,,,,,,,, -121900,7.943033,3.7824018,,,,,,,,,,,,,, -122000,2.6752732,3.7593224,,,,,,,,,,,,,, -122100,3.422566,3.7312126,,,,,,,,,,,,,, -122200,3.2170942,3.8356416,,,,,,,,,,,,,, -122300,6.434549,3.9361353,,,,,,,,,,,,,, -122400,4.3436823,3.7894228,,,,,,,,,,,,,, -122500,27.089064,3.9853866,,,,,,,,,,,,,, -122600,5.177217,3.7168698,,,,,,,,,,,,,, -122700,11.616267,3.710192,,,,,,,,,,,,,, -122800,15.874014,3.9392157,,,,,,,,,,,,,, -122900,3.7724404,3.7477667,,,,,,,,,,,,,, -123000,5.1227503,3.8347125,,,,,,,,,,,,,, -123100,2.887475,3.7349963,,,,,,,,,,,,,, -123200,3.8362484,3.846238,,,,,,,,,,,,,, -123300,6.313995,3.858592,,,,,,,,,,,,,, -123400,3.9272094,3.86928,,,,,,,,,,,,,, -123500,10.609228,3.916779,,,,,,,,,,,,,, -123600,10.902922,3.9500902,,,,,,,,,,,,,, -123700,12.306965,3.8594112,,,,,,,,,,,,,, -123800,4.1741543,3.8255281,,,,,,,,,,,,,, -123836,,,0.0015744579,7.5977,0.00174,7.615778,50000.0,0.0019,7.6381392,10000.0,53956.820291996,55911.70957875252,53956.820291996,1949.544005155564,2.9746978282928467,0.0 -123900,10.455456,3.7439103,,,,,,,,,,,,,, -124000,3.0511992,3.8120224,,,,,,,,,,,,,, -124100,4.07148,3.8573456,,,,,,,,,,,,,, -124200,4.924006,3.8181963,,,,,,,,,,,,,, -124300,12.579405,3.876627,,,,,,,,,,,,,, -124400,3.1175878,3.7407773,,,,,,,,,,,,,, -124500,9.269796,3.8457687,,,,,,,,,,,,,, -124600,3.0821624,3.6371832,,,,,,,,,,,,,, -124700,21.211935,3.8681612,,,,,,,,,,,,,, -124800,5.950034,3.8787365,,,,,,,,,,,,,, -124900,4.49684,3.846447,,,,,,,,,,,,,, -125000,3.7403393,3.9419875,,,,,,,,,,,,,, -125100,4.447743,3.7944674,,,,,,,,,,,,,, -125200,3.397232,3.7243636,,,,,,,,,,,,,, -125300,11.197062,3.8184288,,,,,,,,,,,,,, -125400,2.6906972,3.7622778,,,,,,,,,,,,,, -125500,3.7688205,3.8464904,,,,,,,,,,,,,, -125600,3.3049324,3.7647893,,,,,,,,,,,,,, -125700,4.2222767,3.7181392,,,,,,,,,,,,,, -125800,4.557241,3.8364043,,,,,,,,,,,,,, -125900,4.884914,3.6980176,,,,,,,,,,,,,, -126000,6.7227736,3.92307,,,,,,,,,,,,,, -126100,4.241055,3.6775475,,,,,,,,,,,,,, -126200,3.499955,3.7502038,,,,,,,,,,,,,, -126300,4.542598,3.7865655,,,,,,,,,,,,,, -126400,4.3612976,3.9679422,,,,,,,,,,,,,, -126500,10.77136,4.007995,,,,,,,,,,,,,, -126600,2.9848304,3.773301,,,,,,,,,,,,,, -126700,4.078251,3.825758,,,,,,,,,,,,,, -126748,,,0.002192283,7.5961895,0.0017799999,7.6027384,50000.0,0.0019,7.6265955,10000.0,55952.328662633896,57943.46052527428,55952.328662633896,1985.0605597496035,3.459739923477173,0.0 -126800,4.6065326,3.8750064,,,,,,,,,,,,,, -126900,3.9841883,3.8666985,,,,,,,,,,,,,, -127000,4.19587,3.9528966,,,,,,,,,,,,,, -127100,5.826696,3.8324783,,,,,,,,,,,,,, -127200,17.807783,3.82713,,,,,,,,,,,,,, -127300,4.1498322,3.7210634,,,,,,,,,,,,,, -127400,4.0499625,3.802925,,,,,,,,,,,,,, -127500,4.8336654,3.785916,,,,,,,,,,,,,, -127600,2.4317706,3.9065185,,,,,,,,,,,,,, -127700,3.3372576,3.8618894,,,,,,,,,,,,,, -127800,9.562402,3.962144,,,,,,,,,,,,,, -127900,3.5000615,3.9105558,,,,,,,,,,,,,, -128000,6.8920016,3.924448,,,,,,,,,,,,,, -128100,12.77795,3.825415,,,,,,,,,,,,,, -128200,19.338078,3.8517718,,,,,,,,,,,,,, -128300,5.1319556,3.9616008,,,,,,,,,,,,,, -128400,24.784185,3.933617,,,,,,,,,,,,,, -128500,8.217952,3.937379,,,,,,,,,,,,,, -128600,5.6154876,3.9927263,,,,,,,,,,,,,, -128700,25.944843,3.9131403,,,,,,,,,,,,,, -128800,6.1151004,3.8181145,,,,,,,,,,,,,, -128900,3.8512988,3.8204195,,,,,,,,,,,,,, -129000,5.8273554,3.9977965,,,,,,,,,,,,,, -129100,4.9329987,3.8429203,,,,,,,,,,,,,, -129200,18.143944,3.8571548,,,,,,,,,,,,,, -129300,5.595682,3.9041507,,,,,,,,,,,,,, -129400,4.007479,3.8578212,,,,,,,,,,,,,, -129500,5.6567554,3.915813,,,,,,,,,,,,,, -129600,9.009786,4.060838,,,,,,,,,,,,,, -129664,,,0.0017338967,7.580449,0.00182,7.594372,50000.0,0.0019,7.61915,10000.0,57949.51488518715,59969.13365912437,57949.51488518715,2012.879814863205,3.8696322441101074,0.0 -129700,3.3517041,3.8489535,,,,,,,,,,,,,, -129800,4.209185,3.9517598,,,,,,,,,,,,,, -129900,3.5753055,3.7060328,,,,,,,,,,,,,, -130000,5.282336,3.8749697,,,,,,,,,,,,,, -130100,37.602802,4.048226,,,,,,,,,,,,,, -130200,3.63432,3.8021939,,,,,,,,,,,,,, -130300,4.0353847,3.9017956,,,,,,,,,,,,,, -130400,6.2187753,4.1330185,,,,,,,,,,,,,, -130500,8.197782,4.052623,,,,,,,,,,,,,, -130600,6.8042593,3.9336784,,,,,,,,,,,,,, -130700,5.4732256,3.8324423,,,,,,,,,,,,,, -130800,3.0168664,3.6159778,,,,,,,,,,,,,, -130900,5.9715424,3.8426807,,,,,,,,,,,,,, -131000,5.1567726,3.874363,,,,,,,,,,,,,, -131100,2.889144,3.7527375,,,,,,,,,,,,,, -131200,5.5135827,3.812678,,,,,,,,,,,,,, -131300,7.1405706,3.8648164,,,,,,,,,,,,,, -131400,5.2732534,3.9160786,,,,,,,,,,,,,, -131500,3.8169348,3.79936,,,,,,,,,,,,,, -131600,4.6745896,3.7790623,,,,,,,,,,,,,, -131700,3.4323125,3.7781577,,,,,,,,,,,,,, -131800,3.676134,3.8370671,,,,,,,,,,,,,, -131900,3.499526,3.744359,,,,,,,,,,,,,, -132000,4.9093046,3.9428804,,,,,,,,,,,,,, -132100,4.4625115,3.7666528,,,,,,,,,,,,,, -132200,8.081251,3.7677574,,,,,,,,,,,,,, -132300,4.3300858,3.9088376,,,,,,,,,,,,,, -132400,3.6348166,3.7447648,,,,,,,,,,,,,, -132500,4.6311784,3.9242504,,,,,,,,,,,,,, -132600,5.528411,3.9177155,,,,,,,,,,,,,, -132610,,,0.001813616,7.5625515,0.0018,7.579831,50000.0,0.0018000001,7.605457,10000.0,59946.0051176548,62008.08043074608,59946.0051176548,2054.632348537445,4.288118839263916,0.0 -132700,4.6349735,3.7707348,,,,,,,,,,,,,, -132800,4.7975154,3.7419162,,,,,,,,,,,,,, -132900,4.546232,3.925696,,,,,,,,,,,,,, -133000,4.129563,3.9811711,,,,,,,,,,,,,, -133100,27.67613,3.926815,,,,,,,,,,,,,, -133200,6.153795,3.957784,,,,,,,,,,,,,, -133300,3.3190844,3.8944592,,,,,,,,,,,,,, -133400,6.1725016,3.9338212,,,,,,,,,,,,,, -133500,6.154134,3.994706,,,,,,,,,,,,,, -133600,11.731233,3.8466702,,,,,,,,,,,,,, -133700,6.1500564,3.921942,,,,,,,,,,,,,, -133800,15.1175585,3.8392432,,,,,,,,,,,,,, -133900,6.694962,3.8467731,,,,,,,,,,,,,, -134000,7.819849,4.0930233,,,,,,,,,,,,,, -134100,4.031584,3.9195158,,,,,,,,,,,,,, -134200,13.054775,3.9789762,,,,,,,,,,,,,, -134300,6.653199,3.944149,,,,,,,,,,,,,, -134400,6.414671,3.8172336,,,,,,,,,,,,,, -134500,7.89964,3.9058867,,,,,,,,,,,,,, -134600,5.462031,3.8989003,,,,,,,,,,,,,, -134700,3.062308,3.7560372,,,,,,,,,,,,,, -134800,4.7382016,3.814735,,,,,,,,,,,,,, -134900,4.2175264,3.7313404,,,,,,,,,,,,,, -135000,3.9357486,3.8367722,,,,,,,,,,,,,, -135100,4.8702245,3.7801714,,,,,,,,,,,,,, -135200,4.392237,3.7925894,,,,,,,,,,,,,, -135300,3.6274323,3.7758002,,,,,,,,,,,,,, -135400,11.151893,3.8981698,,,,,,,,,,,,,, -135448,,,0.0017737563,7.552586,0.00204,7.5581985,50000.0,0.0019,7.5847716,10000.0,61943.662088871,64048.056831121445,61943.662088871,2096.377536058426,4.6587910652160645,0.0 -135500,2.2517211,3.7219882,,,,,,,,,,,,,, -135600,3.5304663,3.947412,,,,,,,,,,,,,, -135700,3.6491961,3.8470793,,,,,,,,,,,,,, -135800,3.504812,3.9701824,,,,,,,,,,,,,, -135900,5.9954543,3.9988873,,,,,,,,,,,,,, -136000,25.904045,3.8604927,,,,,,,,,,,,,, -136100,8.213358,4.0181265,,,,,,,,,,,,,, -136200,4.9171133,3.8106866,,,,,,,,,,,,,, -136300,6.219894,3.9318824,,,,,,,,,,,,,, -136400,7.151308,3.9189653,,,,,,,,,,,,,, -136500,3.9935062,3.8526607,,,,,,,,,,,,,, -136600,4.93948,3.8053665,,,,,,,,,,,,,, -136700,3.2404213,3.8049748,,,,,,,,,,,,,, -136800,4.219491,3.7178602,,,,,,,,,,,,,, -136900,5.423903,3.896631,,,,,,,,,,,,,, -137000,6.5687137,3.7315526,,,,,,,,,,,,,, -137100,4.99497,3.9801855,,,,,,,,,,,,,, -137200,5.71572,3.8224187,,,,,,,,,,,,,, -137300,2.9901032,3.855413,,,,,,,,,,,,,, -137400,3.6423774,3.8682897,,,,,,,,,,,,,, -137500,8.655993,4.034668,,,,,,,,,,,,,, -137600,8.7465925,4.1233954,,,,,,,,,,,,,, -137700,4.8780413,3.896852,,,,,,,,,,,,,, -137800,3.7128158,3.8314338,,,,,,,,,,,,,, -137900,4.846003,3.9532673,,,,,,,,,,,,,, -138000,4.7727594,3.8014815,,,,,,,,,,,,,, -138100,6.4787602,3.8645902,,,,,,,,,,,,,, -138200,4.562845,3.8523617,,,,,,,,,,,,,, -138300,3.5587213,3.7978323,,,,,,,,,,,,,, -138357,,,0.0021324935,7.5125346,0.0020599999,7.53964,50000.0,0.002,7.5671916,10000.0,63938.85355734825,66071.15640091896,63938.85355734825,2123.342755794525,5.124191522598267,0.0 -138400,5.55637,3.7755787,,,,,,,,,,,,,, -138500,3.3718374,3.8791933,,,,,,,,,,,,,, -138600,22.835024,3.8973696,,,,,,,,,,,,,, -138700,3.3655233,3.8858745,,,,,,,,,,,,,, -138800,7.2266088,3.9636269,,,,,,,,,,,,,, -138900,20.090502,3.910851,,,,,,,,,,,,,, -139000,6.6730204,4.0319347,,,,,,,,,,,,,, -139100,5.4175215,3.9209085,,,,,,,,,,,,,, -139200,5.785498,3.8992267,,,,,,,,,,,,,, -139300,10.166209,3.8999686,,,,,,,,,,,,,, -139400,9.924324,4.029359,,,,,,,,,,,,,, -139500,5.184852,3.7248063,,,,,,,,,,,,,, -139600,4.376903,3.771514,,,,,,,,,,,,,, -139700,5.826095,3.747323,,,,,,,,,,,,,, -139800,10.620085,4.00409,,,,,,,,,,,,,, -139900,6.008486,3.7739477,,,,,,,,,,,,,, -140000,5.750551,3.8834944,,,,,,,,,,,,,, -140100,11.466199,3.8390493,,,,,,,,,,,,,, -140200,4.5005183,3.948241,,,,,,,,,,,,,, -140300,16.083975,3.914906,,,,,,,,,,,,,, -140400,3.316736,3.8398662,,,,,,,,,,,,,, -140500,16.605505,3.6711276,,,,,,,,,,,,,, -140600,30.907335,3.9662852,,,,,,,,,,,,,, -140700,4.0948086,3.7786148,,,,,,,,,,,,,, -140800,10.223881,4.0885763,,,,,,,,,,,,,, -140900,5.088145,3.814241,,,,,,,,,,,,,, -141000,6.4651556,4.042456,,,,,,,,,,,,,, -141100,6.886142,3.7587595,,,,,,,,,,,,,, -141200,51.13762,4.26091,,,,,,,,,,,,,, -141279,,,0.0019730548,7.5101995,0.00222,7.5270987,50000.0,0.0019,7.5543923,10000.0,65934.31674814224,68100.40691971779,65934.31674814224,2156.486003637314,5.550399541854858,0.0 -141300,4.916695,3.920254,,,,,,,,,,,,,, -141400,5.3441114,4.035045,,,,,,,,,,,,,, -141500,6.056302,3.661153,,,,,,,,,,,,,, -141600,11.05118,3.9600246,,,,,,,,,,,,,, -141700,5.7601295,3.9735744,,,,,,,,,,,,,, -141800,3.6796074,3.8766103,,,,,,,,,,,,,, -141900,5.914992,4.0436497,,,,,,,,,,,,,, -142000,3.8678403,3.9334168,,,,,,,,,,,,,, -142100,4.6479526,3.8349125,,,,,,,,,,,,,, -142200,5.142815,4.0386586,,,,,,,,,,,,,, -142300,9.013766,3.9974341,,,,,,,,,,,,,, -142400,3.5983305,3.7415643,,,,,,,,,,,,,, -142500,45.369972,3.8918815,,,,,,,,,,,,,, -142600,4.8703604,3.8174307,,,,,,,,,,,,,, -142700,3.6371644,3.806127,,,,,,,,,,,,,, -142800,4.315119,3.9530263,,,,,,,,,,,,,, -142900,8.890266,3.9717948,,,,,,,,,,,,,, -143000,10.819942,3.9218936,,,,,,,,,,,,,, -143100,4.4896646,3.8029666,,,,,,,,,,,,,, -143200,3.0291765,3.797175,,,,,,,,,,,,,, -143300,3.1117814,3.71223,,,,,,,,,,,,,, -143400,4.7589154,3.8630419,,,,,,,,,,,,,, -143500,7.9758544,3.8621502,,,,,,,,,,,,,, -143600,4.1292925,3.8461058,,,,,,,,,,,,,, -143700,5.006647,3.849973,,,,,,,,,,,,,, -143800,7.449296,3.825643,,,,,,,,,,,,,, -143900,10.432781,3.9143276,,,,,,,,,,,,,, -144000,4.502899,3.8878417,,,,,,,,,,,,,, -144100,3.0637622,3.9866912,,,,,,,,,,,,,, -144193,,,0.002351722,7.513111,0.00256,7.514347,50000.0,0.0019,7.5415044,10000.0,67929.82773637772,70136.11622142792,67929.82773637772,2195.9613015651703,6.055431127548218,0.0 -144200,4.148107,3.8893437,,,,,,,,,,,,,, -144300,26.393806,3.9268622,,,,,,,,,,,,,, -144400,3.846163,3.8257942,,,,,,,,,,,,,, -144500,7.601764,3.8658042,,,,,,,,,,,,,, -144600,8.486637,3.8613698,,,,,,,,,,,,,, -144700,4.2302637,3.8233047,,,,,,,,,,,,,, -144800,2.453479,3.90842,,,,,,,,,,,,,, -144900,4.661877,3.9667075,,,,,,,,,,,,,, -145000,3.5486996,3.8808932,,,,,,,,,,,,,, -145100,4.59585,3.9970164,,,,,,,,,,,,,, -145200,4.136439,3.8016682,,,,,,,,,,,,,, -145300,6.00421,3.84082,,,,,,,,,,,,,, -145400,4.6914287,3.8578777,,,,,,,,,,,,,, -145500,8.417787,3.8393512,,,,,,,,,,,,,, -145600,5.6041117,3.9657788,,,,,,,,,,,,,, -145700,25.089436,4.0288153,,,,,,,,,,,,,, -145800,6.8527694,4.037016,,,,,,,,,,,,,, -145900,4.5162215,4.0277348,,,,,,,,,,,,,, -146000,10.67943,3.9644775,,,,,,,,,,,,,, -146100,3.464183,3.8558645,,,,,,,,,,,,,, -146200,5.4988155,3.886444,,,,,,,,,,,,,, -146300,3.6709526,3.8529081,,,,,,,,,,,,,, -146400,12.727401,3.9691634,,,,,,,,,,,,,, -146500,5.4271717,3.9526758,,,,,,,,,,,,,, -146600,12.044347,4.180284,,,,,,,,,,,,,, -146700,3.3213959,3.8617756,,,,,,,,,,,,,, -146800,2.9264646,3.9622335,,,,,,,,,,,,,, -146900,6.3585896,3.8714356,,,,,,,,,,,,,, -147000,5.0888457,3.8260703,,,,,,,,,,,,,, -147094,,,0.0018734055,7.492316,0.00244,7.5076427,50000.0,0.0021000002,7.5356674,10000.0,69926.05036187172,72188.20590472221,69926.05036187172,2251.2275352478027,6.4613258838653564,0.0 -147100,12.511688,3.9799967,,,,,,,,,,,,,, -147200,3.8981714,3.7231696,,,,,,,,,,,,,, -147300,6.6749673,4.110593,,,,,,,,,,,,,, -147400,5.2004633,3.6665242,,,,,,,,,,,,,, -147500,4.6903143,3.888262,,,,,,,,,,,,,, -147600,4.4121666,3.981841,,,,,,,,,,,,,, -147700,4.063638,3.8720703,,,,,,,,,,,,,, -147800,4.3871384,3.7562695,,,,,,,,,,,,,, -147900,4.1956773,3.7515187,,,,,,,,,,,,,, -148000,3.6951506,3.724639,,,,,,,,,,,,,, -148100,3.3839467,3.6832376,,,,,,,,,,,,,, -148200,4.2071323,3.9001,,,,,,,,,,,,,, -148300,9.754958,3.9203672,,,,,,,,,,,,,, -148400,5.1017733,3.9192672,,,,,,,,,,,,,, -148500,4.5545187,3.6867814,,,,,,,,,,,,,, -148600,3.0557969,3.8194761,,,,,,,,,,,,,, -148700,5.148833,3.7640877,,,,,,,,,,,,,, -148800,3.2718174,3.7621694,,,,,,,,,,,,,, -148900,4.229083,3.8816133,,,,,,,,,,,,,, -149000,4.8595433,3.8750353,,,,,,,,,,,,,, -149100,2.9685616,3.851181,,,,,,,,,,,,,, -149200,1.9591914,3.5909672,,,,,,,,,,,,,, -149300,5.084506,4.031536,,,,,,,,,,,,,, -149400,4.5329566,3.7850688,,,,,,,,,,,,,, -149500,3.1831021,3.897552,,,,,,,,,,,,,, -149600,75.85711,3.9243429,,,,,,,,,,,,,, -149700,6.5815907,3.8394368,,,,,,,,,,,,,, -149800,2.840412,3.7560227,,,,,,,,,,,,,, -149900,5.33743,3.884756,,,,,,,,,,,,,, -149980,,,0.002471301,7.497335,0.0025,7.5091887,50000.0,0.0018000001,7.536077,10000.0,71922.94839406013,74219.54028224945,71922.94839406013,2284.97248506546,6.899955034255981,0.0 -150000,5.351403,3.977597,,,,,,,,,,,,,, -150100,4.1424794,3.9277654,,,,,,,,,,,,,, -150200,13.881202,3.9683948,,,,,,,,,,,,,, -150300,2.8658798,3.808566,,,,,,,,,,,,,, -150400,11.834768,3.9765933,,,,,,,,,,,,,, -150500,3.531948,3.7456522,,,,,,,,,,,,,, -150600,4.9298306,3.842797,,,,,,,,,,,,,, -150700,2.9383664,3.8158917,,,,,,,,,,,,,, -150800,3.687251,3.9081478,,,,,,,,,,,,,, -150900,24.97658,3.8210855,,,,,,,,,,,,,, -151000,11.597968,3.855148,,,,,,,,,,,,,, -151100,3.575751,3.6894393,,,,,,,,,,,,,, -151200,4.606237,3.9495873,,,,,,,,,,,,,, -151300,11.574319,3.9363236,,,,,,,,,,,,,, -151400,3.4424481,3.7879083,,,,,,,,,,,,,, -151500,5.8509927,3.8674917,,,,,,,,,,,,,, -151600,4.484781,3.9271898,,,,,,,,,,,,,, -151700,3.676269,3.8335054,,,,,,,,,,,,,, -151800,3.7917354,3.9043458,,,,,,,,,,,,,, -151900,4.0945654,3.8184648,,,,,,,,,,,,,, -152000,3.509455,3.863084,,,,,,,,,,,,,, -152100,2.940356,3.709189,,,,,,,,,,,,,, -152200,5.3101377,3.9768846,,,,,,,,,,,,,, -152300,4.6023836,3.8807883,,,,,,,,,,,,,, -152400,5.9561286,3.8538032,,,,,,,,,,,,,, -152500,4.6170974,3.946331,,,,,,,,,,,,,, -152600,4.8838553,3.9479642,,,,,,,,,,,,,, -152700,3.8989985,3.8737729,,,,,,,,,,,,,, -152800,5.071679,3.7823086,,,,,,,,,,,,,, -152900,4.2372017,3.6983323,,,,,,,,,,,,,, -152933,,,0.0026705994,7.502706,0.00278,7.5088005,50000.0,0.002,7.536695,10000.0,73918.37983560562,76248.95145559311,73918.37983560562,2318.299463748932,7.3489696979522705,0.0 -153000,3.7413447,3.8248968,,,,,,,,,,,,,, -153100,3.7417607,3.8072686,,,,,,,,,,,,,, -153200,3.2579947,3.7740605,,,,,,,,,,,,,, -153300,3.8761103,3.7306802,,,,,,,,,,,,,, -153400,4.5520678,3.8243692,,,,,,,,,,,,,, -153500,3.16826,3.8680336,,,,,,,,,,,,,, -153600,4.8688197,3.7464905,,,,,,,,,,,,,, -153700,6.4317217,3.9145036,,,,,,,,,,,,,, -153800,8.599949,3.9195647,,,,,,,,,,,,,, -153900,4.56088,3.813144,,,,,,,,,,,,,, -154000,2.7458622,3.6829123,,,,,,,,,,,,,, -154100,9.315207,4.050851,,,,,,,,,,,,,, -154200,4.950338,3.8840559,,,,,,,,,,,,,, -154300,10.534608,3.7739942,,,,,,,,,,,,,, -154400,3.7198246,3.8218725,,,,,,,,,,,,,, -154500,5.419073,3.9301095,,,,,,,,,,,,,, -154600,2.5316293,3.7607892,,,,,,,,,,,,,, -154700,3.2851186,3.615681,,,,,,,,,,,,,, -154800,3.3182745,3.721899,,,,,,,,,,,,,, -154900,11.172554,3.882389,,,,,,,,,,,,,, -155000,6.346722,3.8571784,,,,,,,,,,,,,, -155100,4.8329473,3.836457,,,,,,,,,,,,,, -155200,8.38208,3.8235443,,,,,,,,,,,,,, -155300,2.8770084,3.8691883,,,,,,,,,,,,,, -155400,24.736267,4.035109,,,,,,,,,,,,,, -155500,3.1201658,3.801094,,,,,,,,,,,,,, -155600,3.3281465,3.971894,,,,,,,,,,,,,, -155700,3.185469,3.7953002,,,,,,,,,,,,,, -155746,,,,,,,,,,,75915.5999929905,,,,,0.0 +0,0.67443293,6.9142685,,,,,,,,,,,,,, +1,,,0.0010762117,6.9101987,0.0011,6.910505,50000.0,0.0013000001,6.910612,10000.0,53.80131030082703,118.50065970420836,53.80131030082703,64.69926476478577,0.0,0.0 +100,0.68602806,6.7963324,,,,,,,,,,,,,, +200,0.84898096,6.4846125,,,,,,,,,,,,,, +300,1.0689995,6.2019396,,,,,,,,,,,,,, +400,2.0354378,5.8884478,,,,,,,,,,,,,, +500,1.8697954,5.808592,,,,,,,,,,,,,, +600,3.9706166,5.630149,,,,,,,,,,,,,, +700,4.338664,5.4395885,,,,,,,,,,,,,, +800,3.2051055,5.242498,,,,,,,,,,,,,, +900,4.9889445,5.1435213,,,,,,,,,,,,,, +1000,4.439693,4.9768543,,,,,,,,,,,,,, +1100,5.249515,4.9530735,,,,,,,,,,,,,, +1200,5.5972114,4.8129625,,,,,,,,,,,,,, +1300,4.0549026,4.606395,,,,,,,,,,,,,, +1400,4.3791575,4.6074896,,,,,,,,,,,,,, +1500,3.115973,4.377636,,,,,,,,,,,,,, +1600,4.363248,4.354099,,,,,,,,,,,,,, +1700,5.2609315,4.3621297,,,,,,,,,,,,,, +1800,5.447823,4.281497,,,,,,,,,,,,,, +1900,4.8260546,4.0915337,,,,,,,,,,,,,, +2000,4.0353265,4.1480136,,,,,,,,,,,,,, +2100,5.71244,4.054986,,,,,,,,,,,,,, +2200,4.179816,3.8286698,,,,,,,,,,,,,, +2300,3.2952952,3.9000711,,,,,,,,,,,,,, +2400,5.1456447,3.8688622,,,,,,,,,,,,,, +2500,3.3030648,3.7171175,,,,,,,,,,,,,, +2600,4.0796523,3.5586543,,,,,,,,,,,,,, +2700,4.2946362,3.5536337,,,,,,,,,,,,,, +2800,3.4494917,3.4041233,,,,,,,,,,,,,, +2900,3.6309683,3.4493475,,,,,,,,,,,,,, +3000,4.5455294,3.3227613,,,,,,,,,,,,,, +3100,3.6274889,3.4171093,,,,,,,,,,,,,, +3200,2.8654137,3.2201793,,,,,,,,,,,,,, +3300,3.6515083,3.1870098,,,,,,,,,,,,,, +3400,3.556906,3.120121,,,,,,,,,,,,,, +3500,3.3634865,3.0035715,,,,,,,,,,,,,, +3600,3.3887494,3.0160909,,,,,,,,,,,,,, +3700,2.7773454,3.1409807,,,,,,,,,,,,,, +3800,3.0214849,2.888189,,,,,,,,,,,,,, +3900,3.201739,2.9495168,,,,,,,,,,,,,, +4000,2.9251468,2.7658713,,,,,,,,,,,,,, +4100,3.714607,2.8475485,,,,,,,,,,,,,, +4200,2.4411838,2.8455422,,,,,,,,,,,,,, +4300,3.9159207,2.7667332,,,,,,,,,,,,,, +4400,2.2171683,2.6653366,,,,,,,,,,,,,, +4500,3.2516491,2.6382704,,,,,,,,,,,,,, +4600,2.9541252,2.5777576,,,,,,,,,,,,,, +4700,2.4442534,2.6245637,,,,,,,,,,,,,, +4800,2.2688074,2.6449327,,,,,,,,,,,,,, +4900,2.9135249,2.580424,,,,,,,,,,,,,, +5000,3.0289237,2.5000772,,,,,,,,,,,,,, +5100,2.3814201,2.496798,,,,,,,,,,,,,, +5200,2.5973732,2.364229,,,,,,,,,,,,,, +5300,2.1166618,2.4033532,,,,,,,,,,,,,, +5400,2.644618,2.462356,,,,,,,,,,,,,, +5500,2.4153526,2.2745779,,,,,,,,,,,,,, +5600,2.4079883,2.4579966,,,,,,,,,,,,,, +5700,2.4223523,2.315286,,,,,,,,,,,,,, +5800,2.5749023,2.3781397,,,,,,,,,,,,,, +5900,2.4993236,2.2629445,,,,,,,,,,,,,, +6000,2.5902412,2.2581728,,,,,,,,,,,,,, +6100,2.0480173,2.2383144,,,,,,,,,,,,,, +6200,2.4927285,2.2575788,,,,,,,,,,,,,, +6300,2.9313102,2.2117476,,,,,,,,,,,,,, +6395,,,0.6256776,1.5689651,0.56868,1.8507421,50000.0,0.43960002,2.5965824,10000.0,2049.826003551483,2136.365298271179,2049.826003551483,86.44905090332031,0.0397703647613525,0.0 +6400,2.3953588,2.382955,,,,,,,,,,,,,, +6500,2.4519715,2.1553526,,,,,,,,,,,,,, +6600,2.8765345,2.1879113,,,,,,,,,,,,,, +6700,2.594625,2.20575,,,,,,,,,,,,,, +6800,1.8977156,2.0885458,,,,,,,,,,,,,, +6900,2.2655647,2.1886926,,,,,,,,,,,,,, +7000,2.686015,2.0966115,,,,,,,,,,,,,, +7100,2.738533,2.06679,,,,,,,,,,,,,, +7200,2.5492182,2.0317254,,,,,,,,,,,,,, +7300,2.3847198,2.13496,,,,,,,,,,,,,, +7400,3.0261738,2.1404996,,,,,,,,,,,,,, +7500,2.278535,2.0511916,,,,,,,,,,,,,, +7600,2.905259,2.020915,,,,,,,,,,,,,, +7700,2.2522068,2.131455,,,,,,,,,,,,,, +7800,2.392212,2.1360214,,,,,,,,,,,,,, +7900,2.0255013,1.9763675,,,,,,,,,,,,,, +8000,2.2612088,1.9049281,,,,,,,,,,,,,, +8100,3.1012897,2.070899,,,,,,,,,,,,,, +8200,2.6022587,1.9978676,,,,,,,,,,,,,, +8300,2.656414,1.9280245,,,,,,,,,,,,,, +8400,2.0691564,2.0323985,,,,,,,,,,,,,, +8500,2.704132,1.9894758,,,,,,,,,,,,,, +8600,2.29189,1.8938234,,,,,,,,,,,,,, +8700,2.863911,1.8969879,,,,,,,,,,,,,, +8800,2.5532844,2.015044,,,,,,,,,,,,,, +8900,3.2368155,1.9592766,,,,,,,,,,,,,, +9000,2.3949528,1.885783,,,,,,,,,,,,,, +9100,2.3167434,1.8994564,,,,,,,,,,,,,, +9200,2.2682154,1.9154273,,,,,,,,,,,,,, +9300,1.7895999,1.7792566,,,,,,,,,,,,,, +9400,2.224772,1.8861719,,,,,,,,,,,,,, +9500,2.1679478,1.923791,,,,,,,,,,,,,, +9600,2.2273233,1.8005538,,,,,,,,,,,,,, +9700,1.8557897,1.798264,,,,,,,,,,,,,, +9800,2.2306025,1.8128015,,,,,,,,,,,,,, +9900,1.8674036,1.850894,,,,,,,,,,,,,, +10000,2.0161433,1.832628,,,,,,,,,,,,,, +10100,2.6098204,1.8939179,,,,,,,,,,,,,, +10200,2.4863982,1.8298337,,,,,,,,,,,,,, +10300,2.119717,1.7870027,,,,,,,,,,,,,, +10400,2.5513494,1.8501476,,,,,,,,,,,,,, +10500,2.1618452,1.7364336,,,,,,,,,,,,,, +10600,2.595633,1.7848547,,,,,,,,,,,,,, +10700,2.3014553,1.7393578,,,,,,,,,,,,,, +10800,2.168651,1.7244539,,,,,,,,,,,,,, +10900,2.1490104,1.7005545,,,,,,,,,,,,,, +11000,2.2833068,1.766995,,,,,,,,,,,,,, +11100,2.2306118,1.7051346,,,,,,,,,,,,,, +11200,2.110808,1.8363931,,,,,,,,,,,,,, +11300,2.2637734,1.7684381,,,,,,,,,,,,,, +11400,2.2170448,1.7492545,,,,,,,,,,,,,, +11500,2.2927787,1.7742791,,,,,,,,,,,,,, +11600,1.9491804,1.6919945,,,,,,,,,,,,,, +11700,2.3708863,1.6370404,,,,,,,,,,,,,, +11800,2.9548175,1.7830241,,,,,,,,,,,,,, +11900,2.252831,1.7220582,,,,,,,,,,,,,, +12000,2.3853114,1.7560424,,,,,,,,,,,,,, +12100,2.2072022,1.7237346,,,,,,,,,,,,,, +12200,2.3745134,1.6723918,,,,,,,,,,,,,, +12300,2.9911559,1.6431628,,,,,,,,,,,,,, +12400,2.4188967,1.7261488,,,,,,,,,,,,,, +12500,2.5256724,1.6804295,,,,,,,,,,,,,, +12600,2.2774715,1.7492124,,,,,,,,,,,,,, +12700,2.005841,1.6487701,,,,,,,,,,,,,, +12800,2.4042256,1.7982483,,,,,,,,,,,,,, +12900,2.1351812,1.6885521,,,,,,,,,,,,,, +13000,2.6341794,1.6708947,,,,,,,,,,,,,, +13100,2.1891923,1.5263959,,,,,,,,,,,,,, +13200,2.1122904,1.6569252,,,,,,,,,,,,,, +13300,2.6336148,1.6350236,,,,,,,,,,,,,, +13400,2.468625,1.5682065,,,,,,,,,,,,,, +13500,2.2449703,1.5940901,,,,,,,,,,,,,, +13600,2.6835952,1.667227,,,,,,,,,,,,,, +13700,2.424216,1.6533635,,,,,,,,,,,,,, +13740,,,0.75713485,0.9526457,0.66587996,1.383565,50000.0,0.5343,2.1114912,10000.0,4045.7496836185455,4154.4920818805695,4045.7496836185455,108.49990367889404,0.1349964141845703,0.0 +13800,2.9247355,1.6157665,,,,,,,,,,,,,, +13900,2.487575,1.7263991,,,,,,,,,,,,,, +14000,2.3142705,1.5998975,,,,,,,,,,,,,, +14100,2.5474887,1.5780584,,,,,,,,,,,,,, +14200,2.2402697,1.6310773,,,,,,,,,,,,,, +14300,2.3761141,1.6258518,,,,,,,,,,,,,, +14400,2.4127097,1.5797105,,,,,,,,,,,,,, +14500,2.642176,1.6048231,,,,,,,,,,,,,, +14600,2.765029,1.6098766,,,,,,,,,,,,,, +14700,2.5833793,1.6327356,,,,,,,,,,,,,, +14800,2.366205,1.55035,,,,,,,,,,,,,, +14900,2.9328175,1.5539522,,,,,,,,,,,,,, +15000,2.5427914,1.5619895,,,,,,,,,,,,,, +15100,2.7473245,1.6411395,,,,,,,,,,,,,, +15200,2.8432212,1.5915171,,,,,,,,,,,,,, +15300,2.4642565,1.6505644,,,,,,,,,,,,,, +15400,2.5879326,1.6801741,,,,,,,,,,,,,, +15500,2.9296286,1.7180151,,,,,,,,,,,,,, +15600,2.6715076,1.5879294,,,,,,,,,,,,,, +15700,2.5941472,1.5102106,,,,,,,,,,,,,, +15800,2.5684788,1.66449,,,,,,,,,,,,,, +15900,2.932039,1.5827501,,,,,,,,,,,,,, +16000,2.9114008,1.6061723,,,,,,,,,,,,,, +16100,3.0662012,1.5521305,,,,,,,,,,,,,, +16200,2.7371128,1.5358162,,,,,,,,,,,,,, +16300,2.929517,1.5253419,,,,,,,,,,,,,, +16400,2.8044367,1.476129,,,,,,,,,,,,,, +16500,3.4758956,1.6188034,,,,,,,,,,,,,, +16600,2.8436186,1.4405282,,,,,,,,,,,,,, +16700,2.9456184,1.6323066,,,,,,,,,,,,,, +16800,3.1221704,1.5898852,,,,,,,,,,,,,, +16900,3.4443002,1.6145122,,,,,,,,,,,,,, +17000,3.114633,1.6092869,,,,,,,,,,,,,, +17100,2.8715944,1.4155016,,,,,,,,,,,,,, +17200,3.3373907,1.5409317,,,,,,,,,,,,,, +17300,2.9945621,1.4538702,,,,,,,,,,,,,, +17400,3.0368295,1.4843043,,,,,,,,,,,,,, +17500,3.0779912,1.4751956,,,,,,,,,,,,,, +17600,3.2113743,1.5701604,,,,,,,,,,,,,, +17700,3.4370143,1.6494673,,,,,,,,,,,,,, +17800,3.4256706,1.5263238,,,,,,,,,,,,,, +17900,3.5184605,1.5226723,,,,,,,,,,,,,, +18000,3.4431999,1.5015571,,,,,,,,,,,,,, +18100,3.6929986,1.4996834,,,,,,,,,,,,,, +18200,3.5989232,1.4192752,,,,,,,,,,,,,, +18300,3.999493,1.5664269,,,,,,,,,,,,,, +18400,4.0275784,1.5067592,,,,,,,,,,,,,, +18500,3.6771107,1.4374093,,,,,,,,,,,,,, +18600,4.272152,1.6163733,,,,,,,,,,,,,, +18700,4.046474,1.5734844,,,,,,,,,,,,,, +18800,4.3629484,1.5825697,,,,,,,,,,,,,, +18900,4.6702614,1.4524597,,,,,,,,,,,,,, +19000,4.4184666,1.5249208,,,,,,,,,,,,,, +19100,4.497393,1.4230453,,,,,,,,,,,,,, +19200,4.585563,1.4709618,,,,,,,,,,,,,, +19300,5.2894516,1.6386707,,,,,,,,,,,,,, +19400,4.795242,1.4674356,,,,,,,,,,,,,, +19500,5.482722,1.4683492,,,,,,,,,,,,,, +19600,5.387566,1.437956,,,,,,,,,,,,,, +19700,5.4920583,1.4219333,,,,,,,,,,,,,, +19800,5.954037,1.5483303,,,,,,,,,,,,,, +19900,6.324367,1.4933286,,,,,,,,,,,,,, +20000,6.676006,1.5251149,,,,,,,,,,,,,, +20100,7.1804256,1.6376086,,,,,,,,,,,,,, +20200,6.8913126,1.5392935,,,,,,,,,,,,,, +20300,6.9660645,1.4771256,,,,,,,,,,,,,, +20400,7.2111435,1.5515683,,,,,,,,,,,,,, +20500,7.457107,1.4731723,,,,,,,,,,,,,, +20600,7.4303656,1.4786524,,,,,,,,,,,,,, +20700,7.629821,1.4592505,,,,,,,,,,,,,, +20800,8.020671,1.4321133,,,,,,,,,,,,,, +20900,8.064123,1.4215634,,,,,,,,,,,,,, +21000,8.373721,1.4359391,,,,,,,,,,,,,, +21092,,,0.80626196,0.7468492,0.69604,1.2423426,50000.0,0.5665,1.966288,10000.0,6041.659068584442,6173.354122161865,6041.659068584442,131.3270001411438,0.2016239166259765,0.0 +21100,9.557383,1.4255931,,,,,,,,,,,,,, +21200,9.634326,1.546961,,,,,,,,,,,,,, +21300,9.685497,1.6118386,,,,,,,,,,,,,, +21400,9.886862,1.5240147,,,,,,,,,,,,,, +21500,9.7975235,1.4754592,,,,,,,,,,,,,, +21600,10.452477,1.5948453,,,,,,,,,,,,,, +21700,11.104937,1.4480739,,,,,,,,,,,,,, +21800,11.443251,1.567525,,,,,,,,,,,,,, +21900,11.143262,1.4766202,,,,,,,,,,,,,, +22000,11.234718,1.4688125,,,,,,,,,,,,,, +22100,11.976441,1.5527772,,,,,,,,,,,,,, +22200,11.653145,1.5262766,,,,,,,,,,,,,, +22300,12.515902,1.4616196,,,,,,,,,,,,,, +22400,12.398321,1.5515943,,,,,,,,,,,,,, +22500,11.966619,1.5430887,,,,,,,,,,,,,, +22600,12.671228,1.5162046,,,,,,,,,,,,,, +22700,12.063148,1.43737,,,,,,,,,,,,,, +22800,12.623193,1.5195613,,,,,,,,,,,,,, +22900,12.841315,1.5209558,,,,,,,,,,,,,, +23000,12.794352,1.5264404,,,,,,,,,,,,,, +23100,14.1313505,1.5459316,,,,,,,,,,,,,, +23200,13.10057,1.5468302,,,,,,,,,,,,,, +23300,13.00319,1.4513229,,,,,,,,,,,,,, +23400,12.7899065,1.497438,,,,,,,,,,,,,, +23500,13.094699,1.4613187,,,,,,,,,,,,,, +23600,12.8479395,1.6071806,,,,,,,,,,,,,, +23700,13.217056,1.451887,,,,,,,,,,,,,, +23800,13.095944,1.526212,,,,,,,,,,,,,, +23900,12.245737,1.4961656,,,,,,,,,,,,,, +24000,13.137688,1.4849154,,,,,,,,,,,,,, +24100,12.375477,1.4814997,,,,,,,,,,,,,, +24200,12.37206,1.426897,,,,,,,,,,,,,, +24300,12.036983,1.3881594,,,,,,,,,,,,,, +24400,11.92641,1.4728968,,,,,,,,,,,,,, +24500,11.771098,1.4304404,,,,,,,,,,,,,, +24600,11.649035,1.5236187,,,,,,,,,,,,,, +24700,11.907681,1.4861745,,,,,,,,,,,,,, +24800,10.68249,1.3228707,,,,,,,,,,,,,, +24900,11.951889,1.5780934,,,,,,,,,,,,,, +25000,11.263922,1.5151592,,,,,,,,,,,,,, +25100,10.915551,1.4973717,,,,,,,,,,,,,, +25200,11.098624,1.4847349,,,,,,,,,,,,,, +25300,10.891446,1.4778204,,,,,,,,,,,,,, +25400,10.664565,1.4708545,,,,,,,,,,,,,, +25500,10.126752,1.4110153,,,,,,,,,,,,,, +25600,10.253349,1.383716,,,,,,,,,,,,,, +25700,9.889932,1.3123438,,,,,,,,,,,,,, +25800,9.558504,1.4940456,,,,,,,,,,,,,, +25900,9.501413,1.448665,,,,,,,,,,,,,, +26000,9.6988,1.4464637,,,,,,,,,,,,,, +26100,9.408886,1.447676,,,,,,,,,,,,,, +26200,9.650246,1.3825984,,,,,,,,,,,,,, +26300,9.363245,1.4882677,,,,,,,,,,,,,, +26400,9.168091,1.5331602,,,,,,,,,,,,,, +26500,9.169416,1.4329627,,,,,,,,,,,,,, +26600,9.049761,1.4842701,,,,,,,,,,,,,, +26700,8.756247,1.4040155,,,,,,,,,,,,,, +26800,9.146037,1.5803797,,,,,,,,,,,,,, +26900,8.695668,1.4477241,,,,,,,,,,,,,, +27000,8.426908,1.4665682,,,,,,,,,,,,,, +27100,8.216556,1.4632483,,,,,,,,,,,,,, +27200,8.094534,1.36522,,,,,,,,,,,,,, +27300,8.046776,1.351177,,,,,,,,,,,,,, +27400,7.874637,1.3922625,,,,,,,,,,,,,, +27500,7.0431967,1.3832042,,,,,,,,,,,,,, +27600,7.395363,1.4512374,,,,,,,,,,,,,, +27700,7.6031857,1.4768431,,,,,,,,,,,,,, +27800,7.716082,1.4309497,,,,,,,,,,,,,, +27900,7.284078,1.3665826,,,,,,,,,,,,,, +28000,7.0502796,1.3327098,,,,,,,,,,,,,, +28100,6.550111,1.2366924,,,,,,,,,,,,,, +28200,6.895546,1.3160229,,,,,,,,,,,,,, +28300,7.538845,1.4293623,,,,,,,,,,,,,, +28400,6.9126673,1.3876631,,,,,,,,,,,,,, +28443,,,0.82924104,0.63326895,0.71349996,1.1654677,50000.0,0.58580005,1.8870754,10000.0,8037.608729124069,8190.942818403244,8037.608729124069,152.86800980567932,0.2448570728302002,0.0 +28500,6.8568554,1.3344162,,,,,,,,,,,,,, +28600,7.0242634,1.4856795,,,,,,,,,,,,,, +28700,6.6532207,1.4719828,,,,,,,,,,,,,, +28800,6.2598543,1.3663869,,,,,,,,,,,,,, +28900,6.5984015,1.3714626,,,,,,,,,,,,,, +29000,6.4800963,1.3752112,,,,,,,,,,,,,, +29100,6.5518923,1.385139,,,,,,,,,,,,,, +29200,6.235161,1.389632,,,,,,,,,,,,,, +29300,6.6323204,1.4930956,,,,,,,,,,,,,, +29400,5.875154,1.3274324,,,,,,,,,,,,,, +29500,6.2344327,1.3566339,,,,,,,,,,,,,, +29600,6.3889647,1.4252231,,,,,,,,,,,,,, +29700,5.952165,1.3801355,,,,,,,,,,,,,, +29800,6.346652,1.4509933,,,,,,,,,,,,,, +29900,5.93223,1.3031559,,,,,,,,,,,,,, +30000,5.5059834,1.1938716,,,,,,,,,,,,,, +30100,5.7550116,1.3663808,,,,,,,,,,,,,, +30200,5.7394986,1.3611279,,,,,,,,,,,,,, +30300,5.9506474,1.3464634,,,,,,,,,,,,,, +30400,5.5565534,1.3372282,,,,,,,,,,,,,, +30500,5.97622,1.312081,,,,,,,,,,,,,, +30600,5.419107,1.2911379,,,,,,,,,,,,,, +30700,5.577155,1.2933437,,,,,,,,,,,,,, +30800,5.660573,1.4083804,,,,,,,,,,,,,, +30900,5.1301847,1.2073398,,,,,,,,,,,,,, +31000,5.482844,1.4166117,,,,,,,,,,,,,, +31100,5.173871,1.2612526,,,,,,,,,,,,,, +31200,5.4598465,1.3581263,,,,,,,,,,,,,, +31300,5.2973504,1.3610317,,,,,,,,,,,,,, +31400,5.414575,1.346878,,,,,,,,,,,,,, +31500,5.036264,1.2264533,,,,,,,,,,,,,, +31600,5.2699614,1.346164,,,,,,,,,,,,,, +31700,5.4402323,1.2917888,,,,,,,,,,,,,, +31800,5.334868,1.3250829,,,,,,,,,,,,,, +31900,5.132401,1.4447329,,,,,,,,,,,,,, +32000,5.207422,1.3273389,,,,,,,,,,,,,, +32100,4.952977,1.2982454,,,,,,,,,,,,,, +32200,4.7783484,1.2664382,,,,,,,,,,,,,, +32300,4.936419,1.2475731,,,,,,,,,,,,,, +32400,5.110879,1.3842224,,,,,,,,,,,,,, +32500,4.8682923,1.2670761,,,,,,,,,,,,,, +32600,5.0672626,1.3139801,,,,,,,,,,,,,, +32700,5.010432,1.3385253,,,,,,,,,,,,,, +32800,4.581557,1.243595,,,,,,,,,,,,,, +32900,5.0270963,1.3686929,,,,,,,,,,,,,, +33000,5.2375674,1.37363,,,,,,,,,,,,,, +33100,4.7983227,1.4210114,,,,,,,,,,,,,, +33200,4.6094522,1.3469799,,,,,,,,,,,,,, +33300,4.5315213,1.3296075,,,,,,,,,,,,,, +33400,4.7060895,1.3575106,,,,,,,,,,,,,, +33500,4.759145,1.4054676,,,,,,,,,,,,,, +33600,4.575412,1.2726943,,,,,,,,,,,,,, +33700,4.5807476,1.2526944,,,,,,,,,,,,,, +33800,4.4947453,1.3171679,,,,,,,,,,,,,, +33900,4.5195184,1.2414218,,,,,,,,,,,,,, +34000,4.661246,1.2922688,,,,,,,,,,,,,, +34100,4.629753,1.3563446,,,,,,,,,,,,,, +34200,4.885884,1.301853,,,,,,,,,,,,,, +34300,4.390519,1.2564466,,,,,,,,,,,,,, +34400,4.2090483,1.2834184,,,,,,,,,,,,,, +34500,4.7766275,1.2690743,,,,,,,,,,,,,, +34600,4.430151,1.2343917,,,,,,,,,,,,,, +34700,4.621642,1.2821805,,,,,,,,,,,,,, +34800,4.6897693,1.3922069,,,,,,,,,,,,,, +34900,4.467527,1.3079548,,,,,,,,,,,,,, +35000,4.3583736,1.3244609,,,,,,,,,,,,,, +35100,4.3338265,1.2724457,,,,,,,,,,,,,, +35200,4.317161,1.2240956,,,,,,,,,,,,,, +35300,4.5857625,1.24436,,,,,,,,,,,,,, +35400,4.247714,1.3141418,,,,,,,,,,,,,, +35500,4.3560786,1.2321714,,,,,,,,,,,,,, +35600,4.1245217,1.1889051,,,,,,,,,,,,,, +35700,4.432794,1.2831471,,,,,,,,,,,,,, +35796,,,0.8547313,0.5380873,0.72433996,1.1205891,50000.0,0.6003,1.8511142,10000.0,10033.702872037888,10209.286883592606,10033.702872037888,174.9927954673767,0.3151042461395263,0.0 +35800,4.720263,1.3348483,,,,,,,,,,,,,, +35900,4.24461,1.2421958,,,,,,,,,,,,,, +36000,4.55009,1.3328924,,,,,,,,,,,,,, +36100,4.172215,1.2736306,,,,,,,,,,,,,, +36200,4.5076027,1.4264014,,,,,,,,,,,,,, +36300,4.25438,1.2949915,,,,,,,,,,,,,, +36400,4.2810583,1.325943,,,,,,,,,,,,,, +36500,4.1195726,1.1970545,,,,,,,,,,,,,, +36600,4.1709924,1.268266,,,,,,,,,,,,,, +36700,4.2175097,1.2188754,,,,,,,,,,,,,, +36800,4.251856,1.2839425,,,,,,,,,,,,,, +36900,4.392795,1.2740784,,,,,,,,,,,,,, +37000,4.3447466,1.3322309,,,,,,,,,,,,,, +37100,4.1499,1.2270205,,,,,,,,,,,,,, +37200,4.3094363,1.2717118,,,,,,,,,,,,,, +37300,4.1960526,1.4266336,,,,,,,,,,,,,, +37400,4.1613145,1.2849329,,,,,,,,,,,,,, +37500,3.8893168,1.2044969,,,,,,,,,,,,,, +37600,4.477601,1.2356882,,,,,,,,,,,,,, +37700,4.040525,1.2609394,,,,,,,,,,,,,, +37800,4.063315,1.3086318,,,,,,,,,,,,,, +37900,4.1932855,1.2433574,,,,,,,,,,,,,, +38000,4.101161,1.3293921,,,,,,,,,,,,,, +38100,3.9735196,1.2837492,,,,,,,,,,,,,, +38200,3.9089363,1.1842885,,,,,,,,,,,,,, +38300,4.223768,1.2818279,,,,,,,,,,,,,, +38400,4.141959,1.287684,,,,,,,,,,,,,, +38500,4.3084264,1.3033929,,,,,,,,,,,,,, +38600,4.0902386,1.2827363,,,,,,,,,,,,,, +38700,4.0125623,1.2983997,,,,,,,,,,,,,, +38800,4.2160506,1.306865,,,,,,,,,,,,,, +38900,3.9346266,1.1754285,,,,,,,,,,,,,, +39000,3.8839405,1.250699,,,,,,,,,,,,,, +39100,4.1025615,1.3252113,,,,,,,,,,,,,, +39200,3.6854343,1.1950884,,,,,,,,,,,,,, +39300,3.9448538,1.267603,,,,,,,,,,,,,, +39400,4.269415,1.2683179,,,,,,,,,,,,,, +39500,3.9312425,1.193228,,,,,,,,,,,,,, +39600,4.0284758,1.2327142,,,,,,,,,,,,,, +39700,3.7256799,1.211329,,,,,,,,,,,,,, +39800,3.8251305,1.2429365,,,,,,,,,,,,,, +39900,3.7393782,1.2069926,,,,,,,,,,,,,, +40000,4.148505,1.274508,,,,,,,,,,,,,, +40100,4.037039,1.2353258,,,,,,,,,,,,,, +40200,3.8967595,1.2146053,,,,,,,,,,,,,, +40300,4.2701206,1.3373218,,,,,,,,,,,,,, +40400,3.6732392,1.1482446,,,,,,,,,,,,,, +40500,3.8309827,1.2190422,,,,,,,,,,,,,, +40600,4.0349503,1.2846187,,,,,,,,,,,,,, +40700,3.9151433,1.2480208,,,,,,,,,,,,,, +40800,3.9472353,1.2184482,,,,,,,,,,,,,, +40900,3.802208,1.1542077,,,,,,,,,,,,,, +41000,3.997094,1.2504532,,,,,,,,,,,,,, +41100,4.144732,1.2411139,,,,,,,,,,,,,, +41200,3.8910005,1.3247635,,,,,,,,,,,,,, +41300,3.8678012,1.322186,,,,,,,,,,,,,, +41400,3.790781,1.2397532,,,,,,,,,,,,,, +41500,3.8935764,1.316673,,,,,,,,,,,,,, +41600,3.8190613,1.2116675,,,,,,,,,,,,,, +41700,4.1057153,1.2219505,,,,,,,,,,,,,, +41800,3.9582517,1.2603569,,,,,,,,,,,,,, +41900,3.9973683,1.3004413,,,,,,,,,,,,,, +42000,3.8785512,1.3307416,,,,,,,,,,,,,, +42100,3.9055479,1.2393316,,,,,,,,,,,,,, +42200,4.066608,1.2006257,,,,,,,,,,,,,, +42300,3.814457,1.2279122,,,,,,,,,,,,,, +42400,4.1929526,1.1712867,,,,,,,,,,,,,, +42500,3.854583,1.1758785,,,,,,,,,,,,,, +42600,3.9803956,1.202991,,,,,,,,,,,,,, +42700,3.9088147,1.239029,,,,,,,,,,,,,, +42800,4.031144,1.2118125,,,,,,,,,,,,,, +42900,3.9245791,1.1612183,,,,,,,,,,,,,, +43000,3.6625364,1.1592093,,,,,,,,,,,,,, +43100,4.054053,1.1741749,,,,,,,,,,,,,, +43147,,,0.871313,0.46765828,0.72818,1.1009811,50000.0,0.6056,1.8409466,10000.0,12029.607981204988,12226.716282129288,12029.607981204988,196.35537314414967,0.4192686080932617,0.0 +43200,3.878725,1.1969955,,,,,,,,,,,,,, +43300,3.9285033,1.1863334,,,,,,,,,,,,,, +43400,3.9330485,1.2109174,,,,,,,,,,,,,, +43500,4.3333216,1.2290897,,,,,,,,,,,,,, +43600,4.000523,1.3087717,,,,,,,,,,,,,, +43700,3.8765874,1.2098255,,,,,,,,,,,,,, +43800,3.9283855,1.3042747,,,,,,,,,,,,,, +43900,3.8389394,1.2098432,,,,,,,,,,,,,, +44000,3.8067799,1.157721,,,,,,,,,,,,,, +44100,3.9657662,1.2061086,,,,,,,,,,,,,, +44200,4.240312,1.2089682,,,,,,,,,,,,,, +44300,3.5896022,1.1802446,,,,,,,,,,,,,, +44400,3.4798841,1.0906482,,,,,,,,,,,,,, +44500,3.727335,1.1844299,,,,,,,,,,,,,, +44600,3.6461813,1.1912581,,,,,,,,,,,,,, +44700,4.1520276,1.22844,,,,,,,,,,,,,, +44800,3.6859236,1.2416346,,,,,,,,,,,,,, +44900,4.000932,1.2352442,,,,,,,,,,,,,, +45000,4.4288845,1.3219464,,,,,,,,,,,,,, +45100,3.5970035,1.1054151,,,,,,,,,,,,,, +45200,3.5812685,1.1448531,,,,,,,,,,,,,, +45300,4.2007127,1.2616429,,,,,,,,,,,,,, +45400,3.523731,1.1928092,,,,,,,,,,,,,, +45500,3.8673406,1.241123,,,,,,,,,,,,,, +45600,3.750519,1.2503902,,,,,,,,,,,,,, +45700,3.8615806,1.1768751,,,,,,,,,,,,,, +45800,3.8128326,1.1711885,,,,,,,,,,,,,, +45900,3.73112,1.2256352,,,,,,,,,,,,,, +46000,3.6924758,1.1618521,,,,,,,,,,,,,, +46100,3.806561,1.1767731,,,,,,,,,,,,,, +46200,3.9275806,1.2225554,,,,,,,,,,,,,, +46300,3.90748,1.176085,,,,,,,,,,,,,, +46400,3.7055101,1.1467383,,,,,,,,,,,,,, +46500,3.7572443,1.1796646,,,,,,,,,,,,,, +46600,3.9603329,1.3142532,,,,,,,,,,,,,, +46700,3.718492,1.335825,,,,,,,,,,,,,, +46800,4.0999575,1.2506394,,,,,,,,,,,,,, +46900,3.6402297,1.1925693,,,,,,,,,,,,,, +47000,3.8582342,1.1832149,,,,,,,,,,,,,, +47100,3.6134398,1.1921444,,,,,,,,,,,,,, +47200,3.48638,1.1003852,,,,,,,,,,,,,, +47300,3.907638,1.1034411,,,,,,,,,,,,,, +47400,3.609619,1.1048025,,,,,,,,,,,,,, +47500,4.064024,1.2876694,,,,,,,,,,,,,, +47600,3.8037753,1.234202,,,,,,,,,,,,,, +47700,3.8401964,1.2875459,,,,,,,,,,,,,, +47800,3.7443476,1.2515287,,,,,,,,,,,,,, +47900,3.7716393,1.1102865,,,,,,,,,,,,,, +48000,3.7963939,1.2691946,,,,,,,,,,,,,, +48100,3.6345663,1.2292393,,,,,,,,,,,,,, +48200,3.9440987,1.2371404,,,,,,,,,,,,,, +48300,3.7003124,1.1114054,,,,,,,,,,,,,, +48400,3.803473,1.2181332,,,,,,,,,,,,,, +48500,3.8531678,1.154243,,,,,,,,,,,,,, +48600,3.8309348,1.2079558,,,,,,,,,,,,,, +48700,3.8702538,1.141825,,,,,,,,,,,,,, +48800,3.6278715,1.2293398,,,,,,,,,,,,,, +48900,3.8061736,1.166743,,,,,,,,,,,,,, +49000,4.0323386,1.2822518,,,,,,,,,,,,,, +49100,4.057734,1.3354738,,,,,,,,,,,,,, +49200,3.8730445,1.2915927,,,,,,,,,,,,,, +49300,3.554855,1.081322,,,,,,,,,,,,,, +49400,3.8711817,1.2754942,,,,,,,,,,,,,, +49500,3.7301388,1.2191672,,,,,,,,,,,,,, +49600,3.763734,1.166129,,,,,,,,,,,,,, +49700,3.7555523,1.1636212,,,,,,,,,,,,,, +49800,4.0949926,1.2108474,,,,,,,,,,,,,, +49900,3.79191,1.1921856,,,,,,,,,,,,,, +50000,3.902987,1.2289088,,,,,,,,,,,,,, +50100,3.8318844,1.2076378,,,,,,,,,,,,,, +50200,3.9105802,1.214437,,,,,,,,,,,,,, +50300,4.017617,1.2429807,,,,,,,,,,,,,, +50400,4.0535426,1.1752005,,,,,,,,,,,,,, +50500,3.8965697,1.188702,,,,,,,,,,,,,, +50502,,,0.8854432,0.41532198,0.7328,1.0901387,50000.0,0.61230004,1.8371395,10000.0,14025.61386179924,14243.602019548416,14025.61386179924,217.0973782539368,0.5025744438171387,0.0 +50600,3.8640666,1.166022,,,,,,,,,,,,,, +50700,4.0280957,1.2646285,,,,,,,,,,,,,, +50800,3.8004036,1.2110083,,,,,,,,,,,,,, +50900,4.1712875,1.221666,,,,,,,,,,,,,, +51000,3.6798563,1.2483498,,,,,,,,,,,,,, +51100,4.0799065,1.2520487,,,,,,,,,,,,,, +51200,3.7118425,1.1630173,,,,,,,,,,,,,, +51300,3.960953,1.1060419,,,,,,,,,,,,,, +51400,3.8259892,1.2153922,,,,,,,,,,,,,, +51500,3.918326,1.1542542,,,,,,,,,,,,,, +51600,3.7876623,1.2338355,,,,,,,,,,,,,, +51700,3.6581404,1.1320767,,,,,,,,,,,,,, +51800,3.8201616,1.0641094,,,,,,,,,,,,,, +51900,4.0517793,1.1623816,,,,,,,,,,,,,, +52000,4.00863,1.2615082,,,,,,,,,,,,,, +52100,4.1107445,1.1734048,,,,,,,,,,,,,, +52200,4.136712,1.2130182,,,,,,,,,,,,,, +52300,4.2606363,1.2912178,,,,,,,,,,,,,, +52400,3.8009052,1.0987062,,,,,,,,,,,,,, +52500,3.7023284,1.2087016,,,,,,,,,,,,,, +52600,4.2564816,1.276298,,,,,,,,,,,,,, +52700,3.839111,1.2562456,,,,,,,,,,,,,, +52800,3.7856464,1.1561302,,,,,,,,,,,,,, +52900,3.9264734,1.1729922,,,,,,,,,,,,,, +53000,4.021919,1.2395172,,,,,,,,,,,,,, +53100,3.881984,1.0995145,,,,,,,,,,,,,, +53200,4.120936,1.2103832,,,,,,,,,,,,,, +53300,4.0209293,1.3110694,,,,,,,,,,,,,, +53400,4.068738,1.198065,,,,,,,,,,,,,, +53500,3.9213939,1.112296,,,,,,,,,,,,,, +53600,3.740293,1.1453027,,,,,,,,,,,,,, +53700,3.9754703,1.167479,,,,,,,,,,,,,, +53800,3.921827,1.1551557,,,,,,,,,,,,,, +53900,3.861023,1.2060766,,,,,,,,,,,,,, +54000,3.7028055,1.1905566,,,,,,,,,,,,,, +54100,3.826962,1.1583554,,,,,,,,,,,,,, +54200,3.9701865,1.1822278,,,,,,,,,,,,,, +54300,3.6424434,1.0892633,,,,,,,,,,,,,, +54400,4.111021,1.2418785,,,,,,,,,,,,,, +54500,3.9327185,1.1298925,,,,,,,,,,,,,, +54600,3.8848248,1.1510432,,,,,,,,,,,,,, +54700,3.9794075,1.1345435,,,,,,,,,,,,,, +54800,4.4841633,1.2013738,,,,,,,,,,,,,, +54900,3.7404857,1.0867016,,,,,,,,,,,,,, +55000,3.9123907,1.2236469,,,,,,,,,,,,,, +55100,4.3716426,1.2296147,,,,,,,,,,,,,, +55200,3.7047322,1.1476979,,,,,,,,,,,,,, +55300,3.816472,1.2530737,,,,,,,,,,,,,, +55400,3.771028,1.1119556,,,,,,,,,,,,,, +55500,3.8951535,1.1224585,,,,,,,,,,,,,, +55600,4.360356,1.2300556,,,,,,,,,,,,,, +55700,3.8989036,1.2129848,,,,,,,,,,,,,, +55800,3.8862464,1.1935923,,,,,,,,,,,,,, +55900,3.786191,1.1556485,,,,,,,,,,,,,, +56000,4.165676,1.2139237,,,,,,,,,,,,,, +56100,4.0112224,1.151546,,,,,,,,,,,,,, +56200,4.1312814,1.2806158,,,,,,,,,,,,,, +56300,3.8533664,1.0993503,,,,,,,,,,,,,, +56400,4.043814,1.1235061,,,,,,,,,,,,,, +56500,4.089488,1.2926569,,,,,,,,,,,,,, +56600,3.7214637,1.2155168,,,,,,,,,,,,,, +56700,3.8828092,1.1416777,,,,,,,,,,,,,, +56800,4.081435,1.2013619,,,,,,,,,,,,,, +56900,3.9855316,1.2298651,,,,,,,,,,,,,, +57000,3.9699361,1.1900685,,,,,,,,,,,,,, +57100,4.2284517,1.2142601,,,,,,,,,,,,,, +57200,4.074763,1.2213061,,,,,,,,,,,,,, +57300,4.171146,1.1951604,,,,,,,,,,,,,, +57400,3.9170423,1.141633,,,,,,,,,,,,,, +57500,3.8793526,1.1436837,,,,,,,,,,,,,, +57600,4.2993712,1.2836177,,,,,,,,,,,,,, +57700,3.9988062,1.1603652,,,,,,,,,,,,,, +57800,3.765184,1.1006163,,,,,,,,,,,,,, +57855,,,0.89253825,0.3850234,0.73662,1.0837384,50000.0,0.609,1.840022,10000.0,16021.685701847076,16260.028512001038,16021.685701847076,237.31938290596008,0.5800423622131348,0.0 +57900,4.0411186,1.2027671,,,,,,,,,,,,,, +58000,4.1641936,1.2515063,,,,,,,,,,,,,, +58100,3.8585994,1.1270262,,,,,,,,,,,,,, +58200,3.9867733,1.1456878,,,,,,,,,,,,,, +58300,4.2866106,1.2606316,,,,,,,,,,,,,, +58400,3.9092772,1.1484383,,,,,,,,,,,,,, +58500,4.1088157,1.1697139,,,,,,,,,,,,,, +58600,4.694065,1.2007215,,,,,,,,,,,,,, +58700,3.8600073,1.1486828,,,,,,,,,,,,,, +58800,3.7497315,1.0931408,,,,,,,,,,,,,, +58900,4.0204844,1.2458053,,,,,,,,,,,,,, +59000,4.5567074,1.228139,,,,,,,,,,,,,, +59100,4.1439185,1.0857952,,,,,,,,,,,,,, +59200,4.0446568,1.1223049,,,,,,,,,,,,,, +59300,4.290148,1.1408072,,,,,,,,,,,,,, +59400,4.181192,1.2643784,,,,,,,,,,,,,, +59500,4.127741,1.2178974,,,,,,,,,,,,,, +59600,3.9982023,1.0730668,,,,,,,,,,,,,, +59700,4.51132,1.2135177,,,,,,,,,,,,,, +59800,4.1749825,1.2108475,,,,,,,,,,,,,, +59900,4.146835,1.1923497,,,,,,,,,,,,,, +60000,3.7286127,1.1063026,,,,,,,,,,,,,, +60100,4.0545516,1.1324818,,,,,,,,,,,,,, +60200,4.18634,1.1639116,,,,,,,,,,,,,, +60300,4.0274634,1.1428424,,,,,,,,,,,,,, +60400,4.0977106,1.1667376,,,,,,,,,,,,,, +60500,4.1667304,1.2456499,,,,,,,,,,,,,, +60600,4.436457,1.1989213,,,,,,,,,,,,,, +60700,4.265892,1.2106388,,,,,,,,,,,,,, +60800,4.073124,1.1143284,,,,,,,,,,,,,, +60900,4.290305,1.1338751,,,,,,,,,,,,,, +61000,4.1896877,1.1182227,,,,,,,,,,,,,, +61100,4.0874505,1.1407794,,,,,,,,,,,,,, +61200,3.8766754,1.0659885,,,,,,,,,,,,,, +61300,4.286856,1.1069067,,,,,,,,,,,,,, +61400,4.276536,1.175674,,,,,,,,,,,,,, +61500,4.2735295,1.1471272,,,,,,,,,,,,,, +61600,4.194719,1.1037203,,,,,,,,,,,,,, +61700,4.015727,1.0487634,,,,,,,,,,,,,, +61800,4.5725975,1.1253948,,,,,,,,,,,,,, +61900,4.238475,1.1076355,,,,,,,,,,,,,, +62000,4.157261,1.1408571,,,,,,,,,,,,,, +62100,4.37157,1.263484,,,,,,,,,,,,,, +62200,4.365885,1.1170647,,,,,,,,,,,,,, +62300,4.323107,1.1869495,,,,,,,,,,,,,, +62400,4.6430287,1.2196738,,,,,,,,,,,,,, +62500,4.5388174,1.2222493,,,,,,,,,,,,,, +62600,3.9685667,1.0584579,,,,,,,,,,,,,, +62700,4.538475,1.1615205,,,,,,,,,,,,,, +62800,4.8223515,1.1555804,,,,,,,,,,,,,, +62900,4.3452463,1.0453455,,,,,,,,,,,,,, +63000,4.433429,1.2152523,,,,,,,,,,,,,, +63100,4.2220216,1.1546389,,,,,,,,,,,,,, +63200,4.4907217,1.1182039,,,,,,,,,,,,,, +63300,4.523039,1.0728405,,,,,,,,,,,,,, +63400,4.37359,1.1755,,,,,,,,,,,,,, +63500,4.2480044,1.0903472,,,,,,,,,,,,,, +63600,4.273742,1.1364958,,,,,,,,,,,,,, +63700,4.640417,1.1735934,,,,,,,,,,,,,, +63800,4.2348747,1.0893021,,,,,,,,,,,,,, +63900,4.455886,1.1877308,,,,,,,,,,,,,, +64000,4.4386563,1.139983,,,,,,,,,,,,,, +64100,4.408264,1.1444511,,,,,,,,,,,,,, +64200,4.317006,1.1207876,,,,,,,,,,,,,, +64300,4.4671597,1.1381729,,,,,,,,,,,,,, +64400,4.6067905,1.0970186,,,,,,,,,,,,,, +64500,4.512941,1.2085509,,,,,,,,,,,,,, +64600,4.5796494,1.1796603,,,,,,,,,,,,,, +64700,4.497619,1.2076114,,,,,,,,,,,,,, +64800,4.4055586,1.156302,,,,,,,,,,,,,, +64900,4.502172,1.0322864,,,,,,,,,,,,,, +65000,4.5340347,1.2412741,,,,,,,,,,,,,, +65100,4.6257486,1.1614686,,,,,,,,,,,,,, +65200,4.7350383,1.147482,,,,,,,,,,,,,, +65208,,,0.90124756,0.35447967,0.73748,1.0785924,50000.0,0.6104,1.8365499,10000.0,18017.576412677765,18276.52331066132,18017.576412677765,257.79344487190247,0.6483263969421387,0.0 +65300,4.4084673,1.1686174,,,,,,,,,,,,,, +65400,4.625331,1.161104,,,,,,,,,,,,,, +65500,4.851145,1.2499616,,,,,,,,,,,,,, +65600,4.8405247,1.1195295,,,,,,,,,,,,,, +65700,4.6255918,1.1100072,,,,,,,,,,,,,, +65800,4.627632,1.1573273,,,,,,,,,,,,,, +65900,4.6867137,1.0334947,,,,,,,,,,,,,, +66000,4.385501,1.037397,,,,,,,,,,,,,, +66100,4.7385035,1.1736593,,,,,,,,,,,,,, +66200,4.7864466,1.1743108,,,,,,,,,,,,,, +66300,4.6372094,1.168529,,,,,,,,,,,,,, +66400,4.6159496,1.0552441,,,,,,,,,,,,,, +66500,4.2832046,1.115658,,,,,,,,,,,,,, +66600,4.3917117,1.0595005,,,,,,,,,,,,,, +66700,5.0789943,1.220778,,,,,,,,,,,,,, +66800,4.6484976,1.1116327,,,,,,,,,,,,,, +66900,4.7546296,1.1105635,,,,,,,,,,,,,, +67000,5.1559033,1.2200499,,,,,,,,,,,,,, +67100,4.4857783,1.0552616,,,,,,,,,,,,,, +67200,4.5000124,1.0578072,,,,,,,,,,,,,, +67300,4.6727366,1.1248643,,,,,,,,,,,,,, +67400,4.8452716,1.1648583,,,,,,,,,,,,,, +67500,4.82269,1.1714895,,,,,,,,,,,,,, +67600,5.0031233,1.1086652,,,,,,,,,,,,,, +67700,4.752123,1.1819581,,,,,,,,,,,,,, +67800,4.759059,1.1221896,,,,,,,,,,,,,, +67900,4.8130493,1.1828184,,,,,,,,,,,,,, +68000,4.950571,1.1001052,,,,,,,,,,,,,, +68100,4.8005114,1.1944144,,,,,,,,,,,,,, +68200,4.589383,1.043499,,,,,,,,,,,,,, +68300,4.951854,1.1029888,,,,,,,,,,,,,, +68400,4.902244,1.056746,,,,,,,,,,,,,, +68500,4.8468347,1.1349437,,,,,,,,,,,,,, +68600,4.81879,1.0974411,,,,,,,,,,,,,, +68700,4.9317784,1.1991532,,,,,,,,,,,,,, +68800,5.0510364,1.1476306,,,,,,,,,,,,,, +68900,4.8798795,1.1401823,,,,,,,,,,,,,, +69000,4.779599,1.1636689,,,,,,,,,,,,,, +69100,4.7336864,1.138608,,,,,,,,,,,,,, +69200,4.913373,1.0821601,,,,,,,,,,,,,, +69300,4.538521,1.0056187,,,,,,,,,,,,,, +69400,4.6859126,1.0542536,,,,,,,,,,,,,, +69500,5.2250147,1.2228357,,,,,,,,,,,,,, +69600,4.832305,1.0613253,,,,,,,,,,,,,, +69700,4.997441,1.0843691,,,,,,,,,,,,,, +69800,5.0918427,1.1361145,,,,,,,,,,,,,, +69900,5.0581775,1.0965068,,,,,,,,,,,,,, +70000,5.1162868,1.1150178,,,,,,,,,,,,,, +70100,5.034357,1.0822905,,,,,,,,,,,,,, +70200,5.0786953,1.2017645,,,,,,,,,,,,,, +70300,4.842858,1.0396072,,,,,,,,,,,,,, +70400,5.2811875,1.1442655,,,,,,,,,,,,,, +70500,4.997383,1.1571869,,,,,,,,,,,,,, +70600,5.00278,1.129809,,,,,,,,,,,,,, +70700,5.0608587,1.1777391,,,,,,,,,,,,,, +70800,4.895029,1.094354,,,,,,,,,,,,,, +70900,4.892038,1.1062815,,,,,,,,,,,,,, +71000,4.9099116,1.0888063,,,,,,,,,,,,,, +71100,4.90444,1.0683061,,,,,,,,,,,,,, +71200,5.0220814,1.1104038,,,,,,,,,,,,,, +71300,4.9022326,1.0629605,,,,,,,,,,,,,, +71400,4.8655915,1.0599872,,,,,,,,,,,,,, +71500,4.8503723,1.0402113,,,,,,,,,,,,,, +71600,5.1391363,1.1360918,,,,,,,,,,,,,, +71700,5.084836,1.119205,,,,,,,,,,,,,, +71800,5.1427546,1.0479193,,,,,,,,,,,,,, +71900,5.1192737,1.0521946,,,,,,,,,,,,,, +72000,5.2237387,1.1881578,,,,,,,,,,,,,, +72100,4.8104186,0.9835866,,,,,,,,,,,,,, +72200,4.9167366,1.0845034,,,,,,,,,,,,,, +72300,4.7851768,1.0774384,,,,,,,,,,,,,, +72400,5.1861305,1.0957171,,,,,,,,,,,,,, +72500,5.191675,1.2011657,,,,,,,,,,,,,, +72561,,,0.90824294,0.32824045,0.74074,1.0726177,50000.0,0.6108,1.8346378,10000.0,20013.38549757004,20291.634605646133,20013.38549757004,276.8868010044098,0.7893733978271484,0.0 +72600,5.258336,1.1007556,,,,,,,,,,,,,, +72700,5.0652895,1.0494859,,,,,,,,,,,,,, +72800,5.3739157,1.0879582,,,,,,,,,,,,,, +72900,4.803307,1.0320947,,,,,,,,,,,,,, +73000,5.0737,1.1762636,,,,,,,,,,,,,, +73100,5.1512933,1.1518657,,,,,,,,,,,,,, +73200,4.900606,1.0374982,,,,,,,,,,,,,, +73300,5.0069017,1.0804923,,,,,,,,,,,,,, +73400,5.1760178,1.1005182,,,,,,,,,,,,,, +73500,4.8772206,1.0573528,,,,,,,,,,,,,, +73600,5.2385345,1.1672783,,,,,,,,,,,,,, +73700,5.4541426,1.1482525,,,,,,,,,,,,,, +73800,4.8899713,1.0302114,,,,,,,,,,,,,, +73900,5.2844167,1.1735618,,,,,,,,,,,,,, +74000,4.7198114,0.9710991,,,,,,,,,,,,,, +74100,5.173821,1.126299,,,,,,,,,,,,,, +74200,5.0290017,1.0381804,,,,,,,,,,,,,, +74300,4.8899302,1.1397886,,,,,,,,,,,,,, +74400,5.2251163,1.0749009,,,,,,,,,,,,,, +74500,5.0995727,1.0224376,,,,,,,,,,,,,, +74600,5.0235305,1.0671968,,,,,,,,,,,,,, +74700,5.1563416,1.127058,,,,,,,,,,,,,, +74800,5.0603385,1.1326286,,,,,,,,,,,,,, +74900,5.07063,1.0973759,,,,,,,,,,,,,, +75000,4.683627,0.99576885,,,,,,,,,,,,,, +75100,5.0944557,1.1114229,,,,,,,,,,,,,, +75200,4.899845,1.0775337,,,,,,,,,,,,,, +75300,5.0012064,1.1131666,,,,,,,,,,,,,, +75400,5.3528857,1.0816067,,,,,,,,,,,,,, +75500,5.195472,1.1168345,,,,,,,,,,,,,, +75600,5.0340986,1.0960524,,,,,,,,,,,,,, +75700,5.151156,1.1211767,,,,,,,,,,,,,, +75800,5.0416307,1.1007633,,,,,,,,,,,,,, +75900,5.4216394,1.099067,,,,,,,,,,,,,, +76000,5.1055465,1.1071829,,,,,,,,,,,,,, +76100,5.046035,1.131532,,,,,,,,,,,,,, +76200,5.1604214,1.1579454,,,,,,,,,,,,,, +76300,5.071198,1.0973086,,,,,,,,,,,,,, +76400,5.1009293,1.0875586,,,,,,,,,,,,,, +76500,4.8908205,1.0608569,,,,,,,,,,,,,, +76600,4.9293113,1.1246624,,,,,,,,,,,,,, +76700,4.898761,1.0388465,,,,,,,,,,,,,, +76800,5.1338625,1.1213703,,,,,,,,,,,,,, +76900,5.2416825,1.0789068,,,,,,,,,,,,,, +77000,4.978885,1.0525554,,,,,,,,,,,,,, +77100,5.049512,1.1005328,,,,,,,,,,,,,, +77200,4.9482617,1.0380824,,,,,,,,,,,,,, +77300,4.8550277,1.0551497,,,,,,,,,,,,,, +77400,5.11069,1.0439122,,,,,,,,,,,,,, +77500,5.245868,1.1588771,,,,,,,,,,,,,, +77600,5.253383,1.152617,,,,,,,,,,,,,, +77700,5.149882,1.1658428,,,,,,,,,,,,,, +77800,5.53518,1.1289995,,,,,,,,,,,,,, +77900,4.7782845,1.0319041,,,,,,,,,,,,,, +78000,5.1952286,1.0831815,,,,,,,,,,,,,, +78100,5.2418427,1.1263652,,,,,,,,,,,,,, +78200,5.0983324,1.1317368,,,,,,,,,,,,,, +78300,4.9898553,1.0627614,,,,,,,,,,,,,, +78400,5.1592226,1.1542898,,,,,,,,,,,,,, +78500,5.45102,1.1827098,,,,,,,,,,,,,, +78600,4.7515507,1.0387801,,,,,,,,,,,,,, +78700,5.461549,1.1650281,,,,,,,,,,,,,, +78800,5.046883,1.0634094,,,,,,,,,,,,,, +78900,5.1359444,1.1858069,,,,,,,,,,,,,, +79000,5.1330605,1.1476011,,,,,,,,,,,,,, +79100,5.210806,1.0379138,,,,,,,,,,,,,, +79200,5.2280936,1.0903552,,,,,,,,,,,,,, +79300,5.2425747,1.0695446,,,,,,,,,,,,,, +79400,4.6560235,0.9886162,,,,,,,,,,,,,, +79500,5.136066,1.1352735,,,,,,,,,,,,,, +79600,5.278378,1.1470916,,,,,,,,,,,,,, +79700,4.837241,1.0685762,,,,,,,,,,,,,, +79800,5.0131407,1.0356762,,,,,,,,,,,,,, +79900,5.1560364,1.0813891,,,,,,,,,,,,,, +79915,,,0.9136041,0.3116816,0.74292,1.0638095,50000.0,0.61490005,1.8260653,10000.0,22009.43017864228,22306.47315478325,22009.43017864228,295.5284707546234,0.8865344524383545,0.0 +80000,4.7440934,0.93678975,,,,,,,,,,,,,, +80100,5.22695,1.0818908,,,,,,,,,,,,,, +80200,5.130754,1.0885625,,,,,,,,,,,,,, +80300,5.0934825,1.0991147,,,,,,,,,,,,,, +80400,5.289252,1.1376756,,,,,,,,,,,,,, +80500,5.154628,1.0619131,,,,,,,,,,,,,, +80600,5.1983066,1.1179153,,,,,,,,,,,,,, +80700,5.296384,1.109895,,,,,,,,,,,,,, +80800,5.3396597,1.1425439,,,,,,,,,,,,,, +80900,5.1314273,1.0713518,,,,,,,,,,,,,, +81000,5.270387,1.1071651,,,,,,,,,,,,,, +81100,4.7585826,1.0342025,,,,,,,,,,,,,, +81200,5.3766675,1.1756318,,,,,,,,,,,,,, +81300,4.837803,1.0759118,,,,,,,,,,,,,, +81400,5.1637363,1.1063645,,,,,,,,,,,,,, +81500,5.1806397,1.1112509,,,,,,,,,,,,,, +81600,5.208389,1.1078935,,,,,,,,,,,,,, +81700,5.0255475,1.0409093,,,,,,,,,,,,,, +81800,5.0255723,1.0438917,,,,,,,,,,,,,, +81900,5.1054144,1.1133626,,,,,,,,,,,,,, +82000,4.910178,0.953773,,,,,,,,,,,,,, +82100,4.8155484,1.0600984,,,,,,,,,,,,,, +82200,5.2933755,1.1484216,,,,,,,,,,,,,, +82300,5.317773,1.1451888,,,,,,,,,,,,,, +82400,5.1617775,1.0146054,,,,,,,,,,,,,, +82500,5.2499847,1.1282032,,,,,,,,,,,,,, +82600,5.494203,1.1245166,,,,,,,,,,,,,, +82700,4.862037,1.0237563,,,,,,,,,,,,,, +82800,5.129316,1.066969,,,,,,,,,,,,,, +82900,5.1616926,1.1265204,,,,,,,,,,,,,, +83000,4.93027,1.028163,,,,,,,,,,,,,, +83100,5.094122,1.0673697,,,,,,,,,,,,,, +83200,5.0204782,1.1077751,,,,,,,,,,,,,, +83300,5.2739244,1.126166,,,,,,,,,,,,,, +83400,4.917331,1.1006786,,,,,,,,,,,,,, +83500,4.8003197,1.0908208,,,,,,,,,,,,,, +83600,5.2559657,1.1403259,,,,,,,,,,,,,, +83700,4.9090343,1.0991261,,,,,,,,,,,,,, +83800,5.0416512,1.095657,,,,,,,,,,,,,, +83900,4.911158,1.0243654,,,,,,,,,,,,,, +84000,5.0859456,1.1317463,,,,,,,,,,,,,, +84100,4.9579935,1.0104325,,,,,,,,,,,,,, +84200,5.3667154,1.0703992,,,,,,,,,,,,,, +84300,5.2011833,1.0934944,,,,,,,,,,,,,, +84400,5.12627,1.1197023,,,,,,,,,,,,,, +84500,5.1120615,1.103268,,,,,,,,,,,,,, +84600,5.010058,1.0136958,,,,,,,,,,,,,, +84700,5.1651187,1.055744,,,,,,,,,,,,,, +84800,5.1577106,1.038613,,,,,,,,,,,,,, +84900,4.876096,1.1368645,,,,,,,,,,,,,, +85000,5.1595664,1.0806019,,,,,,,,,,,,,, +85100,5.0640054,1.0726846,,,,,,,,,,,,,, +85200,4.852942,1.0412897,,,,,,,,,,,,,, +85300,4.9304585,0.99379075,,,,,,,,,,,,,, +85400,5.173736,1.127548,,,,,,,,,,,,,, +85500,5.2388883,1.088276,,,,,,,,,,,,,, +85600,5.3128777,1.1555178,,,,,,,,,,,,,, +85700,4.9872317,1.1068981,,,,,,,,,,,,,, +85800,5.158889,1.0830414,,,,,,,,,,,,,, +85900,5.0162764,1.0491431,,,,,,,,,,,,,, +86000,4.799073,1.0986054,,,,,,,,,,,,,, +86100,4.9541516,1.0928743,,,,,,,,,,,,,, +86200,5.214902,1.0880508,,,,,,,,,,,,,, +86300,5.2477617,1.0971215,,,,,,,,,,,,,, +86400,5.1699476,1.1557715,,,,,,,,,,,,,, +86500,5.33322,1.1058385,,,,,,,,,,,,,, +86600,5.159791,1.0301131,,,,,,,,,,,,,, +86700,5.1269426,1.1231868,,,,,,,,,,,,,, +86800,5.1101766,1.0677825,,,,,,,,,,,,,, +86900,5.4838333,1.1991355,,,,,,,,,,,,,, +87000,4.789301,0.92575836,,,,,,,,,,,,,, +87100,5.1198225,1.1940184,,,,,,,,,,,,,, +87200,5.058359,1.0590508,,,,,,,,,,,,,, +87269,,,0.9151985,0.29835552,0.74513996,1.0610034,50000.0,0.6208,1.8269247,10000.0,24005.31016421318,24320.700825214382,24005.31016421318,313.69797587394714,0.9892408847808838,0.0 +87300,4.862928,1.1477861,,,,,,,,,,,,,, +87400,5.1098256,1.054623,,,,,,,,,,,,,, +87500,5.0553155,1.0578077,,,,,,,,,,,,,, +87600,5.1514597,1.0553305,,,,,,,,,,,,,, +87700,5.165578,0.99610174,,,,,,,,,,,,,, +87800,5.2004113,1.1454986,,,,,,,,,,,,,, +87900,4.9592147,1.0364349,,,,,,,,,,,,,, +88000,5.0385513,1.0307946,,,,,,,,,,,,,, +88100,5.1889114,1.0788046,,,,,,,,,,,,,, +88200,5.192467,1.0401576,,,,,,,,,,,,,, +88300,4.826591,1.0877863,,,,,,,,,,,,,, +88400,4.988775,1.0236487,,,,,,,,,,,,,, +88500,4.84382,1.0654933,,,,,,,,,,,,,, +88600,5.049118,1.1243466,,,,,,,,,,,,,, +88700,4.9958515,1.0561825,,,,,,,,,,,,,, +88800,5.0792933,1.1354634,,,,,,,,,,,,,, +88900,4.8743825,1.0750494,,,,,,,,,,,,,, +89000,5.3274713,1.0396729,,,,,,,,,,,,,, +89100,5.02671,1.1076357,,,,,,,,,,,,,, +89200,5.340488,1.1309959,,,,,,,,,,,,,, +89300,4.976593,1.0540323,,,,,,,,,,,,,, +89400,5.0271416,0.9904266,,,,,,,,,,,,,, +89500,5.018741,1.0614214,,,,,,,,,,,,,, +89600,5.032221,1.0859385,,,,,,,,,,,,,, +89700,5.29699,1.1314268,,,,,,,,,,,,,, +89800,4.9846234,1.0617743,,,,,,,,,,,,,, +89900,5.439864,1.1386751,,,,,,,,,,,,,, +90000,5.298143,1.059113,,,,,,,,,,,,,, +90100,5.362574,1.1434476,,,,,,,,,,,,,, +90200,4.9388375,1.004425,,,,,,,,,,,,,, +90300,4.978389,1.0280626,,,,,,,,,,,,,, +90400,4.8184676,1.0145558,,,,,,,,,,,,,, +90500,4.9261484,1.0545144,,,,,,,,,,,,,, +90600,4.8675275,1.0103247,,,,,,,,,,,,,, +90700,4.8688664,1.0130285,,,,,,,,,,,,,, +90800,5.3724403,1.1501596,,,,,,,,,,,,,, +90900,4.8598576,0.9395515,,,,,,,,,,,,,, +91000,4.980844,1.0504813,,,,,,,,,,,,,, +91100,4.975203,1.0149034,,,,,,,,,,,,,, +91200,5.175694,1.0313646,,,,,,,,,,,,,, +91300,5.2188654,1.1493603,,,,,,,,,,,,,, +91400,5.179587,1.015121,,,,,,,,,,,,,, +91500,5.2081127,1.016558,,,,,,,,,,,,,, +91600,5.002887,1.0540466,,,,,,,,,,,,,, +91700,4.868175,1.0223155,,,,,,,,,,,,,, +91800,5.098212,1.1375864,,,,,,,,,,,,,, +91900,5.0538087,1.1145582,,,,,,,,,,,,,, +92000,4.5978723,0.9242792,,,,,,,,,,,,,, +92100,5.1257715,1.0540128,,,,,,,,,,,,,, +92200,5.078255,1.045655,,,,,,,,,,,,,, +92300,4.9044504,1.0198277,,,,,,,,,,,,,, +92400,5.063964,1.0782282,,,,,,,,,,,,,, +92500,4.838296,0.9733791,,,,,,,,,,,,,, +92600,5.093166,1.0777786,,,,,,,,,,,,,, +92700,4.637843,0.94434977,,,,,,,,,,,,,, +92800,4.997883,1.1145238,,,,,,,,,,,,,, +92900,5.2261677,1.1063461,,,,,,,,,,,,,, +93000,4.7695765,1.0636566,,,,,,,,,,,,,, +93100,4.958585,1.0632446,,,,,,,,,,,,,, +93200,4.8113713,1.010061,,,,,,,,,,,,,, +93300,5.170283,1.1700065,,,,,,,,,,,,,, +93400,4.8548226,1.0168759,,,,,,,,,,,,,, +93500,5.015551,1.0143862,,,,,,,,,,,,,, +93600,5.473679,1.0423561,,,,,,,,,,,,,, +93700,4.961535,1.0193022,,,,,,,,,,,,,, +93800,4.8922453,1.05266,,,,,,,,,,,,,, +93900,5.5182257,1.1046516,,,,,,,,,,,,,, +94000,4.83738,1.0223505,,,,,,,,,,,,,, +94100,4.932234,0.95822966,,,,,,,,,,,,,, +94200,5.0288033,0.95833933,,,,,,,,,,,,,, +94300,4.760734,0.99722826,,,,,,,,,,,,,, +94400,4.854794,0.97203505,,,,,,,,,,,,,, +94500,4.7557325,0.94368273,,,,,,,,,,,,,, +94600,4.9997635,1.1419841,,,,,,,,,,,,,, +94622,,,0.92352915,0.27080315,0.74688,1.0591964,50000.0,0.62310004,1.8336954,10000.0,26001.21587538719,26335.716311454773,26001.21587538719,332.67340207099915,1.0685644149780271,0.0 +94700,4.9691696,1.0850737,,,,,,,,,,,,,, +94800,4.7277737,1.0311708,,,,,,,,,,,,,, +94900,4.923572,1.0282841,,,,,,,,,,,,,, +95000,5.050494,1.0497317,,,,,,,,,,,,,, +95100,4.819559,1.0059688,,,,,,,,,,,,,, +95200,4.90468,1.0477782,,,,,,,,,,,,,, +95300,5.0531363,1.0719063,,,,,,,,,,,,,, +95400,4.9477987,1.0192688,,,,,,,,,,,,,, +95500,4.88097,1.0594423,,,,,,,,,,,,,, +95600,4.917268,1.101958,,,,,,,,,,,,,, +95700,4.710489,0.99272484,,,,,,,,,,,,,, +95800,4.8233223,1.0685754,,,,,,,,,,,,,, +95900,5.090761,1.0772393,,,,,,,,,,,,,, +96000,4.746955,1.0085937,,,,,,,,,,,,,, +96100,4.6274376,1.0695044,,,,,,,,,,,,,, +96200,4.8477855,1.0295919,,,,,,,,,,,,,, +96300,4.771572,0.98099893,,,,,,,,,,,,,, +96400,4.886624,1.0275398,,,,,,,,,,,,,, +96500,5.03067,1.0479249,,,,,,,,,,,,,, +96600,5.052288,1.0560776,,,,,,,,,,,,,, +96700,5.1236095,1.0479873,,,,,,,,,,,,,, +96800,5.0657268,1.1321898,,,,,,,,,,,,,, +96900,4.683469,0.95377314,,,,,,,,,,,,,, +97000,4.904041,1.0468429,,,,,,,,,,,,,, +97100,4.7190385,1.0360827,,,,,,,,,,,,,, +97200,5.009496,1.0432553,,,,,,,,,,,,,, +97300,4.8171864,0.9815118,,,,,,,,,,,,,, +97400,5.0458364,1.1241887,,,,,,,,,,,,,, +97500,5.092387,1.1006452,,,,,,,,,,,,,, +97600,5.0740333,1.050188,,,,,,,,,,,,,, +97700,5.3151093,1.0676798,,,,,,,,,,,,,, +97800,4.849457,1.044003,,,,,,,,,,,,,, +97900,5.0650854,1.1033505,,,,,,,,,,,,,, +98000,4.7362,0.98365736,,,,,,,,,,,,,, +98100,5.020538,0.99303716,,,,,,,,,,,,,, +98200,5.194108,1.042319,,,,,,,,,,,,,, +98300,5.0406446,1.0631154,,,,,,,,,,,,,, +98400,4.684608,0.9492117,,,,,,,,,,,,,, +98500,4.9476743,1.0623978,,,,,,,,,,,,,, +98600,5.1444364,1.0962319,,,,,,,,,,,,,, +98700,4.6965632,0.98370206,,,,,,,,,,,,,, +98800,4.9901056,1.1075429,,,,,,,,,,,,,, +98900,4.737512,0.9301794,,,,,,,,,,,,,, +99000,5.16146,1.1184072,,,,,,,,,,,,,, +99100,5.05285,1.0586447,,,,,,,,,,,,,, +99200,4.9914284,0.99453163,,,,,,,,,,,,,, +99300,4.913522,1.0111423,,,,,,,,,,,,,, +99400,4.920143,0.9721001,,,,,,,,,,,,,, +99500,5.242607,1.0570776,,,,,,,,,,,,,, +99600,5.134933,1.0076547,,,,,,,,,,,,,, +99700,4.9838767,1.0705804,,,,,,,,,,,,,, +99800,4.4909244,0.9185853,,,,,,,,,,,,,, +99900,4.9475822,1.0613029,,,,,,,,,,,,,, +100000,5.1796713,1.046685,,,,,,,,,,,,,, +100100,5.167573,1.0297633,,,,,,,,,,,,,, +100200,4.947313,1.0366803,,,,,,,,,,,,,, +100300,5.108886,1.1061466,,,,,,,,,,,,,, +100400,5.1959453,1.1637647,,,,,,,,,,,,,, +100500,5.007674,1.066778,,,,,,,,,,,,,, +100600,4.8467016,1.0005419,,,,,,,,,,,,,, +100700,4.866467,1.0246458,,,,,,,,,,,,,, +100800,4.8468046,0.9701067,,,,,,,,,,,,,, +100900,4.77389,1.0356143,,,,,,,,,,,,,, +101000,5.1323776,1.0424933,,,,,,,,,,,,,, +101100,5.1154957,1.0250137,,,,,,,,,,,,,, +101200,4.720861,0.98213506,,,,,,,,,,,,,, +101300,5.073286,1.1464223,,,,,,,,,,,,,, +101400,4.4470882,0.9089693,,,,,,,,,,,,,, +101500,5.157323,0.9695748,,,,,,,,,,,,,, +101600,4.66383,0.95951235,,,,,,,,,,,,,, +101700,5.1077166,1.0426953,,,,,,,,,,,,,, +101800,4.8859043,1.03865,,,,,,,,,,,,,, +101900,4.858632,0.9904083,,,,,,,,,,,,,, +101975,,,0.92490435,0.26681402,0.74736,1.0602218,50000.0,0.62320006,1.8342712,10000.0,27997.165731191635,28350.35700964928,27997.165731191635,351.16429591178894,1.2089686393737793,0.0 +102000,4.645664,0.9264372,,,,,,,,,,,,,, +102100,4.621422,1.036573,,,,,,,,,,,,,, +102200,5.0070853,0.9960358,,,,,,,,,,,,,, +102300,4.6336083,0.9069032,,,,,,,,,,,,,, +102400,4.990733,1.0679249,,,,,,,,,,,,,, +102500,5.2709804,1.088825,,,,,,,,,,,,,, +102600,4.9403863,0.9794909,,,,,,,,,,,,,, +102700,4.885138,0.9967936,,,,,,,,,,,,,, +102800,5.2392893,1.109434,,,,,,,,,,,,,, +102900,5.30014,1.0616695,,,,,,,,,,,,,, +103000,4.9516687,0.9739927,,,,,,,,,,,,,, +103100,5.196241,1.0499144,,,,,,,,,,,,,, +103200,4.5120735,0.94372284,,,,,,,,,,,,,, +103300,4.836172,1.1054214,,,,,,,,,,,,,, +103400,4.7342057,0.92712086,,,,,,,,,,,,,, +103500,5.192615,1.0347824,,,,,,,,,,,,,, +103600,4.911137,1.0092695,,,,,,,,,,,,,, +103700,4.8963842,0.97579145,,,,,,,,,,,,,, +103800,4.9961576,0.9974159,,,,,,,,,,,,,, +103900,4.7413955,1.0038204,,,,,,,,,,,,,, +104000,5.213546,1.0410833,,,,,,,,,,,,,, +104100,5.3066316,1.1378026,,,,,,,,,,,,,, +104200,5.1526685,1.0453546,,,,,,,,,,,,,, +104300,5.3145833,1.0001748,,,,,,,,,,,,,, +104400,5.121183,1.0586731,,,,,,,,,,,,,, +104500,4.836161,0.9595436,,,,,,,,,,,,,, +104600,5.1765924,1.1110246,,,,,,,,,,,,,, +104700,4.946266,0.98004025,,,,,,,,,,,,,, +104800,5.3294272,1.0546253,,,,,,,,,,,,,, +104900,5.0048714,0.96395385,,,,,,,,,,,,,, +105000,4.7124143,0.98645484,,,,,,,,,,,,,, +105100,4.9421115,1.0402874,,,,,,,,,,,,,, +105200,4.787493,1.0334847,,,,,,,,,,,,,, +105300,5.0030394,1.0254104,,,,,,,,,,,,,, +105400,5.2040896,1.1215822,,,,,,,,,,,,,, +105500,4.9027815,1.0522324,,,,,,,,,,,,,, +105600,4.746074,0.9647051,,,,,,,,,,,,,, +105700,4.6597223,1.0360446,,,,,,,,,,,,,, +105800,5.0231895,1.0234199,,,,,,,,,,,,,, +105900,5.107797,1.0907558,,,,,,,,,,,,,, +106000,5.2650723,1.0823858,,,,,,,,,,,,,, +106100,4.7014832,0.96844673,,,,,,,,,,,,,, +106200,4.8782535,1.0006012,,,,,,,,,,,,,, +106300,4.766509,0.99414396,,,,,,,,,,,,,, +106400,4.9070287,1.0326787,,,,,,,,,,,,,, +106500,4.875236,1.0046732,,,,,,,,,,,,,, +106600,5.2157555,1.0618668,,,,,,,,,,,,,, +106700,4.7685456,1.0029347,,,,,,,,,,,,,, +106800,4.507076,0.87173724,,,,,,,,,,,,,, +106900,4.9515243,1.0194901,,,,,,,,,,,,,, +107000,4.987048,1.0884076,,,,,,,,,,,,,, +107100,4.9011645,0.97172713,,,,,,,,,,,,,, +107200,4.888549,1.0147302,,,,,,,,,,,,,, +107300,4.9862623,1.0300823,,,,,,,,,,,,,, +107400,5.1018615,1.0646132,,,,,,,,,,,,,, +107500,4.7701735,0.9912981,,,,,,,,,,,,,, +107600,5.021111,1.079919,,,,,,,,,,,,,, +107700,5.038873,0.9219493,,,,,,,,,,,,,, +107800,4.963327,1.0245019,,,,,,,,,,,,,, +107900,5.3294125,1.0346425,,,,,,,,,,,,,, +108000,5.0048985,1.0173415,,,,,,,,,,,,,, +108100,4.9960732,1.0307862,,,,,,,,,,,,,, +108200,5.0001545,0.9875686,,,,,,,,,,,,,, +108300,4.5788946,0.9480152,,,,,,,,,,,,,, +108400,5.1168337,1.1127241,,,,,,,,,,,,,, +108500,4.9642553,1.0086931,,,,,,,,,,,,,, +108600,5.0293264,1.0711213,,,,,,,,,,,,,, +108700,4.65714,0.897216,,,,,,,,,,,,,, +108800,4.812422,0.93550104,,,,,,,,,,,,,, +108900,4.8442907,0.98310304,,,,,,,,,,,,,, +109000,5.013382,1.0371623,,,,,,,,,,,,,, +109100,5.1433935,1.0966556,,,,,,,,,,,,,, +109200,5.021624,0.99756014,,,,,,,,,,,,,, +109300,4.7771244,0.9853638,,,,,,,,,,,,,, +109331,,,0.9286312,0.24832252,0.7486,1.0602936,50000.0,0.6223,1.8452394,10000.0,29993.1374783516,30364.70661520958,29993.1374783516,369.42036747932434,1.274552583694458,0.0 +109400,4.96728,1.0043333,,,,,,,,,,,,,, +109500,4.9576254,1.013037,,,,,,,,,,,,,, +109600,5.1412787,0.99784017,,,,,,,,,,,,,, +109700,4.833191,1.0193137,,,,,,,,,,,,,, +109800,4.864041,0.9834493,,,,,,,,,,,,,, +109900,4.7909245,0.9928283,,,,,,,,,,,,,, +110000,4.717839,1.0193734,,,,,,,,,,,,,, +110100,4.740448,0.97521764,,,,,,,,,,,,,, +110200,4.813823,1.0471191,,,,,,,,,,,,,, +110300,5.021089,1.0582752,,,,,,,,,,,,,, +110400,4.842183,1.0451837,,,,,,,,,,,,,, +110500,4.8494954,0.9751569,,,,,,,,,,,,,, +110600,4.3335795,0.81750965,,,,,,,,,,,,,, +110700,4.839282,1.0116155,,,,,,,,,,,,,, +110800,4.907031,1.0448642,,,,,,,,,,,,,, +110900,5.065103,1.0579635,,,,,,,,,,,,,, +111000,5.0447054,0.9934646,,,,,,,,,,,,,, +111100,4.8179727,1.0597149,,,,,,,,,,,,,, +111200,4.9289317,0.9584872,,,,,,,,,,,,,, +111300,5.229253,1.0294745,,,,,,,,,,,,,, +111400,5.0176597,1.0524282,,,,,,,,,,,,,, +111500,4.8487687,0.9565358,,,,,,,,,,,,,, +111600,4.630929,1.0005667,,,,,,,,,,,,,, +111700,5.05268,1.1076342,,,,,,,,,,,,,, +111800,5.1228495,1.0270578,,,,,,,,,,,,,, +111900,4.810208,0.95798445,,,,,,,,,,,,,, +112000,4.718741,1.0581936,,,,,,,,,,,,,, +112100,4.772494,0.9249016,,,,,,,,,,,,,, +112200,4.7397985,1.0316896,,,,,,,,,,,,,, +112300,4.75501,0.9556118,,,,,,,,,,,,,, +112400,5.3783255,1.0141996,,,,,,,,,,,,,, +112500,4.8861213,0.9980276,,,,,,,,,,,,,, +112600,4.76714,0.95725334,,,,,,,,,,,,,, +112700,5.0381923,1.026045,,,,,,,,,,,,,, +112800,5.116514,1.0376782,,,,,,,,,,,,,, +112900,4.992646,1.0483359,,,,,,,,,,,,,, +113000,4.848923,1.0418324,,,,,,,,,,,,,, +113100,5.0808325,1.0326092,,,,,,,,,,,,,, +113200,5.1217103,1.0551822,,,,,,,,,,,,,, +113300,4.8500824,0.98522496,,,,,,,,,,,,,, +113400,5.154114,1.0434041,,,,,,,,,,,,,, +113500,4.929071,0.97339314,,,,,,,,,,,,,, +113600,4.786121,1.0046653,,,,,,,,,,,,,, +113700,5.234892,1.0781702,,,,,,,,,,,,,, +113800,4.7960463,1.0695703,,,,,,,,,,,,,, +113900,5.1159163,1.0030917,,,,,,,,,,,,,, +114000,4.933364,1.0189523,,,,,,,,,,,,,, +114100,4.623847,1.0023763,,,,,,,,,,,,,, +114200,4.90004,1.0126176,,,,,,,,,,,,,, +114300,5.068505,1.1597884,,,,,,,,,,,,,, +114400,4.5535517,0.93160933,,,,,,,,,,,,,, +114500,4.8633084,1.0404313,,,,,,,,,,,,,, +114600,5.022075,1.0157124,,,,,,,,,,,,,, +114700,5.1019998,1.0295761,,,,,,,,,,,,,, +114800,4.7895436,0.9683779,,,,,,,,,,,,,, +114900,4.8700833,0.9704714,,,,,,,,,,,,,, +115000,5.077556,1.050014,,,,,,,,,,,,,, +115100,4.8273654,0.9857801,,,,,,,,,,,,,, +115200,4.79947,0.9691099,,,,,,,,,,,,,, +115300,5.004415,1.0687587,,,,,,,,,,,,,, +115400,4.708355,0.99877894,,,,,,,,,,,,,, +115500,4.9552655,0.9781686,,,,,,,,,,,,,, +115600,4.895992,0.95186394,,,,,,,,,,,,,, +115700,5.1625514,0.996286,,,,,,,,,,,,,, +115800,4.937767,1.0064547,,,,,,,,,,,,,, +115900,4.8091683,1.0216733,,,,,,,,,,,,,, +116000,5.3392725,1.0284873,,,,,,,,,,,,,, +116100,5.0206466,1.0193744,,,,,,,,,,,,,, +116200,4.8958063,1.0607626,,,,,,,,,,,,,, +116300,4.915205,1.0027375,,,,,,,,,,,,,, +116400,4.7652287,1.0136722,,,,,,,,,,,,,, +116500,4.9740157,0.9715111,,,,,,,,,,,,,, +116600,4.9411163,0.97437906,,,,,,,,,,,,,, +116686,,,0.93215877,0.23511541,0.74862,1.062171,50000.0,0.62310004,1.8545954,10000.0,31989.12622189521,32378.873791217804,31989.12622189521,387.4781656265259,1.3403329849243164,0.0 +116700,4.850866,0.924694,,,,,,,,,,,,,, +116800,4.985006,0.96968275,,,,,,,,,,,,,, +116900,5.280363,1.0387361,,,,,,,,,,,,,, +117000,4.7266912,1.0884373,,,,,,,,,,,,,, +117100,4.657216,0.9450543,,,,,,,,,,,,,, +117200,4.849338,0.94724226,,,,,,,,,,,,,, +117300,4.7362237,0.98486173,,,,,,,,,,,,,, +117400,5.037086,1.0426064,,,,,,,,,,,,,, +117500,4.5763474,0.98888505,,,,,,,,,,,,,, +117600,5.159463,1.0366747,,,,,,,,,,,,,, +117700,4.8435826,1.0207684,,,,,,,,,,,,,, +117800,4.790757,1.0427068,,,,,,,,,,,,,, +117900,5.0247645,1.0566423,,,,,,,,,,,,,, +118000,5.0752654,1.0940913,,,,,,,,,,,,,, +118100,4.832977,0.8915735,,,,,,,,,,,,,, +118200,4.826442,0.9139772,,,,,,,,,,,,,, +118300,4.5998735,1.003947,,,,,,,,,,,,,, +118400,4.629039,0.9524575,,,,,,,,,,,,,, +118500,4.7909894,0.97438323,,,,,,,,,,,,,, +118600,4.7742586,1.0503862,,,,,,,,,,,,,, +118700,5.0690823,1.0420067,,,,,,,,,,,,,, +118800,5.2089214,1.0600954,,,,,,,,,,,,,, +118900,5.0889206,1.003732,,,,,,,,,,,,,, +119000,4.895788,1.0548812,,,,,,,,,,,,,, +119100,5.0212507,1.1313481,,,,,,,,,,,,,, +119200,5.0524583,1.0196854,,,,,,,,,,,,,, +119300,4.855984,1.032133,,,,,,,,,,,,,, +119400,4.8821883,0.9793161,,,,,,,,,,,,,, +119500,4.6826897,0.9560535,,,,,,,,,,,,,, +119600,5.3284545,1.133991,,,,,,,,,,,,,, +119700,4.931461,1.0062449,,,,,,,,,,,,,, +119800,4.8169236,0.979719,,,,,,,,,,,,,, +119900,5.1749744,1.0633664,,,,,,,,,,,,,, +120000,4.7674556,0.94644296,,,,,,,,,,,,,, +120100,4.982606,1.0156686,,,,,,,,,,,,,, +120200,4.888096,1.0156635,,,,,,,,,,,,,, +120300,4.875648,1.0243328,,,,,,,,,,,,,, +120400,5.199045,1.0804758,,,,,,,,,,,,,, +120500,5.0990853,1.05281,,,,,,,,,,,,,, +120600,5.0723586,1.1073802,,,,,,,,,,,,,, +120700,4.9479284,1.0029414,,,,,,,,,,,,,, +120800,5.0070677,1.0194285,,,,,,,,,,,,,, +120900,5.1572165,1.0549673,,,,,,,,,,,,,, +121000,4.975783,0.98916626,,,,,,,,,,,,,, +121100,4.987039,0.997976,,,,,,,,,,,,,, +121200,4.709308,0.95398337,,,,,,,,,,,,,, +121300,5.0469213,1.0125489,,,,,,,,,,,,,, +121400,4.8991303,0.97600126,,,,,,,,,,,,,, +121500,5.0322065,0.99742687,,,,,,,,,,,,,, +121600,5.0671477,0.9801316,,,,,,,,,,,,,, +121700,4.6738524,0.95026845,,,,,,,,,,,,,, +121800,4.7347016,0.9925257,,,,,,,,,,,,,, +121900,4.7652555,0.9965633,,,,,,,,,,,,,, +122000,5.245432,1.0742058,,,,,,,,,,,,,, +122100,5.1195865,0.991964,,,,,,,,,,,,,, +122200,5.021031,1.0741068,,,,,,,,,,,,,, +122300,4.888718,1.0632461,,,,,,,,,,,,,, +122400,5.155271,1.0432822,,,,,,,,,,,,,, +122500,4.75963,0.9929557,,,,,,,,,,,,,, +122600,4.835897,0.96253103,,,,,,,,,,,,,, +122700,4.894546,0.957472,,,,,,,,,,,,,, +122800,4.9423523,1.0315545,,,,,,,,,,,,,, +122900,4.8338995,0.9488271,,,,,,,,,,,,,, +123000,5.066993,1.0502018,,,,,,,,,,,,,, +123100,4.823648,1.0576289,,,,,,,,,,,,,, +123200,4.729123,0.9157568,,,,,,,,,,,,,, +123300,5.039592,1.0762786,,,,,,,,,,,,,, +123400,4.5189824,0.8206817,,,,,,,,,,,,,, +123500,4.942296,1.0297703,,,,,,,,,,,,,, +123600,4.629935,0.9277737,,,,,,,,,,,,,, +123700,4.9961586,1.0691513,,,,,,,,,,,,,, +123800,4.7485204,0.94534343,,,,,,,,,,,,,, +123900,4.9032364,0.9822508,,,,,,,,,,,,,, +124000,5.017119,1.0271432,,,,,,,,,,,,,, +124041,,,0.9368423,0.22361873,0.75012,1.0643718,50000.0,0.62140006,1.8547386,10000.0,33985.024151563644,34392.77033543587,33985.024151563644,405.35461807250977,1.407243728637695,0.0 +124100,4.9597397,0.9920551,,,,,,,,,,,,,, +124200,4.7248783,0.95136106,,,,,,,,,,,,,, +124300,5.189523,1.1107758,,,,,,,,,,,,,, +124400,4.622126,0.9888363,,,,,,,,,,,,,, +124500,5.255875,0.9745399,,,,,,,,,,,,,, +124600,4.630988,0.9354743,,,,,,,,,,,,,, +124700,5.1358166,0.9921715,,,,,,,,,,,,,, +124800,4.9833026,1.0238392,,,,,,,,,,,,,, +124900,4.6937723,0.95476276,,,,,,,,,,,,,, +125000,4.958782,1.0629839,,,,,,,,,,,,,, +125100,5.2824063,1.0854396,,,,,,,,,,,,,, +125200,4.6630993,0.92152685,,,,,,,,,,,,,, +125300,4.6496344,0.9554889,,,,,,,,,,,,,, +125400,5.2143497,0.9384632,,,,,,,,,,,,,, +125500,4.932324,1.0501761,,,,,,,,,,,,,, +125600,4.680226,1.0079079,,,,,,,,,,,,,, +125700,4.7950587,0.9400353,,,,,,,,,,,,,, +125800,4.9333425,1.0056416,,,,,,,,,,,,,, +125900,4.708019,0.93551457,,,,,,,,,,,,,, +126000,4.999929,0.9737023,,,,,,,,,,,,,, +126100,4.926108,0.9490304,,,,,,,,,,,,,, +126200,4.8146534,0.9340658,,,,,,,,,,,,,, +126300,4.775745,0.96694434,,,,,,,,,,,,,, +126400,4.821079,0.9750714,,,,,,,,,,,,,, +126500,5.4730406,1.1141996,,,,,,,,,,,,,, +126600,5.07151,0.99188095,,,,,,,,,,,,,, +126700,4.7341676,0.9224956,,,,,,,,,,,,,, +126800,5.002106,1.0041763,,,,,,,,,,,,,, +126900,4.5831475,0.9289351,,,,,,,,,,,,,, +127000,4.907226,1.0300579,,,,,,,,,,,,,, +127100,5.2305956,1.0330253,,,,,,,,,,,,,, +127200,5.226077,1.0034273,,,,,,,,,,,,,, +127300,5.1292744,0.96000695,,,,,,,,,,,,,, +127400,4.566081,0.90412414,,,,,,,,,,,,,, +127500,4.8500886,0.94777226,,,,,,,,,,,,,, +127600,5.0995226,1.0162574,,,,,,,,,,,,,, +127700,5.2207727,1.0599383,,,,,,,,,,,,,, +127800,5.225055,1.0208986,,,,,,,,,,,,,, +127900,4.915736,0.9938912,,,,,,,,,,,,,, +128000,5.0379596,1.0021892,,,,,,,,,,,,,, +128100,4.761623,0.91196036,,,,,,,,,,,,,, +128200,5.160921,0.9853943,,,,,,,,,,,,,, +128300,4.720266,0.98202884,,,,,,,,,,,,,, +128400,4.8942814,0.92959917,,,,,,,,,,,,,, +128500,4.8365145,0.9874048,,,,,,,,,,,,,, +128600,4.7843843,0.9857383,,,,,,,,,,,,,, +128700,4.663295,0.95996463,,,,,,,,,,,,,, +128800,4.876372,1.0406973,,,,,,,,,,,,,, +128900,4.8326845,0.9893048,,,,,,,,,,,,,, +129000,5.0877404,0.9998128,,,,,,,,,,,,,, +129100,4.804135,0.97647995,,,,,,,,,,,,,, +129200,4.7708507,0.9617882,,,,,,,,,,,,,, +129300,5.032071,0.98279595,,,,,,,,,,,,,, +129400,5.333802,1.0095701,,,,,,,,,,,,,, +129500,4.8801146,1.0176944,,,,,,,,,,,,,, +129600,5.064136,1.061135,,,,,,,,,,,,,, +129700,5.0579967,1.0527387,,,,,,,,,,,,,, +129800,5.0375795,1.0226257,,,,,,,,,,,,,, +129900,4.582227,0.92814493,,,,,,,,,,,,,, +130000,4.9030395,0.9968441,,,,,,,,,,,,,, +130100,4.816661,1.0385652,,,,,,,,,,,,,, +130200,5.0727177,1.0414743,,,,,,,,,,,,,, +130300,5.1611953,1.058078,,,,,,,,,,,,,, +130400,4.9665074,1.0647693,,,,,,,,,,,,,, +130500,4.9490123,0.9918737,,,,,,,,,,,,,, +130600,4.729351,0.98637354,,,,,,,,,,,,,, +130700,4.9164696,0.90843236,,,,,,,,,,,,,, +130800,4.961255,0.98212695,,,,,,,,,,,,,, +130900,4.8681164,0.86855245,,,,,,,,,,,,,, +131000,4.792482,0.98988056,,,,,,,,,,,,,, +131100,5.009235,0.96191347,,,,,,,,,,,,,, +131200,4.6443815,0.9350121,,,,,,,,,,,,,, +131300,4.69996,0.8987126,,,,,,,,,,,,,, +131395,,,0.93939334,0.21290366,0.75088,1.0653294,50000.0,0.62280005,1.852586,10000.0,35980.980189561844,36406.970715522766,35980.980189561844,423.4676520824432,1.4830548763275146,0.0 +131400,5.1747794,1.0998032,,,,,,,,,,,,,, +131500,4.8814726,0.9606101,,,,,,,,,,,,,, +131600,5.015367,0.94809824,,,,,,,,,,,,,, +131700,4.75671,0.9282273,,,,,,,,,,,,,, +131800,4.5502534,0.90294516,,,,,,,,,,,,,, +131900,4.6937284,0.90611315,,,,,,,,,,,,,, +132000,5.185933,1.1230513,,,,,,,,,,,,,, +132100,4.817813,0.9043565,,,,,,,,,,,,,, +132200,4.9089212,0.9540932,,,,,,,,,,,,,, +132300,4.884382,0.971203,,,,,,,,,,,,,, +132400,4.7487,0.9776037,,,,,,,,,,,,,, +132500,4.9269204,0.9731859,,,,,,,,,,,,,, +132600,4.861461,1.0152194,,,,,,,,,,,,,, +132700,5.075225,1.0098729,,,,,,,,,,,,,, +132800,4.839644,0.9747336,,,,,,,,,,,,,, +132900,4.9513845,0.9948178,,,,,,,,,,,,,, +133000,5.130737,1.0611969,,,,,,,,,,,,,, +133100,4.9789743,1.0231566,,,,,,,,,,,,,, +133200,5.06498,1.0556695,,,,,,,,,,,,,, +133300,4.9380293,0.9623418,,,,,,,,,,,,,, +133400,4.9530964,0.98247886,,,,,,,,,,,,,, +133500,4.558867,0.90809983,,,,,,,,,,,,,, +133600,4.6267333,0.9675575,,,,,,,,,,,,,, +133700,4.6693354,0.9381409,,,,,,,,,,,,,, +133800,5.0144677,0.98396873,,,,,,,,,,,,,, +133900,5.3446407,0.9836799,,,,,,,,,,,,,, +134000,4.737255,0.9720773,,,,,,,,,,,,,, +134100,4.7850847,0.99708056,,,,,,,,,,,,,, +134200,4.786213,1.0162544,,,,,,,,,,,,,, +134300,4.8675485,1.0083871,,,,,,,,,,,,,, +134400,4.8439765,0.98041755,,,,,,,,,,,,,, +134500,5.219333,1.008295,,,,,,,,,,,,,, +134600,4.9980903,0.97670937,,,,,,,,,,,,,, +134700,4.984532,1.0002126,,,,,,,,,,,,,, +134800,5.0577745,0.97026765,,,,,,,,,,,,,, +134900,4.8183475,0.9613228,,,,,,,,,,,,,, +135000,4.8867836,0.89829206,,,,,,,,,,,,,, +135100,4.7825046,1.0202123,,,,,,,,,,,,,, +135200,5.052027,1.0283599,,,,,,,,,,,,,, +135300,4.920061,0.9737001,,,,,,,,,,,,,, +135400,5.008422,0.968822,,,,,,,,,,,,,, +135500,4.7384715,0.8799198,,,,,,,,,,,,,, +135600,4.7763286,0.96937025,,,,,,,,,,,,,, +135700,5.0017643,0.9568836,,,,,,,,,,,,,, +135800,4.951026,1.0468212,,,,,,,,,,,,,, +135900,4.9958844,0.98721325,,,,,,,,,,,,,, +136000,5.0418997,0.9912666,,,,,,,,,,,,,, +136100,4.6972113,1.0048337,,,,,,,,,,,,,, +136200,4.8468723,0.9596787,,,,,,,,,,,,,, +136300,4.814752,0.9751401,,,,,,,,,,,,,, +136400,4.8599243,0.96838444,,,,,,,,,,,,,, +136500,4.7757616,0.99881935,,,,,,,,,,,,,, +136600,4.6547413,0.9687707,,,,,,,,,,,,,, +136700,5.1533966,0.9524313,,,,,,,,,,,,,, +136800,4.529002,0.90112495,,,,,,,,,,,,,, +136900,4.8090205,0.9874518,,,,,,,,,,,,,, +137000,4.746774,0.97569835,,,,,,,,,,,,,, +137100,4.814104,0.9646883,,,,,,,,,,,,,, +137200,4.543583,0.84824437,,,,,,,,,,,,,, +137300,4.9231887,1.02197,,,,,,,,,,,,,, +137400,4.854905,0.98111457,,,,,,,,,,,,,, +137500,4.905315,0.99713075,,,,,,,,,,,,,, +137600,5.0919056,1.0305725,,,,,,,,,,,,,, +137700,4.949118,0.9926078,,,,,,,,,,,,,, +137800,5.16623,1.0907676,,,,,,,,,,,,,, +137900,4.6921473,0.9595529,,,,,,,,,,,,,, +138000,4.794547,0.94504553,,,,,,,,,,,,,, +138100,4.727849,0.9876723,,,,,,,,,,,,,, +138200,4.7012596,0.91477424,,,,,,,,,,,,,, +138300,5.00467,0.9913434,,,,,,,,,,,,,, +138400,4.869284,0.9866593,,,,,,,,,,,,,, +138500,4.787543,0.97756207,,,,,,,,,,,,,, +138600,4.576642,0.8719697,,,,,,,,,,,,,, +138700,4.7889895,0.9597725,,,,,,,,,,,,,, +138749,,,0.93939334,0.21290293,0.75162,1.0644616,50000.0,0.6233,1.8521956,10000.0,37976.9183151722,38421.72024726868,37976.9183151722,442.1525115966797,1.554464340209961,0.0 +138800,4.8717036,0.9902591,,,,,,,,,,,,,, +138900,4.8640165,0.9534199,,,,,,,,,,,,,, +139000,4.943868,1.013802,,,,,,,,,,,,,, +139100,4.882792,0.9475374,,,,,,,,,,,,,, +139200,5.117842,0.89593744,,,,,,,,,,,,,, +139300,4.8448205,0.9967928,,,,,,,,,,,,,, +139400,5.0009527,0.9310852,,,,,,,,,,,,,, +139500,5.344364,1.000068,,,,,,,,,,,,,, +139600,4.8953104,1.0049301,,,,,,,,,,,,,, +139700,4.812715,0.9203932,,,,,,,,,,,,,, +139800,4.8625474,0.96680605,,,,,,,,,,,,,, +139900,4.9193335,0.955411,,,,,,,,,,,,,, +140000,5.143237,0.97933555,,,,,,,,,,,,,, +140100,4.6883864,0.8688985,,,,,,,,,,,,,, +140200,4.8848243,0.9146054,,,,,,,,,,,,,, +140300,4.703448,0.9397494,,,,,,,,,,,,,, +140400,4.777933,0.95152897,,,,,,,,,,,,,, +140500,4.760321,0.88183403,,,,,,,,,,,,,, +140600,4.550999,0.89681685,,,,,,,,,,,,,, +140700,4.7918997,0.88125885,,,,,,,,,,,,,, +140800,5.1857934,1.0263419,,,,,,,,,,,,,, +140900,4.806961,0.986444,,,,,,,,,,,,,, +141000,4.9310784,1.022622,,,,,,,,,,,,,, +141100,4.7147126,0.8815237,,,,,,,,,,,,,, +141200,5.120867,1.0595367,,,,,,,,,,,,,, +141300,4.7986665,0.97729766,,,,,,,,,,,,,, +141400,5.009456,1.0006597,,,,,,,,,,,,,, +141500,4.8912573,0.92474955,,,,,,,,,,,,,, +141600,5.019778,1.0210985,,,,,,,,,,,,,, +141700,4.8890743,0.9506537,,,,,,,,,,,,,, +141800,4.7921286,0.90477765,,,,,,,,,,,,,, +141900,5.19428,1.0638078,,,,,,,,,,,,,, +142000,5.1026564,1.0972714,,,,,,,,,,,,,, +142100,4.7848997,0.8872174,,,,,,,,,,,,,, +142200,4.783771,1.0330267,,,,,,,,,,,,,, +142300,5.1482735,1.0364956,,,,,,,,,,,,,, +142400,4.9430556,0.94572777,,,,,,,,,,,,,, +142500,4.8587465,1.0003998,,,,,,,,,,,,,, +142600,4.9623494,1.0550195,,,,,,,,,,,,,, +142700,4.758121,0.9100339,,,,,,,,,,,,,, +142800,4.948895,1.0123061,,,,,,,,,,,,,, +142900,4.8999333,0.9546158,,,,,,,,,,,,,, +143000,4.8834243,0.9496175,,,,,,,,,,,,,, +143100,4.7923107,0.9300961,,,,,,,,,,,,,, +143200,4.9125767,0.9672505,,,,,,,,,,,,,, +143300,4.87308,0.90973693,,,,,,,,,,,,,, +143400,4.949477,0.97593343,,,,,,,,,,,,,, +143500,4.497638,0.9046101,,,,,,,,,,,,,, +143600,5.1304774,1.0155139,,,,,,,,,,,,,, +143700,5.00399,0.9180908,,,,,,,,,,,,,, +143800,4.9964314,0.9707123,,,,,,,,,,,,,, +143900,4.718487,0.992277,,,,,,,,,,,,,, +144000,5.0917745,0.9691386,,,,,,,,,,,,,, +144100,4.956174,1.018227,,,,,,,,,,,,,, +144200,5.032228,1.0013444,,,,,,,,,,,,,, +144300,4.616175,0.9184095,,,,,,,,,,,,,, +144400,4.902664,0.99306583,,,,,,,,,,,,,, +144500,4.8350472,0.91948694,,,,,,,,,,,,,, +144600,4.9404616,1.0046405,,,,,,,,,,,,,, +144700,4.536959,0.8735369,,,,,,,,,,,,,, +144800,5.001752,1.0408049,,,,,,,,,,,,,, +144900,4.9982414,0.9106251,,,,,,,,,,,,,, +145000,4.887908,0.9948021,,,,,,,,,,,,,, +145100,4.8067937,1.0001886,,,,,,,,,,,,,, +145200,5.2143536,1.0018325,,,,,,,,,,,,,, +145300,4.890803,0.9896284,,,,,,,,,,,,,, +145400,4.898375,1.0072888,,,,,,,,,,,,,, +145500,4.8903604,0.9560223,,,,,,,,,,,,,, +145600,4.8799415,1.0323952,,,,,,,,,,,,,, +145700,5.0522375,0.99542636,,,,,,,,,,,,,, +145800,5.049267,1.0674572,,,,,,,,,,,,,, +145900,4.948715,1.0172759,,,,,,,,,,,,,, +146000,5.1561975,1.0982478,,,,,,,,,,,,,, +146100,4.9815774,0.9525747,,,,,,,,,,,,,, +146103,,,0.9429807,0.20439605,0.7515,1.066416,50000.0,0.62600005,1.8532897,10000.0,39972.81428909302,40436.67267847061,39972.81428909302,461.0949785709381,1.613271951675415,0.0 +146200,4.8038955,0.95624435,,,,,,,,,,,,,, +146300,4.9134593,0.9702866,,,,,,,,,,,,,, +146400,5.0551896,0.9564162,,,,,,,,,,,,,, +146500,5.2766466,0.91624653,,,,,,,,,,,,,, +146600,5.106797,1.0166812,,,,,,,,,,,,,, +146700,4.7569604,0.874712,,,,,,,,,,,,,, +146800,5.2172456,1.0309994,,,,,,,,,,,,,, +146900,5.080925,0.92441356,,,,,,,,,,,,,, +147000,5.0213284,1.0281836,,,,,,,,,,,,,, +147100,5.00534,0.9713713,,,,,,,,,,,,,, +147200,4.9248886,0.94448876,,,,,,,,,,,,,, +147300,4.7263513,0.92531,,,,,,,,,,,,,, +147400,4.7506776,0.9350608,,,,,,,,,,,,,, +147500,4.8813634,0.92773765,,,,,,,,,,,,,, +147600,5.020035,1.0242423,,,,,,,,,,,,,, +147700,4.9795218,0.9865715,,,,,,,,,,,,,, +147800,4.8781495,0.8695657,,,,,,,,,,,,,, +147900,5.064438,0.9221641,,,,,,,,,,,,,, +148000,5.002826,0.95648885,,,,,,,,,,,,,, +148100,4.8189783,0.9855359,,,,,,,,,,,,,, +148200,5.116714,0.99981964,,,,,,,,,,,,,, +148300,5.0236015,0.97416306,,,,,,,,,,,,,, +148400,5.208729,0.97097945,,,,,,,,,,,,,, +148500,4.755453,0.92025614,,,,,,,,,,,,,, +148600,4.887612,0.9601861,,,,,,,,,,,,,, +148700,4.778486,0.91506916,,,,,,,,,,,,,, +148800,4.9427943,0.9174061,,,,,,,,,,,,,, +148900,4.9997463,0.9616287,,,,,,,,,,,,,, +149000,4.6048717,0.89654386,,,,,,,,,,,,,, +149100,5.144637,0.9518136,,,,,,,,,,,,,, +149200,4.869402,0.89132726,,,,,,,,,,,,,, +149300,5.179067,1.0208237,,,,,,,,,,,,,, +149400,5.0708556,0.96085435,,,,,,,,,,,,,, +149500,5.0293846,0.9893565,,,,,,,,,,,,,, +149600,4.937864,0.91909885,,,,,,,,,,,,,, +149700,4.742585,0.9082565,,,,,,,,,,,,,, +149800,5.142239,0.9967792,,,,,,,,,,,,,, +149900,5.3548846,0.98734266,,,,,,,,,,,,,, +150000,4.6390815,0.94588614,,,,,,,,,,,,,, +150100,4.672127,0.9393328,,,,,,,,,,,,,, +150200,4.948758,0.98376966,,,,,,,,,,,,,, +150300,4.8815455,0.95514023,,,,,,,,,,,,,, +150400,5.178323,1.1292624,,,,,,,,,,,,,, +150500,4.8781,1.0007592,,,,,,,,,,,,,, +150600,5.0270824,0.92736584,,,,,,,,,,,,,, +150700,4.808567,0.9700811,,,,,,,,,,,,,, +150800,4.8211904,0.9482806,,,,,,,,,,,,,, +150900,5.027789,1.010655,,,,,,,,,,,,,, +151000,5.0265183,0.94293374,,,,,,,,,,,,,, +151100,5.158107,0.88594484,,,,,,,,,,,,,, +151200,4.905797,0.9547635,,,,,,,,,,,,,, +151300,4.807074,0.9206084,,,,,,,,,,,,,, +151400,4.8243117,0.9498526,,,,,,,,,,,,,, +151500,4.9171968,0.9390757,,,,,,,,,,,,,, +151600,5.0971723,0.9732065,,,,,,,,,,,,,, +151700,4.9729095,0.9788637,,,,,,,,,,,,,, +151800,4.8625555,0.9667123,,,,,,,,,,,,,, +151900,5.4912972,1.0652008,,,,,,,,,,,,,, +152000,4.955933,0.9750007,,,,,,,,,,,,,, +152100,4.49873,0.92735547,,,,,,,,,,,,,, +152200,4.7530427,0.9542336,,,,,,,,,,,,,, +152300,5.6092243,0.94556206,,,,,,,,,,,,,, +152400,5.0814023,0.9581666,,,,,,,,,,,,,, +152500,4.9004316,1.0175868,,,,,,,,,,,,,, +152600,5.101755,1.0552105,,,,,,,,,,,,,, +152700,5.0449004,0.94166934,,,,,,,,,,,,,, +152800,5.051442,0.9287148,,,,,,,,,,,,,, +152900,4.61209,0.8741386,,,,,,,,,,,,,, +153000,4.9225035,0.9936746,,,,,,,,,,,,,, +153100,4.998672,1.0102415,,,,,,,,,,,,,, +153200,4.8431015,0.82440287,,,,,,,,,,,,,, +153300,5.432654,1.0285845,,,,,,,,,,,,,, +153400,4.879928,0.9396484,,,,,,,,,,,,,, +153459,,,0.9436384,0.19484542,0.7524,1.0662144,50000.0,0.6261,1.8529103,10000.0,41968.87982940674,42451.374479055405,41968.87982940674,479.6059491634369,1.682976484298706,0.0 +153500,5.0841136,1.0130479,,,,,,,,,,,,,, +153600,5.0804777,1.008426,,,,,,,,,,,,,, +153700,5.0858965,0.99277216,,,,,,,,,,,,,, +153800,4.9493628,0.9818705,,,,,,,,,,,,,, +153900,4.858595,0.9233124,,,,,,,,,,,,,, +154000,4.657139,0.87085885,,,,,,,,,,,,,, +154100,5.0590425,1.0081186,,,,,,,,,,,,,, +154200,5.4274087,1.0356488,,,,,,,,,,,,,, +154300,4.7981467,0.9370342,,,,,,,,,,,,,, +154400,4.6711125,0.8765662,,,,,,,,,,,,,, +154500,4.8528304,0.9968103,,,,,,,,,,,,,, +154600,4.988565,0.96603477,,,,,,,,,,,,,, +154700,4.607136,0.8240559,,,,,,,,,,,,,, +154800,4.7729225,0.88685834,,,,,,,,,,,,,, +154900,4.715194,0.9271548,,,,,,,,,,,,,, +155000,5.4120555,0.95471305,,,,,,,,,,,,,, +155100,5.200059,0.98184913,,,,,,,,,,,,,, +155200,4.8689237,0.9547142,,,,,,,,,,,,,, +155300,5.25053,1.0281522,,,,,,,,,,,,,, +155400,5.072221,1.065803,,,,,,,,,,,,,, +155500,5.155465,1.0150226,,,,,,,,,,,,,, +155600,5.028554,1.0055058,,,,,,,,,,,,,, +155700,4.7288203,0.94165444,,,,,,,,,,,,,, +155800,5.025787,0.9812809,,,,,,,,,,,,,, +155900,4.7158856,0.87852114,,,,,,,,,,,,,, +156000,5.6016517,1.0576088,,,,,,,,,,,,,, +156100,4.8797855,0.95251256,,,,,,,,,,,,,, +156200,4.7306504,0.96325415,,,,,,,,,,,,,, +156300,5.4593425,1.0482864,,,,,,,,,,,,,, +156400,5.184152,0.9838723,,,,,,,,,,,,,, +156500,5.0618687,0.90943366,,,,,,,,,,,,,, +156600,5.1330366,0.9722123,,,,,,,,,,,,,, +156700,5.04715,0.86827683,,,,,,,,,,,,,, +156800,4.84176,0.90902734,,,,,,,,,,,,,, +156900,5.3051186,0.9929329,,,,,,,,,,,,,, +157000,4.8152227,0.91003025,,,,,,,,,,,,,, +157100,4.7937055,0.970716,,,,,,,,,,,,,, +157200,4.8465967,0.92580765,,,,,,,,,,,,,, +157300,4.93716,0.9036329,,,,,,,,,,,,,, +157400,5.080944,1.0339336,,,,,,,,,,,,,, +157500,5.148934,0.98465395,,,,,,,,,,,,,, +157600,5.1306186,0.99608517,,,,,,,,,,,,,, +157700,5.2707496,1.0861318,,,,,,,,,,,,,, +157800,5.1779695,1.0285476,,,,,,,,,,,,,, +157900,4.8627353,0.9337355,,,,,,,,,,,,,, +158000,5.14918,0.9658537,,,,,,,,,,,,,, +158100,5.3401794,1.0032299,,,,,,,,,,,,,, +158200,5.1848583,0.9616742,,,,,,,,,,,,,, +158300,4.9770937,0.94022596,,,,,,,,,,,,,, +158400,5.1106796,1.0157773,,,,,,,,,,,,,, +158500,4.6623416,0.83289486,,,,,,,,,,,,,, +158600,5.057219,1.0304258,,,,,,,,,,,,,, +158700,4.930997,0.96001375,,,,,,,,,,,,,, +158800,5.361905,1.0094714,,,,,,,,,,,,,, +158900,5.0315576,0.91560423,,,,,,,,,,,,,, +159000,5.041659,0.97364795,,,,,,,,,,,,,, +159100,4.904654,0.9062264,,,,,,,,,,,,,, +159200,5.0043793,0.9613056,,,,,,,,,,,,,, +159300,4.8167806,0.9794607,,,,,,,,,,,,,, +159400,5.0650764,0.94073236,,,,,,,,,,,,,, +159500,5.1277013,0.9812685,,,,,,,,,,,,,, +159600,5.0197616,0.91981506,,,,,,,,,,,,,, +159700,4.7904115,0.96662843,,,,,,,,,,,,,, +159800,5.233886,0.95171833,,,,,,,,,,,,,, +159900,5.0686293,0.99520797,,,,,,,,,,,,,, +160000,4.8609715,0.94438887,,,,,,,,,,,,,, +160100,4.9655395,0.93812776,,,,,,,,,,,,,, +160200,5.004106,0.9026545,,,,,,,,,,,,,, +160300,5.2449565,1.0446203,,,,,,,,,,,,,, +160400,4.8209677,0.89405495,,,,,,,,,,,,,, +160500,4.867817,0.91132677,,,,,,,,,,,,,, +160600,5.097232,0.9306005,,,,,,,,,,,,,, +160700,4.905991,0.91726714,,,,,,,,,,,,,, +160800,5.203261,0.93875927,,,,,,,,,,,,,, +160803,,,0.94782364,0.18040912,0.75266,1.0689249,50000.0,0.62670004,1.859921,10000.0,43964.87986493111,44467.10048651695,43964.87986493111,499.2032101154328,1.7564749717712402,0.0 +160900,5.037517,1.014205,,,,,,,,,,,,,, +161000,4.8889546,0.93165946,,,,,,,,,,,,,, +161100,5.1638823,0.93860114,,,,,,,,,,,,,, +161200,4.968267,0.90621024,,,,,,,,,,,,,, +161300,4.931264,0.98480034,,,,,,,,,,,,,, +161400,4.6721807,0.9350362,,,,,,,,,,,,,, +161500,4.6825995,0.8460815,,,,,,,,,,,,,, +161600,4.86004,0.8850826,,,,,,,,,,,,,, +161700,5.053499,0.95577586,,,,,,,,,,,,,, +161800,4.9206643,0.9078186,,,,,,,,,,,,,, +161900,5.013975,0.97029567,,,,,,,,,,,,,, +162000,5.036552,1.0147963,,,,,,,,,,,,,, +162100,5.0309877,0.9494567,,,,,,,,,,,,,, +162200,4.8815303,0.9614452,,,,,,,,,,,,,, +162300,5.1803293,0.98833287,,,,,,,,,,,,,, +162400,4.720633,0.9268122,,,,,,,,,,,,,, +162500,5.08119,0.97253275,,,,,,,,,,,,,, +162600,4.925674,0.91654664,,,,,,,,,,,,,, +162700,5.0532994,0.9621924,,,,,,,,,,,,,, +162800,5.2428427,1.0133158,,,,,,,,,,,,,, +162900,5.092437,0.9443873,,,,,,,,,,,,,, +163000,4.980007,0.9467978,,,,,,,,,,,,,, +163100,5.3279657,1.0417017,,,,,,,,,,,,,, +163200,4.902069,0.9210684,,,,,,,,,,,,,, +163300,4.991594,0.9212215,,,,,,,,,,,,,, +163400,5.028981,0.9454931,,,,,,,,,,,,,, +163500,5.0800457,0.93853545,,,,,,,,,,,,,, +163600,4.892621,0.9650777,,,,,,,,,,,,,, +163700,4.910309,0.9428973,,,,,,,,,,,,,, +163800,5.1525016,0.9682492,,,,,,,,,,,,,, +163900,5.2023206,0.9867276,,,,,,,,,,,,,, +164000,5.076255,0.9135781,,,,,,,,,,,,,, +164100,4.870944,0.9574759,,,,,,,,,,,,,, +164200,5.0169096,0.9958412,,,,,,,,,,,,,, +164300,4.996651,0.91204846,,,,,,,,,,,,,, +164400,4.6682386,0.853723,,,,,,,,,,,,,, +164500,5.567542,1.0777037,,,,,,,,,,,,,, +164600,5.2475615,1.0412563,,,,,,,,,,,,,, +164700,4.9088554,1.0104761,,,,,,,,,,,,,, +164800,5.1441245,0.95668614,,,,,,,,,,,,,, +164900,5.1075997,0.89963275,,,,,,,,,,,,,, +165000,5.097186,1.0018722,,,,,,,,,,,,,, +165100,5.0312595,0.9734193,,,,,,,,,,,,,, +165200,5.0014367,0.89739746,,,,,,,,,,,,,, +165300,5.1592975,0.9280201,,,,,,,,,,,,,, +165400,5.284389,1.0294299,,,,,,,,,,,,,, +165500,4.7996607,0.8889109,,,,,,,,,,,,,, +165600,5.0639925,0.94946563,,,,,,,,,,,,,, +165700,4.989663,0.98080516,,,,,,,,,,,,,, +165800,5.2751727,0.9803785,,,,,,,,,,,,,, +165900,5.366087,1.02744,,,,,,,,,,,,,, +166000,4.8082786,0.9155481,,,,,,,,,,,,,, +166100,5.000519,0.95377505,,,,,,,,,,,,,, +166200,5.0391293,0.9233691,,,,,,,,,,,,,, +166300,5.154555,0.934816,,,,,,,,,,,,,, +166400,5.1776624,0.9203296,,,,,,,,,,,,,, +166500,5.1614647,0.99396014,,,,,,,,,,,,,, +166600,5.198626,0.97853285,,,,,,,,,,,,,, +166700,5.176138,1.0715743,,,,,,,,,,,,,, +166800,5.161009,0.9538119,,,,,,,,,,,,,, +166900,5.189356,0.91116726,,,,,,,,,,,,,, +167000,5.27495,0.9589703,,,,,,,,,,,,,, +167100,5.406421,0.885055,,,,,,,,,,,,,, +167200,5.0418983,1.0291266,,,,,,,,,,,,,, +167300,5.2643137,0.92863333,,,,,,,,,,,,,, +167400,5.0245924,0.9725859,,,,,,,,,,,,,, +167500,4.800239,0.9462224,,,,,,,,,,,,,, +167600,4.937391,0.98827696,,,,,,,,,,,,,, +167700,4.8945403,0.90044415,,,,,,,,,,,,,, +167800,5.142029,0.8912182,,,,,,,,,,,,,, +167900,4.999546,0.97196114,,,,,,,,,,,,,, +167972,,,0.9487404,0.18130454,0.75286,1.0714159,50000.0,0.6251,1.8615223,10000.0,45960.7723968029,46516.528638362885,45960.7723968029,552.6016957759857,1.8376588821411133,0.0 +168000,4.9667206,0.90568006,,,,,,,,,,,,,, +168100,4.865715,0.9361111,,,,,,,,,,,,,, +168200,5.3461957,1.0585035,,,,,,,,,,,,,, +168300,5.1579113,0.9996544,,,,,,,,,,,,,, +168400,4.634342,0.87394446,,,,,,,,,,,,,, +168500,5.052712,0.9680425,,,,,,,,,,,,,, +168600,5.0203314,0.8990182,,,,,,,,,,,,,, +168700,4.8133707,0.89501846,,,,,,,,,,,,,, +168800,4.8332806,0.94627297,,,,,,,,,,,,,, +168900,5.2632256,1.0312617,,,,,,,,,,,,,, +169000,4.633801,0.8590571,,,,,,,,,,,,,, +169100,4.5570154,0.88942057,,,,,,,,,,,,,, +169200,5.018108,0.898605,,,,,,,,,,,,,, +169300,4.9812727,0.9693943,,,,,,,,,,,,,, +169400,5.1240463,0.8985845,,,,,,,,,,,,,, +169500,4.8031077,0.992644,,,,,,,,,,,,,, +169600,5.298022,0.9705309,,,,,,,,,,,,,, +169700,5.0749555,0.94725716,,,,,,,,,,,,,, +169800,4.8649845,0.92673326,,,,,,,,,,,,,, +169900,4.832165,1.0206891,,,,,,,,,,,,,, +170000,4.9505887,0.9472445,,,,,,,,,,,,,, +170100,4.8602667,0.95381796,,,,,,,,,,,,,, +170200,4.934642,0.9659865,,,,,,,,,,,,,, +170300,4.848309,0.9136912,,,,,,,,,,,,,, +170400,4.8304887,0.9700785,,,,,,,,,,,,,, +170500,5.0400515,0.9494431,,,,,,,,,,,,,, +170600,4.8526382,0.9527172,,,,,,,,,,,,,, +170700,4.900259,0.9513532,,,,,,,,,,,,,, +170800,5.142589,0.9712235,,,,,,,,,,,,,, +170900,5.310713,0.97291136,,,,,,,,,,,,,, +171000,5.082428,0.90812534,,,,,,,,,,,,,, +171100,4.91624,0.939294,,,,,,,,,,,,,, +171200,4.984757,0.882973,,,,,,,,,,,,,, +171300,5.198333,0.952021,,,,,,,,,,,,,, +171400,4.9439793,0.95827705,,,,,,,,,,,,,, +171500,5.0407186,0.90364087,,,,,,,,,,,,,, +171600,4.8954487,0.8646008,,,,,,,,,,,,,, +171700,4.9831486,0.98488456,,,,,,,,,,,,,, +171800,4.9244237,0.9151436,,,,,,,,,,,,,, +171900,5.0485964,0.8485209,,,,,,,,,,,,,, +172000,5.5621595,0.98969007,,,,,,,,,,,,,, +172100,5.0107727,0.94153994,,,,,,,,,,,,,, +172200,4.9562173,0.8833257,,,,,,,,,,,,,, +172300,5.030798,0.97043717,,,,,,,,,,,,,, +172400,5.1187468,0.96976376,,,,,,,,,,,,,, +172500,5.2063456,0.9677541,,,,,,,,,,,,,, +172600,5.2370067,0.9463266,,,,,,,,,,,,,, +172658,,,0.94868064,0.17847335,0.75325996,1.0750405,50000.0,0.625,1.8682417,10000.0,47956.8501329422,48535.025040864944,47956.8501329422,574.862462759018,1.9537787437438965,0.0 +172700,4.940844,0.87018794,,,,,,,,,,,,,, +172800,5.219758,0.9864238,,,,,,,,,,,,,, +172900,5.1726384,0.9566007,,,,,,,,,,,,,, +173000,4.9370575,0.9513858,,,,,,,,,,,,,, +173100,5.188773,1.0096965,,,,,,,,,,,,,, +173200,5.5011144,0.93564594,,,,,,,,,,,,,, +173300,5.1147885,1.0001646,,,,,,,,,,,,,, +173400,4.9102592,0.907464,,,,,,,,,,,,,, +173500,5.3325634,1.0502313,,,,,,,,,,,,,, +173600,4.833886,0.91871285,,,,,,,,,,,,,, +173700,5.04689,0.98549616,,,,,,,,,,,,,, +173800,4.963953,0.95398504,,,,,,,,,,,,,, +173900,5.0499053,0.89519405,,,,,,,,,,,,,, +174000,5.104269,0.95325047,,,,,,,,,,,,,, +174100,5.2159505,0.9204903,,,,,,,,,,,,,, +174200,5.1894455,0.9562033,,,,,,,,,,,,,, +174300,5.115742,0.9965088,,,,,,,,,,,,,, +174400,4.7565117,0.9150069,,,,,,,,,,,,,, +174500,5.12375,0.9469827,,,,,,,,,,,,,, +174600,4.8595777,0.8403811,,,,,,,,,,,,,, +174700,5.1539993,0.94129926,,,,,,,,,,,,,, +174800,5.2303467,0.95297694,,,,,,,,,,,,,, +174900,4.8271403,0.88603956,,,,,,,,,,,,,, +175000,4.649547,0.87792677,,,,,,,,,,,,,, +175100,4.9934826,0.9112837,,,,,,,,,,,,,, +175200,4.8498554,0.8608945,,,,,,,,,,,,,, +175300,5.3136797,0.97176397,,,,,,,,,,,,,, +175400,4.7701316,0.84880847,,,,,,,,,,,,,, +175500,5.0794625,0.9661242,,,,,,,,,,,,,, +175600,4.9380674,0.8713828,,,,,,,,,,,,,, +175700,5.0373516,0.89052814,,,,,,,,,,,,,, +175800,5.253012,0.9890426,,,,,,,,,,,,,, +175900,5.051005,0.86877954,,,,,,,,,,,,,, +176000,4.952812,0.900445,,,,,,,,,,,,,, +176100,4.9722347,0.92272365,,,,,,,,,,,,,, +176200,4.9814053,0.9230484,,,,,,,,,,,,,, +176300,5.3722734,0.9524778,,,,,,,,,,,,,, +176400,5.109136,0.9017881,,,,,,,,,,,,,, +176500,4.788116,0.8571448,,,,,,,,,,,,,, +176600,5.1439433,0.9455743,,,,,,,,,,,,,, +176700,5.3004303,0.97190344,,,,,,,,,,,,,, +176800,4.8940396,0.88531625,,,,,,,,,,,,,, +176900,5.3910832,0.9754826,,,,,,,,,,,,,, +177000,4.90068,0.95891047,,,,,,,,,,,,,, +177063,,,0.9486607,0.17960994,0.7531,1.074318,50000.0,0.62600005,1.8686966,10000.0,49952.47198367119,50553.51904273033,49952.47198367119,597.1016075611115,2.3890323638916016,0.0 +177100,5.083138,0.87299967,,,,,,,,,,,,,, +177200,5.0645165,0.98934907,,,,,,,,,,,,,, +177300,5.158038,0.97984385,,,,,,,,,,,,,, +177400,4.817815,0.8882735,,,,,,,,,,,,,, +177500,5.0585246,0.9602864,,,,,,,,,,,,,, +177600,5.0784893,1.0060945,,,,,,,,,,,,,, +177700,5.125732,0.9528493,,,,,,,,,,,,,, +177800,5.3231044,0.9729432,,,,,,,,,,,,,, +177900,5.591818,1.1219244,,,,,,,,,,,,,, +178000,5.0245876,0.9472587,,,,,,,,,,,,,, +178100,5.0327716,0.9264954,,,,,,,,,,,,,, +178200,4.873851,0.86888224,,,,,,,,,,,,,, +178300,5.066472,0.96364945,,,,,,,,,,,,,, +178400,4.8289022,0.86538017,,,,,,,,,,,,,, +178500,5.3049726,0.91750336,,,,,,,,,,,,,, +178600,5.0984864,0.93344146,,,,,,,,,,,,,, +178700,5.105941,0.9509733,,,,,,,,,,,,,, +178800,4.9821053,0.9802786,,,,,,,,,,,,,, +178900,5.475266,1.0458037,,,,,,,,,,,,,, +179000,5.247165,0.9071815,,,,,,,,,,,,,, +179100,5.0495133,0.96467304,,,,,,,,,,,,,, +179200,4.89781,0.88850945,,,,,,,,,,,,,, +179300,4.8708215,0.9540775,,,,,,,,,,,,,, +179400,5.363085,1.0525954,,,,,,,,,,,,,, +179500,4.866728,0.84505343,,,,,,,,,,,,,, +179600,5.0781813,0.97084004,,,,,,,,,,,,,, +179700,5.0709515,0.8997239,,,,,,,,,,,,,, +179800,5.2242165,0.9835059,,,,,,,,,,,,,, +179900,4.855396,0.9756489,,,,,,,,,,,,,, +180000,4.952844,0.8926188,,,,,,,,,,,,,, +180100,5.116674,0.8991693,,,,,,,,,,,,,, +180200,5.0964427,0.9224396,,,,,,,,,,,,,, +180300,4.917978,0.9203802,,,,,,,,,,,,,, +180400,5.192339,0.94506943,,,,,,,,,,,,,, +180500,4.6337194,0.8726009,,,,,,,,,,,,,, +180600,5.154436,0.9764581,,,,,,,,,,,,,, +180700,5.0797834,0.880682,,,,,,,,,,,,,, +180800,5.22919,0.97558826,,,,,,,,,,,,,, +180900,4.8038435,0.8953628,,,,,,,,,,,,,, +181000,4.8384213,0.8455546,,,,,,,,,,,,,, +181100,5.175808,0.9235393,,,,,,,,,,,,,, +181200,4.8665957,0.8268726,,,,,,,,,,,,,, +181300,4.944615,0.9143164,,,,,,,,,,,,,, +181400,5.1056776,0.9457406,,,,,,,,,,,,,, +181500,4.8925104,0.9019402,,,,,,,,,,,,,, +181600,,,0.9502152,0.17258252,0.75378,1.0762789,50000.0,0.62560004,1.8692299,10000.0,51947.90822029114,52589.320518255234,51947.90822029114,636.8429207801819,2.812356948852539,0.0 +181600,5.3622847,0.99221677,,,,,,,,,,,,,, +181700,4.872026,0.92770183,,,,,,,,,,,,,, +181800,5.053481,0.8754605,,,,,,,,,,,,,, +181900,5.119721,0.94041973,,,,,,,,,,,,,, +182000,4.9610143,0.9360741,,,,,,,,,,,,,, +182100,5.327876,1.000843,,,,,,,,,,,,,, +182200,4.8366427,0.8613524,,,,,,,,,,,,,, +182300,4.9496202,0.9395437,,,,,,,,,,,,,, +182400,5.1615825,0.97577775,,,,,,,,,,,,,, +182500,4.9494348,0.930107,,,,,,,,,,,,,, +182600,5.127139,0.9452971,,,,,,,,,,,,,, +182700,4.730691,0.86393654,,,,,,,,,,,,,, +182800,5.299928,1.0132122,,,,,,,,,,,,,, +182900,5.237121,0.9050325,,,,,,,,,,,,,, +183000,5.451279,0.95761466,,,,,,,,,,,,,, +183100,4.8097863,0.80775845,,,,,,,,,,,,,, +183200,4.9937406,0.9438605,,,,,,,,,,,,,, +183300,5.275067,0.98696387,,,,,,,,,,,,,, +183400,5.3849554,1.0459772,,,,,,,,,,,,,, +183500,5.105806,0.9955981,,,,,,,,,,,,,, +183600,5.012576,0.9461968,,,,,,,,,,,,,, +183700,5.06446,0.8691977,,,,,,,,,,,,,, +183800,5.1540456,0.98678195,,,,,,,,,,,,,, +183900,5.318316,1.0325848,,,,,,,,,,,,,, +184000,5.238353,0.9603834,,,,,,,,,,,,,, +184100,5.2783756,0.9775616,,,,,,,,,,,,,, +184200,5.2675366,1.0023177,,,,,,,,,,,,,, +184300,5.4626455,0.9463537,,,,,,,,,,,,,, +184400,5.30923,0.9429004,,,,,,,,,,,,,, +184500,4.915128,0.9007395,,,,,,,,,,,,,, +184600,4.8828964,0.86510515,,,,,,,,,,,,,, +184700,4.9021835,0.8582039,,,,,,,,,,,,,, +184800,5.1856465,0.9449484,,,,,,,,,,,,,, +184900,4.9250855,0.90337765,,,,,,,,,,,,,, +185000,4.9524736,0.848644,,,,,,,,,,,,,, +185100,5.4493804,1.0216895,,,,,,,,,,,,,, +185200,5.1339765,0.9937183,,,,,,,,,,,,,, +185300,4.6884546,0.8530359,,,,,,,,,,,,,, +185400,5.02691,0.8916329,,,,,,,,,,,,,, +185500,4.8696003,0.88487995,,,,,,,,,,,,,, +185600,5.6215706,0.972126,,,,,,,,,,,,,, +185700,5.1800814,0.8848181,,,,,,,,,,,,,, +185800,5.286177,0.96173334,,,,,,,,,,,,,, +185900,5.3633227,0.99414754,,,,,,,,,,,,,, +186000,4.957035,0.8803667,,,,,,,,,,,,,, +186073,,,0.9526666,0.16909547,0.7543,1.0771687,50000.0,0.62570006,1.869183,10000.0,53943.446397304535,54627.26826930046,53943.446397304535,678.5254654884338,3.305555582046509,0.0 +186100,4.806111,0.92890656,,,,,,,,,,,,,, +186200,5.5011334,0.9749775,,,,,,,,,,,,,, +186300,4.9684978,0.94679016,,,,,,,,,,,,,, +186400,5.216027,0.8990528,,,,,,,,,,,,,, +186500,5.426215,1.0138754,,,,,,,,,,,,,, +186600,5.139348,0.9900546,,,,,,,,,,,,,, +186700,4.8914943,0.9578847,,,,,,,,,,,,,, +186800,4.9220734,0.8251274,,,,,,,,,,,,,, +186900,5.27943,0.9530755,,,,,,,,,,,,,, +187000,4.8181367,0.88959026,,,,,,,,,,,,,, +187100,5.0971355,0.93919635,,,,,,,,,,,,,, +187200,5.1756725,0.99206024,,,,,,,,,,,,,, +187300,4.809934,0.91358733,,,,,,,,,,,,,, +187400,5.164145,0.92586744,,,,,,,,,,,,,, +187500,5.077636,0.90955406,,,,,,,,,,,,,, +187600,5.17591,0.93907887,,,,,,,,,,,,,, +187700,5.1735115,1.0058075,,,,,,,,,,,,,, +187800,5.4866014,0.9529781,,,,,,,,,,,,,, +187900,5.3130198,0.9315965,,,,,,,,,,,,,, +188000,5.065372,0.961065,,,,,,,,,,,,,, +188100,5.4068046,1.0355916,,,,,,,,,,,,,, +188200,4.739466,0.85454494,,,,,,,,,,,,,, +188300,5.3673916,0.95074797,,,,,,,,,,,,,, +188400,5.148793,0.99228877,,,,,,,,,,,,,, +188500,5.1867743,1.0059931,,,,,,,,,,,,,, +188600,5.488479,1.0157907,,,,,,,,,,,,,, +188700,5.2515006,0.9174964,,,,,,,,,,,,,, +188800,5.016391,0.89245325,,,,,,,,,,,,,, +188900,5.1955414,0.9385728,,,,,,,,,,,,,, +189000,5.28581,1.0060282,,,,,,,,,,,,,, +189100,5.346924,0.9576769,,,,,,,,,,,,,, +189200,5.2016892,0.9643958,,,,,,,,,,,,,, +189300,5.363741,0.979339,,,,,,,,,,,,,, +189400,4.6979003,0.8371133,,,,,,,,,,,,,, +189500,5.0393295,0.91196537,,,,,,,,,,,,,, +189600,5.3069577,0.9534136,,,,,,,,,,,,,, +189700,5.4665856,0.9276847,,,,,,,,,,,,,, +189800,5.1327167,0.91830456,,,,,,,,,,,,,, +189900,5.4839506,1.0313573,,,,,,,,,,,,,, +190000,5.0383034,0.8895503,,,,,,,,,,,,,, +190100,5.2414007,0.9587568,,,,,,,,,,,,,, +190200,5.140752,0.9453199,,,,,,,,,,,,,, +190300,5.034205,0.8800328,,,,,,,,,,,,,, +190400,5.2709827,0.9857955,,,,,,,,,,,,,, +190500,5.0481987,0.8981488,,,,,,,,,,,,,, +190600,4.8335714,0.83755136,,,,,,,,,,,,,, +190646,,,0.9512117,0.16966121,0.75439996,1.0768428,50000.0,0.628,1.866245,10000.0,55938.98239707947,56643.01434922218,55938.98239707947,698.0477120876312,3.8167998790740967,0.0 +190700,5.12762,0.92933977,,,,,,,,,,,,,, +190800,5.5687165,0.9613124,,,,,,,,,,,,,, +190900,5.2257323,0.89820874,,,,,,,,,,,,,, +191000,5.011116,0.93470347,,,,,,,,,,,,,, +191100,5.09822,0.940812,,,,,,,,,,,,,, +191200,5.050279,0.90928435,,,,,,,,,,,,,, +191300,5.2563705,0.9844318,,,,,,,,,,,,,, +191400,4.946031,0.93832785,,,,,,,,,,,,,, +191500,5.29481,0.93308747,,,,,,,,,,,,,, +191600,5.085925,0.9546743,,,,,,,,,,,,,, +191700,5.3063464,0.9143281,,,,,,,,,,,,,, +191800,4.9300823,0.9501845,,,,,,,,,,,,,, +191900,5.524637,0.9391636,,,,,,,,,,,,,, +192000,4.9932084,0.9396571,,,,,,,,,,,,,, +192100,5.27013,0.9398738,,,,,,,,,,,,,, +192200,4.7685385,0.87414956,,,,,,,,,,,,,, +192300,5.0682607,0.878118,,,,,,,,,,,,,, +192400,5.16956,0.96260196,,,,,,,,,,,,,, +192500,5.0937977,0.94473517,,,,,,,,,,,,,, +192600,4.9406357,0.82841796,,,,,,,,,,,,,, +192700,5.1646237,0.935638,,,,,,,,,,,,,, +192800,5.2365656,0.97320145,,,,,,,,,,,,,, +192900,5.301438,0.89290154,,,,,,,,,,,,,, +193000,5.0841575,0.86986995,,,,,,,,,,,,,, +193100,5.1571226,0.9262805,,,,,,,,,,,,,, +193200,4.918476,0.9366667,,,,,,,,,,,,,, +193300,5.151684,0.98081183,,,,,,,,,,,,,, +193400,5.007638,0.8987859,,,,,,,,,,,,,, +193500,4.9236155,0.8764314,,,,,,,,,,,,,, +193600,4.9784536,0.903165,,,,,,,,,,,,,, +193700,4.623264,0.817997,,,,,,,,,,,,,, +193800,5.2015805,0.96508807,,,,,,,,,,,,,, +193900,5.1869817,0.89804745,,,,,,,,,,,,,, +194000,5.27906,0.8967749,,,,,,,,,,,,,, +194100,5.283215,0.8775999,,,,,,,,,,,,,, +194200,5.1260552,0.9229586,,,,,,,,,,,,,, +194300,5.099612,0.8545802,,,,,,,,,,,,,, +194400,5.2809057,0.9352467,,,,,,,,,,,,,, +194500,5.1532326,0.9404499,,,,,,,,,,,,,, +194600,5.1550264,0.8865392,,,,,,,,,,,,,, +194700,5.1605744,0.9380514,,,,,,,,,,,,,, +194747,,,0.95336413,0.16278976,0.75426,1.0784762,50000.0,0.62960005,1.8696264,10000.0,57934.57347178459,58682.21131968498,57934.57347178459,740.9094491004944,4.328880071640015,0.0 +194800,4.9595065,0.9924592,,,,,,,,,,,,,, +194900,5.2913914,1.012589,,,,,,,,,,,,,, +195000,5.354171,0.9368958,,,,,,,,,,,,,, +195100,5.3229213,0.98429227,,,,,,,,,,,,,, +195200,5.432538,0.90209436,,,,,,,,,,,,,, +195300,5.183285,0.8722374,,,,,,,,,,,,,, +195400,5.262555,0.95591605,,,,,,,,,,,,,, +195500,5.0275655,0.85013604,,,,,,,,,,,,,, +195600,5.0412455,0.9120882,,,,,,,,,,,,,, +195700,5.0854917,0.9187383,,,,,,,,,,,,,, +195800,4.833142,0.8784838,,,,,,,,,,,,,, +195900,4.9262476,0.86095774,,,,,,,,,,,,,, +196000,4.961814,0.9240525,,,,,,,,,,,,,, +196100,5.0567865,0.85896134,,,,,,,,,,,,,, +196200,4.812472,0.8642634,,,,,,,,,,,,,, +196300,5.1877766,0.93556845,,,,,,,,,,,,,, +196400,5.144963,0.9077987,,,,,,,,,,,,,, +196500,4.840472,0.8996868,,,,,,,,,,,,,, +196600,5.287327,0.946203,,,,,,,,,,,,,, +196700,5.3017445,0.95965517,,,,,,,,,,,,,, +196800,5.1461077,0.883515,,,,,,,,,,,,,, +196900,5.3053365,0.92152965,,,,,,,,,,,,,, +197000,5.017615,0.9195654,,,,,,,,,,,,,, +197100,5.0629888,0.8928728,,,,,,,,,,,,,, +197200,5.0612216,0.869267,,,,,,,,,,,,,, +197300,5.258444,0.9018351,,,,,,,,,,,,,, +197400,5.12415,0.9129344,,,,,,,,,,,,,, +197500,5.2068152,0.93061745,,,,,,,,,,,,,, +197600,5.14704,0.9926903,,,,,,,,,,,,,, +197700,4.949328,0.8642858,,,,,,,,,,,,,, +197800,5.0652146,0.95076144,,,,,,,,,,,,,, +197900,5.192774,0.9224343,,,,,,,,,,,,,, +198000,5.211414,0.88953817,,,,,,,,,,,,,, +198100,5.280054,0.87856174,,,,,,,,,,,,,, +198200,4.8758073,0.8343443,,,,,,,,,,,,,, +198300,5.231434,0.8687576,,,,,,,,,,,,,, +198400,5.2156715,0.9282453,,,,,,,,,,,,,, +198500,5.3147883,0.9560634,,,,,,,,,,,,,, +198524,,,0.95336413,0.16411076,0.75376,1.0798633,50000.0,0.62740004,1.8716532,10000.0,59930.2302134037,60715.90223026276,59930.2302134037,778.1086664199829,4.950573444366455,0.0 +198600,5.133053,0.92875296,,,,,,,,,,,,,, +198700,5.2628565,0.9024969,,,,,,,,,,,,,, +198800,5.0888453,0.9036923,,,,,,,,,,,,,, +198900,5.255682,0.894297,,,,,,,,,,,,,, +199000,5.0226808,0.8936321,,,,,,,,,,,,,, +199100,5.193354,0.9531555,,,,,,,,,,,,,, +199200,5.160801,0.921322,,,,,,,,,,,,,, +199300,5.307107,0.9299092,,,,,,,,,,,,,, +199400,5.2866254,1.0115434,,,,,,,,,,,,,, +199500,5.4391394,0.9880577,,,,,,,,,,,,,, +199600,4.7511578,0.80689144,,,,,,,,,,,,,, +199700,4.882023,0.8971051,,,,,,,,,,,,,, +199800,5.3260326,0.9369749,,,,,,,,,,,,,, +199900,5.408931,0.9897218,,,,,,,,,,,,,, +200000,4.966437,0.82147586,,,,,,,,,,,,,, +200100,5.4767866,0.9113569,,,,,,,,,,,,,, +200200,5.2087917,0.92814857,,,,,,,,,,,,,, +200300,5.2099924,0.8563293,,,,,,,,,,,,,, +200400,5.0004306,0.925491,,,,,,,,,,,,,, +200500,5.0872517,0.9056902,,,,,,,,,,,,,, +200600,5.1423836,0.92560124,,,,,,,,,,,,,, +200700,5.265192,0.9384474,,,,,,,,,,,,,, +200800,5.086089,0.8874867,,,,,,,,,,,,,, +200900,5.225386,0.9089085,,,,,,,,,,,,,, +201000,5.085196,0.86008996,,,,,,,,,,,,,, +201100,5.925341,1.0941434,,,,,,,,,,,,,, +201200,5.199377,0.87072325,,,,,,,,,,,,,, +201300,5.1465263,0.9521708,,,,,,,,,,,,,, +201400,5.14943,0.93107176,,,,,,,,,,,,,, +201500,5.063366,0.84711754,,,,,,,,,,,,,, +201600,5.153683,0.8930843,,,,,,,,,,,,,, +201700,4.9915323,0.8199549,,,,,,,,,,,,,, +201800,5.29566,0.91758853,,,,,,,,,,,,,, +201900,5.5136347,0.9488141,,,,,,,,,,,,,, +202000,5.1116557,0.91496813,,,,,,,,,,,,,, +202100,5.138202,0.9242304,,,,,,,,,,,,,, +202200,5.155074,0.98455656,,,,,,,,,,,,,, +202300,5.196497,0.90852,,,,,,,,,,,,,, +202362,,,0.952826,0.16415659,0.75398,1.0797056,50000.0,0.62560004,1.869789,10000.0,61926.093482255936,62742.43178105354,61926.093482255936,808.0945663452148,5.439246416091919,0.0 +202400,5.1221447,0.8791224,,,,,,,,,,,,,, +202500,5.0357637,0.8733915,,,,,,,,,,,,,, +202600,5.0075636,0.86243105,,,,,,,,,,,,,, +202700,4.734544,0.8612898,,,,,,,,,,,,,, +202800,4.852019,0.8549655,,,,,,,,,,,,,, +202900,5.375417,0.9270315,,,,,,,,,,,,,, +203000,5.422729,0.96705645,,,,,,,,,,,,,, +203100,5.0465183,0.9126134,,,,,,,,,,,,,, +203200,4.9898834,0.84088314,,,,,,,,,,,,,, +203300,4.943655,0.8968619,,,,,,,,,,,,,, +203400,4.7934556,0.7936607,,,,,,,,,,,,,, +203500,5.137631,0.8348447,,,,,,,,,,,,,, +203600,5.381656,0.895494,,,,,,,,,,,,,, +203700,4.9634905,0.9135277,,,,,,,,,,,,,, +203800,4.922465,0.8915773,,,,,,,,,,,,,, +203900,4.9560447,0.8133481,,,,,,,,,,,,,, +204000,5.0296316,0.87705815,,,,,,,,,,,,,, +204100,4.7747293,0.8761779,,,,,,,,,,,,,, +204200,4.8862424,0.8930196,,,,,,,,,,,,,, +204300,5.239153,0.90828204,,,,,,,,,,,,,, +204400,5.526373,0.98873806,,,,,,,,,,,,,, +204500,5.089106,0.9347708,,,,,,,,,,,,,, +204600,5.0705743,0.9702394,,,,,,,,,,,,,, +204700,5.0587535,0.91574323,,,,,,,,,,,,,, +204800,5.5886393,0.92564154,,,,,,,,,,,,,, +204900,5.469031,0.9574637,,,,,,,,,,,,,, +205000,5.0657516,0.9036243,,,,,,,,,,,,,, +205100,5.214702,0.9680282,,,,,,,,,,,,,, +205200,5.1035004,0.95822585,,,,,,,,,,,,,, +205300,5.0223727,0.88984895,,,,,,,,,,,,,, +205400,4.9859653,0.94779134,,,,,,,,,,,,,, +205500,5.104971,0.8947805,,,,,,,,,,,,,, +205600,5.211467,0.8741761,,,,,,,,,,,,,, +205700,5.073029,0.9650537,,,,,,,,,,,,,, +205800,4.9412413,0.86551404,,,,,,,,,,,,,, +205900,4.8720527,0.8863571,,,,,,,,,,,,,, +206000,5.331925,0.9609878,,,,,,,,,,,,,, +206100,5.2458935,0.96304,,,,,,,,,,,,,, +206200,5.1905055,0.8632318,,,,,,,,,,,,,, +206300,5.197038,0.9479949,,,,,,,,,,,,,, +206400,5.357416,0.89375925,,,,,,,,,,,,,, +206423,,,0.9546197,0.15765059,0.75356,1.0794744,50000.0,0.62600005,1.8705279,10000.0,63922.77934002876,64782.85824751854,63922.77934002876,851.0713455677032,5.971511363983154,0.0 +206500,5.202617,0.9641247,,,,,,,,,,,,,, +206600,5.0297523,0.8831148,,,,,,,,,,,,,, +206700,5.016756,0.91999304,,,,,,,,,,,,,, +206800,5.523676,1.0044433,,,,,,,,,,,,,, +206900,5.0952888,0.86605144,,,,,,,,,,,,,, +207000,5.3254056,0.93756,,,,,,,,,,,,,, +207100,5.393423,0.9074574,,,,,,,,,,,,,, +207200,5.4114547,0.95175636,,,,,,,,,,,,,, +207300,5.2044616,0.90429926,,,,,,,,,,,,,, +207400,5.037196,0.86758554,,,,,,,,,,,,,, +207500,4.9890313,0.8920212,,,,,,,,,,,,,, +207600,5.3708296,1.0028641,,,,,,,,,,,,,, +207700,5.0735116,0.8668697,,,,,,,,,,,,,, +207800,5.030136,0.8774216,,,,,,,,,,,,,, +207900,5.343798,0.9863862,,,,,,,,,,,,,, +208000,5.057969,0.8948505,,,,,,,,,,,,,, +208100,5.2199235,0.9331372,,,,,,,,,,,,,, +208200,5.3312054,0.8658391,,,,,,,,,,,,,, +208300,5.390745,0.937488,,,,,,,,,,,,,, +208400,4.948375,0.87297475,,,,,,,,,,,,,, +208500,5.174789,0.882633,,,,,,,,,,,,,, +208600,5.17386,0.9117546,,,,,,,,,,,,,, +208700,5.0974035,0.8253373,,,,,,,,,,,,,, +208800,5.0444393,0.9232882,,,,,,,,,,,,,, +208900,5.1759224,0.9277604,,,,,,,,,,,,,, +209000,5.3943624,0.9403527,,,,,,,,,,,,,, +209100,5.283135,0.90747947,,,,,,,,,,,,,, +209200,5.3166466,0.91779137,,,,,,,,,,,,,, +209300,5.2632537,0.97015977,,,,,,,,,,,,,, +209400,5.1502895,0.86308366,,,,,,,,,,,,,, +209500,5.359417,0.93403816,,,,,,,,,,,,,, +209600,5.0812573,0.86940795,,,,,,,,,,,,,, +209700,5.250354,0.9725811,,,,,,,,,,,,,, +209800,4.957873,0.8607099,,,,,,,,,,,,,, +209900,5.0096216,0.93337053,,,,,,,,,,,,,, +210000,5.106468,0.91204023,,,,,,,,,,,,,, +210100,5.2892833,0.9164852,,,,,,,,,,,,,, +210200,4.9267,0.89021426,,,,,,,,,,,,,, +210203,,,0.9555166,0.15610996,0.75497997,1.0815759,50000.0,0.62740004,1.8746012,10000.0,65918.245708704,66816.03911733627,65918.245708704,887.9483625888824,6.512280464172363,0.0 +210300,5.0991006,0.9033667,,,,,,,,,,,,,, +210400,4.975542,0.8824818,,,,,,,,,,,,,, +210500,4.950563,0.8531773,,,,,,,,,,,,,, +210600,5.150695,0.92282754,,,,,,,,,,,,,, +210700,5.025703,0.8377451,,,,,,,,,,,,,, +210800,5.244545,0.9158619,,,,,,,,,,,,,, +210900,4.9298596,0.8668631,,,,,,,,,,,,,, +211000,5.3297276,0.947093,,,,,,,,,,,,,, +211100,5.256029,0.944921,,,,,,,,,,,,,, +211200,5.128743,0.88420767,,,,,,,,,,,,,, +211300,4.7562246,0.85114443,,,,,,,,,,,,,, +211400,4.8288183,0.8599194,,,,,,,,,,,,,, +211500,5.161366,0.9285765,,,,,,,,,,,,,, +211600,5.312283,0.9032131,,,,,,,,,,,,,, +211700,5.2471294,0.95161283,,,,,,,,,,,,,, +211800,4.6070576,0.78942215,,,,,,,,,,,,,, +211900,4.8693075,0.86423904,,,,,,,,,,,,,, +212000,5.008451,0.9228736,,,,,,,,,,,,,, +212100,5.093581,0.9539192,,,,,,,,,,,,,, +212200,5.2038674,0.84322804,,,,,,,,,,,,,, +212300,4.937999,0.8450959,,,,,,,,,,,,,, +212400,5.025061,0.87524486,,,,,,,,,,,,,, +212500,5.1533237,0.9192015,,,,,,,,,,,,,, +212600,5.2352915,0.89203435,,,,,,,,,,,,,, +212700,4.9230843,0.9090395,,,,,,,,,,,,,, +212800,5.1715546,0.97210497,,,,,,,,,,,,,, +212900,5.6140428,1.0218351,,,,,,,,,,,,,, +213000,5.3892365,0.9390974,,,,,,,,,,,,,, +213100,5.2904835,0.92497176,,,,,,,,,,,,,, +213200,5.2461147,0.9054976,,,,,,,,,,,,,, +213300,5.0451307,0.9203429,,,,,,,,,,,,,, +213400,5.138781,0.9174758,,,,,,,,,,,,,, +213500,5.1516633,0.8971946,,,,,,,,,,,,,, +213600,5.1948037,0.847513,,,,,,,,,,,,,, +213700,5.5992546,0.9677801,,,,,,,,,,,,,, +213800,5.409973,1.0112439,,,,,,,,,,,,,, +213900,5.297396,0.94353426,,,,,,,,,,,,,, +214000,5.1148205,0.84335357,,,,,,,,,,,,,, +214100,5.1926603,0.9423616,,,,,,,,,,,,,, +214153,,,0.95475924,0.15814532,0.755,1.0817531,50000.0,0.6273,1.8760587,10000.0,67914.09925222397,68851.28239226341,67914.09925222397,926.6902208328248,6.9625163078308105,0.0 +214200,5.2917466,0.9927454,,,,,,,,,,,,,, +214300,5.6022754,1.0039091,,,,,,,,,,,,,, +214400,5.1976666,0.93714803,,,,,,,,,,,,,, +214500,5.1773973,0.89852417,,,,,,,,,,,,,, +214600,5.1690435,0.88738334,,,,,,,,,,,,,, +214700,4.870616,0.9196385,,,,,,,,,,,,,, +214800,5.2861586,0.9480227,,,,,,,,,,,,,, +214900,5.2074304,0.949932,,,,,,,,,,,,,, +215000,4.9909267,0.82817924,,,,,,,,,,,,,, +215100,5.1498685,0.93578124,,,,,,,,,,,,,, +215200,5.2484097,0.96252596,,,,,,,,,,,,,, +215300,5.1366496,0.8769013,,,,,,,,,,,,,, +215400,4.964019,1.0062419,,,,,,,,,,,,,, +215500,5.2173567,0.9428893,,,,,,,,,,,,,, +215600,4.984161,0.9021591,,,,,,,,,,,,,, +215700,4.7887807,0.8354585,,,,,,,,,,,,,, +215800,5.3459225,0.9986333,,,,,,,,,,,,,, +215900,5.213535,0.968246,,,,,,,,,,,,,, +216000,5.1514626,0.87382996,,,,,,,,,,,,,, +216100,5.0868134,0.9186927,,,,,,,,,,,,,, +216200,4.8486238,0.87616646,,,,,,,,,,,,,, +216300,5.241093,0.9542791,,,,,,,,,,,,,, +216400,5.071016,0.91855276,,,,,,,,,,,,,, +216500,4.9845977,0.8720291,,,,,,,,,,,,,, +216600,5.314584,0.9751673,,,,,,,,,,,,,, +216700,5.4660068,0.9345519,,,,,,,,,,,,,, +216800,4.942785,0.89656883,,,,,,,,,,,,,, +216900,5.5095687,0.95720077,,,,,,,,,,,,,, +217000,5.091161,0.8252367,,,,,,,,,,,,,, +217100,5.339502,0.99270403,,,,,,,,,,,,,, +217200,5.141699,0.8244808,,,,,,,,,,,,,, +217300,5.125891,0.8821708,,,,,,,,,,,,,, +217400,5.2151203,0.9937551,,,,,,,,,,,,,, +217500,5.033257,0.9200014,,,,,,,,,,,,,, +217600,5.2486687,0.91758037,,,,,,,,,,,,,, +217700,5.173307,0.9010352,,,,,,,,,,,,,, +217800,5.019109,0.8735447,,,,,,,,,,,,,, +217900,5.224678,0.972515,,,,,,,,,,,,,, +218000,4.9832873,0.83363205,,,,,,,,,,,,,, +218100,5.2645698,0.9488639,,,,,,,,,,,,,, +218200,5.136441,0.9232617,,,,,,,,,,,,,, +218254,,,0.9552176,0.1577096,0.75531995,1.0797553,50000.0,0.6268,1.8753902,10000.0,69909.51558828354,70886.89947223663,69909.51558828354,966.0391397476196,7.543886661529541,0.0 +218300,5.4268985,0.94647425,,,,,,,,,,,,,, +218400,5.343848,0.97700936,,,,,,,,,,,,,, +218500,4.8014956,0.8362555,,,,,,,,,,,,,, +218600,4.985343,0.8454726,,,,,,,,,,,,,, +218700,5.2321396,0.9690249,,,,,,,,,,,,,, +218800,5.3343315,0.9308944,,,,,,,,,,,,,, +218900,4.9964848,0.85133684,,,,,,,,,,,,,, +219000,4.888138,0.86753213,,,,,,,,,,,,,, +219100,4.959918,0.87138593,,,,,,,,,,,,,, +219200,5.2617183,0.9448763,,,,,,,,,,,,,, +219300,5.0698447,0.9205812,,,,,,,,,,,,,, +219400,5.3723154,0.9106943,,,,,,,,,,,,,, +219500,4.9379086,0.83488,,,,,,,,,,,,,, +219600,4.8095202,0.8052026,,,,,,,,,,,,,, +219700,5.252196,0.9424095,,,,,,,,,,,,,, +219800,5.2140045,0.9356333,,,,,,,,,,,,,, +219900,4.871048,0.85361814,,,,,,,,,,,,,, +220000,5.3911495,0.90551126,,,,,,,,,,,,,, +220100,5.3659606,0.98565114,,,,,,,,,,,,,, +220200,5.1067834,0.94768775,,,,,,,,,,,,,, +220300,5.1661787,0.8866359,,,,,,,,,,,,,, +220400,5.1295385,0.816411,,,,,,,,,,,,,, +220500,5.283845,0.9678706,,,,,,,,,,,,,, +220600,5.3216476,0.9417847,,,,,,,,,,,,,, +220700,5.230552,0.9197606,,,,,,,,,,,,,, +220800,4.8966675,0.85373527,,,,,,,,,,,,,, +220900,5.252561,0.92070806,,,,,,,,,,,,,, +221000,4.995283,0.9387848,,,,,,,,,,,,,, +221100,5.125157,0.87103236,,,,,,,,,,,,,, +221200,5.203973,0.9590159,,,,,,,,,,,,,, +221300,5.342173,0.968451,,,,,,,,,,,,,, +221400,5.202019,0.91973424,,,,,,,,,,,,,, +221500,5.204736,0.9113171,,,,,,,,,,,,,, +221600,5.098733,0.86884457,,,,,,,,,,,,,, +221700,5.2412615,0.9510134,,,,,,,,,,,,,, +221800,5.0428386,0.79964787,,,,,,,,,,,,,, +221900,4.9967546,0.8620137,,,,,,,,,,,,,, +222000,5.202873,0.90192646,,,,,,,,,,,,,, +222100,5.173943,0.9671159,,,,,,,,,,,,,, +222200,5.0981956,0.883572,,,,,,,,,,,,,, +222300,5.321023,0.94663143,,,,,,,,,,,,,, +222400,5.031403,0.8883345,,,,,,,,,,,,,, +222500,5.149908,0.9178548,,,,,,,,,,,,,, +222600,5.175168,0.9811259,,,,,,,,,,,,,, +222700,5.2673497,0.9873952,,,,,,,,,,,,,, +222739,,,0.9566725,0.15274695,0.75558,1.0793678,50000.0,0.62960005,1.8752167,10000.0,71905.1672899723,72943.8052880764,71905.1672899723,1026.6069040298462,8.02806806564331,0.0 +222800,5.141963,0.9152887,,,,,,,,,,,,,, +222900,5.4617386,0.93590474,,,,,,,,,,,,,, +223000,5.5231447,1.0342784,,,,,,,,,,,,,, +223100,5.295791,0.8724558,,,,,,,,,,,,,, +223200,5.2390695,0.98032534,,,,,,,,,,,,,, +223300,5.258751,0.9197937,,,,,,,,,,,,,, +223400,5.413966,0.8461739,,,,,,,,,,,,,, +223500,5.1575365,0.8890392,,,,,,,,,,,,,, +223600,5.0507317,0.7874511,,,,,,,,,,,,,, +223700,5.3357506,0.9271593,,,,,,,,,,,,,, +223800,5.2322583,0.8846102,,,,,,,,,,,,,, +223900,4.881684,0.8007798,,,,,,,,,,,,,, +224000,5.07036,0.9312651,,,,,,,,,,,,,, +224100,5.3233924,0.90660787,,,,,,,,,,,,,, +224200,5.2659435,0.8690326,,,,,,,,,,,,,, +224300,4.8875666,0.874951,,,,,,,,,,,,,, +224400,5.4927917,0.95420754,,,,,,,,,,,,,, +224500,5.074966,0.9694768,,,,,,,,,,,,,, +224600,5.061837,0.8681415,,,,,,,,,,,,,, +224700,4.8266916,0.9221982,,,,,,,,,,,,,, +224800,5.0937104,0.8912603,,,,,,,,,,,,,, +224900,5.3226247,0.9293591,,,,,,,,,,,,,, +225000,5.0970793,0.91157293,,,,,,,,,,,,,, +225100,5.352235,0.87344575,,,,,,,,,,,,,, +225200,5.176937,0.9769225,,,,,,,,,,,,,, +225300,5.121141,0.8887336,,,,,,,,,,,,,, +225400,5.0264564,0.83654314,,,,,,,,,,,,,, +225500,5.1422844,0.8570677,,,,,,,,,,,,,, +225600,4.682515,0.8323725,,,,,,,,,,,,,, +225700,5.272064,0.9946851,,,,,,,,,,,,,, +225800,5.2402496,0.95334065,,,,,,,,,,,,,, +225900,4.85673,0.8267614,,,,,,,,,,,,,, +226000,5.1493626,0.848969,,,,,,,,,,,,,, +226100,5.2821727,0.89263916,,,,,,,,,,,,,, +226200,4.9914985,0.92471546,,,,,,,,,,,,,, +226300,4.9621434,0.87933826,,,,,,,,,,,,,, +226400,5.3296995,0.9033814,,,,,,,,,,,,,, +226500,4.7214913,0.82227784,,,,,,,,,,,,,, +226600,5.4829183,0.9126917,,,,,,,,,,,,,, +226700,5.2204704,0.92302835,,,,,,,,,,,,,, +226800,5.1331015,0.8413694,,,,,,,,,,,,,, +226900,5.2668324,0.9465717,,,,,,,,,,,,,, +227000,5.1211357,0.90181327,,,,,,,,,,,,,, +227100,4.928009,0.80934536,,,,,,,,,,,,,, +227200,5.2099333,0.9842455,,,,,,,,,,,,,, +227300,5.702542,0.96097934,,,,,,,,,,,,,, +227400,5.278942,0.8967426,,,,,,,,,,,,,, +227500,5.137077,0.8735411,,,,,,,,,,,,,, +227600,5.0428905,0.9307991,,,,,,,,,,,,,, +227700,4.9755206,0.8814067,,,,,,,,,,,,,, +227800,5.2408104,0.92633194,,,,,,,,,,,,,, +227900,4.9232216,0.85464,,,,,,,,,,,,,, +228000,5.2129774,0.9212754,,,,,,,,,,,,,, +228100,4.9102592,0.8728088,,,,,,,,,,,,,, +228137,,,0.9572704,0.15046822,0.75606,1.0797607,50000.0,0.62750006,1.8712056,10000.0,73900.43143868446,74975.94643807411,73900.43143868446,1062.7228429317474,8.600809574127197,0.0 +228200,5.1091957,0.8472911,,,,,,,,,,,,,, +228300,5.1608415,0.8259518,,,,,,,,,,,,,, +228400,5.53806,0.88425875,,,,,,,,,,,,,, +228500,5.0425005,0.8998209,,,,,,,,,,,,,, +228600,5.172793,0.8756397,,,,,,,,,,,,,, +228700,5.0258646,0.92053074,,,,,,,,,,,,,, +228800,5.226393,0.89777696,,,,,,,,,,,,,, +228900,5.2634997,0.92194855,,,,,,,,,,,,,, +229000,5.470121,1.0005717,,,,,,,,,,,,,, +229100,5.2195783,0.9156084,,,,,,,,,,,,,, +229200,5.2831693,0.8812978,,,,,,,,,,,,,, +229300,5.058711,0.88246477,,,,,,,,,,,,,, +229400,5.296087,0.94772744,,,,,,,,,,,,,, +229500,4.9236646,0.8533449,,,,,,,,,,,,,, +229600,5.516074,0.92220116,,,,,,,,,,,,,, +229700,5.2705626,0.9427632,,,,,,,,,,,,,, +229800,5.4425116,0.88754857,,,,,,,,,,,,,, +229900,5.295756,0.92271644,,,,,,,,,,,,,, +230000,5.1278286,0.8649423,,,,,,,,,,,,,, +230100,5.3026915,0.8674869,,,,,,,,,,,,,, +230200,5.287762,0.9079925,,,,,,,,,,,,,, +230300,5.022534,0.88804805,,,,,,,,,,,,,, +230400,5.0455403,0.8570906,,,,,,,,,,,,,, +230500,5.3729205,0.91844285,,,,,,,,,,,,,, +230600,5.5879207,0.9265112,,,,,,,,,,,,,, +230700,5.2360554,1.0084689,,,,,,,,,,,,,, +230800,5.7318,0.9543077,,,,,,,,,,,,,, +230900,4.967644,0.8564641,,,,,,,,,,,,,, +231000,5.014541,0.8100097,,,,,,,,,,,,,, +231100,5.3106923,0.9450332,,,,,,,,,,,,,, +231200,4.8723235,0.8590765,,,,,,,,,,,,,, +231300,4.9945297,0.8368402,,,,,,,,,,,,,, +231400,5.3622727,0.91525435,,,,,,,,,,,,,, +231500,5.3965096,0.93379396,,,,,,,,,,,,,, +231600,5.255953,0.899748,,,,,,,,,,,,,, +231700,4.880582,0.86575806,,,,,,,,,,,,,, +231800,5.106797,0.8131549,,,,,,,,,,,,,, +231900,5.183786,0.88945377,,,,,,,,,,,,,, +232000,5.1100135,0.88869536,,,,,,,,,,,,,, +232100,5.211853,1.00143,,,,,,,,,,,,,, +232200,5.416667,0.92512167,,,,,,,,,,,,,, +232300,5.3013425,0.84346706,,,,,,,,,,,,,, +232400,5.119885,0.87672114,,,,,,,,,,,,,, +232500,5.257817,0.8869935,,,,,,,,,,,,,, +232600,5.246023,0.95179915,,,,,,,,,,,,,, +232700,5.0456595,0.89704335,,,,,,,,,,,,,, +232800,5.2052245,0.9111097,,,,,,,,,,,,,, +232900,5.3666906,0.8691924,,,,,,,,,,,,,, +233000,5.299233,0.87305593,,,,,,,,,,,,,, +233100,5.417654,0.9946203,,,,,,,,,,,,,, +233200,5.1496534,0.9320715,,,,,,,,,,,,,, +233300,5.062997,0.8981925,,,,,,,,,,,,,, +233400,5.3445034,0.94381523,,,,,,,,,,,,,, +233500,5.3762116,0.952442,,,,,,,,,,,,,, +233600,5.2252254,0.8655857,,,,,,,,,,,,,, +233700,5.041961,0.9015316,,,,,,,,,,,,,, +233800,5.2122483,0.8341057,,,,,,,,,,,,,, +233900,5.126074,0.8901229,,,,,,,,,,,,,, +234000,5.7034073,1.0235174,,,,,,,,,,,,,, +234100,4.9448104,0.8790635,,,,,,,,,,,,,, +234169,,,,,,,,,,,75895.78674530983,,,,,0.0 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/meta_data_0.json b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/meta_data_0.json index 788c68ac7..e70877f80 100644 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/meta_data_0.json +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_0/imagenet_resnet_jax/trial_1/meta_data_0.json @@ -15,35 +15,35 @@ "workload.use_gelu": false, "workload.use_silu": false, "workload.validation_target_value": 0.77431, - "cpu.util.avg_percent_since_last": 2.5, - "cpu.freq.current": 2200.161999999999, + "cpu.util.avg_percent_since_last": 4.8, + "cpu.freq.current": 2200.207999999999, "mem.total": 359053524992, - "mem.available": 348489388032, - "mem.used": 7315013632, + "mem.available": 348510711808, + "mem.used": 7293693952, "mem.percent_used": 2.9, - "mem.read_bytes_since_boot": 6104005176832, - "mem.write_bytes_since_boot": 42550642176, - "net.bytes_sent_since_boot": 15405, - "net.bytes_recv_since_boot": 707726, + "mem.read_bytes_since_boot": 1128152576, + "mem.write_bytes_since_boot": 26312869376, + "net.bytes_sent_since_boot": 14613, + "net.bytes_recv_since_boot": 739476, "gpu.count": 4, "gpu.0.compute.util": 0.01, "gpu.0.mem.util": 0.7490478515625, "gpu.0.mem.total": 40960.0, "gpu.0.mem.used": 30681.0, "gpu.0.mem.free": 9646.0, - "gpu.0.temp.current": 33.0, + "gpu.0.temp.current": 35.0, "gpu.1.compute.util": 0.0, "gpu.1.mem.util": 0.7488525390625, "gpu.1.mem.total": 40960.0, "gpu.1.mem.used": 30673.0, "gpu.1.mem.free": 9654.0, - "gpu.1.temp.current": 31.0, + "gpu.1.temp.current": 33.0, "gpu.2.compute.util": 0.0, "gpu.2.mem.util": 0.7488525390625, "gpu.2.mem.total": 40960.0, "gpu.2.mem.used": 30673.0, "gpu.2.mem.free": 9654.0, - "gpu.2.temp.current": 32.0, + "gpu.2.temp.current": 34.0, "gpu.3.compute.util": 0.0, "gpu.3.mem.util": 0.7488525390625, "gpu.3.mem.total": 40960.0, @@ -55,7 +55,7 @@ "gpu.avg.mem.total": 40960.0, "gpu.avg.mem.used": 30675.0, "gpu.avg.mem.free": 9652.0, - "gpu.avg.temp.current": 32.5, + "gpu.avg.temp.current": 34.0, "os_platform": "Linux-6.1.0-44-cloud-amd64-x86_64-with-glibc2.31", "python_version": "3.11.10", "python_compiler": "GCC 9.4.0", diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/imagenet_resnet_jax_05-12-2026-19-45-09.log b/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/imagenet_resnet_jax_05-12-2026-19-45-09.log deleted file mode 100644 index 4cf28a3dd..000000000 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/imagenet_resnet_jax_05-12-2026-19-45-09.log +++ /dev/null @@ -1,1766 +0,0 @@ -python submission_runner.py --framework=jax --workload=imagenet_resnet --submission_path=submissions_algorithms/submissions/self_tuning/schedule_free_adamw_jax_v2/submission.py --data_dir=/data/imagenet/jax --experiment_dir=/experiment_runs --experiment_name=submissions_a100/schedule_free_adamw_jax_v2/study_1 --overwrite=True --save_checkpoints=False --rng_seed=-304980929 --imagenet_v2_data_dir=/data/imagenet/jax --tuning_ruleset=self 2>&1 | tee -a /logs/imagenet_resnet_jax_05-12-2026-19-45-09.log -2026-05-12 19:45:19.727645: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered -WARNING: All log messages before absl::InitializeLog() is called are written to STDERR -E0000 00:00:1778615120.149914 13 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered -E0000 00:00:1778615120.266543 13 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered -W0000 00:00:1778615121.715225 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -W0000 00:00:1778615121.715269 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -W0000 00:00:1778615121.715272 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -W0000 00:00:1778615121.715274 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -/usr/local/lib/python3.11/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.) - _C._set_float32_matmul_precision(precision) -/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'repr' attribute with value False was provided to the `Field()` function, which has no effect in the context it was used. 'repr' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. - warnings.warn( -/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'frozen' attribute with value True was provided to the `Field()` function, which has no effect in the context it was used. 'frozen' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. - warnings.warn( -INFO:2026-05-12 19:46:11,614:jax._src.xla_bridge:830: Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory -I0512 19:46:11.614608 140678261474496 xla_bridge.py:830] Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory -I0512 19:46:12.418826 140678261474496 logger_utils.py:84] Creating experiment directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax. -I0512 19:46:15.816436 140678261474496 submission_runner.py:741] Creating directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1. -I0512 19:46:16.177634 140678261474496 submission_runner.py:242] Initializing dataset. -I0512 19:46:17.446465 140678261474496 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:46:17.524975 140678261474496 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. -I0512 19:46:17.886300 140678261474496 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. -I0512 19:46:18.231502 140678261474496 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:46:20.128663 140678261474496 submission_runner.py:251] Initializing model. -I0512 19:46:45.178234 140678261474496 submission_runner.py:294] Initializing optimizer. -I0512 19:46:46.964578 140678261474496 submission_runner.py:299] Initializing metrics bundle. -I0512 19:46:46.964801 140678261474496 submission_runner.py:321] Initializing checkpoint and logger. -I0512 19:46:46.967656 140678261474496 checkpoints.py:1098] Found no checkpoint files in /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1 with prefix checkpoint_ -I0512 19:46:46.967789 140678261474496 submission_runner.py:344] Saving meta data to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/meta_data_0.json. -I0512 19:46:47.499606 140678261474496 submission_runner.py:348] Saving flags to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/flags_0.json. -I0512 19:46:47.510744 140678261474496 submission_runner.py:359] Starting training loop. -/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. -See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. - warnings.warn("Some donated buffers were not usable:" -I0512 19:47:46.813019 140663127500544 logging_writer.py:48] [0] global_step=0, grad_norm=0.31780773401260376, loss=6.910696029663086 -I0512 19:47:48.430732 140678261474496 spec.py:333] Evaluating on the training split. -I0512 19:47:48.722782 140678261474496 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:47:48.728329 140678261474496 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. -I0512 19:47:48.746811 140678261474496 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. -I0512 19:47:48.790859 140678261474496 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:48:27.196235 140678261474496 spec.py:346] Evaluating on the validation split. -I0512 19:48:27.206236 140678261474496 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:48:27.223032 140678261474496 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. -I0512 19:48:27.227835 140678261474496 reader.py:262] Creating a tf.data.Dataset reading 64 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. -I0512 19:48:27.500035 140678261474496 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split validation, from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 19:48:59.792788 140678261474496 spec.py:363] Evaluating on the test split. -I0512 19:48:59.914207 140678261474496 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 -I0512 19:48:59.953375 140678261474496 reader.py:262] Creating a tf.data.Dataset reading 16 files located in folders: /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0. -I0512 19:48:59.990376 140678261474496 logging_logger.py:49] Constructing tf.data.Dataset imagenet_v2 for split test, from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 -I0512 19:49:16.987286 140678261474496 submission_runner.py:516] Time since start: 149.47s, Step: 1, {'train/accuracy': Array(0.00187341, dtype=float32), 'train/loss': Array(6.911738, dtype=float32), 'validation/accuracy': Array(0.00126, dtype=float32), 'validation/loss': Array(6.9117756, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(6.9119825, dtype=float32), 'test/num_examples': 10000, 'score': 60.91907024383545, 'total_duration': 149.47478890419006, 'accumulated_submission_time': 60.91907024383545, 'accumulated_eval_time': 88.55480742454529, 'accumulated_logging_time': 0} -I0512 19:49:17.007068 140460890760960 logging_writer.py:48] [1] accumulated_eval_time=88.5548, accumulated_logging_time=0, accumulated_submission_time=60.9191, global_step=1, preemption_count=0, score=60.9191, test/accuracy=0.0012000000569969416, test/loss=6.911982536315918, test/num_examples=10000, total_duration=149.475, train/accuracy=0.0018734055338427424, train/loss=6.91173791885376, validation/accuracy=0.0012599999317899346, validation/loss=6.911775588989258, validation/num_examples=50000 -/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. -See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. - warnings.warn("Some donated buffers were not usable:" -I0512 19:50:22.446625 140460773328640 logging_writer.py:48] [100] global_step=100, grad_norm=0.9209260940551758, loss=6.758374214172363 -I0512 19:51:09.843364 140460781721344 logging_writer.py:48] [200] global_step=200, grad_norm=1.7862359285354614, loss=6.402630805969238 -I0512 19:51:51.097754 140460773328640 logging_writer.py:48] [300] global_step=300, grad_norm=1.9424149990081787, loss=6.294833183288574 -I0512 19:52:30.977406 140460781721344 logging_writer.py:48] [400] global_step=400, grad_norm=1.5394320487976074, loss=6.158416748046875 -I0512 19:53:30.319487 140460773328640 logging_writer.py:48] [500] global_step=500, grad_norm=2.153167486190796, loss=6.1391377449035645 -I0512 19:54:12.028908 140460781721344 logging_writer.py:48] [600] global_step=600, grad_norm=2.1498806476593018, loss=6.02904748916626 -I0512 19:54:54.120903 140460773328640 logging_writer.py:48] [700] global_step=700, grad_norm=2.6038315296173096, loss=5.998392105102539 -I0512 19:55:38.368784 140460781721344 logging_writer.py:48] [800] global_step=800, grad_norm=4.8983354568481445, loss=5.9576263427734375 -I0512 19:56:35.090499 140460773328640 logging_writer.py:48] [900] global_step=900, grad_norm=2.462722063064575, loss=5.8905863761901855 -I0512 19:57:15.276333 140460781721344 logging_writer.py:48] [1000] global_step=1000, grad_norm=3.1774728298187256, loss=5.876762390136719 -I0512 19:57:55.068993 140460773328640 logging_writer.py:48] [1100] global_step=1100, grad_norm=4.397616863250732, loss=5.933290004730225 -I0512 19:58:50.533303 140460781721344 logging_writer.py:48] [1200] global_step=1200, grad_norm=4.786351203918457, loss=5.821152210235596 -I0512 19:59:36.095281 140460773328640 logging_writer.py:48] [1300] global_step=1300, grad_norm=4.235178470611572, loss=5.871098518371582 -I0512 20:00:29.993118 140460781721344 logging_writer.py:48] [1400] global_step=1400, grad_norm=2.5822339057922363, loss=5.8124895095825195 -I0512 20:01:43.714264 140460773328640 logging_writer.py:48] [1500] global_step=1500, grad_norm=2.3210041522979736, loss=5.782279968261719 -I0512 20:03:02.494947 140460781721344 logging_writer.py:48] [1600] global_step=1600, grad_norm=5.4346699714660645, loss=5.693648338317871 -I0512 20:04:20.133145 140460773328640 logging_writer.py:48] [1700] global_step=1700, grad_norm=1.7878713607788086, loss=5.612153053283691 -I0512 20:05:27.407155 140460781721344 logging_writer.py:48] [1800] global_step=1800, grad_norm=1.2357970476150513, loss=5.6849493980407715 -I0512 20:06:41.016298 140460773328640 logging_writer.py:48] [1900] global_step=1900, grad_norm=2.2307262420654297, loss=5.5674543380737305 -I0512 20:07:53.682655 140460781721344 logging_writer.py:48] [2000] global_step=2000, grad_norm=2.0173752307891846, loss=5.630041122436523 -I0512 20:08:59.642420 140460773328640 logging_writer.py:48] [2100] global_step=2100, grad_norm=4.279491424560547, loss=5.73594856262207 -I0512 20:09:51.068983 140460781721344 logging_writer.py:48] [2200] global_step=2200, grad_norm=7.170375347137451, loss=5.846158027648926 -I0512 20:10:34.999115 140460773328640 logging_writer.py:48] [2300] global_step=2300, grad_norm=1.9347198009490967, loss=5.542311191558838 -I0512 20:11:19.104545 140460781721344 logging_writer.py:48] [2400] global_step=2400, grad_norm=2.947378635406494, loss=5.5498762130737305 -I0512 20:12:10.687939 140460773328640 logging_writer.py:48] [2500] global_step=2500, grad_norm=6.011758804321289, loss=5.644463062286377 -I0512 20:13:10.330129 140460781721344 logging_writer.py:48] [2600] global_step=2600, grad_norm=2.896782398223877, loss=5.500473499298096 -I0512 20:14:07.173732 140460773328640 logging_writer.py:48] [2700] global_step=2700, grad_norm=2.2719264030456543, loss=5.529545783996582 -I0512 20:15:05.385360 140460781721344 logging_writer.py:48] [2800] global_step=2800, grad_norm=1.6686444282531738, loss=5.3805646896362305 -I0512 20:15:51.664134 140460773328640 logging_writer.py:48] [2900] global_step=2900, grad_norm=2.6559486389160156, loss=5.471421241760254 -I0512 20:16:46.916864 140460781721344 logging_writer.py:48] [3000] global_step=3000, grad_norm=3.9281198978424072, loss=5.513998508453369 -I0512 20:17:45.331166 140460773328640 logging_writer.py:48] [3100] global_step=3100, grad_norm=2.740149736404419, loss=5.505005836486816 -I0512 20:18:36.397890 140460781721344 logging_writer.py:48] [3200] global_step=3200, grad_norm=3.8676249980926514, loss=5.419878005981445 -I0512 20:19:21.087880 140460773328640 logging_writer.py:48] [3300] global_step=3300, grad_norm=4.024202346801758, loss=5.407713890075684 -I0512 20:20:11.461849 140460781721344 logging_writer.py:48] [3400] global_step=3400, grad_norm=3.1994333267211914, loss=5.26511287689209 -I0512 20:21:05.916535 140460773328640 logging_writer.py:48] [3500] global_step=3500, grad_norm=6.510839939117432, loss=5.661543846130371 -I0512 20:22:01.072032 140460781721344 logging_writer.py:48] [3600] global_step=3600, grad_norm=2.931042194366455, loss=5.282142639160156 -I0512 20:22:34.323236 140678261474496 spec.py:333] Evaluating on the training split. -I0512 20:22:49.286610 140678261474496 spec.py:346] Evaluating on the validation split. -I0512 20:23:28.056051 140678261474496 spec.py:363] Evaluating on the test split. -I0512 20:23:29.037368 140678261474496 submission_runner.py:516] Time since start: 2201.52s, Step: 3667, {'train/accuracy': Array(0.04308833, dtype=float32), 'train/loss': Array(5.8607073, dtype=float32), 'validation/accuracy': Array(0.04014, dtype=float32), 'validation/loss': Array(5.9290338, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0299, dtype=float32), 'test/loss': Array(6.208869, dtype=float32), 'test/num_examples': 10000, 'score': 2058.181893348694, 'total_duration': 2201.5248374938965, 'accumulated_submission_time': 2058.181893348694, 'accumulated_eval_time': 143.2671549320221, 'accumulated_logging_time': 0.028658628463745117} -I0512 20:23:29.061352 140460899153664 logging_writer.py:48] [3667] accumulated_eval_time=143.267, accumulated_logging_time=0.0286586, accumulated_submission_time=2058.18, global_step=3667, preemption_count=0, score=2058.18, test/accuracy=0.029900001361966133, test/loss=6.208868980407715, test/num_examples=10000, total_duration=2201.52, train/accuracy=0.04308832809329033, train/loss=5.8607072830200195, validation/accuracy=0.040139999240636826, validation/loss=5.9290337562561035, validation/num_examples=50000 -I0512 20:23:58.407016 140460907546368 logging_writer.py:48] [3700] global_step=3700, grad_norm=2.330073595046997, loss=5.177009582519531 -I0512 20:24:55.705896 140460899153664 logging_writer.py:48] [3800] global_step=3800, grad_norm=2.7880139350891113, loss=5.211492538452148 -I0512 20:25:42.457170 140460907546368 logging_writer.py:48] [3900] global_step=3900, grad_norm=3.9142653942108154, loss=5.286319732666016 -I0512 20:26:20.636309 140460899153664 logging_writer.py:48] [4000] global_step=4000, grad_norm=3.070784568786621, loss=5.2371931076049805 -I0512 20:27:07.410858 140460907546368 logging_writer.py:48] [4100] global_step=4100, grad_norm=3.063202381134033, loss=5.085653305053711 -I0512 20:27:45.114641 140460899153664 logging_writer.py:48] [4200] global_step=4200, grad_norm=3.0917298793792725, loss=5.246720790863037 -I0512 20:28:22.397699 140460907546368 logging_writer.py:48] [4300] global_step=4300, grad_norm=2.002202033996582, loss=5.068938255310059 -I0512 20:29:00.005320 140460899153664 logging_writer.py:48] [4400] global_step=4400, grad_norm=2.3945631980895996, loss=5.029738426208496 -I0512 20:29:37.725781 140460907546368 logging_writer.py:48] [4500] global_step=4500, grad_norm=2.878852128982544, loss=4.95932149887085 -I0512 20:30:14.840414 140460899153664 logging_writer.py:48] [4600] global_step=4600, grad_norm=1.9406287670135498, loss=4.962429046630859 -I0512 20:30:52.006147 140460907546368 logging_writer.py:48] [4700] global_step=4700, grad_norm=4.393433094024658, loss=4.982895374298096 -I0512 20:31:30.113925 140460899153664 logging_writer.py:48] [4800] global_step=4800, grad_norm=4.191000461578369, loss=5.155977725982666 -I0512 20:32:07.317590 140460907546368 logging_writer.py:48] [4900] global_step=4900, grad_norm=4.022045135498047, loss=5.160102367401123 -I0512 20:32:44.857022 140460899153664 logging_writer.py:48] [5000] global_step=5000, grad_norm=2.2287185192108154, loss=5.080767631530762 -I0512 20:33:22.563528 140460907546368 logging_writer.py:48] [5100] global_step=5100, grad_norm=4.156472682952881, loss=5.010293960571289 -I0512 20:34:08.771589 140460899153664 logging_writer.py:48] [5200] global_step=5200, grad_norm=4.829965114593506, loss=5.078292369842529 -I0512 20:35:04.717835 140460907546368 logging_writer.py:48] [5300] global_step=5300, grad_norm=2.250473737716675, loss=4.905556678771973 -I0512 20:35:51.329630 140460899153664 logging_writer.py:48] [5400] global_step=5400, grad_norm=2.402311325073242, loss=4.819714546203613 -I0512 20:36:28.587434 140460907546368 logging_writer.py:48] [5500] global_step=5500, grad_norm=2.2166330814361572, loss=4.928332328796387 -I0512 20:37:06.265967 140460899153664 logging_writer.py:48] [5600] global_step=5600, grad_norm=2.725045919418335, loss=4.848555564880371 -I0512 20:37:44.098592 140460907546368 logging_writer.py:48] [5700] global_step=5700, grad_norm=4.15472936630249, loss=4.864901065826416 -I0512 20:38:21.322703 140460899153664 logging_writer.py:48] [5800] global_step=5800, grad_norm=3.7074079513549805, loss=4.971633434295654 -I0512 20:38:58.805812 140460907546368 logging_writer.py:48] [5900] global_step=5900, grad_norm=3.082153081893921, loss=4.920456886291504 -I0512 20:39:36.500574 140460899153664 logging_writer.py:48] [6000] global_step=6000, grad_norm=1.974237322807312, loss=4.887185096740723 -I0512 20:40:13.651970 140460907546368 logging_writer.py:48] [6100] global_step=6100, grad_norm=2.9673547744750977, loss=4.9324469566345215 -I0512 20:40:51.076911 140460899153664 logging_writer.py:48] [6200] global_step=6200, grad_norm=3.2579281330108643, loss=4.949971675872803 -I0512 20:41:54.883623 140460907546368 logging_writer.py:48] [6300] global_step=6300, grad_norm=3.8114044666290283, loss=4.880708694458008 -I0512 20:42:43.050990 140460899153664 logging_writer.py:48] [6400] global_step=6400, grad_norm=2.629997968673706, loss=4.829401016235352 -I0512 20:43:27.408040 140460907546368 logging_writer.py:48] [6500] global_step=6500, grad_norm=4.1177496910095215, loss=4.997490882873535 -I0512 20:44:07.111748 140460899153664 logging_writer.py:48] [6600] global_step=6600, grad_norm=3.34161376953125, loss=4.76191520690918 -I0512 20:44:44.291271 140460907546368 logging_writer.py:48] [6700] global_step=6700, grad_norm=3.428161144256592, loss=4.8025102615356445 -I0512 20:45:21.535966 140460899153664 logging_writer.py:48] [6800] global_step=6800, grad_norm=3.11376690864563, loss=4.894753932952881 -I0512 20:45:59.453462 140460907546368 logging_writer.py:48] [6900] global_step=6900, grad_norm=4.401673793792725, loss=4.996303558349609 -I0512 20:46:36.655530 140460899153664 logging_writer.py:48] [7000] global_step=7000, grad_norm=10.988923072814941, loss=5.095341205596924 -I0512 20:47:14.199062 140460907546368 logging_writer.py:48] [7100] global_step=7100, grad_norm=1.6486014127731323, loss=4.6057329177856445 -I0512 20:47:51.828799 140460899153664 logging_writer.py:48] [7200] global_step=7200, grad_norm=4.070767402648926, loss=4.662487030029297 -I0512 20:48:29.010292 140460907546368 logging_writer.py:48] [7300] global_step=7300, grad_norm=2.391486406326294, loss=4.6851935386657715 -I0512 20:49:06.119452 140460899153664 logging_writer.py:48] [7400] global_step=7400, grad_norm=3.828043222427368, loss=4.789220333099365 -I0512 20:49:52.923765 140460907546368 logging_writer.py:48] [7500] global_step=7500, grad_norm=2.748373031616211, loss=4.697661876678467 -I0512 20:50:38.808803 140460899153664 logging_writer.py:48] [7600] global_step=7600, grad_norm=3.6261303424835205, loss=4.704741477966309 -I0512 20:51:25.299695 140460907546368 logging_writer.py:48] [7700] global_step=7700, grad_norm=3.362413167953491, loss=4.805017471313477 -I0512 20:52:03.315537 140460899153664 logging_writer.py:48] [7800] global_step=7800, grad_norm=2.6272597312927246, loss=4.693347930908203 -I0512 20:52:40.478955 140460907546368 logging_writer.py:48] [7900] global_step=7900, grad_norm=4.023525714874268, loss=4.792562484741211 -I0512 20:53:17.640092 140460899153664 logging_writer.py:48] [8000] global_step=8000, grad_norm=4.13770055770874, loss=4.626368999481201 -I0512 20:53:55.586855 140460907546368 logging_writer.py:48] [8100] global_step=8100, grad_norm=2.4451162815093994, loss=4.569980144500732 -I0512 20:54:32.824386 140460899153664 logging_writer.py:48] [8200] global_step=8200, grad_norm=5.381283760070801, loss=4.753225326538086 -I0512 20:55:10.059433 140460907546368 logging_writer.py:48] [8300] global_step=8300, grad_norm=2.6548166275024414, loss=4.617557525634766 -I0512 20:55:47.950650 140460899153664 logging_writer.py:48] [8400] global_step=8400, grad_norm=5.110328197479248, loss=4.636364459991455 -I0512 20:56:25.089496 140460907546368 logging_writer.py:48] [8500] global_step=8500, grad_norm=4.3993377685546875, loss=4.670101642608643 -I0512 20:56:45.092706 140678261474496 spec.py:333] Evaluating on the training split. -I0512 20:56:59.851049 140678261474496 spec.py:346] Evaluating on the validation split. -I0512 20:57:22.906427 140678261474496 spec.py:363] Evaluating on the test split. -I0512 20:57:23.791894 140678261474496 submission_runner.py:516] Time since start: 4236.28s, Step: 8555, {'train/accuracy': Array(0.01749841, dtype=float32), 'train/loss': Array(6.8538775, dtype=float32), 'validation/accuracy': Array(0.01714, dtype=float32), 'validation/loss': Array(6.8836646, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0133, dtype=float32), 'test/loss': Array(7.0473847, dtype=float32), 'test/num_examples': 10000, 'score': 4054.160504579544, 'total_duration': 4236.279563426971, 'accumulated_submission_time': 4054.160504579544, 'accumulated_eval_time': 181.96476244926453, 'accumulated_logging_time': 0.06117844581604004} -I0512 20:57:23.813073 140460899153664 logging_writer.py:48] [8555] accumulated_eval_time=181.965, accumulated_logging_time=0.0611784, accumulated_submission_time=4054.16, global_step=8555, preemption_count=0, score=4054.16, test/accuracy=0.013300000689923763, test/loss=7.047384738922119, test/num_examples=10000, total_duration=4236.28, train/accuracy=0.017498405650258064, train/loss=6.853877544403076, validation/accuracy=0.017139999195933342, validation/loss=6.883664608001709, validation/num_examples=50000 -I0512 20:57:40.836336 140460907546368 logging_writer.py:48] [8600] global_step=8600, grad_norm=3.9190006256103516, loss=4.615328788757324 -I0512 20:58:18.864657 140460899153664 logging_writer.py:48] [8700] global_step=8700, grad_norm=4.807989120483398, loss=4.6643524169921875 -I0512 20:58:55.872211 140460907546368 logging_writer.py:48] [8800] global_step=8800, grad_norm=4.05600643157959, loss=4.66378116607666 -I0512 20:59:32.981278 140460899153664 logging_writer.py:48] [8900] global_step=8900, grad_norm=2.3713645935058594, loss=4.560507774353027 -I0512 21:00:10.742297 140460907546368 logging_writer.py:48] [9000] global_step=9000, grad_norm=3.131056308746338, loss=4.735022068023682 -I0512 21:00:47.897166 140460899153664 logging_writer.py:48] [9100] global_step=9100, grad_norm=2.0196094512939453, loss=4.600934028625488 -I0512 21:01:24.986332 140460907546368 logging_writer.py:48] [9200] global_step=9200, grad_norm=4.5948076248168945, loss=4.711893558502197 -I0512 21:02:02.869709 140460899153664 logging_writer.py:48] [9300] global_step=9300, grad_norm=4.440432071685791, loss=4.558143615722656 -I0512 21:02:40.011691 140460907546368 logging_writer.py:48] [9400] global_step=9400, grad_norm=4.095041275024414, loss=4.547077178955078 -I0512 21:03:17.185255 140460899153664 logging_writer.py:48] [9500] global_step=9500, grad_norm=6.096622943878174, loss=4.59396505355835 -I0512 21:03:54.685740 140460907546368 logging_writer.py:48] [9600] global_step=9600, grad_norm=4.022310733795166, loss=4.5412702560424805 -I0512 21:04:32.058819 140460899153664 logging_writer.py:48] [9700] global_step=9700, grad_norm=6.654725074768066, loss=4.493906021118164 -I0512 21:05:09.202158 140460907546368 logging_writer.py:48] [9800] global_step=9800, grad_norm=3.048213005065918, loss=4.515395164489746 -I0512 21:05:47.051781 140460899153664 logging_writer.py:48] [9900] global_step=9900, grad_norm=7.127708911895752, loss=4.562563896179199 -I0512 21:06:24.045254 140460907546368 logging_writer.py:48] [10000] global_step=10000, grad_norm=4.984849452972412, loss=4.562511444091797 -I0512 21:07:01.096151 140460899153664 logging_writer.py:48] [10100] global_step=10100, grad_norm=4.010753154754639, loss=4.462884426116943 -I0512 21:07:39.039536 140460907546368 logging_writer.py:48] [10200] global_step=10200, grad_norm=4.357707977294922, loss=4.653226852416992 -I0512 21:08:16.147528 140460899153664 logging_writer.py:48] [10300] global_step=10300, grad_norm=4.799588680267334, loss=4.47319221496582 -I0512 21:08:53.234997 140460907546368 logging_writer.py:48] [10400] global_step=10400, grad_norm=4.516970157623291, loss=4.418220043182373 -I0512 21:09:31.128842 140460899153664 logging_writer.py:48] [10500] global_step=10500, grad_norm=7.413406848907471, loss=4.596695899963379 -I0512 21:10:08.194634 140460907546368 logging_writer.py:48] [10600] global_step=10600, grad_norm=3.730695962905884, loss=4.499547958374023 -I0512 21:10:45.276914 140460899153664 logging_writer.py:48] [10700] global_step=10700, grad_norm=4.48500919342041, loss=4.31428337097168 -I0512 21:11:23.181513 140460907546368 logging_writer.py:48] [10800] global_step=10800, grad_norm=4.910727024078369, loss=4.581844329833984 -I0512 21:12:00.165488 140460899153664 logging_writer.py:48] [10900] global_step=10900, grad_norm=7.567192554473877, loss=4.563248634338379 -I0512 21:12:37.299735 140460907546368 logging_writer.py:48] [11000] global_step=11000, grad_norm=4.8110432624816895, loss=4.488776206970215 -I0512 21:13:15.271070 140460899153664 logging_writer.py:48] [11100] global_step=11100, grad_norm=3.3063619136810303, loss=4.360965251922607 -I0512 21:13:52.344080 140460907546368 logging_writer.py:48] [11200] global_step=11200, grad_norm=4.44964075088501, loss=4.458347320556641 -I0512 21:14:29.402105 140460899153664 logging_writer.py:48] [11300] global_step=11300, grad_norm=5.923408508300781, loss=4.503227710723877 -I0512 21:15:07.250269 140460907546368 logging_writer.py:48] [11400] global_step=11400, grad_norm=3.808549165725708, loss=4.454941749572754 -I0512 21:15:44.378232 140460899153664 logging_writer.py:48] [11500] global_step=11500, grad_norm=5.5770487785339355, loss=4.598136901855469 -I0512 21:16:21.425731 140460907546368 logging_writer.py:48] [11600] global_step=11600, grad_norm=10.108150482177734, loss=4.594392776489258 -I0512 21:16:58.930430 140460899153664 logging_writer.py:48] [11700] global_step=11700, grad_norm=2.535419225692749, loss=4.322304725646973 -I0512 21:17:36.409460 140460907546368 logging_writer.py:48] [11800] global_step=11800, grad_norm=5.014636516571045, loss=4.367799758911133 -I0512 21:18:13.554445 140460899153664 logging_writer.py:48] [11900] global_step=11900, grad_norm=11.726548194885254, loss=4.468168258666992 -I0512 21:18:51.553725 140460907546368 logging_writer.py:48] [12000] global_step=12000, grad_norm=3.943453073501587, loss=4.463531494140625 -I0512 21:19:28.582507 140460899153664 logging_writer.py:48] [12100] global_step=12100, grad_norm=5.380733489990234, loss=4.319850444793701 -I0512 21:20:05.693182 140460907546368 logging_writer.py:48] [12200] global_step=12200, grad_norm=3.4766383171081543, loss=4.387754440307617 -I0512 21:20:43.656437 140460899153664 logging_writer.py:48] [12300] global_step=12300, grad_norm=10.568528175354004, loss=4.267388343811035 -I0512 21:21:20.721928 140460907546368 logging_writer.py:48] [12400] global_step=12400, grad_norm=6.175517559051514, loss=4.318788528442383 -I0512 21:21:57.791555 140460899153664 logging_writer.py:48] [12500] global_step=12500, grad_norm=3.1115105152130127, loss=4.270290374755859 -I0512 21:22:35.691357 140460907546368 logging_writer.py:48] [12600] global_step=12600, grad_norm=15.378402709960938, loss=4.819411277770996 -I0512 21:23:12.901382 140460899153664 logging_writer.py:48] [12700] global_step=12700, grad_norm=5.902534484863281, loss=4.329270362854004 -I0512 21:23:50.076521 140460907546368 logging_writer.py:48] [12800] global_step=12800, grad_norm=4.55104398727417, loss=4.449499607086182 -I0512 21:24:27.949125 140460899153664 logging_writer.py:48] [12900] global_step=12900, grad_norm=5.164992332458496, loss=4.3645172119140625 -I0512 21:25:04.949453 140460907546368 logging_writer.py:48] [13000] global_step=13000, grad_norm=3.7780516147613525, loss=4.359182357788086 -I0512 21:25:42.111806 140460899153664 logging_writer.py:48] [13100] global_step=13100, grad_norm=3.6919257640838623, loss=4.319684982299805 -I0512 21:26:19.917155 140460907546368 logging_writer.py:48] [13200] global_step=13200, grad_norm=3.495734691619873, loss=4.3371076583862305 -I0512 21:26:56.956990 140460899153664 logging_writer.py:48] [13300] global_step=13300, grad_norm=6.217592716217041, loss=4.546796798706055 -I0512 21:27:34.081207 140460907546368 logging_writer.py:48] [13400] global_step=13400, grad_norm=4.022764205932617, loss=4.300336837768555 -I0512 21:28:12.112867 140460899153664 logging_writer.py:48] [13500] global_step=13500, grad_norm=4.375720977783203, loss=4.390005588531494 -I0512 21:28:49.227886 140460907546368 logging_writer.py:48] [13600] global_step=13600, grad_norm=4.762039661407471, loss=4.246867656707764 -I0512 21:29:26.290856 140460899153664 logging_writer.py:48] [13700] global_step=13700, grad_norm=6.674348831176758, loss=4.337204456329346 -I0512 21:30:03.916648 140460907546368 logging_writer.py:48] [13800] global_step=13800, grad_norm=5.102276802062988, loss=4.27059268951416 -I0512 21:30:39.987092 140678261474496 spec.py:333] Evaluating on the training split. -I0512 21:30:57.965682 140678261474496 spec.py:346] Evaluating on the validation split. -I0512 21:32:26.640871 140678261474496 spec.py:363] Evaluating on the test split. -I0512 21:32:27.536268 140678261474496 submission_runner.py:516] Time since start: 6340.02s, Step: 13898, {'train/accuracy': Array(0.00548071, dtype=float32), 'train/loss': Array(7.792888, dtype=float32), 'validation/accuracy': Array(0.00558, dtype=float32), 'validation/loss': Array(7.8234067, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0048, dtype=float32), 'test/loss': Array(7.8949347, dtype=float32), 'test/num_examples': 10000, 'score': 6050.281397342682, 'total_duration': 6340.023873090744, 'accumulated_submission_time': 6050.281397342682, 'accumulated_eval_time': 289.51229429244995, 'accumulated_logging_time': 0.09067535400390625} -I0512 21:32:27.560742 140460899153664 logging_writer.py:48] [13898] accumulated_eval_time=289.512, accumulated_logging_time=0.0906754, accumulated_submission_time=6050.28, global_step=13898, preemption_count=0, score=6050.28, test/accuracy=0.004800000227987766, test/loss=7.89493465423584, test/num_examples=10000, total_duration=6340.02, train/accuracy=0.005480707623064518, train/loss=7.792888164520264, validation/accuracy=0.005579999648034573, validation/loss=7.82340669631958, validation/num_examples=50000 -I0512 21:32:28.804743 140460907546368 logging_writer.py:48] [13900] global_step=13900, grad_norm=4.661243438720703, loss=4.360517978668213 -I0512 21:33:05.867975 140460899153664 logging_writer.py:48] [14000] global_step=14000, grad_norm=9.671588897705078, loss=4.370909690856934 -I0512 21:33:43.876498 140460907546368 logging_writer.py:48] [14100] global_step=14100, grad_norm=6.675647258758545, loss=4.263453483581543 -I0512 21:34:20.888692 140460899153664 logging_writer.py:48] [14200] global_step=14200, grad_norm=4.18839693069458, loss=4.301290988922119 -I0512 21:34:57.951947 140460907546368 logging_writer.py:48] [14300] global_step=14300, grad_norm=6.384646415710449, loss=4.292076587677002 -I0512 21:35:35.462467 140460899153664 logging_writer.py:48] [14400] global_step=14400, grad_norm=7.729719161987305, loss=4.4502668380737305 -I0512 21:36:12.857618 140460907546368 logging_writer.py:48] [14500] global_step=14500, grad_norm=4.940676212310791, loss=4.288954734802246 -I0512 21:36:49.964285 140460899153664 logging_writer.py:48] [14600] global_step=14600, grad_norm=5.685422897338867, loss=4.302354335784912 -I0512 21:37:27.833714 140460907546368 logging_writer.py:48] [14700] global_step=14700, grad_norm=6.0255208015441895, loss=4.23175573348999 -I0512 21:38:04.989147 140460899153664 logging_writer.py:48] [14800] global_step=14800, grad_norm=5.5504961013793945, loss=4.385960578918457 -I0512 21:38:42.092712 140460907546368 logging_writer.py:48] [14900] global_step=14900, grad_norm=6.955899238586426, loss=4.346348285675049 -I0512 21:39:19.968170 140460899153664 logging_writer.py:48] [15000] global_step=15000, grad_norm=4.179986953735352, loss=4.264588356018066 -I0512 21:39:57.011686 140460907546368 logging_writer.py:48] [15100] global_step=15100, grad_norm=8.461177825927734, loss=4.342894077301025 -I0512 21:40:34.096003 140460899153664 logging_writer.py:48] [15200] global_step=15200, grad_norm=6.066428184509277, loss=4.26975154876709 -I0512 21:41:11.610327 140460907546368 logging_writer.py:48] [15300] global_step=15300, grad_norm=8.384297370910645, loss=4.302689552307129 -I0512 21:41:49.033259 140460899153664 logging_writer.py:48] [15400] global_step=15400, grad_norm=13.157255172729492, loss=4.478536128997803 -I0512 21:42:26.118835 140460907546368 logging_writer.py:48] [15500] global_step=15500, grad_norm=5.105961322784424, loss=4.214382171630859 -I0512 21:43:04.126653 140460899153664 logging_writer.py:48] [15600] global_step=15600, grad_norm=6.574802875518799, loss=4.1996331214904785 -I0512 21:43:41.245481 140460907546368 logging_writer.py:48] [15700] global_step=15700, grad_norm=3.8020753860473633, loss=4.220542907714844 -I0512 21:44:18.324295 140460899153664 logging_writer.py:48] [15800] global_step=15800, grad_norm=6.580254554748535, loss=4.211867332458496 -I0512 21:44:55.790135 140460907546368 logging_writer.py:48] [15900] global_step=15900, grad_norm=8.78123950958252, loss=4.510676383972168 -I0512 21:45:33.143430 140460899153664 logging_writer.py:48] [16000] global_step=16000, grad_norm=6.020956993103027, loss=4.235291481018066 -I0512 21:46:10.230318 140460907546368 logging_writer.py:48] [16100] global_step=16100, grad_norm=6.290526390075684, loss=4.0996503829956055 -I0512 21:46:48.115260 140460899153664 logging_writer.py:48] [16200] global_step=16200, grad_norm=6.284870624542236, loss=4.225966453552246 -I0512 21:47:25.197239 140460907546368 logging_writer.py:48] [16300] global_step=16300, grad_norm=4.476515769958496, loss=4.229104518890381 -I0512 21:48:02.229368 140460899153664 logging_writer.py:48] [16400] global_step=16400, grad_norm=4.895669460296631, loss=4.06449031829834 -I0512 21:48:39.890382 140460907546368 logging_writer.py:48] [16500] global_step=16500, grad_norm=4.950253486633301, loss=4.1710968017578125 -I0512 21:49:17.349116 140460899153664 logging_writer.py:48] [16600] global_step=16600, grad_norm=6.756086349487305, loss=4.06992244720459 -I0512 21:49:54.438058 140460907546368 logging_writer.py:48] [16700] global_step=16700, grad_norm=12.722112655639648, loss=4.260487079620361 -I0512 21:50:32.098863 140460899153664 logging_writer.py:48] [16800] global_step=16800, grad_norm=5.917260646820068, loss=4.2189531326293945 -I0512 21:51:09.449855 140460907546368 logging_writer.py:48] [16900] global_step=16900, grad_norm=7.482336044311523, loss=4.341756820678711 -I0512 21:51:46.582879 140460899153664 logging_writer.py:48] [17000] global_step=17000, grad_norm=6.045823097229004, loss=4.099071979522705 -I0512 21:52:24.161803 140460907546368 logging_writer.py:48] [17100] global_step=17100, grad_norm=4.9697465896606445, loss=4.179292678833008 -I0512 21:53:01.599653 140460899153664 logging_writer.py:48] [17200] global_step=17200, grad_norm=4.448553562164307, loss=4.167402267456055 -I0512 21:53:38.718037 140460907546368 logging_writer.py:48] [17300] global_step=17300, grad_norm=7.331707954406738, loss=4.11563777923584 -I0512 21:54:16.401544 140460899153664 logging_writer.py:48] [17400] global_step=17400, grad_norm=6.134143352508545, loss=4.244125843048096 -I0512 21:54:53.698284 140460907546368 logging_writer.py:48] [17500] global_step=17500, grad_norm=6.7266316413879395, loss=4.217532157897949 -I0512 21:55:30.828795 140460899153664 logging_writer.py:48] [17600] global_step=17600, grad_norm=5.217182159423828, loss=4.161099433898926 -I0512 21:56:08.453693 140460907546368 logging_writer.py:48] [17700] global_step=17700, grad_norm=7.654499053955078, loss=4.335679054260254 -I0512 21:56:45.891412 140460899153664 logging_writer.py:48] [17800] global_step=17800, grad_norm=2.969372510910034, loss=4.113663673400879 -I0512 21:57:23.039035 140460907546368 logging_writer.py:48] [17900] global_step=17900, grad_norm=8.657811164855957, loss=4.241668701171875 -I0512 21:58:00.597323 140460899153664 logging_writer.py:48] [18000] global_step=18000, grad_norm=5.29068660736084, loss=4.1098480224609375 -I0512 21:58:37.680685 140460907546368 logging_writer.py:48] [18100] global_step=18100, grad_norm=4.804022789001465, loss=4.188257217407227 -I0512 21:59:15.070666 140460899153664 logging_writer.py:48] [18200] global_step=18200, grad_norm=11.38203239440918, loss=4.2105607986450195 -I0512 21:59:52.549940 140460907546368 logging_writer.py:48] [18300] global_step=18300, grad_norm=7.902659893035889, loss=4.0707526206970215 -I0512 22:00:30.109672 140460899153664 logging_writer.py:48] [18400] global_step=18400, grad_norm=3.8684325218200684, loss=4.231892108917236 -I0512 22:01:07.157568 140460907546368 logging_writer.py:48] [18500] global_step=18500, grad_norm=6.293272972106934, loss=4.210627555847168 -I0512 22:01:44.681501 140460899153664 logging_writer.py:48] [18600] global_step=18600, grad_norm=4.282491207122803, loss=4.10167932510376 -I0512 22:02:22.097557 140460907546368 logging_writer.py:48] [18700] global_step=18700, grad_norm=9.689860343933105, loss=4.062150478363037 -I0512 22:02:59.126121 140460899153664 logging_writer.py:48] [18800] global_step=18800, grad_norm=7.613733291625977, loss=4.16509485244751 -I0512 22:03:36.744341 140460907546368 logging_writer.py:48] [18900] global_step=18900, grad_norm=5.683633804321289, loss=4.170206069946289 -I0512 22:04:14.070185 140460899153664 logging_writer.py:48] [19000] global_step=19000, grad_norm=6.760346412658691, loss=4.178253650665283 -I0512 22:04:51.195245 140460907546368 logging_writer.py:48] [19100] global_step=19100, grad_norm=8.425312995910645, loss=4.250405788421631 -I0512 22:05:28.700258 140460899153664 logging_writer.py:48] [19200] global_step=19200, grad_norm=5.696276664733887, loss=4.097387790679932 -I0512 22:05:43.611909 140678261474496 spec.py:333] Evaluating on the training split. -I0512 22:05:57.767787 140678261474496 spec.py:346] Evaluating on the validation split. -I0512 22:07:51.379855 140678261474496 spec.py:363] Evaluating on the test split. -I0512 22:07:52.272495 140678261474496 submission_runner.py:516] Time since start: 8464.76s, Step: 19241, {'train/accuracy': Array(0.00261081, dtype=float32), 'train/loss': Array(8.562036, dtype=float32), 'validation/accuracy': Array(0.00278, dtype=float32), 'validation/loss': Array(8.59723, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0022, dtype=float32), 'test/loss': Array(8.61545, dtype=float32), 'test/num_examples': 10000, 'score': 8046.279819965363, 'total_duration': 8464.760076522827, 'accumulated_submission_time': 8046.279819965363, 'accumulated_eval_time': 418.1712100505829, 'accumulated_logging_time': 0.12340927124023438} -I0512 22:07:52.296056 140460907546368 logging_writer.py:48] [19241] accumulated_eval_time=418.171, accumulated_logging_time=0.123409, accumulated_submission_time=8046.28, global_step=19241, preemption_count=0, score=8046.28, test/accuracy=0.002199999988079071, test/loss=8.615449905395508, test/num_examples=10000, total_duration=8464.76, train/accuracy=0.0026108098682016134, train/loss=8.56203556060791, validation/accuracy=0.0027799999807029963, validation/loss=8.597229957580566, validation/num_examples=50000 -I0512 22:08:23.689247 140460899153664 logging_writer.py:48] [19300] global_step=19300, grad_norm=4.7088303565979, loss=4.109893798828125 -I0512 22:09:07.737943 140460907546368 logging_writer.py:48] [19400] global_step=19400, grad_norm=817.17626953125, loss=5.164997577667236 -I0512 22:09:54.051434 140460899153664 logging_writer.py:48] [19500] global_step=19500, grad_norm=5.698324680328369, loss=4.256600379943848 -I0512 22:10:35.956869 140460907546368 logging_writer.py:48] [19600] global_step=19600, grad_norm=6.223128318786621, loss=4.164681434631348 -I0512 22:11:19.501851 140460899153664 logging_writer.py:48] [19700] global_step=19700, grad_norm=7.645541667938232, loss=4.133942127227783 -I0512 22:12:04.191225 140460907546368 logging_writer.py:48] [19800] global_step=19800, grad_norm=12.623966217041016, loss=4.068239688873291 -I0512 22:12:50.012063 140460899153664 logging_writer.py:48] [19900] global_step=19900, grad_norm=9.16297721862793, loss=4.122015953063965 -I0512 22:13:35.920263 140460907546368 logging_writer.py:48] [20000] global_step=20000, grad_norm=5.315370082855225, loss=4.2006072998046875 -I0512 22:14:31.365416 140460899153664 logging_writer.py:48] [20100] global_step=20100, grad_norm=5.702026844024658, loss=4.178712844848633 -I0512 22:15:17.745119 140460907546368 logging_writer.py:48] [20200] global_step=20200, grad_norm=5.904375076293945, loss=4.195683002471924 -I0512 22:16:03.693935 140460899153664 logging_writer.py:48] [20300] global_step=20300, grad_norm=4.181166648864746, loss=4.1588921546936035 -I0512 22:16:50.277175 140460907546368 logging_writer.py:48] [20400] global_step=20400, grad_norm=5.298942565917969, loss=4.1779890060424805 -I0512 22:17:36.445697 140460899153664 logging_writer.py:48] [20500] global_step=20500, grad_norm=6.689449787139893, loss=4.108107566833496 -I0512 22:18:31.556657 140460907546368 logging_writer.py:48] [20600] global_step=20600, grad_norm=8.670472145080566, loss=4.183460712432861 -I0512 22:19:18.117274 140460899153664 logging_writer.py:48] [20700] global_step=20700, grad_norm=3.4487926959991455, loss=4.052158355712891 -I0512 22:20:13.865839 140460907546368 logging_writer.py:48] [20800] global_step=20800, grad_norm=18.61138153076172, loss=4.169209003448486 -I0512 22:21:00.205700 140460899153664 logging_writer.py:48] [20900] global_step=20900, grad_norm=4.343446254730225, loss=4.120068073272705 -I0512 22:21:46.863050 140460907546368 logging_writer.py:48] [21000] global_step=21000, grad_norm=4.535013675689697, loss=4.037717342376709 -I0512 22:22:38.802746 140460899153664 logging_writer.py:48] [21100] global_step=21100, grad_norm=7.844866752624512, loss=4.172211647033691 -I0512 22:23:20.851716 140460907546368 logging_writer.py:48] [21200] global_step=21200, grad_norm=4.775367259979248, loss=4.113109588623047 -I0512 22:24:06.381311 140460899153664 logging_writer.py:48] [21300] global_step=21300, grad_norm=7.043260097503662, loss=4.145713806152344 -I0512 22:24:51.379566 140460907546368 logging_writer.py:48] [21400] global_step=21400, grad_norm=6.772409915924072, loss=4.210803985595703 -I0512 22:25:37.690397 140460899153664 logging_writer.py:48] [21500] global_step=21500, grad_norm=6.213203430175781, loss=4.195460319519043 -I0512 22:26:24.281705 140460907546368 logging_writer.py:48] [21600] global_step=21600, grad_norm=6.209829330444336, loss=4.050710201263428 -I0512 22:27:10.649393 140460899153664 logging_writer.py:48] [21700] global_step=21700, grad_norm=5.746896743774414, loss=3.9863052368164062 -I0512 22:27:56.968319 140460907546368 logging_writer.py:48] [21800] global_step=21800, grad_norm=11.739855766296387, loss=4.122081756591797 -I0512 22:28:53.234381 140460899153664 logging_writer.py:48] [21900] global_step=21900, grad_norm=4.314589500427246, loss=4.210794448852539 -I0512 22:29:39.521852 140460907546368 logging_writer.py:48] [22000] global_step=22000, grad_norm=5.187862873077393, loss=4.096171855926514 -I0512 22:30:25.707116 140460899153664 logging_writer.py:48] [22100] global_step=22100, grad_norm=14.661001205444336, loss=4.149279594421387 -I0512 22:31:12.355630 140460907546368 logging_writer.py:48] [22200] global_step=22200, grad_norm=2.698119640350342, loss=4.073481559753418 -I0512 22:31:58.807876 140460899153664 logging_writer.py:48] [22300] global_step=22300, grad_norm=6.580981731414795, loss=4.180966377258301 -I0512 22:32:54.116318 140460907546368 logging_writer.py:48] [22400] global_step=22400, grad_norm=4.270268440246582, loss=4.000158309936523 -I0512 22:33:50.289753 140460899153664 logging_writer.py:48] [22500] global_step=22500, grad_norm=4.235983371734619, loss=4.029404163360596 -I0512 22:34:32.603842 140460907546368 logging_writer.py:48] [22600] global_step=22600, grad_norm=7.435615539550781, loss=4.1462507247924805 -I0512 22:35:22.333519 140460899153664 logging_writer.py:48] [22700] global_step=22700, grad_norm=4.446115970611572, loss=3.9997012615203857 -I0512 22:36:08.966088 140460907546368 logging_writer.py:48] [22800] global_step=22800, grad_norm=6.8676581382751465, loss=4.132075309753418 -I0512 22:37:04.716379 140460899153664 logging_writer.py:48] [22900] global_step=22900, grad_norm=5.168275833129883, loss=4.132658958435059 -I0512 22:38:00.165582 140460907546368 logging_writer.py:48] [23000] global_step=23000, grad_norm=6.008204460144043, loss=4.144842147827148 -I0512 22:38:56.230572 140460899153664 logging_writer.py:48] [23100] global_step=23100, grad_norm=8.001880645751953, loss=4.112642288208008 -I0512 22:39:42.824933 140460907546368 logging_writer.py:48] [23200] global_step=23200, grad_norm=5.636435031890869, loss=4.0726447105407715 -I0512 22:40:28.360188 140460899153664 logging_writer.py:48] [23300] global_step=23300, grad_norm=4.658445358276367, loss=3.916762590408325 -I0512 22:41:08.476908 140678261474496 spec.py:333] Evaluating on the training split. -I0512 22:41:23.613133 140678261474496 spec.py:346] Evaluating on the validation split. -I0512 22:42:46.940881 140678261474496 spec.py:363] Evaluating on the test split. -I0512 22:42:47.835228 140678261474496 submission_runner.py:516] Time since start: 10560.32s, Step: 23383, {'train/accuracy': Array(0.0017339, dtype=float32), 'train/loss': Array(9.085137, dtype=float32), 'validation/accuracy': Array(0.00208, dtype=float32), 'validation/loss': Array(9.102534, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0016, dtype=float32), 'test/loss': Array(9.104018, dtype=float32), 'test/num_examples': 10000, 'score': 10042.387595891953, 'total_duration': 10560.322937726974, 'accumulated_submission_time': 10042.387595891953, 'accumulated_eval_time': 517.5279848575592, 'accumulated_logging_time': 0.15534067153930664} -I0512 22:42:47.859179 140460907546368 logging_writer.py:48] [23383] accumulated_eval_time=517.528, accumulated_logging_time=0.155341, accumulated_submission_time=10042.4, global_step=23383, preemption_count=0, score=10042.4, test/accuracy=0.001600000075995922, test/loss=9.104018211364746, test/num_examples=10000, total_duration=10560.3, train/accuracy=0.001733896671794355, train/loss=9.085137367248535, validation/accuracy=0.0020800000056624413, validation/loss=9.102534294128418, validation/num_examples=50000 -I0512 22:42:54.587152 140460899153664 logging_writer.py:48] [23400] global_step=23400, grad_norm=7.469196796417236, loss=4.020120620727539 -I0512 22:43:31.846406 140460907546368 logging_writer.py:48] [23500] global_step=23500, grad_norm=4.324219226837158, loss=4.098857402801514 -I0512 22:44:08.919120 140460899153664 logging_writer.py:48] [23600] global_step=23600, grad_norm=6.944002151489258, loss=4.1543145179748535 -I0512 22:44:46.708348 140460907546368 logging_writer.py:48] [23700] global_step=23700, grad_norm=3.6725425720214844, loss=3.9369683265686035 -I0512 22:45:23.642656 140460899153664 logging_writer.py:48] [23800] global_step=23800, grad_norm=6.247439384460449, loss=4.092560291290283 -I0512 22:46:00.696901 140460907546368 logging_writer.py:48] [23900] global_step=23900, grad_norm=8.95309829711914, loss=4.071827411651611 -I0512 22:46:38.594116 140460899153664 logging_writer.py:48] [24000] global_step=24000, grad_norm=4.959518909454346, loss=3.9737188816070557 -I0512 22:47:15.585571 140460907546368 logging_writer.py:48] [24100] global_step=24100, grad_norm=5.33604621887207, loss=4.170638084411621 -I0512 22:47:52.629798 140460899153664 logging_writer.py:48] [24200] global_step=24200, grad_norm=5.086591720581055, loss=3.9502077102661133 -I0512 22:48:30.198372 140460907546368 logging_writer.py:48] [24300] global_step=24300, grad_norm=6.998744487762451, loss=4.06788969039917 -I0512 22:49:07.449250 140460899153664 logging_writer.py:48] [24400] global_step=24400, grad_norm=6.581602573394775, loss=4.038023471832275 -I0512 22:49:44.468358 140460907546368 logging_writer.py:48] [24500] global_step=24500, grad_norm=5.784135818481445, loss=4.0264058113098145 -I0512 22:50:22.319516 140460899153664 logging_writer.py:48] [24600] global_step=24600, grad_norm=11.75537395477295, loss=3.9837288856506348 -I0512 22:50:59.327781 140460907546368 logging_writer.py:48] [24700] global_step=24700, grad_norm=4.898009300231934, loss=3.9188036918640137 -I0512 22:51:36.354942 140460899153664 logging_writer.py:48] [24800] global_step=24800, grad_norm=6.857207298278809, loss=4.0317301750183105 -I0512 22:52:13.853923 140460907546368 logging_writer.py:48] [24900] global_step=24900, grad_norm=7.8823981285095215, loss=4.092549800872803 -I0512 22:52:51.118758 140460899153664 logging_writer.py:48] [25000] global_step=25000, grad_norm=6.732275009155273, loss=4.044760704040527 -I0512 22:53:28.241915 140460907546368 logging_writer.py:48] [25100] global_step=25100, grad_norm=5.680588722229004, loss=3.969468593597412 -I0512 22:54:05.779047 140460899153664 logging_writer.py:48] [25200] global_step=25200, grad_norm=5.693339824676514, loss=3.89259934425354 -I0512 22:54:43.100411 140460907546368 logging_writer.py:48] [25300] global_step=25300, grad_norm=5.094283580780029, loss=3.987377405166626 -I0512 22:55:20.073786 140460899153664 logging_writer.py:48] [25400] global_step=25400, grad_norm=4.301829814910889, loss=4.112467288970947 -I0512 22:55:57.502305 140460907546368 logging_writer.py:48] [25500] global_step=25500, grad_norm=4.522195816040039, loss=4.003249168395996 -I0512 22:56:34.857099 140460899153664 logging_writer.py:48] [25600] global_step=25600, grad_norm=3.757065534591675, loss=3.950075626373291 -I0512 22:57:11.852909 140460907546368 logging_writer.py:48] [25700] global_step=25700, grad_norm=5.550106048583984, loss=3.9580740928649902 -I0512 22:57:49.319200 140460899153664 logging_writer.py:48] [25800] global_step=25800, grad_norm=6.632765769958496, loss=4.022561073303223 -I0512 22:58:26.604060 140460907546368 logging_writer.py:48] [25900] global_step=25900, grad_norm=3.975090980529785, loss=3.8840675354003906 -I0512 22:59:03.624452 140460899153664 logging_writer.py:48] [26000] global_step=26000, grad_norm=4.639017581939697, loss=3.8661935329437256 -I0512 22:59:41.077447 140460907546368 logging_writer.py:48] [26100] global_step=26100, grad_norm=10.61921215057373, loss=7.592290878295898 -I0512 23:00:18.370862 140460899153664 logging_writer.py:48] [26200] global_step=26200, grad_norm=3.985041856765747, loss=4.046590805053711 -I0512 23:00:55.346307 140460907546368 logging_writer.py:48] [26300] global_step=26300, grad_norm=4.68715763092041, loss=3.8844759464263916 -I0512 23:01:32.928237 140460899153664 logging_writer.py:48] [26400] global_step=26400, grad_norm=7.104184150695801, loss=4.084194183349609 -I0512 23:02:10.181522 140460907546368 logging_writer.py:48] [26500] global_step=26500, grad_norm=12.174339294433594, loss=3.951478958129883 -I0512 23:02:47.286231 140460899153664 logging_writer.py:48] [26600] global_step=26600, grad_norm=5.5463480949401855, loss=3.94816255569458 -I0512 23:03:24.691524 140460907546368 logging_writer.py:48] [26700] global_step=26700, grad_norm=8.468329429626465, loss=4.0959391593933105 -I0512 23:04:02.041142 140460899153664 logging_writer.py:48] [26800] global_step=26800, grad_norm=8.281891822814941, loss=4.1078410148620605 -I0512 23:04:39.129235 140460907546368 logging_writer.py:48] [26900] global_step=26900, grad_norm=6.346573352813721, loss=4.047309875488281 -I0512 23:05:16.616204 140460899153664 logging_writer.py:48] [27000] global_step=27000, grad_norm=7.297499179840088, loss=4.273834228515625 -I0512 23:05:53.643256 140460907546368 logging_writer.py:48] [27100] global_step=27100, grad_norm=4.51987886428833, loss=4.060060501098633 -I0512 23:06:30.941808 140460899153664 logging_writer.py:48] [27200] global_step=27200, grad_norm=3.925402879714966, loss=3.896911144256592 -I0512 23:07:08.412904 140460907546368 logging_writer.py:48] [27300] global_step=27300, grad_norm=5.0999016761779785, loss=4.005029678344727 -I0512 23:07:45.790868 140460899153664 logging_writer.py:48] [27400] global_step=27400, grad_norm=5.39810037612915, loss=3.9265379905700684 -I0512 23:08:22.777321 140460907546368 logging_writer.py:48] [27500] global_step=27500, grad_norm=7.381557464599609, loss=4.0014214515686035 -I0512 23:09:00.311771 140460899153664 logging_writer.py:48] [27600] global_step=27600, grad_norm=8.746315956115723, loss=3.9310121536254883 -I0512 23:09:37.642983 140460907546368 logging_writer.py:48] [27700] global_step=27700, grad_norm=6.3469557762146, loss=4.061192512512207 -I0512 23:10:14.695376 140460899153664 logging_writer.py:48] [27800] global_step=27800, grad_norm=6.541885852813721, loss=3.9785218238830566 -I0512 23:10:52.189216 140460907546368 logging_writer.py:48] [27900] global_step=27900, grad_norm=5.535500526428223, loss=4.124883651733398 -I0512 23:11:29.561392 140460899153664 logging_writer.py:48] [28000] global_step=28000, grad_norm=10.865156173706055, loss=4.179146766662598 -I0512 23:12:06.662478 140460907546368 logging_writer.py:48] [28100] global_step=28100, grad_norm=6.633310794830322, loss=4.112743854522705 -I0512 23:12:44.102836 140460899153664 logging_writer.py:48] [28200] global_step=28200, grad_norm=5.747229099273682, loss=3.8209877014160156 -I0512 23:13:21.410777 140460907546368 logging_writer.py:48] [28300] global_step=28300, grad_norm=5.310873508453369, loss=4.021536827087402 -I0512 23:13:58.373076 140460899153664 logging_writer.py:48] [28400] global_step=28400, grad_norm=3.441260814666748, loss=3.9195613861083984 -I0512 23:14:38.343540 140460907546368 logging_writer.py:48] [28500] global_step=28500, grad_norm=3.8672032356262207, loss=3.8840372562408447 -I0512 23:15:15.545162 140460899153664 logging_writer.py:48] [28600] global_step=28600, grad_norm=10.93833065032959, loss=4.148106098175049 -I0512 23:15:52.616312 140460907546368 logging_writer.py:48] [28700] global_step=28700, grad_norm=6.39409065246582, loss=3.898073673248291 -I0512 23:16:04.317744 140678261474496 spec.py:333] Evaluating on the training split. -I0512 23:16:17.551745 140678261474496 spec.py:346] Evaluating on the validation split. -I0512 23:18:27.701342 140678261474496 spec.py:363] Evaluating on the test split. -I0512 23:18:28.595062 140678261474496 submission_runner.py:516] Time since start: 12701.08s, Step: 28732, {'train/accuracy': Array(0.00169404, dtype=float32), 'train/loss': Array(9.291868, dtype=float32), 'validation/accuracy': Array(0.00156, dtype=float32), 'validation/loss': Array(9.292005, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0017, dtype=float32), 'test/loss': Array(9.287293, dtype=float32), 'test/num_examples': 10000, 'score': 12038.792675971985, 'total_duration': 12701.082419395447, 'accumulated_submission_time': 12038.792675971985, 'accumulated_eval_time': 661.8034045696259, 'accumulated_logging_time': 0.18757104873657227} -I0512 23:18:28.620528 140460899153664 logging_writer.py:48] [28732] accumulated_eval_time=661.803, accumulated_logging_time=0.187571, accumulated_submission_time=12038.8, global_step=28732, preemption_count=0, score=12038.8, test/accuracy=0.0017000001389533281, test/loss=9.287293434143066, test/num_examples=10000, total_duration=12701.1, train/accuracy=0.001694036996923387, train/loss=9.291868209838867, validation/accuracy=0.001560000004246831, validation/loss=9.292004585266113, validation/num_examples=50000 -I0512 23:18:54.296909 140460907546368 logging_writer.py:48] [28800] global_step=28800, grad_norm=4.593683242797852, loss=3.9742445945739746 -I0512 23:19:31.570414 140460899153664 logging_writer.py:48] [28900] global_step=28900, grad_norm=2.6776022911071777, loss=3.92946720123291 -I0512 23:20:08.527748 140460907546368 logging_writer.py:48] [29000] global_step=29000, grad_norm=4.102603912353516, loss=3.7856805324554443 -I0512 23:20:46.010860 140460899153664 logging_writer.py:48] [29100] global_step=29100, grad_norm=6.687881946563721, loss=4.086498260498047 -I0512 23:21:23.152200 140460907546368 logging_writer.py:48] [29200] global_step=29200, grad_norm=5.806509017944336, loss=3.9896740913391113 -I0512 23:22:00.184600 140460899153664 logging_writer.py:48] [29300] global_step=29300, grad_norm=4.775323390960693, loss=3.9140219688415527 -I0512 23:22:37.689434 140460907546368 logging_writer.py:48] [29400] global_step=29400, grad_norm=8.361307144165039, loss=4.182333946228027 -I0512 23:23:14.995522 140460899153664 logging_writer.py:48] [29500] global_step=29500, grad_norm=15.305659294128418, loss=3.8728184700012207 -I0512 23:23:51.977704 140460907546368 logging_writer.py:48] [29600] global_step=29600, grad_norm=6.309682846069336, loss=4.018271446228027 -I0512 23:24:29.477912 140460899153664 logging_writer.py:48] [29700] global_step=29700, grad_norm=9.141960144042969, loss=4.079906463623047 -I0512 23:25:06.436199 140460907546368 logging_writer.py:48] [29800] global_step=29800, grad_norm=4.466619491577148, loss=3.948256015777588 -I0512 23:25:43.759557 140460899153664 logging_writer.py:48] [29900] global_step=29900, grad_norm=5.159104824066162, loss=3.8990321159362793 -I0512 23:26:21.275111 140460907546368 logging_writer.py:48] [30000] global_step=30000, grad_norm=5.917943000793457, loss=4.011873245239258 -I0512 23:26:58.598469 140460899153664 logging_writer.py:48] [30100] global_step=30100, grad_norm=5.876935005187988, loss=3.879673957824707 -I0512 23:27:35.764792 140460907546368 logging_writer.py:48] [30200] global_step=30200, grad_norm=334.0513610839844, loss=4.729025363922119 -I0512 23:28:13.355149 140460899153664 logging_writer.py:48] [30300] global_step=30300, grad_norm=4.567579746246338, loss=3.9803857803344727 -I0512 23:28:50.377415 140460907546368 logging_writer.py:48] [30400] global_step=30400, grad_norm=110.10150909423828, loss=4.211209774017334 -I0512 23:29:27.768006 140460899153664 logging_writer.py:48] [30500] global_step=30500, grad_norm=4.694270610809326, loss=3.8841476440429688 -I0512 23:30:05.349211 140460907546368 logging_writer.py:48] [30600] global_step=30600, grad_norm=4.290919780731201, loss=3.9688446521759033 -I0512 23:30:42.369229 140460899153664 logging_writer.py:48] [30700] global_step=30700, grad_norm=6.181513786315918, loss=3.870084285736084 -I0512 23:31:19.717257 140460907546368 logging_writer.py:48] [30800] global_step=30800, grad_norm=13.71870231628418, loss=3.9024641513824463 -I0512 23:31:57.196894 140460899153664 logging_writer.py:48] [30900] global_step=30900, grad_norm=11.360831260681152, loss=4.241281986236572 -I0512 23:32:34.254319 140460907546368 logging_writer.py:48] [31000] global_step=31000, grad_norm=5.000714302062988, loss=3.8956685066223145 -I0512 23:33:11.616704 140460899153664 logging_writer.py:48] [31100] global_step=31100, grad_norm=5.461056709289551, loss=3.9804091453552246 -I0512 23:33:49.176294 140460907546368 logging_writer.py:48] [31200] global_step=31200, grad_norm=6.810269832611084, loss=4.078872203826904 -I0512 23:34:26.145868 140460899153664 logging_writer.py:48] [31300] global_step=31300, grad_norm=4.890822887420654, loss=3.927492618560791 -I0512 23:35:03.517937 140460907546368 logging_writer.py:48] [31400] global_step=31400, grad_norm=3.603114604949951, loss=3.908430576324463 -I0512 23:35:40.993666 140460899153664 logging_writer.py:48] [31500] global_step=31500, grad_norm=4.079023361206055, loss=3.8883867263793945 -I0512 23:36:26.779870 140460907546368 logging_writer.py:48] [31600] global_step=31600, grad_norm=4.129029273986816, loss=3.7547812461853027 -I0512 23:37:12.933854 140460899153664 logging_writer.py:48] [31700] global_step=31700, grad_norm=5.629909515380859, loss=3.932936191558838 -I0512 23:37:59.213669 140460907546368 logging_writer.py:48] [31800] global_step=31800, grad_norm=4.61163854598999, loss=3.880995988845825 -I0512 23:38:36.284182 140460899153664 logging_writer.py:48] [31900] global_step=31900, grad_norm=4.990808010101318, loss=3.8670578002929688 -I0512 23:39:13.674161 140460907546368 logging_writer.py:48] [32000] global_step=32000, grad_norm=4.495237350463867, loss=3.946864604949951 -I0512 23:39:51.156318 140460899153664 logging_writer.py:48] [32100] global_step=32100, grad_norm=4.441437721252441, loss=3.791142225265503 -I0512 23:40:28.246507 140460907546368 logging_writer.py:48] [32200] global_step=32200, grad_norm=38.57019805908203, loss=4.039306163787842 -I0512 23:41:05.627715 140460899153664 logging_writer.py:48] [32300] global_step=32300, grad_norm=18.168357849121094, loss=3.980849504470825 -I0512 23:41:43.185694 140460907546368 logging_writer.py:48] [32400] global_step=32400, grad_norm=4.509559631347656, loss=3.9073166847229004 -I0512 23:42:20.211133 140460899153664 logging_writer.py:48] [32500] global_step=32500, grad_norm=5.841487884521484, loss=4.043783187866211 -I0512 23:42:57.656558 140460907546368 logging_writer.py:48] [32600] global_step=32600, grad_norm=5.6468586921691895, loss=3.7931935787200928 -I0512 23:43:52.605919 140460899153664 logging_writer.py:48] [32700] global_step=32700, grad_norm=7.193542957305908, loss=3.9888951778411865 -I0512 23:44:47.029262 140460907546368 logging_writer.py:48] [32800] global_step=32800, grad_norm=4.288009166717529, loss=3.8480312824249268 -I0512 23:45:24.369811 140460899153664 logging_writer.py:48] [32900] global_step=32900, grad_norm=11.305368423461914, loss=3.8756489753723145 -I0512 23:46:10.520007 140460907546368 logging_writer.py:48] [33000] global_step=33000, grad_norm=6.238190174102783, loss=3.986046314239502 -I0512 23:46:56.150789 140460899153664 logging_writer.py:48] [33100] global_step=33100, grad_norm=5.3226318359375, loss=3.8669161796569824 -I0512 23:47:42.256354 140460907546368 logging_writer.py:48] [33200] global_step=33200, grad_norm=5.93214750289917, loss=3.9449024200439453 -I0512 23:48:28.525422 140460899153664 logging_writer.py:48] [33300] global_step=33300, grad_norm=5.085150718688965, loss=3.77662992477417 -I0512 23:49:06.023018 140460907546368 logging_writer.py:48] [33400] global_step=33400, grad_norm=8.001176834106445, loss=4.092648506164551 -I0512 23:49:43.393059 140460899153664 logging_writer.py:48] [33500] global_step=33500, grad_norm=7.774834632873535, loss=3.817535877227783 -I0512 23:50:21.109521 140460907546368 logging_writer.py:48] [33600] global_step=33600, grad_norm=7.218889236450195, loss=3.9749560356140137 -I0512 23:50:58.206124 140460899153664 logging_writer.py:48] [33700] global_step=33700, grad_norm=4.285229682922363, loss=3.99122953414917 -I0512 23:51:35.689217 140460907546368 logging_writer.py:48] [33800] global_step=33800, grad_norm=5.674108982086182, loss=4.0168890953063965 -I0512 23:51:45.220688 140678261474496 spec.py:333] Evaluating on the training split. -I0512 23:51:59.934881 140678261474496 spec.py:346] Evaluating on the validation split. -I0512 23:53:30.792691 140678261474496 spec.py:363] Evaluating on the test split. -I0512 23:53:31.688662 140678261474496 submission_runner.py:516] Time since start: 14804.18s, Step: 33826, {'train/accuracy': Array(0.00129544, dtype=float32), 'train/loss': Array(9.402161, dtype=float32), 'validation/accuracy': Array(0.0014, dtype=float32), 'validation/loss': Array(9.399457, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0016, dtype=float32), 'test/loss': Array(9.391503, dtype=float32), 'test/num_examples': 10000, 'score': 14035.339556455612, 'total_duration': 14804.176125764847, 'accumulated_submission_time': 14035.339556455612, 'accumulated_eval_time': 768.2695868015289, 'accumulated_logging_time': 0.2213890552520752} -I0512 23:53:31.712617 140460899153664 logging_writer.py:48] [33826] accumulated_eval_time=768.27, accumulated_logging_time=0.221389, accumulated_submission_time=14035.3, global_step=33826, preemption_count=0, score=14035.3, test/accuracy=0.001600000075995922, test/loss=9.39150333404541, test/num_examples=10000, total_duration=14804.2, train/accuracy=0.0012954400153830647, train/loss=9.40216064453125, validation/accuracy=0.00139999995008111, validation/loss=9.399456977844238, validation/num_examples=50000 -I0512 23:53:59.574857 140460907546368 logging_writer.py:48] [33900] global_step=33900, grad_norm=5.2740864753723145, loss=3.943526268005371 -I0512 23:54:36.589154 140460899153664 logging_writer.py:48] [34000] global_step=34000, grad_norm=3.731858730316162, loss=3.8649230003356934 -I0512 23:55:13.998659 140460907546368 logging_writer.py:48] [34100] global_step=34100, grad_norm=4.9052300453186035, loss=3.8892178535461426 -I0512 23:55:51.653637 140460899153664 logging_writer.py:48] [34200] global_step=34200, grad_norm=4.647274971008301, loss=4.016456604003906 -I0512 23:56:28.932076 140460907546368 logging_writer.py:48] [34300] global_step=34300, grad_norm=3.6712818145751953, loss=4.004286289215088 -I0512 23:57:06.843609 140460899153664 logging_writer.py:48] [34400] global_step=34400, grad_norm=9.553580284118652, loss=4.054805755615234 -I0512 23:57:45.056600 140460907546368 logging_writer.py:48] [34500] global_step=34500, grad_norm=5.086247444152832, loss=4.015503406524658 -I0512 23:58:27.831474 140460899153664 logging_writer.py:48] [34600] global_step=34600, grad_norm=14.760353088378906, loss=3.9023690223693848 -I0512 23:59:08.120293 140460907546368 logging_writer.py:48] [34700] global_step=34700, grad_norm=9.274499893188477, loss=4.014214515686035 -I0512 23:59:48.004845 140460899153664 logging_writer.py:48] [34800] global_step=34800, grad_norm=7.706545829772949, loss=3.956058979034424 -I0513 00:00:29.594012 140460907546368 logging_writer.py:48] [34900] global_step=34900, grad_norm=8.262212753295898, loss=3.9548563957214355 -I0513 00:01:14.640623 140460899153664 logging_writer.py:48] [35000] global_step=35000, grad_norm=4.767133712768555, loss=3.9101474285125732 -I0513 00:01:56.527508 140460907546368 logging_writer.py:48] [35100] global_step=35100, grad_norm=3.416614055633545, loss=3.7841081619262695 -I0513 00:02:36.453205 140460899153664 logging_writer.py:48] [35200] global_step=35200, grad_norm=11.924470901489258, loss=4.188222408294678 -I0513 00:03:16.772622 140460907546368 logging_writer.py:48] [35300] global_step=35300, grad_norm=4.628430366516113, loss=3.8528409004211426 -I0513 00:03:57.201361 140460899153664 logging_writer.py:48] [35400] global_step=35400, grad_norm=9.179483413696289, loss=3.7647600173950195 -I0513 00:04:39.925024 140460907546368 logging_writer.py:48] [35500] global_step=35500, grad_norm=5.30332088470459, loss=4.019226551055908 -I0513 00:05:19.816790 140460899153664 logging_writer.py:48] [35600] global_step=35600, grad_norm=6.130948543548584, loss=3.7886786460876465 -I0513 00:06:06.350121 140460907546368 logging_writer.py:48] [35700] global_step=35700, grad_norm=5.50038480758667, loss=3.800386428833008 -I0513 00:06:52.090049 140460899153664 logging_writer.py:48] [35800] global_step=35800, grad_norm=4.889102458953857, loss=3.9283413887023926 -I0513 00:07:32.345787 140460907546368 logging_writer.py:48] [35900] global_step=35900, grad_norm=4.193506240844727, loss=3.8270788192749023 -I0513 00:08:10.964926 140460899153664 logging_writer.py:48] [36000] global_step=36000, grad_norm=4.161726951599121, loss=3.820859432220459 -I0513 00:08:48.632276 140460907546368 logging_writer.py:48] [36100] global_step=36100, grad_norm=5.181065559387207, loss=3.7603752613067627 -I0513 00:09:27.084012 140460899153664 logging_writer.py:48] [36200] global_step=36200, grad_norm=9.60279369354248, loss=4.012823104858398 -I0513 00:10:06.395496 140460907546368 logging_writer.py:48] [36300] global_step=36300, grad_norm=5.730980396270752, loss=3.8713431358337402 -I0513 00:10:59.307677 140460899153664 logging_writer.py:48] [36400] global_step=36400, grad_norm=6.932391166687012, loss=3.802305221557617 -I0513 00:11:45.433556 140460907546368 logging_writer.py:48] [36500] global_step=36500, grad_norm=6.05096435546875, loss=3.814488172531128 -I0513 00:12:31.595372 140460899153664 logging_writer.py:48] [36600] global_step=36600, grad_norm=7.017299175262451, loss=3.993959426879883 -I0513 00:13:17.342878 140460907546368 logging_writer.py:48] [36700] global_step=36700, grad_norm=4.305166721343994, loss=3.810619354248047 -I0513 00:13:57.297036 140460899153664 logging_writer.py:48] [36800] global_step=36800, grad_norm=7.476315498352051, loss=3.795881748199463 -I0513 00:14:43.832438 140460907546368 logging_writer.py:48] [36900] global_step=36900, grad_norm=6.789056777954102, loss=3.8704824447631836 -I0513 00:15:38.205013 140460899153664 logging_writer.py:48] [37000] global_step=37000, grad_norm=4.217155456542969, loss=3.7961649894714355 -I0513 00:16:23.923819 140460907546368 logging_writer.py:48] [37100] global_step=37100, grad_norm=4.723061561584473, loss=3.8952507972717285 -I0513 00:17:17.220560 140460899153664 logging_writer.py:48] [37200] global_step=37200, grad_norm=7.963853359222412, loss=3.729318141937256 -I0513 00:18:04.422263 140460907546368 logging_writer.py:48] [37300] global_step=37300, grad_norm=6.8284101486206055, loss=4.016778469085693 -I0513 00:18:58.930936 140460899153664 logging_writer.py:48] [37400] global_step=37400, grad_norm=2.38495135307312, loss=3.7102012634277344 -I0513 00:19:54.226351 140460907546368 logging_writer.py:48] [37500] global_step=37500, grad_norm=4.019250869750977, loss=3.8829078674316406 -I0513 00:20:39.962672 140460899153664 logging_writer.py:48] [37600] global_step=37600, grad_norm=5.876768589019775, loss=3.779147148132324 -I0513 00:21:25.727631 140460907546368 logging_writer.py:48] [37700] global_step=37700, grad_norm=10.435040473937988, loss=3.79048490524292 -I0513 00:22:12.196297 140460899153664 logging_writer.py:48] [37800] global_step=37800, grad_norm=8.809161186218262, loss=4.113216400146484 -I0513 00:22:58.026731 140460907546368 logging_writer.py:48] [37900] global_step=37900, grad_norm=3.737663984298706, loss=4.010284900665283 -I0513 00:23:43.830417 140460899153664 logging_writer.py:48] [38000] global_step=38000, grad_norm=4.482631206512451, loss=3.845797061920166 -I0513 00:24:30.255160 140460907546368 logging_writer.py:48] [38100] global_step=38100, grad_norm=4.094576835632324, loss=3.920849561691284 -I0513 00:25:16.002614 140460899153664 logging_writer.py:48] [38200] global_step=38200, grad_norm=5.627108573913574, loss=3.767085075378418 -I0513 00:26:01.773025 140460907546368 logging_writer.py:48] [38300] global_step=38300, grad_norm=4.655827045440674, loss=3.7716126441955566 -I0513 00:26:48.139120 140460899153664 logging_writer.py:48] [38400] global_step=38400, grad_norm=3.1659157276153564, loss=3.742828845977783 -I0513 00:26:48.152704 140678261474496 spec.py:333] Evaluating on the training split. -I0513 00:27:00.294796 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 00:28:29.281609 140678261474496 spec.py:363] Evaluating on the test split. -I0513 00:28:30.177480 140678261474496 submission_runner.py:516] Time since start: 16902.66s, Step: 38401, {'train/accuracy': Array(0.00149474, dtype=float32), 'train/loss': Array(9.477643, dtype=float32), 'validation/accuracy': Array(0.00126, dtype=float32), 'validation/loss': Array(9.454873, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0014, dtype=float32), 'test/loss': Array(9.444064, dtype=float32), 'test/num_examples': 10000, 'score': 16031.727242708206, 'total_duration': 16902.664816856384, 'accumulated_submission_time': 16031.727242708206, 'accumulated_eval_time': 870.2924361228943, 'accumulated_logging_time': 0.2536473274230957} -I0513 00:28:30.203893 140460907546368 logging_writer.py:48] [38401] accumulated_eval_time=870.292, accumulated_logging_time=0.253647, accumulated_submission_time=16031.7, global_step=38401, preemption_count=0, score=16031.7, test/accuracy=0.0014000000664964318, test/loss=9.444064140319824, test/num_examples=10000, total_duration=16902.7, train/accuracy=0.001494738506153226, train/loss=9.477643013000488, validation/accuracy=0.0012599999317899346, validation/loss=9.454873085021973, validation/num_examples=50000 -I0513 00:29:25.525971 140460899153664 logging_writer.py:48] [38500] global_step=38500, grad_norm=74.31021118164062, loss=3.9889984130859375 -I0513 00:30:11.726873 140460907546368 logging_writer.py:48] [38600] global_step=38600, grad_norm=13.02324390411377, loss=3.8376333713531494 -I0513 00:31:07.319283 140460899153664 logging_writer.py:48] [38700] global_step=38700, grad_norm=7.254849910736084, loss=3.7825369834899902 -I0513 00:31:53.270737 140460907546368 logging_writer.py:48] [38800] global_step=38800, grad_norm=5.694000720977783, loss=3.6525745391845703 -I0513 00:32:39.349922 140460899153664 logging_writer.py:48] [38900] global_step=38900, grad_norm=4.509198188781738, loss=3.8009040355682373 -I0513 00:33:25.878992 140460907546368 logging_writer.py:48] [39000] global_step=39000, grad_norm=3.8916172981262207, loss=3.884906053543091 -I0513 00:34:12.345507 140460899153664 logging_writer.py:48] [39100] global_step=39100, grad_norm=8.689702987670898, loss=4.134483814239502 -I0513 00:34:58.517104 140460907546368 logging_writer.py:48] [39200] global_step=39200, grad_norm=6.129915714263916, loss=3.8923306465148926 -I0513 00:35:45.409996 140460899153664 logging_writer.py:48] [39300] global_step=39300, grad_norm=3.6771059036254883, loss=3.8799681663513184 -I0513 00:36:31.380934 140460907546368 logging_writer.py:48] [39400] global_step=39400, grad_norm=4.965310096740723, loss=3.986145257949829 -I0513 00:37:17.462362 140460899153664 logging_writer.py:48] [39500] global_step=39500, grad_norm=4.43911600112915, loss=3.8144195079803467 -I0513 00:38:04.484318 140460907546368 logging_writer.py:48] [39600] global_step=39600, grad_norm=6.219709396362305, loss=3.92643141746521 -I0513 00:38:43.924220 140460899153664 logging_writer.py:48] [39700] global_step=39700, grad_norm=2.9233217239379883, loss=3.9433810710906982 -I0513 00:39:22.759408 140460907546368 logging_writer.py:48] [39800] global_step=39800, grad_norm=4.938845157623291, loss=3.8950366973876953 -I0513 00:40:04.046181 140460899153664 logging_writer.py:48] [39900] global_step=39900, grad_norm=6.098181247711182, loss=3.977607250213623 -I0513 00:40:43.831804 140460907546368 logging_writer.py:48] [40000] global_step=40000, grad_norm=6.1158246994018555, loss=3.777393341064453 -I0513 00:41:23.405110 140460899153664 logging_writer.py:48] [40100] global_step=40100, grad_norm=5.440559387207031, loss=3.814018726348877 -I0513 00:42:03.848608 140460907546368 logging_writer.py:48] [40200] global_step=40200, grad_norm=11.337740898132324, loss=3.7536184787750244 -I0513 00:42:49.633190 140460899153664 logging_writer.py:48] [40300] global_step=40300, grad_norm=4.05787992477417, loss=3.9406161308288574 -I0513 00:43:35.459354 140460907546368 logging_writer.py:48] [40400] global_step=40400, grad_norm=3.3508381843566895, loss=3.8101859092712402 -I0513 00:44:28.365446 140460899153664 logging_writer.py:48] [40500] global_step=40500, grad_norm=7.090555667877197, loss=3.848144292831421 -I0513 00:45:09.912545 140460907546368 logging_writer.py:48] [40600] global_step=40600, grad_norm=3.4256021976470947, loss=3.8727519512176514 -I0513 00:45:55.616063 140460899153664 logging_writer.py:48] [40700] global_step=40700, grad_norm=3.5055387020111084, loss=3.820167064666748 -I0513 00:46:41.912285 140460907546368 logging_writer.py:48] [40800] global_step=40800, grad_norm=3.2566163539886475, loss=3.763641119003296 -I0513 00:47:25.768216 140460899153664 logging_writer.py:48] [40900] global_step=40900, grad_norm=6.402026653289795, loss=3.9613943099975586 -I0513 00:48:07.610019 140460907546368 logging_writer.py:48] [41000] global_step=41000, grad_norm=5.726372718811035, loss=3.850922107696533 -I0513 00:48:52.577376 140460899153664 logging_writer.py:48] [41100] global_step=41100, grad_norm=3.383744239807129, loss=3.945131778717041 -I0513 00:49:35.801748 140460907546368 logging_writer.py:48] [41200] global_step=41200, grad_norm=5.0277910232543945, loss=3.8080291748046875 -I0513 00:50:21.632838 140460899153664 logging_writer.py:48] [41300] global_step=41300, grad_norm=4.711620807647705, loss=3.8125572204589844 -I0513 00:51:08.078857 140460907546368 logging_writer.py:48] [41400] global_step=41400, grad_norm=3.9551613330841064, loss=3.7780237197875977 -I0513 00:51:54.352777 140460899153664 logging_writer.py:48] [41500] global_step=41500, grad_norm=4.2779765129089355, loss=3.9844212532043457 -I0513 00:52:50.298785 140460907546368 logging_writer.py:48] [41600] global_step=41600, grad_norm=3.6253175735473633, loss=3.777400493621826 -I0513 00:53:37.188353 140460899153664 logging_writer.py:48] [41700] global_step=41700, grad_norm=3.196223497390747, loss=3.8400285243988037 -I0513 00:54:23.825684 140460907546368 logging_writer.py:48] [41800] global_step=41800, grad_norm=6.3026814460754395, loss=3.842402935028076 -I0513 00:55:10.087446 140460899153664 logging_writer.py:48] [41900] global_step=41900, grad_norm=5.438497066497803, loss=3.839418411254883 -I0513 00:55:52.791182 140460907546368 logging_writer.py:48] [42000] global_step=42000, grad_norm=2.415339946746826, loss=3.710728406906128 -I0513 00:56:52.205621 140460899153664 logging_writer.py:48] [42100] global_step=42100, grad_norm=7.6265482902526855, loss=3.946192741394043 -I0513 00:57:38.454648 140460907546368 logging_writer.py:48] [42200] global_step=42200, grad_norm=4.1967854499816895, loss=3.7314062118530273 -I0513 00:58:15.953626 140460899153664 logging_writer.py:48] [42300] global_step=42300, grad_norm=7.703219890594482, loss=4.019425392150879 -I0513 00:59:02.536717 140460907546368 logging_writer.py:48] [42400] global_step=42400, grad_norm=5.010728359222412, loss=3.906130790710449 -I0513 00:59:39.572506 140460899153664 logging_writer.py:48] [42500] global_step=42500, grad_norm=4.89564847946167, loss=3.725747585296631 -I0513 01:00:17.076235 140460907546368 logging_writer.py:48] [42600] global_step=42600, grad_norm=2.032186269760132, loss=3.7578349113464355 -I0513 01:01:03.366013 140460899153664 logging_writer.py:48] [42700] global_step=42700, grad_norm=3.802997350692749, loss=3.805964231491089 -I0513 01:01:46.275435 140678261474496 spec.py:333] Evaluating on the training split. -I0513 01:01:59.938798 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 01:03:35.249082 140678261474496 spec.py:363] Evaluating on the test split. -I0513 01:03:36.150338 140678261474496 submission_runner.py:516] Time since start: 19008.63s, Step: 42791, {'train/accuracy': Array(0.00119579, dtype=float32), 'train/loss': Array(9.3953495, dtype=float32), 'validation/accuracy': Array(0.00122, dtype=float32), 'validation/loss': Array(9.384472, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0017, dtype=float32), 'test/loss': Array(9.3738985, dtype=float32), 'test/num_examples': 10000, 'score': 18027.7465133667, 'total_duration': 19008.6331281662, 'accumulated_submission_time': 18027.7465133667, 'accumulated_eval_time': 980.1608846187592, 'accumulated_logging_time': 0.2882544994354248} -I0513 01:03:36.257581 140460907546368 logging_writer.py:48] [42791] accumulated_eval_time=980.161, accumulated_logging_time=0.288254, accumulated_submission_time=18027.7, global_step=42791, preemption_count=0, score=18027.7, test/accuracy=0.0017000001389533281, test/loss=9.37389850616455, test/num_examples=10000, total_duration=19008.6, train/accuracy=0.001195790828205645, train/loss=9.395349502563477, validation/accuracy=0.0012199999764561653, validation/loss=9.384471893310547, validation/num_examples=50000 -I0513 01:03:43.924766 140460899153664 logging_writer.py:48] [42800] global_step=42800, grad_norm=5.447981834411621, loss=3.8727145195007324 -I0513 01:04:35.475013 140460907546368 logging_writer.py:48] [42900] global_step=42900, grad_norm=4.2652812004089355, loss=3.8318958282470703 -I0513 01:05:21.613171 140460899153664 logging_writer.py:48] [43000] global_step=43000, grad_norm=6.427865982055664, loss=3.817960023880005 -I0513 01:06:16.416326 140460907546368 logging_writer.py:48] [43100] global_step=43100, grad_norm=6.252673625946045, loss=3.8169503211975098 -I0513 01:07:09.876638 140460899153664 logging_writer.py:48] [43200] global_step=43200, grad_norm=3.6733779907226562, loss=3.7597343921661377 -I0513 01:07:55.772716 140460907546368 logging_writer.py:48] [43300] global_step=43300, grad_norm=8.347646713256836, loss=3.9216599464416504 -I0513 01:08:41.961760 140460899153664 logging_writer.py:48] [43400] global_step=43400, grad_norm=4.719368934631348, loss=3.7944419384002686 -I0513 01:09:37.247562 140460907546368 logging_writer.py:48] [43500] global_step=43500, grad_norm=5.887572765350342, loss=3.902421712875366 -I0513 01:10:20.492174 140460899153664 logging_writer.py:48] [43600] global_step=43600, grad_norm=3.906403064727783, loss=3.814117908477783 -I0513 01:11:03.520511 140460907546368 logging_writer.py:48] [43700] global_step=43700, grad_norm=3.461521625518799, loss=3.8109028339385986 -I0513 01:11:47.944446 140460899153664 logging_writer.py:48] [43800] global_step=43800, grad_norm=5.0287909507751465, loss=3.938202381134033 -I0513 01:12:31.896665 140460907546368 logging_writer.py:48] [43900] global_step=43900, grad_norm=4.088149547576904, loss=3.6917402744293213 -I0513 01:13:17.980838 140460899153664 logging_writer.py:48] [44000] global_step=44000, grad_norm=4.954939842224121, loss=3.8018863201141357 -I0513 01:14:04.427150 140460907546368 logging_writer.py:48] [44100] global_step=44100, grad_norm=2.7949957847595215, loss=3.7891721725463867 -I0513 01:14:50.361468 140460899153664 logging_writer.py:48] [44200] global_step=44200, grad_norm=3.8294789791107178, loss=3.752429485321045 -I0513 01:15:45.410938 140460907546368 logging_writer.py:48] [44300] global_step=44300, grad_norm=6.283016681671143, loss=3.8685967922210693 -I0513 01:16:31.953468 140460899153664 logging_writer.py:48] [44400] global_step=44400, grad_norm=4.190720081329346, loss=3.8178813457489014 -I0513 01:17:18.301238 140460907546368 logging_writer.py:48] [44500] global_step=44500, grad_norm=5.454296588897705, loss=3.8579771518707275 -I0513 01:18:04.285004 140460899153664 logging_writer.py:48] [44600] global_step=44600, grad_norm=4.728999614715576, loss=3.760692834854126 -I0513 01:18:43.679039 140460907546368 logging_writer.py:48] [44700] global_step=44700, grad_norm=5.1004791259765625, loss=3.81071400642395 -I0513 01:19:25.405631 140460899153664 logging_writer.py:48] [44800] global_step=44800, grad_norm=3.4921669960021973, loss=3.7554502487182617 -I0513 01:20:04.175639 140460907546368 logging_writer.py:48] [44900] global_step=44900, grad_norm=10.498441696166992, loss=4.073153495788574 -I0513 01:20:42.351800 140460899153664 logging_writer.py:48] [45000] global_step=45000, grad_norm=5.8211588859558105, loss=3.9375100135803223 -I0513 01:21:20.774581 140460907546368 logging_writer.py:48] [45100] global_step=45100, grad_norm=8.555037498474121, loss=3.884633779525757 -I0513 01:22:00.158367 140460899153664 logging_writer.py:48] [45200] global_step=45200, grad_norm=11.142605781555176, loss=3.945122480392456 -I0513 01:22:39.002873 140460907546368 logging_writer.py:48] [45300] global_step=45300, grad_norm=5.695137977600098, loss=3.8348207473754883 -I0513 01:23:17.363241 140460899153664 logging_writer.py:48] [45400] global_step=45400, grad_norm=6.746660232543945, loss=3.992673873901367 -I0513 01:24:13.910064 140460907546368 logging_writer.py:48] [45500] global_step=45500, grad_norm=4.641014099121094, loss=3.7507617473602295 -I0513 01:25:00.691915 140460899153664 logging_writer.py:48] [45600] global_step=45600, grad_norm=3.7546286582946777, loss=3.702951669692993 -I0513 01:25:42.388267 140460907546368 logging_writer.py:48] [45700] global_step=45700, grad_norm=4.868431091308594, loss=3.730161428451538 -I0513 01:26:23.871581 140460899153664 logging_writer.py:48] [45800] global_step=45800, grad_norm=2.769636869430542, loss=3.784857749938965 -I0513 01:27:10.343620 140460907546368 logging_writer.py:48] [45900] global_step=45900, grad_norm=3.288658380508423, loss=3.82543683052063 -I0513 01:27:56.659734 140460899153664 logging_writer.py:48] [46000] global_step=46000, grad_norm=5.128949165344238, loss=3.847562789916992 -I0513 01:28:52.484023 140460907546368 logging_writer.py:48] [46100] global_step=46100, grad_norm=4.568378925323486, loss=3.838993787765503 -I0513 01:29:48.746792 140460899153664 logging_writer.py:48] [46200] global_step=46200, grad_norm=4.484776973724365, loss=3.770026206970215 -I0513 01:30:34.998055 140460907546368 logging_writer.py:48] [46300] global_step=46300, grad_norm=5.377119541168213, loss=3.8253211975097656 -I0513 01:31:21.549924 140460899153664 logging_writer.py:48] [46400] global_step=46400, grad_norm=4.32159423828125, loss=3.914940357208252 -I0513 01:32:08.342595 140460907546368 logging_writer.py:48] [46500] global_step=46500, grad_norm=7.089667797088623, loss=3.8648884296417236 -I0513 01:32:54.650915 140460899153664 logging_writer.py:48] [46600] global_step=46600, grad_norm=3.6378302574157715, loss=3.6471128463745117 -I0513 01:33:41.217675 140460907546368 logging_writer.py:48] [46700] global_step=46700, grad_norm=3.886183261871338, loss=3.772902727127075 -I0513 01:34:28.060725 140460899153664 logging_writer.py:48] [46800] global_step=46800, grad_norm=4.5372748374938965, loss=3.7516403198242188 -I0513 01:35:10.210720 140460907546368 logging_writer.py:48] [46900] global_step=46900, grad_norm=4.272179126739502, loss=3.662111282348633 -I0513 01:35:54.224220 140460899153664 logging_writer.py:48] [47000] global_step=47000, grad_norm=3.396986961364746, loss=3.6620638370513916 -I0513 01:36:41.092548 140460907546368 logging_writer.py:48] [47100] global_step=47100, grad_norm=4.989297866821289, loss=3.759035110473633 -I0513 01:36:52.166783 140678261474496 spec.py:333] Evaluating on the training split. -I0513 01:37:05.366801 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 01:39:14.929204 140678261474496 spec.py:363] Evaluating on the test split. -I0513 01:39:15.824239 140678261474496 submission_runner.py:516] Time since start: 21148.31s, Step: 47116, {'train/accuracy': Array(0.00111607, dtype=float32), 'train/loss': Array(9.273086, dtype=float32), 'validation/accuracy': Array(0.00122, dtype=float32), 'validation/loss': Array(9.26029, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(9.251919, dtype=float32), 'test/num_examples': 10000, 'score': 20023.560445785522, 'total_duration': 21148.311631679535, 'accumulated_submission_time': 20023.560445785522, 'accumulated_eval_time': 1123.8164768218994, 'accumulated_logging_time': 0.44599175453186035} -I0513 01:39:15.851527 140460899153664 logging_writer.py:48] [47116] accumulated_eval_time=1123.82, accumulated_logging_time=0.445992, accumulated_submission_time=20023.6, global_step=47116, preemption_count=0, score=20023.6, test/accuracy=0.0012000000569969416, test/loss=9.25191879272461, test/num_examples=10000, total_duration=21148.3, train/accuracy=0.0011160713620483875, train/loss=9.273085594177246, validation/accuracy=0.0012199999764561653, validation/loss=9.260290145874023, validation/num_examples=50000 -I0513 01:39:54.009303 140460907546368 logging_writer.py:48] [47200] global_step=47200, grad_norm=6.832404136657715, loss=3.744917631149292 -I0513 01:40:36.872855 140460899153664 logging_writer.py:48] [47300] global_step=47300, grad_norm=4.326817989349365, loss=3.565221071243286 -I0513 01:41:21.575852 140460907546368 logging_writer.py:48] [47400] global_step=47400, grad_norm=4.624227046966553, loss=3.74918532371521 -I0513 01:42:05.526163 140460899153664 logging_writer.py:48] [47500] global_step=47500, grad_norm=6.337460517883301, loss=3.861135482788086 -I0513 01:42:51.463295 140460907546368 logging_writer.py:48] [47600] global_step=47600, grad_norm=4.345408916473389, loss=3.805842876434326 -I0513 01:43:38.371739 140460899153664 logging_writer.py:48] [47700] global_step=47700, grad_norm=18.72226333618164, loss=3.9768338203430176 -I0513 01:44:24.546525 140460907546368 logging_writer.py:48] [47800] global_step=47800, grad_norm=5.529379844665527, loss=3.859529495239258 -I0513 01:45:11.093474 140460899153664 logging_writer.py:48] [47900] global_step=47900, grad_norm=4.125874996185303, loss=3.9091618061065674 -I0513 01:45:57.796562 140460907546368 logging_writer.py:48] [48000] global_step=48000, grad_norm=3.4280951023101807, loss=3.7869906425476074 -I0513 01:46:44.047320 140460899153664 logging_writer.py:48] [48100] global_step=48100, grad_norm=4.180973052978516, loss=3.772630214691162 -I0513 01:47:30.691539 140460907546368 logging_writer.py:48] [48200] global_step=48200, grad_norm=5.769481182098389, loss=3.900367259979248 -I0513 01:48:15.146354 140460899153664 logging_writer.py:48] [48300] global_step=48300, grad_norm=3.8640503883361816, loss=3.9260201454162598 -I0513 01:49:18.783205 140460907546368 logging_writer.py:48] [48400] global_step=48400, grad_norm=4.092870235443115, loss=3.7737746238708496 -I0513 01:50:05.099589 140460899153664 logging_writer.py:48] [48500] global_step=48500, grad_norm=4.551623344421387, loss=3.863722324371338 -I0513 01:50:52.180829 140460907546368 logging_writer.py:48] [48600] global_step=48600, grad_norm=3.9432148933410645, loss=3.822449207305908 -I0513 01:51:45.509042 140460899153664 logging_writer.py:48] [48700] global_step=48700, grad_norm=4.103368759155273, loss=3.9234111309051514 -I0513 01:52:27.955424 140460907546368 logging_writer.py:48] [48800] global_step=48800, grad_norm=2.8575756549835205, loss=3.744086742401123 -I0513 01:53:10.869830 140460899153664 logging_writer.py:48] [48900] global_step=48900, grad_norm=4.628213882446289, loss=3.5815529823303223 -I0513 01:53:57.243566 140460907546368 logging_writer.py:48] [49000] global_step=49000, grad_norm=3.7608063220977783, loss=3.768815755844116 -I0513 01:54:52.865375 140460899153664 logging_writer.py:48] [49100] global_step=49100, grad_norm=2.960357427597046, loss=3.890984058380127 -I0513 01:55:39.839241 140460907546368 logging_writer.py:48] [49200] global_step=49200, grad_norm=3.722598075866699, loss=3.8259811401367188 -I0513 01:56:26.117118 140460899153664 logging_writer.py:48] [49300] global_step=49300, grad_norm=5.308701515197754, loss=3.8098692893981934 -I0513 01:57:12.784891 140460907546368 logging_writer.py:48] [49400] global_step=49400, grad_norm=6.365087509155273, loss=3.7662572860717773 -I0513 01:57:59.605659 140460899153664 logging_writer.py:48] [49500] global_step=49500, grad_norm=4.216049671173096, loss=3.8929853439331055 -I0513 01:58:55.325668 140460907546368 logging_writer.py:48] [49600] global_step=49600, grad_norm=4.866794109344482, loss=3.7965149879455566 -I0513 01:59:44.182487 140460899153664 logging_writer.py:48] [49700] global_step=49700, grad_norm=5.226081848144531, loss=3.9116032123565674 -I0513 02:00:30.040988 140460907546368 logging_writer.py:48] [49800] global_step=49800, grad_norm=6.308370113372803, loss=3.676058530807495 -I0513 02:01:23.529717 140460899153664 logging_writer.py:48] [49900] global_step=49900, grad_norm=5.711421966552734, loss=3.855337381362915 -I0513 02:02:09.880549 140460907546368 logging_writer.py:48] [50000] global_step=50000, grad_norm=4.361016273498535, loss=3.805171251296997 -I0513 02:03:06.243531 140460899153664 logging_writer.py:48] [50100] global_step=50100, grad_norm=7.596615791320801, loss=3.948539972305298 -I0513 02:04:01.879985 140460907546368 logging_writer.py:48] [50200] global_step=50200, grad_norm=5.141757488250732, loss=3.916696071624756 -I0513 02:04:48.217025 140460899153664 logging_writer.py:48] [50300] global_step=50300, grad_norm=9.912802696228027, loss=3.8143527507781982 -I0513 02:05:35.286169 140460907546368 logging_writer.py:48] [50400] global_step=50400, grad_norm=4.818367004394531, loss=3.808784246444702 -I0513 02:06:21.629231 140460899153664 logging_writer.py:48] [50500] global_step=50500, grad_norm=5.312907695770264, loss=3.683786153793335 -I0513 02:07:07.607119 140460907546368 logging_writer.py:48] [50600] global_step=50600, grad_norm=5.0388970375061035, loss=3.752208709716797 -I0513 02:08:03.993884 140460899153664 logging_writer.py:48] [50700] global_step=50700, grad_norm=4.804111003875732, loss=3.7868432998657227 -I0513 02:08:50.333788 140460907546368 logging_writer.py:48] [50800] global_step=50800, grad_norm=6.113360404968262, loss=3.822706699371338 -I0513 02:09:36.560109 140460899153664 logging_writer.py:48] [50900] global_step=50900, grad_norm=3.9908323287963867, loss=3.730393648147583 -I0513 02:10:23.637460 140460907546368 logging_writer.py:48] [51000] global_step=51000, grad_norm=4.214468955993652, loss=3.779170036315918 -I0513 02:11:10.038782 140460899153664 logging_writer.py:48] [51100] global_step=51100, grad_norm=6.8298540115356445, loss=3.640273332595825 -I0513 02:11:56.207425 140460907546368 logging_writer.py:48] [51200] global_step=51200, grad_norm=6.695694446563721, loss=3.7152891159057617 -I0513 02:12:32.259439 140678261474496 spec.py:333] Evaluating on the training split. -I0513 02:12:47.054980 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 02:13:49.722086 140678261474496 spec.py:363] Evaluating on the test split. -I0513 02:13:50.618864 140678261474496 submission_runner.py:516] Time since start: 23223.11s, Step: 51274, {'train/accuracy': Array(0.00105628, dtype=float32), 'train/loss': Array(9.120835, dtype=float32), 'validation/accuracy': Array(0.0012, dtype=float32), 'validation/loss': Array(9.070607, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(9.064775, dtype=float32), 'test/num_examples': 10000, 'score': 22019.915491104126, 'total_duration': 23223.106118679047, 'accumulated_submission_time': 22019.915491104126, 'accumulated_eval_time': 1202.174001455307, 'accumulated_logging_time': 0.4815821647644043} -I0513 02:13:50.648167 140460899153664 logging_writer.py:48] [51274] accumulated_eval_time=1202.17, accumulated_logging_time=0.481582, accumulated_submission_time=22019.9, global_step=51274, preemption_count=0, score=22019.9, test/accuracy=0.0010999999940395355, test/loss=9.064775466918945, test/num_examples=10000, total_duration=23223.1, train/accuracy=0.0010562818497419357, train/loss=9.120835304260254, validation/accuracy=0.0011999999405816197, validation/loss=9.07060718536377, validation/num_examples=50000 -I0513 02:14:00.993164 140460907546368 logging_writer.py:48] [51300] global_step=51300, grad_norm=5.997415542602539, loss=3.753206729888916 -I0513 02:14:46.828469 140460899153664 logging_writer.py:48] [51400] global_step=51400, grad_norm=3.6373751163482666, loss=3.827653646469116 -I0513 02:15:32.533103 140460907546368 logging_writer.py:48] [51500] global_step=51500, grad_norm=5.401013374328613, loss=3.722116231918335 -I0513 02:16:18.864876 140460899153664 logging_writer.py:48] [51600] global_step=51600, grad_norm=3.2981479167938232, loss=3.72971773147583 -I0513 02:17:13.416109 140460907546368 logging_writer.py:48] [51700] global_step=51700, grad_norm=7.08255672454834, loss=3.7786943912506104 -I0513 02:17:59.230949 140460899153664 logging_writer.py:48] [51800] global_step=51800, grad_norm=3.8073360919952393, loss=3.871431827545166 -I0513 02:18:45.903555 140460907546368 logging_writer.py:48] [51900] global_step=51900, grad_norm=6.482097148895264, loss=3.8963961601257324 -I0513 02:19:31.785724 140460899153664 logging_writer.py:48] [52000] global_step=52000, grad_norm=4.484093189239502, loss=3.88588285446167 -I0513 02:20:17.968564 140460907546368 logging_writer.py:48] [52100] global_step=52100, grad_norm=8.881542205810547, loss=3.9123291969299316 -I0513 02:21:04.837759 140460899153664 logging_writer.py:48] [52200] global_step=52200, grad_norm=17.42207908630371, loss=4.194055557250977 -I0513 02:21:51.010604 140460907546368 logging_writer.py:48] [52300] global_step=52300, grad_norm=6.307420253753662, loss=3.8497257232666016 -I0513 02:22:33.116815 140460899153664 logging_writer.py:48] [52400] global_step=52400, grad_norm=4.290313243865967, loss=3.8483197689056396 -I0513 02:23:16.797995 140460907546368 logging_writer.py:48] [52500] global_step=52500, grad_norm=3.574615716934204, loss=3.682048797607422 -I0513 02:24:00.332526 140460899153664 logging_writer.py:48] [52600] global_step=52600, grad_norm=7.169369697570801, loss=3.78555965423584 -I0513 02:24:46.538028 140460907546368 logging_writer.py:48] [52700] global_step=52700, grad_norm=6.748843193054199, loss=3.829657793045044 -I0513 02:25:33.232781 140460899153664 logging_writer.py:48] [52800] global_step=52800, grad_norm=2.6603050231933594, loss=3.7936818599700928 -I0513 02:26:19.710261 140460907546368 logging_writer.py:48] [52900] global_step=52900, grad_norm=3.650083303451538, loss=3.693831443786621 -I0513 02:27:06.011509 140460899153664 logging_writer.py:48] [53000] global_step=53000, grad_norm=4.181650638580322, loss=3.9851155281066895 -I0513 02:27:53.055302 140460907546368 logging_writer.py:48] [53100] global_step=53100, grad_norm=4.238804340362549, loss=3.776782512664795 -I0513 02:28:39.432559 140460899153664 logging_writer.py:48] [53200] global_step=53200, grad_norm=3.7153117656707764, loss=3.8432109355926514 -I0513 02:29:25.684997 140460907546368 logging_writer.py:48] [53300] global_step=53300, grad_norm=8.696478843688965, loss=3.965743064880371 -I0513 02:30:12.431402 140460899153664 logging_writer.py:48] [53400] global_step=53400, grad_norm=4.420249938964844, loss=3.854874849319458 -I0513 02:31:07.901233 140460907546368 logging_writer.py:48] [53500] global_step=53500, grad_norm=3.8035025596618652, loss=3.7050204277038574 -I0513 02:31:54.115651 140460899153664 logging_writer.py:48] [53600] global_step=53600, grad_norm=5.1033549308776855, loss=3.730306625366211 -I0513 02:32:41.135411 140460907546368 logging_writer.py:48] [53700] global_step=53700, grad_norm=4.677891254425049, loss=3.6803321838378906 -I0513 02:33:27.196168 140460899153664 logging_writer.py:48] [53800] global_step=53800, grad_norm=3.7854371070861816, loss=3.6196327209472656 -I0513 02:34:18.601960 140460907546368 logging_writer.py:48] [53900] global_step=53900, grad_norm=4.047572612762451, loss=3.6655197143554688 -I0513 02:35:03.035782 140460899153664 logging_writer.py:48] [54000] global_step=54000, grad_norm=6.06397008895874, loss=3.760516881942749 -I0513 02:35:46.574518 140460907546368 logging_writer.py:48] [54100] global_step=54100, grad_norm=5.2587432861328125, loss=3.804701089859009 -I0513 02:36:31.347620 140460899153664 logging_writer.py:48] [54200] global_step=54200, grad_norm=2.614716053009033, loss=3.831608533859253 -I0513 02:37:18.087084 140460907546368 logging_writer.py:48] [54300] global_step=54300, grad_norm=3.236624002456665, loss=3.7220349311828613 -I0513 02:38:13.649316 140460899153664 logging_writer.py:48] [54400] global_step=54400, grad_norm=4.265085697174072, loss=3.8793833255767822 -I0513 02:38:59.850718 140460907546368 logging_writer.py:48] [54500] global_step=54500, grad_norm=4.0412917137146, loss=3.6837329864501953 -I0513 02:39:46.402661 140460899153664 logging_writer.py:48] [54600] global_step=54600, grad_norm=3.2047574520111084, loss=3.8146450519561768 -I0513 02:40:32.825307 140460907546368 logging_writer.py:48] [54700] global_step=54700, grad_norm=4.472233772277832, loss=3.7942001819610596 -I0513 02:41:18.976198 140460899153664 logging_writer.py:48] [54800] global_step=54800, grad_norm=3.752873659133911, loss=3.6256046295166016 -I0513 02:42:05.730255 140460907546368 logging_writer.py:48] [54900] global_step=54900, grad_norm=4.205151557922363, loss=3.7745814323425293 -I0513 02:42:52.182307 140460899153664 logging_writer.py:48] [55000] global_step=55000, grad_norm=5.878551006317139, loss=3.764331102371216 -I0513 02:43:38.276055 140460907546368 logging_writer.py:48] [55100] global_step=55100, grad_norm=4.837996482849121, loss=3.8974976539611816 -I0513 02:44:24.885222 140460899153664 logging_writer.py:48] [55200] global_step=55200, grad_norm=5.153189659118652, loss=3.9138541221618652 -I0513 02:45:11.355227 140460907546368 logging_writer.py:48] [55300] global_step=55300, grad_norm=4.76096773147583, loss=3.895617723464966 -I0513 02:45:57.517031 140460899153664 logging_writer.py:48] [55400] global_step=55400, grad_norm=3.4384098052978516, loss=3.7294609546661377 -I0513 02:46:44.080971 140460907546368 logging_writer.py:48] [55500] global_step=55500, grad_norm=4.807402610778809, loss=3.85074520111084 -I0513 02:47:08.582380 140678261474496 spec.py:333] Evaluating on the training split. -I0513 02:47:24.317003 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 02:49:08.925206 140678261474496 spec.py:363] Evaluating on the test split. -I0513 02:49:09.820318 140678261474496 submission_runner.py:516] Time since start: 25342.31s, Step: 55564, {'train/accuracy': Array(0.00123565, dtype=float32), 'train/loss': Array(8.889726, dtype=float32), 'validation/accuracy': Array(0.00114, dtype=float32), 'validation/loss': Array(8.86706, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(8.862023, dtype=float32), 'test/num_examples': 10000, 'score': 24017.79511475563, 'total_duration': 25342.307918787003, 'accumulated_submission_time': 24017.79511475563, 'accumulated_eval_time': 1323.4103920459747, 'accumulated_logging_time': 0.5189347267150879} -I0513 02:49:09.848482 140460899153664 logging_writer.py:48] [55564] accumulated_eval_time=1323.41, accumulated_logging_time=0.518935, accumulated_submission_time=24017.8, global_step=55564, preemption_count=0, score=24017.8, test/accuracy=0.0010999999940395355, test/loss=8.86202335357666, test/num_examples=10000, total_duration=25342.3, train/accuracy=0.001235650503076613, train/loss=8.889725685119629, validation/accuracy=0.0011399999493733048, validation/loss=8.867059707641602, validation/num_examples=50000 -I0513 02:49:32.572975 140460907546368 logging_writer.py:48] [55600] global_step=55600, grad_norm=4.4634857177734375, loss=3.7740249633789062 -I0513 02:50:18.670835 140460899153664 logging_writer.py:48] [55700] global_step=55700, grad_norm=2.7351174354553223, loss=3.6790554523468018 -I0513 02:51:05.090251 140460907546368 logging_writer.py:48] [55800] global_step=55800, grad_norm=5.315035820007324, loss=3.7407140731811523 -I0513 02:51:45.311812 140460899153664 logging_writer.py:48] [55900] global_step=55900, grad_norm=6.3456549644470215, loss=4.134793281555176 -I0513 02:52:51.107295 140460907546368 logging_writer.py:48] [56000] global_step=56000, grad_norm=2.9733872413635254, loss=3.683293342590332 -I0513 02:53:46.100439 140460899153664 logging_writer.py:48] [56100] global_step=56100, grad_norm=3.6270089149475098, loss=3.840183734893799 -I0513 02:54:31.990758 140460907546368 logging_writer.py:48] [56200] global_step=56200, grad_norm=3.083057403564453, loss=3.800469398498535 -I0513 02:55:18.122631 140460899153664 logging_writer.py:48] [56300] global_step=56300, grad_norm=3.1570892333984375, loss=3.856088161468506 -I0513 02:56:04.490120 140460907546368 logging_writer.py:48] [56400] global_step=56400, grad_norm=4.666184425354004, loss=3.705714225769043 -I0513 02:56:55.915575 140460899153664 logging_writer.py:48] [56500] global_step=56500, grad_norm=3.97646164894104, loss=3.778745651245117 -I0513 02:57:37.251243 140460907546368 logging_writer.py:48] [56600] global_step=56600, grad_norm=6.380198955535889, loss=3.946932554244995 -I0513 02:58:22.467580 140460899153664 logging_writer.py:48] [56700] global_step=56700, grad_norm=2.279340982437134, loss=3.611992359161377 -I0513 02:59:08.630929 140460907546368 logging_writer.py:48] [56800] global_step=56800, grad_norm=15.81696891784668, loss=3.7256007194519043 -I0513 02:59:54.438263 140460899153664 logging_writer.py:48] [56900] global_step=56900, grad_norm=5.744612216949463, loss=3.8674263954162598 -I0513 03:00:40.718140 140460907546368 logging_writer.py:48] [57000] global_step=57000, grad_norm=5.404701232910156, loss=3.7397842407226562 -I0513 03:01:26.456470 140460899153664 logging_writer.py:48] [57100] global_step=57100, grad_norm=11.748214721679688, loss=3.904668092727661 -I0513 03:02:12.434564 140460907546368 logging_writer.py:48] [57200] global_step=57200, grad_norm=3.3702237606048584, loss=3.804692268371582 -I0513 03:02:58.707854 140460899153664 logging_writer.py:48] [57300] global_step=57300, grad_norm=4.166262626647949, loss=3.743884563446045 -I0513 03:03:44.781401 140460907546368 logging_writer.py:48] [57400] global_step=57400, grad_norm=2.439687728881836, loss=3.703348159790039 -I0513 03:04:30.584134 140460899153664 logging_writer.py:48] [57500] global_step=57500, grad_norm=4.68045711517334, loss=3.7472949028015137 -I0513 03:05:16.879769 140460907546368 logging_writer.py:48] [57600] global_step=57600, grad_norm=3.643768548965454, loss=3.5432920455932617 -I0513 03:06:02.578012 140460899153664 logging_writer.py:48] [57700] global_step=57700, grad_norm=4.630345344543457, loss=3.64481520652771 -I0513 03:06:48.531803 140460907546368 logging_writer.py:48] [57800] global_step=57800, grad_norm=3.5432982444763184, loss=3.8304555416107178 -I0513 03:07:43.595034 140460899153664 logging_writer.py:48] [57900] global_step=57900, grad_norm=6.250283241271973, loss=3.9331765174865723 -I0513 03:08:29.899630 140460907546368 logging_writer.py:48] [58000] global_step=58000, grad_norm=4.890139102935791, loss=3.812495708465576 -I0513 03:09:15.620361 140460899153664 logging_writer.py:48] [58100] global_step=58100, grad_norm=4.8640360832214355, loss=3.7665772438049316 -I0513 03:10:01.966044 140460907546368 logging_writer.py:48] [58200] global_step=58200, grad_norm=4.699387073516846, loss=3.6437220573425293 -I0513 03:10:47.671343 140460899153664 logging_writer.py:48] [58300] global_step=58300, grad_norm=3.3976552486419678, loss=3.8009238243103027 -I0513 03:11:33.896931 140460907546368 logging_writer.py:48] [58400] global_step=58400, grad_norm=2.862072229385376, loss=3.6326563358306885 -I0513 03:12:20.084682 140460899153664 logging_writer.py:48] [58500] global_step=58500, grad_norm=4.855384349822998, loss=3.8007121086120605 -I0513 03:13:05.640686 140460907546368 logging_writer.py:48] [58600] global_step=58600, grad_norm=3.910191297531128, loss=3.6625099182128906 -I0513 03:13:51.659805 140460899153664 logging_writer.py:48] [58700] global_step=58700, grad_norm=4.3660783767700195, loss=3.7696330547332764 -I0513 03:14:41.725401 140460907546368 logging_writer.py:48] [58800] global_step=58800, grad_norm=5.028444766998291, loss=3.8393783569335938 -I0513 03:15:24.258992 140460899153664 logging_writer.py:48] [58900] global_step=58900, grad_norm=3.836911916732788, loss=3.622610569000244 -I0513 03:16:08.998301 140460907546368 logging_writer.py:48] [59000] global_step=59000, grad_norm=3.866872549057007, loss=3.7101352214813232 -I0513 03:16:55.297270 140460899153664 logging_writer.py:48] [59100] global_step=59100, grad_norm=2.9768505096435547, loss=3.766498565673828 -I0513 03:17:41.154295 140460907546368 logging_writer.py:48] [59200] global_step=59200, grad_norm=5.2808332443237305, loss=3.701561689376831 -I0513 03:18:35.943044 140460899153664 logging_writer.py:48] [59300] global_step=59300, grad_norm=3.63069224357605, loss=3.8178389072418213 -I0513 03:19:22.228562 140460907546368 logging_writer.py:48] [59400] global_step=59400, grad_norm=3.7755401134490967, loss=3.7449190616607666 -I0513 03:20:07.971770 140460899153664 logging_writer.py:48] [59500] global_step=59500, grad_norm=5.561481475830078, loss=3.697779655456543 -I0513 03:20:54.218606 140460907546368 logging_writer.py:48] [59600] global_step=59600, grad_norm=3.6673524379730225, loss=3.7922568321228027 -I0513 03:21:40.407632 140460899153664 logging_writer.py:48] [59700] global_step=59700, grad_norm=3.8533496856689453, loss=3.8511834144592285 -I0513 03:22:25.943982 140678261474496 spec.py:333] Evaluating on the training split. -I0513 03:22:38.421871 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 03:24:06.534821 140678261474496 spec.py:363] Evaluating on the test split. -I0513 03:24:07.430325 140678261474496 submission_runner.py:516] Time since start: 27439.92s, Step: 59800, {'train/accuracy': Array(0.00115593, dtype=float32), 'train/loss': Array(8.759222, dtype=float32), 'validation/accuracy': Array(0.00118, dtype=float32), 'validation/loss': Array(8.7175255, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(8.709208, dtype=float32), 'test/num_examples': 10000, 'score': 26013.837624788284, 'total_duration': 27439.917838335037, 'accumulated_submission_time': 26013.837624788284, 'accumulated_eval_time': 1424.8949935436249, 'accumulated_logging_time': 0.5556581020355225} -I0513 03:24:07.458022 140460907546368 logging_writer.py:48] [59800] accumulated_eval_time=1424.89, accumulated_logging_time=0.555658, accumulated_submission_time=26013.8, global_step=59800, preemption_count=0, score=26013.8, test/accuracy=0.0012000000569969416, test/loss=8.709207534790039, test/num_examples=10000, total_duration=27439.9, train/accuracy=0.0011559311533346772, train/loss=8.759222030639648, validation/accuracy=0.001180000021122396, validation/loss=8.717525482177734, validation/num_examples=50000 -I0513 03:24:07.736692 140460899153664 logging_writer.py:48] [59800] global_step=59800, grad_norm=4.204706192016602, loss=3.7292628288269043 -I0513 03:25:03.074087 140460907546368 logging_writer.py:48] [59900] global_step=59900, grad_norm=3.799926519393921, loss=3.8159544467926025 -I0513 03:25:49.967028 140460899153664 logging_writer.py:48] [60000] global_step=60000, grad_norm=7.622490406036377, loss=3.9677414894104004 -I0513 03:26:45.486905 140460907546368 logging_writer.py:48] [60100] global_step=60100, grad_norm=4.663120269775391, loss=3.917421340942383 -I0513 03:27:32.065788 140460899153664 logging_writer.py:48] [60200] global_step=60200, grad_norm=3.5187582969665527, loss=3.6891870498657227 -I0513 03:28:18.858431 140460907546368 logging_writer.py:48] [60300] global_step=60300, grad_norm=4.310642242431641, loss=3.770503520965576 -I0513 03:29:05.059346 140460899153664 logging_writer.py:48] [60400] global_step=60400, grad_norm=3.864774703979492, loss=3.7426414489746094 -I0513 03:29:51.511772 140460907546368 logging_writer.py:48] [60500] global_step=60500, grad_norm=4.029377460479736, loss=3.8578402996063232 -I0513 03:30:47.466226 140460899153664 logging_writer.py:48] [60600] global_step=60600, grad_norm=5.539844512939453, loss=3.7441580295562744 -I0513 03:31:33.304947 140460907546368 logging_writer.py:48] [60700] global_step=60700, grad_norm=4.614203929901123, loss=3.8338327407836914 -I0513 03:32:17.530127 140460899153664 logging_writer.py:48] [60800] global_step=60800, grad_norm=3.4088134765625, loss=3.8135251998901367 -I0513 03:33:03.956914 140460907546368 logging_writer.py:48] [60900] global_step=60900, grad_norm=3.5422768592834473, loss=3.503687620162964 -I0513 03:33:52.223192 140460899153664 logging_writer.py:48] [61000] global_step=61000, grad_norm=6.456978797912598, loss=3.930964231491089 -I0513 03:34:33.888045 140460907546368 logging_writer.py:48] [61100] global_step=61100, grad_norm=7.9468770027160645, loss=3.606156349182129 -I0513 03:35:40.973992 140460899153664 logging_writer.py:48] [61200] global_step=61200, grad_norm=4.303416728973389, loss=3.860654354095459 -I0513 03:36:27.340309 140460907546368 logging_writer.py:48] [61300] global_step=61300, grad_norm=4.894606590270996, loss=3.9042303562164307 -I0513 03:37:13.828854 140460899153664 logging_writer.py:48] [61400] global_step=61400, grad_norm=4.026729583740234, loss=3.721092462539673 -I0513 03:38:00.612858 140460907546368 logging_writer.py:48] [61500] global_step=61500, grad_norm=5.102956771850586, loss=3.8561363220214844 -I0513 03:38:47.175009 140460899153664 logging_writer.py:48] [61600] global_step=61600, grad_norm=5.003881931304932, loss=3.675478935241699 -I0513 03:39:33.928445 140460907546368 logging_writer.py:48] [61700] global_step=61700, grad_norm=2.3613154888153076, loss=3.4683690071105957 -I0513 03:40:20.802944 140460899153664 logging_writer.py:48] [61800] global_step=61800, grad_norm=3.015258550643921, loss=3.7269201278686523 -I0513 03:41:07.212747 140460907546368 logging_writer.py:48] [61900] global_step=61900, grad_norm=4.79572057723999, loss=3.6979150772094727 -I0513 03:41:53.543453 140460899153664 logging_writer.py:48] [62000] global_step=62000, grad_norm=3.59104323387146, loss=3.6855976581573486 -I0513 03:42:40.507769 140460907546368 logging_writer.py:48] [62100] global_step=62100, grad_norm=4.2161383628845215, loss=3.8890209197998047 -I0513 03:43:22.692959 140460899153664 logging_writer.py:48] [62200] global_step=62200, grad_norm=3.109468460083008, loss=3.7002079486846924 -I0513 03:44:03.725793 140460907546368 logging_writer.py:48] [62300] global_step=62300, grad_norm=3.8023016452789307, loss=3.806339979171753 -I0513 03:44:50.001465 140460899153664 logging_writer.py:48] [62400] global_step=62400, grad_norm=3.8758509159088135, loss=3.6326420307159424 -I0513 03:45:36.297317 140460907546368 logging_writer.py:48] [62500] global_step=62500, grad_norm=7.9723219871521, loss=3.7662007808685303 -I0513 03:46:23.149676 140460899153664 logging_writer.py:48] [62600] global_step=62600, grad_norm=4.664536952972412, loss=3.680272102355957 -I0513 03:47:10.019820 140460907546368 logging_writer.py:48] [62700] global_step=62700, grad_norm=3.5968661308288574, loss=3.769742250442505 -I0513 03:47:56.438044 140460899153664 logging_writer.py:48] [62800] global_step=62800, grad_norm=3.265793800354004, loss=3.6144964694976807 -I0513 03:48:43.008448 140460907546368 logging_writer.py:48] [62900] global_step=62900, grad_norm=4.606043815612793, loss=3.6293740272521973 -I0513 03:49:29.775199 140460899153664 logging_writer.py:48] [63000] global_step=63000, grad_norm=3.969820022583008, loss=3.834901809692383 -I0513 03:50:16.181755 140460907546368 logging_writer.py:48] [63100] global_step=63100, grad_norm=3.7590115070343018, loss=3.8531126976013184 -I0513 03:51:02.877076 140460899153664 logging_writer.py:48] [63200] global_step=63200, grad_norm=6.993534088134766, loss=3.9065935611724854 -I0513 03:51:49.818367 140460907546368 logging_writer.py:48] [63300] global_step=63300, grad_norm=3.7658069133758545, loss=3.815110683441162 -I0513 03:52:45.530539 140460899153664 logging_writer.py:48] [63400] global_step=63400, grad_norm=4.632556438446045, loss=3.903975009918213 -I0513 03:53:32.208548 140460907546368 logging_writer.py:48] [63500] global_step=63500, grad_norm=5.610090732574463, loss=3.712120294570923 -I0513 03:54:16.594125 140460899153664 logging_writer.py:48] [63600] global_step=63600, grad_norm=2.7733194828033447, loss=3.7926137447357178 -I0513 03:55:14.291579 140460907546368 logging_writer.py:48] [63700] global_step=63700, grad_norm=4.795338153839111, loss=3.8294944763183594 -I0513 03:56:00.838511 140460899153664 logging_writer.py:48] [63800] global_step=63800, grad_norm=4.222810745239258, loss=3.8840267658233643 -I0513 03:56:47.686277 140460907546368 logging_writer.py:48] [63900] global_step=63900, grad_norm=9.03267765045166, loss=3.7262558937072754 -I0513 03:57:25.604200 140678261474496 spec.py:333] Evaluating on the training split. -I0513 03:57:39.244434 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 03:59:01.623477 140678261474496 spec.py:363] Evaluating on the test split. -I0513 03:59:02.515921 140678261474496 submission_runner.py:516] Time since start: 29535.00s, Step: 63964, {'train/accuracy': Array(0.00111607, dtype=float32), 'train/loss': Array(8.594898, dtype=float32), 'validation/accuracy': Array(0.00118, dtype=float32), 'validation/loss': Array(8.576055, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(8.567521, dtype=float32), 'test/num_examples': 10000, 'score': 28011.93083548546, 'total_duration': 29535.00321173668, 'accumulated_submission_time': 28011.93083548546, 'accumulated_eval_time': 1521.804844379425, 'accumulated_logging_time': 0.5917441844940186} -I0513 03:59:02.546496 140460899153664 logging_writer.py:48] [63964] accumulated_eval_time=1521.8, accumulated_logging_time=0.591744, accumulated_submission_time=28011.9, global_step=63964, preemption_count=0, score=28011.9, test/accuracy=0.0010999999940395355, test/loss=8.567521095275879, test/num_examples=10000, total_duration=29535, train/accuracy=0.0011160713620483875, train/loss=8.594898223876953, validation/accuracy=0.001180000021122396, validation/loss=8.576054573059082, validation/num_examples=50000 -I0513 03:59:24.941248 140460907546368 logging_writer.py:48] [64000] global_step=64000, grad_norm=4.048264980316162, loss=3.8158226013183594 -I0513 04:00:04.840629 140460899153664 logging_writer.py:48] [64100] global_step=64100, grad_norm=3.566314697265625, loss=3.7770402431488037 -I0513 04:00:51.049752 140460907546368 logging_writer.py:48] [64200] global_step=64200, grad_norm=3.584942579269409, loss=3.8664166927337646 -I0513 04:01:45.420659 140460899153664 logging_writer.py:48] [64300] global_step=64300, grad_norm=3.529787302017212, loss=3.8055853843688965 -I0513 04:02:31.232782 140460907546368 logging_writer.py:48] [64400] global_step=64400, grad_norm=3.6388275623321533, loss=3.7487406730651855 -I0513 04:03:17.419739 140460899153664 logging_writer.py:48] [64500] global_step=64500, grad_norm=5.141243934631348, loss=3.8210458755493164 -I0513 04:04:03.233556 140460907546368 logging_writer.py:48] [64600] global_step=64600, grad_norm=6.031074523925781, loss=3.845547676086426 -I0513 04:04:48.902084 140460899153664 logging_writer.py:48] [64700] global_step=64700, grad_norm=7.433409214019775, loss=3.8090343475341797 -I0513 04:05:44.041492 140460907546368 logging_writer.py:48] [64800] global_step=64800, grad_norm=3.7980964183807373, loss=3.6756014823913574 -I0513 04:06:29.905572 140460899153664 logging_writer.py:48] [64900] global_step=64900, grad_norm=6.219339847564697, loss=3.8385000228881836 -I0513 04:07:07.794327 140460907546368 logging_writer.py:48] [65000] global_step=65000, grad_norm=7.985232830047607, loss=3.776777744293213 -I0513 04:07:45.847901 140460899153664 logging_writer.py:48] [65100] global_step=65100, grad_norm=8.585768699645996, loss=3.9834442138671875 -I0513 04:08:41.155647 140460907546368 logging_writer.py:48] [65200] global_step=65200, grad_norm=4.419251918792725, loss=3.8014626502990723 -I0513 04:09:27.373234 140460899153664 logging_writer.py:48] [65300] global_step=65300, grad_norm=3.4499142169952393, loss=3.818742513656616 -I0513 04:10:13.814723 140460907546368 logging_writer.py:48] [65400] global_step=65400, grad_norm=5.642630100250244, loss=3.7007460594177246 -I0513 04:10:59.834942 140460899153664 logging_writer.py:48] [65500] global_step=65500, grad_norm=3.696897506713867, loss=3.8837571144104004 -I0513 04:11:55.233279 140460907546368 logging_writer.py:48] [65600] global_step=65600, grad_norm=5.461510181427002, loss=3.709198474884033 -I0513 04:12:42.112448 140460899153664 logging_writer.py:48] [65700] global_step=65700, grad_norm=3.0618791580200195, loss=3.6945230960845947 -I0513 04:13:28.244452 140460907546368 logging_writer.py:48] [65800] global_step=65800, grad_norm=4.177756309509277, loss=3.7573211193084717 -I0513 04:14:14.038808 140460899153664 logging_writer.py:48] [65900] global_step=65900, grad_norm=3.5920650959014893, loss=3.634028196334839 -I0513 04:15:00.748399 140460907546368 logging_writer.py:48] [66000] global_step=66000, grad_norm=3.3164358139038086, loss=3.8125672340393066 -I0513 04:15:46.947020 140460899153664 logging_writer.py:48] [66100] global_step=66100, grad_norm=5.272975921630859, loss=3.7531018257141113 -I0513 04:16:33.057293 140460907546368 logging_writer.py:48] [66200] global_step=66200, grad_norm=2.8587307929992676, loss=3.718118906021118 -I0513 04:17:20.054669 140460899153664 logging_writer.py:48] [66300] global_step=66300, grad_norm=5.677445411682129, loss=3.7502689361572266 -I0513 04:18:06.294971 140460907546368 logging_writer.py:48] [66400] global_step=66400, grad_norm=4.8133320808410645, loss=3.8094987869262695 -I0513 04:19:02.053453 140460899153664 logging_writer.py:48] [66500] global_step=66500, grad_norm=5.028270721435547, loss=3.704474925994873 -I0513 04:19:48.705387 140460907546368 logging_writer.py:48] [66600] global_step=66600, grad_norm=5.443061351776123, loss=3.7748091220855713 -I0513 04:20:34.911128 140460899153664 logging_writer.py:48] [66700] global_step=66700, grad_norm=8.276126861572266, loss=3.76920485496521 -I0513 04:21:21.051786 140460907546368 logging_writer.py:48] [66800] global_step=66800, grad_norm=3.7410905361175537, loss=3.7418787479400635 -I0513 04:22:08.026700 140460899153664 logging_writer.py:48] [66900] global_step=66900, grad_norm=3.8876821994781494, loss=3.7240805625915527 -I0513 04:22:54.180085 140460907546368 logging_writer.py:48] [67000] global_step=67000, grad_norm=5.28666353225708, loss=3.8732807636260986 -I0513 04:23:40.402351 140460899153664 logging_writer.py:48] [67100] global_step=67100, grad_norm=4.207780361175537, loss=3.727139472961426 -I0513 04:24:27.412671 140460907546368 logging_writer.py:48] [67200] global_step=67200, grad_norm=6.599878311157227, loss=3.8276405334472656 -I0513 04:25:22.829386 140460899153664 logging_writer.py:48] [67300] global_step=67300, grad_norm=4.221244812011719, loss=3.7887351512908936 -I0513 04:26:08.486755 140460907546368 logging_writer.py:48] [67400] global_step=67400, grad_norm=3.804572105407715, loss=3.76881742477417 -I0513 04:26:55.028655 140460899153664 logging_writer.py:48] [67500] global_step=67500, grad_norm=3.5643599033355713, loss=3.6097774505615234 -I0513 04:27:38.558176 140460907546368 logging_writer.py:48] [67600] global_step=67600, grad_norm=23.789852142333984, loss=3.861812114715576 -I0513 04:28:21.133064 140460899153664 logging_writer.py:48] [67700] global_step=67700, grad_norm=58.84666442871094, loss=3.9760637283325195 -I0513 04:29:07.352475 140460907546368 logging_writer.py:48] [67800] global_step=67800, grad_norm=4.154022216796875, loss=3.7539961338043213 -I0513 04:29:49.643556 140460899153664 logging_writer.py:48] [67900] global_step=67900, grad_norm=5.0382890701293945, loss=3.6225218772888184 -I0513 04:30:35.296504 140460907546368 logging_writer.py:48] [68000] global_step=68000, grad_norm=4.902912616729736, loss=3.8099708557128906 -I0513 04:31:22.217030 140460899153664 logging_writer.py:48] [68100] global_step=68100, grad_norm=4.587691307067871, loss=3.6438677310943604 -I0513 04:32:08.357784 140460907546368 logging_writer.py:48] [68200] global_step=68200, grad_norm=4.345426082611084, loss=3.820115327835083 -I0513 04:32:18.561894 140678261474496 spec.py:333] Evaluating on the training split. -I0513 04:32:33.640392 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 04:34:12.634945 140678261474496 spec.py:363] Evaluating on the test split. -I0513 04:34:13.526051 140678261474496 submission_runner.py:516] Time since start: 31646.01s, Step: 68219, {'train/accuracy': Array(0.00109614, dtype=float32), 'train/loss': Array(8.4948015, dtype=float32), 'validation/accuracy': Array(0.00116, dtype=float32), 'validation/loss': Array(8.460207, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(8.452573, dtype=float32), 'test/num_examples': 10000, 'score': 30007.89012002945, 'total_duration': 31646.01367497444, 'accumulated_submission_time': 30007.89012002945, 'accumulated_eval_time': 1636.7673811912537, 'accumulated_logging_time': 0.634131669998169} -I0513 04:34:13.556080 140460899153664 logging_writer.py:48] [68219] accumulated_eval_time=1636.77, accumulated_logging_time=0.634132, accumulated_submission_time=30007.9, global_step=68219, preemption_count=0, score=30007.9, test/accuracy=0.0013000001199543476, test/loss=8.4525728225708, test/num_examples=10000, total_duration=31646, train/accuracy=0.0010961415246129036, train/loss=8.49480152130127, validation/accuracy=0.0011599999852478504, validation/loss=8.460206985473633, validation/num_examples=50000 -I0513 04:34:45.378769 140460907546368 logging_writer.py:48] [68300] global_step=68300, grad_norm=6.456268787384033, loss=3.8530197143554688 -I0513 04:35:26.293247 140460899153664 logging_writer.py:48] [68400] global_step=68400, grad_norm=4.196746349334717, loss=3.7709760665893555 -I0513 04:36:19.498997 140460907546368 logging_writer.py:48] [68500] global_step=68500, grad_norm=14.037670135498047, loss=4.396938800811768 -I0513 04:37:05.697050 140460899153664 logging_writer.py:48] [68600] global_step=68600, grad_norm=25.198284149169922, loss=3.8518214225769043 -I0513 04:37:52.884270 140460907546368 logging_writer.py:48] [68700] global_step=68700, grad_norm=5.746915340423584, loss=3.865098714828491 -I0513 04:38:41.336701 140460899153664 logging_writer.py:48] [68800] global_step=68800, grad_norm=5.172417640686035, loss=3.7555530071258545 -I0513 04:39:34.940364 140460907546368 logging_writer.py:48] [68900] global_step=68900, grad_norm=4.711776256561279, loss=3.8937463760375977 -I0513 04:40:12.911865 140460899153664 logging_writer.py:48] [69000] global_step=69000, grad_norm=4.458269119262695, loss=3.892524003982544 -I0513 04:40:57.195884 140460907546368 logging_writer.py:48] [69100] global_step=69100, grad_norm=4.589003086090088, loss=3.7854061126708984 -I0513 04:41:42.900934 140460899153664 logging_writer.py:48] [69200] global_step=69200, grad_norm=2.936908006668091, loss=3.662104606628418 -I0513 04:42:29.374726 140460907546368 logging_writer.py:48] [69300] global_step=69300, grad_norm=3.924611806869507, loss=3.706124782562256 -I0513 04:43:13.385520 140460899153664 logging_writer.py:48] [69400] global_step=69400, grad_norm=6.079455375671387, loss=3.8365232944488525 -I0513 04:43:57.389363 140460907546368 logging_writer.py:48] [69500] global_step=69500, grad_norm=6.102921009063721, loss=3.6591341495513916 -I0513 04:44:42.225801 140460899153664 logging_writer.py:48] [69600] global_step=69600, grad_norm=2.7949719429016113, loss=3.705169916152954 -I0513 04:45:26.244968 140460907546368 logging_writer.py:48] [69700] global_step=69700, grad_norm=4.458542346954346, loss=3.657015085220337 -I0513 04:46:10.197953 140460899153664 logging_writer.py:48] [69800] global_step=69800, grad_norm=3.2554099559783936, loss=3.6579856872558594 -I0513 04:46:56.355839 140460907546368 logging_writer.py:48] [69900] global_step=69900, grad_norm=4.315677165985107, loss=3.795191526412964 -I0513 04:47:41.757829 140460899153664 logging_writer.py:48] [70000] global_step=70000, grad_norm=6.058112144470215, loss=3.7772717475891113 -I0513 04:48:27.514148 140460907546368 logging_writer.py:48] [70100] global_step=70100, grad_norm=6.8558783531188965, loss=3.810713052749634 -I0513 04:49:12.378256 140460899153664 logging_writer.py:48] [70200] global_step=70200, grad_norm=5.599267482757568, loss=3.7263646125793457 -I0513 04:49:54.039910 140460907546368 logging_writer.py:48] [70300] global_step=70300, grad_norm=4.144339561462402, loss=3.779921531677246 -I0513 04:50:35.601670 140460899153664 logging_writer.py:48] [70400] global_step=70400, grad_norm=11.134873390197754, loss=3.8345460891723633 -I0513 04:51:16.165416 140460907546368 logging_writer.py:48] [70500] global_step=70500, grad_norm=7.600071907043457, loss=3.8803491592407227 -I0513 04:52:00.409375 140460899153664 logging_writer.py:48] [70600] global_step=70600, grad_norm=5.554253578186035, loss=3.9285120964050293 -I0513 04:52:39.841438 140460907546368 logging_writer.py:48] [70700] global_step=70700, grad_norm=5.30936336517334, loss=3.759115695953369 -I0513 04:53:19.634781 140460899153664 logging_writer.py:48] [70800] global_step=70800, grad_norm=3.8918778896331787, loss=3.872467279434204 -I0513 04:53:58.511859 140460907546368 logging_writer.py:48] [70900] global_step=70900, grad_norm=4.754245758056641, loss=3.7908120155334473 -I0513 04:54:37.891293 140460899153664 logging_writer.py:48] [71000] global_step=71000, grad_norm=4.992252349853516, loss=3.7777955532073975 -I0513 04:55:23.982043 140460907546368 logging_writer.py:48] [71100] global_step=71100, grad_norm=5.4390645027160645, loss=3.76515531539917 -I0513 04:56:09.941351 140460899153664 logging_writer.py:48] [71200] global_step=71200, grad_norm=6.451941967010498, loss=3.866053581237793 -I0513 04:56:55.597547 140460907546368 logging_writer.py:48] [71300] global_step=71300, grad_norm=10.574663162231445, loss=3.7253551483154297 -I0513 04:57:41.698736 140460899153664 logging_writer.py:48] [71400] global_step=71400, grad_norm=8.87796688079834, loss=3.947399616241455 -I0513 04:58:20.828101 140460907546368 logging_writer.py:48] [71500] global_step=71500, grad_norm=3.1193971633911133, loss=3.731412887573242 -I0513 04:59:00.654691 140460899153664 logging_writer.py:48] [71600] global_step=71600, grad_norm=4.121596336364746, loss=3.8004984855651855 -I0513 04:59:40.750765 140460907546368 logging_writer.py:48] [71700] global_step=71700, grad_norm=4.250677108764648, loss=3.903663396835327 -I0513 05:00:20.503557 140460899153664 logging_writer.py:48] [71800] global_step=71800, grad_norm=3.676924705505371, loss=3.769554615020752 -I0513 05:00:59.844934 140460907546368 logging_writer.py:48] [71900] global_step=71900, grad_norm=3.579944133758545, loss=3.7754082679748535 -I0513 05:01:39.791258 140460899153664 logging_writer.py:48] [72000] global_step=72000, grad_norm=3.859290599822998, loss=3.88533353805542 -I0513 05:02:19.428753 140460907546368 logging_writer.py:48] [72100] global_step=72100, grad_norm=6.856759548187256, loss=4.025289535522461 -I0513 05:02:58.751906 140460899153664 logging_writer.py:48] [72200] global_step=72200, grad_norm=3.3386552333831787, loss=3.677476167678833 -I0513 05:03:38.052157 140460907546368 logging_writer.py:48] [72300] global_step=72300, grad_norm=10.570206642150879, loss=3.72707200050354 -I0513 05:04:18.144305 140460899153664 logging_writer.py:48] [72400] global_step=72400, grad_norm=5.859450340270996, loss=3.7569565773010254 -I0513 05:04:59.543612 140460907546368 logging_writer.py:48] [72500] global_step=72500, grad_norm=4.95604133605957, loss=3.808838367462158 -I0513 05:05:39.797562 140460899153664 logging_writer.py:48] [72600] global_step=72600, grad_norm=4.928428649902344, loss=3.819654941558838 -I0513 05:06:19.468836 140460907546368 logging_writer.py:48] [72700] global_step=72700, grad_norm=4.129063129425049, loss=3.805621385574341 -I0513 05:06:59.454168 140460899153664 logging_writer.py:48] [72800] global_step=72800, grad_norm=6.9960222244262695, loss=3.8580751419067383 -I0513 05:07:29.660711 140678261474496 spec.py:333] Evaluating on the training split. -I0513 05:07:42.758689 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 05:08:43.070502 140678261474496 spec.py:363] Evaluating on the test split. -I0513 05:08:43.967678 140678261474496 submission_runner.py:516] Time since start: 33716.45s, Step: 72875, {'train/accuracy': Array(0.00111607, dtype=float32), 'train/loss': Array(8.365174, dtype=float32), 'validation/accuracy': Array(0.00116, dtype=float32), 'validation/loss': Array(8.342442, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0016, dtype=float32), 'test/loss': Array(8.336133, dtype=float32), 'test/num_examples': 10000, 'score': 32003.9418694973, 'total_duration': 33716.45394611359, 'accumulated_submission_time': 32003.9418694973, 'accumulated_eval_time': 1711.0713665485382, 'accumulated_logging_time': 0.6724264621734619} -I0513 05:08:44.003050 140460907546368 logging_writer.py:48] [72875] accumulated_eval_time=1711.07, accumulated_logging_time=0.672426, accumulated_submission_time=32003.9, global_step=72875, preemption_count=0, score=32003.9, test/accuracy=0.001600000075995922, test/loss=8.336133003234863, test/num_examples=10000, total_duration=33716.5, train/accuracy=0.0011160713620483875, train/loss=8.365174293518066, validation/accuracy=0.0011599999852478504, validation/loss=8.34244155883789, validation/num_examples=50000 -I0513 05:08:54.095791 140460899153664 logging_writer.py:48] [72900] global_step=72900, grad_norm=6.949901580810547, loss=3.73588228225708 -I0513 05:09:40.278829 140460907546368 logging_writer.py:48] [73000] global_step=73000, grad_norm=5.747148036956787, loss=4.002292156219482 -I0513 05:10:26.038344 140460899153664 logging_writer.py:48] [73100] global_step=73100, grad_norm=4.2536749839782715, loss=3.772477865219116 -I0513 05:11:12.164683 140460907546368 logging_writer.py:48] [73200] global_step=73200, grad_norm=3.6861319541931152, loss=3.722146987915039 -I0513 05:11:58.121515 140460899153664 logging_writer.py:48] [73300] global_step=73300, grad_norm=4.988461494445801, loss=3.691791534423828 -I0513 05:12:43.906285 140460907546368 logging_writer.py:48] [73400] global_step=73400, grad_norm=5.622732162475586, loss=3.84716796875 -I0513 05:13:30.044382 140460899153664 logging_writer.py:48] [73500] global_step=73500, grad_norm=5.591344833374023, loss=3.779764413833618 -I0513 05:14:15.869214 140460907546368 logging_writer.py:48] [73600] global_step=73600, grad_norm=5.8094377517700195, loss=3.6656336784362793 -I0513 05:14:58.919337 140460899153664 logging_writer.py:48] [73700] global_step=73700, grad_norm=4.527812480926514, loss=3.990389108657837 -I0513 05:15:43.918076 140460907546368 logging_writer.py:48] [73800] global_step=73800, grad_norm=4.641282081604004, loss=3.7413108348846436 -I0513 05:16:41.862379 140460899153664 logging_writer.py:48] [73900] global_step=73900, grad_norm=3.6104986667633057, loss=3.7720069885253906 -I0513 05:17:20.892402 140460907546368 logging_writer.py:48] [74000] global_step=74000, grad_norm=4.16671085357666, loss=3.735189437866211 -I0513 05:18:23.097781 140460899153664 logging_writer.py:48] [74100] global_step=74100, grad_norm=6.472297191619873, loss=3.8148915767669678 -I0513 05:18:59.994098 140460907546368 logging_writer.py:48] [74200] global_step=74200, grad_norm=3.5125489234924316, loss=3.797060489654541 -I0513 05:19:37.436551 140460899153664 logging_writer.py:48] [74300] global_step=74300, grad_norm=4.1699442863464355, loss=3.645820140838623 -I0513 05:20:24.239800 140460907546368 logging_writer.py:48] [74400] global_step=74400, grad_norm=4.987982273101807, loss=3.8461716175079346 -I0513 05:21:03.329883 140460899153664 logging_writer.py:48] [74500] global_step=74500, grad_norm=6.044218063354492, loss=3.759408950805664 -I0513 05:21:56.928805 140460907546368 logging_writer.py:48] [74600] global_step=74600, grad_norm=9.224708557128906, loss=3.894533634185791 -I0513 05:22:43.837608 140460899153664 logging_writer.py:48] [74700] global_step=74700, grad_norm=3.8542237281799316, loss=3.7806248664855957 -I0513 05:23:30.096470 140460907546368 logging_writer.py:48] [74800] global_step=74800, grad_norm=4.215198040008545, loss=3.757687568664551 -I0513 05:24:16.762986 140460899153664 logging_writer.py:48] [74900] global_step=74900, grad_norm=3.1167471408843994, loss=3.6235177516937256 -I0513 05:25:13.067045 140460907546368 logging_writer.py:48] [75000] global_step=75000, grad_norm=5.0558977127075195, loss=3.712160110473633 -I0513 05:25:59.769124 140460899153664 logging_writer.py:48] [75100] global_step=75100, grad_norm=9.418214797973633, loss=3.6743907928466797 -I0513 05:26:36.855652 140460907546368 logging_writer.py:48] [75200] global_step=75200, grad_norm=3.135939121246338, loss=3.597532272338867 -I0513 05:27:32.746575 140460899153664 logging_writer.py:48] [75300] global_step=75300, grad_norm=5.780406951904297, loss=3.806438446044922 -I0513 05:28:28.218406 140460907546368 logging_writer.py:48] [75400] global_step=75400, grad_norm=4.310510158538818, loss=3.7414283752441406 -I0513 05:29:14.707458 140460899153664 logging_writer.py:48] [75500] global_step=75500, grad_norm=5.57172966003418, loss=3.903456926345825 -I0513 05:30:01.391161 140460907546368 logging_writer.py:48] [75600] global_step=75600, grad_norm=4.065433502197266, loss=3.6935176849365234 -I0513 05:30:47.500012 140460899153664 logging_writer.py:48] [75700] global_step=75700, grad_norm=19.315914154052734, loss=3.779259443283081 -I0513 05:31:33.940203 140460907546368 logging_writer.py:48] [75800] global_step=75800, grad_norm=5.179622173309326, loss=3.702481985092163 -I0513 05:32:20.489731 140460899153664 logging_writer.py:48] [75900] global_step=75900, grad_norm=4.559787273406982, loss=3.688620090484619 -I0513 05:33:12.053818 140460907546368 logging_writer.py:48] [76000] global_step=76000, grad_norm=3.384814500808716, loss=3.814387083053589 -I0513 05:33:54.731439 140460899153664 logging_writer.py:48] [76100] global_step=76100, grad_norm=4.967829704284668, loss=3.9207234382629395 -I0513 05:34:48.435425 140460907546368 logging_writer.py:48] [76200] global_step=76200, grad_norm=3.952404022216797, loss=3.5896480083465576 -I0513 05:35:43.815541 140460899153664 logging_writer.py:48] [76300] global_step=76300, grad_norm=3.872671127319336, loss=3.8463361263275146 -I0513 05:36:48.635609 140460907546368 logging_writer.py:48] [76400] global_step=76400, grad_norm=11.8170166015625, loss=3.9491515159606934 -I0513 05:37:35.371964 140460899153664 logging_writer.py:48] [76500] global_step=76500, grad_norm=4.259308815002441, loss=3.8898963928222656 -I0513 05:38:20.312853 140460907546368 logging_writer.py:48] [76600] global_step=76600, grad_norm=3.6978678703308105, loss=3.7325637340545654 -I0513 05:39:04.129939 140460899153664 logging_writer.py:48] [76700] global_step=76700, grad_norm=5.282975196838379, loss=4.0141706466674805 -I0513 05:39:49.745592 140460907546368 logging_writer.py:48] [76800] global_step=76800, grad_norm=7.9687371253967285, loss=3.9030373096466064 -I0513 05:40:33.539092 140460899153664 logging_writer.py:48] [76900] global_step=76900, grad_norm=5.983922958374023, loss=4.073177814483643 -I0513 05:41:16.812742 140460907546368 logging_writer.py:48] [77000] global_step=77000, grad_norm=3.6110568046569824, loss=3.672572612762451 -I0513 05:42:00.029782 140678261474496 spec.py:333] Evaluating on the training split. -I0513 05:42:13.412168 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 05:43:34.564100 140678261474496 spec.py:363] Evaluating on the test split. -I0513 05:43:35.459910 140678261474496 submission_runner.py:516] Time since start: 35807.95s, Step: 77093, {'train/accuracy': Array(0.00149474, dtype=float32), 'train/loss': Array(8.261976, dtype=float32), 'validation/accuracy': Array(0.00126, dtype=float32), 'validation/loss': Array(8.243793, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(8.239175, dtype=float32), 'test/num_examples': 10000, 'score': 33999.91070771217, 'total_duration': 35807.94622397423, 'accumulated_submission_time': 33999.91070771217, 'accumulated_eval_time': 1806.4985573291779, 'accumulated_logging_time': 0.7217459678649902} -I0513 05:43:35.509664 140460899153664 logging_writer.py:48] [77093] accumulated_eval_time=1806.5, accumulated_logging_time=0.721746, accumulated_submission_time=33999.9, global_step=77093, preemption_count=0, score=33999.9, test/accuracy=0.0010999999940395355, test/loss=8.239174842834473, test/num_examples=10000, total_duration=35807.9, train/accuracy=0.001494738506153226, train/loss=8.26197624206543, validation/accuracy=0.0012599999317899346, validation/loss=8.243792533874512, validation/num_examples=50000 -I0513 05:43:42.279968 140460907546368 logging_writer.py:48] [77100] global_step=77100, grad_norm=4.0061187744140625, loss=3.9393503665924072 -I0513 05:44:25.113517 140460899153664 logging_writer.py:48] [77200] global_step=77200, grad_norm=5.567539215087891, loss=3.8308002948760986 -I0513 05:45:02.577443 140460907546368 logging_writer.py:48] [77300] global_step=77300, grad_norm=3.732372522354126, loss=3.7761569023132324 -I0513 05:45:40.120137 140460899153664 logging_writer.py:48] [77400] global_step=77400, grad_norm=8.266767501831055, loss=3.7788262367248535 -I0513 05:46:17.186541 140460907546368 logging_writer.py:48] [77500] global_step=77500, grad_norm=4.249536991119385, loss=3.772766590118408 -I0513 05:46:54.322369 140460899153664 logging_writer.py:48] [77600] global_step=77600, grad_norm=5.213112831115723, loss=3.791496753692627 -I0513 05:47:32.176122 140460907546368 logging_writer.py:48] [77700] global_step=77700, grad_norm=5.609199523925781, loss=3.890740156173706 -I0513 05:48:09.341383 140460899153664 logging_writer.py:48] [77800] global_step=77800, grad_norm=3.7171757221221924, loss=3.876363515853882 -I0513 05:48:54.215315 140460907546368 logging_writer.py:48] [77900] global_step=77900, grad_norm=5.051131725311279, loss=3.850496530532837 -I0513 05:49:33.179147 140460899153664 logging_writer.py:48] [78000] global_step=78000, grad_norm=6.632018089294434, loss=3.803096294403076 -I0513 05:50:10.989227 140460907546368 logging_writer.py:48] [78100] global_step=78100, grad_norm=3.6527652740478516, loss=3.702903985977173 -I0513 05:50:57.671356 140460899153664 logging_writer.py:48] [78200] global_step=78200, grad_norm=5.702215671539307, loss=3.895993709564209 -I0513 05:51:44.463948 140460907546368 logging_writer.py:48] [78300] global_step=78300, grad_norm=3.3551313877105713, loss=3.826117753982544 -I0513 05:52:30.707135 140460899153664 logging_writer.py:48] [78400] global_step=78400, grad_norm=4.12289571762085, loss=3.8400087356567383 -I0513 05:53:16.872113 140460907546368 logging_writer.py:48] [78500] global_step=78500, grad_norm=5.364712715148926, loss=3.7787978649139404 -I0513 05:54:03.901433 140460899153664 logging_writer.py:48] [78600] global_step=78600, grad_norm=4.4424357414245605, loss=3.837165117263794 -I0513 05:54:50.043056 140460907546368 logging_writer.py:48] [78700] global_step=78700, grad_norm=4.547253131866455, loss=3.8533284664154053 -I0513 05:55:36.269841 140460899153664 logging_writer.py:48] [78800] global_step=78800, grad_norm=3.5617353916168213, loss=3.8047053813934326 -I0513 05:56:32.078053 140460907546368 logging_writer.py:48] [78900] global_step=78900, grad_norm=4.631094932556152, loss=3.783069133758545 -I0513 05:57:18.225993 140460899153664 logging_writer.py:48] [79000] global_step=79000, grad_norm=4.313591957092285, loss=3.7169127464294434 -I0513 05:58:04.470849 140460907546368 logging_writer.py:48] [79100] global_step=79100, grad_norm=4.6931233406066895, loss=3.925509452819824 -I0513 05:59:00.420446 140460899153664 logging_writer.py:48] [79200] global_step=79200, grad_norm=4.5737080574035645, loss=3.97444486618042 -I0513 05:59:46.580927 140460907546368 logging_writer.py:48] [79300] global_step=79300, grad_norm=4.574536323547363, loss=3.8855702877044678 -I0513 06:00:33.093176 140460899153664 logging_writer.py:48] [79400] global_step=79400, grad_norm=4.767889499664307, loss=3.840047836303711 -I0513 06:01:19.731844 140460907546368 logging_writer.py:48] [79500] global_step=79500, grad_norm=7.482603073120117, loss=3.866945266723633 -I0513 06:02:05.682444 140460899153664 logging_writer.py:48] [79600] global_step=79600, grad_norm=2.990741014480591, loss=3.6678104400634766 -I0513 06:02:51.724369 140460907546368 logging_writer.py:48] [79700] global_step=79700, grad_norm=7.998231410980225, loss=3.8196487426757812 -I0513 06:03:38.697956 140460899153664 logging_writer.py:48] [79800] global_step=79800, grad_norm=9.618823051452637, loss=3.879157543182373 -I0513 06:04:33.436308 140460907546368 logging_writer.py:48] [79900] global_step=79900, grad_norm=3.4859859943389893, loss=3.6942906379699707 -I0513 06:05:14.661848 140460899153664 logging_writer.py:48] [80000] global_step=80000, grad_norm=6.446134090423584, loss=3.754537582397461 -I0513 06:06:00.271446 140460907546368 logging_writer.py:48] [80100] global_step=80100, grad_norm=4.396010398864746, loss=3.665456533432007 -I0513 06:07:00.582364 140460899153664 logging_writer.py:48] [80200] global_step=80200, grad_norm=5.095090866088867, loss=3.8724732398986816 -I0513 06:07:55.698914 140460907546368 logging_writer.py:48] [80300] global_step=80300, grad_norm=5.54566764831543, loss=3.9476521015167236 -I0513 06:08:58.807000 140460899153664 logging_writer.py:48] [80400] global_step=80400, grad_norm=8.814923286437988, loss=3.876269578933716 -I0513 06:09:40.849434 140460907546368 logging_writer.py:48] [80500] global_step=80500, grad_norm=3.367208242416382, loss=3.8054630756378174 -I0513 06:10:23.428430 140460899153664 logging_writer.py:48] [80600] global_step=80600, grad_norm=4.560641288757324, loss=3.9828379154205322 -I0513 06:11:19.361315 140460907546368 logging_writer.py:48] [80700] global_step=80700, grad_norm=6.500066757202148, loss=3.795757293701172 -I0513 06:12:05.362841 140460899153664 logging_writer.py:48] [80800] global_step=80800, grad_norm=4.093310356140137, loss=3.6848297119140625 -I0513 06:12:51.477352 140460907546368 logging_writer.py:48] [80900] global_step=80900, grad_norm=5.555255889892578, loss=3.7888662815093994 -I0513 06:13:38.265593 140460899153664 logging_writer.py:48] [81000] global_step=81000, grad_norm=5.961996555328369, loss=3.856295585632324 -I0513 06:14:24.279375 140460907546368 logging_writer.py:48] [81100] global_step=81100, grad_norm=5.398944854736328, loss=3.747192621231079 -I0513 06:15:10.446858 140460899153664 logging_writer.py:48] [81200] global_step=81200, grad_norm=4.620109558105469, loss=3.6451025009155273 -I0513 06:16:06.199000 140460907546368 logging_writer.py:48] [81300] global_step=81300, grad_norm=5.923137664794922, loss=3.8634426593780518 -I0513 06:16:51.510217 140678261474496 spec.py:333] Evaluating on the training split. -I0513 06:17:01.888964 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 06:18:33.629726 140678261474496 spec.py:363] Evaluating on the test split. -I0513 06:18:34.521039 140678261474496 submission_runner.py:516] Time since start: 37907.01s, Step: 81399, {'train/accuracy': Array(0.00115593, dtype=float32), 'train/loss': Array(8.150389, dtype=float32), 'validation/accuracy': Array(0.00128, dtype=float32), 'validation/loss': Array(8.144362, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(8.141392, dtype=float32), 'test/num_examples': 10000, 'score': 35995.8516535759, 'total_duration': 37907.00846672058, 'accumulated_submission_time': 35995.8516535759, 'accumulated_eval_time': 1909.5075569152832, 'accumulated_logging_time': 0.7861449718475342} -I0513 06:18:34.553887 140460899153664 logging_writer.py:48] [81399] accumulated_eval_time=1909.51, accumulated_logging_time=0.786145, accumulated_submission_time=35995.9, global_step=81399, preemption_count=0, score=35995.9, test/accuracy=0.0010999999940395355, test/loss=8.14139175415039, test/num_examples=10000, total_duration=37907, train/accuracy=0.0011559311533346772, train/loss=8.150388717651367, validation/accuracy=0.0012799999676644802, validation/loss=8.144362449645996, validation/num_examples=50000 -I0513 06:18:35.629360 140460907546368 logging_writer.py:48] [81400] global_step=81400, grad_norm=5.341067790985107, loss=3.816748857498169 -I0513 06:19:13.807397 140460899153664 logging_writer.py:48] [81500] global_step=81500, grad_norm=3.6715586185455322, loss=3.7415332794189453 -I0513 06:19:52.261662 140460907546368 logging_writer.py:48] [81600] global_step=81600, grad_norm=4.312809944152832, loss=4.0052571296691895 -I0513 06:20:29.281309 140460899153664 logging_writer.py:48] [81700] global_step=81700, grad_norm=3.927884817123413, loss=3.737581729888916 -I0513 06:21:18.260170 140460907546368 logging_writer.py:48] [81800] global_step=81800, grad_norm=5.220919609069824, loss=3.8600544929504395 -I0513 06:22:00.968420 140460899153664 logging_writer.py:48] [81900] global_step=81900, grad_norm=4.561461448669434, loss=3.919743299484253 -I0513 06:22:48.026714 140460907546368 logging_writer.py:48] [82000] global_step=82000, grad_norm=5.440866947174072, loss=3.859485387802124 -I0513 06:23:27.809759 140460899153664 logging_writer.py:48] [82100] global_step=82100, grad_norm=3.8421339988708496, loss=3.63350772857666 -I0513 06:24:06.980804 140460907546368 logging_writer.py:48] [82200] global_step=82200, grad_norm=3.8495941162109375, loss=3.8454983234405518 -I0513 06:24:45.415679 140460899153664 logging_writer.py:48] [82300] global_step=82300, grad_norm=4.388421535491943, loss=3.7761178016662598 -I0513 06:25:23.647510 140460907546368 logging_writer.py:48] [82400] global_step=82400, grad_norm=5.664334297180176, loss=3.8744542598724365 -I0513 06:26:02.668395 140460899153664 logging_writer.py:48] [82500] global_step=82500, grad_norm=3.577474355697632, loss=3.7874221801757812 -I0513 06:26:41.123220 140460907546368 logging_writer.py:48] [82600] global_step=82600, grad_norm=4.516883850097656, loss=3.7836897373199463 -I0513 06:27:19.132080 140460899153664 logging_writer.py:48] [82700] global_step=82700, grad_norm=4.3106465339660645, loss=3.779022693634033 -I0513 06:28:16.332841 140460907546368 logging_writer.py:48] [82800] global_step=82800, grad_norm=4.437597274780273, loss=3.8962650299072266 -I0513 06:28:53.812081 140460899153664 logging_writer.py:48] [82900] global_step=82900, grad_norm=3.9036998748779297, loss=3.7411158084869385 -I0513 06:29:40.401462 140460907546368 logging_writer.py:48] [83000] global_step=83000, grad_norm=6.563032150268555, loss=4.019347190856934 -I0513 06:30:27.861324 140460899153664 logging_writer.py:48] [83100] global_step=83100, grad_norm=3.670644760131836, loss=3.5514285564422607 -I0513 06:31:05.039845 140460907546368 logging_writer.py:48] [83200] global_step=83200, grad_norm=4.466805934906006, loss=3.7972464561462402 -I0513 06:31:51.573225 140460899153664 logging_writer.py:48] [83300] global_step=83300, grad_norm=5.690506458282471, loss=3.837528944015503 -I0513 06:32:38.339052 140460907546368 logging_writer.py:48] [83400] global_step=83400, grad_norm=6.895279884338379, loss=3.898256778717041 -I0513 06:33:24.942294 140460899153664 logging_writer.py:48] [83500] global_step=83500, grad_norm=9.322198867797852, loss=3.8653793334960938 -I0513 06:34:11.153125 140460907546368 logging_writer.py:48] [83600] global_step=83600, grad_norm=3.8551347255706787, loss=3.8223979473114014 -I0513 06:35:07.470879 140460899153664 logging_writer.py:48] [83700] global_step=83700, grad_norm=5.547417640686035, loss=3.815887928009033 -I0513 06:36:02.817995 140460907546368 logging_writer.py:48] [83800] global_step=83800, grad_norm=3.9503729343414307, loss=3.870173215866089 -I0513 06:36:49.214733 140460899153664 logging_writer.py:48] [83900] global_step=83900, grad_norm=3.565140724182129, loss=3.826611042022705 -I0513 06:37:35.984703 140460907546368 logging_writer.py:48] [84000] global_step=84000, grad_norm=10.331446647644043, loss=3.910172462463379 -I0513 06:38:21.941551 140460899153664 logging_writer.py:48] [84100] global_step=84100, grad_norm=4.205031394958496, loss=3.915855884552002 -I0513 06:39:08.813798 140460907546368 logging_writer.py:48] [84200] global_step=84200, grad_norm=3.2751152515411377, loss=3.6102774143218994 -I0513 06:39:55.617237 140460899153664 logging_writer.py:48] [84300] global_step=84300, grad_norm=9.87294864654541, loss=4.02012825012207 -I0513 06:40:42.160139 140460907546368 logging_writer.py:48] [84400] global_step=84400, grad_norm=3.3246209621429443, loss=3.755236864089966 -I0513 06:41:28.407881 140460899153664 logging_writer.py:48] [84500] global_step=84500, grad_norm=4.427681922912598, loss=3.7134337425231934 -I0513 06:42:15.028228 140460907546368 logging_writer.py:48] [84600] global_step=84600, grad_norm=5.738009929656982, loss=3.8219690322875977 -I0513 06:43:01.808368 140460899153664 logging_writer.py:48] [84700] global_step=84700, grad_norm=4.279715061187744, loss=3.8322789669036865 -I0513 06:43:48.118573 140460907546368 logging_writer.py:48] [84800] global_step=84800, grad_norm=3.5175702571868896, loss=3.7908005714416504 -I0513 06:44:44.266672 140460899153664 logging_writer.py:48] [84900] global_step=84900, grad_norm=3.5420901775360107, loss=3.815145492553711 -I0513 06:45:30.688709 140460907546368 logging_writer.py:48] [85000] global_step=85000, grad_norm=10.650847434997559, loss=3.8852732181549072 -I0513 06:46:17.009227 140460899153664 logging_writer.py:48] [85100] global_step=85100, grad_norm=4.873795032501221, loss=3.945984363555908 -I0513 06:47:03.814081 140460907546368 logging_writer.py:48] [85200] global_step=85200, grad_norm=8.832684516906738, loss=3.9998085498809814 -I0513 06:47:50.466935 140460899153664 logging_writer.py:48] [85300] global_step=85300, grad_norm=3.5896687507629395, loss=3.7085561752319336 -I0513 06:48:34.976464 140460907546368 logging_writer.py:48] [85400] global_step=85400, grad_norm=4.208330154418945, loss=3.85217547416687 -I0513 06:49:22.878821 140460899153664 logging_writer.py:48] [85500] global_step=85500, grad_norm=5.741873264312744, loss=3.7686078548431396 -I0513 06:50:09.165138 140460907546368 logging_writer.py:48] [85600] global_step=85600, grad_norm=7.447808742523193, loss=3.8215160369873047 -I0513 06:51:04.962948 140460899153664 logging_writer.py:48] [85700] global_step=85700, grad_norm=3.547058582305908, loss=3.920888900756836 -I0513 06:51:50.661874 140678261474496 spec.py:333] Evaluating on the training split. -I0513 06:52:01.355528 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 06:53:22.084512 140678261474496 spec.py:363] Evaluating on the test split. -I0513 06:53:22.975699 140678261474496 submission_runner.py:516] Time since start: 39995.46s, Step: 85799, {'train/accuracy': Array(0.00129544, dtype=float32), 'train/loss': Array(8.063519, dtype=float32), 'validation/accuracy': Array(0.00134, dtype=float32), 'validation/loss': Array(8.04747, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.001, dtype=float32), 'test/loss': Array(8.047056, dtype=float32), 'test/num_examples': 10000, 'score': 37991.905416965485, 'total_duration': 39995.46316099167, 'accumulated_submission_time': 37991.905416965485, 'accumulated_eval_time': 2001.8195931911469, 'accumulated_logging_time': 0.8286004066467285} -I0513 06:53:23.007067 140460907546368 logging_writer.py:48] [85799] accumulated_eval_time=2001.82, accumulated_logging_time=0.8286, accumulated_submission_time=37991.9, global_step=85799, preemption_count=0, score=37991.9, test/accuracy=0.0010000000474974513, test/loss=8.047056198120117, test/num_examples=10000, total_duration=39995.5, train/accuracy=0.0012954400153830647, train/loss=8.063518524169922, validation/accuracy=0.001339999958872795, validation/loss=8.047470092773438, validation/num_examples=50000 -I0513 06:53:24.097423 140460899153664 logging_writer.py:48] [85800] global_step=85800, grad_norm=4.224855422973633, loss=3.8662939071655273 -I0513 06:54:17.252139 140460907546368 logging_writer.py:48] [85900] global_step=85900, grad_norm=3.9824235439300537, loss=3.8303585052490234 -I0513 06:54:59.262190 140460899153664 logging_writer.py:48] [86000] global_step=86000, grad_norm=4.528426647186279, loss=3.6199207305908203 -I0513 06:55:43.683142 140460907546368 logging_writer.py:48] [86100] global_step=86100, grad_norm=4.29439640045166, loss=3.6850638389587402 -I0513 06:56:35.893352 140460899153664 logging_writer.py:48] [86200] global_step=86200, grad_norm=6.041114330291748, loss=3.8112685680389404 -I0513 06:57:22.671324 140460907546368 logging_writer.py:48] [86300] global_step=86300, grad_norm=4.337866306304932, loss=3.759674549102783 -I0513 06:58:05.850435 140460899153664 logging_writer.py:48] [86400] global_step=86400, grad_norm=4.733765125274658, loss=3.684037685394287 -I0513 06:58:44.731074 140460907546368 logging_writer.py:48] [86500] global_step=86500, grad_norm=5.24194860458374, loss=3.857696056365967 -I0513 06:59:28.785772 140460899153664 logging_writer.py:48] [86600] global_step=86600, grad_norm=5.429234504699707, loss=3.886012554168701 -I0513 07:00:12.951426 140460907546368 logging_writer.py:48] [86700] global_step=86700, grad_norm=5.633275508880615, loss=3.7594809532165527 -I0513 07:00:57.405229 140460899153664 logging_writer.py:48] [86800] global_step=86800, grad_norm=3.9000532627105713, loss=3.7699007987976074 -I0513 07:01:39.461067 140460907546368 logging_writer.py:48] [86900] global_step=86900, grad_norm=4.303974151611328, loss=3.8436131477355957 -I0513 07:02:25.462249 140460899153664 logging_writer.py:48] [87000] global_step=87000, grad_norm=2.9964187145233154, loss=3.767550468444824 -I0513 07:03:12.545368 140460907546368 logging_writer.py:48] [87100] global_step=87100, grad_norm=6.035670280456543, loss=3.858394145965576 -I0513 07:03:50.096033 140460899153664 logging_writer.py:48] [87200] global_step=87200, grad_norm=4.123244762420654, loss=3.8190858364105225 -I0513 07:04:37.271335 140460907546368 logging_writer.py:48] [87300] global_step=87300, grad_norm=3.0768039226531982, loss=3.7213211059570312 -I0513 07:05:24.249277 140460899153664 logging_writer.py:48] [87400] global_step=87400, grad_norm=4.980648517608643, loss=3.9287946224212646 -I0513 07:06:10.747673 140460907546368 logging_writer.py:48] [87500] global_step=87500, grad_norm=4.243657112121582, loss=3.8454201221466064 -I0513 07:06:57.625673 140460899153664 logging_writer.py:48] [87600] global_step=87600, grad_norm=3.7439019680023193, loss=3.8849265575408936 -I0513 07:07:53.438271 140460907546368 logging_writer.py:48] [87700] global_step=87700, grad_norm=3.159721612930298, loss=3.764211654663086 -I0513 07:08:49.671281 140460899153664 logging_writer.py:48] [87800] global_step=87800, grad_norm=3.977562189102173, loss=3.7142868041992188 -I0513 07:09:35.675674 140460907546368 logging_writer.py:48] [87900] global_step=87900, grad_norm=12.469419479370117, loss=3.6953916549682617 -I0513 07:10:16.868453 140460899153664 logging_writer.py:48] [88000] global_step=88000, grad_norm=3.033334970474243, loss=3.8219003677368164 -I0513 07:10:59.376046 140460907546368 logging_writer.py:48] [88100] global_step=88100, grad_norm=11.65747356414795, loss=3.8355469703674316 -I0513 07:11:46.139460 140460899153664 logging_writer.py:48] [88200] global_step=88200, grad_norm=4.076336860656738, loss=3.698057174682617 -I0513 07:12:32.496477 140460907546368 logging_writer.py:48] [88300] global_step=88300, grad_norm=5.6452460289001465, loss=3.7777440547943115 -I0513 07:13:19.059734 140460899153664 logging_writer.py:48] [88400] global_step=88400, grad_norm=6.808701038360596, loss=3.7643516063690186 -I0513 07:14:15.172166 140460907546368 logging_writer.py:48] [88500] global_step=88500, grad_norm=9.564594268798828, loss=4.012497901916504 -I0513 07:15:01.507421 140460899153664 logging_writer.py:48] [88600] global_step=88600, grad_norm=4.330405235290527, loss=3.7575857639312744 -I0513 07:15:47.998466 140460907546368 logging_writer.py:48] [88700] global_step=88700, grad_norm=3.1850171089172363, loss=3.6566483974456787 -I0513 07:16:34.734076 140460899153664 logging_writer.py:48] [88800] global_step=88800, grad_norm=4.46567964553833, loss=3.8495423793792725 -I0513 07:17:21.209497 140460907546368 logging_writer.py:48] [88900] global_step=88900, grad_norm=4.646286964416504, loss=3.933137893676758 -I0513 07:18:16.308377 140460899153664 logging_writer.py:48] [89000] global_step=89000, grad_norm=6.875772476196289, loss=3.8803672790527344 -I0513 07:18:59.445574 140460907546368 logging_writer.py:48] [89100] global_step=89100, grad_norm=2.891200542449951, loss=3.8028640747070312 -I0513 07:19:40.363098 140460899153664 logging_writer.py:48] [89200] global_step=89200, grad_norm=4.066144943237305, loss=3.8847885131835938 -I0513 07:20:26.278333 140460907546368 logging_writer.py:48] [89300] global_step=89300, grad_norm=4.078550338745117, loss=3.804953098297119 -I0513 07:21:13.159543 140460899153664 logging_writer.py:48] [89400] global_step=89400, grad_norm=5.99378776550293, loss=3.90647554397583 -I0513 07:21:59.526000 140460907546368 logging_writer.py:48] [89500] global_step=89500, grad_norm=4.645710468292236, loss=3.662891149520874 -I0513 07:22:46.291105 140460899153664 logging_writer.py:48] [89600] global_step=89600, grad_norm=7.230001449584961, loss=3.8156332969665527 -I0513 07:23:33.210793 140460907546368 logging_writer.py:48] [89700] global_step=89700, grad_norm=4.563052654266357, loss=3.9181132316589355 -I0513 07:24:19.540375 140460899153664 logging_writer.py:48] [89800] global_step=89800, grad_norm=4.733450412750244, loss=3.694173812866211 -I0513 07:25:05.951342 140460907546368 logging_writer.py:48] [89900] global_step=89900, grad_norm=4.272140979766846, loss=3.6506197452545166 -I0513 07:25:53.039677 140460899153664 logging_writer.py:48] [90000] global_step=90000, grad_norm=41.589359283447266, loss=3.934208869934082 -I0513 07:26:39.110073 140678261474496 spec.py:333] Evaluating on the training split. -I0513 07:26:50.447838 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 07:28:17.314564 140678261474496 spec.py:363] Evaluating on the test split. -I0513 07:28:18.205927 140678261474496 submission_runner.py:516] Time since start: 42090.69s, Step: 90100, {'train/accuracy': Array(0.00131537, dtype=float32), 'train/loss': Array(7.9802356, dtype=float32), 'validation/accuracy': Array(0.0014, dtype=float32), 'validation/loss': Array(7.960504, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(7.9610558, dtype=float32), 'test/num_examples': 10000, 'score': 39987.95483374596, 'total_duration': 42090.69328832626, 'accumulated_submission_time': 39987.95483374596, 'accumulated_eval_time': 2100.913553237915, 'accumulated_logging_time': 0.8685033321380615} -I0513 07:28:18.236319 140460907546368 logging_writer.py:48] [90100] accumulated_eval_time=2100.91, accumulated_logging_time=0.868503, accumulated_submission_time=39988, global_step=90100, preemption_count=0, score=39988, test/accuracy=0.0010999999940395355, test/loss=7.961055755615234, test/num_examples=10000, total_duration=42090.7, train/accuracy=0.0013153698528185487, train/loss=7.980235576629639, validation/accuracy=0.00139999995008111, validation/loss=7.960504055023193, validation/num_examples=50000 -I0513 07:28:18.504506 140460899153664 logging_writer.py:48] [90100] global_step=90100, grad_norm=10.948570251464844, loss=3.913022756576538 -I0513 07:29:14.546554 140460907546368 logging_writer.py:48] [90200] global_step=90200, grad_norm=4.509075164794922, loss=3.816344738006592 -I0513 07:30:01.236550 140460899153664 logging_writer.py:48] [90300] global_step=90300, grad_norm=5.141840934753418, loss=3.780602216720581 -I0513 07:30:54.388052 140460907546368 logging_writer.py:48] [90400] global_step=90400, grad_norm=4.677832126617432, loss=3.9111263751983643 -I0513 07:31:36.979624 140460899153664 logging_writer.py:48] [90500] global_step=90500, grad_norm=4.139318466186523, loss=3.712047815322876 -I0513 07:32:23.244439 140460907546368 logging_writer.py:48] [90600] global_step=90600, grad_norm=8.189085006713867, loss=3.7315409183502197 -I0513 07:33:06.146157 140460899153664 logging_writer.py:48] [90700] global_step=90700, grad_norm=3.3804819583892822, loss=3.797423839569092 -I0513 07:33:58.712534 140460907546368 logging_writer.py:48] [90800] global_step=90800, grad_norm=4.920464992523193, loss=3.8274569511413574 -I0513 07:34:38.657103 140460899153664 logging_writer.py:48] [90900] global_step=90900, grad_norm=3.9284539222717285, loss=3.8261749744415283 -I0513 07:35:16.487098 140460907546368 logging_writer.py:48] [91000] global_step=91000, grad_norm=4.5098042488098145, loss=3.780270576477051 -I0513 07:35:54.551532 140460899153664 logging_writer.py:48] [91100] global_step=91100, grad_norm=4.957934856414795, loss=3.6634726524353027 -I0513 07:36:33.723943 140460907546368 logging_writer.py:48] [91200] global_step=91200, grad_norm=4.768673896789551, loss=3.700324296951294 -I0513 07:37:19.848201 140460899153664 logging_writer.py:48] [91300] global_step=91300, grad_norm=4.933558940887451, loss=3.7424392700195312 -I0513 07:37:59.125276 140460907546368 logging_writer.py:48] [91400] global_step=91400, grad_norm=5.555438041687012, loss=3.8752834796905518 -I0513 07:38:36.819938 140460899153664 logging_writer.py:48] [91500] global_step=91500, grad_norm=5.5240092277526855, loss=3.869260311126709 -I0513 07:39:14.175217 140460907546368 logging_writer.py:48] [91600] global_step=91600, grad_norm=5.941207408905029, loss=3.7706010341644287 -I0513 07:39:51.760317 140460899153664 logging_writer.py:48] [91700] global_step=91700, grad_norm=8.529118537902832, loss=3.754472255706787 -I0513 07:40:29.533739 140460907546368 logging_writer.py:48] [91800] global_step=91800, grad_norm=5.000514507293701, loss=3.704399585723877 -I0513 07:41:06.847147 140460899153664 logging_writer.py:48] [91900] global_step=91900, grad_norm=5.442907810211182, loss=3.910405158996582 -I0513 07:41:44.064099 140460907546368 logging_writer.py:48] [92000] global_step=92000, grad_norm=5.440472602844238, loss=3.792483329772949 -I0513 07:42:21.955488 140460899153664 logging_writer.py:48] [92100] global_step=92100, grad_norm=4.812552452087402, loss=3.87166428565979 -I0513 07:43:08.437758 140460907546368 logging_writer.py:48] [92200] global_step=92200, grad_norm=4.086339473724365, loss=3.8696224689483643 -I0513 07:43:45.763434 140460899153664 logging_writer.py:48] [92300] global_step=92300, grad_norm=3.9890873432159424, loss=3.905237913131714 -I0513 07:44:23.289177 140460907546368 logging_writer.py:48] [92400] global_step=92400, grad_norm=6.703495979309082, loss=3.8337721824645996 -I0513 07:45:00.222510 140460899153664 logging_writer.py:48] [92500] global_step=92500, grad_norm=5.70967435836792, loss=3.846621513366699 -I0513 07:45:37.699200 140460907546368 logging_writer.py:48] [92600] global_step=92600, grad_norm=5.7875075340271, loss=3.7894341945648193 -I0513 07:46:15.150492 140460899153664 logging_writer.py:48] [92700] global_step=92700, grad_norm=8.184720993041992, loss=3.8709945678710938 -I0513 07:46:52.139105 140460907546368 logging_writer.py:48] [92800] global_step=92800, grad_norm=4.61423921585083, loss=3.8477606773376465 -I0513 07:47:29.455945 140460899153664 logging_writer.py:48] [92900] global_step=92900, grad_norm=3.956357717514038, loss=3.8248634338378906 -I0513 07:48:17.548476 140460907546368 logging_writer.py:48] [93000] global_step=93000, grad_norm=3.7475476264953613, loss=3.714282989501953 -I0513 07:48:54.949093 140460899153664 logging_writer.py:48] [93100] global_step=93100, grad_norm=6.458767890930176, loss=3.8477678298950195 -I0513 07:49:32.484347 140460907546368 logging_writer.py:48] [93200] global_step=93200, grad_norm=4.47292947769165, loss=3.682412624359131 -I0513 07:50:10.189485 140460899153664 logging_writer.py:48] [93300] global_step=93300, grad_norm=3.734076738357544, loss=3.83250093460083 -I0513 07:50:47.334067 140460907546368 logging_writer.py:48] [93400] global_step=93400, grad_norm=7.129732608795166, loss=3.8429007530212402 -I0513 07:51:24.956437 140460899153664 logging_writer.py:48] [93500] global_step=93500, grad_norm=10.599674224853516, loss=4.2624125480651855 -I0513 07:52:02.640142 140460907546368 logging_writer.py:48] [93600] global_step=93600, grad_norm=6.92296028137207, loss=3.814478874206543 -I0513 07:52:39.801099 140460899153664 logging_writer.py:48] [93700] global_step=93700, grad_norm=7.7157111167907715, loss=3.7428677082061768 -I0513 07:53:17.354708 140460907546368 logging_writer.py:48] [93800] global_step=93800, grad_norm=5.946841239929199, loss=3.9010515213012695 -I0513 07:53:55.014625 140460899153664 logging_writer.py:48] [93900] global_step=93900, grad_norm=4.29172945022583, loss=3.9677467346191406 -I0513 07:54:32.223661 140460907546368 logging_writer.py:48] [94000] global_step=94000, grad_norm=5.395374298095703, loss=3.9360275268554688 -I0513 07:55:09.420059 140460899153664 logging_writer.py:48] [94100] global_step=94100, grad_norm=3.836005926132202, loss=3.8784711360931396 -I0513 07:55:47.234442 140460907546368 logging_writer.py:48] [94200] global_step=94200, grad_norm=6.399783134460449, loss=3.8942184448242188 -I0513 07:56:24.386933 140460899153664 logging_writer.py:48] [94300] global_step=94300, grad_norm=6.888943195343018, loss=4.00918436050415 -I0513 07:57:01.888834 140460907546368 logging_writer.py:48] [94400] global_step=94400, grad_norm=12.643881797790527, loss=4.075257301330566 -I0513 07:57:39.431530 140460899153664 logging_writer.py:48] [94500] global_step=94500, grad_norm=3.469191551208496, loss=3.7757301330566406 -I0513 07:58:16.661285 140460907546368 logging_writer.py:48] [94600] global_step=94600, grad_norm=21.919151306152344, loss=3.9179792404174805 -I0513 07:58:54.178519 140460899153664 logging_writer.py:48] [94700] global_step=94700, grad_norm=3.708085060119629, loss=3.787489891052246 -I0513 07:59:31.795850 140460907546368 logging_writer.py:48] [94800] global_step=94800, grad_norm=5.35345458984375, loss=3.62014102935791 -I0513 08:00:08.969798 140460899153664 logging_writer.py:48] [94900] global_step=94900, grad_norm=4.0392842292785645, loss=3.852752685546875 -I0513 08:00:46.522398 140460907546368 logging_writer.py:48] [95000] global_step=95000, grad_norm=7.490811824798584, loss=3.8384194374084473 -I0513 08:01:24.056098 140460899153664 logging_writer.py:48] [95100] global_step=95100, grad_norm=7.877013206481934, loss=3.8732519149780273 -I0513 08:01:34.230965 140678261474496 spec.py:333] Evaluating on the training split. -I0513 08:01:46.465407 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 08:03:15.331581 140678261474496 spec.py:363] Evaluating on the test split. -I0513 08:03:16.225642 140678261474496 submission_runner.py:516] Time since start: 44188.71s, Step: 95129, {'train/accuracy': Array(0.00109614, dtype=float32), 'train/loss': Array(7.89192, dtype=float32), 'validation/accuracy': Array(0.00134, dtype=float32), 'validation/loss': Array(7.8832655, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0009, dtype=float32), 'test/loss': Array(7.8850675, dtype=float32), 'test/num_examples': 10000, 'score': 41983.89597201347, 'total_duration': 44188.71274614334, 'accumulated_submission_time': 41983.89597201347, 'accumulated_eval_time': 2202.9060847759247, 'accumulated_logging_time': 0.9071593284606934} -I0513 08:03:16.263988 140460907546368 logging_writer.py:48] [95129] accumulated_eval_time=2202.91, accumulated_logging_time=0.907159, accumulated_submission_time=41983.9, global_step=95129, preemption_count=0, score=41983.9, test/accuracy=0.0009000000427477062, test/loss=7.885067462921143, test/num_examples=10000, total_duration=44188.7, train/accuracy=0.0010961415246129036, train/loss=7.89192008972168, validation/accuracy=0.001339999958872795, validation/loss=7.883265495300293, validation/num_examples=50000 -I0513 08:03:43.027014 140460899153664 logging_writer.py:48] [95200] global_step=95200, grad_norm=3.354382276535034, loss=3.8114922046661377 -I0513 08:04:20.364913 140460907546368 logging_writer.py:48] [95300] global_step=95300, grad_norm=4.808167457580566, loss=3.8455333709716797 -I0513 08:05:01.937754 140460899153664 logging_writer.py:48] [95400] global_step=95400, grad_norm=7.138449668884277, loss=3.958710193634033 -I0513 08:05:41.613908 140460907546368 logging_writer.py:48] [95500] global_step=95500, grad_norm=3.6178700923919678, loss=3.77113676071167 -I0513 08:06:21.107240 140460899153664 logging_writer.py:48] [95600] global_step=95600, grad_norm=4.977309703826904, loss=3.8573155403137207 -I0513 08:07:01.661644 140460907546368 logging_writer.py:48] [95700] global_step=95700, grad_norm=6.241224765777588, loss=3.72729229927063 -I0513 08:07:45.464900 140460899153664 logging_writer.py:48] [95800] global_step=95800, grad_norm=4.171980857849121, loss=3.7638003826141357 -I0513 08:08:25.625959 140460907546368 logging_writer.py:48] [95900] global_step=95900, grad_norm=5.155831813812256, loss=3.664594888687134 -I0513 08:09:17.331924 140460899153664 logging_writer.py:48] [96000] global_step=96000, grad_norm=6.023154258728027, loss=3.603090286254883 -I0513 08:10:11.602197 140460907546368 logging_writer.py:48] [96100] global_step=96100, grad_norm=4.299440383911133, loss=3.7093887329101562 -I0513 08:11:02.024841 140460899153664 logging_writer.py:48] [96200] global_step=96200, grad_norm=7.624329566955566, loss=3.810959577560425 -I0513 08:12:02.131484 140460907546368 logging_writer.py:48] [96300] global_step=96300, grad_norm=6.7118730545043945, loss=3.83454966545105 -I0513 08:13:02.022970 140460899153664 logging_writer.py:48] [96400] global_step=96400, grad_norm=8.083127975463867, loss=3.940143585205078 -I0513 08:13:46.574661 140460907546368 logging_writer.py:48] [96500] global_step=96500, grad_norm=4.3435797691345215, loss=3.782449245452881 -I0513 08:14:32.034845 140460899153664 logging_writer.py:48] [96600] global_step=96600, grad_norm=9.470343589782715, loss=3.7630910873413086 -I0513 08:15:17.149464 140460907546368 logging_writer.py:48] [96700] global_step=96700, grad_norm=9.088248252868652, loss=3.8050761222839355 -I0513 08:15:56.276051 140460899153664 logging_writer.py:48] [96800] global_step=96800, grad_norm=4.150111675262451, loss=3.895420551300049 -I0513 08:16:36.040871 140460907546368 logging_writer.py:48] [96900] global_step=96900, grad_norm=3.810899257659912, loss=3.7804946899414062 -I0513 08:17:24.787189 140460899153664 logging_writer.py:48] [97000] global_step=97000, grad_norm=6.819541931152344, loss=3.8287148475646973 -I0513 08:18:11.143256 140460907546368 logging_writer.py:48] [97100] global_step=97100, grad_norm=4.364023208618164, loss=3.8176097869873047 -I0513 08:18:57.484166 140460899153664 logging_writer.py:48] [97200] global_step=97200, grad_norm=4.284811019897461, loss=3.8856594562530518 -I0513 08:19:35.887659 140460907546368 logging_writer.py:48] [97300] global_step=97300, grad_norm=3.881463050842285, loss=3.7122490406036377 -I0513 08:20:23.360846 140460899153664 logging_writer.py:48] [97400] global_step=97400, grad_norm=4.675198078155518, loss=3.6778368949890137 -I0513 08:21:01.408993 140460907546368 logging_writer.py:48] [97500] global_step=97500, grad_norm=4.755186080932617, loss=3.819511890411377 -I0513 08:21:39.339506 140460899153664 logging_writer.py:48] [97600] global_step=97600, grad_norm=5.189528465270996, loss=3.8879799842834473 -I0513 08:22:17.160022 140460907546368 logging_writer.py:48] [97700] global_step=97700, grad_norm=8.440134048461914, loss=4.0680131912231445 -I0513 08:22:56.706207 140460899153664 logging_writer.py:48] [97800] global_step=97800, grad_norm=4.429986000061035, loss=3.83591890335083 -I0513 08:23:42.902665 140460907546368 logging_writer.py:48] [97900] global_step=97900, grad_norm=3.766267776489258, loss=3.769105911254883 -I0513 08:24:29.463665 140460899153664 logging_writer.py:48] [98000] global_step=98000, grad_norm=16.422508239746094, loss=3.8327059745788574 -I0513 08:25:21.160832 140460907546368 logging_writer.py:48] [98100] global_step=98100, grad_norm=4.530926704406738, loss=3.7325081825256348 -I0513 08:26:04.957390 140460899153664 logging_writer.py:48] [98200] global_step=98200, grad_norm=86.1595458984375, loss=4.139312744140625 -I0513 08:26:48.006861 140460907546368 logging_writer.py:48] [98300] global_step=98300, grad_norm=3.62532639503479, loss=3.7731308937072754 -I0513 08:27:33.311110 140460899153664 logging_writer.py:48] [98400] global_step=98400, grad_norm=4.351271152496338, loss=3.798084020614624 -I0513 08:28:25.471858 140460907546368 logging_writer.py:48] [98500] global_step=98500, grad_norm=3.585325241088867, loss=3.866652250289917 -I0513 08:29:09.928947 140460899153664 logging_writer.py:48] [98600] global_step=98600, grad_norm=4.567863464355469, loss=3.894746780395508 -I0513 08:30:11.246021 140460907546368 logging_writer.py:48] [98700] global_step=98700, grad_norm=5.484220027923584, loss=3.925459146499634 -I0513 08:31:08.128279 140460899153664 logging_writer.py:48] [98800] global_step=98800, grad_norm=6.137061595916748, loss=3.728240728378296 -I0513 08:32:03.679808 140460907546368 logging_writer.py:48] [98900] global_step=98900, grad_norm=5.499675273895264, loss=3.8644447326660156 -I0513 08:33:01.223588 140460899153664 logging_writer.py:48] [99000] global_step=99000, grad_norm=4.801363945007324, loss=3.7155914306640625 -I0513 08:33:52.427301 140460907546368 logging_writer.py:48] [99100] global_step=99100, grad_norm=5.48730993270874, loss=3.7627875804901123 -I0513 08:34:43.787933 140460899153664 logging_writer.py:48] [99200] global_step=99200, grad_norm=6.35140323638916, loss=3.9185256958007812 -I0513 08:35:40.843558 140460907546368 logging_writer.py:48] [99300] global_step=99300, grad_norm=40.34929656982422, loss=3.9201459884643555 -I0513 08:36:33.487705 140678261474496 spec.py:333] Evaluating on the training split. -I0513 08:36:47.521517 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 08:38:10.434312 140678261474496 spec.py:363] Evaluating on the test split. -I0513 08:38:11.463634 140678261474496 submission_runner.py:516] Time since start: 46283.82s, Step: 99394, {'train/accuracy': Array(0.00097656, dtype=float32), 'train/loss': Array(7.833911, dtype=float32), 'validation/accuracy': Array(0.00114, dtype=float32), 'validation/loss': Array(7.8286967, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(7.8307037, dtype=float32), 'test/num_examples': 10000, 'score': 43981.064588308334, 'total_duration': 46283.816162109375, 'accumulated_submission_time': 43981.064588308334, 'accumulated_eval_time': 2300.74538230896, 'accumulated_logging_time': 0.9539778232574463} -I0513 08:38:11.921288 140460899153664 logging_writer.py:48] [99394] accumulated_eval_time=2300.75, accumulated_logging_time=0.953978, accumulated_submission_time=43981.1, global_step=99394, preemption_count=0, score=43981.1, test/accuracy=0.0010999999940395355, test/loss=7.8307037353515625, test/num_examples=10000, total_duration=46283.8, train/accuracy=0.0009765625, train/loss=7.833910942077637, validation/accuracy=0.0011399999493733048, validation/loss=7.8286967277526855, validation/num_examples=50000 -I0513 08:38:14.627749 140460907546368 logging_writer.py:48] [99400] global_step=99400, grad_norm=6.849164962768555, loss=3.8251545429229736 -I0513 08:38:59.171271 140460899153664 logging_writer.py:48] [99500] global_step=99500, grad_norm=4.586352348327637, loss=3.8029768466949463 -I0513 08:39:41.987125 140460907546368 logging_writer.py:48] [99600] global_step=99600, grad_norm=5.305855751037598, loss=3.7803871631622314 -I0513 08:40:29.285961 140460899153664 logging_writer.py:48] [99700] global_step=99700, grad_norm=8.913647651672363, loss=3.7342731952667236 -I0513 08:41:25.282482 140460907546368 logging_writer.py:48] [99800] global_step=99800, grad_norm=4.155928134918213, loss=3.9605932235717773 -I0513 08:42:08.285049 140460899153664 logging_writer.py:48] [99900] global_step=99900, grad_norm=3.533277988433838, loss=3.8324999809265137 -I0513 08:43:03.313524 140460907546368 logging_writer.py:48] [100000] global_step=100000, grad_norm=6.277818202972412, loss=3.9499635696411133 -I0513 08:44:02.207909 140460899153664 logging_writer.py:48] [100100] global_step=100100, grad_norm=3.576676607131958, loss=3.848503589630127 -I0513 08:44:47.188951 140460907546368 logging_writer.py:48] [100200] global_step=100200, grad_norm=14.48756217956543, loss=3.8832590579986572 -I0513 08:45:35.555730 140460899153664 logging_writer.py:48] [100300] global_step=100300, grad_norm=5.687304973602295, loss=3.7929046154022217 -I0513 08:46:21.328441 140460907546368 logging_writer.py:48] [100400] global_step=100400, grad_norm=5.055975437164307, loss=3.9039082527160645 -I0513 08:47:16.893154 140460899153664 logging_writer.py:48] [100500] global_step=100500, grad_norm=4.6007914543151855, loss=3.8486454486846924 -I0513 08:48:02.573055 140460907546368 logging_writer.py:48] [100600] global_step=100600, grad_norm=10.004302024841309, loss=4.018870830535889 -I0513 08:48:46.984180 140460899153664 logging_writer.py:48] [100700] global_step=100700, grad_norm=8.064719200134277, loss=3.8895323276519775 -I0513 08:49:25.147973 140460907546368 logging_writer.py:48] [100800] global_step=100800, grad_norm=5.413357734680176, loss=3.8739757537841797 -I0513 08:50:02.914989 140460899153664 logging_writer.py:48] [100900] global_step=100900, grad_norm=7.1084160804748535, loss=4.025882720947266 -I0513 08:50:40.516021 140460907546368 logging_writer.py:48] [101000] global_step=101000, grad_norm=3.5982515811920166, loss=3.8280842304229736 -I0513 08:51:19.606410 140460899153664 logging_writer.py:48] [101100] global_step=101100, grad_norm=7.303923606872559, loss=3.8899106979370117 -I0513 08:51:59.366849 140460907546368 logging_writer.py:48] [101200] global_step=101200, grad_norm=5.437741756439209, loss=3.8602232933044434 -I0513 08:52:45.534067 140460899153664 logging_writer.py:48] [101300] global_step=101300, grad_norm=6.442915916442871, loss=3.990473747253418 -I0513 08:53:40.298756 140460907546368 logging_writer.py:48] [101400] global_step=101400, grad_norm=4.117714881896973, loss=3.929281711578369 -I0513 08:54:26.479444 140460899153664 logging_writer.py:48] [101500] global_step=101500, grad_norm=5.609199523925781, loss=3.7813572883605957 -I0513 08:55:13.468622 140460907546368 logging_writer.py:48] [101600] global_step=101600, grad_norm=3.1657440662384033, loss=3.9256441593170166 -I0513 08:56:00.147009 140460899153664 logging_writer.py:48] [101700] global_step=101700, grad_norm=6.038615703582764, loss=3.8847527503967285 -I0513 08:56:47.729552 140460907546368 logging_writer.py:48] [101800] global_step=101800, grad_norm=4.150689601898193, loss=3.767083168029785 -I0513 08:57:30.132106 140460899153664 logging_writer.py:48] [101900] global_step=101900, grad_norm=7.592629432678223, loss=3.983919382095337 -I0513 08:58:16.130585 140460907546368 logging_writer.py:48] [102000] global_step=102000, grad_norm=18.14057159423828, loss=3.949997901916504 -I0513 08:58:58.811496 140460899153664 logging_writer.py:48] [102100] global_step=102100, grad_norm=4.020564556121826, loss=3.7209715843200684 -I0513 08:59:44.500860 140460907546368 logging_writer.py:48] [102200] global_step=102200, grad_norm=5.306018829345703, loss=3.7323241233825684 -I0513 09:00:31.700152 140460899153664 logging_writer.py:48] [102300] global_step=102300, grad_norm=3.5669164657592773, loss=3.762629747390747 -I0513 09:01:21.476897 140460907546368 logging_writer.py:48] [102400] global_step=102400, grad_norm=6.86998176574707, loss=3.858736515045166 -I0513 09:02:04.101200 140460899153664 logging_writer.py:48] [102500] global_step=102500, grad_norm=3.892204761505127, loss=3.8480544090270996 -I0513 09:02:53.645239 140460907546368 logging_writer.py:48] [102600] global_step=102600, grad_norm=5.380331993103027, loss=3.7225747108459473 -I0513 09:03:46.978888 140460899153664 logging_writer.py:48] [102700] global_step=102700, grad_norm=6.820309638977051, loss=3.753641366958618 -I0513 09:04:29.783235 140460907546368 logging_writer.py:48] [102800] global_step=102800, grad_norm=5.804520130157471, loss=3.919973134994507 -I0513 09:05:17.526557 140460899153664 logging_writer.py:48] [102900] global_step=102900, grad_norm=4.6263885498046875, loss=3.7958836555480957 -I0513 09:06:00.137383 140460907546368 logging_writer.py:48] [103000] global_step=103000, grad_norm=3.044297456741333, loss=3.660358428955078 -I0513 09:06:45.982035 140460899153664 logging_writer.py:48] [103100] global_step=103100, grad_norm=7.713465213775635, loss=3.9784226417541504 -I0513 09:07:36.430421 140460907546368 logging_writer.py:48] [103200] global_step=103200, grad_norm=9.819127082824707, loss=3.7930827140808105 -I0513 09:08:23.172173 140460899153664 logging_writer.py:48] [103300] global_step=103300, grad_norm=5.170093059539795, loss=3.800729751586914 -I0513 09:09:11.392995 140460907546368 logging_writer.py:48] [103400] global_step=103400, grad_norm=4.859818935394287, loss=3.807338237762451 -I0513 09:09:54.546564 140460899153664 logging_writer.py:48] [103500] global_step=103500, grad_norm=4.221146583557129, loss=3.7915050983428955 -I0513 09:10:41.771285 140460907546368 logging_writer.py:48] [103600] global_step=103600, grad_norm=5.814962387084961, loss=3.871877908706665 -I0513 09:11:23.572858 140460899153664 logging_writer.py:48] [103700] global_step=103700, grad_norm=3.8086273670196533, loss=3.8059024810791016 -I0513 09:11:27.920216 140678261474496 spec.py:333] Evaluating on the training split. -I0513 09:11:42.004076 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 09:12:49.796660 140678261474496 spec.py:363] Evaluating on the test split. -I0513 09:12:50.887417 140678261474496 submission_runner.py:516] Time since start: 48363.14s, Step: 103708, {'train/accuracy': Array(0.00169404, dtype=float32), 'train/loss': Array(7.7662373, dtype=float32), 'validation/accuracy': Array(0.00136, dtype=float32), 'validation/loss': Array(7.7699013, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(7.772891, dtype=float32), 'test/num_examples': 10000, 'score': 45976.930940151215, 'total_duration': 48363.140311956406, 'accumulated_submission_time': 45976.930940151215, 'accumulated_eval_time': 2383.4762213230133, 'accumulated_logging_time': 1.4857048988342285} -I0513 09:12:51.143414 140460907546368 logging_writer.py:48] [103708] accumulated_eval_time=2383.48, accumulated_logging_time=1.4857, accumulated_submission_time=45976.9, global_step=103708, preemption_count=0, score=45976.9, test/accuracy=0.0010999999940395355, test/loss=7.772891044616699, test/num_examples=10000, total_duration=48363.1, train/accuracy=0.001694036996923387, train/loss=7.766237258911133, validation/accuracy=0.0013599999947473407, validation/loss=7.769901275634766, validation/num_examples=50000 -I0513 09:13:46.008794 140460899153664 logging_writer.py:48] [103800] global_step=103800, grad_norm=3.791799545288086, loss=3.7798755168914795 -I0513 09:14:50.898667 140460907546368 logging_writer.py:48] [103900] global_step=103900, grad_norm=3.511486053466797, loss=3.714195966720581 -I0513 09:15:43.330635 140460899153664 logging_writer.py:48] [104000] global_step=104000, grad_norm=3.901045083999634, loss=3.7973995208740234 -I0513 09:16:29.820963 140460907546368 logging_writer.py:48] [104100] global_step=104100, grad_norm=3.582700490951538, loss=3.777087688446045 -I0513 09:17:15.014058 140460899153664 logging_writer.py:48] [104200] global_step=104200, grad_norm=6.234501361846924, loss=3.8231358528137207 -I0513 09:18:08.135549 140460907546368 logging_writer.py:48] [104300] global_step=104300, grad_norm=4.201540470123291, loss=3.8543992042541504 -I0513 09:19:03.869861 140460899153664 logging_writer.py:48] [104400] global_step=104400, grad_norm=4.389229774475098, loss=3.8518307209014893 -I0513 09:19:55.715376 140460907546368 logging_writer.py:48] [104500] global_step=104500, grad_norm=5.585868835449219, loss=3.7659382820129395 -I0513 09:20:50.488746 140460899153664 logging_writer.py:48] [104600] global_step=104600, grad_norm=6.658158779144287, loss=3.921970844268799 -I0513 09:21:46.763227 140460907546368 logging_writer.py:48] [104700] global_step=104700, grad_norm=6.974661350250244, loss=3.7434895038604736 -I0513 09:22:40.797488 140460899153664 logging_writer.py:48] [104800] global_step=104800, grad_norm=4.939453125, loss=3.77407169342041 -I0513 09:23:34.264974 140460907546368 logging_writer.py:48] [104900] global_step=104900, grad_norm=14.094151496887207, loss=3.793348550796509 -I0513 09:24:27.897006 140460899153664 logging_writer.py:48] [105000] global_step=105000, grad_norm=4.736960411071777, loss=3.801128387451172 -I0513 09:25:22.893005 140460907546368 logging_writer.py:48] [105100] global_step=105100, grad_norm=4.719320774078369, loss=3.7999978065490723 -I0513 09:26:05.301260 140460899153664 logging_writer.py:48] [105200] global_step=105200, grad_norm=6.018013954162598, loss=3.8536245822906494 -I0513 09:26:51.831508 140460907546368 logging_writer.py:48] [105300] global_step=105300, grad_norm=4.3824238777160645, loss=3.840116024017334 -I0513 09:27:37.842781 140460899153664 logging_writer.py:48] [105400] global_step=105400, grad_norm=8.889808654785156, loss=3.9887704849243164 -I0513 09:28:26.593012 140460907546368 logging_writer.py:48] [105500] global_step=105500, grad_norm=4.012938022613525, loss=3.7513463497161865 -I0513 09:29:18.985199 140460899153664 logging_writer.py:48] [105600] global_step=105600, grad_norm=12.915979385375977, loss=3.7323193550109863 -I0513 09:30:15.068216 140460907546368 logging_writer.py:48] [105700] global_step=105700, grad_norm=4.008749485015869, loss=3.8466925621032715 -I0513 09:31:08.808896 140460899153664 logging_writer.py:48] [105800] global_step=105800, grad_norm=12.496825218200684, loss=3.7859511375427246 -I0513 09:32:03.438216 140460907546368 logging_writer.py:48] [105900] global_step=105900, grad_norm=5.793846607208252, loss=3.9013891220092773 -I0513 09:32:54.126013 140460899153664 logging_writer.py:48] [106000] global_step=106000, grad_norm=3.7226641178131104, loss=3.840048313140869 -I0513 09:33:47.004726 140460907546368 logging_writer.py:48] [106100] global_step=106100, grad_norm=4.282467842102051, loss=3.686673402786255 -I0513 09:34:37.593435 140460899153664 logging_writer.py:48] [106200] global_step=106200, grad_norm=6.168426513671875, loss=3.872377872467041 -I0513 09:35:30.684010 140460907546368 logging_writer.py:48] [106300] global_step=106300, grad_norm=6.401358127593994, loss=3.8281731605529785 -I0513 09:36:14.625079 140460899153664 logging_writer.py:48] [106400] global_step=106400, grad_norm=4.904049396514893, loss=3.758202314376831 -I0513 09:37:03.647283 140460907546368 logging_writer.py:48] [106500] global_step=106500, grad_norm=3.1958072185516357, loss=3.7847938537597656 -I0513 09:37:49.599213 140460899153664 logging_writer.py:48] [106600] global_step=106600, grad_norm=4.729272842407227, loss=3.8970866203308105 -I0513 09:38:40.752761 140460907546368 logging_writer.py:48] [106700] global_step=106700, grad_norm=3.7540974617004395, loss=3.784024715423584 -I0513 09:39:34.678696 140460899153664 logging_writer.py:48] [106800] global_step=106800, grad_norm=7.697290420532227, loss=3.993462085723877 -I0513 09:40:29.682343 140460907546368 logging_writer.py:48] [106900] global_step=106900, grad_norm=7.127755641937256, loss=3.8478779792785645 -I0513 09:41:23.051416 140460899153664 logging_writer.py:48] [107000] global_step=107000, grad_norm=8.760859489440918, loss=3.8773207664489746 -I0513 09:42:20.554656 140460907546368 logging_writer.py:48] [107100] global_step=107100, grad_norm=4.847646236419678, loss=3.7123119831085205 -I0513 09:43:10.768983 140460899153664 logging_writer.py:48] [107200] global_step=107200, grad_norm=5.025604248046875, loss=3.863089084625244 -I0513 09:44:00.664084 140460907546368 logging_writer.py:48] [107300] global_step=107300, grad_norm=3.486504316329956, loss=3.7198550701141357 -I0513 09:44:52.852356 140460899153664 logging_writer.py:48] [107400] global_step=107400, grad_norm=4.538763046264648, loss=3.9173710346221924 -I0513 09:45:42.099827 140460907546368 logging_writer.py:48] [107500] global_step=107500, grad_norm=3.9842236042022705, loss=3.7497849464416504 -I0513 09:46:06.840885 140678261474496 spec.py:333] Evaluating on the training split. -I0513 09:46:22.854856 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 09:46:43.943943 140678261474496 spec.py:363] Evaluating on the test split. -I0513 09:46:45.057501 140678261474496 submission_runner.py:516] Time since start: 50397.33s, Step: 107550, {'train/accuracy': Array(0.00145488, dtype=float32), 'train/loss': Array(7.7392006, dtype=float32), 'validation/accuracy': Array(0.00136, dtype=float32), 'validation/loss': Array(7.724337, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.001, dtype=float32), 'test/loss': Array(7.727155, dtype=float32), 'test/num_examples': 10000, 'score': 47972.51910805702, 'total_duration': 50397.326488018036, 'accumulated_submission_time': 47972.51910805702, 'accumulated_eval_time': 2421.4725844860077, 'accumulated_logging_time': 1.7976596355438232} -I0513 09:46:45.792356 140460899153664 logging_writer.py:48] [107550] accumulated_eval_time=2421.47, accumulated_logging_time=1.79766, accumulated_submission_time=47972.5, global_step=107550, preemption_count=0, score=47972.5, test/accuracy=0.0010000000474974513, test/loss=7.7271552085876465, test/num_examples=10000, total_duration=50397.3, train/accuracy=0.001454878831282258, train/loss=7.739200592041016, validation/accuracy=0.0013599999947473407, validation/loss=7.724337100982666, validation/num_examples=50000 -I0513 09:47:07.762918 140460907546368 logging_writer.py:48] [107600] global_step=107600, grad_norm=4.8747429847717285, loss=3.899897813796997 -I0513 09:47:54.421104 140460899153664 logging_writer.py:48] [107700] global_step=107700, grad_norm=3.9105734825134277, loss=3.8328323364257812 -I0513 09:48:43.493439 140460907546368 logging_writer.py:48] [107800] global_step=107800, grad_norm=6.699464321136475, loss=3.749650239944458 -I0513 09:49:30.323408 140460899153664 logging_writer.py:48] [107900] global_step=107900, grad_norm=4.894131183624268, loss=3.7485618591308594 -I0513 09:50:22.116173 140460907546368 logging_writer.py:48] [108000] global_step=108000, grad_norm=4.87468147277832, loss=3.8815321922302246 -I0513 09:51:12.367134 140460899153664 logging_writer.py:48] [108100] global_step=108100, grad_norm=3.6567413806915283, loss=3.6761369705200195 -I0513 09:52:10.499263 140460907546368 logging_writer.py:48] [108200] global_step=108200, grad_norm=3.687344551086426, loss=3.814520835876465 -I0513 09:53:10.424264 140460899153664 logging_writer.py:48] [108300] global_step=108300, grad_norm=4.357717037200928, loss=3.731874942779541 -I0513 09:54:05.031501 140460907546368 logging_writer.py:48] [108400] global_step=108400, grad_norm=10.989595413208008, loss=3.986215114593506 -I0513 09:54:57.538093 140460899153664 logging_writer.py:48] [108500] global_step=108500, grad_norm=3.4526166915893555, loss=3.9142303466796875 -I0513 09:55:56.133013 140460907546368 logging_writer.py:48] [108600] global_step=108600, grad_norm=3.7263100147247314, loss=3.890279769897461 -I0513 09:56:50.209956 140460899153664 logging_writer.py:48] [108700] global_step=108700, grad_norm=4.3336615562438965, loss=3.8466458320617676 -I0513 09:57:38.165505 140460907546368 logging_writer.py:48] [108800] global_step=108800, grad_norm=4.263465881347656, loss=3.746877670288086 -I0513 09:58:35.208068 140460899153664 logging_writer.py:48] [108900] global_step=108900, grad_norm=3.5314197540283203, loss=3.637552261352539 -I0513 09:59:27.808755 140460907546368 logging_writer.py:48] [109000] global_step=109000, grad_norm=4.757903099060059, loss=3.8238162994384766 -I0513 10:00:19.714089 140460899153664 logging_writer.py:48] [109100] global_step=109100, grad_norm=9.606913566589355, loss=3.9764676094055176 -I0513 10:01:12.714395 140460907546368 logging_writer.py:48] [109200] global_step=109200, grad_norm=8.370362281799316, loss=3.777028799057007 -I0513 10:02:00.310318 140460899153664 logging_writer.py:48] [109300] global_step=109300, grad_norm=4.7246527671813965, loss=3.8257508277893066 -I0513 10:02:53.095280 140460907546368 logging_writer.py:48] [109400] global_step=109400, grad_norm=5.2756123542785645, loss=3.872981071472168 -I0513 10:03:51.965609 140460899153664 logging_writer.py:48] [109500] global_step=109500, grad_norm=3.793165683746338, loss=3.758946180343628 -I0513 10:04:48.284610 140460907546368 logging_writer.py:48] [109600] global_step=109600, grad_norm=5.455047607421875, loss=3.8866865634918213 -I0513 10:05:47.450433 140460899153664 logging_writer.py:48] [109700] global_step=109700, grad_norm=3.2544755935668945, loss=3.6385269165039062 -I0513 10:06:48.923144 140460907546368 logging_writer.py:48] [109800] global_step=109800, grad_norm=4.579129219055176, loss=3.951742172241211 -I0513 10:07:42.120087 140460899153664 logging_writer.py:48] [109900] global_step=109900, grad_norm=4.822222709655762, loss=3.8741455078125 -I0513 10:08:32.446669 140460907546368 logging_writer.py:48] [110000] global_step=110000, grad_norm=6.971320152282715, loss=3.811682939529419 -I0513 10:09:23.415158 140460899153664 logging_writer.py:48] [110100] global_step=110100, grad_norm=6.0194993019104, loss=3.9000730514526367 -I0513 10:10:10.267459 140460907546368 logging_writer.py:48] [110200] global_step=110200, grad_norm=4.5593719482421875, loss=3.8584742546081543 -I0513 10:11:02.096600 140460899153664 logging_writer.py:48] [110300] global_step=110300, grad_norm=8.31136417388916, loss=3.9821934700012207 -I0513 10:11:58.229788 140460907546368 logging_writer.py:48] [110400] global_step=110400, grad_norm=3.396399974822998, loss=3.741450071334839 -I0513 10:12:46.040858 140460899153664 logging_writer.py:48] [110500] global_step=110500, grad_norm=3.6275269985198975, loss=3.885382652282715 -I0513 10:13:35.428447 140460907546368 logging_writer.py:48] [110600] global_step=110600, grad_norm=3.6115617752075195, loss=3.9216763973236084 -I0513 10:14:26.186056 140460899153664 logging_writer.py:48] [110700] global_step=110700, grad_norm=5.254197120666504, loss=3.8007194995880127 -I0513 10:15:17.486263 140460907546368 logging_writer.py:48] [110800] global_step=110800, grad_norm=4.214099407196045, loss=3.757906198501587 -I0513 10:16:10.325120 140460899153664 logging_writer.py:48] [110900] global_step=110900, grad_norm=4.669919490814209, loss=3.8001770973205566 -I0513 10:17:01.246916 140460907546368 logging_writer.py:48] [111000] global_step=111000, grad_norm=8.200010299682617, loss=3.8763840198516846 -I0513 10:17:51.281149 140460899153664 logging_writer.py:48] [111100] global_step=111100, grad_norm=2.9364399909973145, loss=3.77557635307312 -I0513 10:18:44.174780 140460907546368 logging_writer.py:48] [111200] global_step=111200, grad_norm=11.575361251831055, loss=3.884000301361084 -I0513 10:19:40.716816 140460899153664 logging_writer.py:48] [111300] global_step=111300, grad_norm=8.84481430053711, loss=3.8905224800109863 -I0513 10:20:00.970034 140678261474496 spec.py:333] Evaluating on the training split. -I0513 10:20:09.358201 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 10:20:33.453137 140678261474496 spec.py:363] Evaluating on the test split. -I0513 10:20:34.481721 140678261474496 submission_runner.py:516] Time since start: 52426.83s, Step: 111343, {'train/accuracy': Array(0.00141502, dtype=float32), 'train/loss': Array(7.680137, dtype=float32), 'validation/accuracy': Array(0.00136, dtype=float32), 'validation/loss': Array(7.674657, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.001, dtype=float32), 'test/loss': Array(7.677686, dtype=float32), 'test/num_examples': 10000, 'score': 49967.550864458084, 'total_duration': 52426.83486223221, 'accumulated_submission_time': 49967.550864458084, 'accumulated_eval_time': 2454.8481583595276, 'accumulated_logging_time': 2.6185977458953857} -I0513 10:20:34.799387 140460907546368 logging_writer.py:48] [111343] accumulated_eval_time=2454.85, accumulated_logging_time=2.6186, accumulated_submission_time=49967.6, global_step=111343, preemption_count=0, score=49967.6, test/accuracy=0.0010000000474974513, test/loss=7.6776862144470215, test/num_examples=10000, total_duration=52426.8, train/accuracy=0.0014150191564112902, train/loss=7.6801371574401855, validation/accuracy=0.0013599999947473407, validation/loss=7.674656867980957, validation/num_examples=50000 -I0513 10:21:19.569605 140460899153664 logging_writer.py:48] [111400] global_step=111400, grad_norm=4.558562755584717, loss=3.5818238258361816 -I0513 10:22:25.672505 140460907546368 logging_writer.py:48] [111500] global_step=111500, grad_norm=3.5310542583465576, loss=3.8304152488708496 -I0513 10:23:22.635987 140460899153664 logging_writer.py:48] [111600] global_step=111600, grad_norm=7.097387313842773, loss=3.9712424278259277 -I0513 10:24:14.784283 140460907546368 logging_writer.py:48] [111700] global_step=111700, grad_norm=7.160634517669678, loss=3.908433437347412 -I0513 10:25:04.614826 140460899153664 logging_writer.py:48] [111800] global_step=111800, grad_norm=3.853031635284424, loss=3.9092254638671875 -I0513 10:25:57.149419 140460907546368 logging_writer.py:48] [111900] global_step=111900, grad_norm=5.164126873016357, loss=3.960165023803711 -I0513 10:26:49.788756 140460899153664 logging_writer.py:48] [112000] global_step=112000, grad_norm=4.634699821472168, loss=3.83396577835083 -I0513 10:27:44.595041 140460907546368 logging_writer.py:48] [112100] global_step=112100, grad_norm=10.173685073852539, loss=3.903892993927002 -I0513 10:28:35.574846 140460899153664 logging_writer.py:48] [112200] global_step=112200, grad_norm=6.096891403198242, loss=3.7418618202209473 -I0513 10:29:31.169225 140460907546368 logging_writer.py:48] [112300] global_step=112300, grad_norm=24.954254150390625, loss=3.9198126792907715 -I0513 10:30:26.066084 140460899153664 logging_writer.py:48] [112400] global_step=112400, grad_norm=3.627126932144165, loss=3.7555675506591797 -I0513 10:31:31.108951 140460907546368 logging_writer.py:48] [112500] global_step=112500, grad_norm=5.839941501617432, loss=3.945071220397949 -2026-05-13 10:32:10.852252: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 10:32:22.611848 140460899153664 logging_writer.py:48] [112600] global_step=112600, grad_norm=5.378479957580566, loss=4.009416580200195 -I0513 10:33:13.196921 140460907546368 logging_writer.py:48] [112700] global_step=112700, grad_norm=3.3897125720977783, loss=3.8341355323791504 -I0513 10:34:16.346039 140460899153664 logging_writer.py:48] [112800] global_step=112800, grad_norm=3.5607964992523193, loss=3.6959035396575928 -I0513 10:35:26.326002 140460907546368 logging_writer.py:48] [112900] global_step=112900, grad_norm=3.434936285018921, loss=3.790040969848633 -I0513 10:36:33.885020 140460899153664 logging_writer.py:48] [113000] global_step=113000, grad_norm=4.321892261505127, loss=3.743619203567505 -I0513 10:37:29.777425 140460907546368 logging_writer.py:48] [113100] global_step=113100, grad_norm=8.68310260772705, loss=4.041650295257568 -I0513 10:38:42.862982 140460899153664 logging_writer.py:48] [113200] global_step=113200, grad_norm=4.616185188293457, loss=3.8269927501678467 -I0513 10:39:34.505625 140460907546368 logging_writer.py:48] [113300] global_step=113300, grad_norm=3.811770439147949, loss=3.7935166358947754 -I0513 10:40:25.225098 140460899153664 logging_writer.py:48] [113400] global_step=113400, grad_norm=7.415616989135742, loss=3.9786534309387207 -I0513 10:41:16.155177 140460907546368 logging_writer.py:48] [113500] global_step=113500, grad_norm=11.612405776977539, loss=4.056801795959473 -I0513 10:42:04.745977 140460899153664 logging_writer.py:48] [113600] global_step=113600, grad_norm=5.437489986419678, loss=3.814197063446045 -I0513 10:43:06.938196 140460907546368 logging_writer.py:48] [113700] global_step=113700, grad_norm=7.785743236541748, loss=3.84440541267395 -I0513 10:43:57.203288 140460899153664 logging_writer.py:48] [113800] global_step=113800, grad_norm=6.713485240936279, loss=3.891310214996338 -2026-05-13 10:44:09.697750: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 10:44:47.567129 140460907546368 logging_writer.py:48] [113900] global_step=113900, grad_norm=4.372097015380859, loss=3.760233163833618 -I0513 10:45:37.818326 140460899153664 logging_writer.py:48] [114000] global_step=114000, grad_norm=4.9440016746521, loss=3.9106457233428955 -I0513 10:46:29.254822 140460907546368 logging_writer.py:48] [114100] global_step=114100, grad_norm=5.532032489776611, loss=3.8592467308044434 -I0513 10:47:22.344196 140460899153664 logging_writer.py:48] [114200] global_step=114200, grad_norm=6.165235996246338, loss=3.801546096801758 -I0513 10:48:27.183024 140460907546368 logging_writer.py:48] [114300] global_step=114300, grad_norm=4.749051094055176, loss=3.885162830352783 -I0513 10:49:18.433058 140460899153664 logging_writer.py:48] [114400] global_step=114400, grad_norm=3.228023052215576, loss=3.801393508911133 -I0513 10:50:05.937324 140460907546368 logging_writer.py:48] [114500] global_step=114500, grad_norm=12.149544715881348, loss=3.7691285610198975 -I0513 10:50:58.540942 140460899153664 logging_writer.py:48] [114600] global_step=114600, grad_norm=5.477750778198242, loss=4.016438961029053 -I0513 10:52:04.687189 140460907546368 logging_writer.py:48] [114700] global_step=114700, grad_norm=4.650347709655762, loss=3.88940691947937 -I0513 10:53:06.695180 140460899153664 logging_writer.py:48] [114800] global_step=114800, grad_norm=42.974365234375, loss=3.967489719390869 -I0513 10:53:50.660349 140678261474496 spec.py:333] Evaluating on the training split. -I0513 10:53:58.465894 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 10:54:33.706970 140678261474496 spec.py:363] Evaluating on the test split. -I0513 10:54:34.785135 140678261474496 submission_runner.py:516] Time since start: 54467.09s, Step: 114877, {'train/accuracy': Array(0.00137516, dtype=float32), 'train/loss': Array(7.636404, dtype=float32), 'validation/accuracy': Array(0.00144, dtype=float32), 'validation/loss': Array(7.6376934, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.001, dtype=float32), 'test/loss': Array(7.640008, dtype=float32), 'test/num_examples': 10000, 'score': 51963.30955648422, 'total_duration': 54467.09082698822, 'accumulated_submission_time': 51963.30955648422, 'accumulated_eval_time': 2498.7893834114075, 'accumulated_logging_time': 2.9784514904022217} -I0513 10:54:35.127146 140460907546368 logging_writer.py:48] [114877] accumulated_eval_time=2498.79, accumulated_logging_time=2.97845, accumulated_submission_time=51963.3, global_step=114877, preemption_count=0, score=51963.3, test/accuracy=0.0010000000474974513, test/loss=7.640007972717285, test/num_examples=10000, total_duration=54467.1, train/accuracy=0.0013751593651250005, train/loss=7.636404037475586, validation/accuracy=0.0014400000218302011, validation/loss=7.637693405151367, validation/num_examples=50000 -I0513 10:54:47.083644 140460899153664 logging_writer.py:48] [114900] global_step=114900, grad_norm=5.260801315307617, loss=3.794264793395996 -I0513 10:55:57.374940 140460907546368 logging_writer.py:48] [115000] global_step=115000, grad_norm=3.1204230785369873, loss=3.819068193435669 -I0513 10:57:02.198166 140460899153664 logging_writer.py:48] [115100] global_step=115100, grad_norm=4.697772026062012, loss=3.996001720428467 -I0513 10:57:57.674497 140460907546368 logging_writer.py:48] [115200] global_step=115200, grad_norm=2.941760540008545, loss=3.7378435134887695 -I0513 10:58:53.145982 140460899153664 logging_writer.py:48] [115300] global_step=115300, grad_norm=3.9566121101379395, loss=3.899348735809326 -I0513 10:59:45.333063 140460907546368 logging_writer.py:48] [115400] global_step=115400, grad_norm=8.20370101928711, loss=3.997238874435425 -I0513 11:00:35.422116 140460899153664 logging_writer.py:48] [115500] global_step=115500, grad_norm=4.898102283477783, loss=3.793091297149658 -I0513 11:01:29.722356 140460907546368 logging_writer.py:48] [115600] global_step=115600, grad_norm=3.5041568279266357, loss=3.872262716293335 -I0513 11:02:35.198498 140460899153664 logging_writer.py:48] [115700] global_step=115700, grad_norm=3.2715988159179688, loss=3.8190836906433105 -I0513 11:03:27.698020 140460907546368 logging_writer.py:48] [115800] global_step=115800, grad_norm=3.984297752380371, loss=3.690706968307495 -I0513 11:04:14.510299 140460899153664 logging_writer.py:48] [115900] global_step=115900, grad_norm=8.452103614807129, loss=3.839488983154297 -I0513 11:05:04.728581 140460907546368 logging_writer.py:48] [116000] global_step=116000, grad_norm=3.1625516414642334, loss=3.8826098442077637 -I0513 11:05:56.841776 140460899153664 logging_writer.py:48] [116100] global_step=116100, grad_norm=7.055395603179932, loss=3.7837181091308594 -I0513 11:06:47.855728 140460907546368 logging_writer.py:48] [116200] global_step=116200, grad_norm=4.061038494110107, loss=3.8738746643066406 -I0513 11:07:37.679347 140460899153664 logging_writer.py:48] [116300] global_step=116300, grad_norm=3.7737953662872314, loss=3.7913658618927 -2026-05-13 11:07:52.586177: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 11:08:32.616525 140460907546368 logging_writer.py:48] [116400] global_step=116400, grad_norm=4.785337924957275, loss=3.8094539642333984 -I0513 11:09:22.783750 140460899153664 logging_writer.py:48] [116500] global_step=116500, grad_norm=22.975156784057617, loss=4.022401809692383 -I0513 11:10:12.589682 140460907546368 logging_writer.py:48] [116600] global_step=116600, grad_norm=19.519323348999023, loss=3.9278411865234375 -I0513 11:11:08.918143 140460899153664 logging_writer.py:48] [116700] global_step=116700, grad_norm=5.11399507522583, loss=3.750986099243164 -I0513 11:12:03.403896 140460907546368 logging_writer.py:48] [116800] global_step=116800, grad_norm=3.6949081420898438, loss=3.906851291656494 -I0513 11:12:49.569920 140460899153664 logging_writer.py:48] [116900] global_step=116900, grad_norm=3.2508182525634766, loss=3.7153542041778564 -I0513 11:13:40.477466 140460907546368 logging_writer.py:48] [117000] global_step=117000, grad_norm=6.804659366607666, loss=3.819336175918579 -I0513 11:14:28.596365 140460899153664 logging_writer.py:48] [117100] global_step=117100, grad_norm=5.239307403564453, loss=3.635225772857666 -I0513 11:15:16.764062 140460907546368 logging_writer.py:48] [117200] global_step=117200, grad_norm=4.1766815185546875, loss=3.970816135406494 -I0513 11:16:04.377401 140460899153664 logging_writer.py:48] [117300] global_step=117300, grad_norm=6.134846210479736, loss=3.90327525138855 -I0513 11:16:57.056768 140460907546368 logging_writer.py:48] [117400] global_step=117400, grad_norm=6.616902828216553, loss=3.692502498626709 -I0513 11:18:02.445416 140460899153664 logging_writer.py:48] [117500] global_step=117500, grad_norm=4.665860176086426, loss=3.8944239616394043 -2026-05-13 11:18:46.099713: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 11:18:55.613661 140460907546368 logging_writer.py:48] [117600] global_step=117600, grad_norm=3.7157442569732666, loss=3.7687888145446777 -I0513 11:20:19.648736 140460899153664 logging_writer.py:48] [117700] global_step=117700, grad_norm=4.489870071411133, loss=3.8047571182250977 -I0513 11:21:09.017704 140460907546368 logging_writer.py:48] [117800] global_step=117800, grad_norm=5.529723644256592, loss=3.9669337272644043 -I0513 11:22:00.537839 140460899153664 logging_writer.py:48] [117900] global_step=117900, grad_norm=6.093075752258301, loss=3.8389368057250977 -I0513 11:22:53.577207 140460907546368 logging_writer.py:48] [118000] global_step=118000, grad_norm=4.34080696105957, loss=3.809009075164795 -I0513 11:23:50.298700 140460899153664 logging_writer.py:48] [118100] global_step=118100, grad_norm=10.439270973205566, loss=3.931492328643799 -I0513 11:24:42.882781 140460907546368 logging_writer.py:48] [118200] global_step=118200, grad_norm=9.607900619506836, loss=4.087333679199219 -I0513 11:25:30.739178 140460899153664 logging_writer.py:48] [118300] global_step=118300, grad_norm=7.08602237701416, loss=3.9159746170043945 -I0513 11:26:22.877356 140460907546368 logging_writer.py:48] [118400] global_step=118400, grad_norm=5.136185169219971, loss=3.9216256141662598 -I0513 11:27:24.394539 140460899153664 logging_writer.py:48] [118500] global_step=118500, grad_norm=5.217095851898193, loss=3.7804489135742188 -I0513 11:27:50.979077 140678261474496 spec.py:333] Evaluating on the training split. -I0513 11:27:59.016637 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 11:28:22.948596 140678261474496 spec.py:363] Evaluating on the test split. -I0513 11:28:24.120006 140678261474496 submission_runner.py:516] Time since start: 56496.33s, Step: 118552, {'train/accuracy': Array(0.0013353, dtype=float32), 'train/loss': Array(7.5854464, dtype=float32), 'validation/accuracy': Array(0.00138, dtype=float32), 'validation/loss': Array(7.598124, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.001, dtype=float32), 'test/loss': Array(7.600594, dtype=float32), 'test/num_examples': 10000, 'score': 53959.05523252487, 'total_duration': 56496.327038526535, 'accumulated_submission_time': 53959.05523252487, 'accumulated_eval_time': 2531.6480898857117, 'accumulated_logging_time': 3.3708062171936035} -I0513 11:28:24.460448 140460907546368 logging_writer.py:48] [118552] accumulated_eval_time=2531.65, accumulated_logging_time=3.37081, accumulated_submission_time=53959.1, global_step=118552, preemption_count=0, score=53959.1, test/accuracy=0.0010000000474974513, test/loss=7.6005940437316895, test/num_examples=10000, total_duration=56496.3, train/accuracy=0.0013352996902540326, train/loss=7.585446357727051, validation/accuracy=0.0013799999142065644, validation/loss=7.598124027252197, validation/num_examples=50000 -I0513 11:28:56.240951 140460899153664 logging_writer.py:48] [118600] global_step=118600, grad_norm=2.91705060005188, loss=3.7853336334228516 -I0513 11:29:58.258959 140460907546368 logging_writer.py:48] [118700] global_step=118700, grad_norm=7.2469000816345215, loss=3.7243497371673584 -I0513 11:30:56.715489 140460899153664 logging_writer.py:48] [118800] global_step=118800, grad_norm=5.357443332672119, loss=3.7737488746643066 -I0513 11:31:49.826179 140460907546368 logging_writer.py:48] [118900] global_step=118900, grad_norm=5.116738319396973, loss=3.7952497005462646 -I0513 11:32:37.081884 140460899153664 logging_writer.py:48] [119000] global_step=119000, grad_norm=34.149845123291016, loss=3.92466402053833 -I0513 11:33:28.449249 140460907546368 logging_writer.py:48] [119100] global_step=119100, grad_norm=6.782682418823242, loss=3.9249234199523926 -I0513 11:34:17.797299 140460899153664 logging_writer.py:48] [119200] global_step=119200, grad_norm=50.14460754394531, loss=3.932544708251953 -I0513 11:35:13.532245 140460907546368 logging_writer.py:48] [119300] global_step=119300, grad_norm=4.047405242919922, loss=3.9120934009552 -I0513 11:36:12.037802 140460899153664 logging_writer.py:48] [119400] global_step=119400, grad_norm=5.097559452056885, loss=3.926788330078125 -I0513 11:36:59.602034 140460907546368 logging_writer.py:48] [119500] global_step=119500, grad_norm=6.362913608551025, loss=3.858896017074585 -I0513 11:37:56.011667 140460899153664 logging_writer.py:48] [119600] global_step=119600, grad_norm=5.3494110107421875, loss=3.8209609985351562 -I0513 11:38:51.106351 140460907546368 logging_writer.py:48] [119700] global_step=119700, grad_norm=4.310811996459961, loss=3.774264335632324 -I0513 11:39:55.401710 140460899153664 logging_writer.py:48] [119800] global_step=119800, grad_norm=6.742356300354004, loss=3.7501072883605957 -I0513 11:40:44.112846 140460907546368 logging_writer.py:48] [119900] global_step=119900, grad_norm=8.4390287399292, loss=3.8113813400268555 -I0513 11:41:38.734397 140460899153664 logging_writer.py:48] [120000] global_step=120000, grad_norm=3.1274843215942383, loss=3.8376145362854004 -I0513 11:42:26.236552 140460907546368 logging_writer.py:48] [120100] global_step=120100, grad_norm=9.984424591064453, loss=3.7386038303375244 -I0513 11:43:15.682718 140460899153664 logging_writer.py:48] [120200] global_step=120200, grad_norm=10.522790908813477, loss=3.829829216003418 -I0513 11:44:07.363137 140460907546368 logging_writer.py:48] [120300] global_step=120300, grad_norm=13.487242698669434, loss=4.0220537185668945 -I0513 11:44:59.056282 140460899153664 logging_writer.py:48] [120400] global_step=120400, grad_norm=3.413917064666748, loss=3.8758246898651123 -I0513 11:45:52.982663 140460907546368 logging_writer.py:48] [120500] global_step=120500, grad_norm=3.6048500537872314, loss=3.9735639095306396 -I0513 11:46:44.259530 140460899153664 logging_writer.py:48] [120600] global_step=120600, grad_norm=8.38576602935791, loss=3.8088488578796387 -I0513 11:47:33.969798 140460907546368 logging_writer.py:48] [120700] global_step=120700, grad_norm=2.6121182441711426, loss=3.8589305877685547 -I0513 11:48:22.114336 140460899153664 logging_writer.py:48] [120800] global_step=120800, grad_norm=6.1229023933410645, loss=3.8469619750976562 -I0513 11:49:11.755295 140460907546368 logging_writer.py:48] [120900] global_step=120900, grad_norm=5.071491718292236, loss=3.807886838912964 -I0513 11:50:00.506171 140460899153664 logging_writer.py:48] [121000] global_step=121000, grad_norm=3.681756019592285, loss=3.8346261978149414 -I0513 11:50:48.908247 140460907546368 logging_writer.py:48] [121100] global_step=121100, grad_norm=3.9362430572509766, loss=3.853741407394409 -I0513 11:51:41.091326 140460899153664 logging_writer.py:48] [121200] global_step=121200, grad_norm=6.682533264160156, loss=3.765807628631592 -I0513 11:52:32.694773 140460907546368 logging_writer.py:48] [121300] global_step=121300, grad_norm=5.030050277709961, loss=3.96640682220459 -2026-05-13 11:52:48.264241: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 11:53:57.025497 140460899153664 logging_writer.py:48] [121400] global_step=121400, grad_norm=6.9828033447265625, loss=3.709566831588745 -I0513 11:54:54.588490 140460907546368 logging_writer.py:48] [121500] global_step=121500, grad_norm=3.832500457763672, loss=3.769562244415283 -I0513 11:55:42.443257 140460899153664 logging_writer.py:48] [121600] global_step=121600, grad_norm=4.083047389984131, loss=3.7805018424987793 -I0513 11:56:30.089905 140460907546368 logging_writer.py:48] [121700] global_step=121700, grad_norm=8.026565551757812, loss=3.958927631378174 -I0513 11:57:21.816985 140460899153664 logging_writer.py:48] [121800] global_step=121800, grad_norm=8.401272773742676, loss=3.899522542953491 -I0513 11:58:13.575093 140460907546368 logging_writer.py:48] [121900] global_step=121900, grad_norm=5.786273956298828, loss=3.8663506507873535 -I0513 11:59:15.117306 140460899153664 logging_writer.py:48] [122000] global_step=122000, grad_norm=7.407585620880127, loss=3.8134336471557617 -I0513 12:00:11.724967 140460907546368 logging_writer.py:48] [122100] global_step=122100, grad_norm=3.1879637241363525, loss=3.8665192127227783 -I0513 12:01:25.378377 140460899153664 logging_writer.py:48] [122200] global_step=122200, grad_norm=50.078521728515625, loss=3.9355392456054688 -I0513 12:01:40.324237 140678261474496 spec.py:333] Evaluating on the training split. -I0513 12:01:47.535844 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 12:02:13.964872 140678261474496 spec.py:363] Evaluating on the test split. -I0513 12:02:15.020335 140678261474496 submission_runner.py:516] Time since start: 58527.35s, Step: 122227, {'train/accuracy': Array(0.00157446, dtype=float32), 'train/loss': Array(7.5687747, dtype=float32), 'validation/accuracy': Array(0.00144, dtype=float32), 'validation/loss': Array(7.560353, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(7.5624433, dtype=float32), 'test/num_examples': 10000, 'score': 55954.80316567421, 'total_duration': 58527.34854841232, 'accumulated_submission_time': 55954.80316567421, 'accumulated_eval_time': 2566.183147907257, 'accumulated_logging_time': 3.765636682510376} -I0513 12:02:15.361861 140460907546368 logging_writer.py:48] [122227] accumulated_eval_time=2566.18, accumulated_logging_time=3.76564, accumulated_submission_time=55954.8, global_step=122227, preemption_count=0, score=55954.8, test/accuracy=0.0010999999940395355, test/loss=7.562443256378174, test/num_examples=10000, total_duration=58527.3, train/accuracy=0.0015744578558951616, train/loss=7.568774700164795, validation/accuracy=0.0014400000218302011, validation/loss=7.560352802276611, validation/num_examples=50000 -I0513 12:03:00.375711 140460899153664 logging_writer.py:48] [122300] global_step=122300, grad_norm=13.227065086364746, loss=3.7695200443267822 -I0513 12:03:57.739548 140460907546368 logging_writer.py:48] [122400] global_step=122400, grad_norm=15.801810264587402, loss=4.055057048797607 -I0513 12:04:57.205477 140460899153664 logging_writer.py:48] [122500] global_step=122500, grad_norm=5.372288703918457, loss=3.8980822563171387 -I0513 12:06:07.005947 140460907546368 logging_writer.py:48] [122600] global_step=122600, grad_norm=3.996983766555786, loss=3.7590222358703613 -I0513 12:06:59.441546 140460899153664 logging_writer.py:48] [122700] global_step=122700, grad_norm=5.162598133087158, loss=3.9335126876831055 -I0513 12:07:51.132786 140460907546368 logging_writer.py:48] [122800] global_step=122800, grad_norm=3.751394748687744, loss=3.8516573905944824 -I0513 12:08:42.359944 140460899153664 logging_writer.py:48] [122900] global_step=122900, grad_norm=227.5292510986328, loss=4.626815319061279 -I0513 12:09:34.703748 140460907546368 logging_writer.py:48] [123000] global_step=123000, grad_norm=3.798992872238159, loss=3.941772937774658 -I0513 12:10:31.860920 140460899153664 logging_writer.py:48] [123100] global_step=123100, grad_norm=4.517099857330322, loss=3.79453706741333 -I0513 12:11:21.184283 140460907546368 logging_writer.py:48] [123200] global_step=123200, grad_norm=3.936816453933716, loss=3.9833714962005615 -I0513 12:12:14.533768 140460899153664 logging_writer.py:48] [123300] global_step=123300, grad_norm=3.2724320888519287, loss=3.732659339904785 -I0513 12:13:08.973056 140460907546368 logging_writer.py:48] [123400] global_step=123400, grad_norm=3.482522487640381, loss=3.8364622592926025 -I0513 12:14:01.717103 140460899153664 logging_writer.py:48] [123500] global_step=123500, grad_norm=5.915541648864746, loss=3.8025097846984863 -I0513 12:14:49.065544 140460907546368 logging_writer.py:48] [123600] global_step=123600, grad_norm=6.736523628234863, loss=3.9164657592773438 -I0513 12:15:58.917777 140460899153664 logging_writer.py:48] [123700] global_step=123700, grad_norm=4.480875015258789, loss=3.699223041534424 -I0513 12:17:01.897138 140460907546368 logging_writer.py:48] [123800] global_step=123800, grad_norm=3.9372828006744385, loss=3.6785528659820557 -I0513 12:18:10.423371 140460899153664 logging_writer.py:48] [123900] global_step=123900, grad_norm=4.44668436050415, loss=3.7098114490509033 -I0513 12:19:16.647397 140460907546368 logging_writer.py:48] [124000] global_step=124000, grad_norm=6.025877952575684, loss=3.944643974304199 -I0513 12:20:15.055301 140460899153664 logging_writer.py:48] [124100] global_step=124100, grad_norm=4.650511264801025, loss=3.832779884338379 -I0513 12:21:10.576508 140460907546368 logging_writer.py:48] [124200] global_step=124200, grad_norm=30.97605323791504, loss=3.9006989002227783 -I0513 12:22:00.124969 140460899153664 logging_writer.py:48] [124300] global_step=124300, grad_norm=3.409669876098633, loss=4.033926486968994 -I0513 12:22:49.241014 140460907546368 logging_writer.py:48] [124400] global_step=124400, grad_norm=5.105112552642822, loss=3.8226027488708496 -I0513 12:23:47.673975 140460899153664 logging_writer.py:48] [124500] global_step=124500, grad_norm=4.850347995758057, loss=3.7961647510528564 -I0513 12:24:37.014903 140460907546368 logging_writer.py:48] [124600] global_step=124600, grad_norm=4.334569931030273, loss=3.8054215908050537 -I0513 12:25:28.448145 140460899153664 logging_writer.py:48] [124700] global_step=124700, grad_norm=6.322381973266602, loss=3.9499711990356445 -I0513 12:26:22.655098 140460907546368 logging_writer.py:48] [124800] global_step=124800, grad_norm=10.476749420166016, loss=3.9227676391601562 -I0513 12:27:18.968153 140460899153664 logging_writer.py:48] [124900] global_step=124900, grad_norm=4.898588180541992, loss=4.030034065246582 -I0513 12:28:13.261421 140460907546368 logging_writer.py:48] [125000] global_step=125000, grad_norm=8.93137264251709, loss=3.890315532684326 -I0513 12:29:07.445710 140460899153664 logging_writer.py:48] [125100] global_step=125100, grad_norm=3.942125082015991, loss=3.830170154571533 -I0513 12:30:01.691911 140460907546368 logging_writer.py:48] [125200] global_step=125200, grad_norm=6.800615310668945, loss=4.034303188323975 -I0513 12:30:51.527055 140460899153664 logging_writer.py:48] [125300] global_step=125300, grad_norm=6.582274913787842, loss=3.8847298622131348 -I0513 12:31:40.574818 140460907546368 logging_writer.py:48] [125400] global_step=125400, grad_norm=8.590882301330566, loss=3.9814069271087646 -I0513 12:32:35.497751 140460899153664 logging_writer.py:48] [125500] global_step=125500, grad_norm=5.0694098472595215, loss=3.783996105194092 -I0513 12:33:26.543046 140460907546368 logging_writer.py:48] [125600] global_step=125600, grad_norm=4.5061798095703125, loss=3.773998260498047 -I0513 12:34:16.311612 140460899153664 logging_writer.py:48] [125700] global_step=125700, grad_norm=4.729523181915283, loss=3.7853894233703613 -I0513 12:35:17.092553 140460907546368 logging_writer.py:48] [125800] global_step=125800, grad_norm=5.674349784851074, loss=3.7327632904052734 -I0513 12:35:31.595340 140678261474496 spec.py:333] Evaluating on the training split. -I0513 12:35:41.317151 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 12:36:43.627938 140678261474496 spec.py:363] Evaluating on the test split. -I0513 12:36:44.717082 140678261474496 submission_runner.py:516] Time since start: 60597.01s, Step: 125812, {'train/accuracy': Array(0.00165418, dtype=float32), 'train/loss': Array(7.515965, dtype=float32), 'validation/accuracy': Array(0.00158, dtype=float32), 'validation/loss': Array(7.518538, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(7.5200286, dtype=float32), 'test/num_examples': 10000, 'score': 57950.92701792717, 'total_duration': 60597.006568431854, 'accumulated_submission_time': 57950.92701792717, 'accumulated_eval_time': 2639.1051247119904, 'accumulated_logging_time': 4.160260915756226} -I0513 12:36:45.059647 140460899153664 logging_writer.py:48] [125812] accumulated_eval_time=2639.11, accumulated_logging_time=4.16026, accumulated_submission_time=57950.9, global_step=125812, preemption_count=0, score=57950.9, test/accuracy=0.0013000001199543476, test/loss=7.520028591156006, test/num_examples=10000, total_duration=60597, train/accuracy=0.0016541772056370974, train/loss=7.515964984893799, validation/accuracy=0.0015799999237060547, validation/loss=7.518537998199463, validation/num_examples=50000 -I0513 12:37:42.220446 140460907546368 logging_writer.py:48] [125900] global_step=125900, grad_norm=5.043823719024658, loss=3.8086071014404297 -I0513 12:38:47.546800 140460899153664 logging_writer.py:48] [126000] global_step=126000, grad_norm=3.493056535720825, loss=3.758488655090332 -I0513 12:39:57.818825 140460907546368 logging_writer.py:48] [126100] global_step=126100, grad_norm=7.24391508102417, loss=3.848954677581787 -I0513 12:41:03.591325 140460899153664 logging_writer.py:48] [126200] global_step=126200, grad_norm=5.679778575897217, loss=3.844296932220459 -I0513 12:41:53.652917 140460907546368 logging_writer.py:48] [126300] global_step=126300, grad_norm=3.688568353652954, loss=3.8645262718200684 -I0513 12:42:45.170801 140460899153664 logging_writer.py:48] [126400] global_step=126400, grad_norm=10.036218643188477, loss=3.794710397720337 -I0513 12:43:34.512718 140460907546368 logging_writer.py:48] [126500] global_step=126500, grad_norm=5.371102809906006, loss=3.809680461883545 -I0513 12:44:31.217214 140460899153664 logging_writer.py:48] [126600] global_step=126600, grad_norm=4.38192892074585, loss=3.8689870834350586 -I0513 12:45:31.267066 140460907546368 logging_writer.py:48] [126700] global_step=126700, grad_norm=4.184635639190674, loss=3.747105598449707 -I0513 12:46:33.444672 140460899153664 logging_writer.py:48] [126800] global_step=126800, grad_norm=4.702724933624268, loss=3.835885524749756 -I0513 12:47:25.321907 140460907546368 logging_writer.py:48] [126900] global_step=126900, grad_norm=3.6686577796936035, loss=3.856055736541748 -I0513 12:48:12.908367 140460899153664 logging_writer.py:48] [127000] global_step=127000, grad_norm=3.3572001457214355, loss=3.8894128799438477 -I0513 12:49:02.679642 140460907546368 logging_writer.py:48] [127100] global_step=127100, grad_norm=4.5430731773376465, loss=4.044495582580566 -I0513 12:50:16.197960 140460899153664 logging_writer.py:48] [127200] global_step=127200, grad_norm=12.015634536743164, loss=4.196597099304199 -I0513 12:51:05.650485 140460907546368 logging_writer.py:48] [127300] global_step=127300, grad_norm=3.096425771713257, loss=3.7656073570251465 -I0513 12:51:59.016935 140460899153664 logging_writer.py:48] [127400] global_step=127400, grad_norm=5.554234027862549, loss=3.9830684661865234 -I0513 12:53:02.592504 140460907546368 logging_writer.py:48] [127500] global_step=127500, grad_norm=6.128091812133789, loss=3.8519415855407715 -I0513 12:53:49.073302 140460899153664 logging_writer.py:48] [127600] global_step=127600, grad_norm=5.302657127380371, loss=4.006052494049072 -I0513 12:55:09.984532 140460907546368 logging_writer.py:48] [127700] global_step=127700, grad_norm=6.072127342224121, loss=3.9541890621185303 -I0513 12:55:56.695903 140460899153664 logging_writer.py:48] [127800] global_step=127800, grad_norm=3.6639785766601562, loss=3.9268548488616943 -I0513 12:56:53.800312 140460907546368 logging_writer.py:48] [127900] global_step=127900, grad_norm=24.150754928588867, loss=3.8506312370300293 -I0513 12:57:44.460754 140460899153664 logging_writer.py:48] [128000] global_step=128000, grad_norm=7.920095920562744, loss=3.905643939971924 -I0513 12:58:33.699166 140460907546368 logging_writer.py:48] [128100] global_step=128100, grad_norm=21.680850982666016, loss=3.85709285736084 -I0513 12:59:25.431252 140460899153664 logging_writer.py:48] [128200] global_step=128200, grad_norm=5.716482639312744, loss=3.805119037628174 -I0513 13:00:18.586418 140460907546368 logging_writer.py:48] [128300] global_step=128300, grad_norm=18.817855834960938, loss=3.7607336044311523 -I0513 13:01:27.237644 140460899153664 logging_writer.py:48] [128400] global_step=128400, grad_norm=5.59943151473999, loss=3.813216209411621 -I0513 13:02:30.310479 140460907546368 logging_writer.py:48] [128500] global_step=128500, grad_norm=3.9648501873016357, loss=3.938842296600342 -I0513 13:03:20.731007 140460899153664 logging_writer.py:48] [128600] global_step=128600, grad_norm=2.963606357574463, loss=3.9173953533172607 -I0513 13:04:11.283480 140460907546368 logging_writer.py:48] [128700] global_step=128700, grad_norm=3.4109556674957275, loss=3.772733449935913 -I0513 13:04:59.213904 140460899153664 logging_writer.py:48] [128800] global_step=128800, grad_norm=4.009189605712891, loss=3.823751211166382 -I0513 13:05:51.476642 140460907546368 logging_writer.py:48] [128900] global_step=128900, grad_norm=4.078551769256592, loss=3.8295769691467285 -I0513 13:06:44.398288 140460899153664 logging_writer.py:48] [129000] global_step=129000, grad_norm=4.278016090393066, loss=3.7555441856384277 -I0513 13:07:39.957009 140460907546368 logging_writer.py:48] [129100] global_step=129100, grad_norm=8.153244018554688, loss=3.903920888900757 -I0513 13:08:40.806110 140460899153664 logging_writer.py:48] [129200] global_step=129200, grad_norm=11.554072380065918, loss=3.9686994552612305 -I0513 13:09:31.327780 140460907546368 logging_writer.py:48] [129300] global_step=129300, grad_norm=6.335620403289795, loss=3.7912817001342773 -I0513 13:10:01.004162 140678261474496 spec.py:333] Evaluating on the training split. -I0513 13:10:07.842300 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 13:10:27.063799 140678261474496 spec.py:363] Evaluating on the test split. -I0513 13:10:28.145095 140678261474496 submission_runner.py:516] Time since start: 62620.44s, Step: 129361, {'train/accuracy': Array(0.0017339, dtype=float32), 'train/loss': Array(7.4779243, dtype=float32), 'validation/accuracy': Array(0.00164, dtype=float32), 'validation/loss': Array(7.4817243, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(7.48373, dtype=float32), 'test/num_examples': 10000, 'score': 59946.77712726593, 'total_duration': 62620.444516420364, 'accumulated_submission_time': 59946.77712726593, 'accumulated_eval_time': 2666.0562329292297, 'accumulated_logging_time': 4.544161796569824} -I0513 13:10:28.467172 140460899153664 logging_writer.py:48] [129361] accumulated_eval_time=2666.06, accumulated_logging_time=4.54416, accumulated_submission_time=59946.8, global_step=129361, preemption_count=0, score=59946.8, test/accuracy=0.0012000000569969416, test/loss=7.483729839324951, test/num_examples=10000, total_duration=62620.4, train/accuracy=0.001733896671794355, train/loss=7.477924346923828, validation/accuracy=0.0016399999149143696, validation/loss=7.481724262237549, validation/num_examples=50000 -I0513 13:10:47.460742 140460907546368 logging_writer.py:48] [129400] global_step=129400, grad_norm=9.589828491210938, loss=3.810466766357422 -I0513 13:11:52.556451 140460899153664 logging_writer.py:48] [129500] global_step=129500, grad_norm=5.032624244689941, loss=3.914412498474121 -I0513 13:13:00.815974 140460907546368 logging_writer.py:48] [129600] global_step=129600, grad_norm=3.190286636352539, loss=3.8353724479675293 -I0513 13:13:54.250490 140460899153664 logging_writer.py:48] [129700] global_step=129700, grad_norm=3.3245201110839844, loss=3.8216073513031006 -I0513 13:14:44.541944 140460907546368 logging_writer.py:48] [129800] global_step=129800, grad_norm=6.688719749450684, loss=4.019801616668701 -I0513 13:15:37.326099 140460899153664 logging_writer.py:48] [129900] global_step=129900, grad_norm=4.181392669677734, loss=3.899049758911133 -I0513 13:16:29.252148 140460907546368 logging_writer.py:48] [130000] global_step=130000, grad_norm=5.86929178237915, loss=3.966883420944214 -I0513 13:17:18.551431 140460899153664 logging_writer.py:48] [130100] global_step=130100, grad_norm=4.653183937072754, loss=3.844064235687256 -I0513 13:18:18.385847 140460907546368 logging_writer.py:48] [130200] global_step=130200, grad_norm=5.888641834259033, loss=3.8371291160583496 -I0513 13:19:09.905462 140460899153664 logging_writer.py:48] [130300] global_step=130300, grad_norm=5.130902290344238, loss=3.8001492023468018 -I0513 13:19:58.394870 140460907546368 logging_writer.py:48] [130400] global_step=130400, grad_norm=4.828049182891846, loss=3.98614239692688 -I0513 13:20:52.007070 140460899153664 logging_writer.py:48] [130500] global_step=130500, grad_norm=4.2463884353637695, loss=3.677258253097534 -I0513 13:22:13.597310 140460907546368 logging_writer.py:48] [130600] global_step=130600, grad_norm=4.44193172454834, loss=3.6655962467193604 -I0513 13:23:05.757222 140460899153664 logging_writer.py:48] [130700] global_step=130700, grad_norm=30.239601135253906, loss=4.037126541137695 -I0513 13:23:58.134577 140460907546368 logging_writer.py:48] [130800] global_step=130800, grad_norm=13.230003356933594, loss=3.753317356109619 -I0513 13:24:47.864001 140460899153664 logging_writer.py:48] [130900] global_step=130900, grad_norm=4.964049339294434, loss=3.936901807785034 -I0513 13:25:37.032636 140460907546368 logging_writer.py:48] [131000] global_step=131000, grad_norm=5.590752601623535, loss=3.9772605895996094 -I0513 13:26:30.323443 140460899153664 logging_writer.py:48] [131100] global_step=131100, grad_norm=3.2691256999969482, loss=3.81335711479187 -I0513 13:27:22.819055 140460907546368 logging_writer.py:48] [131200] global_step=131200, grad_norm=13.333988189697266, loss=3.976792812347412 -I0513 13:28:18.862400 140460899153664 logging_writer.py:48] [131300] global_step=131300, grad_norm=5.587927341461182, loss=3.838860273361206 -I0513 13:29:06.305016 140460907546368 logging_writer.py:48] [131400] global_step=131400, grad_norm=4.532068729400635, loss=3.786989212036133 -I0513 13:29:56.206572 140460899153664 logging_writer.py:48] [131500] global_step=131500, grad_norm=5.532174587249756, loss=4.024261474609375 -I0513 13:30:47.186253 140460907546368 logging_writer.py:48] [131600] global_step=131600, grad_norm=6.476302146911621, loss=3.8614258766174316 -I0513 13:31:43.189686 140460899153664 logging_writer.py:48] [131700] global_step=131700, grad_norm=7.863617420196533, loss=3.8917341232299805 -I0513 13:32:40.250785 140460907546368 logging_writer.py:48] [131800] global_step=131800, grad_norm=5.029913425445557, loss=3.8724870681762695 -I0513 13:33:27.198305 140460899153664 logging_writer.py:48] [131900] global_step=131900, grad_norm=5.730442523956299, loss=3.7248053550720215 -I0513 13:34:19.514822 140460907546368 logging_writer.py:48] [132000] global_step=132000, grad_norm=5.498828411102295, loss=3.8421630859375 -I0513 13:35:13.274777 140460899153664 logging_writer.py:48] [132100] global_step=132100, grad_norm=4.62940788269043, loss=3.6746692657470703 -I0513 13:36:09.682206 140460907546368 logging_writer.py:48] [132200] global_step=132200, grad_norm=5.462438106536865, loss=4.160462379455566 -I0513 13:37:03.375875 140460899153664 logging_writer.py:48] [132300] global_step=132300, grad_norm=3.9967546463012695, loss=3.9152626991271973 -I0513 13:37:56.279860 140460907546368 logging_writer.py:48] [132400] global_step=132400, grad_norm=7.0374226570129395, loss=3.969850778579712 -I0513 13:38:55.815896 140460899153664 logging_writer.py:48] [132500] global_step=132500, grad_norm=8.431406021118164, loss=4.07099723815918 -2026-05-13 13:39:56.840167: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 13:40:01.743021 140460907546368 logging_writer.py:48] [132600] global_step=132600, grad_norm=4.596796035766602, loss=3.786829710006714 -I0513 13:41:01.656645 140460899153664 logging_writer.py:48] [132700] global_step=132700, grad_norm=4.3696818351745605, loss=3.9264769554138184 -I0513 13:41:53.884788 140460907546368 logging_writer.py:48] [132800] global_step=132800, grad_norm=4.1141357421875, loss=3.9291388988494873 -I0513 13:42:43.719244 140460899153664 logging_writer.py:48] [132900] global_step=132900, grad_norm=4.11135196685791, loss=3.8383255004882812 -I0513 13:43:40.718996 140460907546368 logging_writer.py:48] [133000] global_step=133000, grad_norm=4.2912211418151855, loss=3.76182222366333 -I0513 13:43:44.562015 140678261474496 spec.py:333] Evaluating on the training split. -I0513 13:43:52.716769 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 13:44:38.977861 140678261474496 spec.py:363] Evaluating on the test split. -I0513 13:44:40.020931 140678261474496 submission_runner.py:516] Time since start: 64672.36s, Step: 133006, {'train/accuracy': Array(0.00131537, dtype=float32), 'train/loss': Array(7.440227, dtype=float32), 'validation/accuracy': Array(0.00158, dtype=float32), 'validation/loss': Array(7.4432135, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(7.444616, dtype=float32), 'test/num_examples': 10000, 'score': 61942.77612376213, 'total_duration': 64672.361290216446, 'accumulated_submission_time': 61942.77612376213, 'accumulated_eval_time': 2721.3663465976715, 'accumulated_logging_time': 4.906957149505615} -I0513 13:44:40.316851 140460899153664 logging_writer.py:48] [133006] accumulated_eval_time=2721.37, accumulated_logging_time=4.90696, accumulated_submission_time=61942.8, global_step=133006, preemption_count=0, score=61942.8, test/accuracy=0.0013000001199543476, test/loss=7.444615840911865, test/num_examples=10000, total_duration=64672.4, train/accuracy=0.0013153698528185487, train/loss=7.440227031707764, validation/accuracy=0.0015799999237060547, validation/loss=7.44321346282959, validation/num_examples=50000 -I0513 13:45:25.864124 140460907546368 logging_writer.py:48] [133100] global_step=133100, grad_norm=5.724856376647949, loss=3.9611756801605225 -I0513 13:46:16.973987 140460899153664 logging_writer.py:48] [133200] global_step=133200, grad_norm=3.593883514404297, loss=3.933133125305176 -I0513 13:47:06.384172 140460907546368 logging_writer.py:48] [133300] global_step=133300, grad_norm=8.028440475463867, loss=3.811744213104248 -I0513 13:47:53.836171 140460899153664 logging_writer.py:48] [133400] global_step=133400, grad_norm=7.558205604553223, loss=3.78446102142334 -I0513 13:48:46.831035 140460907546368 logging_writer.py:48] [133500] global_step=133500, grad_norm=7.113080024719238, loss=3.9171550273895264 -I0513 13:49:35.724987 140460899153664 logging_writer.py:48] [133600] global_step=133600, grad_norm=6.240009784698486, loss=3.9000444412231445 -I0513 13:50:26.735132 140460907546368 logging_writer.py:48] [133700] global_step=133700, grad_norm=5.359076499938965, loss=3.9538559913635254 -I0513 13:51:21.527603 140460899153664 logging_writer.py:48] [133800] global_step=133800, grad_norm=5.565454959869385, loss=3.9080564975738525 -I0513 13:52:36.559023 140460907546368 logging_writer.py:48] [133900] global_step=133900, grad_norm=4.225979804992676, loss=3.772817611694336 -I0513 13:53:41.085165 140460899153664 logging_writer.py:48] [134000] global_step=134000, grad_norm=5.352962017059326, loss=3.8572020530700684 -I0513 13:54:45.913871 140460907546368 logging_writer.py:48] [134100] global_step=134100, grad_norm=6.583169937133789, loss=3.9290215969085693 -I0513 13:55:36.679400 140460899153664 logging_writer.py:48] [134200] global_step=134200, grad_norm=4.3751091957092285, loss=3.6266603469848633 -I0513 13:56:25.603698 140460907546368 logging_writer.py:48] [134300] global_step=134300, grad_norm=23.193920135498047, loss=3.8027267456054688 -I0513 13:57:15.203021 140460899153664 logging_writer.py:48] [134400] global_step=134400, grad_norm=15.171771049499512, loss=3.9835610389709473 -I0513 13:58:03.800531 140460907546368 logging_writer.py:48] [134500] global_step=134500, grad_norm=4.172743320465088, loss=3.7151870727539062 -I0513 13:58:54.716768 140460899153664 logging_writer.py:48] [134600] global_step=134600, grad_norm=3.718759775161743, loss=3.9050376415252686 -I0513 13:59:46.192952 140460907546368 logging_writer.py:48] [134700] global_step=134700, grad_norm=5.292100429534912, loss=3.8870770931243896 -I0513 14:00:37.504552 140460899153664 logging_writer.py:48] [134800] global_step=134800, grad_norm=3.2830119132995605, loss=3.8337202072143555 -I0513 14:01:27.144583 140460907546368 logging_writer.py:48] [134900] global_step=134900, grad_norm=3.1182444095611572, loss=3.7640233039855957 -I0513 14:02:20.781397 140460899153664 logging_writer.py:48] [135000] global_step=135000, grad_norm=7.246944904327393, loss=3.9275615215301514 -2026-05-13 14:03:11.522843: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 14:03:12.924837 140460907546368 logging_writer.py:48] [135100] global_step=135100, grad_norm=8.295702934265137, loss=3.8766982555389404 -I0513 14:04:13.708487 140460899153664 logging_writer.py:48] [135200] global_step=135200, grad_norm=4.3458991050720215, loss=3.8061330318450928 -I0513 14:05:30.858172 140460907546368 logging_writer.py:48] [135300] global_step=135300, grad_norm=3.068587064743042, loss=3.7786619663238525 -I0513 14:06:23.344620 140460899153664 logging_writer.py:48] [135400] global_step=135400, grad_norm=5.717787265777588, loss=3.8535759449005127 -I0513 14:07:19.035224 140460907546368 logging_writer.py:48] [135500] global_step=135500, grad_norm=10.99756145477295, loss=3.968228340148926 -I0513 14:08:18.808224 140460899153664 logging_writer.py:48] [135600] global_step=135600, grad_norm=4.975460052490234, loss=3.8391127586364746 -I0513 14:09:07.387612 140460907546368 logging_writer.py:48] [135700] global_step=135700, grad_norm=10.63094711303711, loss=3.9109597206115723 -I0513 14:10:03.551620 140460899153664 logging_writer.py:48] [135800] global_step=135800, grad_norm=6.474438667297363, loss=4.018261909484863 -I0513 14:11:04.160129 140460907546368 logging_writer.py:48] [135900] global_step=135900, grad_norm=5.20443058013916, loss=3.9670088291168213 -I0513 14:12:06.473502 140460899153664 logging_writer.py:48] [136000] global_step=136000, grad_norm=6.563535213470459, loss=3.8858492374420166 -I0513 14:12:53.054327 140460907546368 logging_writer.py:48] [136100] global_step=136100, grad_norm=3.000497579574585, loss=3.825113296508789 -I0513 14:13:50.902908 140460899153664 logging_writer.py:48] [136200] global_step=136200, grad_norm=7.613068580627441, loss=4.083691120147705 -I0513 14:14:44.029948 140460907546368 logging_writer.py:48] [136300] global_step=136300, grad_norm=5.56748104095459, loss=3.8125228881835938 -I0513 14:15:34.358236 140460899153664 logging_writer.py:48] [136400] global_step=136400, grad_norm=5.9646477699279785, loss=3.9043426513671875 -I0513 14:16:26.795097 140460907546368 logging_writer.py:48] [136500] global_step=136500, grad_norm=4.199394226074219, loss=3.9345388412475586 -I0513 14:17:14.039082 140460899153664 logging_writer.py:48] [136600] global_step=136600, grad_norm=6.614282608032227, loss=3.9635283946990967 -I0513 14:17:56.524148 140678261474496 spec.py:333] Evaluating on the training split. -I0513 14:18:04.912963 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 14:18:32.487894 140678261474496 spec.py:363] Evaluating on the test split. -I0513 14:18:33.528884 140678261474496 submission_runner.py:516] Time since start: 66705.87s, Step: 136687, {'train/accuracy': Array(0.00179369, dtype=float32), 'train/loss': Array(7.3960743, dtype=float32), 'validation/accuracy': Array(0.00172, dtype=float32), 'validation/loss': Array(7.398652, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(7.4008193, dtype=float32), 'test/num_examples': 10000, 'score': 63938.894350767136, 'total_duration': 66705.86566662788, 'accumulated_submission_time': 63938.894350767136, 'accumulated_eval_time': 2758.2186217308044, 'accumulated_logging_time': 5.240548133850098} -I0513 14:18:33.793822 140460907546368 logging_writer.py:48] [136687] accumulated_eval_time=2758.22, accumulated_logging_time=5.24055, accumulated_submission_time=63938.9, global_step=136687, preemption_count=0, score=63938.9, test/accuracy=0.0013000001199543476, test/loss=7.400819301605225, test/num_examples=10000, total_duration=66705.9, train/accuracy=0.0017936861841008067, train/loss=7.396074295043945, validation/accuracy=0.00171999994199723, validation/loss=7.398652076721191, validation/num_examples=50000 -I0513 14:18:39.127351 140460899153664 logging_writer.py:48] [136700] global_step=136700, grad_norm=5.015724182128906, loss=3.7936277389526367 -I0513 14:19:33.696259 140460907546368 logging_writer.py:48] [136800] global_step=136800, grad_norm=4.307075500488281, loss=3.766801357269287 -I0513 14:20:33.452693 140460899153664 logging_writer.py:48] [136900] global_step=136900, grad_norm=4.005608081817627, loss=3.9586291313171387 -I0513 14:21:28.412021 140460907546368 logging_writer.py:48] [137000] global_step=137000, grad_norm=5.973000526428223, loss=3.7881808280944824 -I0513 14:22:17.655438 140460899153664 logging_writer.py:48] [137100] global_step=137100, grad_norm=7.3922529220581055, loss=3.9164485931396484 -I0513 14:23:04.283707 140460907546368 logging_writer.py:48] [137200] global_step=137200, grad_norm=4.160150051116943, loss=3.8864316940307617 -I0513 14:23:56.734131 140460899153664 logging_writer.py:48] [137300] global_step=137300, grad_norm=6.211120128631592, loss=3.7433977127075195 -I0513 14:24:49.251807 140460907546368 logging_writer.py:48] [137400] global_step=137400, grad_norm=4.743871212005615, loss=3.892524003982544 -I0513 14:25:49.768748 140460899153664 logging_writer.py:48] [137500] global_step=137500, grad_norm=4.448116302490234, loss=3.8343842029571533 -I0513 14:26:50.269348 140460907546368 logging_writer.py:48] [137600] global_step=137600, grad_norm=5.877606391906738, loss=3.7699270248413086 -I0513 14:28:14.776965 140460899153664 logging_writer.py:48] [137700] global_step=137700, grad_norm=3.961019515991211, loss=3.8233494758605957 -I0513 14:29:11.061230 140460907546368 logging_writer.py:48] [137800] global_step=137800, grad_norm=5.731451511383057, loss=3.8261616230010986 -I0513 14:30:24.502449 140460899153664 logging_writer.py:48] [137900] global_step=137900, grad_norm=12.683074951171875, loss=4.041749000549316 -I0513 14:31:32.773831 140460907546368 logging_writer.py:48] [138000] global_step=138000, grad_norm=6.477255821228027, loss=3.857654094696045 -I0513 14:33:17.828966 140460899153664 logging_writer.py:48] [138100] global_step=138100, grad_norm=5.837165355682373, loss=3.9317352771759033 -I0513 14:34:22.452222 140460907546368 logging_writer.py:48] [138200] global_step=138200, grad_norm=3.3714637756347656, loss=3.7765679359436035 -I0513 14:35:11.293446 140460899153664 logging_writer.py:48] [138300] global_step=138300, grad_norm=6.5850324630737305, loss=3.822859048843384 -I0513 14:36:05.385766 140460907546368 logging_writer.py:48] [138400] global_step=138400, grad_norm=4.730902671813965, loss=3.986905574798584 -I0513 14:37:23.657261 140460899153664 logging_writer.py:48] [138500] global_step=138500, grad_norm=3.6973273754119873, loss=3.8859450817108154 -I0513 14:38:17.226200 140460907546368 logging_writer.py:48] [138600] global_step=138600, grad_norm=5.232942581176758, loss=4.029339790344238 -I0513 14:39:05.733041 140460899153664 logging_writer.py:48] [138700] global_step=138700, grad_norm=7.165712356567383, loss=4.070328235626221 -I0513 14:40:06.675301 140460907546368 logging_writer.py:48] [138800] global_step=138800, grad_norm=3.2491986751556396, loss=3.887312173843384 -I0513 14:41:02.435203 140460899153664 logging_writer.py:48] [138900] global_step=138900, grad_norm=3.231795310974121, loss=3.866117000579834 -I0513 14:41:56.719191 140460907546368 logging_writer.py:48] [139000] global_step=139000, grad_norm=3.9476635456085205, loss=3.9324605464935303 -I0513 14:42:46.185626 140460899153664 logging_writer.py:48] [139100] global_step=139100, grad_norm=10.039617538452148, loss=4.108737468719482 -I0513 14:43:41.484436 140460907546368 logging_writer.py:48] [139200] global_step=139200, grad_norm=7.156628131866455, loss=4.02711296081543 -I0513 14:44:32.201253 140460899153664 logging_writer.py:48] [139300] global_step=139300, grad_norm=4.728353500366211, loss=3.917379856109619 -I0513 14:45:22.259932 140460907546368 logging_writer.py:48] [139400] global_step=139400, grad_norm=5.281749725341797, loss=4.006739139556885 -I0513 14:46:10.755958 140460899153664 logging_writer.py:48] [139500] global_step=139500, grad_norm=5.883159160614014, loss=3.915339469909668 -I0513 14:47:10.238704 140460907546368 logging_writer.py:48] [139600] global_step=139600, grad_norm=4.7542548179626465, loss=3.9651358127593994 -I0513 14:47:58.456748 140460899153664 logging_writer.py:48] [139700] global_step=139700, grad_norm=5.592064380645752, loss=3.870737075805664 -I0513 14:48:46.866142 140460907546368 logging_writer.py:48] [139800] global_step=139800, grad_norm=3.545985460281372, loss=3.785560369491577 -I0513 14:49:35.461712 140460899153664 logging_writer.py:48] [139900] global_step=139900, grad_norm=9.908478736877441, loss=4.066740989685059 -I0513 14:50:28.295313 140460907546368 logging_writer.py:48] [140000] global_step=140000, grad_norm=5.968626499176025, loss=3.9765992164611816 -I0513 14:51:18.722570 140460899153664 logging_writer.py:48] [140100] global_step=140100, grad_norm=5.0973968505859375, loss=3.7162094116210938 -I0513 14:51:49.422859 140678261474496 spec.py:333] Evaluating on the training split. -I0513 14:51:57.635128 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 14:52:16.835496 140678261474496 spec.py:363] Evaluating on the test split. -I0513 14:52:17.872926 140678261474496 submission_runner.py:516] Time since start: 68730.22s, Step: 140156, {'train/accuracy': Array(0.00191327, dtype=float32), 'train/loss': Array(7.353606, dtype=float32), 'validation/accuracy': Array(0.00188, dtype=float32), 'validation/loss': Array(7.3550563, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(7.357977, dtype=float32), 'test/num_examples': 10000, 'score': 65934.44663023949, 'total_duration': 68730.21747016907, 'accumulated_submission_time': 65934.44663023949, 'accumulated_eval_time': 2786.5239927768707, 'accumulated_logging_time': 5.537599086761475} -I0513 14:52:18.145564 140460907546368 logging_writer.py:48] [140156] accumulated_eval_time=2786.52, accumulated_logging_time=5.5376, accumulated_submission_time=65934.4, global_step=140156, preemption_count=0, score=65934.4, test/accuracy=0.0012000000569969416, test/loss=7.357976913452148, test/num_examples=10000, total_duration=68730.2, train/accuracy=0.0019132653251290321, train/loss=7.353606224060059, validation/accuracy=0.001879999996162951, validation/loss=7.355056285858154, validation/num_examples=50000 -I0513 14:52:41.620174 140460899153664 logging_writer.py:48] [140200] global_step=140200, grad_norm=3.5806798934936523, loss=3.8337037563323975 -I0513 14:53:37.956047 140460907546368 logging_writer.py:48] [140300] global_step=140300, grad_norm=3.462242841720581, loss=3.7809853553771973 -I0513 14:54:38.731550 140460899153664 logging_writer.py:48] [140400] global_step=140400, grad_norm=3.122346878051758, loss=3.9179329872131348 -I0513 14:56:19.866497 140460907546368 logging_writer.py:48] [140500] global_step=140500, grad_norm=12.766247749328613, loss=3.9580788612365723 -I0513 14:57:11.812216 140460899153664 logging_writer.py:48] [140600] global_step=140600, grad_norm=7.822099208831787, loss=3.8285350799560547 -I0513 14:58:13.893489 140460907546368 logging_writer.py:48] [140700] global_step=140700, grad_norm=4.239649295806885, loss=3.8641347885131836 -I0513 14:59:05.689312 140460899153664 logging_writer.py:48] [140800] global_step=140800, grad_norm=8.336674690246582, loss=3.8946950435638428 -I0513 14:59:55.273536 140460907546368 logging_writer.py:48] [140900] global_step=140900, grad_norm=8.079732894897461, loss=4.078408241271973 -I0513 15:00:42.642755 140460899153664 logging_writer.py:48] [141000] global_step=141000, grad_norm=3.7554211616516113, loss=3.906400680541992 -I0513 15:01:37.869740 140460907546368 logging_writer.py:48] [141100] global_step=141100, grad_norm=3.966156005859375, loss=3.960049867630005 -I0513 15:02:45.302729 140460899153664 logging_writer.py:48] [141200] global_step=141200, grad_norm=6.018372535705566, loss=3.7457408905029297 -I0513 15:03:35.335522 140460907546368 logging_writer.py:48] [141300] global_step=141300, grad_norm=6.027738571166992, loss=3.8453235626220703 -I0513 15:04:23.092614 140460899153664 logging_writer.py:48] [141400] global_step=141400, grad_norm=5.153494358062744, loss=3.9011898040771484 -I0513 15:05:17.332823 140460907546368 logging_writer.py:48] [141500] global_step=141500, grad_norm=8.409449577331543, loss=3.87636137008667 -I0513 15:06:11.397860 140460899153664 logging_writer.py:48] [141600] global_step=141600, grad_norm=5.5606536865234375, loss=3.8208932876586914 -I0513 15:07:15.039377 140460907546368 logging_writer.py:48] [141700] global_step=141700, grad_norm=20.658796310424805, loss=3.966008186340332 -I0513 15:08:42.712545 140460899153664 logging_writer.py:48] [141800] global_step=141800, grad_norm=3.1075141429901123, loss=3.783130645751953 -I0513 15:10:39.467856 140460907546368 logging_writer.py:48] [141900] global_step=141900, grad_norm=4.169227600097656, loss=3.8672828674316406 -I0513 15:11:48.809115 140460899153664 logging_writer.py:48] [142000] global_step=142000, grad_norm=5.226704120635986, loss=3.8456013202667236 -I0513 15:13:23.527030 140460907546368 logging_writer.py:48] [142100] global_step=142100, grad_norm=5.707348346710205, loss=3.952298879623413 -I0513 15:14:26.533027 140460899153664 logging_writer.py:48] [142200] global_step=142200, grad_norm=14.40700912475586, loss=3.833812952041626 -I0513 15:15:24.594976 140460907546368 logging_writer.py:48] [142300] global_step=142300, grad_norm=10.701654434204102, loss=3.9399514198303223 -I0513 15:16:17.629727 140460899153664 logging_writer.py:48] [142400] global_step=142400, grad_norm=4.138687610626221, loss=3.77583384513855 -I0513 15:17:27.142755 140460907546368 logging_writer.py:48] [142500] global_step=142500, grad_norm=6.811947345733643, loss=3.86965012550354 -I0513 15:18:33.128827 140460899153664 logging_writer.py:48] [142600] global_step=142600, grad_norm=7.538802623748779, loss=3.9722683429718018 -2026-05-13 15:18:42.615370: W tensorflow/core/kernels/data/prefetch_autotuner.cc:52] Prefetch autotuner tried to allocate 91453 bytes after encountering the first element of size 91453 bytes.This already causes the autotune ram budget to be exceeded. To stay within the ram budget, either increase the ram budget or reduce element size -I0513 15:20:07.425995 140460907546368 logging_writer.py:48] [142700] global_step=142700, grad_norm=5.080458164215088, loss=3.8709182739257812 -I0513 15:21:14.666670 140460899153664 logging_writer.py:48] [142800] global_step=142800, grad_norm=3.8276538848876953, loss=3.9125914573669434 -I0513 15:22:54.226166 140460907546368 logging_writer.py:48] [142900] global_step=142900, grad_norm=6.304086208343506, loss=3.9886627197265625 -I0513 15:23:46.330050 140460899153664 logging_writer.py:48] [143000] global_step=143000, grad_norm=3.6600334644317627, loss=3.7939343452453613 -I0513 15:25:34.806180 140678261474496 spec.py:333] Evaluating on the training split. -I0513 15:25:43.511456 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 15:26:25.327491 140678261474496 spec.py:363] Evaluating on the test split. -I0513 15:26:26.353573 140678261474496 submission_runner.py:516] Time since start: 70778.71s, Step: 143098, {'train/accuracy': Array(0.00163425, dtype=float32), 'train/loss': Array(7.3302083, dtype=float32), 'validation/accuracy': Array(0.00194, dtype=float32), 'validation/loss': Array(7.3216434, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(7.324862, dtype=float32), 'test/num_examples': 10000, 'score': 67931.0115404129, 'total_duration': 70778.71036839485, 'accumulated_submission_time': 67931.0115404129, 'accumulated_eval_time': 2837.938948869705, 'accumulated_logging_time': 5.867739677429199} -I0513 15:26:26.610344 140460907546368 logging_writer.py:48] [143098] accumulated_eval_time=2837.94, accumulated_logging_time=5.86774, accumulated_submission_time=67931, global_step=143098, preemption_count=0, score=67931, test/accuracy=0.0013000001199543476, test/loss=7.324862003326416, test/num_examples=10000, total_duration=70778.7, train/accuracy=0.0016342473682016134, train/loss=7.3302083015441895, validation/accuracy=0.0019399999873712659, validation/loss=7.321643352508545, validation/num_examples=50000 -I0513 15:26:27.857958 140460899153664 logging_writer.py:48] [143100] global_step=143100, grad_norm=3.7670576572418213, loss=3.817470073699951 -I0513 15:28:14.992105 140460907546368 logging_writer.py:48] [143200] global_step=143200, grad_norm=6.069732666015625, loss=3.86025071144104 -I0513 15:29:14.866043 140460899153664 logging_writer.py:48] [143300] global_step=143300, grad_norm=3.4529967308044434, loss=3.862619161605835 -I0513 15:30:05.955566 140460907546368 logging_writer.py:48] [143400] global_step=143400, grad_norm=4.940078258514404, loss=3.799927234649658 -I0513 15:30:56.619019 140460899153664 logging_writer.py:48] [143500] global_step=143500, grad_norm=6.976273536682129, loss=3.970039129257202 -I0513 15:31:46.014250 140460907546368 logging_writer.py:48] [143600] global_step=143600, grad_norm=5.565046787261963, loss=4.085651874542236 -I0513 15:32:42.331969 140460899153664 logging_writer.py:48] [143700] global_step=143700, grad_norm=5.359874725341797, loss=3.7801012992858887 -I0513 15:33:46.837204 140460907546368 logging_writer.py:48] [143800] global_step=143800, grad_norm=4.324221611022949, loss=3.862678289413452 -I0513 15:34:49.028888 140460899153664 logging_writer.py:48] [143900] global_step=143900, grad_norm=4.1137518882751465, loss=3.958458662033081 -I0513 15:35:48.728083 140460907546368 logging_writer.py:48] [144000] global_step=144000, grad_norm=3.791752338409424, loss=3.911513090133667 -I0513 15:36:49.526379 140460899153664 logging_writer.py:48] [144100] global_step=144100, grad_norm=7.6322479248046875, loss=3.9354796409606934 -I0513 15:38:30.204693 140460907546368 logging_writer.py:48] [144200] global_step=144200, grad_norm=8.856834411621094, loss=3.966865062713623 -I0513 15:39:40.917842 140460899153664 logging_writer.py:48] [144300] global_step=144300, grad_norm=31.689516067504883, loss=4.198694705963135 -I0513 15:40:40.185950 140460907546368 logging_writer.py:48] [144400] global_step=144400, grad_norm=4.297616004943848, loss=3.8428735733032227 -I0513 15:41:51.383329 140460899153664 logging_writer.py:48] [144500] global_step=144500, grad_norm=5.802192687988281, loss=3.9255309104919434 -I0513 15:42:59.264495 140460907546368 logging_writer.py:48] [144600] global_step=144600, grad_norm=5.394393444061279, loss=3.9808948040008545 -I0513 15:44:22.607523 140460899153664 logging_writer.py:48] [144700] global_step=144700, grad_norm=5.354955673217773, loss=3.805161952972412 -I0513 15:45:35.509138 140460907546368 logging_writer.py:48] [144800] global_step=144800, grad_norm=5.035033226013184, loss=3.8230836391448975 -I0513 15:46:31.951520 140460899153664 logging_writer.py:48] [144900] global_step=144900, grad_norm=5.4955644607543945, loss=3.9391376972198486 -I0513 15:47:25.789005 140460907546368 logging_writer.py:48] [145000] global_step=145000, grad_norm=5.542181491851807, loss=3.951209545135498 -I0513 15:49:05.633712 140460899153664 logging_writer.py:48] [145100] global_step=145100, grad_norm=6.468028545379639, loss=3.8677992820739746 -I0513 15:50:26.025413 140460907546368 logging_writer.py:48] [145200] global_step=145200, grad_norm=5.7126054763793945, loss=3.996070146560669 -I0513 15:51:21.358907 140460899153664 logging_writer.py:48] [145300] global_step=145300, grad_norm=3.5684516429901123, loss=3.727790594100952 -I0513 15:52:20.536117 140460907546368 logging_writer.py:48] [145400] global_step=145400, grad_norm=5.808554649353027, loss=3.942150115966797 -I0513 15:53:31.509848 140460899153664 logging_writer.py:48] [145500] global_step=145500, grad_norm=5.230720043182373, loss=3.915574550628662 -I0513 15:54:45.604959 140460907546368 logging_writer.py:48] [145600] global_step=145600, grad_norm=19.34010887145996, loss=3.858691692352295 -I0513 15:58:15.203621 140460899153664 logging_writer.py:48] [145700] global_step=145700, grad_norm=4.4447126388549805, loss=3.852400541305542 -I0513 15:59:44.283432 140678261474496 spec.py:333] Evaluating on the training split. -I0513 15:59:51.693073 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 16:00:32.408323 140678261474496 spec.py:363] Evaluating on the test split. -I0513 16:00:33.474943 140678261474496 submission_runner.py:516] Time since start: 72825.79s, Step: 145779, {'train/accuracy': Array(0.00203284, dtype=float32), 'train/loss': Array(7.2840333, dtype=float32), 'validation/accuracy': Array(0.00206, dtype=float32), 'validation/loss': Array(7.286662, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(7.291023, dtype=float32), 'test/num_examples': 10000, 'score': 69928.60928463936, 'total_duration': 72825.79156255722, 'accumulated_submission_time': 69928.60928463936, 'accumulated_eval_time': 2886.9578301906586, 'accumulated_logging_time': 6.165061712265015} -I0513 16:00:33.782856 140460907546368 logging_writer.py:48] [145779] accumulated_eval_time=2886.96, accumulated_logging_time=6.16506, accumulated_submission_time=69928.6, global_step=145779, preemption_count=0, score=69928.6, test/accuracy=0.0012000000569969416, test/loss=7.291022777557373, test/num_examples=10000, total_duration=72825.8, train/accuracy=0.0020328443497419357, train/loss=7.284033298492432, validation/accuracy=0.002059999853372574, validation/loss=7.2866621017456055, validation/num_examples=50000 -I0513 16:00:52.791287 140460899153664 logging_writer.py:48] [145800] global_step=145800, grad_norm=9.128424644470215, loss=4.04818058013916 -I0513 16:01:51.186697 140460907546368 logging_writer.py:48] [145900] global_step=145900, grad_norm=4.217911720275879, loss=3.747523546218872 -I0513 16:02:56.882737 140460899153664 logging_writer.py:48] [146000] global_step=146000, grad_norm=11.381990432739258, loss=3.8543143272399902 -I0513 16:03:51.472880 140460907546368 logging_writer.py:48] [146100] global_step=146100, grad_norm=4.753579139709473, loss=3.898552894592285 -I0513 16:04:44.805780 140460899153664 logging_writer.py:48] [146200] global_step=146200, grad_norm=5.211374282836914, loss=3.876208782196045 -I0513 16:06:31.085731 140460907546368 logging_writer.py:48] [146300] global_step=146300, grad_norm=5.562703609466553, loss=3.767148971557617 -I0513 16:08:18.192880 140460899153664 logging_writer.py:48] [146400] global_step=146400, grad_norm=11.663446426391602, loss=3.811128854751587 -I0513 16:09:07.633651 140460907546368 logging_writer.py:48] [146500] global_step=146500, grad_norm=4.298900127410889, loss=3.8060553073883057 -I0513 16:10:13.307659 140460899153664 logging_writer.py:48] [146600] global_step=146600, grad_norm=3.968970775604248, loss=3.840163230895996 -I0513 16:11:17.341451 140460907546368 logging_writer.py:48] [146700] global_step=146700, grad_norm=3.931307077407837, loss=3.8730063438415527 -I0513 16:12:07.517066 140460899153664 logging_writer.py:48] [146800] global_step=146800, grad_norm=15.233014106750488, loss=3.858748435974121 -I0513 16:12:58.407978 140460907546368 logging_writer.py:48] [146900] global_step=146900, grad_norm=13.58024787902832, loss=4.558838844299316 -I0513 16:14:42.884094 140460899153664 logging_writer.py:48] [147000] global_step=147000, grad_norm=3.597449779510498, loss=3.976567268371582 -I0513 16:16:59.468597 140460907546368 logging_writer.py:48] [147100] global_step=147100, grad_norm=4.0270843505859375, loss=3.824723720550537 -I0513 16:18:25.597321 140460899153664 logging_writer.py:48] [147200] global_step=147200, grad_norm=4.010537624359131, loss=3.884422540664673 -I0513 16:19:30.925181 140460907546368 logging_writer.py:48] [147300] global_step=147300, grad_norm=3.827925682067871, loss=3.901451587677002 -I0513 16:21:00.240532 140460899153664 logging_writer.py:48] [147400] global_step=147400, grad_norm=4.373264312744141, loss=3.8109378814697266 -I0513 16:23:03.642296 140460907546368 logging_writer.py:48] [147500] global_step=147500, grad_norm=4.5306525230407715, loss=3.800367832183838 -I0513 16:25:19.536203 140460899153664 logging_writer.py:48] [147600] global_step=147600, grad_norm=2.729860782623291, loss=3.8314809799194336 -I0513 16:26:34.838194 140460907546368 logging_writer.py:48] [147700] global_step=147700, grad_norm=3.668325662612915, loss=3.8132033348083496 -I0513 16:27:44.516472 140460899153664 logging_writer.py:48] [147800] global_step=147800, grad_norm=2.4022762775421143, loss=3.7591733932495117 -I0513 16:28:53.360982 140460907546368 logging_writer.py:48] [147900] global_step=147900, grad_norm=3.6037325859069824, loss=3.90327525138855 -I0513 16:30:00.776848 140460899153664 logging_writer.py:48] [148000] global_step=148000, grad_norm=33.73482131958008, loss=3.8885977268218994 -I0513 16:31:26.843081 140460907546368 logging_writer.py:48] [148100] global_step=148100, grad_norm=4.761085510253906, loss=4.115481853485107 -I0513 16:32:53.107056 140460899153664 logging_writer.py:48] [148200] global_step=148200, grad_norm=13.045433044433594, loss=4.138762474060059 -I0513 16:33:49.414436 140678261474496 spec.py:333] Evaluating on the training split. -I0513 16:33:58.429932 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 16:34:41.229784 140678261474496 spec.py:363] Evaluating on the test split. -I0513 16:34:42.287041 140678261474496 submission_runner.py:516] Time since start: 74874.61s, Step: 148284, {'train/accuracy': Array(0.00159439, dtype=float32), 'train/loss': Array(7.25432, dtype=float32), 'validation/accuracy': Array(0.0021, dtype=float32), 'validation/loss': Array(7.256729, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(7.262055, dtype=float32), 'test/num_examples': 10000, 'score': 71924.17374968529, 'total_duration': 74874.6106865406, 'accumulated_submission_time': 71924.17374968529, 'accumulated_eval_time': 2939.664833545685, 'accumulated_logging_time': 6.508132457733154} -I0513 16:34:42.565684 140460907546368 logging_writer.py:48] [148284] accumulated_eval_time=2939.66, accumulated_logging_time=6.50813, accumulated_submission_time=71924.2, global_step=148284, preemption_count=0, score=71924.2, test/accuracy=0.0010999999940395355, test/loss=7.262054920196533, test/num_examples=10000, total_duration=74874.6, train/accuracy=0.0015943876933306456, train/loss=7.25432014465332, validation/accuracy=0.002099999925121665, validation/loss=7.2567291259765625, validation/num_examples=50000 -I0513 16:34:49.564604 140460899153664 logging_writer.py:48] [148300] global_step=148300, grad_norm=4.903435230255127, loss=3.8956403732299805 -I0513 16:35:53.528373 140460907546368 logging_writer.py:48] [148400] global_step=148400, grad_norm=4.018662452697754, loss=4.054006099700928 -I0513 16:36:51.871749 140460899153664 logging_writer.py:48] [148500] global_step=148500, grad_norm=6.06927490234375, loss=3.89327073097229 -I0513 16:38:16.579580 140460907546368 logging_writer.py:48] [148600] global_step=148600, grad_norm=4.2077250480651855, loss=3.897268772125244 -I0513 16:39:56.872621 140460899153664 logging_writer.py:48] [148700] global_step=148700, grad_norm=4.585819721221924, loss=3.838303565979004 -I0513 16:41:12.009523 140460907546368 logging_writer.py:48] [148800] global_step=148800, grad_norm=15.406478881835938, loss=3.756746768951416 -I0513 16:42:19.132166 140460899153664 logging_writer.py:48] [148900] global_step=148900, grad_norm=4.335975646972656, loss=3.9710183143615723 -I0513 16:43:24.357453 140460907546368 logging_writer.py:48] [149000] global_step=149000, grad_norm=4.640788555145264, loss=3.6784255504608154 -I0513 16:45:07.698291 140460899153664 logging_writer.py:48] [149100] global_step=149100, grad_norm=3.5620744228363037, loss=3.867169141769409 -I0513 16:46:20.635216 140460907546368 logging_writer.py:48] [149200] global_step=149200, grad_norm=4.374199390411377, loss=3.7970070838928223 -I0513 16:47:34.123652 140460899153664 logging_writer.py:48] [149300] global_step=149300, grad_norm=7.410396099090576, loss=3.979830503463745 -I0513 16:51:17.649476 140460907546368 logging_writer.py:48] [149400] global_step=149400, grad_norm=3.520967483520508, loss=3.834955930709839 -I0513 16:53:47.258860 140460899153664 logging_writer.py:48] [149500] global_step=149500, grad_norm=9.484121322631836, loss=3.965811252593994 -I0513 16:56:45.123395 140460907546368 logging_writer.py:48] [149600] global_step=149600, grad_norm=5.860807418823242, loss=3.890944004058838 -I0513 16:58:09.390960 140460899153664 logging_writer.py:48] [149700] global_step=149700, grad_norm=4.050935745239258, loss=3.8236002922058105 -I0513 16:59:29.982854 140460907546368 logging_writer.py:48] [149800] global_step=149800, grad_norm=5.98357629776001, loss=4.024722099304199 -I0513 17:00:18.357015 140460899153664 logging_writer.py:48] [149900] global_step=149900, grad_norm=10.583053588867188, loss=4.0261430740356445 -I0513 17:01:37.509234 140460907546368 logging_writer.py:48] [150000] global_step=150000, grad_norm=4.3185601234436035, loss=3.865976333618164 -I0513 17:03:12.771606 140460899153664 logging_writer.py:48] [150100] global_step=150100, grad_norm=3.418933153152466, loss=3.848090648651123 -I0513 17:05:22.276767 140460907546368 logging_writer.py:48] [150200] global_step=150200, grad_norm=3.9575693607330322, loss=3.769902229309082 -I0513 17:08:00.399825 140678261474496 spec.py:333] Evaluating on the training split. -I0513 17:08:07.982385 140678261474496 spec.py:346] Evaluating on the validation split. -I0513 17:09:25.723210 140678261474496 spec.py:363] Evaluating on the test split. -I0513 17:09:26.760461 140678261474496 submission_runner.py:516] Time since start: 76959.11s, Step: 150291, {'train/accuracy': Array(0.00199298, dtype=float32), 'train/loss': Array(7.249264, dtype=float32), 'validation/accuracy': Array(0.00208, dtype=float32), 'validation/loss': Array(7.241174, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0014, dtype=float32), 'test/loss': Array(7.246727, dtype=float32), 'test/num_examples': 10000, 'score': 73921.95043730736, 'total_duration': 76959.10509991646, 'accumulated_submission_time': 73921.95043730736, 'accumulated_eval_time': 3025.88085770607, 'accumulated_logging_time': 6.818323612213135} -I0513 17:09:27.090000 140460899153664 logging_writer.py:48] [150291] accumulated_eval_time=3025.88, accumulated_logging_time=6.81832, accumulated_submission_time=73922, global_step=150291, preemption_count=0, score=73922, test/accuracy=0.0014000000664964318, test/loss=7.246726989746094, test/num_examples=10000, total_duration=76959.1, train/accuracy=0.001992984674870968, train/loss=7.249263763427734, validation/accuracy=0.0020800000056624413, validation/loss=7.241174221038818, validation/num_examples=50000 -I0513 17:09:31.292601 140460907546368 logging_writer.py:48] [150300] global_step=150300, grad_norm=4.049410343170166, loss=3.8096237182617188 -I0513 17:12:07.042749 140460899153664 logging_writer.py:48] [150400] global_step=150400, grad_norm=6.110259056091309, loss=3.9249372482299805 -I0513 17:13:58.666720 140460907546368 logging_writer.py:48] [150500] global_step=150500, grad_norm=10.180412292480469, loss=3.8809494972229004 -I0513 17:15:22.512068 140460899153664 logging_writer.py:48] [150600] global_step=150600, grad_norm=9.409932136535645, loss=3.7877864837646484 -I0513 17:16:48.656098 140460907546368 logging_writer.py:48] [150700] global_step=150700, grad_norm=3.4193012714385986, loss=3.7256712913513184 -I0513 17:19:04.286564 140460899153664 logging_writer.py:48] [150800] global_step=150800, grad_norm=5.9671196937561035, loss=3.869309663772583 -I0513 17:25:14.012737 140460907546368 logging_writer.py:48] [150900] global_step=150900, grad_norm=6.675341606140137, loss=3.797635555267334 -I0513 17:28:51.462216 140460899153664 logging_writer.py:48] [151000] global_step=151000, grad_norm=8.222138404846191, loss=3.855140209197998 -I0513 17:29:59.093385 140460907546368 logging_writer.py:48] [151100] global_step=151100, grad_norm=5.288030624389648, loss=3.8591935634613037 -I0513 17:31:07.564152 140460899153664 logging_writer.py:48] [151200] global_step=151200, grad_norm=5.123741149902344, loss=3.8579864501953125 -I0513 17:32:40.925191 140460907546368 logging_writer.py:48] [151300] global_step=151300, grad_norm=4.92185640335083, loss=3.9257893562316895 -I0513 17:36:16.025234 140460899153664 logging_writer.py:48] [151400] global_step=151400, grad_norm=4.671258926391602, loss=3.8533236980438232 -I0513 17:39:35.037812 140460907546368 logging_writer.py:48] [151500] global_step=151500, grad_norm=3.5626044273376465, loss=3.6997034549713135 -I0513 17:42:35.738299 140460899153664 logging_writer.py:48] [151600] global_step=151600, grad_norm=18.862258911132812, loss=4.118522644042969 -I0513 17:42:43.851818 140460907546368 logging_writer.py:48] [151603] global_step=151603, preemption_count=0, score=75918.6 -I0513 17:42:44.749156 140678261474496 submission_runner.py:857] Final imagenet_resnet score: 75918.62122917175 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-09-17-19.log b/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-09-17-19.log new file mode 100644 index 000000000..e72283417 --- /dev/null +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-09-17-19.log @@ -0,0 +1,2600 @@ +python submission_runner.py --framework=jax --workload=imagenet_resnet --submission_path=submissions_algorithms/submissions/self_tuning/schedule_free_adamw_jax_v2_bn_fix/submission.py --data_dir=/data/imagenet/jax --experiment_dir=/experiment_runs --experiment_name=submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_1 --overwrite=True --save_checkpoints=False --rng_seed=-304980929 --imagenet_v2_data_dir=/data/imagenet/jax --tuning_ruleset=self 2>&1 | tee -a /logs/imagenet_resnet_jax_08-31-2026-09-17-19.log +2026-08-31 09:17:19.922897: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered +WARNING: All log messages before absl::InitializeLog() is called are written to STDERR +E0000 00:00:1788167839.945634 13 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered +E0000 00:00:1788167839.953178 13 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered +W0000 00:00:1788167839.971543 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +W0000 00:00:1788167839.971569 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +W0000 00:00:1788167839.971572 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +W0000 00:00:1788167839.971574 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +/usr/local/lib/python3.11/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.) + _C._set_float32_matmul_precision(precision) +/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'repr' attribute with value False was provided to the `Field()` function, which has no effect in the context it was used. 'repr' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. + warnings.warn( +/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'frozen' attribute with value True was provided to the `Field()` function, which has no effect in the context it was used. 'frozen' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. + warnings.warn( +INFO:2026-08-31 09:17:30,301:jax._src.xla_bridge:830: Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory +I0831 09:17:30.301846 140659750036672 xla_bridge.py:830] Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory +I0831 09:17:31.001174 140659750036672 logger_utils.py:59] Removing existing experiment directory /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_1/imagenet_resnet_jax because --overwrite was set. +I0831 09:17:31.007174 140659750036672 logger_utils.py:84] Creating experiment directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_1/imagenet_resnet_jax. +I0831 09:17:31.603521 140659750036672 submission_runner.py:741] Creating directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_1/imagenet_resnet_jax/trial_1. +I0831 09:17:31.870208 140659750036672 submission_runner.py:242] Initializing dataset. +I0831 09:17:32.374876 140659750036672 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:17:32.433460 140659750036672 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. +I0831 09:17:32.715991 140659750036672 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. +I0831 09:17:32.794457 140659750036672 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:17:33.691803 140659750036672 submission_runner.py:251] Initializing model. +I0831 09:17:56.980294 140659750036672 submission_runner.py:294] Initializing optimizer. +I0831 09:17:58.569765 140659750036672 submission_runner.py:299] Initializing metrics bundle. +I0831 09:17:58.569987 140659750036672 submission_runner.py:321] Initializing checkpoint and logger. +I0831 09:17:58.577660 140659750036672 checkpoints.py:1098] Found no checkpoint files in /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_1/imagenet_resnet_jax/trial_1 with prefix checkpoint_ +I0831 09:17:58.577788 140659750036672 submission_runner.py:344] Saving meta data to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_1/imagenet_resnet_jax/trial_1/meta_data_0.json. +I0831 09:17:58.806552 140659750036672 submission_runner.py:348] Saving flags to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_1/imagenet_resnet_jax/trial_1/flags_0.json. +I0831 09:17:58.822477 140659750036672 submission_runner.py:359] Starting training loop. +/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. +See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. + warnings.warn("Some donated buffers were not usable:" +I0831 09:18:50.797195 140498647889664 logging_writer.py:48] [0] global_step=0, grad_norm=0.6739698052406311, loss=6.925764083862305 +I0831 09:18:51.998569 140659750036672 spec.py:333] Evaluating on the training split. +I0831 09:18:52.271038 140659750036672 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:18:52.276498 140659750036672 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. +I0831 09:18:52.295097 140659750036672 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. +I0831 09:18:52.335187 140659750036672 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:19:19.949734 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 09:19:19.954059 140659750036672 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:19:19.991881 140659750036672 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. +I0831 09:19:19.994528 140659750036672 reader.py:262] Creating a tf.data.Dataset reading 64 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. +I0831 09:19:20.137126 140659750036672 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split validation, from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:19:50.810942 140659750036672 spec.py:363] Evaluating on the test split. +I0831 09:19:50.872813 140659750036672 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 +I0831 09:19:50.923952 140659750036672 reader.py:262] Creating a tf.data.Dataset reading 16 files located in folders: /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0. +I0831 09:19:50.966313 140659750036672 logging_logger.py:49] Constructing tf.data.Dataset imagenet_v2 for split test, from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 +I0831 09:19:55.729682 140659750036672 submission_runner.py:516] Time since start: 116.91s, Step: 1, {'train/accuracy': Array(0.00187341, dtype=float32), 'train/loss': Array(6.9117374, dtype=float32), 'validation/accuracy': Array(0.00126, dtype=float32), 'validation/loss': Array(6.9117756, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(6.9119825, dtype=float32), 'test/num_examples': 10000, 'score': 53.1759774684906, 'total_duration': 116.90558505058289, 'accumulated_submission_time': 53.1759774684906, 'accumulated_eval_time': 63.72952103614807, 'accumulated_logging_time': 0} +I0831 09:19:55.761385 140462199400192 logging_writer.py:48] [1] accumulated_eval_time=63.7295, accumulated_logging_time=0, accumulated_submission_time=53.176, global_step=1, preemption_count=0, score=53.176, test/accuracy=0.0012000000569969416, test/loss=6.911982536315918, test/num_examples=10000, total_duration=116.906, train/accuracy=0.0018734055338427424, train/loss=6.911737442016602, validation/accuracy=0.0012599999317899346, validation/loss=6.911775588989258, validation/num_examples=50000 +/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. +See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. + warnings.warn("Some donated buffers were not usable:" +I0831 09:20:37.129345 140462157436672 logging_writer.py:48] [100] global_step=100, grad_norm=0.6964315176010132, loss=6.796125411987305 +I0831 09:21:22.395779 140462165829376 logging_writer.py:48] [200] global_step=200, grad_norm=0.8121169805526733, loss=6.479639530181885 +I0831 09:22:07.220787 140462157436672 logging_writer.py:48] [300] global_step=300, grad_norm=1.0721325874328613, loss=6.212225914001465 +I0831 09:22:51.343698 140462165829376 logging_writer.py:48] [400] global_step=400, grad_norm=1.1262391805648804, loss=5.954605579376221 +I0831 09:23:36.214612 140462157436672 logging_writer.py:48] [500] global_step=500, grad_norm=3.1064703464508057, loss=5.799692153930664 +I0831 09:24:21.359216 140462165829376 logging_writer.py:48] [600] global_step=600, grad_norm=3.630962371826172, loss=5.558352470397949 +I0831 09:25:05.521713 140462157436672 logging_writer.py:48] [700] global_step=700, grad_norm=4.996706008911133, loss=5.42868709564209 +I0831 09:25:49.019224 140462165829376 logging_writer.py:48] [800] global_step=800, grad_norm=4.612509250640869, loss=5.279544830322266 +I0831 09:26:33.658786 140462157436672 logging_writer.py:48] [900] global_step=900, grad_norm=4.8454437255859375, loss=5.071611404418945 +I0831 09:27:17.543823 140462165829376 logging_writer.py:48] [1000] global_step=1000, grad_norm=3.928295373916626, loss=4.910638332366943 +I0831 09:28:01.348965 140462157436672 logging_writer.py:48] [1100] global_step=1100, grad_norm=4.07260274887085, loss=4.919383525848389 +I0831 09:28:46.132364 140462165829376 logging_writer.py:48] [1200] global_step=1200, grad_norm=3.996246814727783, loss=4.73813533782959 +I0831 09:29:18.501567 140462157436672 logging_writer.py:48] [1300] global_step=1300, grad_norm=5.7899651527404785, loss=4.602542877197266 +I0831 09:29:45.661481 140462165829376 logging_writer.py:48] [1400] global_step=1400, grad_norm=3.528083562850952, loss=4.590331077575684 +I0831 09:30:12.886576 140462157436672 logging_writer.py:48] [1500] global_step=1500, grad_norm=4.655301570892334, loss=4.507709980010986 +I0831 09:30:40.063801 140462165829376 logging_writer.py:48] [1600] global_step=1600, grad_norm=4.822508335113525, loss=4.3325300216674805 +I0831 09:31:07.207689 140462157436672 logging_writer.py:48] [1700] global_step=1700, grad_norm=4.0672454833984375, loss=4.181547164916992 +I0831 09:31:34.420810 140462165829376 logging_writer.py:48] [1800] global_step=1800, grad_norm=6.445084095001221, loss=4.243796348571777 +I0831 09:32:01.591214 140462157436672 logging_writer.py:48] [1900] global_step=1900, grad_norm=4.479231834411621, loss=4.031885147094727 +I0831 09:32:28.708279 140462165829376 logging_writer.py:48] [2000] global_step=2000, grad_norm=2.974898338317871, loss=4.090665817260742 +I0831 09:32:55.952490 140462157436672 logging_writer.py:48] [2100] global_step=2100, grad_norm=4.390645980834961, loss=4.0698394775390625 +I0831 09:33:23.303023 140462165829376 logging_writer.py:48] [2200] global_step=2200, grad_norm=5.457085132598877, loss=3.878929853439331 +I0831 09:33:50.455903 140462157436672 logging_writer.py:48] [2300] global_step=2300, grad_norm=3.368861198425293, loss=3.776527166366577 +I0831 09:34:17.690873 140462165829376 logging_writer.py:48] [2400] global_step=2400, grad_norm=4.0132060050964355, loss=3.8108394145965576 +I0831 09:34:44.853113 140462157436672 logging_writer.py:48] [2500] global_step=2500, grad_norm=3.5858957767486572, loss=3.5933480262756348 +I0831 09:35:12.008490 140462165829376 logging_writer.py:48] [2600] global_step=2600, grad_norm=4.1020588874816895, loss=3.5579185485839844 +I0831 09:35:39.246356 140462157436672 logging_writer.py:48] [2700] global_step=2700, grad_norm=4.0402326583862305, loss=3.5771539211273193 +I0831 09:36:06.411015 140462165829376 logging_writer.py:48] [2800] global_step=2800, grad_norm=3.9309868812561035, loss=3.433032989501953 +I0831 09:36:33.551494 140462157436672 logging_writer.py:48] [2900] global_step=2900, grad_norm=3.9465811252593994, loss=3.4345099925994873 +I0831 09:37:00.785952 140462165829376 logging_writer.py:48] [3000] global_step=3000, grad_norm=3.957076072692871, loss=3.3551878929138184 +I0831 09:37:27.942860 140462157436672 logging_writer.py:48] [3100] global_step=3100, grad_norm=4.952515125274658, loss=3.2890000343322754 +I0831 09:37:55.302911 140462165829376 logging_writer.py:48] [3200] global_step=3200, grad_norm=2.8561599254608154, loss=3.232318639755249 +I0831 09:38:22.526938 140462157436672 logging_writer.py:48] [3300] global_step=3300, grad_norm=3.490729808807373, loss=3.1476402282714844 +I0831 09:38:49.678438 140462165829376 logging_writer.py:48] [3400] global_step=3400, grad_norm=3.2990849018096924, loss=3.040374755859375 +I0831 09:39:16.860544 140462157436672 logging_writer.py:48] [3500] global_step=3500, grad_norm=3.2475318908691406, loss=3.1476457118988037 +I0831 09:39:44.117379 140462165829376 logging_writer.py:48] [3600] global_step=3600, grad_norm=2.7855331897735596, loss=3.0557751655578613 +I0831 09:40:11.282425 140462157436672 logging_writer.py:48] [3700] global_step=3700, grad_norm=3.509517192840576, loss=3.112171173095703 +I0831 09:40:38.429498 140462165829376 logging_writer.py:48] [3800] global_step=3800, grad_norm=2.916825294494629, loss=2.828758955001831 +I0831 09:41:05.666673 140462157436672 logging_writer.py:48] [3900] global_step=3900, grad_norm=4.093054294586182, loss=2.860326051712036 +I0831 09:41:32.820900 140462165829376 logging_writer.py:48] [4000] global_step=4000, grad_norm=2.7748939990997314, loss=2.7570395469665527 +I0831 09:41:59.977961 140462157436672 logging_writer.py:48] [4100] global_step=4100, grad_norm=2.2204842567443848, loss=2.773515462875366 +I0831 09:42:27.161726 140462165829376 logging_writer.py:48] [4200] global_step=4200, grad_norm=3.0503320693969727, loss=2.8372161388397217 +I0831 09:42:54.543531 140462157436672 logging_writer.py:48] [4300] global_step=4300, grad_norm=2.2261972427368164, loss=2.8358631134033203 +I0831 09:43:21.688625 140462165829376 logging_writer.py:48] [4400] global_step=4400, grad_norm=2.8291518688201904, loss=2.5504839420318604 +I0831 09:43:48.888868 140462157436672 logging_writer.py:48] [4500] global_step=4500, grad_norm=3.1905479431152344, loss=2.6903815269470215 +I0831 09:44:16.028227 140462165829376 logging_writer.py:48] [4600] global_step=4600, grad_norm=2.9553496837615967, loss=2.456031322479248 +I0831 09:44:43.181112 140462157436672 logging_writer.py:48] [4700] global_step=4700, grad_norm=2.751591920852661, loss=2.5452725887298584 +I0831 09:45:10.425918 140462165829376 logging_writer.py:48] [4800] global_step=4800, grad_norm=2.232351303100586, loss=2.561889171600342 +I0831 09:45:37.595349 140462157436672 logging_writer.py:48] [4900] global_step=4900, grad_norm=2.9586894512176514, loss=2.5514352321624756 +I0831 09:46:04.738105 140462165829376 logging_writer.py:48] [5000] global_step=5000, grad_norm=2.6541900634765625, loss=2.579103946685791 +I0831 09:46:31.961678 140462157436672 logging_writer.py:48] [5100] global_step=5100, grad_norm=2.7490530014038086, loss=2.5441694259643555 +I0831 09:46:59.116663 140462165829376 logging_writer.py:48] [5200] global_step=5200, grad_norm=2.850781202316284, loss=2.5506644248962402 +I0831 09:47:26.276400 140462157436672 logging_writer.py:48] [5300] global_step=5300, grad_norm=2.353989601135254, loss=2.4298558235168457 +I0831 09:47:53.691962 140462165829376 logging_writer.py:48] [5400] global_step=5400, grad_norm=2.9130468368530273, loss=2.391958713531494 +I0831 09:48:20.843053 140462157436672 logging_writer.py:48] [5500] global_step=5500, grad_norm=2.571042060852051, loss=2.3410792350769043 +I0831 09:48:47.992897 140462165829376 logging_writer.py:48] [5600] global_step=5600, grad_norm=2.8292415142059326, loss=2.2800538539886475 +I0831 09:49:15.180390 140462157436672 logging_writer.py:48] [5700] global_step=5700, grad_norm=1.9617319107055664, loss=2.2136735916137695 +I0831 09:49:42.300506 140462165829376 logging_writer.py:48] [5800] global_step=5800, grad_norm=2.4758846759796143, loss=2.227445602416992 +I0831 09:50:09.442220 140462157436672 logging_writer.py:48] [5900] global_step=5900, grad_norm=3.0113320350646973, loss=2.4582290649414062 +I0831 09:50:36.671079 140462165829376 logging_writer.py:48] [6000] global_step=6000, grad_norm=2.4725732803344727, loss=2.3466849327087402 +I0831 09:51:03.858357 140462157436672 logging_writer.py:48] [6100] global_step=6100, grad_norm=2.351223945617676, loss=2.2409472465515137 +I0831 09:51:31.004857 140462165829376 logging_writer.py:48] [6200] global_step=6200, grad_norm=2.3518292903900146, loss=2.287602663040161 +I0831 09:51:58.183176 140462157436672 logging_writer.py:48] [6300] global_step=6300, grad_norm=2.2440106868743896, loss=2.305797815322876 +I0831 09:52:25.334387 140462165829376 logging_writer.py:48] [6400] global_step=6400, grad_norm=2.167264223098755, loss=2.256558656692505 +I0831 09:52:52.678747 140462157436672 logging_writer.py:48] [6500] global_step=6500, grad_norm=2.5926637649536133, loss=2.3127479553222656 +I0831 09:53:11.870809 140659750036672 spec.py:333] Evaluating on the training split. +I0831 09:53:23.228886 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 09:53:32.696458 140659750036672 spec.py:363] Evaluating on the test split. +I0831 09:53:33.615926 140659750036672 submission_runner.py:516] Time since start: 2134.79s, Step: 6572, {'train/accuracy': Array(0.6306003, dtype=float32), 'train/loss': Array(1.5325073, dtype=float32), 'validation/accuracy': Array(0.57357997, dtype=float32), 'validation/loss': Array(1.8282131, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.44230002, dtype=float32), 'test/loss': Array(2.5904195, dtype=float32), 'test/num_examples': 10000, 'score': 2049.2234110832214, 'total_duration': 2134.791506290436, 'accumulated_submission_time': 2049.2234110832214, 'accumulated_eval_time': 85.47269701957703, 'accumulated_logging_time': 0.040308237075805664} +I0831 09:53:33.663228 140462207792896 logging_writer.py:48] [6572] accumulated_eval_time=85.4727, accumulated_logging_time=0.0403082, accumulated_submission_time=2049.22, global_step=6572, preemption_count=0, score=2049.22, test/accuracy=0.442300021648407, test/loss=2.5904195308685303, test/num_examples=10000, total_duration=2134.79, train/accuracy=0.6306002736091614, train/loss=1.532507300376892, validation/accuracy=0.5735799670219421, validation/loss=1.828213095664978, validation/num_examples=50000 +I0831 09:53:41.690324 140462291654400 logging_writer.py:48] [6600] global_step=6600, grad_norm=3.6754932403564453, loss=2.146148204803467 +I0831 09:54:08.840731 140462207792896 logging_writer.py:48] [6700] global_step=6700, grad_norm=2.1184144020080566, loss=2.284724235534668 +I0831 09:54:36.012640 140462291654400 logging_writer.py:48] [6800] global_step=6800, grad_norm=2.376620292663574, loss=2.2591500282287598 +I0831 09:55:03.227610 140462207792896 logging_writer.py:48] [6900] global_step=6900, grad_norm=2.3602869510650635, loss=2.1125152111053467 +I0831 09:55:30.354702 140462291654400 logging_writer.py:48] [7000] global_step=7000, grad_norm=2.852790594100952, loss=2.1171715259552 +I0831 09:55:57.487476 140462207792896 logging_writer.py:48] [7100] global_step=7100, grad_norm=4.532451629638672, loss=2.000176429748535 +I0831 09:56:24.680791 140462291654400 logging_writer.py:48] [7200] global_step=7200, grad_norm=2.4538440704345703, loss=1.983835220336914 +I0831 09:56:51.839541 140462207792896 logging_writer.py:48] [7300] global_step=7300, grad_norm=2.755748748779297, loss=2.1194794178009033 +I0831 09:57:19.018742 140462291654400 logging_writer.py:48] [7400] global_step=7400, grad_norm=2.9472603797912598, loss=2.119140863418579 +I0831 09:57:46.201977 140462207792896 logging_writer.py:48] [7500] global_step=7500, grad_norm=2.4568779468536377, loss=2.0059566497802734 +I0831 09:58:13.557440 140462291654400 logging_writer.py:48] [7600] global_step=7600, grad_norm=2.2398524284362793, loss=2.028844118118286 +I0831 09:58:40.698849 140462207792896 logging_writer.py:48] [7700] global_step=7700, grad_norm=2.833209276199341, loss=2.1118547916412354 +I0831 09:59:07.900511 140462291654400 logging_writer.py:48] [7800] global_step=7800, grad_norm=2.7344605922698975, loss=2.1323328018188477 +I0831 09:59:35.027035 140462207792896 logging_writer.py:48] [7900] global_step=7900, grad_norm=2.33223819732666, loss=2.188560962677002 +I0831 10:00:02.148133 140462291654400 logging_writer.py:48] [8000] global_step=8000, grad_norm=2.6854770183563232, loss=2.0299248695373535 +I0831 10:00:29.363302 140462207792896 logging_writer.py:48] [8100] global_step=8100, grad_norm=2.240314483642578, loss=1.9469785690307617 +I0831 10:00:56.482617 140462291654400 logging_writer.py:48] [8200] global_step=8200, grad_norm=2.121680736541748, loss=1.9682697057724 +I0831 10:01:23.629285 140462207792896 logging_writer.py:48] [8300] global_step=8300, grad_norm=2.3822858333587646, loss=1.927729606628418 +I0831 10:01:50.846941 140462291654400 logging_writer.py:48] [8400] global_step=8400, grad_norm=2.7464606761932373, loss=2.004028797149658 +I0831 10:02:18.000859 140462207792896 logging_writer.py:48] [8500] global_step=8500, grad_norm=2.2814574241638184, loss=2.055668830871582 +I0831 10:02:45.342330 140462291654400 logging_writer.py:48] [8600] global_step=8600, grad_norm=2.0253264904022217, loss=1.9111089706420898 +I0831 10:03:12.544453 140462207792896 logging_writer.py:48] [8700] global_step=8700, grad_norm=2.228611469268799, loss=1.985327959060669 +I0831 10:03:39.693298 140462291654400 logging_writer.py:48] [8800] global_step=8800, grad_norm=2.6028523445129395, loss=2.0528459548950195 +I0831 10:04:06.868272 140462207792896 logging_writer.py:48] [8900] global_step=8900, grad_norm=2.235804319381714, loss=1.9673993587493896 +I0831 10:04:34.101693 140462291654400 logging_writer.py:48] [9000] global_step=9000, grad_norm=2.3639976978302, loss=1.9105784893035889 +I0831 10:05:01.242782 140462207792896 logging_writer.py:48] [9100] global_step=9100, grad_norm=2.1943278312683105, loss=1.986356496810913 +I0831 10:05:28.375346 140462291654400 logging_writer.py:48] [9200] global_step=9200, grad_norm=2.3464906215667725, loss=1.9924416542053223 +I0831 10:05:55.572153 140462207792896 logging_writer.py:48] [9300] global_step=9300, grad_norm=1.669503092765808, loss=1.7664964199066162 +I0831 10:06:22.708218 140462291654400 logging_writer.py:48] [9400] global_step=9400, grad_norm=2.2533938884735107, loss=1.7622374296188354 +I0831 10:06:49.898433 140462207792896 logging_writer.py:48] [9500] global_step=9500, grad_norm=2.191455841064453, loss=1.8608760833740234 +I0831 10:07:17.125061 140462291654400 logging_writer.py:48] [9600] global_step=9600, grad_norm=2.046034812927246, loss=1.90233314037323 +I0831 10:07:44.469987 140462207792896 logging_writer.py:48] [9700] global_step=9700, grad_norm=2.016108989715576, loss=1.795114517211914 +I0831 10:08:11.599521 140462291654400 logging_writer.py:48] [9800] global_step=9800, grad_norm=2.768373489379883, loss=1.9371118545532227 +I0831 10:08:38.805683 140462207792896 logging_writer.py:48] [9900] global_step=9900, grad_norm=1.8984369039535522, loss=1.7905786037445068 +I0831 10:09:05.976620 140462291654400 logging_writer.py:48] [10000] global_step=10000, grad_norm=2.5216500759124756, loss=1.8121864795684814 +I0831 10:09:33.114503 140462207792896 logging_writer.py:48] [10100] global_step=10100, grad_norm=2.2036962509155273, loss=1.8106070756912231 +I0831 10:10:00.318805 140462291654400 logging_writer.py:48] [10200] global_step=10200, grad_norm=2.080181360244751, loss=1.7938354015350342 +I0831 10:10:27.444236 140462207792896 logging_writer.py:48] [10300] global_step=10300, grad_norm=1.9111698865890503, loss=1.8409600257873535 +I0831 10:10:54.614145 140462291654400 logging_writer.py:48] [10400] global_step=10400, grad_norm=2.1766438484191895, loss=1.7789649963378906 +I0831 10:11:21.801310 140462207792896 logging_writer.py:48] [10500] global_step=10500, grad_norm=2.7298994064331055, loss=1.7708156108856201 +I0831 10:11:48.935449 140462291654400 logging_writer.py:48] [10600] global_step=10600, grad_norm=2.6131584644317627, loss=1.7813756465911865 +I0831 10:12:16.088700 140462207792896 logging_writer.py:48] [10700] global_step=10700, grad_norm=2.171294927597046, loss=1.6262258291244507 +I0831 10:12:43.524435 140462291654400 logging_writer.py:48] [10800] global_step=10800, grad_norm=2.811069965362549, loss=1.7096426486968994 +I0831 10:13:10.678869 140462207792896 logging_writer.py:48] [10900] global_step=10900, grad_norm=2.269869804382324, loss=1.7413790225982666 +I0831 10:13:37.804855 140462291654400 logging_writer.py:48] [11000] global_step=11000, grad_norm=2.712602138519287, loss=1.7849931716918945 +I0831 10:14:05.006917 140462207792896 logging_writer.py:48] [11100] global_step=11100, grad_norm=2.4884133338928223, loss=1.8502622842788696 +I0831 10:14:32.148612 140462291654400 logging_writer.py:48] [11200] global_step=11200, grad_norm=2.8317806720733643, loss=1.8181617259979248 +I0831 10:14:59.265797 140462207792896 logging_writer.py:48] [11300] global_step=11300, grad_norm=2.4868650436401367, loss=1.748432993888855 +I0831 10:15:26.470335 140462291654400 logging_writer.py:48] [11400] global_step=11400, grad_norm=2.510864496231079, loss=1.8518562316894531 +I0831 10:15:53.618223 140462207792896 logging_writer.py:48] [11500] global_step=11500, grad_norm=2.389118194580078, loss=1.8340940475463867 +I0831 10:16:20.763849 140462291654400 logging_writer.py:48] [11600] global_step=11600, grad_norm=3.1187994480133057, loss=1.687792420387268 +I0831 10:16:47.944489 140462207792896 logging_writer.py:48] [11700] global_step=11700, grad_norm=2.4140069484710693, loss=1.8124284744262695 +I0831 10:17:15.091807 140462291654400 logging_writer.py:48] [11800] global_step=11800, grad_norm=2.1857166290283203, loss=1.6492376327514648 +I0831 10:17:42.457854 140462207792896 logging_writer.py:48] [11900] global_step=11900, grad_norm=2.439077377319336, loss=1.5927047729492188 +I0831 10:18:09.656205 140462291654400 logging_writer.py:48] [12000] global_step=12000, grad_norm=2.5857770442962646, loss=1.7065401077270508 +I0831 10:18:36.805850 140462207792896 logging_writer.py:48] [12100] global_step=12100, grad_norm=1.9919244050979614, loss=1.7136585712432861 +I0831 10:19:03.944302 140462291654400 logging_writer.py:48] [12200] global_step=12200, grad_norm=2.260702610015869, loss=1.6966392993927002 +I0831 10:19:31.143750 140462207792896 logging_writer.py:48] [12300] global_step=12300, grad_norm=2.4203598499298096, loss=1.679175615310669 +I0831 10:19:58.266631 140462291654400 logging_writer.py:48] [12400] global_step=12400, grad_norm=2.1513073444366455, loss=1.5677989721298218 +I0831 10:20:25.394826 140462207792896 logging_writer.py:48] [12500] global_step=12500, grad_norm=2.2790517807006836, loss=1.6863300800323486 +I0831 10:20:52.623670 140462291654400 logging_writer.py:48] [12600] global_step=12600, grad_norm=2.408601999282837, loss=1.645574688911438 +I0831 10:21:19.776096 140462207792896 logging_writer.py:48] [12700] global_step=12700, grad_norm=2.409668445587158, loss=1.6535801887512207 +I0831 10:21:46.928018 140462291654400 logging_writer.py:48] [12800] global_step=12800, grad_norm=2.549299478530884, loss=1.8028111457824707 +I0831 10:22:14.305039 140462207792896 logging_writer.py:48] [12900] global_step=12900, grad_norm=2.246994972229004, loss=1.783373236656189 +I0831 10:22:41.446722 140462291654400 logging_writer.py:48] [13000] global_step=13000, grad_norm=2.415525436401367, loss=1.6351895332336426 +I0831 10:23:08.577767 140462207792896 logging_writer.py:48] [13100] global_step=13100, grad_norm=2.2035224437713623, loss=1.6449565887451172 +I0831 10:23:35.771061 140462291654400 logging_writer.py:48] [13200] global_step=13200, grad_norm=2.2959773540496826, loss=1.6004631519317627 +I0831 10:24:02.913881 140462207792896 logging_writer.py:48] [13300] global_step=13300, grad_norm=2.366063117980957, loss=1.7040777206420898 +I0831 10:24:30.069758 140462291654400 logging_writer.py:48] [13400] global_step=13400, grad_norm=2.1199851036071777, loss=1.5748274326324463 +I0831 10:24:57.262408 140462207792896 logging_writer.py:48] [13500] global_step=13500, grad_norm=2.343412160873413, loss=1.6480287313461304 +I0831 10:25:24.392125 140462291654400 logging_writer.py:48] [13600] global_step=13600, grad_norm=2.547119379043579, loss=1.6030848026275635 +I0831 10:25:51.528706 140462207792896 logging_writer.py:48] [13700] global_step=13700, grad_norm=2.3141095638275146, loss=1.6994836330413818 +I0831 10:26:18.728930 140462291654400 logging_writer.py:48] [13800] global_step=13800, grad_norm=2.285571813583374, loss=1.5324442386627197 +I0831 10:26:45.863877 140462207792896 logging_writer.py:48] [13900] global_step=13900, grad_norm=2.379819631576538, loss=1.5746440887451172 +I0831 10:26:49.811743 140659750036672 spec.py:333] Evaluating on the training split. +I0831 10:27:02.462969 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 10:27:11.555723 140659750036672 spec.py:363] Evaluating on the test split. +I0831 10:27:12.430308 140659750036672 submission_runner.py:516] Time since start: 4153.61s, Step: 13916, {'train/accuracy': Array(0.7621771, dtype=float32), 'train/loss': Array(0.9435853, dtype=float32), 'validation/accuracy': Array(0.66736, dtype=float32), 'validation/loss': Array(1.3725033, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.53010005, dtype=float32), 'test/loss': Array(2.1301467, dtype=float32), 'test/num_examples': 10000, 'score': 4045.282567501068, 'total_duration': 4153.60573720932, 'accumulated_submission_time': 4045.282567501068, 'accumulated_eval_time': 108.08916926383972, 'accumulated_logging_time': 0.11857056617736816} +I0831 10:27:12.492893 140462291654400 logging_writer.py:48] [13916] accumulated_eval_time=108.089, accumulated_logging_time=0.118571, accumulated_submission_time=4045.28, global_step=13916, preemption_count=0, score=4045.28, test/accuracy=0.5301000475883484, test/loss=2.1301467418670654, test/num_examples=10000, total_duration=4153.61, train/accuracy=0.7621771097183228, train/loss=0.9435852766036987, validation/accuracy=0.6673600077629089, validation/loss=1.3725032806396484, validation/num_examples=50000 +I0831 10:27:35.969255 140462207792896 logging_writer.py:48] [14000] global_step=14000, grad_norm=2.5680906772613525, loss=1.5774723291397095 +I0831 10:28:03.161202 140462291654400 logging_writer.py:48] [14100] global_step=14100, grad_norm=2.374664306640625, loss=1.719275712966919 +I0831 10:28:30.307424 140462207792896 logging_writer.py:48] [14200] global_step=14200, grad_norm=2.909944534301758, loss=1.648324966430664 +I0831 10:28:57.437073 140462291654400 logging_writer.py:48] [14300] global_step=14300, grad_norm=2.5713629722595215, loss=1.592890739440918 +I0831 10:29:24.619043 140462207792896 logging_writer.py:48] [14400] global_step=14400, grad_norm=2.5559241771698, loss=1.688223123550415 +I0831 10:29:51.765385 140462291654400 logging_writer.py:48] [14500] global_step=14500, grad_norm=2.1649234294891357, loss=1.5837830305099487 +I0831 10:30:18.915499 140462207792896 logging_writer.py:48] [14600] global_step=14600, grad_norm=2.5722508430480957, loss=1.6536203622817993 +I0831 10:30:46.152587 140462291654400 logging_writer.py:48] [14700] global_step=14700, grad_norm=3.484571695327759, loss=1.5803334712982178 +I0831 10:31:13.300446 140462207792896 logging_writer.py:48] [14800] global_step=14800, grad_norm=2.7065367698669434, loss=1.669950008392334 +I0831 10:31:40.410676 140462291654400 logging_writer.py:48] [14900] global_step=14900, grad_norm=2.473562240600586, loss=1.694442629814148 +I0831 10:32:07.609059 140462207792896 logging_writer.py:48] [15000] global_step=15000, grad_norm=2.5298497676849365, loss=1.6152448654174805 +I0831 10:32:34.924647 140462291654400 logging_writer.py:48] [15100] global_step=15100, grad_norm=3.1821839809417725, loss=1.5178382396697998 +I0831 10:33:02.040521 140462207792896 logging_writer.py:48] [15200] global_step=15200, grad_norm=2.971757650375366, loss=1.5309734344482422 +I0831 10:33:29.260599 140462291654400 logging_writer.py:48] [15300] global_step=15300, grad_norm=2.6300106048583984, loss=1.578116774559021 +I0831 10:33:56.395581 140462207792896 logging_writer.py:48] [15400] global_step=15400, grad_norm=3.0556576251983643, loss=1.6937977075576782 +I0831 10:34:23.536761 140462291654400 logging_writer.py:48] [15500] global_step=15500, grad_norm=2.7657971382141113, loss=1.591247797012329 +I0831 10:34:50.714614 140462207792896 logging_writer.py:48] [15600] global_step=15600, grad_norm=2.7367196083068848, loss=1.6313602924346924 +I0831 10:35:17.861080 140462291654400 logging_writer.py:48] [15700] global_step=15700, grad_norm=3.1936049461364746, loss=1.5898640155792236 +I0831 10:35:45.009954 140462207792896 logging_writer.py:48] [15800] global_step=15800, grad_norm=2.6657676696777344, loss=1.5066211223602295 +I0831 10:36:12.233393 140462291654400 logging_writer.py:48] [15900] global_step=15900, grad_norm=2.8907532691955566, loss=1.612101435661316 +I0831 10:36:39.370964 140462207792896 logging_writer.py:48] [16000] global_step=16000, grad_norm=2.8402769565582275, loss=1.5437281131744385 +I0831 10:37:06.509610 140462291654400 logging_writer.py:48] [16100] global_step=16100, grad_norm=2.55635142326355, loss=1.424768090248108 +I0831 10:37:33.958329 140462207792896 logging_writer.py:48] [16200] global_step=16200, grad_norm=2.514019727706909, loss=1.6056722402572632 +I0831 10:38:01.064362 140462291654400 logging_writer.py:48] [16300] global_step=16300, grad_norm=3.4673306941986084, loss=1.6122403144836426 +I0831 10:38:28.178598 140462207792896 logging_writer.py:48] [16400] global_step=16400, grad_norm=2.7446064949035645, loss=1.4478479623794556 +I0831 10:38:55.378334 140462291654400 logging_writer.py:48] [16500] global_step=16500, grad_norm=2.861154317855835, loss=1.4633314609527588 +I0831 10:39:22.528518 140462207792896 logging_writer.py:48] [16600] global_step=16600, grad_norm=2.89803147315979, loss=1.4998345375061035 +I0831 10:39:49.672561 140462291654400 logging_writer.py:48] [16700] global_step=16700, grad_norm=2.9622349739074707, loss=1.414170742034912 +I0831 10:40:16.851900 140462207792896 logging_writer.py:48] [16800] global_step=16800, grad_norm=3.2053935527801514, loss=1.4510498046875 +I0831 10:40:44.009473 140462291654400 logging_writer.py:48] [16900] global_step=16900, grad_norm=2.9549124240875244, loss=1.5469920635223389 +I0831 10:41:11.141363 140462207792896 logging_writer.py:48] [17000] global_step=17000, grad_norm=3.36299729347229, loss=1.5127586126327515 +I0831 10:41:38.338854 140462291654400 logging_writer.py:48] [17100] global_step=17100, grad_norm=2.978367567062378, loss=1.4687614440917969 +I0831 10:42:05.506339 140462207792896 logging_writer.py:48] [17200] global_step=17200, grad_norm=3.1030945777893066, loss=1.5492079257965088 +I0831 10:42:32.880960 140462291654400 logging_writer.py:48] [17300] global_step=17300, grad_norm=3.175999402999878, loss=1.504704236984253 +I0831 10:43:00.111091 140462207792896 logging_writer.py:48] [17400] global_step=17400, grad_norm=3.1675915718078613, loss=1.4444739818572998 +I0831 10:43:27.250918 140462291654400 logging_writer.py:48] [17500] global_step=17500, grad_norm=3.2506039142608643, loss=1.493762731552124 +I0831 10:43:54.364377 140462207792896 logging_writer.py:48] [17600] global_step=17600, grad_norm=3.2194998264312744, loss=1.490762710571289 +I0831 10:44:21.552435 140462291654400 logging_writer.py:48] [17700] global_step=17700, grad_norm=3.3054392337799072, loss=1.4821350574493408 +I0831 10:44:48.658061 140462207792896 logging_writer.py:48] [17800] global_step=17800, grad_norm=3.4288899898529053, loss=1.496376395225525 +I0831 10:45:15.783975 140462291654400 logging_writer.py:48] [17900] global_step=17900, grad_norm=3.809194803237915, loss=1.4803627729415894 +I0831 10:45:43.026231 140462207792896 logging_writer.py:48] [18000] global_step=18000, grad_norm=3.376085042953491, loss=1.5223851203918457 +I0831 10:46:10.154620 140462291654400 logging_writer.py:48] [18100] global_step=18100, grad_norm=3.6558287143707275, loss=1.5483472347259521 +I0831 10:46:37.321955 140462207792896 logging_writer.py:48] [18200] global_step=18200, grad_norm=3.7369649410247803, loss=1.5370577573776245 +I0831 10:47:04.758229 140462291654400 logging_writer.py:48] [18300] global_step=18300, grad_norm=3.928854465484619, loss=1.5301045179367065 +I0831 10:47:31.881158 140462207792896 logging_writer.py:48] [18400] global_step=18400, grad_norm=4.132620811462402, loss=1.5771081447601318 +I0831 10:47:59.047307 140462291654400 logging_writer.py:48] [18500] global_step=18500, grad_norm=4.37750768661499, loss=1.6002705097198486 +I0831 10:48:26.260580 140462207792896 logging_writer.py:48] [18600] global_step=18600, grad_norm=4.076304912567139, loss=1.4915862083435059 +I0831 10:48:53.420092 140462291654400 logging_writer.py:48] [18700] global_step=18700, grad_norm=4.026027202606201, loss=1.446304202079773 +I0831 10:49:20.571087 140462207792896 logging_writer.py:48] [18800] global_step=18800, grad_norm=4.3599066734313965, loss=1.5809693336486816 +I0831 10:49:47.747143 140462291654400 logging_writer.py:48] [18900] global_step=18900, grad_norm=4.713173866271973, loss=1.470792293548584 +I0831 10:50:14.877334 140462207792896 logging_writer.py:48] [19000] global_step=19000, grad_norm=4.568814277648926, loss=1.5624616146087646 +I0831 10:50:42.001276 140462291654400 logging_writer.py:48] [19100] global_step=19100, grad_norm=4.695838928222656, loss=1.531742811203003 +I0831 10:51:09.221976 140462207792896 logging_writer.py:48] [19200] global_step=19200, grad_norm=4.969592094421387, loss=1.5617458820343018 +I0831 10:51:36.354297 140462291654400 logging_writer.py:48] [19300] global_step=19300, grad_norm=5.660805702209473, loss=1.5578421354293823 +I0831 10:52:03.711159 140462207792896 logging_writer.py:48] [19400] global_step=19400, grad_norm=5.130966663360596, loss=1.5209615230560303 +I0831 10:52:30.895575 140462291654400 logging_writer.py:48] [19500] global_step=19500, grad_norm=5.260776996612549, loss=1.4526710510253906 +I0831 10:52:58.046101 140462207792896 logging_writer.py:48] [19600] global_step=19600, grad_norm=5.340883731842041, loss=1.4840925931930542 +I0831 10:53:25.178373 140462291654400 logging_writer.py:48] [19700] global_step=19700, grad_norm=5.665178298950195, loss=1.5533970594406128 +I0831 10:53:52.385903 140462207792896 logging_writer.py:48] [19800] global_step=19800, grad_norm=5.40863561630249, loss=1.5447629690170288 +I0831 10:54:19.511529 140462291654400 logging_writer.py:48] [19900] global_step=19900, grad_norm=6.1310648918151855, loss=1.4513378143310547 +I0831 10:54:46.645640 140462207792896 logging_writer.py:48] [20000] global_step=20000, grad_norm=6.391512393951416, loss=1.4875731468200684 +I0831 10:55:13.837154 140462291654400 logging_writer.py:48] [20100] global_step=20100, grad_norm=6.586771488189697, loss=1.4764924049377441 +I0831 10:55:40.958710 140462207792896 logging_writer.py:48] [20200] global_step=20200, grad_norm=7.218862056732178, loss=1.6000481843948364 +I0831 10:56:08.103439 140462291654400 logging_writer.py:48] [20300] global_step=20300, grad_norm=7.027608871459961, loss=1.6368716955184937 +I0831 10:56:35.327436 140462207792896 logging_writer.py:48] [20400] global_step=20400, grad_norm=6.813013076782227, loss=1.5145292282104492 +I0831 10:57:02.684443 140462291654400 logging_writer.py:48] [20500] global_step=20500, grad_norm=7.27241849899292, loss=1.5426185131072998 +I0831 10:57:29.837340 140462207792896 logging_writer.py:48] [20600] global_step=20600, grad_norm=8.036831855773926, loss=1.5630038976669312 +I0831 10:57:57.042104 140462291654400 logging_writer.py:48] [20700] global_step=20700, grad_norm=7.873350143432617, loss=1.5745408535003662 +I0831 10:58:24.202287 140462207792896 logging_writer.py:48] [20800] global_step=20800, grad_norm=8.035379409790039, loss=1.4845781326293945 +I0831 10:58:51.346819 140462291654400 logging_writer.py:48] [20900] global_step=20900, grad_norm=9.4005765914917, loss=1.4661574363708496 +I0831 10:59:18.527436 140462207792896 logging_writer.py:48] [21000] global_step=21000, grad_norm=8.914730072021484, loss=1.4609622955322266 +I0831 10:59:45.660397 140462291654400 logging_writer.py:48] [21100] global_step=21100, grad_norm=9.12794303894043, loss=1.479184865951538 +I0831 11:00:12.837373 140462207792896 logging_writer.py:48] [21200] global_step=21200, grad_norm=9.464282989501953, loss=1.5207958221435547 +I0831 11:00:28.508384 140659750036672 spec.py:333] Evaluating on the training split. +I0831 11:00:39.961244 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 11:00:48.962906 140659750036672 spec.py:363] Evaluating on the test split. +I0831 11:00:49.839258 140659750036672 submission_runner.py:516] Time since start: 6171.01s, Step: 21259, {'train/accuracy': Array(0.8075972, dtype=float32), 'train/loss': Array(0.7395382, dtype=float32), 'validation/accuracy': Array(0.69585997, dtype=float32), 'validation/loss': Array(1.2365702, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.5655, dtype=float32), 'test/loss': Array(1.971104, dtype=float32), 'test/num_examples': 10000, 'score': 6041.208413362503, 'total_duration': 6171.006928682327, 'accumulated_submission_time': 6041.208413362503, 'accumulated_eval_time': 129.41019129753113, 'accumulated_logging_time': 0.21257448196411133} +I0831 11:00:49.886009 140462291654400 logging_writer.py:48] [21259] accumulated_eval_time=129.41, accumulated_logging_time=0.212574, accumulated_submission_time=6041.21, global_step=21259, preemption_count=0, score=6041.21, test/accuracy=0.565500020980835, test/loss=1.9711040258407593, test/num_examples=10000, total_duration=6171.01, train/accuracy=0.8075972199440002, train/loss=0.7395381927490234, validation/accuracy=0.695859968662262, validation/loss=1.2365702390670776, validation/num_examples=50000 +I0831 11:01:01.431082 140462207792896 logging_writer.py:48] [21300] global_step=21300, grad_norm=9.959440231323242, loss=1.5244860649108887 +I0831 11:01:28.542243 140462291654400 logging_writer.py:48] [21400] global_step=21400, grad_norm=9.996405601501465, loss=1.5710407495498657 +I0831 11:01:55.672532 140462207792896 logging_writer.py:48] [21500] global_step=21500, grad_norm=9.766855239868164, loss=1.4647681713104248 +I0831 11:02:23.132272 140462291654400 logging_writer.py:48] [21600] global_step=21600, grad_norm=9.829764366149902, loss=1.4652414321899414 +I0831 11:02:50.291771 140462207792896 logging_writer.py:48] [21700] global_step=21700, grad_norm=10.630622863769531, loss=1.455362319946289 +I0831 11:03:17.407003 140462291654400 logging_writer.py:48] [21800] global_step=21800, grad_norm=11.314786911010742, loss=1.5204124450683594 +I0831 11:03:44.601691 140462207792896 logging_writer.py:48] [21900] global_step=21900, grad_norm=10.795660972595215, loss=1.5607848167419434 +I0831 11:04:11.729961 140462291654400 logging_writer.py:48] [22000] global_step=22000, grad_norm=11.20372486114502, loss=1.477311134338379 +I0831 11:04:38.879221 140462207792896 logging_writer.py:48] [22100] global_step=22100, grad_norm=12.30726146697998, loss=1.4531530141830444 +I0831 11:05:06.088661 140462291654400 logging_writer.py:48] [22200] global_step=22200, grad_norm=11.807415008544922, loss=1.5424872636795044 +I0831 11:05:33.222708 140462207792896 logging_writer.py:48] [22300] global_step=22300, grad_norm=11.279500007629395, loss=1.480666995048523 +I0831 11:06:00.354058 140462291654400 logging_writer.py:48] [22400] global_step=22400, grad_norm=11.795482635498047, loss=1.478236436843872 +I0831 11:06:27.573340 140462207792896 logging_writer.py:48] [22500] global_step=22500, grad_norm=12.360130310058594, loss=1.4908270835876465 +I0831 11:06:54.986969 140462291654400 logging_writer.py:48] [22600] global_step=22600, grad_norm=12.287235260009766, loss=1.5347745418548584 +I0831 11:07:22.145145 140462207792896 logging_writer.py:48] [22700] global_step=22700, grad_norm=12.216634750366211, loss=1.4644726514816284 +I0831 11:07:49.352163 140462291654400 logging_writer.py:48] [22800] global_step=22800, grad_norm=13.242913246154785, loss=1.5811470746994019 +I0831 11:08:16.472045 140462207792896 logging_writer.py:48] [22900] global_step=22900, grad_norm=12.785090446472168, loss=1.5857042074203491 +I0831 11:08:43.602873 140462291654400 logging_writer.py:48] [23000] global_step=23000, grad_norm=12.755589485168457, loss=1.4981820583343506 +I0831 11:09:10.831646 140462207792896 logging_writer.py:48] [23100] global_step=23100, grad_norm=12.544475555419922, loss=1.2978429794311523 +I0831 11:09:37.951270 140462291654400 logging_writer.py:48] [23200] global_step=23200, grad_norm=12.887504577636719, loss=1.4520010948181152 +I0831 11:10:05.064209 140462207792896 logging_writer.py:48] [23300] global_step=23300, grad_norm=12.713252067565918, loss=1.4144994020462036 +I0831 11:10:32.254339 140462291654400 logging_writer.py:48] [23400] global_step=23400, grad_norm=12.721389770507812, loss=1.5767252445220947 +I0831 11:10:59.423676 140462207792896 logging_writer.py:48] [23500] global_step=23500, grad_norm=13.253745079040527, loss=1.5312401056289673 +I0831 11:11:26.550789 140462291654400 logging_writer.py:48] [23600] global_step=23600, grad_norm=12.978544235229492, loss=1.5911860466003418 +I0831 11:11:53.972831 140462207792896 logging_writer.py:48] [23700] global_step=23700, grad_norm=12.83498764038086, loss=1.4739487171173096 +I0831 11:12:21.104138 140462291654400 logging_writer.py:48] [23800] global_step=23800, grad_norm=12.575363159179688, loss=1.4737350940704346 +I0831 11:12:48.227841 140462207792896 logging_writer.py:48] [23900] global_step=23900, grad_norm=12.437125205993652, loss=1.4368994235992432 +I0831 11:13:15.453807 140462291654400 logging_writer.py:48] [24000] global_step=24000, grad_norm=13.498224258422852, loss=1.483588695526123 +I0831 11:13:42.590928 140462207792896 logging_writer.py:48] [24100] global_step=24100, grad_norm=12.053511619567871, loss=1.5489437580108643 +I0831 11:14:09.721834 140462291654400 logging_writer.py:48] [24200] global_step=24200, grad_norm=12.17817211151123, loss=1.4189302921295166 +I0831 11:14:36.933991 140462207792896 logging_writer.py:48] [24300] global_step=24300, grad_norm=11.979373931884766, loss=1.4417623281478882 +I0831 11:15:04.053517 140462291654400 logging_writer.py:48] [24400] global_step=24400, grad_norm=11.905501365661621, loss=1.4199943542480469 +I0831 11:15:31.190570 140462207792896 logging_writer.py:48] [24500] global_step=24500, grad_norm=12.559513092041016, loss=1.4650168418884277 +I0831 11:15:58.387575 140462291654400 logging_writer.py:48] [24600] global_step=24600, grad_norm=11.29023551940918, loss=1.4878263473510742 +I0831 11:16:25.542918 140462207792896 logging_writer.py:48] [24700] global_step=24700, grad_norm=11.05226993560791, loss=1.408846378326416 +I0831 11:16:52.888616 140462291654400 logging_writer.py:48] [24800] global_step=24800, grad_norm=11.937566757202148, loss=1.4488484859466553 +I0831 11:17:20.094238 140462207792896 logging_writer.py:48] [24900] global_step=24900, grad_norm=11.518203735351562, loss=1.5162503719329834 +I0831 11:17:47.226829 140462291654400 logging_writer.py:48] [25000] global_step=25000, grad_norm=10.729233741760254, loss=1.3275067806243896 +I0831 11:18:14.360632 140462207792896 logging_writer.py:48] [25100] global_step=25100, grad_norm=11.072562217712402, loss=1.3683048486709595 +I0831 11:18:41.584938 140462291654400 logging_writer.py:48] [25200] global_step=25200, grad_norm=11.189452171325684, loss=1.343193531036377 +I0831 11:19:08.741148 140462207792896 logging_writer.py:48] [25300] global_step=25300, grad_norm=9.992819786071777, loss=1.3159159421920776 +I0831 11:19:35.902997 140462291654400 logging_writer.py:48] [25400] global_step=25400, grad_norm=10.870250701904297, loss=1.5098881721496582 +I0831 11:20:03.096642 140462207792896 logging_writer.py:48] [25500] global_step=25500, grad_norm=10.800687789916992, loss=1.3990739583969116 +I0831 11:20:30.232256 140462291654400 logging_writer.py:48] [25600] global_step=25600, grad_norm=10.095355987548828, loss=1.4783849716186523 +I0831 11:20:57.365722 140462207792896 logging_writer.py:48] [25700] global_step=25700, grad_norm=10.299860000610352, loss=1.389349341392517 +I0831 11:21:24.550699 140462291654400 logging_writer.py:48] [25800] global_step=25800, grad_norm=10.027989387512207, loss=1.5135093927383423 +I0831 11:21:51.907984 140462207792896 logging_writer.py:48] [25900] global_step=25900, grad_norm=9.517210960388184, loss=1.367726445198059 +I0831 11:22:19.034924 140462291654400 logging_writer.py:48] [26000] global_step=26000, grad_norm=9.319551467895508, loss=1.381082534790039 +I0831 11:22:46.253885 140462207792896 logging_writer.py:48] [26100] global_step=26100, grad_norm=9.650248527526855, loss=1.5731110572814941 +I0831 11:23:13.384060 140462291654400 logging_writer.py:48] [26200] global_step=26200, grad_norm=9.121173858642578, loss=1.4309334754943848 +I0831 11:23:40.547261 140462207792896 logging_writer.py:48] [26300] global_step=26300, grad_norm=9.441349983215332, loss=1.463244080543518 +I0831 11:24:07.749104 140462291654400 logging_writer.py:48] [26400] global_step=26400, grad_norm=9.39389705657959, loss=1.4581501483917236 +I0831 11:24:34.878141 140462207792896 logging_writer.py:48] [26500] global_step=26500, grad_norm=8.666021347045898, loss=1.464780330657959 +I0831 11:25:01.998836 140462291654400 logging_writer.py:48] [26600] global_step=26600, grad_norm=8.975469589233398, loss=1.4505524635314941 +I0831 11:25:29.200004 140462207792896 logging_writer.py:48] [26700] global_step=26700, grad_norm=8.526373863220215, loss=1.338141679763794 +I0831 11:25:56.341402 140462291654400 logging_writer.py:48] [26800] global_step=26800, grad_norm=8.272428512573242, loss=1.4030585289001465 +I0831 11:26:23.689275 140462207792896 logging_writer.py:48] [26900] global_step=26900, grad_norm=8.31014347076416, loss=1.3618639707565308 +I0831 11:26:50.882823 140462291654400 logging_writer.py:48] [27000] global_step=27000, grad_norm=8.155004501342773, loss=1.5082733631134033 +I0831 11:27:18.029777 140462207792896 logging_writer.py:48] [27100] global_step=27100, grad_norm=7.714046001434326, loss=1.3244388103485107 +I0831 11:27:45.188159 140462291654400 logging_writer.py:48] [27200] global_step=27200, grad_norm=8.172981262207031, loss=1.3869025707244873 +I0831 11:28:12.389500 140462207792896 logging_writer.py:48] [27300] global_step=27300, grad_norm=8.20667839050293, loss=1.3995213508605957 +I0831 11:28:39.509066 140462291654400 logging_writer.py:48] [27400] global_step=27400, grad_norm=7.779822826385498, loss=1.4149588346481323 +I0831 11:29:06.634688 140462207792896 logging_writer.py:48] [27500] global_step=27500, grad_norm=7.360105991363525, loss=1.3499209880828857 +I0831 11:29:33.834489 140462291654400 logging_writer.py:48] [27600] global_step=27600, grad_norm=7.7369866371154785, loss=1.4029148817062378 +I0831 11:30:00.984694 140462207792896 logging_writer.py:48] [27700] global_step=27700, grad_norm=7.408384323120117, loss=1.4167379140853882 +I0831 11:30:28.134196 140462291654400 logging_writer.py:48] [27800] global_step=27800, grad_norm=7.556114673614502, loss=1.4562536478042603 +I0831 11:30:55.343249 140462207792896 logging_writer.py:48] [27900] global_step=27900, grad_norm=7.5156660079956055, loss=1.406574010848999 +I0831 11:31:22.734411 140462291654400 logging_writer.py:48] [28000] global_step=28000, grad_norm=7.731326580047607, loss=1.4265847206115723 +I0831 11:31:49.868740 140462207792896 logging_writer.py:48] [28100] global_step=28100, grad_norm=7.687921524047852, loss=1.452186942100525 +I0831 11:32:17.083545 140462291654400 logging_writer.py:48] [28200] global_step=28200, grad_norm=7.403524875640869, loss=1.2904845476150513 +I0831 11:32:44.205160 140462207792896 logging_writer.py:48] [28300] global_step=28300, grad_norm=7.246540069580078, loss=1.4337832927703857 +I0831 11:33:11.317532 140462291654400 logging_writer.py:48] [28400] global_step=28400, grad_norm=7.029806137084961, loss=1.3245986700057983 +I0831 11:33:38.529500 140462207792896 logging_writer.py:48] [28500] global_step=28500, grad_norm=6.706068515777588, loss=1.3703325986862183 +I0831 11:34:05.678953 140462291654400 logging_writer.py:48] [28600] global_step=28600, grad_norm=6.515596389770508, loss=1.3315939903259277 +I0831 11:34:05.848270 140659750036672 spec.py:333] Evaluating on the training split. +I0831 11:34:17.394324 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 11:34:26.400084 140659750036672 spec.py:363] Evaluating on the test split. +I0831 11:34:27.278815 140659750036672 submission_runner.py:516] Time since start: 8188.45s, Step: 28602, {'train/accuracy': Array(0.83296794, dtype=float32), 'train/loss': Array(0.6281649, dtype=float32), 'validation/accuracy': Array(0.7123, dtype=float32), 'validation/loss': Array(1.1660333, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.5809, dtype=float32), 'test/loss': Array(1.8867964, dtype=float32), 'test/num_examples': 10000, 'score': 8037.070692062378, 'total_duration': 8188.448412895203, 'accumulated_submission_time': 8037.070692062378, 'accumulated_eval_time': 150.83281111717224, 'accumulated_logging_time': 0.3011040687561035} +I0831 11:34:27.326356 140462207792896 logging_writer.py:48] [28602] accumulated_eval_time=150.833, accumulated_logging_time=0.301104, accumulated_submission_time=8037.07, global_step=28602, preemption_count=0, score=8037.07, test/accuracy=0.5809000134468079, test/loss=1.8867963552474976, test/num_examples=10000, total_duration=8188.45, train/accuracy=0.8329679369926453, train/loss=0.6281648874282837, validation/accuracy=0.7123000025749207, validation/loss=1.1660332679748535, validation/num_examples=50000 +I0831 11:34:54.306293 140462291654400 logging_writer.py:48] [28700] global_step=28700, grad_norm=6.609522819519043, loss=1.2933446168899536 +I0831 11:35:21.491109 140462207792896 logging_writer.py:48] [28800] global_step=28800, grad_norm=6.837456703186035, loss=1.4635859727859497 +I0831 11:35:48.637130 140462291654400 logging_writer.py:48] [28900] global_step=28900, grad_norm=6.472423076629639, loss=1.373159646987915 +I0831 11:36:15.758714 140462207792896 logging_writer.py:48] [29000] global_step=29000, grad_norm=6.2627692222595215, loss=1.3141703605651855 +I0831 11:36:43.193438 140462291654400 logging_writer.py:48] [29100] global_step=29100, grad_norm=6.40470552444458, loss=1.360843539237976 +I0831 11:37:10.344481 140462207792896 logging_writer.py:48] [29200] global_step=29200, grad_norm=6.335663318634033, loss=1.2863843441009521 +I0831 11:37:37.467824 140462291654400 logging_writer.py:48] [29300] global_step=29300, grad_norm=6.271585941314697, loss=1.3671941757202148 +I0831 11:38:04.665549 140462207792896 logging_writer.py:48] [29400] global_step=29400, grad_norm=6.345093250274658, loss=1.441563367843628 +I0831 11:38:31.808360 140462291654400 logging_writer.py:48] [29500] global_step=29500, grad_norm=6.163066864013672, loss=1.3192434310913086 +I0831 11:38:58.920741 140462207792896 logging_writer.py:48] [29600] global_step=29600, grad_norm=6.090878963470459, loss=1.3199868202209473 +I0831 11:39:26.130238 140462291654400 logging_writer.py:48] [29700] global_step=29700, grad_norm=6.361289978027344, loss=1.400546669960022 +I0831 11:39:53.264173 140462207792896 logging_writer.py:48] [29800] global_step=29800, grad_norm=5.969298362731934, loss=1.355963110923767 +I0831 11:40:20.400139 140462291654400 logging_writer.py:48] [29900] global_step=29900, grad_norm=5.802892208099365, loss=1.354963779449463 +I0831 11:40:47.610675 140462207792896 logging_writer.py:48] [30000] global_step=30000, grad_norm=6.323470592498779, loss=1.4635627269744873 +I0831 11:41:14.950966 140462291654400 logging_writer.py:48] [30100] global_step=30100, grad_norm=5.541538238525391, loss=1.2727761268615723 +I0831 11:41:42.067865 140462207792896 logging_writer.py:48] [30200] global_step=30200, grad_norm=5.848811626434326, loss=1.3879365921020508 +I0831 11:42:09.249580 140462291654400 logging_writer.py:48] [30300] global_step=30300, grad_norm=5.698920249938965, loss=1.3646045923233032 +I0831 11:42:36.398546 140462207792896 logging_writer.py:48] [30400] global_step=30400, grad_norm=5.474451065063477, loss=1.2419759035110474 +I0831 11:43:03.567694 140462291654400 logging_writer.py:48] [30500] global_step=30500, grad_norm=5.702555179595947, loss=1.2867686748504639 +I0831 11:43:30.759437 140462207792896 logging_writer.py:48] [30600] global_step=30600, grad_norm=5.694380283355713, loss=1.3768789768218994 +I0831 11:43:57.902389 140462291654400 logging_writer.py:48] [30700] global_step=30700, grad_norm=5.675682544708252, loss=1.3309699296951294 +I0831 11:44:25.036145 140462207792896 logging_writer.py:48] [30800] global_step=30800, grad_norm=5.34798526763916, loss=1.2753090858459473 +I0831 11:44:52.224778 140462291654400 logging_writer.py:48] [30900] global_step=30900, grad_norm=5.702296257019043, loss=1.4533826112747192 +I0831 11:45:19.327594 140462207792896 logging_writer.py:48] [31000] global_step=31000, grad_norm=5.419517993927002, loss=1.294588327407837 +I0831 11:45:46.456686 140462291654400 logging_writer.py:48] [31100] global_step=31100, grad_norm=5.433119773864746, loss=1.4330167770385742 +I0831 11:46:13.845450 140462207792896 logging_writer.py:48] [31200] global_step=31200, grad_norm=5.135880947113037, loss=1.3049097061157227 +I0831 11:46:40.972644 140462291654400 logging_writer.py:48] [31300] global_step=31300, grad_norm=5.267536640167236, loss=1.308225154876709 +I0831 11:47:08.091219 140462207792896 logging_writer.py:48] [31400] global_step=31400, grad_norm=5.228147506713867, loss=1.3142257928848267 +I0831 11:47:35.274390 140462291654400 logging_writer.py:48] [31500] global_step=31500, grad_norm=5.214102268218994, loss=1.3637198209762573 +I0831 11:48:02.418422 140462207792896 logging_writer.py:48] [31600] global_step=31600, grad_norm=5.2773118019104, loss=1.2595891952514648 +I0831 11:48:29.538743 140462291654400 logging_writer.py:48] [31700] global_step=31700, grad_norm=5.086458683013916, loss=1.2794311046600342 +I0831 11:48:56.744116 140462207792896 logging_writer.py:48] [31800] global_step=31800, grad_norm=5.350872993469238, loss=1.273229718208313 +I0831 11:49:23.866777 140462291654400 logging_writer.py:48] [31900] global_step=31900, grad_norm=4.818657875061035, loss=1.1937506198883057 +I0831 11:49:51.011577 140462207792896 logging_writer.py:48] [32000] global_step=32000, grad_norm=5.33039665222168, loss=1.4084193706512451 +I0831 11:50:18.204578 140462291654400 logging_writer.py:48] [32100] global_step=32100, grad_norm=5.446366310119629, loss=1.326916217803955 +I0831 11:50:45.552695 140462207792896 logging_writer.py:48] [32200] global_step=32200, grad_norm=5.165380954742432, loss=1.3364017009735107 +I0831 11:51:12.692987 140462291654400 logging_writer.py:48] [32300] global_step=32300, grad_norm=5.024208068847656, loss=1.302327036857605 +I0831 11:51:39.897646 140462207792896 logging_writer.py:48] [32400] global_step=32400, grad_norm=4.9382643699646, loss=1.3068780899047852 +I0831 11:52:07.045990 140462291654400 logging_writer.py:48] [32500] global_step=32500, grad_norm=5.216855049133301, loss=1.4640514850616455 +I0831 11:52:34.188132 140462207792896 logging_writer.py:48] [32600] global_step=32600, grad_norm=4.882631778717041, loss=1.263021469116211 +I0831 11:53:01.372406 140462291654400 logging_writer.py:48] [32700] global_step=32700, grad_norm=5.153981685638428, loss=1.4195183515548706 +I0831 11:53:28.539622 140462207792896 logging_writer.py:48] [32800] global_step=32800, grad_norm=4.884887218475342, loss=1.3436219692230225 +I0831 11:53:55.673669 140462291654400 logging_writer.py:48] [32900] global_step=32900, grad_norm=4.6607279777526855, loss=1.2893708944320679 +I0831 11:54:22.882302 140462207792896 logging_writer.py:48] [33000] global_step=33000, grad_norm=4.893924713134766, loss=1.3249742984771729 +I0831 11:54:50.020302 140462291654400 logging_writer.py:48] [33100] global_step=33100, grad_norm=4.614312648773193, loss=1.2655611038208008 +I0831 11:55:17.132606 140462207792896 logging_writer.py:48] [33200] global_step=33200, grad_norm=4.547956466674805, loss=1.2842888832092285 +I0831 11:55:44.545397 140462291654400 logging_writer.py:48] [33300] global_step=33300, grad_norm=4.891819477081299, loss=1.321599006652832 +I0831 11:56:11.687173 140462207792896 logging_writer.py:48] [33400] global_step=33400, grad_norm=4.566592693328857, loss=1.231497049331665 +I0831 11:56:38.805274 140462291654400 logging_writer.py:48] [33500] global_step=33500, grad_norm=4.757067680358887, loss=1.2435661554336548 +I0831 11:57:05.991200 140462207792896 logging_writer.py:48] [33600] global_step=33600, grad_norm=5.0101637840271, loss=1.3158655166625977 +I0831 11:57:33.128551 140462291654400 logging_writer.py:48] [33700] global_step=33700, grad_norm=4.710142135620117, loss=1.2853671312332153 +I0831 11:58:00.250411 140462207792896 logging_writer.py:48] [33800] global_step=33800, grad_norm=4.788321018218994, loss=1.2843751907348633 +I0831 11:58:27.431882 140462291654400 logging_writer.py:48] [33900] global_step=33900, grad_norm=4.7062177658081055, loss=1.3888301849365234 +I0831 11:58:54.582756 140462207792896 logging_writer.py:48] [34000] global_step=34000, grad_norm=4.51085090637207, loss=1.2920514345169067 +I0831 11:59:21.742166 140462291654400 logging_writer.py:48] [34100] global_step=34100, grad_norm=4.569221019744873, loss=1.2388501167297363 +I0831 11:59:48.935935 140462207792896 logging_writer.py:48] [34200] global_step=34200, grad_norm=4.682833194732666, loss=1.2989847660064697 +I0831 12:00:16.254518 140462291654400 logging_writer.py:48] [34300] global_step=34300, grad_norm=4.687042713165283, loss=1.3438371419906616 +I0831 12:00:43.541393 140462207792896 logging_writer.py:48] [34400] global_step=34400, grad_norm=4.425777912139893, loss=1.2989332675933838 +I0831 12:01:10.768105 140462291654400 logging_writer.py:48] [34500] global_step=34500, grad_norm=4.410975456237793, loss=1.313439965248108 +I0831 12:01:37.917951 140462207792896 logging_writer.py:48] [34600] global_step=34600, grad_norm=4.774184226989746, loss=1.2852375507354736 +I0831 12:02:05.057895 140462291654400 logging_writer.py:48] [34700] global_step=34700, grad_norm=4.337837219238281, loss=1.3129221200942993 +I0831 12:02:32.258621 140462207792896 logging_writer.py:48] [34800] global_step=34800, grad_norm=4.678781986236572, loss=1.3350584506988525 +I0831 12:02:59.414223 140462291654400 logging_writer.py:48] [34900] global_step=34900, grad_norm=4.368221759796143, loss=1.261006236076355 +I0831 12:03:26.551759 140462207792896 logging_writer.py:48] [35000] global_step=35000, grad_norm=4.6517229080200195, loss=1.2632654905319214 +I0831 12:03:53.754889 140462291654400 logging_writer.py:48] [35100] global_step=35100, grad_norm=4.720946788787842, loss=1.3842995166778564 +I0831 12:04:20.903725 140462207792896 logging_writer.py:48] [35200] global_step=35200, grad_norm=4.3485918045043945, loss=1.315881371498108 +I0831 12:04:48.059821 140462291654400 logging_writer.py:48] [35300] global_step=35300, grad_norm=4.505282878875732, loss=1.296027660369873 +I0831 12:05:15.525852 140462207792896 logging_writer.py:48] [35400] global_step=35400, grad_norm=4.255302429199219, loss=1.2805774211883545 +I0831 12:05:42.687332 140462291654400 logging_writer.py:48] [35500] global_step=35500, grad_norm=4.457056045532227, loss=1.278364896774292 +I0831 12:06:09.827774 140462207792896 logging_writer.py:48] [35600] global_step=35600, grad_norm=4.086075305938721, loss=1.2384512424468994 +I0831 12:06:37.043465 140462291654400 logging_writer.py:48] [35700] global_step=35700, grad_norm=4.1898040771484375, loss=1.2515857219696045 +I0831 12:07:04.180288 140462207792896 logging_writer.py:48] [35800] global_step=35800, grad_norm=4.337493419647217, loss=1.2515852451324463 +I0831 12:07:31.311478 140462291654400 logging_writer.py:48] [35900] global_step=35900, grad_norm=4.218838691711426, loss=1.2193708419799805 +I0831 12:07:43.405236 140659750036672 spec.py:333] Evaluating on the training split. +I0831 12:07:54.518574 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 12:08:04.119040 140659750036672 spec.py:363] Evaluating on the test split. +I0831 12:08:04.994189 140659750036672 submission_runner.py:516] Time since start: 10206.17s, Step: 35946, {'train/accuracy': Array(0.85487086, dtype=float32), 'train/loss': Array(0.53780156, dtype=float32), 'validation/accuracy': Array(0.72341996, dtype=float32), 'validation/loss': Array(1.1221741, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.59190005, dtype=float32), 'test/loss': Array(1.8422476, dtype=float32), 'test/num_examples': 10000, 'score': 10033.056078910828, 'total_duration': 10206.169721841812, 'accumulated_submission_time': 10033.056078910828, 'accumulated_eval_time': 172.41977453231812, 'accumulated_logging_time': 0.3790128231048584} +I0831 12:08:05.048361 140462207792896 logging_writer.py:48] [35946] accumulated_eval_time=172.42, accumulated_logging_time=0.379013, accumulated_submission_time=10033.1, global_step=35946, preemption_count=0, score=10033.1, test/accuracy=0.5919000506401062, test/loss=1.8422476053237915, test/num_examples=10000, total_duration=10206.2, train/accuracy=0.8548708558082581, train/loss=0.5378015637397766, validation/accuracy=0.7234199643135071, validation/loss=1.1221741437911987, validation/num_examples=50000 +I0831 12:08:20.153981 140462291654400 logging_writer.py:48] [36000] global_step=36000, grad_norm=4.1046929359436035, loss=1.2239006757736206 +I0831 12:08:47.286841 140462207792896 logging_writer.py:48] [36100] global_step=36100, grad_norm=4.283592224121094, loss=1.292374610900879 +I0831 12:09:14.445010 140462291654400 logging_writer.py:48] [36200] global_step=36200, grad_norm=4.189538955688477, loss=1.2688446044921875 +I0831 12:09:41.681580 140462207792896 logging_writer.py:48] [36300] global_step=36300, grad_norm=4.159573554992676, loss=1.3402104377746582 +I0831 12:10:09.035661 140462291654400 logging_writer.py:48] [36400] global_step=36400, grad_norm=4.1007914543151855, loss=1.2952494621276855 +I0831 12:10:36.171667 140462207792896 logging_writer.py:48] [36500] global_step=36500, grad_norm=4.732675552368164, loss=1.319555640220642 +I0831 12:11:03.361141 140462291654400 logging_writer.py:48] [36600] global_step=36600, grad_norm=4.217029571533203, loss=1.3168063163757324 +I0831 12:11:30.479653 140462207792896 logging_writer.py:48] [36700] global_step=36700, grad_norm=3.96195125579834, loss=1.2630605697631836 +I0831 12:11:57.660053 140462291654400 logging_writer.py:48] [36800] global_step=36800, grad_norm=4.222769260406494, loss=1.2385406494140625 +I0831 12:12:24.873904 140462207792896 logging_writer.py:48] [36900] global_step=36900, grad_norm=4.170384407043457, loss=1.2850862741470337 +I0831 12:12:52.000686 140462291654400 logging_writer.py:48] [37000] global_step=37000, grad_norm=4.285221099853516, loss=1.2758548259735107 +I0831 12:13:19.131179 140462207792896 logging_writer.py:48] [37100] global_step=37100, grad_norm=4.244308948516846, loss=1.2764737606048584 +I0831 12:13:46.357960 140462291654400 logging_writer.py:48] [37200] global_step=37200, grad_norm=4.338650226593018, loss=1.217721939086914 +I0831 12:14:13.512223 140462207792896 logging_writer.py:48] [37300] global_step=37300, grad_norm=4.207889556884766, loss=1.3382354974746704 +I0831 12:14:40.661645 140462291654400 logging_writer.py:48] [37400] global_step=37400, grad_norm=4.063713550567627, loss=1.2689188718795776 +I0831 12:15:08.135866 140462207792896 logging_writer.py:48] [37500] global_step=37500, grad_norm=4.344985485076904, loss=1.372358798980713 +I0831 12:15:35.273574 140462291654400 logging_writer.py:48] [37600] global_step=37600, grad_norm=4.263857841491699, loss=1.2092657089233398 +I0831 12:16:02.386687 140462207792896 logging_writer.py:48] [37700] global_step=37700, grad_norm=4.013289451599121, loss=1.2246111631393433 +I0831 12:16:29.580536 140462291654400 logging_writer.py:48] [37800] global_step=37800, grad_norm=4.093195915222168, loss=1.2537506818771362 +I0831 12:16:56.743996 140462207792896 logging_writer.py:48] [37900] global_step=37900, grad_norm=4.0982818603515625, loss=1.2758581638336182 +I0831 12:17:23.895006 140462291654400 logging_writer.py:48] [38000] global_step=38000, grad_norm=4.27721643447876, loss=1.3319034576416016 +I0831 12:17:51.119284 140462207792896 logging_writer.py:48] [38100] global_step=38100, grad_norm=4.382910251617432, loss=1.323509931564331 +I0831 12:18:18.254228 140462291654400 logging_writer.py:48] [38200] global_step=38200, grad_norm=4.077047824859619, loss=1.2172138690948486 +I0831 12:18:45.393955 140462207792896 logging_writer.py:48] [38300] global_step=38300, grad_norm=4.0806732177734375, loss=1.2528972625732422 +I0831 12:19:12.581680 140462291654400 logging_writer.py:48] [38400] global_step=38400, grad_norm=4.235974311828613, loss=1.2324590682983398 +I0831 12:19:39.969316 140462207792896 logging_writer.py:48] [38500] global_step=38500, grad_norm=4.312421798706055, loss=1.2781178951263428 +I0831 12:20:07.158238 140462291654400 logging_writer.py:48] [38600] global_step=38600, grad_norm=3.929267406463623, loss=1.2043635845184326 +I0831 12:20:34.348641 140462207792896 logging_writer.py:48] [38700] global_step=38700, grad_norm=3.8541131019592285, loss=1.1693217754364014 +I0831 12:21:01.491415 140462291654400 logging_writer.py:48] [38800] global_step=38800, grad_norm=4.004918575286865, loss=1.181320309638977 +I0831 12:21:28.612240 140462207792896 logging_writer.py:48] [38900] global_step=38900, grad_norm=4.066263675689697, loss=1.302997350692749 +I0831 12:21:55.829802 140462291654400 logging_writer.py:48] [39000] global_step=39000, grad_norm=4.1554694175720215, loss=1.3178744316101074 +I0831 12:22:22.972139 140462207792896 logging_writer.py:48] [39100] global_step=39100, grad_norm=4.274717330932617, loss=1.2175147533416748 +I0831 12:22:50.120994 140462291654400 logging_writer.py:48] [39200] global_step=39200, grad_norm=4.129603862762451, loss=1.352050542831421 +I0831 12:23:17.311149 140462207792896 logging_writer.py:48] [39300] global_step=39300, grad_norm=3.928443670272827, loss=1.2526464462280273 +I0831 12:23:44.473605 140462291654400 logging_writer.py:48] [39400] global_step=39400, grad_norm=3.9357433319091797, loss=1.2659902572631836 +I0831 12:24:11.607980 140462207792896 logging_writer.py:48] [39500] global_step=39500, grad_norm=3.9960148334503174, loss=1.3023923635482788 +I0831 12:24:39.034921 140462291654400 logging_writer.py:48] [39600] global_step=39600, grad_norm=4.07056188583374, loss=1.2092084884643555 +I0831 12:25:06.178774 140462207792896 logging_writer.py:48] [39700] global_step=39700, grad_norm=4.337121963500977, loss=1.3312549591064453 +I0831 12:25:33.318089 140462291654400 logging_writer.py:48] [39800] global_step=39800, grad_norm=3.902848720550537, loss=1.24507474899292 +I0831 12:26:00.498645 140462207792896 logging_writer.py:48] [39900] global_step=39900, grad_norm=3.906527042388916, loss=1.1948095560073853 +I0831 12:26:27.627426 140462291654400 logging_writer.py:48] [40000] global_step=40000, grad_norm=3.6361193656921387, loss=1.1093918085098267 +I0831 12:26:54.745565 140462207792896 logging_writer.py:48] [40100] global_step=40100, grad_norm=3.960379123687744, loss=1.1843857765197754 +I0831 12:27:21.961767 140462291654400 logging_writer.py:48] [40200] global_step=40200, grad_norm=3.8957231044769287, loss=1.2545894384384155 +I0831 12:27:49.081785 140462207792896 logging_writer.py:48] [40300] global_step=40300, grad_norm=4.0353851318359375, loss=1.3614407777786255 +I0831 12:28:16.219094 140462291654400 logging_writer.py:48] [40400] global_step=40400, grad_norm=4.304354190826416, loss=1.333116054534912 +I0831 12:28:43.406469 140462207792896 logging_writer.py:48] [40500] global_step=40500, grad_norm=4.085301399230957, loss=1.2582781314849854 +I0831 12:29:10.719668 140462291654400 logging_writer.py:48] [40600] global_step=40600, grad_norm=4.002737522125244, loss=1.2043240070343018 +I0831 12:29:37.842678 140462207792896 logging_writer.py:48] [40700] global_step=40700, grad_norm=4.176192760467529, loss=1.297956943511963 +I0831 12:30:05.058714 140462291654400 logging_writer.py:48] [40800] global_step=40800, grad_norm=3.791134834289551, loss=1.2197328805923462 +I0831 12:30:32.188993 140462207792896 logging_writer.py:48] [40900] global_step=40900, grad_norm=4.095873832702637, loss=1.3858033418655396 +I0831 12:30:59.305418 140462291654400 logging_writer.py:48] [41000] global_step=41000, grad_norm=3.7459628582000732, loss=1.1500470638275146 +I0831 12:31:26.518000 140462207792896 logging_writer.py:48] [41100] global_step=41100, grad_norm=3.9234347343444824, loss=1.262179970741272 +I0831 12:31:53.666950 140462291654400 logging_writer.py:48] [41200] global_step=41200, grad_norm=3.82421612739563, loss=1.177004098892212 +I0831 12:32:20.786715 140462207792896 logging_writer.py:48] [41300] global_step=41300, grad_norm=3.976911783218384, loss=1.2499884366989136 +I0831 12:32:47.977241 140462291654400 logging_writer.py:48] [41400] global_step=41400, grad_norm=4.104262828826904, loss=1.2586705684661865 +I0831 12:33:15.104129 140462207792896 logging_writer.py:48] [41500] global_step=41500, grad_norm=3.891371726989746, loss=1.3556936979293823 +I0831 12:33:42.251701 140462291654400 logging_writer.py:48] [41600] global_step=41600, grad_norm=3.9457223415374756, loss=1.228129267692566 +I0831 12:34:09.694822 140462207792896 logging_writer.py:48] [41700] global_step=41700, grad_norm=4.019747257232666, loss=1.3081371784210205 +I0831 12:34:36.824691 140462291654400 logging_writer.py:48] [41800] global_step=41800, grad_norm=4.028475761413574, loss=1.3078804016113281 +I0831 12:35:03.941698 140462207792896 logging_writer.py:48] [41900] global_step=41900, grad_norm=3.8546721935272217, loss=1.1664012670516968 +I0831 12:35:31.131933 140462291654400 logging_writer.py:48] [42000] global_step=42000, grad_norm=3.8412907123565674, loss=1.258398175239563 +I0831 12:35:58.271073 140462207792896 logging_writer.py:48] [42100] global_step=42100, grad_norm=3.66617751121521, loss=1.161614179611206 +I0831 12:36:25.402373 140462291654400 logging_writer.py:48] [42200] global_step=42200, grad_norm=3.7477734088897705, loss=1.1960692405700684 +I0831 12:36:52.610644 140462207792896 logging_writer.py:48] [42300] global_step=42300, grad_norm=4.1291422843933105, loss=1.2835004329681396 +I0831 12:37:19.773842 140462291654400 logging_writer.py:48] [42400] global_step=42400, grad_norm=4.244822978973389, loss=1.3108934164047241 +I0831 12:37:46.922426 140462207792896 logging_writer.py:48] [42500] global_step=42500, grad_norm=3.869791030883789, loss=1.1901062726974487 +I0831 12:38:14.123398 140462291654400 logging_writer.py:48] [42600] global_step=42600, grad_norm=4.1261420249938965, loss=1.2717546224594116 +I0831 12:38:41.487538 140462207792896 logging_writer.py:48] [42700] global_step=42700, grad_norm=4.349362373352051, loss=1.3262256383895874 +I0831 12:39:08.625935 140462291654400 logging_writer.py:48] [42800] global_step=42800, grad_norm=3.695712089538574, loss=1.1838074922561646 +I0831 12:39:35.817400 140462207792896 logging_writer.py:48] [42900] global_step=42900, grad_norm=4.0939555168151855, loss=1.2089738845825195 +I0831 12:40:02.942468 140462291654400 logging_writer.py:48] [43000] global_step=43000, grad_norm=3.9512789249420166, loss=1.2504940032958984 +I0831 12:40:30.069161 140462207792896 logging_writer.py:48] [43100] global_step=43100, grad_norm=3.93393874168396, loss=1.1875637769699097 +I0831 12:40:57.285878 140462291654400 logging_writer.py:48] [43200] global_step=43200, grad_norm=3.7739977836608887, loss=1.166163682937622 +I0831 12:41:20.999953 140659750036672 spec.py:333] Evaluating on the training split. +I0831 12:41:30.905028 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 12:41:40.073933 140659750036672 spec.py:363] Evaluating on the test split. +I0831 12:41:40.955981 140659750036672 submission_runner.py:516] Time since start: 12222.13s, Step: 43289, {'train/accuracy': Array(0.87282765, dtype=float32), 'train/loss': Array(0.4651624, dtype=float32), 'validation/accuracy': Array(0.73126, dtype=float32), 'validation/loss': Array(1.0990901, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.60010004, dtype=float32), 'test/loss': Array(1.833945, dtype=float32), 'test/num_examples': 10000, 'score': 12028.914838790894, 'total_duration': 12222.131848573685, 'accumulated_submission_time': 12028.914838790894, 'accumulated_eval_time': 192.3741476535797, 'accumulated_logging_time': 0.4674093723297119} +I0831 12:41:41.013443 140462207792896 logging_writer.py:48] [43289] accumulated_eval_time=192.374, accumulated_logging_time=0.467409, accumulated_submission_time=12028.9, global_step=43289, preemption_count=0, score=12028.9, test/accuracy=0.600100040435791, test/loss=1.8339450359344482, test/num_examples=10000, total_duration=12222.1, train/accuracy=0.8728276491165161, train/loss=0.46516239643096924, validation/accuracy=0.7312600016593933, validation/loss=1.0990900993347168, validation/num_examples=50000 +I0831 12:41:44.400986 140462291654400 logging_writer.py:48] [43300] global_step=43300, grad_norm=3.6253390312194824, loss=1.1812093257904053 +I0831 12:42:11.554096 140462207792896 logging_writer.py:48] [43400] global_step=43400, grad_norm=4.058777809143066, loss=1.1621904373168945 +I0831 12:42:38.746244 140462291654400 logging_writer.py:48] [43500] global_step=43500, grad_norm=4.037467002868652, loss=1.2663241624832153 +I0831 12:43:05.918294 140462207792896 logging_writer.py:48] [43600] global_step=43600, grad_norm=4.001245975494385, loss=1.2758774757385254 +I0831 12:43:33.066456 140462291654400 logging_writer.py:48] [43700] global_step=43700, grad_norm=3.906146764755249, loss=1.23658287525177 +I0831 12:44:00.496480 140462207792896 logging_writer.py:48] [43800] global_step=43800, grad_norm=3.8191163539886475, loss=1.2966890335083008 +I0831 12:44:27.641421 140462291654400 logging_writer.py:48] [43900] global_step=43900, grad_norm=3.64502215385437, loss=1.109588861465454 +I0831 12:44:54.775078 140462207792896 logging_writer.py:48] [44000] global_step=44000, grad_norm=4.0554962158203125, loss=1.1799733638763428 +I0831 12:45:22.015402 140462291654400 logging_writer.py:48] [44100] global_step=44100, grad_norm=3.9012956619262695, loss=1.188908338546753 +I0831 12:45:49.134330 140462207792896 logging_writer.py:48] [44200] global_step=44200, grad_norm=3.747408628463745, loss=1.1372754573822021 +I0831 12:46:16.263308 140462291654400 logging_writer.py:48] [44300] global_step=44300, grad_norm=3.854013204574585, loss=1.192842960357666 +I0831 12:46:43.437591 140462207792896 logging_writer.py:48] [44400] global_step=44400, grad_norm=4.079299449920654, loss=1.2989790439605713 +I0831 12:47:10.562702 140462291654400 logging_writer.py:48] [44500] global_step=44500, grad_norm=3.835484266281128, loss=1.3418947458267212 +I0831 12:47:37.694511 140462207792896 logging_writer.py:48] [44600] global_step=44600, grad_norm=3.907440662384033, loss=1.2875480651855469 +I0831 12:48:04.918940 140462291654400 logging_writer.py:48] [44700] global_step=44700, grad_norm=4.0691351890563965, loss=1.1646970510482788 +I0831 12:48:32.068069 140462207792896 logging_writer.py:48] [44800] global_step=44800, grad_norm=3.8508968353271484, loss=1.2805017232894897 +I0831 12:48:59.435215 140462291654400 logging_writer.py:48] [44900] global_step=44900, grad_norm=3.93062162399292, loss=1.2226883172988892 +I0831 12:49:26.629656 140462207792896 logging_writer.py:48] [45000] global_step=45000, grad_norm=4.236686706542969, loss=1.306670069694519 +I0831 12:49:53.780520 140462291654400 logging_writer.py:48] [45100] global_step=45100, grad_norm=3.6765730381011963, loss=1.157848834991455 +I0831 12:50:20.925786 140462207792896 logging_writer.py:48] [45200] global_step=45200, grad_norm=3.6760103702545166, loss=1.1994596719741821 +I0831 12:50:48.113380 140462291654400 logging_writer.py:48] [45300] global_step=45300, grad_norm=3.7086405754089355, loss=1.2196991443634033 +I0831 12:51:15.227947 140462207792896 logging_writer.py:48] [45400] global_step=45400, grad_norm=4.101501941680908, loss=1.341827154159546 +I0831 12:51:42.380791 140462291654400 logging_writer.py:48] [45500] global_step=45500, grad_norm=3.851835012435913, loss=1.1858985424041748 +I0831 12:52:09.571937 140462207792896 logging_writer.py:48] [45600] global_step=45600, grad_norm=3.8446438312530518, loss=1.1981534957885742 +I0831 12:52:36.698479 140462291654400 logging_writer.py:48] [45700] global_step=45700, grad_norm=3.6646218299865723, loss=1.1981711387634277 +I0831 12:53:03.825706 140462207792896 logging_writer.py:48] [45800] global_step=45800, grad_norm=3.730795383453369, loss=1.130879282951355 +I0831 12:53:31.060880 140462291654400 logging_writer.py:48] [45900] global_step=45900, grad_norm=3.5542752742767334, loss=1.2043970823287964 +I0831 12:53:58.458482 140462207792896 logging_writer.py:48] [46000] global_step=46000, grad_norm=3.8079142570495605, loss=1.2134393453598022 +I0831 12:54:25.568154 140462291654400 logging_writer.py:48] [46100] global_step=46100, grad_norm=3.62117075920105, loss=1.095943808555603 +I0831 12:54:52.776518 140462207792896 logging_writer.py:48] [46200] global_step=46200, grad_norm=3.9310619831085205, loss=1.2078192234039307 +I0831 12:55:19.943398 140462291654400 logging_writer.py:48] [46300] global_step=46300, grad_norm=3.7139406204223633, loss=1.1786866188049316 +I0831 12:55:47.070250 140462207792896 logging_writer.py:48] [46400] global_step=46400, grad_norm=4.062290668487549, loss=1.2887263298034668 +I0831 12:56:14.262563 140462291654400 logging_writer.py:48] [46500] global_step=46500, grad_norm=3.787461757659912, loss=1.1654592752456665 +I0831 12:56:41.397842 140462207792896 logging_writer.py:48] [46600] global_step=46600, grad_norm=3.8743479251861572, loss=1.1525882482528687 +I0831 12:57:08.544818 140462291654400 logging_writer.py:48] [46700] global_step=46700, grad_norm=3.623109817504883, loss=1.1800618171691895 +I0831 12:57:35.754392 140462207792896 logging_writer.py:48] [46800] global_step=46800, grad_norm=3.8647525310516357, loss=1.17124605178833 +I0831 12:58:02.883892 140462291654400 logging_writer.py:48] [46900] global_step=46900, grad_norm=4.042476654052734, loss=1.1093672513961792 +I0831 12:58:30.010981 140462207792896 logging_writer.py:48] [47000] global_step=47000, grad_norm=3.878173828125, loss=1.2014050483703613 +I0831 12:58:57.412872 140462291654400 logging_writer.py:48] [47100] global_step=47100, grad_norm=4.211263656616211, loss=1.216605544090271 +I0831 12:59:24.524979 140462207792896 logging_writer.py:48] [47200] global_step=47200, grad_norm=3.659132957458496, loss=1.1114258766174316 +I0831 12:59:51.623124 140462291654400 logging_writer.py:48] [47300] global_step=47300, grad_norm=3.928593158721924, loss=1.1438319683074951 +I0831 13:00:18.810510 140462207792896 logging_writer.py:48] [47400] global_step=47400, grad_norm=4.002999305725098, loss=1.1652374267578125 +I0831 13:00:45.937301 140462291654400 logging_writer.py:48] [47500] global_step=47500, grad_norm=3.961869716644287, loss=1.216312289237976 +I0831 13:01:13.068742 140462207792896 logging_writer.py:48] [47600] global_step=47600, grad_norm=3.8188652992248535, loss=1.2578771114349365 +I0831 13:01:40.262204 140462291654400 logging_writer.py:48] [47700] global_step=47700, grad_norm=4.0717058181762695, loss=1.1941614151000977 +I0831 13:02:07.396923 140462207792896 logging_writer.py:48] [47800] global_step=47800, grad_norm=3.913347005844116, loss=1.2146433591842651 +I0831 13:02:34.510676 140462291654400 logging_writer.py:48] [47900] global_step=47900, grad_norm=3.671271800994873, loss=1.1214309930801392 +I0831 13:03:01.677966 140462207792896 logging_writer.py:48] [48000] global_step=48000, grad_norm=3.6643059253692627, loss=1.172257661819458 +I0831 13:03:29.049028 140462291654400 logging_writer.py:48] [48100] global_step=48100, grad_norm=3.7128520011901855, loss=1.2290780544281006 +I0831 13:03:56.206564 140462207792896 logging_writer.py:48] [48200] global_step=48200, grad_norm=4.160257816314697, loss=1.2574717998504639 +I0831 13:04:23.405280 140462291654400 logging_writer.py:48] [48300] global_step=48300, grad_norm=3.6360316276550293, loss=1.2098990678787231 +I0831 13:04:50.569400 140462207792896 logging_writer.py:48] [48400] global_step=48400, grad_norm=3.9540855884552, loss=1.2472081184387207 +I0831 13:05:17.693935 140462291654400 logging_writer.py:48] [48500] global_step=48500, grad_norm=4.1653852462768555, loss=1.3295023441314697 +I0831 13:05:44.875444 140462207792896 logging_writer.py:48] [48600] global_step=48600, grad_norm=4.023526191711426, loss=1.246119737625122 +I0831 13:06:12.009905 140462291654400 logging_writer.py:48] [48700] global_step=48700, grad_norm=3.616994857788086, loss=1.197047233581543 +I0831 13:06:39.152379 140462207792896 logging_writer.py:48] [48800] global_step=48800, grad_norm=3.6229426860809326, loss=1.2187148332595825 +I0831 13:07:06.336535 140462291654400 logging_writer.py:48] [48900] global_step=48900, grad_norm=3.8367068767547607, loss=1.1772300004959106 +I0831 13:07:33.451739 140462207792896 logging_writer.py:48] [49000] global_step=49000, grad_norm=3.7014567852020264, loss=1.1451342105865479 +I0831 13:08:00.584984 140462291654400 logging_writer.py:48] [49100] global_step=49100, grad_norm=3.9762394428253174, loss=1.3447818756103516 +I0831 13:08:27.997621 140462207792896 logging_writer.py:48] [49200] global_step=49200, grad_norm=3.645780086517334, loss=1.0814201831817627 +I0831 13:08:55.130551 140462291654400 logging_writer.py:48] [49300] global_step=49300, grad_norm=3.771366834640503, loss=1.174557089805603 +I0831 13:09:22.246713 140462207792896 logging_writer.py:48] [49400] global_step=49400, grad_norm=3.806736707687378, loss=1.1065372228622437 +I0831 13:09:49.432970 140462291654400 logging_writer.py:48] [49500] global_step=49500, grad_norm=4.074043273925781, loss=1.2584071159362793 +I0831 13:10:16.554732 140462207792896 logging_writer.py:48] [49600] global_step=49600, grad_norm=3.9678354263305664, loss=1.231448769569397 +I0831 13:10:43.742847 140462291654400 logging_writer.py:48] [49700] global_step=49700, grad_norm=4.109307765960693, loss=1.257702112197876 +I0831 13:11:10.930041 140462207792896 logging_writer.py:48] [49800] global_step=49800, grad_norm=3.712094306945801, loss=1.1337109804153442 +I0831 13:11:38.060147 140462291654400 logging_writer.py:48] [49900] global_step=49900, grad_norm=4.213749408721924, loss=1.1415842771530151 +I0831 13:12:05.189445 140462207792896 logging_writer.py:48] [50000] global_step=50000, grad_norm=3.8202905654907227, loss=1.073469877243042 +I0831 13:12:32.407823 140462291654400 logging_writer.py:48] [50100] global_step=50100, grad_norm=3.8956265449523926, loss=1.2160636186599731 +I0831 13:12:59.784383 140462207792896 logging_writer.py:48] [50200] global_step=50200, grad_norm=4.0497355461120605, loss=1.2040045261383057 +I0831 13:13:26.974769 140462291654400 logging_writer.py:48] [50300] global_step=50300, grad_norm=4.001854419708252, loss=1.1149406433105469 +I0831 13:13:54.184405 140462207792896 logging_writer.py:48] [50400] global_step=50400, grad_norm=3.968867778778076, loss=1.1384732723236084 +I0831 13:14:21.327136 140462291654400 logging_writer.py:48] [50500] global_step=50500, grad_norm=3.6692941188812256, loss=1.1448373794555664 +I0831 13:14:48.439590 140462207792896 logging_writer.py:48] [50600] global_step=50600, grad_norm=3.925882339477539, loss=1.1200003623962402 +I0831 13:14:57.028257 140659750036672 spec.py:333] Evaluating on the training split. +I0831 13:15:07.036708 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 13:15:16.278114 140659750036672 spec.py:363] Evaluating on the test split. +I0831 13:15:17.169634 140659750036672 submission_runner.py:516] Time since start: 14238.35s, Step: 50633, {'train/accuracy': Array(0.88392854, dtype=float32), 'train/loss': Array(0.41763052, dtype=float32), 'validation/accuracy': Array(0.73604, dtype=float32), 'validation/loss': Array(1.089328, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6078, dtype=float32), 'test/loss': Array(1.8267647, dtype=float32), 'test/num_examples': 10000, 'score': 14024.841076850891, 'total_duration': 14238.34527349472, 'accumulated_submission_time': 14024.841076850891, 'accumulated_eval_time': 212.51364135742188, 'accumulated_logging_time': 0.5555286407470703} +I0831 13:15:17.210221 140462291654400 logging_writer.py:48] [50633] accumulated_eval_time=212.514, accumulated_logging_time=0.555529, accumulated_submission_time=14024.8, global_step=50633, preemption_count=0, score=14024.8, test/accuracy=0.6078000068664551, test/loss=1.826764702796936, test/num_examples=10000, total_duration=14238.3, train/accuracy=0.8839285373687744, train/loss=0.41763052344322205, validation/accuracy=0.7360399961471558, validation/loss=1.0893280506134033, validation/num_examples=50000 +I0831 13:15:35.800657 140462207792896 logging_writer.py:48] [50700] global_step=50700, grad_norm=3.7859373092651367, loss=1.1013820171356201 +I0831 13:16:02.944732 140462291654400 logging_writer.py:48] [50800] global_step=50800, grad_norm=3.861665964126587, loss=1.184069275856018 +I0831 13:16:30.062315 140462207792896 logging_writer.py:48] [50900] global_step=50900, grad_norm=3.637481689453125, loss=1.1670653820037842 +I0831 13:16:57.253473 140462291654400 logging_writer.py:48] [51000] global_step=51000, grad_norm=4.088878631591797, loss=1.1804039478302002 +I0831 13:17:24.375604 140462207792896 logging_writer.py:48] [51100] global_step=51100, grad_norm=3.623582601547241, loss=1.0684077739715576 +I0831 13:17:51.545329 140462291654400 logging_writer.py:48] [51200] global_step=51200, grad_norm=3.8701515197753906, loss=1.155937910079956 +I0831 13:18:18.966570 140462207792896 logging_writer.py:48] [51300] global_step=51300, grad_norm=3.8517415523529053, loss=1.146122694015503 +I0831 13:18:46.126277 140462291654400 logging_writer.py:48] [51400] global_step=51400, grad_norm=3.6112308502197266, loss=1.0899128913879395 +I0831 13:19:13.268574 140462207792896 logging_writer.py:48] [51500] global_step=51500, grad_norm=4.017630577087402, loss=1.2323477268218994 +I0831 13:19:40.484544 140462291654400 logging_writer.py:48] [51600] global_step=51600, grad_norm=3.9595561027526855, loss=1.203701138496399 +I0831 13:20:07.630299 140462207792896 logging_writer.py:48] [51700] global_step=51700, grad_norm=3.678962230682373, loss=1.1494066715240479 +I0831 13:20:34.767513 140462291654400 logging_writer.py:48] [51800] global_step=51800, grad_norm=3.947899341583252, loss=1.2273876667022705 +I0831 13:21:01.974481 140462207792896 logging_writer.py:48] [51900] global_step=51900, grad_norm=3.8055789470672607, loss=1.212195634841919 +I0831 13:21:29.110900 140462291654400 logging_writer.py:48] [52000] global_step=52000, grad_norm=3.5643603801727295, loss=1.2276321649551392 +I0831 13:21:56.229179 140462207792896 logging_writer.py:48] [52100] global_step=52100, grad_norm=3.809922933578491, loss=1.1461362838745117 +I0831 13:22:23.406019 140462291654400 logging_writer.py:48] [52200] global_step=52200, grad_norm=3.61582612991333, loss=1.0541980266571045 +I0831 13:22:50.532177 140462207792896 logging_writer.py:48] [52300] global_step=52300, grad_norm=3.946943521499634, loss=1.228779673576355 +I0831 13:23:17.936751 140462291654400 logging_writer.py:48] [52400] global_step=52400, grad_norm=3.953396797180176, loss=1.184752106666565 +I0831 13:23:45.167261 140462207792896 logging_writer.py:48] [52500] global_step=52500, grad_norm=3.629624366760254, loss=1.2045304775238037 +I0831 13:24:12.312368 140462291654400 logging_writer.py:48] [52600] global_step=52600, grad_norm=4.074431896209717, loss=1.1660223007202148 +I0831 13:24:39.430045 140462207792896 logging_writer.py:48] [52700] global_step=52700, grad_norm=3.581594228744507, loss=1.2217650413513184 +I0831 13:25:06.658115 140462291654400 logging_writer.py:48] [52800] global_step=52800, grad_norm=4.290639400482178, loss=1.1988329887390137 +I0831 13:25:33.794027 140462207792896 logging_writer.py:48] [52900] global_step=52900, grad_norm=4.091780662536621, loss=1.190108299255371 +I0831 13:26:00.946307 140462291654400 logging_writer.py:48] [53000] global_step=53000, grad_norm=3.9183273315429688, loss=1.1465978622436523 +I0831 13:26:28.150576 140462207792896 logging_writer.py:48] [53100] global_step=53100, grad_norm=3.669569253921509, loss=1.2074629068374634 +I0831 13:26:55.294342 140462291654400 logging_writer.py:48] [53200] global_step=53200, grad_norm=3.880448818206787, loss=1.1102255582809448 +I0831 13:27:22.425275 140462207792896 logging_writer.py:48] [53300] global_step=53300, grad_norm=3.6548635959625244, loss=1.250501036643982 +I0831 13:27:49.861276 140462291654400 logging_writer.py:48] [53400] global_step=53400, grad_norm=3.8312387466430664, loss=1.1244142055511475 +I0831 13:28:16.988116 140462207792896 logging_writer.py:48] [53500] global_step=53500, grad_norm=3.8517892360687256, loss=1.1199653148651123 +I0831 13:28:44.114305 140462291654400 logging_writer.py:48] [53600] global_step=53600, grad_norm=3.8043463230133057, loss=1.0632200241088867 +I0831 13:29:11.303995 140462207792896 logging_writer.py:48] [53700] global_step=53700, grad_norm=3.8481569290161133, loss=1.1683844327926636 +I0831 13:29:38.430480 140462291654400 logging_writer.py:48] [53800] global_step=53800, grad_norm=3.6657555103302, loss=1.1066498756408691 +I0831 13:30:05.528975 140462207792896 logging_writer.py:48] [53900] global_step=53900, grad_norm=3.606133222579956, loss=1.1348462104797363 +I0831 13:30:32.738471 140462291654400 logging_writer.py:48] [54000] global_step=54000, grad_norm=4.101473808288574, loss=1.1958285570144653 +I0831 13:30:59.894825 140462207792896 logging_writer.py:48] [54100] global_step=54100, grad_norm=3.9226784706115723, loss=1.1608233451843262 +I0831 13:31:27.024165 140462291654400 logging_writer.py:48] [54200] global_step=54200, grad_norm=4.021599292755127, loss=1.2885723114013672 +I0831 13:31:54.201966 140462207792896 logging_writer.py:48] [54300] global_step=54300, grad_norm=3.797149658203125, loss=1.1397860050201416 +I0831 13:32:21.321215 140462291654400 logging_writer.py:48] [54400] global_step=54400, grad_norm=3.9193239212036133, loss=1.19612455368042 +I0831 13:32:48.677033 140462207792896 logging_writer.py:48] [54500] global_step=54500, grad_norm=3.751234769821167, loss=1.181968331336975 +I0831 13:33:15.886493 140462291654400 logging_writer.py:48] [54600] global_step=54600, grad_norm=3.8687455654144287, loss=1.1720805168151855 +I0831 13:33:43.025739 140462207792896 logging_writer.py:48] [54700] global_step=54700, grad_norm=3.635523557662964, loss=1.078225016593933 +I0831 13:34:10.134963 140462291654400 logging_writer.py:48] [54800] global_step=54800, grad_norm=3.785173177719116, loss=1.1184310913085938 +I0831 13:34:37.334315 140462207792896 logging_writer.py:48] [54900] global_step=54900, grad_norm=3.9152042865753174, loss=1.0947823524475098 +I0831 13:35:04.487913 140462291654400 logging_writer.py:48] [55000] global_step=55000, grad_norm=3.9016854763031006, loss=1.118397831916809 +I0831 13:35:31.622522 140462207792896 logging_writer.py:48] [55100] global_step=55100, grad_norm=3.8000311851501465, loss=1.214634895324707 +I0831 13:35:58.820520 140462291654400 logging_writer.py:48] [55200] global_step=55200, grad_norm=4.158570289611816, loss=1.224477767944336 +I0831 13:36:25.966489 140462207792896 logging_writer.py:48] [55300] global_step=55300, grad_norm=3.9574241638183594, loss=1.209652066230774 +I0831 13:36:53.101068 140462291654400 logging_writer.py:48] [55400] global_step=55400, grad_norm=4.181178569793701, loss=1.0855724811553955 +I0831 13:37:20.313002 140462207792896 logging_writer.py:48] [55500] global_step=55500, grad_norm=3.86507248878479, loss=1.2065331935882568 +I0831 13:37:47.677028 140462291654400 logging_writer.py:48] [55600] global_step=55600, grad_norm=4.158779621124268, loss=1.1715162992477417 +I0831 13:38:14.784157 140462207792896 logging_writer.py:48] [55700] global_step=55700, grad_norm=4.096261024475098, loss=1.1864023208618164 +I0831 13:38:41.968697 140462291654400 logging_writer.py:48] [55800] global_step=55800, grad_norm=3.9497909545898438, loss=1.1878294944763184 +I0831 13:39:09.096106 140462207792896 logging_writer.py:48] [55900] global_step=55900, grad_norm=3.8232734203338623, loss=1.1875354051589966 +I0831 13:39:36.219347 140462291654400 logging_writer.py:48] [56000] global_step=56000, grad_norm=4.047987461090088, loss=1.1254318952560425 +I0831 13:40:03.467440 140462207792896 logging_writer.py:48] [56100] global_step=56100, grad_norm=3.96488618850708, loss=1.1574175357818604 +I0831 13:40:30.607915 140462291654400 logging_writer.py:48] [56200] global_step=56200, grad_norm=3.9945008754730225, loss=1.141599416732788 +I0831 13:40:57.780580 140462207792896 logging_writer.py:48] [56300] global_step=56300, grad_norm=3.949859380722046, loss=1.240176796913147 +I0831 13:41:25.015635 140462291654400 logging_writer.py:48] [56400] global_step=56400, grad_norm=3.9797627925872803, loss=1.0629929304122925 +I0831 13:41:52.168318 140462207792896 logging_writer.py:48] [56500] global_step=56500, grad_norm=4.032841682434082, loss=1.199572205543518 +I0831 13:42:19.298352 140462291654400 logging_writer.py:48] [56600] global_step=56600, grad_norm=4.233610153198242, loss=1.201412320137024 +I0831 13:42:46.751338 140462207792896 logging_writer.py:48] [56700] global_step=56700, grad_norm=4.130585670471191, loss=1.1636669635772705 +I0831 13:43:13.859402 140462291654400 logging_writer.py:48] [56800] global_step=56800, grad_norm=3.7270896434783936, loss=1.1012301445007324 +I0831 13:43:40.963746 140462207792896 logging_writer.py:48] [56900] global_step=56900, grad_norm=4.125523567199707, loss=1.188875436782837 +I0831 13:44:08.142246 140462291654400 logging_writer.py:48] [57000] global_step=57000, grad_norm=4.32472038269043, loss=1.262582540512085 +I0831 13:44:35.261532 140462207792896 logging_writer.py:48] [57100] global_step=57100, grad_norm=4.058272361755371, loss=1.1648154258728027 +I0831 13:45:02.379279 140462291654400 logging_writer.py:48] [57200] global_step=57200, grad_norm=3.9597201347351074, loss=1.2015842199325562 +I0831 13:45:29.558244 140462207792896 logging_writer.py:48] [57300] global_step=57300, grad_norm=3.9307472705841064, loss=1.1039321422576904 +I0831 13:45:56.709715 140462291654400 logging_writer.py:48] [57400] global_step=57400, grad_norm=4.059325218200684, loss=1.119604229927063 +I0831 13:46:23.844132 140462207792896 logging_writer.py:48] [57500] global_step=57500, grad_norm=3.867258071899414, loss=1.23659086227417 +I0831 13:46:51.016087 140462291654400 logging_writer.py:48] [57600] global_step=57600, grad_norm=3.712080478668213, loss=1.0713375806808472 +I0831 13:47:18.369969 140462207792896 logging_writer.py:48] [57700] global_step=57700, grad_norm=3.952075481414795, loss=1.0295724868774414 +I0831 13:47:45.497194 140462291654400 logging_writer.py:48] [57800] global_step=57800, grad_norm=4.220205307006836, loss=1.1891672611236572 +I0831 13:48:12.696619 140462207792896 logging_writer.py:48] [57900] global_step=57900, grad_norm=4.217330455780029, loss=1.274344563484192 +I0831 13:48:33.415342 140659750036672 spec.py:333] Evaluating on the training split. +I0831 13:48:42.497893 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 13:48:51.561762 140659750036672 spec.py:363] Evaluating on the test split. +I0831 13:48:52.438694 140659750036672 submission_runner.py:516] Time since start: 16253.61s, Step: 57978, {'train/accuracy': Array(0.8929368, dtype=float32), 'train/loss': Array(0.38473663, dtype=float32), 'validation/accuracy': Array(0.73706, dtype=float32), 'validation/loss': Array(1.0853631, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.60950005, dtype=float32), 'test/loss': Array(1.8237377, dtype=float32), 'test/num_examples': 10000, 'score': 16020.971867799759, 'total_duration': 16253.614742040634, 'accumulated_submission_time': 16020.971867799759, 'accumulated_eval_time': 231.5355188846588, 'accumulated_logging_time': 0.6126320362091064} +I0831 13:48:52.503669 140462291654400 logging_writer.py:48] [57978] accumulated_eval_time=231.536, accumulated_logging_time=0.612632, accumulated_submission_time=16021, global_step=57978, preemption_count=0, score=16021, test/accuracy=0.6095000505447388, test/loss=1.8237377405166626, test/num_examples=10000, total_duration=16253.6, train/accuracy=0.8929368257522583, train/loss=0.3847366273403168, validation/accuracy=0.737060010433197, validation/loss=1.0853631496429443, validation/num_examples=50000 +I0831 13:48:58.883542 140462207792896 logging_writer.py:48] [58000] global_step=58000, grad_norm=4.0574750900268555, loss=1.2353789806365967 +I0831 13:49:26.009712 140462291654400 logging_writer.py:48] [58100] global_step=58100, grad_norm=3.9186370372772217, loss=1.1001039743423462 +I0831 13:49:53.195768 140462207792896 logging_writer.py:48] [58200] global_step=58200, grad_norm=4.232709884643555, loss=1.2277204990386963 +I0831 13:50:20.316319 140462291654400 logging_writer.py:48] [58300] global_step=58300, grad_norm=4.648276329040527, loss=1.223358392715454 +I0831 13:50:47.442259 140462207792896 logging_writer.py:48] [58400] global_step=58400, grad_norm=4.073798179626465, loss=1.0981905460357666 +I0831 13:51:14.658367 140462291654400 logging_writer.py:48] [58500] global_step=58500, grad_norm=3.94584321975708, loss=1.1303236484527588 +I0831 13:51:41.776161 140462207792896 logging_writer.py:48] [58600] global_step=58600, grad_norm=3.847421407699585, loss=1.1483267545700073 +I0831 13:52:08.925719 140462291654400 logging_writer.py:48] [58700] global_step=58700, grad_norm=4.207337856292725, loss=1.2111364603042603 +I0831 13:52:36.358842 140462207792896 logging_writer.py:48] [58800] global_step=58800, grad_norm=4.129154205322266, loss=1.1786203384399414 +I0831 13:53:03.495699 140462291654400 logging_writer.py:48] [58900] global_step=58900, grad_norm=3.8919084072113037, loss=1.13741135597229 +I0831 13:53:30.619644 140462207792896 logging_writer.py:48] [59000] global_step=59000, grad_norm=4.069941997528076, loss=1.106372356414795 +I0831 13:53:57.825412 140462291654400 logging_writer.py:48] [59100] global_step=59100, grad_norm=4.114039421081543, loss=1.144714117050171 +I0831 13:54:24.945344 140462207792896 logging_writer.py:48] [59200] global_step=59200, grad_norm=4.488448619842529, loss=1.2532100677490234 +I0831 13:54:52.083218 140462291654400 logging_writer.py:48] [59300] global_step=59300, grad_norm=4.21856164932251, loss=1.1614978313446045 +I0831 13:55:19.277261 140462207792896 logging_writer.py:48] [59400] global_step=59400, grad_norm=4.1410040855407715, loss=1.1232411861419678 +I0831 13:55:46.448896 140462291654400 logging_writer.py:48] [59500] global_step=59500, grad_norm=4.294313430786133, loss=1.1341241598129272 +I0831 13:56:13.588469 140462207792896 logging_writer.py:48] [59600] global_step=59600, grad_norm=4.498416900634766, loss=1.289866328239441 +I0831 13:56:40.790091 140462291654400 logging_writer.py:48] [59700] global_step=59700, grad_norm=4.234975814819336, loss=1.1874021291732788 +I0831 13:57:08.197914 140462207792896 logging_writer.py:48] [59800] global_step=59800, grad_norm=4.112311363220215, loss=1.0667126178741455 +I0831 13:57:35.333203 140462291654400 logging_writer.py:48] [59900] global_step=59900, grad_norm=4.129216194152832, loss=1.2047224044799805 +I0831 13:58:02.515774 140462207792896 logging_writer.py:48] [60000] global_step=60000, grad_norm=4.231048583984375, loss=1.1794657707214355 +I0831 13:58:29.640873 140462291654400 logging_writer.py:48] [60100] global_step=60100, grad_norm=4.340626239776611, loss=1.21722412109375 +I0831 13:58:56.758829 140462207792896 logging_writer.py:48] [60200] global_step=60200, grad_norm=4.004168510437012, loss=1.1437658071517944 +I0831 13:59:23.957289 140462291654400 logging_writer.py:48] [60300] global_step=60300, grad_norm=4.256436824798584, loss=1.1482956409454346 +I0831 13:59:51.088067 140462207792896 logging_writer.py:48] [60400] global_step=60400, grad_norm=4.036151885986328, loss=1.110124111175537 +I0831 14:00:18.224655 140462291654400 logging_writer.py:48] [60500] global_step=60500, grad_norm=4.349050521850586, loss=1.2058964967727661 +I0831 14:00:45.418267 140462207792896 logging_writer.py:48] [60600] global_step=60600, grad_norm=3.9427640438079834, loss=1.1109867095947266 +I0831 14:01:12.553056 140462291654400 logging_writer.py:48] [60700] global_step=60700, grad_norm=3.9597513675689697, loss=1.0898019075393677 +I0831 14:01:39.688182 140462207792896 logging_writer.py:48] [60800] global_step=60800, grad_norm=4.077946186065674, loss=1.1935244798660278 +I0831 14:02:07.109884 140462291654400 logging_writer.py:48] [60900] global_step=60900, grad_norm=3.825608253479004, loss=0.9942179918289185 +I0831 14:02:34.259627 140462207792896 logging_writer.py:48] [61000] global_step=61000, grad_norm=4.335648536682129, loss=1.159499168395996 +I0831 14:03:01.402373 140462291654400 logging_writer.py:48] [61100] global_step=61100, grad_norm=3.933427572250366, loss=1.0411491394042969 +I0831 14:03:28.609000 140462207792896 logging_writer.py:48] [61200] global_step=61200, grad_norm=4.394455432891846, loss=1.1468101739883423 +I0831 14:03:55.748989 140462291654400 logging_writer.py:48] [61300] global_step=61300, grad_norm=4.488161087036133, loss=1.2605468034744263 +I0831 14:04:22.886924 140462207792896 logging_writer.py:48] [61400] global_step=61400, grad_norm=4.1989006996154785, loss=1.1437532901763916 +I0831 14:04:50.095343 140462291654400 logging_writer.py:48] [61500] global_step=61500, grad_norm=4.138639450073242, loss=1.1480075120925903 +I0831 14:05:17.220824 140462207792896 logging_writer.py:48] [61600] global_step=61600, grad_norm=3.9928810596466064, loss=1.0494506359100342 +I0831 14:05:44.370078 140462291654400 logging_writer.py:48] [61700] global_step=61700, grad_norm=4.209929943084717, loss=1.0548369884490967 +I0831 14:06:11.558336 140462207792896 logging_writer.py:48] [61800] global_step=61800, grad_norm=4.1351165771484375, loss=1.1320818662643433 +I0831 14:06:38.962699 140462291654400 logging_writer.py:48] [61900] global_step=61900, grad_norm=4.402517795562744, loss=1.174030065536499 +I0831 14:07:06.096302 140462207792896 logging_writer.py:48] [62000] global_step=62000, grad_norm=4.635240077972412, loss=1.14308762550354 +I0831 14:07:33.288693 140462291654400 logging_writer.py:48] [62100] global_step=62100, grad_norm=4.353769779205322, loss=1.202255368232727 +I0831 14:08:00.423377 140462207792896 logging_writer.py:48] [62200] global_step=62200, grad_norm=4.360079765319824, loss=1.1736249923706055 +I0831 14:08:27.587749 140462291654400 logging_writer.py:48] [62300] global_step=62300, grad_norm=4.5913310050964355, loss=1.2308727502822876 +I0831 14:08:54.800509 140462207792896 logging_writer.py:48] [62400] global_step=62400, grad_norm=4.2525224685668945, loss=1.1051523685455322 +I0831 14:09:21.927550 140462291654400 logging_writer.py:48] [62500] global_step=62500, grad_norm=4.346681594848633, loss=1.189514398574829 +I0831 14:09:49.094521 140462207792896 logging_writer.py:48] [62600] global_step=62600, grad_norm=4.452826976776123, loss=1.0993497371673584 +I0831 14:10:16.291604 140462291654400 logging_writer.py:48] [62700] global_step=62700, grad_norm=4.38563346862793, loss=1.2077784538269043 +I0831 14:10:43.439439 140462207792896 logging_writer.py:48] [62800] global_step=62800, grad_norm=4.4207539558410645, loss=1.1171356439590454 +I0831 14:11:10.572555 140462291654400 logging_writer.py:48] [62900] global_step=62900, grad_norm=4.31390380859375, loss=1.0374815464019775 +I0831 14:11:38.008424 140462207792896 logging_writer.py:48] [63000] global_step=63000, grad_norm=4.479304790496826, loss=1.1356697082519531 +I0831 14:12:05.152732 140462291654400 logging_writer.py:48] [63100] global_step=63100, grad_norm=4.319552421569824, loss=1.1420392990112305 +I0831 14:12:32.275512 140462207792896 logging_writer.py:48] [63200] global_step=63200, grad_norm=4.683479309082031, loss=1.1916344165802002 +I0831 14:12:59.482403 140462291654400 logging_writer.py:48] [63300] global_step=63300, grad_norm=4.598440647125244, loss=1.1886094808578491 +I0831 14:13:26.599662 140462207792896 logging_writer.py:48] [63400] global_step=63400, grad_norm=4.342299938201904, loss=1.2049875259399414 +I0831 14:13:53.728135 140462291654400 logging_writer.py:48] [63500] global_step=63500, grad_norm=4.002748012542725, loss=0.9823920726776123 +I0831 14:14:20.946938 140462207792896 logging_writer.py:48] [63600] global_step=63600, grad_norm=4.560853958129883, loss=1.216147541999817 +I0831 14:14:48.078398 140462291654400 logging_writer.py:48] [63700] global_step=63700, grad_norm=4.181757926940918, loss=1.0669641494750977 +I0831 14:15:15.222414 140462207792896 logging_writer.py:48] [63800] global_step=63800, grad_norm=4.614319801330566, loss=1.2537699937820435 +I0831 14:15:42.422091 140462291654400 logging_writer.py:48] [63900] global_step=63900, grad_norm=4.161206245422363, loss=1.1217033863067627 +I0831 14:16:09.793797 140462207792896 logging_writer.py:48] [64000] global_step=64000, grad_norm=4.328306674957275, loss=1.153753399848938 +I0831 14:16:36.932788 140462291654400 logging_writer.py:48] [64100] global_step=64100, grad_norm=4.707287788391113, loss=1.226334571838379 +I0831 14:17:04.134815 140462207792896 logging_writer.py:48] [64200] global_step=64200, grad_norm=4.304986476898193, loss=1.0734068155288696 +I0831 14:17:31.299638 140462291654400 logging_writer.py:48] [64300] global_step=64300, grad_norm=4.589047908782959, loss=1.2880934476852417 +I0831 14:17:58.424725 140462207792896 logging_writer.py:48] [64400] global_step=64400, grad_norm=4.380955219268799, loss=1.1563457250595093 +I0831 14:18:25.640285 140462291654400 logging_writer.py:48] [64500] global_step=64500, grad_norm=4.430264472961426, loss=1.1682482957839966 +I0831 14:18:52.764599 140462207792896 logging_writer.py:48] [64600] global_step=64600, grad_norm=4.740336894989014, loss=1.1271042823791504 +I0831 14:19:19.921661 140462291654400 logging_writer.py:48] [64700] global_step=64700, grad_norm=4.512879848480225, loss=1.128707766532898 +I0831 14:19:47.102000 140462207792896 logging_writer.py:48] [64800] global_step=64800, grad_norm=4.561408042907715, loss=1.121180534362793 +I0831 14:20:14.230034 140462291654400 logging_writer.py:48] [64900] global_step=64900, grad_norm=4.737875461578369, loss=1.2171547412872314 +I0831 14:20:41.373762 140462207792896 logging_writer.py:48] [65000] global_step=65000, grad_norm=4.4257378578186035, loss=1.1137096881866455 +I0831 14:21:08.771758 140462291654400 logging_writer.py:48] [65100] global_step=65100, grad_norm=4.627267837524414, loss=1.1256811618804932 +I0831 14:21:35.884293 140462207792896 logging_writer.py:48] [65200] global_step=65200, grad_norm=4.349369049072266, loss=1.0520504713058472 +I0831 14:22:03.006849 140462291654400 logging_writer.py:48] [65300] global_step=65300, grad_norm=4.512179374694824, loss=1.2575383186340332 +I0831 14:22:08.587691 140659750036672 spec.py:333] Evaluating on the training split. +I0831 14:22:17.867395 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 14:22:26.893923 140659750036672 spec.py:363] Evaluating on the test split. +I0831 14:22:27.744155 140659750036672 submission_runner.py:516] Time since start: 18268.92s, Step: 65322, {'train/accuracy': Array(0.9012675, dtype=float32), 'train/loss': Array(0.35587227, dtype=float32), 'validation/accuracy': Array(0.74094, dtype=float32), 'validation/loss': Array(1.0815302, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.61170006, dtype=float32), 'test/loss': Array(1.827934, dtype=float32), 'test/num_examples': 10000, 'score': 18016.97361254692, 'total_duration': 18268.919723272324, 'accumulated_submission_time': 18016.97361254692, 'accumulated_eval_time': 250.69002890586853, 'accumulated_logging_time': 0.701411247253418} +I0831 14:22:27.788972 140462207792896 logging_writer.py:48] [65322] accumulated_eval_time=250.69, accumulated_logging_time=0.701411, accumulated_submission_time=18017, global_step=65322, preemption_count=0, score=18017, test/accuracy=0.6117000579833984, test/loss=1.8279340267181396, test/num_examples=10000, total_duration=18268.9, train/accuracy=0.9012675285339355, train/loss=0.3558722734451294, validation/accuracy=0.7409399747848511, validation/loss=1.081530213356018, validation/num_examples=50000 +I0831 14:22:49.389630 140462291654400 logging_writer.py:48] [65400] global_step=65400, grad_norm=4.284206390380859, loss=1.0190246105194092 +I0831 14:23:16.542282 140462207792896 logging_writer.py:48] [65500] global_step=65500, grad_norm=4.932464599609375, loss=1.2434974908828735 +I0831 14:23:43.700573 140462291654400 logging_writer.py:48] [65600] global_step=65600, grad_norm=4.53204870223999, loss=1.0806829929351807 +I0831 14:24:10.928452 140462207792896 logging_writer.py:48] [65700] global_step=65700, grad_norm=4.491011619567871, loss=1.096167802810669 +I0831 14:24:38.068518 140462291654400 logging_writer.py:48] [65800] global_step=65800, grad_norm=4.554718971252441, loss=1.133201241493225 +I0831 14:25:05.193167 140462207792896 logging_writer.py:48] [65900] global_step=65900, grad_norm=4.494077682495117, loss=1.0525048971176147 +I0831 14:25:32.416553 140462291654400 logging_writer.py:48] [66000] global_step=66000, grad_norm=4.502111434936523, loss=1.100895643234253 +I0831 14:25:59.536046 140462207792896 logging_writer.py:48] [66100] global_step=66100, grad_norm=4.566282749176025, loss=1.0475738048553467 +I0831 14:26:26.906597 140462291654400 logging_writer.py:48] [66200] global_step=66200, grad_norm=4.419798851013184, loss=1.126304268836975 +I0831 14:26:54.103857 140462207792896 logging_writer.py:48] [66300] global_step=66300, grad_norm=4.399794101715088, loss=1.0503495931625366 +I0831 14:27:21.257030 140462291654400 logging_writer.py:48] [66400] global_step=66400, grad_norm=4.890985012054443, loss=1.1326605081558228 +I0831 14:27:48.404087 140462207792896 logging_writer.py:48] [66500] global_step=66500, grad_norm=4.889479160308838, loss=1.1310876607894897 +I0831 14:28:15.613812 140462291654400 logging_writer.py:48] [66600] global_step=66600, grad_norm=4.340965270996094, loss=1.0289013385772705 +I0831 14:28:42.730699 140462207792896 logging_writer.py:48] [66700] global_step=66700, grad_norm=4.932310104370117, loss=1.1279044151306152 +I0831 14:29:09.848071 140462291654400 logging_writer.py:48] [66800] global_step=66800, grad_norm=4.479370594024658, loss=1.1179966926574707 +I0831 14:29:37.036448 140462207792896 logging_writer.py:48] [66900] global_step=66900, grad_norm=4.530259609222412, loss=1.1143271923065186 +I0831 14:30:04.161231 140462291654400 logging_writer.py:48] [67000] global_step=67000, grad_norm=5.057637691497803, loss=1.1680517196655273 +I0831 14:30:31.293544 140462207792896 logging_writer.py:48] [67100] global_step=67100, grad_norm=4.431516170501709, loss=1.1127394437789917 +I0831 14:30:58.492388 140462291654400 logging_writer.py:48] [67200] global_step=67200, grad_norm=4.792877674102783, loss=1.1247069835662842 +I0831 14:31:25.852335 140462207792896 logging_writer.py:48] [67300] global_step=67300, grad_norm=4.476561069488525, loss=1.1439166069030762 +I0831 14:31:52.970043 140462291654400 logging_writer.py:48] [67400] global_step=67400, grad_norm=4.633327007293701, loss=1.1331901550292969 +I0831 14:32:20.155227 140462207792896 logging_writer.py:48] [67500] global_step=67500, grad_norm=4.84542179107666, loss=1.0841716527938843 +I0831 14:32:47.289683 140462291654400 logging_writer.py:48] [67600] global_step=67600, grad_norm=4.836230278015137, loss=1.1284255981445312 +I0831 14:33:14.435696 140462207792896 logging_writer.py:48] [67700] global_step=67700, grad_norm=4.809790134429932, loss=1.0862503051757812 +I0831 14:33:41.619664 140462291654400 logging_writer.py:48] [67800] global_step=67800, grad_norm=4.821517467498779, loss=1.129023551940918 +I0831 14:34:08.755947 140462207792896 logging_writer.py:48] [67900] global_step=67900, grad_norm=4.542420387268066, loss=1.0678744316101074 +I0831 14:34:35.891966 140462291654400 logging_writer.py:48] [68000] global_step=68000, grad_norm=5.121830463409424, loss=1.2001752853393555 +I0831 14:35:03.117610 140462207792896 logging_writer.py:48] [68100] global_step=68100, grad_norm=4.539107799530029, loss=1.0397202968597412 +I0831 14:35:30.228630 140462291654400 logging_writer.py:48] [68200] global_step=68200, grad_norm=5.082265377044678, loss=1.1701273918151855 +I0831 14:35:57.344394 140462207792896 logging_writer.py:48] [68300] global_step=68300, grad_norm=4.715770244598389, loss=1.1374961137771606 +I0831 14:36:24.756226 140462291654400 logging_writer.py:48] [68400] global_step=68400, grad_norm=4.886198043823242, loss=1.1447570323944092 +I0831 14:36:51.893005 140462207792896 logging_writer.py:48] [68500] global_step=68500, grad_norm=5.014782428741455, loss=1.114876389503479 +I0831 14:37:19.016660 140462291654400 logging_writer.py:48] [68600] global_step=68600, grad_norm=4.8369646072387695, loss=1.105344533920288 +I0831 14:37:46.259731 140462207792896 logging_writer.py:48] [68700] global_step=68700, grad_norm=4.724262714385986, loss=1.1144500970840454 +I0831 14:38:13.407853 140462291654400 logging_writer.py:48] [68800] global_step=68800, grad_norm=4.575559139251709, loss=1.0805474519729614 +I0831 14:38:40.523135 140462207792896 logging_writer.py:48] [68900] global_step=68900, grad_norm=4.845252513885498, loss=1.183036208152771 +I0831 14:39:07.719451 140462291654400 logging_writer.py:48] [69000] global_step=69000, grad_norm=4.726552963256836, loss=1.1148223876953125 +I0831 14:39:34.851024 140462207792896 logging_writer.py:48] [69100] global_step=69100, grad_norm=4.690332412719727, loss=1.059108853340149 +I0831 14:40:01.983876 140462291654400 logging_writer.py:48] [69200] global_step=69200, grad_norm=4.8058671951293945, loss=1.1190779209136963 +I0831 14:40:29.229309 140462207792896 logging_writer.py:48] [69300] global_step=69300, grad_norm=4.794759750366211, loss=1.0835239887237549 +I0831 14:40:56.607501 140462291654400 logging_writer.py:48] [69400] global_step=69400, grad_norm=4.834987640380859, loss=1.153177261352539 +I0831 14:41:23.731445 140462207792896 logging_writer.py:48] [69500] global_step=69500, grad_norm=4.849311351776123, loss=1.097834587097168 +I0831 14:41:50.946708 140462291654400 logging_writer.py:48] [69600] global_step=69600, grad_norm=4.562088966369629, loss=1.1294622421264648 +I0831 14:42:18.079017 140462207792896 logging_writer.py:48] [69700] global_step=69700, grad_norm=4.771415710449219, loss=1.1647746562957764 +I0831 14:42:45.215358 140462291654400 logging_writer.py:48] [69800] global_step=69800, grad_norm=4.971536159515381, loss=1.1076338291168213 +I0831 14:43:12.404544 140462207792896 logging_writer.py:48] [69900] global_step=69900, grad_norm=5.116303443908691, loss=1.1614720821380615 +I0831 14:43:39.531889 140462291654400 logging_writer.py:48] [70000] global_step=70000, grad_norm=4.858510971069336, loss=1.111580729484558 +I0831 14:44:06.688975 140462207792896 logging_writer.py:48] [70100] global_step=70100, grad_norm=4.824286937713623, loss=1.148178219795227 +I0831 14:44:33.876608 140462291654400 logging_writer.py:48] [70200] global_step=70200, grad_norm=5.05073356628418, loss=1.129698395729065 +I0831 14:45:01.021797 140462207792896 logging_writer.py:48] [70300] global_step=70300, grad_norm=4.9951019287109375, loss=1.1462994813919067 +I0831 14:45:28.139046 140462291654400 logging_writer.py:48] [70400] global_step=70400, grad_norm=5.174649715423584, loss=1.19618821144104 +I0831 14:45:55.551306 140462207792896 logging_writer.py:48] [70500] global_step=70500, grad_norm=5.2298126220703125, loss=1.2179630994796753 +I0831 14:46:22.695321 140462291654400 logging_writer.py:48] [70600] global_step=70600, grad_norm=5.161717414855957, loss=1.1740407943725586 +I0831 14:46:49.843053 140462207792896 logging_writer.py:48] [70700] global_step=70700, grad_norm=5.038063049316406, loss=1.1523008346557617 +I0831 14:47:17.038139 140462291654400 logging_writer.py:48] [70800] global_step=70800, grad_norm=4.958276748657227, loss=1.0678048133850098 +I0831 14:47:44.171280 140462207792896 logging_writer.py:48] [70900] global_step=70900, grad_norm=5.025483131408691, loss=1.1770962476730347 +I0831 14:48:11.320039 140462291654400 logging_writer.py:48] [71000] global_step=71000, grad_norm=5.062246799468994, loss=1.124931812286377 +I0831 14:48:38.519197 140462207792896 logging_writer.py:48] [71100] global_step=71100, grad_norm=4.8349080085754395, loss=1.073495626449585 +I0831 14:49:05.643352 140462291654400 logging_writer.py:48] [71200] global_step=71200, grad_norm=5.02469539642334, loss=1.1062854528427124 +I0831 14:49:32.780731 140462207792896 logging_writer.py:48] [71300] global_step=71300, grad_norm=5.048979759216309, loss=1.1153062582015991 +I0831 14:49:59.982401 140462291654400 logging_writer.py:48] [71400] global_step=71400, grad_norm=5.089901447296143, loss=1.193464756011963 +I0831 14:50:27.371221 140462207792896 logging_writer.py:48] [71500] global_step=71500, grad_norm=5.065579414367676, loss=1.1467745304107666 +I0831 14:50:54.487068 140462291654400 logging_writer.py:48] [71600] global_step=71600, grad_norm=4.74937105178833, loss=1.0938713550567627 +I0831 14:51:21.700103 140462207792896 logging_writer.py:48] [71700] global_step=71700, grad_norm=5.357005596160889, loss=1.2373077869415283 +I0831 14:51:48.808971 140462291654400 logging_writer.py:48] [71800] global_step=71800, grad_norm=4.91694974899292, loss=1.0784114599227905 +I0831 14:52:15.936059 140462207792896 logging_writer.py:48] [71900] global_step=71900, grad_norm=4.8602399826049805, loss=1.1049989461898804 +I0831 14:52:43.123491 140462291654400 logging_writer.py:48] [72000] global_step=72000, grad_norm=5.576911926269531, loss=1.2597181797027588 +I0831 14:53:10.298779 140462207792896 logging_writer.py:48] [72100] global_step=72100, grad_norm=4.9126482009887695, loss=1.0772030353546143 +I0831 14:53:37.414741 140462291654400 logging_writer.py:48] [72200] global_step=72200, grad_norm=5.017378330230713, loss=1.1221458911895752 +I0831 14:54:04.593322 140462207792896 logging_writer.py:48] [72300] global_step=72300, grad_norm=4.766191005706787, loss=1.0541183948516846 +I0831 14:54:31.730116 140462291654400 logging_writer.py:48] [72400] global_step=72400, grad_norm=4.713466167449951, loss=1.0424288511276245 +I0831 14:54:58.841064 140462207792896 logging_writer.py:48] [72500] global_step=72500, grad_norm=4.87831449508667, loss=1.0286810398101807 +I0831 14:55:26.304171 140462291654400 logging_writer.py:48] [72600] global_step=72600, grad_norm=5.030486106872559, loss=1.0626521110534668 +I0831 14:55:43.764853 140659750036672 spec.py:333] Evaluating on the training split. +I0831 14:55:52.624987 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 14:56:01.582471 140659750036672 spec.py:363] Evaluating on the test split. +I0831 14:56:02.477355 140659750036672 submission_runner.py:516] Time since start: 20283.65s, Step: 72666, {'train/accuracy': Array(0.9071468, dtype=float32), 'train/loss': Array(0.33349392, dtype=float32), 'validation/accuracy': Array(0.74214, dtype=float32), 'validation/loss': Array(1.0764859, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6153, dtype=float32), 'test/loss': Array(1.8251016, dtype=float32), 'test/num_examples': 10000, 'score': 20012.831142663956, 'total_duration': 20283.65297317505, 'accumulated_submission_time': 20012.831142663956, 'accumulated_eval_time': 269.40062832832336, 'accumulated_logging_time': 0.8057889938354492} +I0831 14:56:02.531687 140462207792896 logging_writer.py:48] [72666] accumulated_eval_time=269.401, accumulated_logging_time=0.805789, accumulated_submission_time=20012.8, global_step=72666, preemption_count=0, score=20012.8, test/accuracy=0.6152999997138977, test/loss=1.825101613998413, test/num_examples=10000, total_duration=20283.7, train/accuracy=0.9071468114852905, train/loss=0.3334939181804657, validation/accuracy=0.742139995098114, validation/loss=1.0764858722686768, validation/num_examples=50000 +I0831 14:56:12.201543 140462291654400 logging_writer.py:48] [72700] global_step=72700, grad_norm=4.988339900970459, loss=1.0567495822906494 +I0831 14:56:39.346813 140462207792896 logging_writer.py:48] [72800] global_step=72800, grad_norm=5.282829761505127, loss=1.1511059999465942 +I0831 14:57:06.554137 140462291654400 logging_writer.py:48] [72900] global_step=72900, grad_norm=4.980942249298096, loss=1.05224609375 +I0831 14:57:33.713636 140462207792896 logging_writer.py:48] [73000] global_step=73000, grad_norm=5.556791305541992, loss=1.2654364109039307 +I0831 14:58:00.846304 140462291654400 logging_writer.py:48] [73100] global_step=73100, grad_norm=5.11857795715332, loss=1.205348014831543 +I0831 14:58:28.070595 140462207792896 logging_writer.py:48] [73200] global_step=73200, grad_norm=5.123295783996582, loss=1.078803539276123 +I0831 14:58:55.198497 140462291654400 logging_writer.py:48] [73300] global_step=73300, grad_norm=4.979549884796143, loss=1.0549975633621216 +I0831 14:59:22.329517 140462207792896 logging_writer.py:48] [73400] global_step=73400, grad_norm=4.953442573547363, loss=1.0956718921661377 +I0831 14:59:49.515853 140462291654400 logging_writer.py:48] [73500] global_step=73500, grad_norm=5.243229389190674, loss=1.1241934299468994 +I0831 15:00:16.636009 140462207792896 logging_writer.py:48] [73600] global_step=73600, grad_norm=4.930180072784424, loss=1.0406070947647095 +I0831 15:00:43.982019 140462291654400 logging_writer.py:48] [73700] global_step=73700, grad_norm=5.154042720794678, loss=1.1904408931732178 +I0831 15:01:11.194699 140462207792896 logging_writer.py:48] [73800] global_step=73800, grad_norm=5.013015270233154, loss=1.07711923122406 +I0831 15:01:38.317161 140462291654400 logging_writer.py:48] [73900] global_step=73900, grad_norm=5.092146873474121, loss=1.100111722946167 +I0831 15:02:05.426794 140462207792896 logging_writer.py:48] [74000] global_step=74000, grad_norm=4.945342063903809, loss=1.0772162675857544 +I0831 15:02:32.600225 140462291654400 logging_writer.py:48] [74100] global_step=74100, grad_norm=5.114232540130615, loss=1.0843480825424194 +I0831 15:02:59.719852 140462207792896 logging_writer.py:48] [74200] global_step=74200, grad_norm=5.4109978675842285, loss=1.2193963527679443 +I0831 15:03:26.842858 140462291654400 logging_writer.py:48] [74300] global_step=74300, grad_norm=5.093811511993408, loss=1.0939356088638306 +I0831 15:03:54.059223 140462207792896 logging_writer.py:48] [74400] global_step=74400, grad_norm=5.186962604522705, loss=1.139366865158081 +I0831 15:04:21.179890 140462291654400 logging_writer.py:48] [74500] global_step=74500, grad_norm=4.879283905029297, loss=1.0728275775909424 +I0831 15:04:48.308450 140462207792896 logging_writer.py:48] [74600] global_step=74600, grad_norm=5.0075225830078125, loss=1.0428383350372314 +I0831 15:05:15.528039 140462291654400 logging_writer.py:48] [74700] global_step=74700, grad_norm=5.003865718841553, loss=1.1002367734909058 +I0831 15:05:42.888578 140462207792896 logging_writer.py:48] [74800] global_step=74800, grad_norm=5.081692218780518, loss=1.0607489347457886 +I0831 15:06:10.024645 140462291654400 logging_writer.py:48] [74900] global_step=74900, grad_norm=5.044943809509277, loss=1.105198860168457 +I0831 15:06:37.204737 140462207792896 logging_writer.py:48] [75000] global_step=75000, grad_norm=5.308656215667725, loss=1.1711395978927612 +I0831 15:07:04.350597 140462291654400 logging_writer.py:48] [75100] global_step=75100, grad_norm=5.252565383911133, loss=1.1370654106140137 +I0831 15:07:31.470876 140462207792896 logging_writer.py:48] [75200] global_step=75200, grad_norm=4.9596710205078125, loss=1.0261977910995483 +I0831 15:07:58.703403 140462291654400 logging_writer.py:48] [75300] global_step=75300, grad_norm=5.137115955352783, loss=1.1482950448989868 +I0831 15:08:25.831673 140462207792896 logging_writer.py:48] [75400] global_step=75400, grad_norm=5.261562824249268, loss=1.1152247190475464 +I0831 15:08:52.980357 140462291654400 logging_writer.py:48] [75500] global_step=75500, grad_norm=4.953219413757324, loss=1.1569592952728271 +I0831 15:09:20.183687 140462207792896 logging_writer.py:48] [75600] global_step=75600, grad_norm=5.322703838348389, loss=1.0978277921676636 +I0831 15:09:47.315649 140462291654400 logging_writer.py:48] [75700] global_step=75700, grad_norm=5.218213081359863, loss=1.141169548034668 +I0831 15:10:14.685340 140462207792896 logging_writer.py:48] [75800] global_step=75800, grad_norm=4.717349052429199, loss=1.0207397937774658 +I0831 15:10:41.865921 140462291654400 logging_writer.py:48] [75900] global_step=75900, grad_norm=5.152605056762695, loss=1.126367211341858 +I0831 15:11:09.008747 140462207792896 logging_writer.py:48] [76000] global_step=76000, grad_norm=5.174790859222412, loss=1.1213984489440918 +I0831 15:11:36.128692 140462291654400 logging_writer.py:48] [76100] global_step=76100, grad_norm=5.230886459350586, loss=1.1371116638183594 +I0831 15:12:03.317011 140462207792896 logging_writer.py:48] [76200] global_step=76200, grad_norm=5.105378150939941, loss=1.078294277191162 +I0831 15:12:30.455310 140462291654400 logging_writer.py:48] [76300] global_step=76300, grad_norm=5.3959856033325195, loss=1.1209861040115356 +I0831 15:12:57.595014 140462207792896 logging_writer.py:48] [76400] global_step=76400, grad_norm=5.312765598297119, loss=1.158987045288086 +I0831 15:13:24.786737 140462291654400 logging_writer.py:48] [76500] global_step=76500, grad_norm=5.122409343719482, loss=1.1337230205535889 +I0831 15:13:51.924569 140462207792896 logging_writer.py:48] [76600] global_step=76600, grad_norm=4.891146183013916, loss=1.0661208629608154 +I0831 15:14:19.087124 140462291654400 logging_writer.py:48] [76700] global_step=76700, grad_norm=5.494735240936279, loss=1.3196933269500732 +I0831 15:14:46.292710 140462207792896 logging_writer.py:48] [76800] global_step=76800, grad_norm=5.1176371574401855, loss=1.1478564739227295 +I0831 15:15:13.662039 140462291654400 logging_writer.py:48] [76900] global_step=76900, grad_norm=5.544063091278076, loss=1.2121343612670898 +I0831 15:15:40.799252 140462207792896 logging_writer.py:48] [77000] global_step=77000, grad_norm=4.994622230529785, loss=1.1006391048431396 +I0831 15:16:08.001233 140462291654400 logging_writer.py:48] [77100] global_step=77100, grad_norm=5.582889556884766, loss=1.210869550704956 +I0831 15:16:35.119962 140462207792896 logging_writer.py:48] [77200] global_step=77200, grad_norm=5.4206132888793945, loss=1.117112398147583 +I0831 15:17:02.280909 140462291654400 logging_writer.py:48] [77300] global_step=77300, grad_norm=5.197893142700195, loss=1.1093621253967285 +I0831 15:17:29.505422 140462207792896 logging_writer.py:48] [77400] global_step=77400, grad_norm=4.984208106994629, loss=1.0767338275909424 +I0831 15:17:56.648786 140462291654400 logging_writer.py:48] [77500] global_step=77500, grad_norm=5.26463508605957, loss=1.196333885192871 +I0831 15:18:23.812944 140462207792896 logging_writer.py:48] [77600] global_step=77600, grad_norm=4.850378036499023, loss=0.9869534969329834 +I0831 15:18:50.998341 140462291654400 logging_writer.py:48] [77700] global_step=77700, grad_norm=5.361964702606201, loss=1.2147841453552246 +I0831 15:19:18.117804 140462207792896 logging_writer.py:48] [77800] global_step=77800, grad_norm=5.195345401763916, loss=1.1594328880310059 +I0831 15:19:45.258813 140462291654400 logging_writer.py:48] [77900] global_step=77900, grad_norm=5.28692102432251, loss=1.106515645980835 +I0831 15:20:12.679407 140462207792896 logging_writer.py:48] [78000] global_step=78000, grad_norm=5.173584938049316, loss=1.1548712253570557 +I0831 15:20:39.840758 140462291654400 logging_writer.py:48] [78100] global_step=78100, grad_norm=4.914551258087158, loss=0.9914030432701111 +I0831 15:21:06.958157 140462207792896 logging_writer.py:48] [78200] global_step=78200, grad_norm=5.167934417724609, loss=1.1004729270935059 +I0831 15:21:34.174460 140462291654400 logging_writer.py:48] [78300] global_step=78300, grad_norm=5.197473049163818, loss=1.0962107181549072 +I0831 15:22:01.276096 140462207792896 logging_writer.py:48] [78400] global_step=78400, grad_norm=5.070916652679443, loss=1.101027011871338 +I0831 15:22:28.412616 140462291654400 logging_writer.py:48] [78500] global_step=78500, grad_norm=4.951089859008789, loss=1.0859774351119995 +I0831 15:22:55.648705 140462207792896 logging_writer.py:48] [78600] global_step=78600, grad_norm=5.2664947509765625, loss=1.173295259475708 +I0831 15:23:22.780879 140462291654400 logging_writer.py:48] [78700] global_step=78700, grad_norm=5.147072792053223, loss=1.0562145709991455 +I0831 15:23:49.910059 140462207792896 logging_writer.py:48] [78800] global_step=78800, grad_norm=5.206952095031738, loss=1.1434600353240967 +I0831 15:24:17.140887 140462291654400 logging_writer.py:48] [78900] global_step=78900, grad_norm=5.333350658416748, loss=1.1164429187774658 +I0831 15:24:44.489534 140462207792896 logging_writer.py:48] [79000] global_step=79000, grad_norm=5.01648473739624, loss=1.0871572494506836 +I0831 15:25:11.637832 140462291654400 logging_writer.py:48] [79100] global_step=79100, grad_norm=5.365006446838379, loss=1.1525310277938843 +I0831 15:25:38.850716 140462207792896 logging_writer.py:48] [79200] global_step=79200, grad_norm=5.134157180786133, loss=1.1608151197433472 +I0831 15:26:05.997452 140462291654400 logging_writer.py:48] [79300] global_step=79300, grad_norm=5.2374677658081055, loss=1.2145328521728516 +I0831 15:26:33.155091 140462207792896 logging_writer.py:48] [79400] global_step=79400, grad_norm=4.954680919647217, loss=1.0658445358276367 +I0831 15:27:00.350726 140462291654400 logging_writer.py:48] [79500] global_step=79500, grad_norm=5.122426986694336, loss=1.0331730842590332 +I0831 15:27:27.517232 140462207792896 logging_writer.py:48] [79600] global_step=79600, grad_norm=5.372981071472168, loss=1.1386553049087524 +I0831 15:27:54.645976 140462291654400 logging_writer.py:48] [79700] global_step=79700, grad_norm=4.980198860168457, loss=1.0352798700332642 +I0831 15:28:21.844126 140462207792896 logging_writer.py:48] [79800] global_step=79800, grad_norm=4.922448635101318, loss=1.1133697032928467 +I0831 15:28:48.992383 140462291654400 logging_writer.py:48] [79900] global_step=79900, grad_norm=4.9384026527404785, loss=0.9916665554046631 +I0831 15:29:16.122765 140462207792896 logging_writer.py:48] [80000] global_step=80000, grad_norm=5.114055156707764, loss=1.071118950843811 +I0831 15:29:18.765294 140659750036672 spec.py:333] Evaluating on the training split. +I0831 15:29:26.674328 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 15:29:35.921750 140659750036672 spec.py:363] Evaluating on the test split. +I0831 15:29:36.795688 140659750036672 submission_runner.py:516] Time since start: 22297.97s, Step: 80011, {'train/accuracy': Array(0.9135244, dtype=float32), 'train/loss': Array(0.30929574, dtype=float32), 'validation/accuracy': Array(0.74368, dtype=float32), 'validation/loss': Array(1.0724281, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6151, dtype=float32), 'test/loss': Array(1.8198733, dtype=float32), 'test/num_examples': 10000, 'score': 22008.974818468094, 'total_duration': 22297.971310138702, 'accumulated_submission_time': 22008.974818468094, 'accumulated_eval_time': 287.4291217327118, 'accumulated_logging_time': 0.8914148807525635} +I0831 15:29:36.850972 140462291654400 logging_writer.py:48] [80011] accumulated_eval_time=287.429, accumulated_logging_time=0.891415, accumulated_submission_time=22009, global_step=80011, preemption_count=0, score=22009, test/accuracy=0.6151000261306763, test/loss=1.819873332977295, test/num_examples=10000, total_duration=22298, train/accuracy=0.9135243892669678, train/loss=0.30929574370384216, validation/accuracy=0.7436800003051758, validation/loss=1.0724281072616577, validation/num_examples=50000 +I0831 15:30:01.617244 140462207792896 logging_writer.py:48] [80100] global_step=80100, grad_norm=4.8027520179748535, loss=1.045777678489685 +I0831 15:30:28.742407 140462291654400 logging_writer.py:48] [80200] global_step=80200, grad_norm=5.046718597412109, loss=1.1183059215545654 +I0831 15:30:55.899578 140462207792896 logging_writer.py:48] [80300] global_step=80300, grad_norm=5.072199821472168, loss=1.0807229280471802 +I0831 15:31:23.081217 140462291654400 logging_writer.py:48] [80400] global_step=80400, grad_norm=5.102008819580078, loss=1.0879298448562622 +I0831 15:31:50.218806 140462207792896 logging_writer.py:48] [80500] global_step=80500, grad_norm=5.056987762451172, loss=1.091522455215454 +I0831 15:32:17.336322 140462291654400 logging_writer.py:48] [80600] global_step=80600, grad_norm=5.346099376678467, loss=1.1192913055419922 +I0831 15:32:44.546762 140462207792896 logging_writer.py:48] [80700] global_step=80700, grad_norm=4.863588809967041, loss=1.0515360832214355 +I0831 15:33:11.678821 140462291654400 logging_writer.py:48] [80800] global_step=80800, grad_norm=5.1354756355285645, loss=1.0947846174240112 +I0831 15:33:38.825002 140462207792896 logging_writer.py:48] [80900] global_step=80900, grad_norm=5.228660583496094, loss=1.0588380098342896 +I0831 15:34:06.012605 140462291654400 logging_writer.py:48] [81000] global_step=81000, grad_norm=5.4167938232421875, loss=1.1546671390533447 +I0831 15:34:33.182551 140462207792896 logging_writer.py:48] [81100] global_step=81100, grad_norm=5.204902648925781, loss=1.1178687810897827 +I0831 15:35:00.537090 140462291654400 logging_writer.py:48] [81200] global_step=81200, grad_norm=5.0533223152160645, loss=1.1036878824234009 +I0831 15:35:27.762565 140462207792896 logging_writer.py:48] [81300] global_step=81300, grad_norm=4.778297424316406, loss=1.0054988861083984 +I0831 15:35:54.919884 140462291654400 logging_writer.py:48] [81400] global_step=81400, grad_norm=5.092770099639893, loss=1.0561141967773438 +I0831 15:36:22.079777 140462207792896 logging_writer.py:48] [81500] global_step=81500, grad_norm=5.282026290893555, loss=1.1500883102416992 +I0831 15:36:49.272172 140462291654400 logging_writer.py:48] [81600] global_step=81600, grad_norm=4.808155059814453, loss=1.019740104675293 +I0831 15:37:16.387844 140462207792896 logging_writer.py:48] [81700] global_step=81700, grad_norm=4.8297200202941895, loss=1.033597707748413 +I0831 15:37:43.517898 140462291654400 logging_writer.py:48] [81800] global_step=81800, grad_norm=5.105103969573975, loss=1.115894079208374 +I0831 15:38:10.699128 140462207792896 logging_writer.py:48] [81900] global_step=81900, grad_norm=5.15748929977417, loss=1.0981687307357788 +I0831 15:38:37.852572 140462291654400 logging_writer.py:48] [82000] global_step=82000, grad_norm=5.11760950088501, loss=1.167235255241394 +I0831 15:39:04.975497 140462207792896 logging_writer.py:48] [82100] global_step=82100, grad_norm=5.21095609664917, loss=1.1144347190856934 +I0831 15:39:32.389033 140462291654400 logging_writer.py:48] [82200] global_step=82200, grad_norm=5.202498435974121, loss=1.1674529314041138 +I0831 15:39:59.522205 140462207792896 logging_writer.py:48] [82300] global_step=82300, grad_norm=5.336050987243652, loss=1.174316167831421 +I0831 15:40:26.658825 140462291654400 logging_writer.py:48] [82400] global_step=82400, grad_norm=5.025356769561768, loss=1.0778332948684692 +I0831 15:40:53.839304 140462207792896 logging_writer.py:48] [82500] global_step=82500, grad_norm=4.990633010864258, loss=1.0369369983673096 +I0831 15:41:20.979122 140462291654400 logging_writer.py:48] [82600] global_step=82600, grad_norm=4.913053035736084, loss=1.0315380096435547 +I0831 15:41:48.103095 140462207792896 logging_writer.py:48] [82700] global_step=82700, grad_norm=5.058833122253418, loss=1.1632330417633057 +I0831 15:42:15.287482 140462291654400 logging_writer.py:48] [82800] global_step=82800, grad_norm=5.111359119415283, loss=1.076669692993164 +I0831 15:42:42.414959 140462207792896 logging_writer.py:48] [82900] global_step=82900, grad_norm=5.015858173370361, loss=1.038299322128296 +I0831 15:43:09.528374 140462291654400 logging_writer.py:48] [83000] global_step=83000, grad_norm=5.048405647277832, loss=1.0787433385849 +I0831 15:43:36.712402 140462207792896 logging_writer.py:48] [83100] global_step=83100, grad_norm=4.75800085067749, loss=0.9571088552474976 +I0831 15:44:03.836605 140462291654400 logging_writer.py:48] [83200] global_step=83200, grad_norm=5.09819221496582, loss=1.1296072006225586 +I0831 15:44:31.234739 140462207792896 logging_writer.py:48] [83300] global_step=83300, grad_norm=5.04482889175415, loss=1.1123828887939453 +I0831 15:44:58.461814 140462291654400 logging_writer.py:48] [83400] global_step=83400, grad_norm=5.234004020690918, loss=1.1535555124282837 +I0831 15:45:25.612139 140462207792896 logging_writer.py:48] [83500] global_step=83500, grad_norm=5.191713333129883, loss=1.1817467212677002 +I0831 15:45:52.733690 140462291654400 logging_writer.py:48] [83600] global_step=83600, grad_norm=5.290879726409912, loss=1.0777533054351807 +I0831 15:46:19.938376 140462207792896 logging_writer.py:48] [83700] global_step=83700, grad_norm=4.93105411529541, loss=1.0401595830917358 +I0831 15:46:47.096294 140462291654400 logging_writer.py:48] [83800] global_step=83800, grad_norm=4.891636371612549, loss=1.0953097343444824 +I0831 15:47:14.231655 140462207792896 logging_writer.py:48] [83900] global_step=83900, grad_norm=5.457226753234863, loss=1.08694589138031 +I0831 15:47:41.430518 140462291654400 logging_writer.py:48] [84000] global_step=84000, grad_norm=5.129451274871826, loss=1.0509101152420044 +I0831 15:48:08.558040 140462207792896 logging_writer.py:48] [84100] global_step=84100, grad_norm=5.379550933837891, loss=1.1043866872787476 +I0831 15:48:35.681452 140462291654400 logging_writer.py:48] [84200] global_step=84200, grad_norm=4.720390319824219, loss=1.033921718597412 +I0831 15:49:02.871552 140462207792896 logging_writer.py:48] [84300] global_step=84300, grad_norm=4.95485782623291, loss=1.033087968826294 +I0831 15:49:30.229374 140462291654400 logging_writer.py:48] [84400] global_step=84400, grad_norm=5.368260383605957, loss=1.086872935295105 +I0831 15:49:57.351355 140462207792896 logging_writer.py:48] [84500] global_step=84500, grad_norm=4.818210601806641, loss=1.0919336080551147 +I0831 15:50:24.571552 140462291654400 logging_writer.py:48] [84600] global_step=84600, grad_norm=5.091428756713867, loss=1.1080976724624634 +I0831 15:50:51.736696 140462207792896 logging_writer.py:48] [84700] global_step=84700, grad_norm=5.3407440185546875, loss=1.1161068677902222 +I0831 15:51:18.892567 140462291654400 logging_writer.py:48] [84800] global_step=84800, grad_norm=5.057868003845215, loss=1.1248582601547241 +I0831 15:51:46.092302 140462207792896 logging_writer.py:48] [84900] global_step=84900, grad_norm=5.131648063659668, loss=1.1121399402618408 +I0831 15:52:13.235661 140462291654400 logging_writer.py:48] [85000] global_step=85000, grad_norm=5.133615493774414, loss=1.1190361976623535 +I0831 15:52:40.366699 140462207792896 logging_writer.py:48] [85100] global_step=85100, grad_norm=5.104923248291016, loss=1.1714637279510498 +I0831 15:53:07.564352 140462291654400 logging_writer.py:48] [85200] global_step=85200, grad_norm=5.167296409606934, loss=1.147531509399414 +I0831 15:53:34.718199 140462207792896 logging_writer.py:48] [85300] global_step=85300, grad_norm=5.2078657150268555, loss=1.0842106342315674 +I0831 15:54:01.866739 140462291654400 logging_writer.py:48] [85400] global_step=85400, grad_norm=5.184271335601807, loss=1.0012669563293457 +I0831 15:54:29.298123 140462207792896 logging_writer.py:48] [85500] global_step=85500, grad_norm=4.995631694793701, loss=1.098233699798584 +I0831 15:54:56.423478 140462291654400 logging_writer.py:48] [85600] global_step=85600, grad_norm=4.8185906410217285, loss=1.0647668838500977 +I0831 15:55:23.588460 140462207792896 logging_writer.py:48] [85700] global_step=85700, grad_norm=5.128810882568359, loss=1.0949413776397705 +I0831 15:55:50.781186 140462291654400 logging_writer.py:48] [85800] global_step=85800, grad_norm=5.428458213806152, loss=1.1746182441711426 +I0831 15:56:17.912009 140462207792896 logging_writer.py:48] [85900] global_step=85900, grad_norm=4.905691146850586, loss=1.0862361192703247 +I0831 15:56:45.037642 140462291654400 logging_writer.py:48] [86000] global_step=86000, grad_norm=5.135137557983398, loss=1.0044636726379395 +I0831 15:57:12.260535 140462207792896 logging_writer.py:48] [86100] global_step=86100, grad_norm=4.895088195800781, loss=1.0311821699142456 +I0831 15:57:39.376469 140462291654400 logging_writer.py:48] [86200] global_step=86200, grad_norm=4.936323642730713, loss=1.0615873336791992 +I0831 15:58:06.492154 140462207792896 logging_writer.py:48] [86300] global_step=86300, grad_norm=4.968416213989258, loss=1.0038541555404663 +I0831 15:58:33.673908 140462291654400 logging_writer.py:48] [86400] global_step=86400, grad_norm=4.9149651527404785, loss=1.0676205158233643 +I0831 15:59:01.015235 140462207792896 logging_writer.py:48] [86500] global_step=86500, grad_norm=4.952779769897461, loss=1.0510683059692383 +I0831 15:59:28.169139 140462291654400 logging_writer.py:48] [86600] global_step=86600, grad_norm=4.891543865203857, loss=1.1048495769500732 +I0831 15:59:55.385012 140462207792896 logging_writer.py:48] [86700] global_step=86700, grad_norm=5.10454797744751, loss=1.0697541236877441 +I0831 16:00:22.531288 140462291654400 logging_writer.py:48] [86800] global_step=86800, grad_norm=4.959455490112305, loss=1.1157441139221191 +I0831 16:00:49.642400 140462207792896 logging_writer.py:48] [86900] global_step=86900, grad_norm=5.217304229736328, loss=1.0510529279708862 +I0831 16:01:16.834686 140462291654400 logging_writer.py:48] [87000] global_step=87000, grad_norm=4.825283050537109, loss=1.0595011711120605 +I0831 16:01:43.961452 140462207792896 logging_writer.py:48] [87100] global_step=87100, grad_norm=5.501576900482178, loss=1.1701405048370361 +I0831 16:02:11.092808 140462291654400 logging_writer.py:48] [87200] global_step=87200, grad_norm=4.97939395904541, loss=1.101808786392212 +I0831 16:02:38.296216 140462207792896 logging_writer.py:48] [87300] global_step=87300, grad_norm=5.287903308868408, loss=1.0951745510101318 +I0831 16:02:52.862508 140659750036672 spec.py:333] Evaluating on the training split. +I0831 16:03:01.144877 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 16:03:10.828894 140659750036672 spec.py:363] Evaluating on the test split. +I0831 16:03:11.724396 140659750036672 submission_runner.py:516] Time since start: 24312.90s, Step: 87355, {'train/accuracy': Array(0.91770965, dtype=float32), 'train/loss': Array(0.29734606, dtype=float32), 'validation/accuracy': Array(0.74516, dtype=float32), 'validation/loss': Array(1.0671043, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.61770004, dtype=float32), 'test/loss': Array(1.810064, dtype=float32), 'test/num_examples': 10000, 'score': 24004.91706252098, 'total_duration': 24312.900218486786, 'accumulated_submission_time': 24004.91706252098, 'accumulated_eval_time': 306.2893109321594, 'accumulated_logging_time': 0.9574494361877441} +I0831 16:03:11.780621 140462291654400 logging_writer.py:48] [87355] accumulated_eval_time=306.289, accumulated_logging_time=0.957449, accumulated_submission_time=24004.9, global_step=87355, preemption_count=0, score=24004.9, test/accuracy=0.6177000403404236, test/loss=1.8100639581680298, test/num_examples=10000, total_duration=24312.9, train/accuracy=0.9177096486091614, train/loss=0.2973460555076599, validation/accuracy=0.7451599836349487, validation/loss=1.0671043395996094, validation/num_examples=50000 +I0831 16:03:24.374741 140462207792896 logging_writer.py:48] [87400] global_step=87400, grad_norm=5.097751617431641, loss=1.0515035390853882 +I0831 16:03:51.529090 140462291654400 logging_writer.py:48] [87500] global_step=87500, grad_norm=5.246204853057861, loss=1.0999469757080078 +I0831 16:04:18.952025 140462207792896 logging_writer.py:48] [87600] global_step=87600, grad_norm=5.357090950012207, loss=1.1811838150024414 +I0831 16:04:46.098130 140462291654400 logging_writer.py:48] [87700] global_step=87700, grad_norm=4.9145660400390625, loss=1.0151646137237549 +I0831 16:05:13.240762 140462207792896 logging_writer.py:48] [87800] global_step=87800, grad_norm=5.027377128601074, loss=1.055638313293457 +I0831 16:05:40.433928 140462291654400 logging_writer.py:48] [87900] global_step=87900, grad_norm=4.883982181549072, loss=1.0394787788391113 +I0831 16:06:07.584380 140462207792896 logging_writer.py:48] [88000] global_step=88000, grad_norm=5.289548397064209, loss=1.1517364978790283 +I0831 16:06:34.706968 140462291654400 logging_writer.py:48] [88100] global_step=88100, grad_norm=5.052430629730225, loss=1.1103522777557373 +I0831 16:07:01.929594 140462207792896 logging_writer.py:48] [88200] global_step=88200, grad_norm=4.837864398956299, loss=1.010026454925537 +I0831 16:07:29.082287 140462291654400 logging_writer.py:48] [88300] global_step=88300, grad_norm=5.097057342529297, loss=1.0487135648727417 +I0831 16:07:56.209052 140462207792896 logging_writer.py:48] [88400] global_step=88400, grad_norm=5.145213603973389, loss=1.0514527559280396 +I0831 16:08:23.380327 140462291654400 logging_writer.py:48] [88500] global_step=88500, grad_norm=4.792759418487549, loss=1.0134515762329102 +I0831 16:08:50.525143 140462207792896 logging_writer.py:48] [88600] global_step=88600, grad_norm=4.7926225662231445, loss=1.0294077396392822 +I0831 16:09:17.876160 140462291654400 logging_writer.py:48] [88700] global_step=88700, grad_norm=4.825915336608887, loss=1.0267112255096436 +I0831 16:09:45.094743 140462207792896 logging_writer.py:48] [88800] global_step=88800, grad_norm=5.055381774902344, loss=1.082041621208191 +I0831 16:10:12.237136 140462291654400 logging_writer.py:48] [88900] global_step=88900, grad_norm=5.174954414367676, loss=1.0908920764923096 +I0831 16:10:39.368040 140462207792896 logging_writer.py:48] [89000] global_step=89000, grad_norm=4.878984451293945, loss=0.9936631321907043 +I0831 16:11:06.566241 140462291654400 logging_writer.py:48] [89100] global_step=89100, grad_norm=5.28227424621582, loss=1.0856448411941528 +I0831 16:11:33.699414 140462207792896 logging_writer.py:48] [89200] global_step=89200, grad_norm=5.098367691040039, loss=1.0847740173339844 +I0831 16:12:00.873578 140462291654400 logging_writer.py:48] [89300] global_step=89300, grad_norm=5.023670673370361, loss=1.06284499168396 +I0831 16:12:28.104704 140462207792896 logging_writer.py:48] [89400] global_step=89400, grad_norm=5.0673346519470215, loss=1.0222907066345215 +I0831 16:12:55.220141 140462291654400 logging_writer.py:48] [89500] global_step=89500, grad_norm=4.917794704437256, loss=1.0454509258270264 +I0831 16:13:22.340825 140462207792896 logging_writer.py:48] [89600] global_step=89600, grad_norm=5.3382568359375, loss=1.089584469795227 +I0831 16:13:49.785715 140462291654400 logging_writer.py:48] [89700] global_step=89700, grad_norm=4.8083343505859375, loss=1.0991663932800293 +I0831 16:14:16.918957 140462207792896 logging_writer.py:48] [89800] global_step=89800, grad_norm=5.218383312225342, loss=1.0366439819335938 +I0831 16:14:44.071511 140462291654400 logging_writer.py:48] [89900] global_step=89900, grad_norm=4.818188667297363, loss=0.9921644926071167 +I0831 16:15:11.293351 140462207792896 logging_writer.py:48] [90000] global_step=90000, grad_norm=5.021610736846924, loss=1.0244442224502563 +I0831 16:15:38.451275 140462291654400 logging_writer.py:48] [90100] global_step=90100, grad_norm=4.8175554275512695, loss=1.0610618591308594 +I0831 16:16:05.600199 140462207792896 logging_writer.py:48] [90200] global_step=90200, grad_norm=5.161630630493164, loss=1.1447056531906128 +I0831 16:16:32.809495 140462291654400 logging_writer.py:48] [90300] global_step=90300, grad_norm=5.004881381988525, loss=1.0498486757278442 +I0831 16:16:59.963400 140462207792896 logging_writer.py:48] [90400] global_step=90400, grad_norm=4.866598606109619, loss=1.0832184553146362 +I0831 16:17:27.112154 140462291654400 logging_writer.py:48] [90500] global_step=90500, grad_norm=4.859375, loss=0.974939227104187 +I0831 16:17:54.319310 140462207792896 logging_writer.py:48] [90600] global_step=90600, grad_norm=4.931149005889893, loss=1.0548481941223145 +I0831 16:18:21.458143 140462291654400 logging_writer.py:48] [90700] global_step=90700, grad_norm=4.776697158813477, loss=1.0218379497528076 +I0831 16:18:48.850522 140462207792896 logging_writer.py:48] [90800] global_step=90800, grad_norm=5.250412940979004, loss=1.1257505416870117 +I0831 16:19:16.056849 140462291654400 logging_writer.py:48] [90900] global_step=90900, grad_norm=5.185441017150879, loss=1.107836127281189 +I0831 16:19:43.216965 140462207792896 logging_writer.py:48] [91000] global_step=91000, grad_norm=4.939932346343994, loss=1.0587016344070435 +I0831 16:20:10.357344 140462291654400 logging_writer.py:48] [91100] global_step=91100, grad_norm=4.819828510284424, loss=1.032155156135559 +I0831 16:20:37.581303 140462207792896 logging_writer.py:48] [91200] global_step=91200, grad_norm=4.939154624938965, loss=0.9641773700714111 +I0831 16:21:04.738502 140462291654400 logging_writer.py:48] [91300] global_step=91300, grad_norm=5.056154727935791, loss=1.0398766994476318 +I0831 16:21:31.877462 140462207792896 logging_writer.py:48] [91400] global_step=91400, grad_norm=4.9779582023620605, loss=1.0471256971359253 +I0831 16:21:59.106608 140462291654400 logging_writer.py:48] [91500] global_step=91500, grad_norm=5.169121742248535, loss=1.115097999572754 +I0831 16:22:26.265974 140462207792896 logging_writer.py:48] [91600] global_step=91600, grad_norm=4.681024551391602, loss=1.0649073123931885 +I0831 16:22:53.405616 140462291654400 logging_writer.py:48] [91700] global_step=91700, grad_norm=5.096673488616943, loss=1.1290011405944824 +I0831 16:23:20.609972 140462207792896 logging_writer.py:48] [91800] global_step=91800, grad_norm=4.86850118637085, loss=0.9855281710624695 +I0831 16:23:47.969150 140462291654400 logging_writer.py:48] [91900] global_step=91900, grad_norm=5.164760589599609, loss=1.125441312789917 +I0831 16:24:15.094579 140462207792896 logging_writer.py:48] [92000] global_step=92000, grad_norm=4.840758323669434, loss=0.9856946468353271 +I0831 16:24:42.286593 140462291654400 logging_writer.py:48] [92100] global_step=92100, grad_norm=5.013010025024414, loss=1.12363862991333 +I0831 16:25:09.448120 140462207792896 logging_writer.py:48] [92200] global_step=92200, grad_norm=4.896224498748779, loss=1.048984408378601 +I0831 16:25:36.564065 140462291654400 logging_writer.py:48] [92300] global_step=92300, grad_norm=4.991527557373047, loss=1.172895908355713 +I0831 16:26:03.773501 140462207792896 logging_writer.py:48] [92400] global_step=92400, grad_norm=5.214746952056885, loss=1.034113883972168 +I0831 16:26:30.888698 140462291654400 logging_writer.py:48] [92500] global_step=92500, grad_norm=4.787045478820801, loss=1.034144401550293 +I0831 16:26:58.046241 140462207792896 logging_writer.py:48] [92600] global_step=92600, grad_norm=5.005835056304932, loss=1.073025107383728 +I0831 16:27:25.257487 140462291654400 logging_writer.py:48] [92700] global_step=92700, grad_norm=4.922765731811523, loss=0.9844890832901001 +I0831 16:27:52.415655 140462207792896 logging_writer.py:48] [92800] global_step=92800, grad_norm=4.429115295410156, loss=1.0079073905944824 +I0831 16:28:19.551599 140462291654400 logging_writer.py:48] [92900] global_step=92900, grad_norm=5.14548397064209, loss=1.1472595930099487 +I0831 16:28:47.008821 140462207792896 logging_writer.py:48] [93000] global_step=93000, grad_norm=4.8970794677734375, loss=0.9901389479637146 +I0831 16:29:14.140476 140462291654400 logging_writer.py:48] [93100] global_step=93100, grad_norm=5.036914348602295, loss=1.1062043905258179 +I0831 16:29:41.257766 140462207792896 logging_writer.py:48] [93200] global_step=93200, grad_norm=5.011218547821045, loss=1.013047456741333 +I0831 16:30:08.457424 140462291654400 logging_writer.py:48] [93300] global_step=93300, grad_norm=4.794943809509277, loss=1.1009881496429443 +I0831 16:30:35.601506 140462207792896 logging_writer.py:48] [93400] global_step=93400, grad_norm=4.894553184509277, loss=1.0569958686828613 +I0831 16:31:02.728389 140462291654400 logging_writer.py:48] [93500] global_step=93500, grad_norm=5.044276714324951, loss=1.031606912612915 +I0831 16:31:29.914551 140462207792896 logging_writer.py:48] [93600] global_step=93600, grad_norm=4.995568752288818, loss=1.0005443096160889 +I0831 16:31:57.057391 140462291654400 logging_writer.py:48] [93700] global_step=93700, grad_norm=4.928337574005127, loss=1.0297882556915283 +I0831 16:32:24.207545 140462207792896 logging_writer.py:48] [93800] global_step=93800, grad_norm=5.06362771987915, loss=1.0843534469604492 +I0831 16:32:51.408195 140462291654400 logging_writer.py:48] [93900] global_step=93900, grad_norm=5.217005729675293, loss=1.1089762449264526 +I0831 16:33:18.751950 140462207792896 logging_writer.py:48] [94000] global_step=94000, grad_norm=5.003041744232178, loss=1.0580289363861084 +I0831 16:33:45.893544 140462291654400 logging_writer.py:48] [94100] global_step=94100, grad_norm=5.149652481079102, loss=1.1499342918395996 +I0831 16:34:13.104323 140462207792896 logging_writer.py:48] [94200] global_step=94200, grad_norm=5.130496501922607, loss=1.0756165981292725 +I0831 16:34:40.240749 140462291654400 logging_writer.py:48] [94300] global_step=94300, grad_norm=5.141345500946045, loss=1.089173436164856 +I0831 16:35:07.390180 140462207792896 logging_writer.py:48] [94400] global_step=94400, grad_norm=4.663873195648193, loss=0.9440207481384277 +I0831 16:35:34.612046 140462291654400 logging_writer.py:48] [94500] global_step=94500, grad_norm=5.062002182006836, loss=1.0581746101379395 +I0831 16:36:01.724695 140462207792896 logging_writer.py:48] [94600] global_step=94600, grad_norm=5.026883602142334, loss=1.084574818611145 +I0831 16:36:27.874607 140659750036672 spec.py:333] Evaluating on the training split. +I0831 16:36:35.168021 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 16:36:44.205514 140659750036672 spec.py:363] Evaluating on the test split. +I0831 16:36:45.085255 140659750036672 submission_runner.py:516] Time since start: 26326.26s, Step: 94698, {'train/accuracy': Array(0.92386794, dtype=float32), 'train/loss': Array(0.27556, dtype=float32), 'validation/accuracy': Array(0.74774, dtype=float32), 'validation/loss': Array(1.0638627, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62390006, dtype=float32), 'test/loss': Array(1.803928, dtype=float32), 'test/num_examples': 10000, 'score': 26000.928428173065, 'total_duration': 26326.260948896408, 'accumulated_submission_time': 26000.928428173065, 'accumulated_eval_time': 323.49813055992126, 'accumulated_logging_time': 1.0380325317382812} +I0831 16:36:45.139284 140462291654400 logging_writer.py:48] [94698] accumulated_eval_time=323.498, accumulated_logging_time=1.03803, accumulated_submission_time=26000.9, global_step=94698, preemption_count=0, score=26000.9, test/accuracy=0.6239000558853149, test/loss=1.803928017616272, test/num_examples=10000, total_duration=26326.3, train/accuracy=0.92386794090271, train/loss=0.2755599915981293, validation/accuracy=0.7477399706840515, validation/loss=1.063862681388855, validation/num_examples=50000 +I0831 16:36:46.109150 140462207792896 logging_writer.py:48] [94700] global_step=94700, grad_norm=4.693626880645752, loss=1.0532925128936768 +I0831 16:37:13.461462 140462291654400 logging_writer.py:48] [94800] global_step=94800, grad_norm=4.879063606262207, loss=1.0160030126571655 +I0831 16:37:40.604357 140462207792896 logging_writer.py:48] [94900] global_step=94900, grad_norm=5.000967502593994, loss=1.0718858242034912 +I0831 16:38:07.745402 140462291654400 logging_writer.py:48] [95000] global_step=95000, grad_norm=4.774775981903076, loss=1.0043113231658936 +I0831 16:38:35.178410 140462207792896 logging_writer.py:48] [95100] global_step=95100, grad_norm=5.064755916595459, loss=0.9909855127334595 +I0831 16:39:02.340879 140462291654400 logging_writer.py:48] [95200] global_step=95200, grad_norm=5.0424699783325195, loss=1.0640532970428467 +I0831 16:39:29.462306 140462207792896 logging_writer.py:48] [95300] global_step=95300, grad_norm=5.249161720275879, loss=1.056574821472168 +I0831 16:39:56.645437 140462291654400 logging_writer.py:48] [95400] global_step=95400, grad_norm=4.9380903244018555, loss=1.0589238405227661 +I0831 16:40:23.790142 140462207792896 logging_writer.py:48] [95500] global_step=95500, grad_norm=4.858255386352539, loss=1.0276727676391602 +I0831 16:40:50.906157 140462291654400 logging_writer.py:48] [95600] global_step=95600, grad_norm=4.797016620635986, loss=0.9872702360153198 +I0831 16:41:18.119009 140462207792896 logging_writer.py:48] [95700] global_step=95700, grad_norm=4.796022415161133, loss=0.9915856719017029 +I0831 16:41:45.248172 140462291654400 logging_writer.py:48] [95800] global_step=95800, grad_norm=5.291432857513428, loss=1.0377278327941895 +I0831 16:42:12.403145 140462207792896 logging_writer.py:48] [95900] global_step=95900, grad_norm=4.802250385284424, loss=0.9897311925888062 +I0831 16:42:39.583899 140462291654400 logging_writer.py:48] [96000] global_step=96000, grad_norm=4.935360908508301, loss=1.0220088958740234 +I0831 16:43:06.956920 140462207792896 logging_writer.py:48] [96100] global_step=96100, grad_norm=5.325441360473633, loss=1.0311238765716553 +I0831 16:43:34.088780 140462291654400 logging_writer.py:48] [96200] global_step=96200, grad_norm=5.183315277099609, loss=1.049618124961853 +I0831 16:44:01.257334 140462207792896 logging_writer.py:48] [96300] global_step=96300, grad_norm=4.870697021484375, loss=1.0344524383544922 +I0831 16:44:28.364415 140462291654400 logging_writer.py:48] [96400] global_step=96400, grad_norm=5.040430545806885, loss=1.0631110668182373 +I0831 16:44:55.507818 140462207792896 logging_writer.py:48] [96500] global_step=96500, grad_norm=4.8309102058410645, loss=1.0081449747085571 +I0831 16:45:22.869040 140462291654400 logging_writer.py:48] [96600] global_step=96600, grad_norm=4.776351451873779, loss=0.9766185879707336 +I0831 16:45:50.002841 140462207792896 logging_writer.py:48] [96700] global_step=96700, grad_norm=5.236298561096191, loss=1.1174023151397705 +I0831 16:46:17.135259 140462291654400 logging_writer.py:48] [96800] global_step=96800, grad_norm=4.806175231933594, loss=0.9855413436889648 +I0831 16:46:44.330199 140462207792896 logging_writer.py:48] [96900] global_step=96900, grad_norm=5.041355133056641, loss=1.0148332118988037 +I0831 16:47:11.478715 140462291654400 logging_writer.py:48] [97000] global_step=97000, grad_norm=4.95532751083374, loss=1.0531220436096191 +I0831 16:47:38.599319 140462207792896 logging_writer.py:48] [97100] global_step=97100, grad_norm=5.015015602111816, loss=1.0621381998062134 +I0831 16:48:06.037150 140462291654400 logging_writer.py:48] [97200] global_step=97200, grad_norm=4.979625225067139, loss=1.049323320388794 +I0831 16:48:33.182907 140462207792896 logging_writer.py:48] [97300] global_step=97300, grad_norm=4.83981990814209, loss=0.983441174030304 +I0831 16:49:00.307317 140462291654400 logging_writer.py:48] [97400] global_step=97400, grad_norm=5.0100507736206055, loss=1.0584750175476074 +I0831 16:49:27.543871 140462207792896 logging_writer.py:48] [97500] global_step=97500, grad_norm=4.880284309387207, loss=0.9604684114456177 +I0831 16:49:54.665338 140462291654400 logging_writer.py:48] [97600] global_step=97600, grad_norm=5.33800745010376, loss=1.0673964023590088 +I0831 16:50:21.797919 140462207792896 logging_writer.py:48] [97700] global_step=97700, grad_norm=5.18308162689209, loss=1.0756897926330566 +I0831 16:50:49.001393 140462291654400 logging_writer.py:48] [97800] global_step=97800, grad_norm=5.10719633102417, loss=1.06125807762146 +I0831 16:51:16.137428 140462207792896 logging_writer.py:48] [97900] global_step=97900, grad_norm=4.70042085647583, loss=0.9745508432388306 +I0831 16:51:43.249206 140462291654400 logging_writer.py:48] [98000] global_step=98000, grad_norm=4.956940650939941, loss=1.0125585794448853 +I0831 16:52:10.454811 140462207792896 logging_writer.py:48] [98100] global_step=98100, grad_norm=5.0980730056762695, loss=1.0574699640274048 +I0831 16:52:37.836997 140462291654400 logging_writer.py:48] [98200] global_step=98200, grad_norm=5.088790416717529, loss=1.0867928266525269 +I0831 16:53:04.944770 140462207792896 logging_writer.py:48] [98300] global_step=98300, grad_norm=4.759167671203613, loss=1.0208320617675781 +I0831 16:53:32.140799 140462291654400 logging_writer.py:48] [98400] global_step=98400, grad_norm=4.6556715965271, loss=0.9473191499710083 +I0831 16:53:59.274152 140462207792896 logging_writer.py:48] [98500] global_step=98500, grad_norm=4.935302257537842, loss=1.1145343780517578 +I0831 16:54:26.391455 140462291654400 logging_writer.py:48] [98600] global_step=98600, grad_norm=4.811570167541504, loss=0.9742432236671448 +I0831 16:54:53.594622 140462207792896 logging_writer.py:48] [98700] global_step=98700, grad_norm=4.968760013580322, loss=1.1077399253845215 +I0831 16:55:20.749787 140462291654400 logging_writer.py:48] [98800] global_step=98800, grad_norm=5.053279876708984, loss=1.0617077350616455 +I0831 16:55:47.862609 140462207792896 logging_writer.py:48] [98900] global_step=98900, grad_norm=5.211219310760498, loss=1.046743631362915 +I0831 16:56:15.079314 140462291654400 logging_writer.py:48] [99000] global_step=99000, grad_norm=4.954123497009277, loss=1.00326669216156 +I0831 16:56:42.203715 140462207792896 logging_writer.py:48] [99100] global_step=99100, grad_norm=5.07562255859375, loss=1.0120887756347656 +I0831 16:57:09.353038 140462291654400 logging_writer.py:48] [99200] global_step=99200, grad_norm=5.284984111785889, loss=1.0593984127044678 +I0831 16:57:36.764540 140462207792896 logging_writer.py:48] [99300] global_step=99300, grad_norm=4.817725658416748, loss=1.02742338180542 +I0831 16:58:03.898365 140462291654400 logging_writer.py:48] [99400] global_step=99400, grad_norm=5.198250770568848, loss=1.1574572324752808 +I0831 16:58:31.034865 140462207792896 logging_writer.py:48] [99500] global_step=99500, grad_norm=5.057433128356934, loss=1.0567950010299683 +I0831 16:58:58.253478 140462291654400 logging_writer.py:48] [99600] global_step=99600, grad_norm=4.997857093811035, loss=1.0588295459747314 +I0831 16:59:25.362139 140462207792896 logging_writer.py:48] [99700] global_step=99700, grad_norm=4.876481056213379, loss=1.0050923824310303 +I0831 16:59:52.478974 140462291654400 logging_writer.py:48] [99800] global_step=99800, grad_norm=5.1738362312316895, loss=1.077665090560913 +I0831 17:00:19.685299 140462207792896 logging_writer.py:48] [99900] global_step=99900, grad_norm=5.254868507385254, loss=1.0590457916259766 +I0831 17:00:46.814943 140462291654400 logging_writer.py:48] [100000] global_step=100000, grad_norm=5.156538963317871, loss=1.1239815950393677 +I0831 17:01:13.999451 140462207792896 logging_writer.py:48] [100100] global_step=100100, grad_norm=5.166530132293701, loss=1.0208284854888916 +I0831 17:01:41.183329 140462291654400 logging_writer.py:48] [100200] global_step=100200, grad_norm=4.9408674240112305, loss=1.014465093612671 +I0831 17:02:08.589393 140462207792896 logging_writer.py:48] [100300] global_step=100300, grad_norm=4.800274848937988, loss=0.9624548554420471 +I0831 17:02:35.769239 140462291654400 logging_writer.py:48] [100400] global_step=100400, grad_norm=4.928806304931641, loss=1.0362461805343628 +I0831 17:03:02.974946 140462207792896 logging_writer.py:48] [100500] global_step=100500, grad_norm=4.899048328399658, loss=1.0363547801971436 +I0831 17:03:30.127040 140462291654400 logging_writer.py:48] [100600] global_step=100600, grad_norm=5.029213905334473, loss=1.0065749883651733 +I0831 17:03:57.253608 140462207792896 logging_writer.py:48] [100700] global_step=100700, grad_norm=4.96403694152832, loss=0.9981410503387451 +I0831 17:04:24.457445 140462291654400 logging_writer.py:48] [100800] global_step=100800, grad_norm=4.981239318847656, loss=1.0644645690917969 +I0831 17:04:51.607629 140462207792896 logging_writer.py:48] [100900] global_step=100900, grad_norm=4.971870422363281, loss=1.040497899055481 +I0831 17:05:18.739074 140462291654400 logging_writer.py:48] [101000] global_step=101000, grad_norm=4.823514938354492, loss=1.059802532196045 +I0831 17:05:45.951379 140462207792896 logging_writer.py:48] [101100] global_step=101100, grad_norm=4.874724388122559, loss=1.0838130712509155 +I0831 17:06:13.095160 140462291654400 logging_writer.py:48] [101200] global_step=101200, grad_norm=5.009035587310791, loss=1.0701756477355957 +I0831 17:06:40.201687 140462207792896 logging_writer.py:48] [101300] global_step=101300, grad_norm=4.979429721832275, loss=1.1795549392700195 +I0831 17:07:07.625720 140462291654400 logging_writer.py:48] [101400] global_step=101400, grad_norm=5.2694010734558105, loss=1.0464379787445068 +I0831 17:07:34.786457 140462207792896 logging_writer.py:48] [101500] global_step=101500, grad_norm=4.92676305770874, loss=1.0811281204223633 +I0831 17:08:01.904951 140462291654400 logging_writer.py:48] [101600] global_step=101600, grad_norm=5.163987636566162, loss=1.1679179668426514 +I0831 17:08:29.140533 140462207792896 logging_writer.py:48] [101700] global_step=101700, grad_norm=5.066330432891846, loss=1.117128849029541 +I0831 17:08:56.246093 140462291654400 logging_writer.py:48] [101800] global_step=101800, grad_norm=4.916522979736328, loss=1.0274510383605957 +I0831 17:09:23.374877 140462207792896 logging_writer.py:48] [101900] global_step=101900, grad_norm=5.014157772064209, loss=1.112300157546997 +I0831 17:09:50.562090 140462291654400 logging_writer.py:48] [102000] global_step=102000, grad_norm=4.947015285491943, loss=1.1072766780853271 +I0831 17:10:01.282083 140659750036672 spec.py:333] Evaluating on the training split. +I0831 17:10:08.208855 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 17:10:17.360839 140659750036672 spec.py:363] Evaluating on the test split. +I0831 17:10:18.247520 140659750036672 submission_runner.py:516] Time since start: 28339.42s, Step: 102041, {'train/accuracy': Array(0.92610013, dtype=float32), 'train/loss': Array(0.263924, dtype=float32), 'validation/accuracy': Array(0.74884, dtype=float32), 'validation/loss': Array(1.0614623, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62560004, dtype=float32), 'test/loss': Array(1.8158712, dtype=float32), 'test/num_examples': 10000, 'score': 27996.96565246582, 'total_duration': 28339.422913074493, 'accumulated_submission_time': 27996.96565246582, 'accumulated_eval_time': 340.46144580841064, 'accumulated_logging_time': 1.1386680603027344} +I0831 17:10:18.298198 140462207792896 logging_writer.py:48] [102041] accumulated_eval_time=340.461, accumulated_logging_time=1.13867, accumulated_submission_time=27997, global_step=102041, preemption_count=0, score=27997, test/accuracy=0.6256000399589539, test/loss=1.815871238708496, test/num_examples=10000, total_duration=28339.4, train/accuracy=0.9261001348495483, train/loss=0.2639240026473999, validation/accuracy=0.7488399744033813, validation/loss=1.0614622831344604, validation/num_examples=50000 +I0831 17:10:34.747519 140462291654400 logging_writer.py:48] [102100] global_step=102100, grad_norm=5.194640159606934, loss=1.0292339324951172 +I0831 17:11:01.912324 140462207792896 logging_writer.py:48] [102200] global_step=102200, grad_norm=4.505799770355225, loss=1.02686607837677 +I0831 17:11:29.117628 140462291654400 logging_writer.py:48] [102300] global_step=102300, grad_norm=5.007357120513916, loss=1.1097944974899292 +I0831 17:11:56.266842 140462207792896 logging_writer.py:48] [102400] global_step=102400, grad_norm=4.783303737640381, loss=1.0392441749572754 +I0831 17:12:23.665950 140462291654400 logging_writer.py:48] [102500] global_step=102500, grad_norm=4.872734546661377, loss=0.9921414852142334 +I0831 17:12:50.844650 140462207792896 logging_writer.py:48] [102600] global_step=102600, grad_norm=5.075994968414307, loss=1.1613836288452148 +I0831 17:13:17.965412 140462291654400 logging_writer.py:48] [102700] global_step=102700, grad_norm=4.865532875061035, loss=0.9526028633117676 +I0831 17:13:45.089545 140462207792896 logging_writer.py:48] [102800] global_step=102800, grad_norm=4.719111442565918, loss=1.0490633249282837 +I0831 17:14:12.303377 140462291654400 logging_writer.py:48] [102900] global_step=102900, grad_norm=5.235545635223389, loss=1.0883257389068604 +I0831 17:14:39.436825 140462207792896 logging_writer.py:48] [103000] global_step=103000, grad_norm=4.637126445770264, loss=0.9903701543807983 +I0831 17:15:06.587292 140462291654400 logging_writer.py:48] [103100] global_step=103100, grad_norm=4.927246570587158, loss=1.0158581733703613 +I0831 17:15:33.786654 140462207792896 logging_writer.py:48] [103200] global_step=103200, grad_norm=4.819151401519775, loss=0.9746272563934326 +I0831 17:16:00.921820 140462291654400 logging_writer.py:48] [103300] global_step=103300, grad_norm=5.181046009063721, loss=1.007714867591858 +I0831 17:16:28.023384 140462207792896 logging_writer.py:48] [103400] global_step=103400, grad_norm=5.090240001678467, loss=1.051032543182373 +I0831 17:16:55.464096 140462291654400 logging_writer.py:48] [103500] global_step=103500, grad_norm=4.667335033416748, loss=0.9786435961723328 +I0831 17:17:22.635556 140462207792896 logging_writer.py:48] [103600] global_step=103600, grad_norm=4.991156578063965, loss=1.1562021970748901 +I0831 17:17:49.812985 140462291654400 logging_writer.py:48] [103700] global_step=103700, grad_norm=4.834035396575928, loss=1.0048906803131104 +I0831 17:18:17.005521 140462207792896 logging_writer.py:48] [103800] global_step=103800, grad_norm=4.886865615844727, loss=1.0323644876480103 +I0831 17:18:44.116787 140462291654400 logging_writer.py:48] [103900] global_step=103900, grad_norm=5.111143112182617, loss=1.031903624534607 +I0831 17:19:11.242412 140462207792896 logging_writer.py:48] [104000] global_step=104000, grad_norm=5.121515274047852, loss=1.0809403657913208 +I0831 17:19:38.426764 140462291654400 logging_writer.py:48] [104100] global_step=104100, grad_norm=5.078837871551514, loss=1.0866223573684692 +I0831 17:20:05.561338 140462207792896 logging_writer.py:48] [104200] global_step=104200, grad_norm=4.878502368927002, loss=1.0404222011566162 +I0831 17:20:32.680645 140462291654400 logging_writer.py:48] [104300] global_step=104300, grad_norm=4.860300540924072, loss=0.9979685544967651 +I0831 17:20:59.882358 140462207792896 logging_writer.py:48] [104400] global_step=104400, grad_norm=4.99242639541626, loss=1.0561573505401611 +I0831 17:21:27.018528 140462291654400 logging_writer.py:48] [104500] global_step=104500, grad_norm=4.977812767028809, loss=1.0323872566223145 +I0831 17:21:54.396751 140462207792896 logging_writer.py:48] [104600] global_step=104600, grad_norm=4.967808246612549, loss=1.0720524787902832 +I0831 17:22:21.607487 140462291654400 logging_writer.py:48] [104700] global_step=104700, grad_norm=4.823157787322998, loss=1.0309654474258423 +I0831 17:22:48.770322 140462207792896 logging_writer.py:48] [104800] global_step=104800, grad_norm=4.807140827178955, loss=0.9666155576705933 +I0831 17:23:15.912987 140462291654400 logging_writer.py:48] [104900] global_step=104900, grad_norm=5.0314788818359375, loss=1.1361863613128662 +I0831 17:23:43.090359 140462207792896 logging_writer.py:48] [105000] global_step=105000, grad_norm=5.017558574676514, loss=1.0930137634277344 +I0831 17:24:10.233392 140462291654400 logging_writer.py:48] [105100] global_step=105100, grad_norm=4.765532493591309, loss=0.921245813369751 +I0831 17:24:37.363541 140462207792896 logging_writer.py:48] [105200] global_step=105200, grad_norm=4.807982921600342, loss=1.035590648651123 +I0831 17:25:04.564896 140462291654400 logging_writer.py:48] [105300] global_step=105300, grad_norm=4.958456993103027, loss=1.0513955354690552 +I0831 17:25:31.723584 140462207792896 logging_writer.py:48] [105400] global_step=105400, grad_norm=4.958673000335693, loss=1.0388975143432617 +I0831 17:25:58.855403 140462291654400 logging_writer.py:48] [105500] global_step=105500, grad_norm=4.99260139465332, loss=1.1168696880340576 +I0831 17:26:26.052479 140462207792896 logging_writer.py:48] [105600] global_step=105600, grad_norm=5.213581562042236, loss=1.0397772789001465 +I0831 17:26:53.377607 140462291654400 logging_writer.py:48] [105700] global_step=105700, grad_norm=4.998309135437012, loss=1.094983458518982 +I0831 17:27:20.520775 140462207792896 logging_writer.py:48] [105800] global_step=105800, grad_norm=4.999780178070068, loss=1.0302159786224365 +I0831 17:27:47.709553 140462291654400 logging_writer.py:48] [105900] global_step=105900, grad_norm=5.054821968078613, loss=1.0754841566085815 +I0831 17:28:14.850373 140462207792896 logging_writer.py:48] [106000] global_step=106000, grad_norm=5.218051910400391, loss=1.0089396238327026 +I0831 17:28:41.977255 140462291654400 logging_writer.py:48] [106100] global_step=106100, grad_norm=4.951213359832764, loss=0.9678333401679993 +I0831 17:29:09.185509 140462207792896 logging_writer.py:48] [106200] global_step=106200, grad_norm=4.724476337432861, loss=1.026360034942627 +I0831 17:29:36.317423 140462291654400 logging_writer.py:48] [106300] global_step=106300, grad_norm=5.0774760246276855, loss=1.0413904190063477 +I0831 17:30:03.473923 140462207792896 logging_writer.py:48] [106400] global_step=106400, grad_norm=4.756595134735107, loss=0.9894619584083557 +I0831 17:30:30.694685 140462291654400 logging_writer.py:48] [106500] global_step=106500, grad_norm=5.214836597442627, loss=1.0083906650543213 +I0831 17:30:57.825230 140462207792896 logging_writer.py:48] [106600] global_step=106600, grad_norm=5.118247032165527, loss=1.0497236251831055 +I0831 17:31:24.950820 140462291654400 logging_writer.py:48] [106700] global_step=106700, grad_norm=4.711628437042236, loss=0.9605365991592407 +I0831 17:31:52.391185 140462207792896 logging_writer.py:48] [106800] global_step=106800, grad_norm=4.794162750244141, loss=1.0743898153305054 +I0831 17:32:19.541456 140462291654400 logging_writer.py:48] [106900] global_step=106900, grad_norm=4.751049518585205, loss=0.9680519104003906 +I0831 17:32:46.660308 140462207792896 logging_writer.py:48] [107000] global_step=107000, grad_norm=5.108840465545654, loss=0.959273099899292 +I0831 17:33:13.877080 140462291654400 logging_writer.py:48] [107100] global_step=107100, grad_norm=4.834384918212891, loss=0.9728933572769165 +I0831 17:33:41.014424 140462207792896 logging_writer.py:48] [107200] global_step=107200, grad_norm=4.594396114349365, loss=0.9436824917793274 +I0831 17:34:08.144736 140462291654400 logging_writer.py:48] [107300] global_step=107300, grad_norm=5.058049201965332, loss=1.004830002784729 +I0831 17:34:35.343692 140462207792896 logging_writer.py:48] [107400] global_step=107400, grad_norm=5.0524067878723145, loss=0.9749428033828735 +I0831 17:35:02.478272 140462291654400 logging_writer.py:48] [107500] global_step=107500, grad_norm=5.125331878662109, loss=1.0057857036590576 +I0831 17:35:29.643676 140462207792896 logging_writer.py:48] [107600] global_step=107600, grad_norm=5.085574150085449, loss=1.1062045097351074 +I0831 17:35:56.847878 140462291654400 logging_writer.py:48] [107700] global_step=107700, grad_norm=4.864752769470215, loss=1.0218298435211182 +I0831 17:36:24.238192 140462207792896 logging_writer.py:48] [107800] global_step=107800, grad_norm=4.866130352020264, loss=0.9534983038902283 +I0831 17:36:51.368727 140462291654400 logging_writer.py:48] [107900] global_step=107900, grad_norm=4.741833686828613, loss=1.0419350862503052 +I0831 17:37:18.578632 140462207792896 logging_writer.py:48] [108000] global_step=108000, grad_norm=5.009776592254639, loss=1.113434076309204 +I0831 17:37:45.712571 140462291654400 logging_writer.py:48] [108100] global_step=108100, grad_norm=4.988548278808594, loss=1.0677930116653442 +I0831 17:38:12.843454 140462207792896 logging_writer.py:48] [108200] global_step=108200, grad_norm=4.7804741859436035, loss=1.0237528085708618 +I0831 17:38:40.063431 140462291654400 logging_writer.py:48] [108300] global_step=108300, grad_norm=4.602092266082764, loss=0.9573655128479004 +I0831 17:39:07.218791 140462207792896 logging_writer.py:48] [108400] global_step=108400, grad_norm=5.151592254638672, loss=1.0584150552749634 +I0831 17:39:34.334255 140462291654400 logging_writer.py:48] [108500] global_step=108500, grad_norm=5.390037536621094, loss=1.126624345779419 +I0831 17:40:01.544644 140462207792896 logging_writer.py:48] [108600] global_step=108600, grad_norm=5.1278157234191895, loss=1.0788651704788208 +I0831 17:40:28.659284 140462291654400 logging_writer.py:48] [108700] global_step=108700, grad_norm=4.837026119232178, loss=1.0168812274932861 +I0831 17:40:55.779689 140462207792896 logging_writer.py:48] [108800] global_step=108800, grad_norm=5.074102878570557, loss=0.9916492104530334 +I0831 17:41:23.194958 140462291654400 logging_writer.py:48] [108900] global_step=108900, grad_norm=4.991849422454834, loss=0.9836401343345642 +I0831 17:41:50.311203 140462207792896 logging_writer.py:48] [109000] global_step=109000, grad_norm=4.822633266448975, loss=0.9902591705322266 +I0831 17:42:17.442381 140462291654400 logging_writer.py:48] [109100] global_step=109100, grad_norm=4.9695234298706055, loss=1.0588710308074951 +I0831 17:42:44.635595 140462207792896 logging_writer.py:48] [109200] global_step=109200, grad_norm=4.868171215057373, loss=1.0034327507019043 +I0831 17:43:11.758167 140462291654400 logging_writer.py:48] [109300] global_step=109300, grad_norm=4.713303565979004, loss=1.026906967163086 +I0831 17:43:34.419709 140659750036672 spec.py:333] Evaluating on the training split. +I0831 17:43:40.796582 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 17:43:50.089042 140659750036672 spec.py:363] Evaluating on the test split. +I0831 17:43:50.956459 140659750036672 submission_runner.py:516] Time since start: 30352.13s, Step: 109385, {'train/accuracy': Array(0.92875075, dtype=float32), 'train/loss': Array(0.25364253, dtype=float32), 'validation/accuracy': Array(0.74917996, dtype=float32), 'validation/loss': Array(1.0621685, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6247, dtype=float32), 'test/loss': Array(1.8259393, dtype=float32), 'test/num_examples': 10000, 'score': 29992.987577676773, 'total_duration': 30352.131969690323, 'accumulated_submission_time': 29992.987577676773, 'accumulated_eval_time': 356.9961848258972, 'accumulated_logging_time': 1.2308220863342285} +I0831 17:43:51.015804 140462207792896 logging_writer.py:48] [109385] accumulated_eval_time=356.996, accumulated_logging_time=1.23082, accumulated_submission_time=29993, global_step=109385, preemption_count=0, score=29993, test/accuracy=0.6247000098228455, test/loss=1.8259392976760864, test/num_examples=10000, total_duration=30352.1, train/accuracy=0.92875075340271, train/loss=0.2536425292491913, validation/accuracy=0.7491799592971802, validation/loss=1.0621684789657593, validation/num_examples=50000 +I0831 17:43:55.458451 140462291654400 logging_writer.py:48] [109400] global_step=109400, grad_norm=5.2348313331604, loss=1.100180745124817 +I0831 17:44:22.755216 140462207792896 logging_writer.py:48] [109500] global_step=109500, grad_norm=4.930593967437744, loss=1.1189113855361938 +I0831 17:44:49.929155 140462291654400 logging_writer.py:48] [109600] global_step=109600, grad_norm=4.952072620391846, loss=1.0543146133422852 +I0831 17:45:17.061769 140462207792896 logging_writer.py:48] [109700] global_step=109700, grad_norm=4.6407294273376465, loss=0.9559882879257202 +I0831 17:45:44.244615 140462291654400 logging_writer.py:48] [109800] global_step=109800, grad_norm=5.113948822021484, loss=1.0318955183029175 +I0831 17:46:11.588959 140462207792896 logging_writer.py:48] [109900] global_step=109900, grad_norm=4.765798091888428, loss=1.0029258728027344 +I0831 17:46:38.720394 140462291654400 logging_writer.py:48] [110000] global_step=110000, grad_norm=4.617291450500488, loss=0.967522919178009 +I0831 17:47:05.916826 140462207792896 logging_writer.py:48] [110100] global_step=110100, grad_norm=5.245910167694092, loss=1.129124402999878 +I0831 17:47:33.030663 140462291654400 logging_writer.py:48] [110200] global_step=110200, grad_norm=5.056057929992676, loss=1.037756085395813 +I0831 17:48:00.155939 140462207792896 logging_writer.py:48] [110300] global_step=110300, grad_norm=5.033859729766846, loss=1.05642569065094 +I0831 17:48:27.347037 140462291654400 logging_writer.py:48] [110400] global_step=110400, grad_norm=4.794990539550781, loss=0.9899466037750244 +I0831 17:48:54.459671 140462207792896 logging_writer.py:48] [110500] global_step=110500, grad_norm=5.125531196594238, loss=1.1509652137756348 +I0831 17:49:21.589198 140462291654400 logging_writer.py:48] [110600] global_step=110600, grad_norm=4.733424663543701, loss=0.9854485988616943 +I0831 17:49:48.759018 140462207792896 logging_writer.py:48] [110700] global_step=110700, grad_norm=4.972171783447266, loss=1.028668761253357 +I0831 17:50:15.923690 140462291654400 logging_writer.py:48] [110800] global_step=110800, grad_norm=4.7240214347839355, loss=1.0347086191177368 +I0831 17:50:43.060151 140462207792896 logging_writer.py:48] [110900] global_step=110900, grad_norm=4.869174480438232, loss=1.0984965562820435 +I0831 17:51:10.453750 140462291654400 logging_writer.py:48] [111000] global_step=111000, grad_norm=5.132298946380615, loss=1.06745445728302 +I0831 17:51:37.581410 140462207792896 logging_writer.py:48] [111100] global_step=111100, grad_norm=4.818721771240234, loss=1.0622485876083374 +I0831 17:52:04.698608 140462291654400 logging_writer.py:48] [111200] global_step=111200, grad_norm=4.660581588745117, loss=1.0363316535949707 +I0831 17:52:31.889219 140462207792896 logging_writer.py:48] [111300] global_step=111300, grad_norm=4.924393177032471, loss=1.0118135213851929 +I0831 17:52:59.029268 140462291654400 logging_writer.py:48] [111400] global_step=111400, grad_norm=4.6164350509643555, loss=0.9029459357261658 +I0831 17:53:26.130348 140462207792896 logging_writer.py:48] [111500] global_step=111500, grad_norm=4.723528861999512, loss=0.9464471340179443 +I0831 17:53:53.333247 140462291654400 logging_writer.py:48] [111600] global_step=111600, grad_norm=4.776853084564209, loss=1.02729332447052 +I0831 17:54:20.461332 140462207792896 logging_writer.py:48] [111700] global_step=111700, grad_norm=4.847368240356445, loss=0.9946446418762207 +I0831 17:54:47.620876 140462291654400 logging_writer.py:48] [111800] global_step=111800, grad_norm=4.907256603240967, loss=0.9636017084121704 +I0831 17:55:14.844494 140462207792896 logging_writer.py:48] [111900] global_step=111900, grad_norm=5.050285816192627, loss=1.0265753269195557 +I0831 17:55:41.953633 140462291654400 logging_writer.py:48] [112000] global_step=112000, grad_norm=4.869489669799805, loss=0.9594745635986328 +I0831 17:56:09.342247 140462207792896 logging_writer.py:48] [112100] global_step=112100, grad_norm=4.814338684082031, loss=0.9044790267944336 +I0831 17:56:36.537792 140462291654400 logging_writer.py:48] [112200] global_step=112200, grad_norm=4.6746039390563965, loss=0.9967349171638489 +I0831 17:57:03.664834 140462207792896 logging_writer.py:48] [112300] global_step=112300, grad_norm=4.903285503387451, loss=0.9686717987060547 +I0831 17:57:30.771883 140462291654400 logging_writer.py:48] [112400] global_step=112400, grad_norm=4.722811698913574, loss=0.9718879461288452 +I0831 17:57:57.960464 140462207792896 logging_writer.py:48] [112500] global_step=112500, grad_norm=5.48336935043335, loss=1.036401629447937 +I0831 17:58:25.099525 140462291654400 logging_writer.py:48] [112600] global_step=112600, grad_norm=4.9437103271484375, loss=1.0585964918136597 +I0831 17:58:52.225186 140462207792896 logging_writer.py:48] [112700] global_step=112700, grad_norm=4.9874958992004395, loss=0.9870588779449463 +I0831 17:59:19.459472 140462291654400 logging_writer.py:48] [112800] global_step=112800, grad_norm=4.929352283477783, loss=0.9841917753219604 +I0831 17:59:46.598329 140462207792896 logging_writer.py:48] [112900] global_step=112900, grad_norm=5.026012897491455, loss=1.0115678310394287 +I0831 18:00:13.749101 140462291654400 logging_writer.py:48] [113000] global_step=113000, grad_norm=4.7503252029418945, loss=1.0415900945663452 +I0831 18:00:41.151221 140462207792896 logging_writer.py:48] [113100] global_step=113100, grad_norm=5.075703144073486, loss=1.0444105863571167 +I0831 18:01:08.299482 140462291654400 logging_writer.py:48] [113200] global_step=113200, grad_norm=5.126249313354492, loss=1.0597550868988037 +I0831 18:01:35.414740 140462207792896 logging_writer.py:48] [113300] global_step=113300, grad_norm=4.64411735534668, loss=1.0029047727584839 +I0831 18:02:02.626124 140462291654400 logging_writer.py:48] [113400] global_step=113400, grad_norm=4.862545013427734, loss=0.9639256596565247 +I0831 18:02:29.746487 140462207792896 logging_writer.py:48] [113500] global_step=113500, grad_norm=5.033511638641357, loss=0.9984768629074097 +I0831 18:02:56.882824 140462291654400 logging_writer.py:48] [113600] global_step=113600, grad_norm=5.020736217498779, loss=1.0135153532028198 +I0831 18:03:24.080148 140462207792896 logging_writer.py:48] [113700] global_step=113700, grad_norm=4.870365142822266, loss=1.0113880634307861 +I0831 18:03:51.265471 140462291654400 logging_writer.py:48] [113800] global_step=113800, grad_norm=5.060053825378418, loss=0.9979097247123718 +I0831 18:04:18.402516 140462207792896 logging_writer.py:48] [113900] global_step=113900, grad_norm=4.44724178314209, loss=0.8935201168060303 +I0831 18:04:45.573301 140462291654400 logging_writer.py:48] [114000] global_step=114000, grad_norm=4.915965557098389, loss=1.0460951328277588 +I0831 18:05:12.691894 140462207792896 logging_writer.py:48] [114100] global_step=114100, grad_norm=5.042156219482422, loss=1.0298449993133545 +I0831 18:05:40.049712 140462291654400 logging_writer.py:48] [114200] global_step=114200, grad_norm=4.825747489929199, loss=0.9783709049224854 +I0831 18:06:07.229893 140462207792896 logging_writer.py:48] [114300] global_step=114300, grad_norm=4.850982666015625, loss=1.0407823324203491 +I0831 18:06:34.347345 140462291654400 logging_writer.py:48] [114400] global_step=114400, grad_norm=4.737436771392822, loss=0.9647795557975769 +I0831 18:07:01.470637 140462207792896 logging_writer.py:48] [114500] global_step=114500, grad_norm=4.80708122253418, loss=1.024256706237793 +I0831 18:07:28.643766 140462291654400 logging_writer.py:48] [114600] global_step=114600, grad_norm=5.192071914672852, loss=1.0884548425674438 +I0831 18:07:55.771061 140462207792896 logging_writer.py:48] [114700] global_step=114700, grad_norm=4.941633701324463, loss=0.9405542016029358 +I0831 18:08:22.911675 140462291654400 logging_writer.py:48] [114800] global_step=114800, grad_norm=5.015111446380615, loss=1.0538872480392456 +I0831 18:08:50.105256 140462207792896 logging_writer.py:48] [114900] global_step=114900, grad_norm=5.086458683013916, loss=0.9753001928329468 +I0831 18:09:17.232752 140462291654400 logging_writer.py:48] [115000] global_step=115000, grad_norm=4.889427661895752, loss=0.9397754073143005 +I0831 18:09:44.396610 140462207792896 logging_writer.py:48] [115100] global_step=115100, grad_norm=4.955015182495117, loss=1.0586707592010498 +I0831 18:10:11.581874 140462291654400 logging_writer.py:48] [115200] global_step=115200, grad_norm=4.954550743103027, loss=0.9569666385650635 +I0831 18:10:38.940376 140462207792896 logging_writer.py:48] [115300] global_step=115300, grad_norm=4.884578704833984, loss=1.0643097162246704 +I0831 18:11:06.041926 140462291654400 logging_writer.py:48] [115400] global_step=115400, grad_norm=4.894836902618408, loss=0.9695751070976257 +I0831 18:11:33.231553 140462207792896 logging_writer.py:48] [115500] global_step=115500, grad_norm=5.12708044052124, loss=1.017775535583496 +I0831 18:12:00.358731 140462291654400 logging_writer.py:48] [115600] global_step=115600, grad_norm=4.726414203643799, loss=1.0506479740142822 +I0831 18:12:27.486851 140462207792896 logging_writer.py:48] [115700] global_step=115700, grad_norm=5.102452278137207, loss=1.111606240272522 +I0831 18:12:54.691140 140462291654400 logging_writer.py:48] [115800] global_step=115800, grad_norm=4.768645763397217, loss=0.9927196502685547 +I0831 18:13:21.827167 140462207792896 logging_writer.py:48] [115900] global_step=115900, grad_norm=4.824051380157471, loss=0.9528430700302124 +I0831 18:13:48.946635 140462291654400 logging_writer.py:48] [116000] global_step=116000, grad_norm=5.006685256958008, loss=1.04258131980896 +I0831 18:14:16.138865 140462207792896 logging_writer.py:48] [116100] global_step=116100, grad_norm=5.072755813598633, loss=1.0347421169281006 +I0831 18:14:43.273213 140462291654400 logging_writer.py:48] [116200] global_step=116200, grad_norm=4.892188549041748, loss=1.0289825201034546 +I0831 18:15:10.391015 140462207792896 logging_writer.py:48] [116300] global_step=116300, grad_norm=5.1061320304870605, loss=1.048293948173523 +I0831 18:15:37.845779 140462291654400 logging_writer.py:48] [116400] global_step=116400, grad_norm=4.726328372955322, loss=0.9641531109809875 +I0831 18:16:04.964305 140462207792896 logging_writer.py:48] [116500] global_step=116500, grad_norm=5.116252899169922, loss=1.0062456130981445 +I0831 18:16:32.095300 140462291654400 logging_writer.py:48] [116600] global_step=116600, grad_norm=5.018049716949463, loss=0.9801557064056396 +I0831 18:16:59.323729 140462207792896 logging_writer.py:48] [116700] global_step=116700, grad_norm=4.790430545806885, loss=0.9435322880744934 +I0831 18:17:07.056774 140659750036672 spec.py:333] Evaluating on the training split. +I0831 18:17:13.462018 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 18:17:22.634654 140659750036672 spec.py:363] Evaluating on the test split. +I0831 18:17:23.516817 140659750036672 submission_runner.py:516] Time since start: 32364.69s, Step: 116730, {'train/accuracy': Array(0.9323382, dtype=float32), 'train/loss': Array(0.23542856, dtype=float32), 'validation/accuracy': Array(0.75158, dtype=float32), 'validation/loss': Array(1.0637244, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62320006, dtype=float32), 'test/loss': Array(1.8310074, dtype=float32), 'test/num_examples': 10000, 'score': 31988.944032669067, 'total_duration': 32364.69255375862, 'accumulated_submission_time': 31988.944032669067, 'accumulated_eval_time': 373.45445227622986, 'accumulated_logging_time': 1.3163821697235107} +I0831 18:17:23.572537 140462291654400 logging_writer.py:48] [116730] accumulated_eval_time=373.454, accumulated_logging_time=1.31638, accumulated_submission_time=31988.9, global_step=116730, preemption_count=0, score=31988.9, test/accuracy=0.6232000589370728, test/loss=1.8310073614120483, test/num_examples=10000, total_duration=32364.7, train/accuracy=0.9323381781578064, train/loss=0.2354285567998886, validation/accuracy=0.751579999923706, validation/loss=1.063724398612976, validation/num_examples=50000 +I0831 18:17:43.061893 140462207792896 logging_writer.py:48] [116800] global_step=116800, grad_norm=5.058164119720459, loss=1.1141021251678467 +I0831 18:18:10.201099 140462291654400 logging_writer.py:48] [116900] global_step=116900, grad_norm=4.602115631103516, loss=0.8961552381515503 +I0831 18:18:37.400765 140462207792896 logging_writer.py:48] [117000] global_step=117000, grad_norm=4.886359691619873, loss=1.0001276731491089 +I0831 18:19:04.533174 140462291654400 logging_writer.py:48] [117100] global_step=117100, grad_norm=4.85435676574707, loss=0.9666292667388916 +I0831 18:19:31.661381 140462207792896 logging_writer.py:48] [117200] global_step=117200, grad_norm=4.9128618240356445, loss=1.0565216541290283 +I0831 18:19:58.850826 140462291654400 logging_writer.py:48] [117300] global_step=117300, grad_norm=5.048862457275391, loss=0.9992228746414185 +I0831 18:20:26.201205 140462207792896 logging_writer.py:48] [117400] global_step=117400, grad_norm=5.0836663246154785, loss=1.0276873111724854 +I0831 18:20:53.333193 140462291654400 logging_writer.py:48] [117500] global_step=117500, grad_norm=4.797100067138672, loss=1.0006670951843262 +I0831 18:21:20.526432 140462207792896 logging_writer.py:48] [117600] global_step=117600, grad_norm=5.013663291931152, loss=1.0083142518997192 +I0831 18:21:47.636953 140462291654400 logging_writer.py:48] [117700] global_step=117700, grad_norm=5.114383697509766, loss=1.0651369094848633 +I0831 18:22:14.756113 140462207792896 logging_writer.py:48] [117800] global_step=117800, grad_norm=5.075742244720459, loss=1.0506207942962646 +I0831 18:22:41.972548 140462291654400 logging_writer.py:48] [117900] global_step=117900, grad_norm=4.777499198913574, loss=0.9389281868934631 +I0831 18:23:09.121435 140462207792896 logging_writer.py:48] [118000] global_step=118000, grad_norm=5.148336887359619, loss=0.9814031720161438 +I0831 18:23:36.287013 140462291654400 logging_writer.py:48] [118100] global_step=118100, grad_norm=4.970338344573975, loss=1.0472642183303833 +I0831 18:24:03.486158 140462207792896 logging_writer.py:48] [118200] global_step=118200, grad_norm=5.108377456665039, loss=1.1723076105117798 +I0831 18:24:30.637889 140462291654400 logging_writer.py:48] [118300] global_step=118300, grad_norm=4.800948619842529, loss=0.9654536247253418 +I0831 18:24:57.764885 140462207792896 logging_writer.py:48] [118400] global_step=118400, grad_norm=5.000363826751709, loss=1.0014458894729614 +I0831 18:25:25.223495 140462291654400 logging_writer.py:48] [118500] global_step=118500, grad_norm=5.206971645355225, loss=1.0721126794815063 +I0831 18:25:52.362612 140462207792896 logging_writer.py:48] [118600] global_step=118600, grad_norm=4.5182576179504395, loss=0.8723965883255005 +I0831 18:26:19.502624 140462291654400 logging_writer.py:48] [118700] global_step=118700, grad_norm=4.7357330322265625, loss=0.9893472194671631 +I0831 18:26:46.733748 140462207792896 logging_writer.py:48] [118800] global_step=118800, grad_norm=5.046602725982666, loss=1.0019222497940063 +I0831 18:27:13.868820 140462291654400 logging_writer.py:48] [118900] global_step=118900, grad_norm=4.974422454833984, loss=0.9600452184677124 +I0831 18:27:41.013953 140462207792896 logging_writer.py:48] [119000] global_step=119000, grad_norm=5.0243706703186035, loss=1.0479531288146973 +I0831 18:28:08.213048 140462291654400 logging_writer.py:48] [119100] global_step=119100, grad_norm=4.823195934295654, loss=1.0052921772003174 +I0831 18:28:35.325400 140462207792896 logging_writer.py:48] [119200] global_step=119200, grad_norm=4.728658676147461, loss=0.9682378172874451 +I0831 18:29:02.468082 140462291654400 logging_writer.py:48] [119300] global_step=119300, grad_norm=5.077601432800293, loss=1.0624881982803345 +I0831 18:29:29.649986 140462207792896 logging_writer.py:48] [119400] global_step=119400, grad_norm=4.881599426269531, loss=1.0165334939956665 +I0831 18:29:57.028962 140462291654400 logging_writer.py:48] [119500] global_step=119500, grad_norm=5.049523830413818, loss=1.0844212770462036 +I0831 18:30:24.147668 140462207792896 logging_writer.py:48] [119600] global_step=119600, grad_norm=4.885377883911133, loss=0.9741007089614868 +I0831 18:30:51.325977 140462291654400 logging_writer.py:48] [119700] global_step=119700, grad_norm=4.723869800567627, loss=0.9625809192657471 +I0831 18:31:18.444834 140462207792896 logging_writer.py:48] [119800] global_step=119800, grad_norm=4.6406755447387695, loss=0.932768702507019 +I0831 18:31:45.549969 140462291654400 logging_writer.py:48] [119900] global_step=119900, grad_norm=4.854034900665283, loss=1.0233194828033447 +I0831 18:32:12.748756 140462207792896 logging_writer.py:48] [120000] global_step=120000, grad_norm=4.93853759765625, loss=1.0610573291778564 +I0831 18:32:39.874849 140462291654400 logging_writer.py:48] [120100] global_step=120100, grad_norm=4.448482513427734, loss=0.9229467511177063 +I0831 18:33:07.004144 140462207792896 logging_writer.py:48] [120200] global_step=120200, grad_norm=5.070688724517822, loss=1.0384318828582764 +I0831 18:33:34.229561 140462291654400 logging_writer.py:48] [120300] global_step=120300, grad_norm=4.839802265167236, loss=1.044087290763855 +I0831 18:34:01.379346 140462207792896 logging_writer.py:48] [120400] global_step=120400, grad_norm=5.240930080413818, loss=1.1075561046600342 +I0831 18:34:28.526514 140462291654400 logging_writer.py:48] [120500] global_step=120500, grad_norm=4.813675880432129, loss=1.0089457035064697 +I0831 18:34:55.948564 140462207792896 logging_writer.py:48] [120600] global_step=120600, grad_norm=4.805734634399414, loss=1.0223191976547241 +I0831 18:35:23.087695 140462291654400 logging_writer.py:48] [120700] global_step=120700, grad_norm=4.912923336029053, loss=0.9996052980422974 +I0831 18:35:50.244693 140462207792896 logging_writer.py:48] [120800] global_step=120800, grad_norm=4.797335624694824, loss=0.9476337432861328 +I0831 18:36:17.439042 140462291654400 logging_writer.py:48] [120900] global_step=120900, grad_norm=4.851179599761963, loss=0.959784984588623 +I0831 18:36:44.577270 140462207792896 logging_writer.py:48] [121000] global_step=121000, grad_norm=4.582058906555176, loss=0.9240178465843201 +I0831 18:37:11.723982 140462291654400 logging_writer.py:48] [121100] global_step=121100, grad_norm=4.771092891693115, loss=0.955331563949585 +I0831 18:37:38.952239 140462207792896 logging_writer.py:48] [121200] global_step=121200, grad_norm=4.553656101226807, loss=0.8787550926208496 +I0831 18:38:06.057974 140462291654400 logging_writer.py:48] [121300] global_step=121300, grad_norm=5.071834087371826, loss=1.0178806781768799 +I0831 18:38:33.219517 140462207792896 logging_writer.py:48] [121400] global_step=121400, grad_norm=4.5512847900390625, loss=0.9243761897087097 +I0831 18:39:00.422232 140462291654400 logging_writer.py:48] [121500] global_step=121500, grad_norm=4.6898651123046875, loss=0.9188967943191528 +I0831 18:39:27.811151 140462207792896 logging_writer.py:48] [121600] global_step=121600, grad_norm=5.013453483581543, loss=1.0483176708221436 +I0831 18:39:54.970608 140462291654400 logging_writer.py:48] [121700] global_step=121700, grad_norm=4.99608850479126, loss=0.9939717054367065 +I0831 18:40:22.202000 140462207792896 logging_writer.py:48] [121800] global_step=121800, grad_norm=4.7887749671936035, loss=0.9667272567749023 +I0831 18:40:49.329799 140462291654400 logging_writer.py:48] [121900] global_step=121900, grad_norm=5.182550430297852, loss=1.165998101234436 +I0831 18:41:16.444746 140462207792896 logging_writer.py:48] [122000] global_step=122000, grad_norm=4.992772102355957, loss=0.9730014204978943 +I0831 18:41:43.648191 140462291654400 logging_writer.py:48] [122100] global_step=122100, grad_norm=4.868677139282227, loss=0.9751788377761841 +I0831 18:42:10.799826 140462207792896 logging_writer.py:48] [122200] global_step=122200, grad_norm=4.906267166137695, loss=1.018440842628479 +I0831 18:42:37.948528 140462291654400 logging_writer.py:48] [122300] global_step=122300, grad_norm=4.791965961456299, loss=1.0067616701126099 +I0831 18:43:05.143853 140462207792896 logging_writer.py:48] [122400] global_step=122400, grad_norm=4.8009233474731445, loss=1.0279536247253418 +I0831 18:43:32.272264 140462291654400 logging_writer.py:48] [122500] global_step=122500, grad_norm=4.578644275665283, loss=0.9688810706138611 +I0831 18:43:59.413134 140462207792896 logging_writer.py:48] [122600] global_step=122600, grad_norm=4.831005096435547, loss=0.9854665994644165 +I0831 18:44:26.822414 140462291654400 logging_writer.py:48] [122700] global_step=122700, grad_norm=5.316127777099609, loss=1.036868929862976 +I0831 18:44:54.011192 140462207792896 logging_writer.py:48] [122800] global_step=122800, grad_norm=4.913704872131348, loss=1.0041658878326416 +I0831 18:45:21.137297 140462291654400 logging_writer.py:48] [122900] global_step=122900, grad_norm=4.830339431762695, loss=1.0465507507324219 +I0831 18:45:48.350311 140462207792896 logging_writer.py:48] [123000] global_step=123000, grad_norm=4.998183250427246, loss=1.0761563777923584 +I0831 18:46:15.504659 140462291654400 logging_writer.py:48] [123100] global_step=123100, grad_norm=4.575785160064697, loss=0.9318074584007263 +I0831 18:46:42.641149 140462207792896 logging_writer.py:48] [123200] global_step=123200, grad_norm=4.792347431182861, loss=1.003838062286377 +I0831 18:47:09.854095 140462291654400 logging_writer.py:48] [123300] global_step=123300, grad_norm=5.0051493644714355, loss=1.0165058374404907 +I0831 18:47:36.994924 140462207792896 logging_writer.py:48] [123400] global_step=123400, grad_norm=5.071453094482422, loss=1.0671257972717285 +I0831 18:48:04.129349 140462291654400 logging_writer.py:48] [123500] global_step=123500, grad_norm=4.932889461517334, loss=1.0281145572662354 +I0831 18:48:31.330734 140462207792896 logging_writer.py:48] [123600] global_step=123600, grad_norm=5.0653204917907715, loss=0.9843783378601074 +I0831 18:48:58.686822 140462291654400 logging_writer.py:48] [123700] global_step=123700, grad_norm=4.647601127624512, loss=1.0230827331542969 +I0831 18:49:25.860027 140462207792896 logging_writer.py:48] [123800] global_step=123800, grad_norm=4.587740898132324, loss=0.8665546774864197 +I0831 18:49:53.053067 140462291654400 logging_writer.py:48] [123900] global_step=123900, grad_norm=4.729403495788574, loss=0.9145935773849487 +I0831 18:50:20.195735 140462207792896 logging_writer.py:48] [124000] global_step=124000, grad_norm=5.089610576629639, loss=1.121161699295044 +I0831 18:50:39.596826 140659750036672 spec.py:333] Evaluating on the training split. +I0831 18:50:45.806874 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 18:50:55.226302 140659750036672 spec.py:363] Evaluating on the test split. +I0831 18:50:56.119573 140659750036672 submission_runner.py:516] Time since start: 34377.30s, Step: 124073, {'train/accuracy': Array(0.9353077, dtype=float32), 'train/loss': Array(0.22499812, dtype=float32), 'validation/accuracy': Array(0.75262, dtype=float32), 'validation/loss': Array(1.0646015, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6204, dtype=float32), 'test/loss': Array(1.841354, dtype=float32), 'test/num_examples': 10000, 'score': 33984.88090109825, 'total_duration': 34377.29505729675, 'accumulated_submission_time': 33984.88090109825, 'accumulated_eval_time': 389.97516417503357, 'accumulated_logging_time': 1.4011025428771973} +I0831 18:50:56.200026 140462291654400 logging_writer.py:48] [124073] accumulated_eval_time=389.975, accumulated_logging_time=1.4011, accumulated_submission_time=33984.9, global_step=124073, preemption_count=0, score=33984.9, test/accuracy=0.6204000115394592, test/loss=1.8413540124893188, test/num_examples=10000, total_duration=34377.3, train/accuracy=0.9353076815605164, train/loss=0.2249981164932251, validation/accuracy=0.7526199817657471, validation/loss=1.0646015405654907, validation/num_examples=50000 +I0831 18:51:03.911171 140462207792896 logging_writer.py:48] [124100] global_step=124100, grad_norm=4.904119968414307, loss=1.0710865259170532 +I0831 18:51:31.084819 140462291654400 logging_writer.py:48] [124200] global_step=124200, grad_norm=5.013574123382568, loss=1.0300567150115967 +I0831 18:51:58.241670 140462207792896 logging_writer.py:48] [124300] global_step=124300, grad_norm=5.103557109832764, loss=1.0337011814117432 +I0831 18:52:25.397814 140462291654400 logging_writer.py:48] [124400] global_step=124400, grad_norm=4.72796106338501, loss=0.9825449585914612 +I0831 18:52:52.622317 140462207792896 logging_writer.py:48] [124500] global_step=124500, grad_norm=4.842215538024902, loss=1.0115368366241455 +I0831 18:53:19.761790 140462291654400 logging_writer.py:48] [124600] global_step=124600, grad_norm=4.6935272216796875, loss=0.9507580399513245 +I0831 18:53:46.885830 140462207792896 logging_writer.py:48] [124700] global_step=124700, grad_norm=5.043552875518799, loss=1.044607400894165 +I0831 18:54:14.347078 140462291654400 logging_writer.py:48] [124800] global_step=124800, grad_norm=4.601502418518066, loss=0.9795895218849182 +I0831 18:54:41.531608 140462207792896 logging_writer.py:48] [124900] global_step=124900, grad_norm=5.110959053039551, loss=1.0955064296722412 +I0831 18:55:08.667826 140462291654400 logging_writer.py:48] [125000] global_step=125000, grad_norm=4.5604023933410645, loss=0.8917099237442017 +I0831 18:55:35.851816 140462207792896 logging_writer.py:48] [125100] global_step=125100, grad_norm=5.1112260818481445, loss=1.052515983581543 +I0831 18:56:03.015535 140462291654400 logging_writer.py:48] [125200] global_step=125200, grad_norm=4.754487991333008, loss=0.988122284412384 +I0831 18:56:30.162299 140462207792896 logging_writer.py:48] [125300] global_step=125300, grad_norm=5.047730922698975, loss=1.006106972694397 +I0831 18:56:57.347147 140462291654400 logging_writer.py:48] [125400] global_step=125400, grad_norm=4.849312782287598, loss=0.9701749682426453 +I0831 18:57:24.465829 140462207792896 logging_writer.py:48] [125500] global_step=125500, grad_norm=4.91546106338501, loss=0.9795525074005127 +I0831 18:57:51.629070 140462291654400 logging_writer.py:48] [125600] global_step=125600, grad_norm=4.918391704559326, loss=0.9865906238555908 +I0831 18:58:18.822149 140462207792896 logging_writer.py:48] [125700] global_step=125700, grad_norm=4.707231521606445, loss=0.9020818471908569 +I0831 18:58:45.939081 140462291654400 logging_writer.py:48] [125800] global_step=125800, grad_norm=4.577415466308594, loss=0.919074535369873 +I0831 18:59:13.314783 140462207792896 logging_writer.py:48] [125900] global_step=125900, grad_norm=4.921111583709717, loss=0.9801174402236938 +I0831 18:59:40.534367 140462291654400 logging_writer.py:48] [126000] global_step=126000, grad_norm=4.858335018157959, loss=0.9574331045150757 +I0831 19:00:07.656168 140462207792896 logging_writer.py:48] [126100] global_step=126100, grad_norm=4.9719953536987305, loss=1.0022038221359253 +I0831 19:00:34.794416 140462291654400 logging_writer.py:48] [126200] global_step=126200, grad_norm=4.838680267333984, loss=0.9495887756347656 +I0831 19:01:02.003033 140462207792896 logging_writer.py:48] [126300] global_step=126300, grad_norm=4.937734603881836, loss=1.0443426370620728 +I0831 19:01:29.130965 140462291654400 logging_writer.py:48] [126400] global_step=126400, grad_norm=4.927275657653809, loss=1.0065414905548096 +I0831 19:01:56.307336 140462207792896 logging_writer.py:48] [126500] global_step=126500, grad_norm=5.108245372772217, loss=1.0423998832702637 +I0831 19:02:23.495838 140462291654400 logging_writer.py:48] [126600] global_step=126600, grad_norm=5.1056060791015625, loss=1.0658725500106812 +I0831 19:02:50.646084 140462207792896 logging_writer.py:48] [126700] global_step=126700, grad_norm=5.036767482757568, loss=1.0083175897598267 +I0831 19:03:17.836750 140462291654400 logging_writer.py:48] [126800] global_step=126800, grad_norm=4.4392781257629395, loss=0.8948415517807007 +I0831 19:03:45.257914 140462207792896 logging_writer.py:48] [126900] global_step=126900, grad_norm=4.8815460205078125, loss=0.9459068179130554 +I0831 19:04:12.376321 140462291654400 logging_writer.py:48] [127000] global_step=127000, grad_norm=5.155577659606934, loss=1.0249789953231812 +I0831 19:04:39.546157 140462207792896 logging_writer.py:48] [127100] global_step=127100, grad_norm=4.884428977966309, loss=1.0642575025558472 +I0831 19:05:06.758373 140462291654400 logging_writer.py:48] [127200] global_step=127200, grad_norm=4.806020736694336, loss=1.0288459062576294 +I0831 19:05:33.901215 140462207792896 logging_writer.py:48] [127300] global_step=127300, grad_norm=4.872886657714844, loss=1.0134962797164917 +I0831 19:06:01.052794 140462291654400 logging_writer.py:48] [127400] global_step=127400, grad_norm=5.027369499206543, loss=1.0318619012832642 +I0831 19:06:28.265936 140462207792896 logging_writer.py:48] [127500] global_step=127500, grad_norm=4.590577125549316, loss=0.9203292727470398 +I0831 19:06:55.421501 140462291654400 logging_writer.py:48] [127600] global_step=127600, grad_norm=5.278386116027832, loss=1.0394244194030762 +I0831 19:07:22.559993 140462207792896 logging_writer.py:48] [127700] global_step=127700, grad_norm=5.157871723175049, loss=1.076653242111206 +I0831 19:07:49.743436 140462291654400 logging_writer.py:48] [127800] global_step=127800, grad_norm=4.806427478790283, loss=1.0104268789291382 +I0831 19:08:16.890180 140462207792896 logging_writer.py:48] [127900] global_step=127900, grad_norm=4.669752597808838, loss=0.9074390530586243 +I0831 19:08:44.259920 140462291654400 logging_writer.py:48] [128000] global_step=128000, grad_norm=4.722379207611084, loss=0.9568477869033813 +I0831 19:09:11.454045 140462207792896 logging_writer.py:48] [128100] global_step=128100, grad_norm=5.040907382965088, loss=1.0083115100860596 +I0831 19:09:38.593007 140462291654400 logging_writer.py:48] [128200] global_step=128200, grad_norm=4.9026336669921875, loss=0.9492679238319397 +I0831 19:10:05.731705 140462207792896 logging_writer.py:48] [128300] global_step=128300, grad_norm=4.72209358215332, loss=0.9167138338088989 +I0831 19:10:32.957316 140462291654400 logging_writer.py:48] [128400] global_step=128400, grad_norm=4.942287921905518, loss=0.9436628818511963 +I0831 19:11:00.108513 140462207792896 logging_writer.py:48] [128500] global_step=128500, grad_norm=4.870534420013428, loss=0.9054669141769409 +I0831 19:11:27.234295 140462291654400 logging_writer.py:48] [128600] global_step=128600, grad_norm=5.337985515594482, loss=1.059817910194397 +I0831 19:11:54.443041 140462207792896 logging_writer.py:48] [128700] global_step=128700, grad_norm=4.894642353057861, loss=1.009624719619751 +I0831 19:12:21.580345 140462291654400 logging_writer.py:48] [128800] global_step=128800, grad_norm=5.057863235473633, loss=1.0513979196548462 +I0831 19:12:48.735001 140462207792896 logging_writer.py:48] [128900] global_step=128900, grad_norm=5.086897850036621, loss=1.036156177520752 +I0831 19:13:15.936834 140462291654400 logging_writer.py:48] [129000] global_step=129000, grad_norm=4.848765850067139, loss=0.9480619430541992 +I0831 19:13:43.278722 140462207792896 logging_writer.py:48] [129100] global_step=129100, grad_norm=5.121601581573486, loss=1.0108648538589478 +I0831 19:14:10.418360 140462291654400 logging_writer.py:48] [129200] global_step=129200, grad_norm=4.956387519836426, loss=0.980854332447052 +I0831 19:14:37.642588 140462207792896 logging_writer.py:48] [129300] global_step=129300, grad_norm=4.840982913970947, loss=0.9268102645874023 +I0831 19:15:04.782050 140462291654400 logging_writer.py:48] [129400] global_step=129400, grad_norm=4.6216816902160645, loss=0.9635359048843384 +I0831 19:15:31.918240 140462207792896 logging_writer.py:48] [129500] global_step=129500, grad_norm=4.726377964019775, loss=0.9992663264274597 +I0831 19:15:59.127114 140462291654400 logging_writer.py:48] [129600] global_step=129600, grad_norm=4.918759822845459, loss=1.0646185874938965 +I0831 19:16:26.265179 140462207792896 logging_writer.py:48] [129700] global_step=129700, grad_norm=5.06077766418457, loss=1.0052297115325928 +I0831 19:16:53.385207 140462291654400 logging_writer.py:48] [129800] global_step=129800, grad_norm=4.991241931915283, loss=1.0191779136657715 +I0831 19:17:20.580752 140462207792896 logging_writer.py:48] [129900] global_step=129900, grad_norm=4.792139530181885, loss=0.9640012383460999 +I0831 19:17:47.722842 140462291654400 logging_writer.py:48] [130000] global_step=130000, grad_norm=5.175104141235352, loss=1.0370745658874512 +I0831 19:18:15.069711 140462207792896 logging_writer.py:48] [130100] global_step=130100, grad_norm=4.886310577392578, loss=1.015156865119934 +I0831 19:18:42.267400 140462291654400 logging_writer.py:48] [130200] global_step=130200, grad_norm=4.42291259765625, loss=0.8750790357589722 +I0831 19:19:09.388000 140462207792896 logging_writer.py:48] [130300] global_step=130300, grad_norm=4.977358341217041, loss=0.9982320666313171 +I0831 19:19:36.517823 140462291654400 logging_writer.py:48] [130400] global_step=130400, grad_norm=5.193551540374756, loss=0.9897265434265137 +I0831 19:20:03.733483 140462207792896 logging_writer.py:48] [130500] global_step=130500, grad_norm=4.715762615203857, loss=0.9419952630996704 +I0831 19:20:30.861227 140462291654400 logging_writer.py:48] [130600] global_step=130600, grad_norm=4.887787818908691, loss=0.949242115020752 +I0831 19:20:57.999439 140462207792896 logging_writer.py:48] [130700] global_step=130700, grad_norm=4.794997692108154, loss=0.9541128873825073 +I0831 19:21:25.217416 140462291654400 logging_writer.py:48] [130800] global_step=130800, grad_norm=4.9515252113342285, loss=0.890883207321167 +I0831 19:21:52.330173 140462207792896 logging_writer.py:48] [130900] global_step=130900, grad_norm=4.670994758605957, loss=0.9170784950256348 +I0831 19:22:19.445137 140462291654400 logging_writer.py:48] [131000] global_step=131000, grad_norm=5.037604331970215, loss=0.9877687096595764 +I0831 19:22:46.643815 140462207792896 logging_writer.py:48] [131100] global_step=131100, grad_norm=5.1962103843688965, loss=0.9980480670928955 +I0831 19:23:14.034998 140462291654400 logging_writer.py:48] [131200] global_step=131200, grad_norm=5.01896858215332, loss=1.0698829889297485 +I0831 19:23:41.145080 140462207792896 logging_writer.py:48] [131300] global_step=131300, grad_norm=4.878316879272461, loss=1.0382630825042725 +I0831 19:24:08.319670 140462291654400 logging_writer.py:48] [131400] global_step=131400, grad_norm=4.544620037078857, loss=0.9346424341201782 +I0831 19:24:12.284154 140659750036672 spec.py:333] Evaluating on the training split. +I0831 19:24:18.596096 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 19:24:28.395749 140659750036672 spec.py:363] Evaluating on the test split. +I0831 19:24:29.286674 140659750036672 submission_runner.py:516] Time since start: 36390.46s, Step: 131416, {'train/accuracy': Array(0.93769926, dtype=float32), 'train/loss': Array(0.21849291, dtype=float32), 'validation/accuracy': Array(0.75236, dtype=float32), 'validation/loss': Array(1.0655642, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6225, dtype=float32), 'test/loss': Array(1.8407513, dtype=float32), 'test/num_examples': 10000, 'score': 35980.86967778206, 'total_duration': 36390.460955142975, 'accumulated_submission_time': 35980.86967778206, 'accumulated_eval_time': 406.9744417667389, 'accumulated_logging_time': 1.5179433822631836} +I0831 19:24:29.343795 140462207792896 logging_writer.py:48] [131416] accumulated_eval_time=406.974, accumulated_logging_time=1.51794, accumulated_submission_time=35980.9, global_step=131416, preemption_count=0, score=35980.9, test/accuracy=0.6225000023841858, test/loss=1.84075129032135, test/num_examples=10000, total_duration=36390.5, train/accuracy=0.9376992583274841, train/loss=0.21849291026592255, validation/accuracy=0.7523599863052368, validation/loss=1.0655641555786133, validation/num_examples=50000 +I0831 19:24:52.464426 140462291654400 logging_writer.py:48] [131500] global_step=131500, grad_norm=5.088639736175537, loss=1.0706193447113037 +I0831 19:25:19.578706 140462207792896 logging_writer.py:48] [131600] global_step=131600, grad_norm=5.138761520385742, loss=0.9675340056419373 +I0831 19:25:46.800252 140462291654400 logging_writer.py:48] [131700] global_step=131700, grad_norm=4.586668968200684, loss=0.9584568738937378 +I0831 19:26:13.943938 140462207792896 logging_writer.py:48] [131800] global_step=131800, grad_norm=4.878078460693359, loss=0.9067360162734985 +I0831 19:26:41.081779 140462291654400 logging_writer.py:48] [131900] global_step=131900, grad_norm=4.727520942687988, loss=0.9361591339111328 +I0831 19:27:08.274572 140462207792896 logging_writer.py:48] [132000] global_step=132000, grad_norm=5.078022003173828, loss=1.0110633373260498 +I0831 19:27:35.402755 140462291654400 logging_writer.py:48] [132100] global_step=132100, grad_norm=4.602766990661621, loss=0.9133206009864807 +I0831 19:28:02.776562 140462207792896 logging_writer.py:48] [132200] global_step=132200, grad_norm=5.087735652923584, loss=1.0595200061798096 +I0831 19:28:29.969807 140462291654400 logging_writer.py:48] [132300] global_step=132300, grad_norm=5.015707969665527, loss=1.062422752380371 +I0831 19:28:57.073860 140462207792896 logging_writer.py:48] [132400] global_step=132400, grad_norm=5.2038068771362305, loss=1.1299234628677368 +I0831 19:29:24.197413 140462291654400 logging_writer.py:48] [132500] global_step=132500, grad_norm=4.7672505378723145, loss=0.9881995916366577 +I0831 19:29:51.392121 140462207792896 logging_writer.py:48] [132600] global_step=132600, grad_norm=4.946400165557861, loss=0.9914063215255737 +I0831 19:30:18.521338 140462291654400 logging_writer.py:48] [132700] global_step=132700, grad_norm=4.880516529083252, loss=0.9837242364883423 +I0831 19:30:45.632685 140462207792896 logging_writer.py:48] [132800] global_step=132800, grad_norm=5.056700706481934, loss=1.0083162784576416 +I0831 19:31:12.821568 140462291654400 logging_writer.py:48] [132900] global_step=132900, grad_norm=4.950362205505371, loss=0.9591640830039978 +I0831 19:31:39.977714 140462207792896 logging_writer.py:48] [133000] global_step=133000, grad_norm=4.748385429382324, loss=0.9933662414550781 +I0831 19:32:07.105887 140462291654400 logging_writer.py:48] [133100] global_step=133100, grad_norm=4.91379976272583, loss=0.9482781887054443 +I0831 19:32:34.286937 140462207792896 logging_writer.py:48] [133200] global_step=133200, grad_norm=5.239889621734619, loss=1.067208170890808 +I0831 19:33:01.626725 140462291654400 logging_writer.py:48] [133300] global_step=133300, grad_norm=4.552188873291016, loss=0.8705334067344666 +I0831 19:33:28.767127 140462207792896 logging_writer.py:48] [133400] global_step=133400, grad_norm=5.007979393005371, loss=0.9735894203186035 +I0831 19:33:55.954976 140462291654400 logging_writer.py:48] [133500] global_step=133500, grad_norm=4.846745014190674, loss=1.0085041522979736 +I0831 19:34:23.076879 140462207792896 logging_writer.py:48] [133600] global_step=133600, grad_norm=5.107703685760498, loss=1.0288716554641724 +I0831 19:34:50.190138 140462291654400 logging_writer.py:48] [133700] global_step=133700, grad_norm=4.77857780456543, loss=0.901441752910614 +I0831 19:35:17.378520 140462207792896 logging_writer.py:48] [133800] global_step=133800, grad_norm=5.0426740646362305, loss=0.991850733757019 +I0831 19:35:44.511348 140462291654400 logging_writer.py:48] [133900] global_step=133900, grad_norm=4.563718795776367, loss=0.9624825119972229 +I0831 19:36:11.646410 140462207792896 logging_writer.py:48] [134000] global_step=134000, grad_norm=5.289568901062012, loss=1.0520305633544922 +I0831 19:36:38.828994 140462291654400 logging_writer.py:48] [134100] global_step=134100, grad_norm=5.263108730316162, loss=1.0118894577026367 +I0831 19:37:05.976790 140462207792896 logging_writer.py:48] [134200] global_step=134200, grad_norm=5.188027858734131, loss=0.9524095058441162 +I0831 19:37:33.102838 140462291654400 logging_writer.py:48] [134300] global_step=134300, grad_norm=4.824306964874268, loss=0.9094181656837463 +I0831 19:38:00.554319 140462207792896 logging_writer.py:48] [134400] global_step=134400, grad_norm=4.741953372955322, loss=0.9899893403053284 +I0831 19:38:27.681951 140462291654400 logging_writer.py:48] [134500] global_step=134500, grad_norm=4.702227592468262, loss=0.9641828536987305 +I0831 19:38:54.829272 140462207792896 logging_writer.py:48] [134600] global_step=134600, grad_norm=4.959468841552734, loss=1.0700668096542358 +I0831 19:39:22.029093 140462291654400 logging_writer.py:48] [134700] global_step=134700, grad_norm=4.6775288581848145, loss=0.9067063927650452 +I0831 19:39:49.193231 140462207792896 logging_writer.py:48] [134800] global_step=134800, grad_norm=4.8790411949157715, loss=0.9351679086685181 +I0831 19:40:16.343873 140462291654400 logging_writer.py:48] [134900] global_step=134900, grad_norm=4.9359354972839355, loss=0.9611467123031616 +I0831 19:40:43.536087 140462207792896 logging_writer.py:48] [135000] global_step=135000, grad_norm=4.77863073348999, loss=1.0016525983810425 +I0831 19:41:10.642688 140462291654400 logging_writer.py:48] [135100] global_step=135100, grad_norm=4.990229606628418, loss=0.9433571100234985 +I0831 19:41:37.741838 140462207792896 logging_writer.py:48] [135200] global_step=135200, grad_norm=4.841803550720215, loss=0.9523303508758545 +I0831 19:42:04.952910 140462291654400 logging_writer.py:48] [135300] global_step=135300, grad_norm=4.918972492218018, loss=0.9787647128105164 +I0831 19:42:32.122677 140462207792896 logging_writer.py:48] [135400] global_step=135400, grad_norm=4.937927722930908, loss=0.9982444643974304 +I0831 19:42:59.492346 140462291654400 logging_writer.py:48] [135500] global_step=135500, grad_norm=4.795772552490234, loss=0.960688054561615 +I0831 19:43:26.720024 140462207792896 logging_writer.py:48] [135600] global_step=135600, grad_norm=5.184144496917725, loss=1.0150196552276611 +I0831 19:43:53.843949 140462291654400 logging_writer.py:48] [135700] global_step=135700, grad_norm=4.941924571990967, loss=1.0416250228881836 +I0831 19:44:20.956830 140462207792896 logging_writer.py:48] [135800] global_step=135800, grad_norm=4.8542866706848145, loss=0.9621334075927734 +I0831 19:44:48.126482 140462291654400 logging_writer.py:48] [135900] global_step=135900, grad_norm=5.036419868469238, loss=0.9928356409072876 +I0831 19:45:15.231724 140462207792896 logging_writer.py:48] [136000] global_step=136000, grad_norm=5.054364204406738, loss=0.9925578832626343 +I0831 19:45:42.365578 140462291654400 logging_writer.py:48] [136100] global_step=136100, grad_norm=4.816431522369385, loss=0.9427075386047363 +I0831 19:46:09.542826 140462207792896 logging_writer.py:48] [136200] global_step=136200, grad_norm=5.158205032348633, loss=1.0210700035095215 +I0831 19:46:36.663465 140462291654400 logging_writer.py:48] [136300] global_step=136300, grad_norm=4.863442420959473, loss=0.9541996121406555 +I0831 19:47:03.784559 140462207792896 logging_writer.py:48] [136400] global_step=136400, grad_norm=4.868241310119629, loss=1.001609206199646 +I0831 19:47:31.243554 140462291654400 logging_writer.py:48] [136500] global_step=136500, grad_norm=4.835646152496338, loss=0.9345439672470093 +I0831 19:47:58.387175 140462207792896 logging_writer.py:48] [136600] global_step=136600, grad_norm=4.943624496459961, loss=1.0194058418273926 +I0831 19:48:25.522573 140462291654400 logging_writer.py:48] [136700] global_step=136700, grad_norm=4.8345441818237305, loss=0.9494859576225281 +I0831 19:48:52.708358 140462207792896 logging_writer.py:48] [136800] global_step=136800, grad_norm=4.58807373046875, loss=0.9247536659240723 +I0831 19:49:19.860191 140462291654400 logging_writer.py:48] [136900] global_step=136900, grad_norm=4.9317946434021, loss=0.9907269477844238 +I0831 19:49:46.964816 140462207792896 logging_writer.py:48] [137000] global_step=137000, grad_norm=4.786767482757568, loss=0.9396387338638306 +I0831 19:50:14.169023 140462291654400 logging_writer.py:48] [137100] global_step=137100, grad_norm=5.002502918243408, loss=0.9896750450134277 +I0831 19:50:41.296794 140462207792896 logging_writer.py:48] [137200] global_step=137200, grad_norm=4.862489223480225, loss=0.9988329410552979 +I0831 19:51:08.433206 140462291654400 logging_writer.py:48] [137300] global_step=137300, grad_norm=4.733452320098877, loss=0.9492261409759521 +I0831 19:51:35.628247 140462207792896 logging_writer.py:48] [137400] global_step=137400, grad_norm=4.9575066566467285, loss=1.03700590133667 +I0831 19:52:02.757204 140462291654400 logging_writer.py:48] [137500] global_step=137500, grad_norm=5.034666538238525, loss=0.9804816246032715 +I0831 19:52:30.120920 140462207792896 logging_writer.py:48] [137600] global_step=137600, grad_norm=4.818413257598877, loss=0.9408241510391235 +I0831 19:52:57.325653 140462291654400 logging_writer.py:48] [137700] global_step=137700, grad_norm=4.968624114990234, loss=1.0151206254959106 +I0831 19:53:24.451770 140462207792896 logging_writer.py:48] [137800] global_step=137800, grad_norm=5.628589153289795, loss=1.0814909934997559 +I0831 19:53:51.572188 140462291654400 logging_writer.py:48] [137900] global_step=137900, grad_norm=4.9721999168396, loss=1.0177054405212402 +I0831 19:54:18.756858 140462207792896 logging_writer.py:48] [138000] global_step=138000, grad_norm=4.969879627227783, loss=0.9964390993118286 +I0831 19:54:45.876119 140462291654400 logging_writer.py:48] [138100] global_step=138100, grad_norm=5.289172172546387, loss=1.025873064994812 +I0831 19:55:13.016450 140462207792896 logging_writer.py:48] [138200] global_step=138200, grad_norm=4.578670978546143, loss=0.9587172865867615 +I0831 19:55:40.206748 140462291654400 logging_writer.py:48] [138300] global_step=138300, grad_norm=4.618259906768799, loss=0.9105334877967834 +I0831 19:56:07.315269 140462207792896 logging_writer.py:48] [138400] global_step=138400, grad_norm=5.059256076812744, loss=1.010119915008545 +I0831 19:56:34.443770 140462291654400 logging_writer.py:48] [138500] global_step=138500, grad_norm=4.6692633628845215, loss=0.9378334879875183 +I0831 19:57:01.642219 140462207792896 logging_writer.py:48] [138600] global_step=138600, grad_norm=5.09670877456665, loss=1.0881261825561523 +I0831 19:57:28.986926 140462291654400 logging_writer.py:48] [138700] global_step=138700, grad_norm=4.74769401550293, loss=1.0420399904251099 +I0831 19:57:45.375704 140659750036672 spec.py:333] Evaluating on the training split. +I0831 19:57:51.542303 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 19:58:02.482608 140659750036672 spec.py:363] Evaluating on the test split. +I0831 19:58:03.369391 140659750036672 submission_runner.py:516] Time since start: 38404.54s, Step: 138762, {'train/accuracy': Array(0.938636, dtype=float32), 'train/loss': Array(0.20914964, dtype=float32), 'validation/accuracy': Array(0.75232, dtype=float32), 'validation/loss': Array(1.0671511, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6244, dtype=float32), 'test/loss': Array(1.843589, dtype=float32), 'test/num_examples': 10000, 'score': 37976.824387311935, 'total_duration': 38404.54493522644, 'accumulated_submission_time': 37976.824387311935, 'accumulated_eval_time': 424.9661509990692, 'accumulated_logging_time': 1.5933632850646973} +I0831 19:58:03.424714 140462207792896 logging_writer.py:48] [138762] accumulated_eval_time=424.966, accumulated_logging_time=1.59336, accumulated_submission_time=37976.8, global_step=138762, preemption_count=0, score=37976.8, test/accuracy=0.6244000196456909, test/loss=1.843588948249817, test/num_examples=10000, total_duration=38404.5, train/accuracy=0.9386360049247742, train/loss=0.20914964377880096, validation/accuracy=0.7523199915885925, validation/loss=1.0671510696411133, validation/num_examples=50000 +I0831 19:58:14.096484 140462291654400 logging_writer.py:48] [138800] global_step=138800, grad_norm=4.830262660980225, loss=0.9987660646438599 +I0831 19:58:41.290149 140462207792896 logging_writer.py:48] [138900] global_step=138900, grad_norm=5.062754154205322, loss=1.0001490116119385 +I0831 19:59:08.442151 140462291654400 logging_writer.py:48] [139000] global_step=139000, grad_norm=5.082733154296875, loss=1.0084928274154663 +I0831 19:59:35.559297 140462207792896 logging_writer.py:48] [139100] global_step=139100, grad_norm=4.917656421661377, loss=1.022098422050476 +I0831 20:00:02.740309 140462291654400 logging_writer.py:48] [139200] global_step=139200, grad_norm=4.917812347412109, loss=0.9034224152565002 +I0831 20:00:29.853378 140462207792896 logging_writer.py:48] [139300] global_step=139300, grad_norm=5.0727458000183105, loss=0.9797490239143372 +I0831 20:00:57.002269 140462291654400 logging_writer.py:48] [139400] global_step=139400, grad_norm=4.932916641235352, loss=0.985655665397644 +I0831 20:01:24.205607 140462207792896 logging_writer.py:48] [139500] global_step=139500, grad_norm=4.877617835998535, loss=0.8744930624961853 +I0831 20:01:51.325957 140462291654400 logging_writer.py:48] [139600] global_step=139600, grad_norm=4.890471935272217, loss=0.9570766091346741 +I0831 20:02:18.677765 140462207792896 logging_writer.py:48] [139700] global_step=139700, grad_norm=5.033680438995361, loss=1.0259199142456055 +I0831 20:02:45.887956 140462291654400 logging_writer.py:48] [139800] global_step=139800, grad_norm=4.885627269744873, loss=1.0002994537353516 +I0831 20:03:13.023017 140462207792896 logging_writer.py:48] [139900] global_step=139900, grad_norm=4.852828502655029, loss=0.9855765700340271 +I0831 20:03:40.146353 140462291654400 logging_writer.py:48] [140000] global_step=140000, grad_norm=4.849849224090576, loss=0.9389604330062866 +I0831 20:04:07.326362 140462207792896 logging_writer.py:48] [140100] global_step=140100, grad_norm=4.544692516326904, loss=0.836556613445282 +I0831 20:04:34.447370 140462291654400 logging_writer.py:48] [140200] global_step=140200, grad_norm=5.138384819030762, loss=0.9527119398117065 +I0831 20:05:01.557655 140462207792896 logging_writer.py:48] [140300] global_step=140300, grad_norm=4.93916654586792, loss=0.960952877998352 +I0831 20:05:28.736950 140462291654400 logging_writer.py:48] [140400] global_step=140400, grad_norm=4.5759429931640625, loss=0.9180654883384705 +I0831 20:05:55.870678 140462207792896 logging_writer.py:48] [140500] global_step=140500, grad_norm=5.137375831604004, loss=1.0176935195922852 +I0831 20:06:23.012442 140462291654400 logging_writer.py:48] [140600] global_step=140600, grad_norm=4.71948766708374, loss=0.9832110404968262 +I0831 20:06:50.229496 140462207792896 logging_writer.py:48] [140700] global_step=140700, grad_norm=4.964005947113037, loss=0.8979731202125549 +I0831 20:07:17.599871 140462291654400 logging_writer.py:48] [140800] global_step=140800, grad_norm=4.699598789215088, loss=0.9553326368331909 +I0831 20:07:44.718358 140462207792896 logging_writer.py:48] [140900] global_step=140900, grad_norm=4.970709800720215, loss=0.9851078391075134 +I0831 20:08:11.918054 140462291654400 logging_writer.py:48] [141000] global_step=141000, grad_norm=4.874389171600342, loss=1.0389052629470825 +I0831 20:08:39.035610 140462207792896 logging_writer.py:48] [141100] global_step=141100, grad_norm=5.419908046722412, loss=1.07694411277771 +I0831 20:09:06.159539 140462291654400 logging_writer.py:48] [141200] global_step=141200, grad_norm=4.892148971557617, loss=0.9373798966407776 +I0831 20:09:33.366446 140462207792896 logging_writer.py:48] [141300] global_step=141300, grad_norm=4.701444625854492, loss=0.9141632318496704 +I0831 20:10:00.515591 140462291654400 logging_writer.py:48] [141400] global_step=141400, grad_norm=4.816829681396484, loss=1.0342016220092773 +I0831 20:10:27.668982 140462207792896 logging_writer.py:48] [141500] global_step=141500, grad_norm=4.506455898284912, loss=0.8862476348876953 +I0831 20:10:54.854794 140462291654400 logging_writer.py:48] [141600] global_step=141600, grad_norm=4.774014949798584, loss=0.9959375858306885 +I0831 20:11:22.006600 140462207792896 logging_writer.py:48] [141700] global_step=141700, grad_norm=4.990034580230713, loss=1.0020654201507568 +I0831 20:11:49.149016 140462291654400 logging_writer.py:48] [141800] global_step=141800, grad_norm=4.684852600097656, loss=0.8957135677337646 +I0831 20:12:16.616536 140462207792896 logging_writer.py:48] [141900] global_step=141900, grad_norm=5.060370922088623, loss=0.9861916303634644 +I0831 20:12:43.765913 140462291654400 logging_writer.py:48] [142000] global_step=142000, grad_norm=4.949754238128662, loss=0.988481879234314 +I0831 20:13:10.909900 140462207792896 logging_writer.py:48] [142100] global_step=142100, grad_norm=5.142743110656738, loss=1.0455578565597534 +I0831 20:13:38.108532 140462291654400 logging_writer.py:48] [142200] global_step=142200, grad_norm=4.890366554260254, loss=0.9067081212997437 +I0831 20:14:05.239831 140462207792896 logging_writer.py:48] [142300] global_step=142300, grad_norm=4.821104526519775, loss=0.9569635391235352 +I0831 20:14:32.374801 140462291654400 logging_writer.py:48] [142400] global_step=142400, grad_norm=4.802005290985107, loss=0.955528974533081 +I0831 20:14:59.555113 140462207792896 logging_writer.py:48] [142500] global_step=142500, grad_norm=4.816959857940674, loss=0.9861581325531006 +I0831 20:15:26.661381 140462291654400 logging_writer.py:48] [142600] global_step=142600, grad_norm=4.727746963500977, loss=0.9175833463668823 +I0831 20:15:53.764877 140462207792896 logging_writer.py:48] [142700] global_step=142700, grad_norm=4.7272443771362305, loss=0.9713278412818909 +I0831 20:16:20.977744 140462291654400 logging_writer.py:48] [142800] global_step=142800, grad_norm=4.8808794021606445, loss=0.9741232395172119 +I0831 20:16:48.100403 140462207792896 logging_writer.py:48] [142900] global_step=142900, grad_norm=5.107519626617432, loss=1.0755395889282227 +I0831 20:17:15.502620 140462291654400 logging_writer.py:48] [143000] global_step=143000, grad_norm=4.692234516143799, loss=0.9500089883804321 +I0831 20:17:42.680952 140462207792896 logging_writer.py:48] [143100] global_step=143100, grad_norm=4.907933235168457, loss=0.943557858467102 +I0831 20:18:09.805939 140462291654400 logging_writer.py:48] [143200] global_step=143200, grad_norm=4.953145503997803, loss=0.919223427772522 +I0831 20:18:36.930718 140462207792896 logging_writer.py:48] [143300] global_step=143300, grad_norm=5.097940444946289, loss=1.036686897277832 +I0831 20:19:04.092277 140462291654400 logging_writer.py:48] [143400] global_step=143400, grad_norm=4.941688537597656, loss=1.063997745513916 +I0831 20:19:31.211153 140462207792896 logging_writer.py:48] [143500] global_step=143500, grad_norm=4.9455766677856445, loss=0.9716359972953796 +I0831 20:19:58.331164 140462291654400 logging_writer.py:48] [143600] global_step=143600, grad_norm=5.312034606933594, loss=1.1513004302978516 +I0831 20:20:25.534459 140462207792896 logging_writer.py:48] [143700] global_step=143700, grad_norm=4.607305526733398, loss=0.8604352474212646 +I0831 20:20:52.660562 140462291654400 logging_writer.py:48] [143800] global_step=143800, grad_norm=4.973117351531982, loss=0.9257693290710449 +I0831 20:21:19.811162 140462207792896 logging_writer.py:48] [143900] global_step=143900, grad_norm=4.9907684326171875, loss=1.026926040649414 +I0831 20:21:47.222787 140462291654400 logging_writer.py:48] [144000] global_step=144000, grad_norm=4.903079509735107, loss=0.9511042833328247 +I0831 20:22:14.335456 140462207792896 logging_writer.py:48] [144100] global_step=144100, grad_norm=5.007538795471191, loss=1.0664609670639038 +I0831 20:22:41.471020 140462291654400 logging_writer.py:48] [144200] global_step=144200, grad_norm=5.338953971862793, loss=1.0264203548431396 +I0831 20:23:08.654305 140462207792896 logging_writer.py:48] [144300] global_step=144300, grad_norm=4.847156524658203, loss=0.9169592261314392 +I0831 20:23:35.800721 140462291654400 logging_writer.py:48] [144400] global_step=144400, grad_norm=4.880401611328125, loss=0.9410930871963501 +I0831 20:24:02.919296 140462207792896 logging_writer.py:48] [144500] global_step=144500, grad_norm=5.09762716293335, loss=0.9685720205307007 +I0831 20:24:30.128725 140462291654400 logging_writer.py:48] [144600] global_step=144600, grad_norm=4.820201873779297, loss=0.9499877691268921 +I0831 20:24:57.244216 140462207792896 logging_writer.py:48] [144700] global_step=144700, grad_norm=4.785480499267578, loss=0.9770574569702148 +I0831 20:25:24.385845 140462291654400 logging_writer.py:48] [144800] global_step=144800, grad_norm=4.669081687927246, loss=0.9857913255691528 +I0831 20:25:51.552381 140462207792896 logging_writer.py:48] [144900] global_step=144900, grad_norm=4.630598545074463, loss=0.8965394496917725 +I0831 20:26:18.678395 140462291654400 logging_writer.py:48] [145000] global_step=145000, grad_norm=5.073329448699951, loss=0.9998706579208374 +I0831 20:26:46.024077 140462207792896 logging_writer.py:48] [145100] global_step=145100, grad_norm=4.988084316253662, loss=0.9402070641517639 +I0831 20:27:13.197692 140462291654400 logging_writer.py:48] [145200] global_step=145200, grad_norm=5.165643692016602, loss=0.9798785448074341 +I0831 20:27:40.306568 140462207792896 logging_writer.py:48] [145300] global_step=145300, grad_norm=4.889545440673828, loss=0.9754876494407654 +I0831 20:28:07.458778 140462291654400 logging_writer.py:48] [145400] global_step=145400, grad_norm=4.912050724029541, loss=0.9446065425872803 +I0831 20:28:34.662860 140462207792896 logging_writer.py:48] [145500] global_step=145500, grad_norm=5.003946304321289, loss=0.9720159769058228 +I0831 20:29:01.770882 140462291654400 logging_writer.py:48] [145600] global_step=145600, grad_norm=4.841989517211914, loss=0.9266313314437866 +I0831 20:29:28.917506 140462207792896 logging_writer.py:48] [145700] global_step=145700, grad_norm=5.062679290771484, loss=0.9196957945823669 +I0831 20:29:56.116684 140462291654400 logging_writer.py:48] [145800] global_step=145800, grad_norm=4.7233662605285645, loss=0.9329549074172974 +I0831 20:30:23.255583 140462207792896 logging_writer.py:48] [145900] global_step=145900, grad_norm=4.674619197845459, loss=0.9059809446334839 +I0831 20:30:50.392326 140462291654400 logging_writer.py:48] [146000] global_step=146000, grad_norm=5.139963626861572, loss=1.0002278089523315 +I0831 20:31:17.574523 140462207792896 logging_writer.py:48] [146100] global_step=146100, grad_norm=4.982588291168213, loss=0.9433948993682861 +I0831 20:31:19.577448 140659750036672 spec.py:333] Evaluating on the training split. +I0831 20:31:25.847848 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 20:31:35.592710 140659750036672 spec.py:363] Evaluating on the test split. +I0831 20:31:36.469550 140659750036672 submission_runner.py:516] Time since start: 40417.65s, Step: 146109, {'train/accuracy': Array(0.9416255, dtype=float32), 'train/loss': Array(0.20841375, dtype=float32), 'validation/accuracy': Array(0.75446, dtype=float32), 'validation/loss': Array(1.0689094, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6243, dtype=float32), 'test/loss': Array(1.8404001, dtype=float32), 'test/num_examples': 10000, 'score': 39972.9033203125, 'total_duration': 40417.645072460175, 'accumulated_submission_time': 39972.9033203125, 'accumulated_eval_time': 441.8562512397766, 'accumulated_logging_time': 1.663456678390503} +I0831 20:31:36.528123 140462291654400 logging_writer.py:48] [146109] accumulated_eval_time=441.856, accumulated_logging_time=1.66346, accumulated_submission_time=39972.9, global_step=146109, preemption_count=0, score=39972.9, test/accuracy=0.6243000030517578, test/loss=1.8404000997543335, test/num_examples=10000, total_duration=40417.6, train/accuracy=0.9416254758834839, train/loss=0.2084137499332428, validation/accuracy=0.7544599771499634, validation/loss=1.0689094066619873, validation/num_examples=50000 +I0831 20:32:01.878280 140462207792896 logging_writer.py:48] [146200] global_step=146200, grad_norm=5.17712926864624, loss=0.9505095481872559 +I0831 20:32:29.027476 140462291654400 logging_writer.py:48] [146300] global_step=146300, grad_norm=5.046927452087402, loss=0.9485172033309937 +I0831 20:32:56.237148 140462207792896 logging_writer.py:48] [146400] global_step=146400, grad_norm=4.728654384613037, loss=0.9205676913261414 +I0831 20:33:23.366979 140462291654400 logging_writer.py:48] [146500] global_step=146500, grad_norm=4.787416934967041, loss=0.9554644823074341 +I0831 20:33:50.477837 140462207792896 logging_writer.py:48] [146600] global_step=146600, grad_norm=4.852389812469482, loss=0.9383794069290161 +I0831 20:34:17.658504 140462291654400 logging_writer.py:48] [146700] global_step=146700, grad_norm=5.047277450561523, loss=0.9799472689628601 +I0831 20:34:44.794566 140462207792896 logging_writer.py:48] [146800] global_step=146800, grad_norm=5.202159881591797, loss=1.0359314680099487 +I0831 20:35:11.952790 140462291654400 logging_writer.py:48] [146900] global_step=146900, grad_norm=5.147257328033447, loss=1.103712558746338 +I0831 20:35:39.159012 140462207792896 logging_writer.py:48] [147000] global_step=147000, grad_norm=4.922194480895996, loss=0.9959367513656616 +I0831 20:36:06.316476 140462291654400 logging_writer.py:48] [147100] global_step=147100, grad_norm=5.103699684143066, loss=1.0427465438842773 +I0831 20:36:33.460478 140462207792896 logging_writer.py:48] [147200] global_step=147200, grad_norm=4.762744903564453, loss=0.8823360204696655 +I0831 20:37:00.906771 140462291654400 logging_writer.py:48] [147300] global_step=147300, grad_norm=4.916979789733887, loss=0.9913369417190552 +I0831 20:37:28.022621 140462207792896 logging_writer.py:48] [147400] global_step=147400, grad_norm=5.077879428863525, loss=1.008123755455017 +I0831 20:37:55.147078 140462291654400 logging_writer.py:48] [147500] global_step=147500, grad_norm=5.049915790557861, loss=0.9502120614051819 +I0831 20:38:22.348618 140462207792896 logging_writer.py:48] [147600] global_step=147600, grad_norm=5.033355236053467, loss=1.02969491481781 +I0831 20:38:49.509623 140462291654400 logging_writer.py:48] [147700] global_step=147700, grad_norm=4.9608306884765625, loss=0.964897632598877 +I0831 20:39:16.644605 140462207792896 logging_writer.py:48] [147800] global_step=147800, grad_norm=5.05487060546875, loss=0.9029403924942017 +I0831 20:39:43.848089 140462291654400 logging_writer.py:48] [147900] global_step=147900, grad_norm=5.197794437408447, loss=1.049597978591919 +I0831 20:40:10.977881 140462207792896 logging_writer.py:48] [148000] global_step=148000, grad_norm=4.57059907913208, loss=0.9002726078033447 +I0831 20:40:38.116564 140462291654400 logging_writer.py:48] [148100] global_step=148100, grad_norm=5.322535991668701, loss=1.113464593887329 +I0831 20:41:05.305842 140462207792896 logging_writer.py:48] [148200] global_step=148200, grad_norm=5.008724212646484, loss=1.0780854225158691 +I0831 20:41:32.674187 140462291654400 logging_writer.py:48] [148300] global_step=148300, grad_norm=5.109135627746582, loss=0.9403414726257324 +I0831 20:41:59.841852 140462207792896 logging_writer.py:48] [148400] global_step=148400, grad_norm=5.178848743438721, loss=0.966275691986084 +I0831 20:42:27.027464 140462291654400 logging_writer.py:48] [148500] global_step=148500, grad_norm=5.01737642288208, loss=1.0300499200820923 +I0831 20:42:54.183230 140462207792896 logging_writer.py:48] [148600] global_step=148600, grad_norm=4.843362808227539, loss=0.9558639526367188 +I0831 20:43:21.333904 140462291654400 logging_writer.py:48] [148700] global_step=148700, grad_norm=4.926650524139404, loss=1.020301103591919 +I0831 20:43:48.519927 140462207792896 logging_writer.py:48] [148800] global_step=148800, grad_norm=4.95529317855835, loss=0.9226106405258179 +I0831 20:44:15.651443 140462291654400 logging_writer.py:48] [148900] global_step=148900, grad_norm=5.418279647827148, loss=0.9709245562553406 +I0831 20:44:42.808407 140462207792896 logging_writer.py:48] [149000] global_step=149000, grad_norm=4.965385437011719, loss=0.8876020312309265 +I0831 20:45:10.003053 140462291654400 logging_writer.py:48] [149100] global_step=149100, grad_norm=5.225776672363281, loss=1.00492262840271 +I0831 20:45:37.151373 140462207792896 logging_writer.py:48] [149200] global_step=149200, grad_norm=5.009291648864746, loss=1.0167897939682007 +I0831 20:46:04.291003 140462291654400 logging_writer.py:48] [149300] global_step=149300, grad_norm=4.997506618499756, loss=0.9744384288787842 +I0831 20:46:31.695221 140462207792896 logging_writer.py:48] [149400] global_step=149400, grad_norm=5.0499043464660645, loss=1.0428186655044556 +I0831 20:46:58.857033 140462291654400 logging_writer.py:48] [149500] global_step=149500, grad_norm=5.010615348815918, loss=1.0024430751800537 +I0831 20:47:25.997692 140462207792896 logging_writer.py:48] [149600] global_step=149600, grad_norm=4.984416961669922, loss=1.037156343460083 +I0831 20:47:53.196340 140462291654400 logging_writer.py:48] [149700] global_step=149700, grad_norm=4.9696736335754395, loss=0.939487636089325 +I0831 20:48:20.345687 140462207792896 logging_writer.py:48] [149800] global_step=149800, grad_norm=5.135555267333984, loss=1.010632872581482 +I0831 20:48:47.470044 140462291654400 logging_writer.py:48] [149900] global_step=149900, grad_norm=5.011598110198975, loss=0.9588530659675598 +I0831 20:49:14.666305 140462207792896 logging_writer.py:48] [150000] global_step=150000, grad_norm=5.107729911804199, loss=0.9452160596847534 +I0831 20:49:41.801273 140462291654400 logging_writer.py:48] [150100] global_step=150100, grad_norm=4.957622528076172, loss=0.9374066591262817 +I0831 20:50:08.927160 140462207792896 logging_writer.py:48] [150200] global_step=150200, grad_norm=4.7109150886535645, loss=0.9575833082199097 +I0831 20:50:36.109305 140462291654400 logging_writer.py:48] [150300] global_step=150300, grad_norm=5.177845478057861, loss=0.9307512640953064 +I0831 20:51:03.485515 140462207792896 logging_writer.py:48] [150400] global_step=150400, grad_norm=5.326398849487305, loss=1.039961814880371 +I0831 20:51:30.603776 140462291654400 logging_writer.py:48] [150500] global_step=150500, grad_norm=5.097043514251709, loss=1.0228372812271118 +I0831 20:51:57.796148 140462207792896 logging_writer.py:48] [150600] global_step=150600, grad_norm=4.763008117675781, loss=0.9322720170021057 +I0831 20:52:24.929771 140462291654400 logging_writer.py:48] [150700] global_step=150700, grad_norm=4.824267864227295, loss=0.9222245216369629 +I0831 20:52:52.064993 140462207792896 logging_writer.py:48] [150800] global_step=150800, grad_norm=4.837645530700684, loss=0.9711809754371643 +I0831 20:53:19.254137 140462291654400 logging_writer.py:48] [150900] global_step=150900, grad_norm=4.991395473480225, loss=0.9337594509124756 +I0831 20:53:46.387294 140462207792896 logging_writer.py:48] [151000] global_step=151000, grad_norm=4.769856929779053, loss=0.8471380472183228 +I0831 20:54:13.544546 140462291654400 logging_writer.py:48] [151100] global_step=151100, grad_norm=5.241861343383789, loss=1.0724267959594727 +I0831 20:54:40.756809 140462207792896 logging_writer.py:48] [151200] global_step=151200, grad_norm=5.06591272354126, loss=0.9868869185447693 +I0831 20:55:07.913612 140462291654400 logging_writer.py:48] [151300] global_step=151300, grad_norm=5.319314479827881, loss=1.042734146118164 +I0831 20:55:35.042843 140462207792896 logging_writer.py:48] [151400] global_step=151400, grad_norm=4.778836250305176, loss=0.9482230544090271 +I0831 20:56:02.507287 140462291654400 logging_writer.py:48] [151500] global_step=151500, grad_norm=4.377245903015137, loss=0.879317045211792 +I0831 20:56:29.643958 140462207792896 logging_writer.py:48] [151600] global_step=151600, grad_norm=4.816533088684082, loss=0.914557933807373 +I0831 20:56:56.770130 140462291654400 logging_writer.py:48] [151700] global_step=151700, grad_norm=5.197831153869629, loss=0.986128032207489 +I0831 20:57:23.947949 140462207792896 logging_writer.py:48] [151800] global_step=151800, grad_norm=4.891530990600586, loss=1.0468089580535889 +I0831 20:57:51.082072 140462291654400 logging_writer.py:48] [151900] global_step=151900, grad_norm=4.74835205078125, loss=0.9557092189788818 +I0831 20:58:18.211731 140462207792896 logging_writer.py:48] [152000] global_step=152000, grad_norm=4.929506778717041, loss=1.0193893909454346 +I0831 20:58:45.401396 140462291654400 logging_writer.py:48] [152100] global_step=152100, grad_norm=5.043182849884033, loss=1.0067133903503418 +I0831 20:59:12.544354 140462207792896 logging_writer.py:48] [152200] global_step=152200, grad_norm=5.046173572540283, loss=0.9986405968666077 +I0831 20:59:39.645484 140462291654400 logging_writer.py:48] [152300] global_step=152300, grad_norm=4.662707805633545, loss=0.9805785417556763 +I0831 21:00:06.836483 140462207792896 logging_writer.py:48] [152400] global_step=152400, grad_norm=4.850884437561035, loss=0.9784605503082275 +I0831 21:00:34.194904 140462291654400 logging_writer.py:48] [152500] global_step=152500, grad_norm=4.854133605957031, loss=0.9607039093971252 +I0831 21:01:01.323429 140462207792896 logging_writer.py:48] [152600] global_step=152600, grad_norm=5.509657382965088, loss=1.0362211465835571 +I0831 21:01:28.503691 140462291654400 logging_writer.py:48] [152700] global_step=152700, grad_norm=5.2319231033325195, loss=0.9220635294914246 +I0831 21:01:55.637205 140462207792896 logging_writer.py:48] [152800] global_step=152800, grad_norm=4.932852745056152, loss=0.936275064945221 +I0831 21:02:22.765516 140462291654400 logging_writer.py:48] [152900] global_step=152900, grad_norm=4.972334384918213, loss=1.0324690341949463 +I0831 21:02:49.945555 140462207792896 logging_writer.py:48] [153000] global_step=153000, grad_norm=5.067378044128418, loss=0.9535118937492371 +I0831 21:03:17.069672 140462291654400 logging_writer.py:48] [153100] global_step=153100, grad_norm=5.121818542480469, loss=0.9970060586929321 +I0831 21:03:44.196805 140462207792896 logging_writer.py:48] [153200] global_step=153200, grad_norm=4.9269609451293945, loss=0.9534643888473511 +I0831 21:04:11.373622 140462291654400 logging_writer.py:48] [153300] global_step=153300, grad_norm=4.714056491851807, loss=0.9931031465530396 +I0831 21:04:38.520738 140462207792896 logging_writer.py:48] [153400] global_step=153400, grad_norm=4.8814568519592285, loss=0.9286468625068665 +I0831 21:04:52.481662 140659750036672 spec.py:333] Evaluating on the training split. +I0831 21:04:58.969117 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 21:05:09.752521 140659750036672 spec.py:363] Evaluating on the test split. +I0831 21:05:10.644979 140659750036672 submission_runner.py:516] Time since start: 42431.82s, Step: 153453, {'train/accuracy': Array(0.9456513, dtype=float32), 'train/loss': Array(0.19136712, dtype=float32), 'validation/accuracy': Array(0.75391996, dtype=float32), 'validation/loss': Array(1.0708171, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6244, dtype=float32), 'test/loss': Array(1.8389771, dtype=float32), 'test/num_examples': 10000, 'score': 41968.78911828995, 'total_duration': 42431.820268154144, 'accumulated_submission_time': 41968.78911828995, 'accumulated_eval_time': 460.0173342227936, 'accumulated_logging_time': 1.7309489250183105} +I0831 21:05:10.705475 140462291654400 logging_writer.py:48] [153453] accumulated_eval_time=460.017, accumulated_logging_time=1.73095, accumulated_submission_time=41968.8, global_step=153453, preemption_count=0, score=41968.8, test/accuracy=0.6244000196456909, test/loss=1.8389770984649658, test/num_examples=10000, total_duration=42431.8, train/accuracy=0.9456512928009033, train/loss=0.19136711955070496, validation/accuracy=0.7539199590682983, validation/loss=1.0708171129226685, validation/num_examples=50000 +I0831 21:05:23.862630 140462207792896 logging_writer.py:48] [153500] global_step=153500, grad_norm=5.056943416595459, loss=0.9110780954360962 +I0831 21:05:51.301095 140462291654400 logging_writer.py:48] [153600] global_step=153600, grad_norm=5.054113388061523, loss=1.0131895542144775 +I0831 21:06:18.430454 140462207792896 logging_writer.py:48] [153700] global_step=153700, grad_norm=4.929725170135498, loss=1.001912236213684 +I0831 21:06:45.566636 140462291654400 logging_writer.py:48] [153800] global_step=153800, grad_norm=4.732777118682861, loss=0.9106921553611755 +I0831 21:07:12.749039 140462207792896 logging_writer.py:48] [153900] global_step=153900, grad_norm=4.943103790283203, loss=1.000882863998413 +I0831 21:07:39.879129 140462291654400 logging_writer.py:48] [154000] global_step=154000, grad_norm=4.717749118804932, loss=0.9211532473564148 +I0831 21:08:07.007514 140462207792896 logging_writer.py:48] [154100] global_step=154100, grad_norm=4.9016523361206055, loss=1.0056395530700684 +I0831 21:08:34.202372 140462291654400 logging_writer.py:48] [154200] global_step=154200, grad_norm=5.074620246887207, loss=0.9689913988113403 +I0831 21:09:07.931459 140462207792896 logging_writer.py:48] [154300] global_step=154300, grad_norm=4.713864803314209, loss=0.8617621660232544 +I0831 21:09:35.053909 140462291654400 logging_writer.py:48] [154400] global_step=154400, grad_norm=4.681466579437256, loss=0.8951507806777954 +I0831 21:10:02.235129 140462207792896 logging_writer.py:48] [154500] global_step=154500, grad_norm=4.941573619842529, loss=0.9789581298828125 +I0831 21:10:29.592882 140462291654400 logging_writer.py:48] [154600] global_step=154600, grad_norm=4.9301958084106445, loss=0.9613468647003174 +I0831 21:10:56.718068 140462207792896 logging_writer.py:48] [154700] global_step=154700, grad_norm=5.1611552238464355, loss=0.9885886907577515 +I0831 21:11:23.923705 140462291654400 logging_writer.py:48] [154800] global_step=154800, grad_norm=5.204541206359863, loss=1.0234973430633545 +I0831 21:11:51.053797 140462207792896 logging_writer.py:48] [154900] global_step=154900, grad_norm=4.958973407745361, loss=1.0006130933761597 +I0831 21:12:18.186157 140462291654400 logging_writer.py:48] [155000] global_step=155000, grad_norm=4.7439775466918945, loss=0.9392033815383911 +I0831 21:12:45.391417 140462207792896 logging_writer.py:48] [155100] global_step=155100, grad_norm=4.950677394866943, loss=0.9754773378372192 +I0831 21:13:12.541332 140462291654400 logging_writer.py:48] [155200] global_step=155200, grad_norm=5.130274295806885, loss=0.9503422975540161 +I0831 21:13:39.675525 140462207792896 logging_writer.py:48] [155300] global_step=155300, grad_norm=4.981634616851807, loss=0.8916758894920349 +I0831 21:14:06.874484 140462291654400 logging_writer.py:48] [155400] global_step=155400, grad_norm=5.156291484832764, loss=0.9766621589660645 +I0831 21:14:34.043181 140462207792896 logging_writer.py:48] [155500] global_step=155500, grad_norm=5.0920610427856445, loss=1.0361781120300293 +I0831 21:15:01.190714 140462291654400 logging_writer.py:48] [155600] global_step=155600, grad_norm=5.353774547576904, loss=1.0434694290161133 +I0831 21:15:28.648644 140462207792896 logging_writer.py:48] [155700] global_step=155700, grad_norm=4.982146739959717, loss=0.9537841081619263 +I0831 21:15:55.771922 140462291654400 logging_writer.py:48] [155800] global_step=155800, grad_norm=4.984724998474121, loss=1.0238637924194336 +I0831 21:16:22.882650 140462207792896 logging_writer.py:48] [155900] global_step=155900, grad_norm=4.907485008239746, loss=0.8829836845397949 +I0831 21:16:50.056948 140462291654400 logging_writer.py:48] [156000] global_step=156000, grad_norm=4.9478230476379395, loss=0.9675800800323486 +I0831 21:17:17.209203 140462207792896 logging_writer.py:48] [156100] global_step=156100, grad_norm=5.466383457183838, loss=1.0970048904418945 +I0831 21:17:44.326246 140462291654400 logging_writer.py:48] [156200] global_step=156200, grad_norm=5.356410980224609, loss=0.971360981464386 +I0831 21:18:11.497326 140462207792896 logging_writer.py:48] [156300] global_step=156300, grad_norm=4.844964027404785, loss=0.964983344078064 +I0831 21:18:38.621993 140462291654400 logging_writer.py:48] [156400] global_step=156400, grad_norm=4.891105651855469, loss=0.939577579498291 +I0831 21:19:05.779719 140462207792896 logging_writer.py:48] [156500] global_step=156500, grad_norm=4.826744079589844, loss=0.9629478454589844 +I0831 21:19:32.980643 140462291654400 logging_writer.py:48] [156600] global_step=156600, grad_norm=5.404359817504883, loss=1.0200448036193848 +I0831 21:20:00.360680 140462207792896 logging_writer.py:48] [156700] global_step=156700, grad_norm=4.716101169586182, loss=0.8877424001693726 +I0831 21:20:27.510772 140462291654400 logging_writer.py:48] [156800] global_step=156800, grad_norm=5.0705647468566895, loss=0.9997568130493164 +I0831 21:20:54.753931 140462207792896 logging_writer.py:48] [156900] global_step=156900, grad_norm=5.045383453369141, loss=0.9201006889343262 +I0831 21:21:21.919795 140462291654400 logging_writer.py:48] [157000] global_step=157000, grad_norm=4.872276782989502, loss=0.9744699001312256 +I0831 21:21:49.074963 140462207792896 logging_writer.py:48] [157100] global_step=157100, grad_norm=5.259357929229736, loss=1.0818557739257812 +I0831 21:22:16.270503 140462291654400 logging_writer.py:48] [157200] global_step=157200, grad_norm=4.742873191833496, loss=0.9044402837753296 +I0831 21:22:43.398856 140462207792896 logging_writer.py:48] [157300] global_step=157300, grad_norm=5.118371486663818, loss=0.959683895111084 +I0831 21:23:10.565423 140462291654400 logging_writer.py:48] [157400] global_step=157400, grad_norm=5.073634624481201, loss=0.9532342553138733 +I0831 21:23:37.786724 140462207792896 logging_writer.py:48] [157500] global_step=157500, grad_norm=4.770040988922119, loss=0.9331149458885193 +I0831 21:24:04.932130 140462291654400 logging_writer.py:48] [157600] global_step=157600, grad_norm=5.167385578155518, loss=0.9532885551452637 +I0831 21:24:32.096435 140462207792896 logging_writer.py:48] [157700] global_step=157700, grad_norm=5.107757091522217, loss=0.9258207082748413 +I0831 21:24:59.511176 140462291654400 logging_writer.py:48] [157800] global_step=157800, grad_norm=4.911567687988281, loss=1.0154297351837158 +I0831 21:25:26.644470 140462207792896 logging_writer.py:48] [157900] global_step=157900, grad_norm=4.925627708435059, loss=0.9427800178527832 +I0831 21:25:53.791404 140462291654400 logging_writer.py:48] [158000] global_step=158000, grad_norm=4.853490352630615, loss=0.9296988248825073 +I0831 21:26:21.008293 140462207792896 logging_writer.py:48] [158100] global_step=158100, grad_norm=4.589003086090088, loss=0.8321968913078308 +I0831 21:26:48.132665 140462291654400 logging_writer.py:48] [158200] global_step=158200, grad_norm=5.013749122619629, loss=0.9791020154953003 +I0831 21:27:15.299751 140462207792896 logging_writer.py:48] [158300] global_step=158300, grad_norm=4.7193450927734375, loss=0.895133376121521 +I0831 21:27:42.502954 140462291654400 logging_writer.py:48] [158400] global_step=158400, grad_norm=4.781099796295166, loss=0.9597982168197632 +I0831 21:28:09.651352 140462207792896 logging_writer.py:48] [158500] global_step=158500, grad_norm=4.635530471801758, loss=0.921591579914093 +I0831 21:28:36.769594 140462291654400 logging_writer.py:48] [158600] global_step=158600, grad_norm=5.455977439880371, loss=1.001595139503479 +I0831 21:29:03.985689 140462207792896 logging_writer.py:48] [158700] global_step=158700, grad_norm=5.101857662200928, loss=0.9711489081382751 +I0831 21:29:31.319957 140462291654400 logging_writer.py:48] [158800] global_step=158800, grad_norm=4.8571624755859375, loss=0.8531947731971741 +I0831 21:29:58.444665 140462207792896 logging_writer.py:48] [158900] global_step=158900, grad_norm=5.20810604095459, loss=1.0924031734466553 +I0831 21:30:25.631657 140462291654400 logging_writer.py:48] [159000] global_step=159000, grad_norm=5.144449234008789, loss=0.9600377082824707 +I0831 21:30:52.817534 140462207792896 logging_writer.py:48] [159100] global_step=159100, grad_norm=4.589337348937988, loss=0.9452648162841797 +I0831 21:31:19.950645 140462291654400 logging_writer.py:48] [159200] global_step=159200, grad_norm=4.936088562011719, loss=1.0741868019104004 +I0831 21:31:47.319860 140462207792896 logging_writer.py:48] [159300] global_step=159300, grad_norm=4.875755786895752, loss=0.9632929563522339 +I0831 21:32:14.482379 140462291654400 logging_writer.py:48] [159400] global_step=159400, grad_norm=4.551681041717529, loss=0.8740323185920715 +I0831 21:32:41.609766 140462207792896 logging_writer.py:48] [159500] global_step=159500, grad_norm=4.919926643371582, loss=0.9471516609191895 +I0831 21:33:08.814949 140462291654400 logging_writer.py:48] [159600] global_step=159600, grad_norm=4.949033260345459, loss=0.9478359818458557 +I0831 21:33:35.956482 140462207792896 logging_writer.py:48] [159700] global_step=159700, grad_norm=5.112330436706543, loss=0.9988484382629395 +I0831 21:34:03.060277 140462291654400 logging_writer.py:48] [159800] global_step=159800, grad_norm=5.1288743019104, loss=0.9346522092819214 +I0831 21:34:30.512587 140462207792896 logging_writer.py:48] [159900] global_step=159900, grad_norm=4.84617280960083, loss=0.9551305174827576 +I0831 21:34:57.639428 140462291654400 logging_writer.py:48] [160000] global_step=160000, grad_norm=5.356376647949219, loss=0.979600727558136 +I0831 21:35:24.776584 140462207792896 logging_writer.py:48] [160100] global_step=160100, grad_norm=5.229974746704102, loss=1.0839356184005737 +I0831 21:35:51.977927 140462291654400 logging_writer.py:48] [160200] global_step=160200, grad_norm=4.77571439743042, loss=0.9437277317047119 +I0831 21:36:19.133797 140462207792896 logging_writer.py:48] [160300] global_step=160300, grad_norm=4.864029884338379, loss=0.9823747277259827 +I0831 21:36:46.280523 140462291654400 logging_writer.py:48] [160400] global_step=160400, grad_norm=5.000840663909912, loss=0.9484063982963562 +I0831 21:37:13.513010 140462207792896 logging_writer.py:48] [160500] global_step=160500, grad_norm=4.963239669799805, loss=0.9105367660522461 +I0831 21:37:40.655997 140462291654400 logging_writer.py:48] [160600] global_step=160600, grad_norm=5.134299278259277, loss=0.9874027967453003 +I0831 21:38:07.798678 140462207792896 logging_writer.py:48] [160700] global_step=160700, grad_norm=4.865184783935547, loss=0.9291127920150757 +I0831 21:38:26.701415 140659750036672 spec.py:333] Evaluating on the training split. +I0831 21:38:33.055192 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 21:38:44.212057 140659750036672 spec.py:363] Evaluating on the test split. +I0831 21:38:45.095850 140659750036672 submission_runner.py:516] Time since start: 44446.27s, Step: 160771, {'train/accuracy': Array(0.94876033, dtype=float32), 'train/loss': Array(0.18166068, dtype=float32), 'validation/accuracy': Array(0.75387996, dtype=float32), 'validation/loss': Array(1.07235, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62450004, dtype=float32), 'test/loss': Array(1.8409966, dtype=float32), 'test/num_examples': 10000, 'score': 43964.716457128525, 'total_duration': 44446.27172589302, 'accumulated_submission_time': 43964.716457128525, 'accumulated_eval_time': 478.41012358665466, 'accumulated_logging_time': 1.80169677734375} +I0831 21:38:45.162418 140462291654400 logging_writer.py:48] [160771] accumulated_eval_time=478.41, accumulated_logging_time=1.8017, accumulated_submission_time=43964.7, global_step=160771, preemption_count=0, score=43964.7, test/accuracy=0.624500036239624, test/loss=1.8409966230392456, test/num_examples=10000, total_duration=44446.3, train/accuracy=0.9487603306770325, train/loss=0.18166068196296692, validation/accuracy=0.753879964351654, validation/loss=1.072350025177002, validation/num_examples=50000 +I0831 21:38:53.357166 140462207792896 logging_writer.py:48] [160800] global_step=160800, grad_norm=5.098038196563721, loss=0.9746597409248352 +I0831 21:39:20.465385 140462291654400 logging_writer.py:48] [160900] global_step=160900, grad_norm=5.095700263977051, loss=1.0124276876449585 +I0831 21:39:55.239290 140462207792896 logging_writer.py:48] [161000] global_step=161000, grad_norm=5.043336868286133, loss=1.0014071464538574 +I0831 21:40:38.513604 140462291654400 logging_writer.py:48] [161100] global_step=161100, grad_norm=5.063284873962402, loss=0.9685930013656616 +I0831 21:41:23.359558 140462207792896 logging_writer.py:48] [161200] global_step=161200, grad_norm=5.114185810089111, loss=0.9995945692062378 +I0831 21:42:07.467256 140462291654400 logging_writer.py:48] [161300] global_step=161300, grad_norm=4.876865386962891, loss=0.9169625043869019 +I0831 21:42:48.845308 140462207792896 logging_writer.py:48] [161400] global_step=161400, grad_norm=4.909302234649658, loss=0.9567756652832031 +I0831 21:43:16.045150 140462291654400 logging_writer.py:48] [161500] global_step=161500, grad_norm=5.066482067108154, loss=0.9473592042922974 +I0831 21:43:43.179838 140462207792896 logging_writer.py:48] [161600] global_step=161600, grad_norm=4.779394149780273, loss=0.8306132555007935 +I0831 21:44:10.375538 140462291654400 logging_writer.py:48] [161700] global_step=161700, grad_norm=5.407292366027832, loss=1.0409095287322998 +I0831 21:44:37.526360 140462207792896 logging_writer.py:48] [161800] global_step=161800, grad_norm=4.9781012535095215, loss=0.9446587562561035 +I0831 21:45:04.678123 140462291654400 logging_writer.py:48] [161900] global_step=161900, grad_norm=4.614935874938965, loss=0.8507758378982544 +I0831 21:45:32.105823 140462207792896 logging_writer.py:48] [162000] global_step=162000, grad_norm=4.996749401092529, loss=0.9923340678215027 +I0831 21:45:59.217583 140462291654400 logging_writer.py:48] [162100] global_step=162100, grad_norm=4.919920921325684, loss=0.9165802001953125 +I0831 21:46:26.359838 140462207792896 logging_writer.py:48] [162200] global_step=162200, grad_norm=5.001339912414551, loss=0.9442132711410522 +I0831 21:46:53.547977 140462291654400 logging_writer.py:48] [162300] global_step=162300, grad_norm=5.357915878295898, loss=0.9715074896812439 +I0831 21:47:24.361074 140462207792896 logging_writer.py:48] [162400] global_step=162400, grad_norm=5.192262649536133, loss=0.9739127159118652 +I0831 21:47:52.472787 140462291654400 logging_writer.py:48] [162500] global_step=162500, grad_norm=5.2324652671813965, loss=0.9767922163009644 +I0831 21:48:19.653805 140462207792896 logging_writer.py:48] [162600] global_step=162600, grad_norm=4.980734825134277, loss=1.0035372972488403 +I0831 21:48:46.782135 140462291654400 logging_writer.py:48] [162700] global_step=162700, grad_norm=4.885928630828857, loss=0.8694019913673401 +I0831 21:49:13.907143 140462207792896 logging_writer.py:48] [162800] global_step=162800, grad_norm=5.127648830413818, loss=0.9809357523918152 +I0831 21:49:41.121764 140462291654400 logging_writer.py:48] [162900] global_step=162900, grad_norm=5.048933029174805, loss=0.9566844701766968 +I0831 21:50:08.255216 140462207792896 logging_writer.py:48] [163000] global_step=163000, grad_norm=4.877925395965576, loss=0.8979131579399109 +I0831 21:50:35.603696 140462291654400 logging_writer.py:48] [163100] global_step=163100, grad_norm=4.637904167175293, loss=0.8578420877456665 +I0831 21:51:02.801772 140462207792896 logging_writer.py:48] [163200] global_step=163200, grad_norm=5.0354413986206055, loss=0.9485828280448914 +I0831 21:51:29.975018 140462291654400 logging_writer.py:48] [163300] global_step=163300, grad_norm=5.110825538635254, loss=0.9954307675361633 +I0831 21:51:57.095324 140462207792896 logging_writer.py:48] [163400] global_step=163400, grad_norm=5.001240253448486, loss=0.9435819983482361 +I0831 21:52:24.289131 140462291654400 logging_writer.py:48] [163500] global_step=163500, grad_norm=5.503946304321289, loss=0.9919255971908569 +I0831 21:52:51.410598 140462207792896 logging_writer.py:48] [163600] global_step=163600, grad_norm=5.035711288452148, loss=1.0071194171905518 +I0831 21:53:18.542016 140462291654400 logging_writer.py:48] [163700] global_step=163700, grad_norm=5.052321910858154, loss=0.9421260952949524 +I0831 21:53:45.735125 140462207792896 logging_writer.py:48] [163800] global_step=163800, grad_norm=5.289350509643555, loss=1.1293590068817139 +I0831 21:54:12.906118 140462291654400 logging_writer.py:48] [163900] global_step=163900, grad_norm=4.6522345542907715, loss=0.8945479393005371 +I0831 21:54:40.028523 140462207792896 logging_writer.py:48] [164000] global_step=164000, grad_norm=4.963842391967773, loss=0.9680155515670776 +I0831 21:55:07.244997 140462291654400 logging_writer.py:48] [164100] global_step=164100, grad_norm=5.109665870666504, loss=0.9674621820449829 +I0831 21:55:34.592551 140462207792896 logging_writer.py:48] [164200] global_step=164200, grad_norm=5.462714672088623, loss=0.9511999487876892 +I0831 21:56:01.718972 140462291654400 logging_writer.py:48] [164300] global_step=164300, grad_norm=5.171063423156738, loss=1.0255486965179443 +I0831 21:56:28.910925 140462207792896 logging_writer.py:48] [164400] global_step=164400, grad_norm=5.282876014709473, loss=1.0285884141921997 +I0831 21:56:56.068378 140462291654400 logging_writer.py:48] [164500] global_step=164500, grad_norm=4.706733703613281, loss=0.8541030287742615 +I0831 21:57:23.201829 140462207792896 logging_writer.py:48] [164600] global_step=164600, grad_norm=5.162103176116943, loss=0.9180993437767029 +I0831 21:57:50.393200 140462291654400 logging_writer.py:48] [164700] global_step=164700, grad_norm=5.020479202270508, loss=0.9766398668289185 +I0831 21:58:17.511663 140462207792896 logging_writer.py:48] [164800] global_step=164800, grad_norm=4.7006731033325195, loss=0.8993335962295532 +I0831 21:58:44.644565 140462291654400 logging_writer.py:48] [164900] global_step=164900, grad_norm=4.851202011108398, loss=0.9187802672386169 +I0831 21:59:11.841597 140462207792896 logging_writer.py:48] [165000] global_step=165000, grad_norm=4.951328754425049, loss=0.9593105912208557 +I0831 21:59:38.977236 140462291654400 logging_writer.py:48] [165100] global_step=165100, grad_norm=5.194088459014893, loss=0.9817239046096802 +I0831 22:00:06.101523 140462207792896 logging_writer.py:48] [165200] global_step=165200, grad_norm=5.032107353210449, loss=0.9262283444404602 +I0831 22:00:33.593572 140462291654400 logging_writer.py:48] [165300] global_step=165300, grad_norm=5.002983093261719, loss=0.9303833246231079 +I0831 22:01:00.729788 140462207792896 logging_writer.py:48] [165400] global_step=165400, grad_norm=5.304816722869873, loss=0.9452857971191406 +I0831 22:01:27.868565 140462291654400 logging_writer.py:48] [165500] global_step=165500, grad_norm=5.339747428894043, loss=1.0003089904785156 +I0831 22:01:55.064239 140462207792896 logging_writer.py:48] [165600] global_step=165600, grad_norm=5.1085004806518555, loss=1.0624154806137085 +I0831 22:02:22.203017 140462291654400 logging_writer.py:48] [165700] global_step=165700, grad_norm=4.9945573806762695, loss=0.9381334185600281 +I0831 22:02:49.308668 140462207792896 logging_writer.py:48] [165800] global_step=165800, grad_norm=4.689522743225098, loss=0.9119595885276794 +I0831 22:03:16.495235 140462291654400 logging_writer.py:48] [165900] global_step=165900, grad_norm=4.885805606842041, loss=0.9081794023513794 +I0831 22:03:43.635086 140462207792896 logging_writer.py:48] [166000] global_step=166000, grad_norm=5.234796047210693, loss=1.0034908056259155 +I0831 22:04:10.776830 140462291654400 logging_writer.py:48] [166100] global_step=166100, grad_norm=5.067605495452881, loss=0.9637454152107239 +I0831 22:04:37.955425 140462207792896 logging_writer.py:48] [166200] global_step=166200, grad_norm=5.16677188873291, loss=1.0152298212051392 +I0831 22:05:05.330614 140462291654400 logging_writer.py:48] [166300] global_step=166300, grad_norm=5.018907070159912, loss=0.9588017463684082 +I0831 22:05:32.490901 140462207792896 logging_writer.py:48] [166400] global_step=166400, grad_norm=5.031796455383301, loss=0.9180569648742676 +I0831 22:05:59.679502 140462291654400 logging_writer.py:48] [166500] global_step=166500, grad_norm=4.960878372192383, loss=0.8948320150375366 +I0831 22:06:26.797247 140462207792896 logging_writer.py:48] [166600] global_step=166600, grad_norm=4.9406256675720215, loss=0.9352167844772339 +I0831 22:06:53.912856 140462291654400 logging_writer.py:48] [166700] global_step=166700, grad_norm=5.076388835906982, loss=0.9608315229415894 +I0831 22:07:21.111330 140462207792896 logging_writer.py:48] [166800] global_step=166800, grad_norm=5.068970680236816, loss=0.9386606812477112 +I0831 22:07:48.240593 140462291654400 logging_writer.py:48] [166900] global_step=166900, grad_norm=4.996463775634766, loss=0.9130566716194153 +I0831 22:08:15.388851 140462207792896 logging_writer.py:48] [167000] global_step=167000, grad_norm=5.340920448303223, loss=1.0733318328857422 +I0831 22:08:42.582410 140462291654400 logging_writer.py:48] [167100] global_step=167100, grad_norm=4.7588725090026855, loss=0.850719690322876 +I0831 22:09:09.719810 140462207792896 logging_writer.py:48] [167200] global_step=167200, grad_norm=5.0988006591796875, loss=0.9281388521194458 +I0831 22:09:36.890285 140462291654400 logging_writer.py:48] [167300] global_step=167300, grad_norm=5.188480854034424, loss=0.9607030153274536 +I0831 22:10:04.303669 140462207792896 logging_writer.py:48] [167400] global_step=167400, grad_norm=5.070847034454346, loss=0.9176207184791565 +I0831 22:10:31.440529 140462291654400 logging_writer.py:48] [167500] global_step=167500, grad_norm=5.078551769256592, loss=0.8954919576644897 +I0831 22:10:58.586200 140462207792896 logging_writer.py:48] [167600] global_step=167600, grad_norm=5.082925796508789, loss=1.0074293613433838 +I0831 22:11:25.785268 140462291654400 logging_writer.py:48] [167700] global_step=167700, grad_norm=5.373117446899414, loss=0.9898949265480042 +I0831 22:11:52.925725 140462207792896 logging_writer.py:48] [167800] global_step=167800, grad_norm=5.000267028808594, loss=0.9166758060455322 +I0831 22:12:01.215854 140659750036672 spec.py:333] Evaluating on the training split. +I0831 22:12:07.611867 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 22:12:19.544370 140659750036672 spec.py:363] Evaluating on the test split. +I0831 22:12:20.441216 140659750036672 submission_runner.py:516] Time since start: 46461.62s, Step: 167832, {'train/accuracy': Array(0.9493782, dtype=float32), 'train/loss': Array(0.18224207, dtype=float32), 'validation/accuracy': Array(0.75308, dtype=float32), 'validation/loss': Array(1.0753248, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6258, dtype=float32), 'test/loss': Array(1.850111, dtype=float32), 'test/num_examples': 10000, 'score': 45960.701127529144, 'total_duration': 46461.61534452438, 'accumulated_submission_time': 45960.701127529144, 'accumulated_eval_time': 497.63209295272827, 'accumulated_logging_time': 1.8793811798095703} +I0831 22:12:20.563740 140462291654400 logging_writer.py:48] [167832] accumulated_eval_time=497.632, accumulated_logging_time=1.87938, accumulated_submission_time=45960.7, global_step=167832, preemption_count=0, score=45960.7, test/accuracy=0.6258000135421753, test/loss=1.8501110076904297, test/num_examples=10000, total_duration=46461.6, train/accuracy=0.9493781924247742, train/loss=0.18224206566810608, validation/accuracy=0.7530800104141235, validation/loss=1.0753247737884521, validation/num_examples=50000 +I0831 22:12:39.361716 140462207792896 logging_writer.py:48] [167900] global_step=167900, grad_norm=5.249515533447266, loss=0.9779113531112671 +I0831 22:13:06.531573 140462291654400 logging_writer.py:48] [168000] global_step=168000, grad_norm=4.947891712188721, loss=0.9121688604354858 +I0831 22:13:37.970142 140462207792896 logging_writer.py:48] [168100] global_step=168100, grad_norm=5.0595221519470215, loss=1.006283164024353 +I0831 22:14:20.050939 140462291654400 logging_writer.py:48] [168200] global_step=168200, grad_norm=4.771828651428223, loss=0.9053659439086914 +I0831 22:15:02.916015 140462207792896 logging_writer.py:48] [168300] global_step=168300, grad_norm=4.841762065887451, loss=0.9043993949890137 +I0831 22:15:46.139372 140462291654400 logging_writer.py:48] [168400] global_step=168400, grad_norm=5.188339710235596, loss=0.9599301218986511 +I0831 22:16:28.596869 140462207792896 logging_writer.py:48] [168500] global_step=168500, grad_norm=5.2734832763671875, loss=1.0038108825683594 +I0831 22:17:11.957236 140462291654400 logging_writer.py:48] [168600] global_step=168600, grad_norm=4.854990482330322, loss=0.9786995053291321 +I0831 22:17:54.846849 140462207792896 logging_writer.py:48] [168700] global_step=168700, grad_norm=5.28982400894165, loss=0.9922857284545898 +I0831 22:18:38.281803 140462291654400 logging_writer.py:48] [168800] global_step=168800, grad_norm=5.279576301574707, loss=1.0417850017547607 +I0831 22:19:17.872385 140462207792896 logging_writer.py:48] [168900] global_step=168900, grad_norm=4.88008975982666, loss=0.9583238363265991 +I0831 22:19:45.962125 140462291654400 logging_writer.py:48] [169000] global_step=169000, grad_norm=5.051194190979004, loss=0.8823865652084351 +I0831 22:20:13.667649 140462207792896 logging_writer.py:48] [169100] global_step=169100, grad_norm=5.224035739898682, loss=0.9260735511779785 +I0831 22:20:41.052379 140462291654400 logging_writer.py:48] [169200] global_step=169200, grad_norm=5.120239734649658, loss=1.0295157432556152 +I0831 22:21:20.155750 140462207792896 logging_writer.py:48] [169300] global_step=169300, grad_norm=4.997237205505371, loss=0.9371074438095093 +I0831 22:22:05.152720 140462291654400 logging_writer.py:48] [169400] global_step=169400, grad_norm=4.812235355377197, loss=0.9006893634796143 +I0831 22:22:48.912215 140462207792896 logging_writer.py:48] [169500] global_step=169500, grad_norm=4.933948993682861, loss=0.9436824917793274 +I0831 22:23:29.916634 140462291654400 logging_writer.py:48] [169600] global_step=169600, grad_norm=4.9962310791015625, loss=0.9614154696464539 +I0831 22:24:10.534910 140462207792896 logging_writer.py:48] [169700] global_step=169700, grad_norm=4.866765022277832, loss=0.9026402831077576 +I0831 22:24:52.484248 140462291654400 logging_writer.py:48] [169800] global_step=169800, grad_norm=4.952950477600098, loss=0.945799708366394 +I0831 22:25:34.941564 140462207792896 logging_writer.py:48] [169900] global_step=169900, grad_norm=5.139213562011719, loss=0.9639832973480225 +I0831 22:26:17.609127 140462291654400 logging_writer.py:48] [170000] global_step=170000, grad_norm=4.9815263748168945, loss=0.880686342716217 +I0831 22:26:59.548337 140462207792896 logging_writer.py:48] [170100] global_step=170100, grad_norm=5.028893947601318, loss=0.9958508014678955 +I0831 22:27:43.940606 140462291654400 logging_writer.py:48] [170200] global_step=170200, grad_norm=5.083122253417969, loss=0.9931449890136719 +I0831 22:28:27.609377 140462207792896 logging_writer.py:48] [170300] global_step=170300, grad_norm=5.182442665100098, loss=0.9561965465545654 +I0831 22:29:10.546769 140462291654400 logging_writer.py:48] [170400] global_step=170400, grad_norm=5.034732818603516, loss=0.9193260669708252 +I0831 22:29:52.377224 140462207792896 logging_writer.py:48] [170500] global_step=170500, grad_norm=5.213207244873047, loss=0.9506373405456543 +I0831 22:30:35.620006 140462291654400 logging_writer.py:48] [170600] global_step=170600, grad_norm=5.281569004058838, loss=0.9350639581680298 +I0831 22:31:20.132289 140462207792896 logging_writer.py:48] [170700] global_step=170700, grad_norm=5.124488353729248, loss=0.9277832508087158 +I0831 22:32:03.582141 140462291654400 logging_writer.py:48] [170800] global_step=170800, grad_norm=4.886213302612305, loss=0.9536986351013184 +I0831 22:32:45.789853 140462207792896 logging_writer.py:48] [170900] global_step=170900, grad_norm=5.06386137008667, loss=0.9789506196975708 +I0831 22:33:28.447354 140462291654400 logging_writer.py:48] [171000] global_step=171000, grad_norm=4.8332390785217285, loss=0.9135371446609497 +I0831 22:34:10.364440 140462207792896 logging_writer.py:48] [171100] global_step=171100, grad_norm=5.0275678634643555, loss=0.9423391222953796 +I0831 22:34:53.801843 140462291654400 logging_writer.py:48] [171200] global_step=171200, grad_norm=4.8156938552856445, loss=0.9010908007621765 +I0831 22:35:38.533423 140462207792896 logging_writer.py:48] [171300] global_step=171300, grad_norm=5.0445404052734375, loss=0.9327936768531799 +I0831 22:36:24.490343 140462291654400 logging_writer.py:48] [171400] global_step=171400, grad_norm=5.077574253082275, loss=0.9776238203048706 +I0831 22:37:08.598884 140462207792896 logging_writer.py:48] [171500] global_step=171500, grad_norm=5.1444478034973145, loss=0.9815625548362732 +I0831 22:37:54.914034 140462291654400 logging_writer.py:48] [171600] global_step=171600, grad_norm=4.952216625213623, loss=0.9028355479240417 +I0831 22:38:41.361954 140462207792896 logging_writer.py:48] [171700] global_step=171700, grad_norm=5.315011978149414, loss=0.9862964153289795 +I0831 22:39:28.852830 140462291654400 logging_writer.py:48] [171800] global_step=171800, grad_norm=4.855005741119385, loss=0.9074143171310425 +I0831 22:40:17.278409 140462207792896 logging_writer.py:48] [171900] global_step=171900, grad_norm=5.167689323425293, loss=0.9997232556343079 +I0831 22:41:09.024457 140462291654400 logging_writer.py:48] [172000] global_step=172000, grad_norm=4.871642112731934, loss=0.9030478000640869 +I0831 22:41:59.241652 140462207792896 logging_writer.py:48] [172100] global_step=172100, grad_norm=4.943066596984863, loss=0.9310522079467773 +I0831 22:42:51.092289 140462291654400 logging_writer.py:48] [172200] global_step=172200, grad_norm=5.299054145812988, loss=0.9719447493553162 +I0831 22:43:41.175223 140462207792896 logging_writer.py:48] [172300] global_step=172300, grad_norm=5.20459508895874, loss=0.9850228428840637 +I0831 22:44:30.510998 140462291654400 logging_writer.py:48] [172400] global_step=172400, grad_norm=5.192038536071777, loss=1.0121500492095947 +I0831 22:45:19.684081 140462207792896 logging_writer.py:48] [172500] global_step=172500, grad_norm=4.864529132843018, loss=0.8867329955101013 +I0831 22:45:36.645315 140659750036672 spec.py:333] Evaluating on the training split. +I0831 22:45:45.035536 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 22:46:14.289816 140659750036672 spec.py:363] Evaluating on the test split. +I0831 22:46:15.369348 140659750036672 submission_runner.py:516] Time since start: 48496.35s, Step: 172536, {'train/accuracy': Array(0.9483219, dtype=float32), 'train/loss': Array(0.18042949, dtype=float32), 'validation/accuracy': Array(0.75468, dtype=float32), 'validation/loss': Array(1.0734031, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62750006, dtype=float32), 'test/loss': Array(1.8528265, dtype=float32), 'test/num_examples': 10000, 'score': 47956.73145079613, 'total_duration': 48496.35363531113, 'accumulated_submission_time': 47956.73145079613, 'accumulated_eval_time': 536.1628928184509, 'accumulated_logging_time': 2.010361909866333} +I0831 22:46:15.873483 140462291654400 logging_writer.py:48] [172536] accumulated_eval_time=536.163, accumulated_logging_time=2.01036, accumulated_submission_time=47956.7, global_step=172536, preemption_count=0, score=47956.7, test/accuracy=0.627500057220459, test/loss=1.852826476097107, test/num_examples=10000, total_duration=48496.4, train/accuracy=0.9483218789100647, train/loss=0.18042948842048645, validation/accuracy=0.7546799778938293, validation/loss=1.0734031200408936, validation/num_examples=50000 +I0831 22:46:42.520467 140462207792896 logging_writer.py:48] [172600] global_step=172600, grad_norm=4.952641010284424, loss=0.9319667816162109 +I0831 22:47:30.214604 140462291654400 logging_writer.py:48] [172700] global_step=172700, grad_norm=4.818584442138672, loss=0.9227604866027832 +I0831 22:48:17.454384 140462207792896 logging_writer.py:48] [172800] global_step=172800, grad_norm=5.173440456390381, loss=1.0223735570907593 +I0831 22:49:06.222964 140462291654400 logging_writer.py:48] [172900] global_step=172900, grad_norm=5.130667209625244, loss=0.9341515302658081 +I0831 22:49:54.743420 140462207792896 logging_writer.py:48] [173000] global_step=173000, grad_norm=4.85006046295166, loss=0.9061501026153564 +I0831 22:50:42.810330 140462291654400 logging_writer.py:48] [173100] global_step=173100, grad_norm=5.253986358642578, loss=1.0068721771240234 +I0831 22:51:29.252436 140462207792896 logging_writer.py:48] [173200] global_step=173200, grad_norm=5.012919902801514, loss=0.917576014995575 +I0831 22:52:15.753137 140462291654400 logging_writer.py:48] [173300] global_step=173300, grad_norm=5.046753883361816, loss=0.9700452089309692 +I0831 22:53:00.476390 140462207792896 logging_writer.py:48] [173400] global_step=173400, grad_norm=4.817607879638672, loss=0.931251585483551 +I0831 22:53:44.061093 140462291654400 logging_writer.py:48] [173500] global_step=173500, grad_norm=5.2205424308776855, loss=0.94095778465271 +I0831 22:54:27.765555 140462207792896 logging_writer.py:48] [173600] global_step=173600, grad_norm=5.071459770202637, loss=0.9952388405799866 +I0831 22:55:12.536483 140462291654400 logging_writer.py:48] [173700] global_step=173700, grad_norm=4.991672992706299, loss=0.9414447546005249 +I0831 22:55:56.837494 140462207792896 logging_writer.py:48] [173800] global_step=173800, grad_norm=5.235715866088867, loss=1.0271356105804443 +I0831 22:56:40.227126 140462291654400 logging_writer.py:48] [173900] global_step=173900, grad_norm=4.8656325340271, loss=0.8749480247497559 +I0831 22:57:25.781465 140462207792896 logging_writer.py:48] [174000] global_step=174000, grad_norm=4.987722873687744, loss=0.9106559157371521 +I0831 22:58:09.783246 140462291654400 logging_writer.py:48] [174100] global_step=174100, grad_norm=5.120002746582031, loss=0.9853039979934692 +I0831 22:58:52.881175 140462207792896 logging_writer.py:48] [174200] global_step=174200, grad_norm=4.902224063873291, loss=0.9015776515007019 +I0831 22:59:35.447818 140462291654400 logging_writer.py:48] [174300] global_step=174300, grad_norm=5.2856831550598145, loss=0.9611312747001648 +I0831 23:00:19.423662 140462207792896 logging_writer.py:48] [174400] global_step=174400, grad_norm=4.720272064208984, loss=0.8351663947105408 +I0831 23:01:04.165485 140462291654400 logging_writer.py:48] [174500] global_step=174500, grad_norm=4.996337413787842, loss=0.9317958950996399 +I0831 23:01:46.671909 140462207792896 logging_writer.py:48] [174600] global_step=174600, grad_norm=4.925080299377441, loss=0.9542815685272217 +I0831 23:02:29.522642 140462291654400 logging_writer.py:48] [174700] global_step=174700, grad_norm=5.25102424621582, loss=1.0014592409133911 +I0831 23:03:11.704177 140462207792896 logging_writer.py:48] [174800] global_step=174800, grad_norm=5.045250415802002, loss=0.933146595954895 +I0831 23:03:54.511243 140462291654400 logging_writer.py:48] [174900] global_step=174900, grad_norm=5.323514461517334, loss=0.9493603706359863 +I0831 23:04:39.342868 140462207792896 logging_writer.py:48] [175000] global_step=175000, grad_norm=5.004557132720947, loss=0.975064218044281 +I0831 23:05:23.392998 140462291654400 logging_writer.py:48] [175100] global_step=175100, grad_norm=4.905547142028809, loss=0.9189598560333252 +I0831 23:06:08.910784 140462207792896 logging_writer.py:48] [175200] global_step=175200, grad_norm=5.169896125793457, loss=0.921867847442627 +I0831 23:06:52.146848 140462291654400 logging_writer.py:48] [175300] global_step=175300, grad_norm=4.7265706062316895, loss=0.824954628944397 +I0831 23:07:37.516496 140462207792896 logging_writer.py:48] [175400] global_step=175400, grad_norm=5.315758228302002, loss=0.9758365750312805 +I0831 23:08:21.836453 140462291654400 logging_writer.py:48] [175500] global_step=175500, grad_norm=5.231678009033203, loss=0.9635101556777954 +I0831 23:09:05.186280 140462207792896 logging_writer.py:48] [175600] global_step=175600, grad_norm=5.154740333557129, loss=0.9458163976669312 +I0831 23:09:49.731126 140462291654400 logging_writer.py:48] [175700] global_step=175700, grad_norm=4.905780792236328, loss=0.938208281993866 +I0831 23:10:32.689327 140462207792896 logging_writer.py:48] [175800] global_step=175800, grad_norm=4.883871078491211, loss=0.9667889475822449 +I0831 23:11:15.946782 140462291654400 logging_writer.py:48] [175900] global_step=175900, grad_norm=4.90940523147583, loss=0.9547457695007324 +I0831 23:11:59.168535 140462207792896 logging_writer.py:48] [176000] global_step=176000, grad_norm=5.070376396179199, loss=0.9594306945800781 +I0831 23:12:43.502179 140462291654400 logging_writer.py:48] [176100] global_step=176100, grad_norm=5.105963230133057, loss=0.9802647829055786 +I0831 23:13:25.875019 140462207792896 logging_writer.py:48] [176200] global_step=176200, grad_norm=5.159963607788086, loss=1.0186784267425537 +I0831 23:14:10.202975 140462291654400 logging_writer.py:48] [176300] global_step=176300, grad_norm=5.227097034454346, loss=0.8960621356964111 +I0831 23:14:55.013187 140462207792896 logging_writer.py:48] [176400] global_step=176400, grad_norm=5.1046576499938965, loss=0.9798374176025391 +I0831 23:15:38.931752 140462291654400 logging_writer.py:48] [176500] global_step=176500, grad_norm=5.162769317626953, loss=0.9073487520217896 +I0831 23:16:23.981596 140462207792896 logging_writer.py:48] [176600] global_step=176600, grad_norm=5.0620598793029785, loss=0.9283930063247681 +I0831 23:17:06.849113 140462291654400 logging_writer.py:48] [176700] global_step=176700, grad_norm=4.8068766593933105, loss=0.9714906811714172 +I0831 23:17:52.655359 140462207792896 logging_writer.py:48] [176800] global_step=176800, grad_norm=4.8712592124938965, loss=0.8779171705245972 +I0831 23:18:37.031484 140462291654400 logging_writer.py:48] [176900] global_step=176900, grad_norm=5.1905837059021, loss=0.9603651165962219 +I0831 23:19:19.323199 140462207792896 logging_writer.py:48] [177000] global_step=177000, grad_norm=5.116184711456299, loss=0.8888077735900879 +I0831 23:19:31.551317 140659750036672 spec.py:333] Evaluating on the training split. +I0831 23:19:39.755292 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 23:20:08.067660 140659750036672 spec.py:363] Evaluating on the test split. +I0831 23:20:09.106439 140659750036672 submission_runner.py:516] Time since start: 50530.09s, Step: 177030, {'train/accuracy': Array(0.94915897, dtype=float32), 'train/loss': Array(0.1775648, dtype=float32), 'validation/accuracy': Array(0.7549, dtype=float32), 'validation/loss': Array(1.0738461, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62630004, dtype=float32), 'test/loss': Array(1.8531942, dtype=float32), 'test/num_examples': 10000, 'score': 49952.29267835617, 'total_duration': 50530.0918970108, 'accumulated_submission_time': 49952.29267835617, 'accumulated_eval_time': 573.5259537696838, 'accumulated_logging_time': 2.5825936794281006} +I0831 23:20:09.540125 140462291654400 logging_writer.py:48] [177030] accumulated_eval_time=573.526, accumulated_logging_time=2.58259, accumulated_submission_time=49952.3, global_step=177030, preemption_count=0, score=49952.3, test/accuracy=0.626300036907196, test/loss=1.853194236755371, test/num_examples=10000, total_duration=50530.1, train/accuracy=0.9491589665412903, train/loss=0.177564799785614, validation/accuracy=0.7548999786376953, validation/loss=1.0738461017608643, validation/num_examples=50000 +I0831 23:20:35.203486 140462207792896 logging_writer.py:48] [177100] global_step=177100, grad_norm=5.074657440185547, loss=0.9274404048919678 +I0831 23:21:19.105588 140462291654400 logging_writer.py:48] [177200] global_step=177200, grad_norm=5.000508785247803, loss=0.9121726751327515 +I0831 23:22:02.503638 140462207792896 logging_writer.py:48] [177300] global_step=177300, grad_norm=4.9493207931518555, loss=0.9564206600189209 +I0831 23:22:46.483117 140462291654400 logging_writer.py:48] [177400] global_step=177400, grad_norm=5.007307529449463, loss=0.9345482587814331 +I0831 23:23:32.552372 140462207792896 logging_writer.py:48] [177500] global_step=177500, grad_norm=5.067407608032227, loss=0.9543004035949707 +I0831 23:24:18.334029 140462291654400 logging_writer.py:48] [177600] global_step=177600, grad_norm=4.986429214477539, loss=0.909703254699707 +I0831 23:25:05.651735 140462207792896 logging_writer.py:48] [177700] global_step=177700, grad_norm=5.141106128692627, loss=0.9536807537078857 +I0831 23:25:53.818519 140462291654400 logging_writer.py:48] [177800] global_step=177800, grad_norm=5.247263431549072, loss=0.932931125164032 +I0831 23:26:42.943144 140462207792896 logging_writer.py:48] [177900] global_step=177900, grad_norm=4.958828449249268, loss=0.9373376369476318 +I0831 23:27:32.294521 140462291654400 logging_writer.py:48] [178000] global_step=178000, grad_norm=5.335000038146973, loss=0.9289266467094421 +I0831 23:28:17.733265 140462207792896 logging_writer.py:48] [178100] global_step=178100, grad_norm=4.823667526245117, loss=0.914096474647522 +I0831 23:29:04.535987 140462291654400 logging_writer.py:48] [178200] global_step=178200, grad_norm=5.229920387268066, loss=0.9330036640167236 +I0831 23:29:50.083599 140462207792896 logging_writer.py:48] [178300] global_step=178300, grad_norm=4.936186790466309, loss=0.9288408756256104 +I0831 23:30:34.946969 140462291654400 logging_writer.py:48] [178400] global_step=178400, grad_norm=4.637578010559082, loss=0.8261798620223999 +I0831 23:31:20.790790 140462207792896 logging_writer.py:48] [178500] global_step=178500, grad_norm=5.266441345214844, loss=0.9718639850616455 +I0831 23:32:06.842786 140462291654400 logging_writer.py:48] [178600] global_step=178600, grad_norm=5.244848251342773, loss=0.9647239446640015 +I0831 23:32:53.262911 140462207792896 logging_writer.py:48] [178700] global_step=178700, grad_norm=4.909234523773193, loss=0.864410400390625 +I0831 23:33:38.938832 140462291654400 logging_writer.py:48] [178800] global_step=178800, grad_norm=4.787827491760254, loss=0.879813551902771 +I0831 23:34:26.841154 140462207792896 logging_writer.py:48] [178900] global_step=178900, grad_norm=4.949629783630371, loss=0.8708364367485046 +I0831 23:35:11.679471 140462291654400 logging_writer.py:48] [179000] global_step=179000, grad_norm=5.292900562286377, loss=0.9856553673744202 +I0831 23:35:57.737233 140462207792896 logging_writer.py:48] [179100] global_step=179100, grad_norm=4.838003158569336, loss=0.8596022129058838 +I0831 23:36:43.077682 140462291654400 logging_writer.py:48] [179200] global_step=179200, grad_norm=5.207801342010498, loss=0.9942983984947205 +I0831 23:37:27.784674 140462207792896 logging_writer.py:48] [179300] global_step=179300, grad_norm=4.871219158172607, loss=0.912481427192688 +I0831 23:38:15.152010 140462291654400 logging_writer.py:48] [179400] global_step=179400, grad_norm=4.915262699127197, loss=0.9612508416175842 +I0831 23:38:58.830628 140462207792896 logging_writer.py:48] [179500] global_step=179500, grad_norm=5.254115581512451, loss=1.0035550594329834 +I0831 23:39:42.004836 140462291654400 logging_writer.py:48] [179600] global_step=179600, grad_norm=5.0665106773376465, loss=0.9190337657928467 +I0831 23:40:25.266832 140462207792896 logging_writer.py:48] [179700] global_step=179700, grad_norm=5.0718674659729, loss=0.9340239763259888 +I0831 23:41:08.883384 140462291654400 logging_writer.py:48] [179800] global_step=179800, grad_norm=5.068124294281006, loss=0.9368919134140015 +I0831 23:41:53.312387 140462207792896 logging_writer.py:48] [179900] global_step=179900, grad_norm=4.964376926422119, loss=0.9363213777542114 +I0831 23:42:37.405147 140462291654400 logging_writer.py:48] [180000] global_step=180000, grad_norm=5.2110700607299805, loss=0.9083186984062195 +I0831 23:43:20.424063 140462207792896 logging_writer.py:48] [180100] global_step=180100, grad_norm=5.152426242828369, loss=0.9668314456939697 +I0831 23:44:03.821759 140462291654400 logging_writer.py:48] [180200] global_step=180200, grad_norm=4.947897434234619, loss=0.863544225692749 +I0831 23:44:46.258082 140462207792896 logging_writer.py:48] [180300] global_step=180300, grad_norm=5.269309043884277, loss=0.9609748721122742 +I0831 23:45:31.399881 140462291654400 logging_writer.py:48] [180400] global_step=180400, grad_norm=4.914673328399658, loss=0.8598454594612122 +I0831 23:46:14.139460 140462207792896 logging_writer.py:48] [180500] global_step=180500, grad_norm=5.121306419372559, loss=0.9007447957992554 +I0831 23:46:57.369444 140462291654400 logging_writer.py:48] [180600] global_step=180600, grad_norm=4.847136497497559, loss=0.9385638236999512 +I0831 23:47:42.880849 140462207792896 logging_writer.py:48] [180700] global_step=180700, grad_norm=5.285668849945068, loss=0.8860064744949341 +I0831 23:48:26.378141 140462291654400 logging_writer.py:48] [180800] global_step=180800, grad_norm=5.47243070602417, loss=1.0369856357574463 +I0831 23:49:09.662722 140462207792896 logging_writer.py:48] [180900] global_step=180900, grad_norm=4.862879276275635, loss=0.8778634071350098 +I0831 23:49:51.632477 140462291654400 logging_writer.py:48] [181000] global_step=181000, grad_norm=5.1417670249938965, loss=0.9430004358291626 +I0831 23:50:34.535393 140462207792896 logging_writer.py:48] [181100] global_step=181100, grad_norm=5.1278204917907715, loss=0.9381250143051147 +I0831 23:51:19.419407 140462291654400 logging_writer.py:48] [181200] global_step=181200, grad_norm=4.989552021026611, loss=0.9107352495193481 +I0831 23:52:02.146924 140462207792896 logging_writer.py:48] [181300] global_step=181300, grad_norm=4.891119956970215, loss=0.9983269572257996 +I0831 23:52:46.875212 140462291654400 logging_writer.py:48] [181400] global_step=181400, grad_norm=4.934650897979736, loss=0.9219993948936462 +I0831 23:53:25.018694 140659750036672 spec.py:333] Evaluating on the training split. +I0831 23:53:33.338346 140659750036672 spec.py:346] Evaluating on the validation split. +I0831 23:54:03.930612 140659750036672 spec.py:363] Evaluating on the test split. +I0831 23:54:05.021245 140659750036672 submission_runner.py:516] Time since start: 52565.99s, Step: 181488, {'train/accuracy': Array(0.9496771, dtype=float32), 'train/loss': Array(0.17702009, dtype=float32), 'validation/accuracy': Array(0.7542, dtype=float32), 'validation/loss': Array(1.0756431, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6247, dtype=float32), 'test/loss': Array(1.8563241, dtype=float32), 'test/num_examples': 10000, 'score': 51947.68952417374, 'total_duration': 52565.99465203285, 'accumulated_submission_time': 51947.68952417374, 'accumulated_eval_time': 613.3243944644928, 'accumulated_logging_time': 3.0541510581970215} +I0831 23:54:05.401913 140462207792896 logging_writer.py:48] [181488] accumulated_eval_time=613.324, accumulated_logging_time=3.05415, accumulated_submission_time=51947.7, global_step=181488, preemption_count=0, score=51947.7, test/accuracy=0.6247000098228455, test/loss=1.8563240766525269, test/num_examples=10000, total_duration=52566, train/accuracy=0.9496771097183228, train/loss=0.1770200878381729, validation/accuracy=0.7541999816894531, validation/loss=1.0756430625915527, validation/num_examples=50000 +I0831 23:54:09.053733 140462291654400 logging_writer.py:48] [181500] global_step=181500, grad_norm=4.777865886688232, loss=0.8965727686882019 +I0831 23:54:52.465248 140462207792896 logging_writer.py:48] [181600] global_step=181600, grad_norm=5.140329360961914, loss=0.9278042316436768 +I0831 23:55:36.186586 140462291654400 logging_writer.py:48] [181700] global_step=181700, grad_norm=4.999294757843018, loss=0.9762254357337952 +I0831 23:56:20.701881 140462207792896 logging_writer.py:48] [181800] global_step=181800, grad_norm=5.07246208190918, loss=0.9769768714904785 +I0831 23:57:04.607062 140462291654400 logging_writer.py:48] [181900] global_step=181900, grad_norm=5.148148059844971, loss=0.9064602851867676 +I0831 23:57:48.534315 140462207792896 logging_writer.py:48] [182000] global_step=182000, grad_norm=5.118179798126221, loss=0.934413492679596 +I0831 23:58:32.118633 140462291654400 logging_writer.py:48] [182100] global_step=182100, grad_norm=4.733865737915039, loss=0.8654574155807495 +I0831 23:59:14.740665 140462207792896 logging_writer.py:48] [182200] global_step=182200, grad_norm=5.078291416168213, loss=0.9450221657752991 +I0831 23:59:56.733263 140462291654400 logging_writer.py:48] [182300] global_step=182300, grad_norm=5.220005035400391, loss=1.0215524435043335 +I0901 00:00:38.850584 140462207792896 logging_writer.py:48] [182400] global_step=182400, grad_norm=5.337954998016357, loss=0.9765850305557251 +I0901 00:01:21.233425 140462291654400 logging_writer.py:48] [182500] global_step=182500, grad_norm=5.019347667694092, loss=0.9512249231338501 +I0901 00:02:06.312148 140462207792896 logging_writer.py:48] [182600] global_step=182600, grad_norm=5.237615585327148, loss=0.9594628214836121 +I0901 00:02:51.786991 140462291654400 logging_writer.py:48] [182700] global_step=182700, grad_norm=5.192789077758789, loss=0.9290973544120789 +I0901 00:03:34.570655 140462207792896 logging_writer.py:48] [182800] global_step=182800, grad_norm=5.269105911254883, loss=0.9676105380058289 +I0901 00:04:19.161897 140462291654400 logging_writer.py:48] [182900] global_step=182900, grad_norm=5.040559768676758, loss=0.9061084389686584 +I0901 00:05:02.478021 140462207792896 logging_writer.py:48] [183000] global_step=183000, grad_norm=5.288152694702148, loss=0.9055905938148499 +I0901 00:05:45.752267 140462291654400 logging_writer.py:48] [183100] global_step=183100, grad_norm=5.114673614501953, loss=0.9928376078605652 +I0901 00:06:29.390506 140462207792896 logging_writer.py:48] [183200] global_step=183200, grad_norm=5.065274238586426, loss=0.9406737089157104 +I0901 00:07:12.406910 140462291654400 logging_writer.py:48] [183300] global_step=183300, grad_norm=5.222174167633057, loss=0.9908910989761353 +I0901 00:07:56.581515 140462207792896 logging_writer.py:48] [183400] global_step=183400, grad_norm=5.186910152435303, loss=1.0142154693603516 +I0901 00:08:39.082929 140462291654400 logging_writer.py:48] [183500] global_step=183500, grad_norm=5.188793182373047, loss=0.9178715944290161 +I0901 00:09:22.360904 140462207792896 logging_writer.py:48] [183600] global_step=183600, grad_norm=5.012269020080566, loss=0.8652242422103882 +I0901 00:10:06.715574 140462291654400 logging_writer.py:48] [183700] global_step=183700, grad_norm=5.013866424560547, loss=0.9137861728668213 +I0901 00:10:49.695429 140462207792896 logging_writer.py:48] [183800] global_step=183800, grad_norm=4.726047992706299, loss=0.9475349187850952 +I0901 00:11:33.319357 140462291654400 logging_writer.py:48] [183900] global_step=183900, grad_norm=5.415104389190674, loss=1.0028718709945679 +I0901 00:12:17.171881 140462207792896 logging_writer.py:48] [184000] global_step=184000, grad_norm=5.02932071685791, loss=0.9207634925842285 +I0901 00:13:00.630497 140462291654400 logging_writer.py:48] [184100] global_step=184100, grad_norm=5.21934175491333, loss=0.9405956864356995 +I0901 00:13:42.884354 140462207792896 logging_writer.py:48] [184200] global_step=184200, grad_norm=5.0374932289123535, loss=0.9122084379196167 +I0901 00:14:24.553991 140462291654400 logging_writer.py:48] [184300] global_step=184300, grad_norm=5.329568386077881, loss=0.9933372139930725 +I0901 00:15:07.854610 140462207792896 logging_writer.py:48] [184400] global_step=184400, grad_norm=5.20517635345459, loss=0.9360822439193726 +I0901 00:15:50.612025 140462291654400 logging_writer.py:48] [184500] global_step=184500, grad_norm=5.2653584480285645, loss=0.943243145942688 +I0901 00:16:31.873199 140462207792896 logging_writer.py:48] [184600] global_step=184600, grad_norm=5.100636005401611, loss=0.9723906517028809 +I0901 00:17:12.911286 140462291654400 logging_writer.py:48] [184700] global_step=184700, grad_norm=5.037036418914795, loss=0.893587589263916 +I0901 00:17:53.410392 140462207792896 logging_writer.py:48] [184800] global_step=184800, grad_norm=4.871624946594238, loss=0.9598718285560608 +I0901 00:18:35.501424 140462291654400 logging_writer.py:48] [184900] global_step=184900, grad_norm=5.006869316101074, loss=0.9689427614212036 +I0901 00:19:17.804289 140462207792896 logging_writer.py:48] [185000] global_step=185000, grad_norm=5.496054649353027, loss=1.030204176902771 +I0901 00:20:01.375579 140462291654400 logging_writer.py:48] [185100] global_step=185100, grad_norm=5.21451997756958, loss=0.927370548248291 +I0901 00:20:45.802628 140462207792896 logging_writer.py:48] [185200] global_step=185200, grad_norm=5.04327917098999, loss=0.9798082709312439 +I0901 00:21:29.182816 140462291654400 logging_writer.py:48] [185300] global_step=185300, grad_norm=5.153430461883545, loss=0.8965625762939453 +I0901 00:22:13.234688 140462207792896 logging_writer.py:48] [185400] global_step=185400, grad_norm=4.855484485626221, loss=0.8252954483032227 +I0901 00:22:55.850781 140462291654400 logging_writer.py:48] [185500] global_step=185500, grad_norm=5.368815898895264, loss=0.9600778222084045 +I0901 00:23:38.282111 140462207792896 logging_writer.py:48] [185600] global_step=185600, grad_norm=5.042879581451416, loss=0.9022982716560364 +I0901 00:24:22.932370 140462291654400 logging_writer.py:48] [185700] global_step=185700, grad_norm=4.767755031585693, loss=0.8865697979927063 +I0901 00:25:06.352398 140462207792896 logging_writer.py:48] [185800] global_step=185800, grad_norm=5.294009208679199, loss=0.9720927476882935 +I0901 00:25:48.083274 140462291654400 logging_writer.py:48] [185900] global_step=185900, grad_norm=5.371466636657715, loss=0.9233159422874451 +I0901 00:26:30.799426 140462207792896 logging_writer.py:48] [186000] global_step=186000, grad_norm=5.047402381896973, loss=0.9401895999908447 +I0901 00:27:13.242962 140462291654400 logging_writer.py:48] [186100] global_step=186100, grad_norm=5.203314304351807, loss=0.9863709211349487 +I0901 00:27:21.226199 140659750036672 spec.py:333] Evaluating on the training split. +I0901 00:27:29.517195 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 00:28:00.629547 140659750036672 spec.py:363] Evaluating on the test split. +I0901 00:28:01.736380 140659750036672 submission_runner.py:516] Time since start: 54602.69s, Step: 186120, {'train/accuracy': Array(0.9524673, dtype=float32), 'train/loss': Array(0.16976719, dtype=float32), 'validation/accuracy': Array(0.75492, dtype=float32), 'validation/loss': Array(1.0755513, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62640005, dtype=float32), 'test/loss': Array(1.8612959, dtype=float32), 'test/num_examples': 10000, 'score': 53943.41852045059, 'total_duration': 54602.69437932968, 'accumulated_submission_time': 53943.41852045059, 'accumulated_eval_time': 653.6150543689728, 'accumulated_logging_time': 3.488328456878662} +I0901 00:28:02.203574 140462207792896 logging_writer.py:48] [186120] accumulated_eval_time=653.615, accumulated_logging_time=3.48833, accumulated_submission_time=53943.4, global_step=186120, preemption_count=0, score=53943.4, test/accuracy=0.6264000535011292, test/loss=1.8612959384918213, test/num_examples=10000, total_duration=54602.7, train/accuracy=0.9524673223495483, train/loss=0.16976718604564667, validation/accuracy=0.7549200057983398, validation/loss=1.0755512714385986, validation/num_examples=50000 +I0901 00:28:33.395898 140462291654400 logging_writer.py:48] [186200] global_step=186200, grad_norm=5.000091552734375, loss=0.9481979012489319 +I0901 00:29:19.280950 140462207792896 logging_writer.py:48] [186300] global_step=186300, grad_norm=5.056436061859131, loss=0.9424130916595459 +I0901 00:30:04.350382 140462291654400 logging_writer.py:48] [186400] global_step=186400, grad_norm=4.96738338470459, loss=1.038466215133667 +I0901 00:30:51.092118 140462207792896 logging_writer.py:48] [186500] global_step=186500, grad_norm=5.094128131866455, loss=0.905540406703949 +I0901 00:31:35.609941 140462291654400 logging_writer.py:48] [186600] global_step=186600, grad_norm=5.158945560455322, loss=0.969817578792572 +I0901 00:32:18.378438 140462207792896 logging_writer.py:48] [186700] global_step=186700, grad_norm=5.110857963562012, loss=0.9462757110595703 +I0901 00:33:01.582189 140462291654400 logging_writer.py:48] [186800] global_step=186800, grad_norm=5.516566276550293, loss=0.9779945611953735 +I0901 00:33:46.901953 140462207792896 logging_writer.py:48] [186900] global_step=186900, grad_norm=5.100681304931641, loss=0.9281010627746582 +I0901 00:34:33.072363 140462291654400 logging_writer.py:48] [187000] global_step=187000, grad_norm=4.957509994506836, loss=0.868954598903656 +I0901 00:35:17.205519 140462207792896 logging_writer.py:48] [187100] global_step=187100, grad_norm=4.993037223815918, loss=0.9799744486808777 +I0901 00:36:00.083300 140462291654400 logging_writer.py:48] [187200] global_step=187200, grad_norm=5.257252216339111, loss=0.9543837904930115 +I0901 00:36:42.837701 140462207792896 logging_writer.py:48] [187300] global_step=187300, grad_norm=4.908782005310059, loss=0.8894467353820801 +I0901 00:37:27.032659 140462291654400 logging_writer.py:48] [187400] global_step=187400, grad_norm=5.254242420196533, loss=0.9473912119865417 +I0901 00:38:12.648122 140462207792896 logging_writer.py:48] [187500] global_step=187500, grad_norm=5.29475212097168, loss=0.9450246095657349 +I0901 00:39:00.972084 140462291654400 logging_writer.py:48] [187600] global_step=187600, grad_norm=4.667696952819824, loss=0.8188464641571045 +I0901 00:39:49.750685 140462207792896 logging_writer.py:48] [187700] global_step=187700, grad_norm=5.277328968048096, loss=0.9283656477928162 +I0901 00:40:37.054332 140462291654400 logging_writer.py:48] [187800] global_step=187800, grad_norm=5.210704803466797, loss=0.9196891188621521 +I0901 00:41:23.422458 140462207792896 logging_writer.py:48] [187900] global_step=187900, grad_norm=5.1303205490112305, loss=0.9339657425880432 +I0901 00:42:10.498245 140462291654400 logging_writer.py:48] [188000] global_step=188000, grad_norm=5.2509446144104, loss=0.915166437625885 +I0901 00:42:56.557884 140462207792896 logging_writer.py:48] [188100] global_step=188100, grad_norm=4.931962490081787, loss=0.8615691661834717 +I0901 00:43:45.013665 140462291654400 logging_writer.py:48] [188200] global_step=188200, grad_norm=5.123899459838867, loss=0.9243561625480652 +I0901 00:44:32.725351 140462207792896 logging_writer.py:48] [188300] global_step=188300, grad_norm=4.96083927154541, loss=0.9249551296234131 +I0901 00:45:20.658367 140462291654400 logging_writer.py:48] [188400] global_step=188400, grad_norm=4.915194511413574, loss=0.9186629056930542 +I0901 00:46:07.936869 140462207792896 logging_writer.py:48] [188500] global_step=188500, grad_norm=5.591803073883057, loss=0.9413820505142212 +I0901 00:46:56.893549 140462291654400 logging_writer.py:48] [188600] global_step=188600, grad_norm=4.745026111602783, loss=0.9119367599487305 +I0901 00:47:47.687024 140462207792896 logging_writer.py:48] [188700] global_step=188700, grad_norm=5.420844078063965, loss=1.0085724592208862 +I0901 00:48:36.033409 140462291654400 logging_writer.py:48] [188800] global_step=188800, grad_norm=4.907855033874512, loss=0.9279282093048096 +I0901 00:49:25.819334 140462207792896 logging_writer.py:48] [188900] global_step=188900, grad_norm=4.800914287567139, loss=0.8871800899505615 +I0901 00:50:17.821497 140462291654400 logging_writer.py:48] [189000] global_step=189000, grad_norm=5.374838352203369, loss=1.019045352935791 +I0901 00:51:09.741248 140462207792896 logging_writer.py:48] [189100] global_step=189100, grad_norm=5.106861114501953, loss=0.9028024077415466 +I0901 00:51:59.969689 140462291654400 logging_writer.py:48] [189200] global_step=189200, grad_norm=5.169831275939941, loss=0.8804554343223572 +I0901 00:52:49.845973 140462207792896 logging_writer.py:48] [189300] global_step=189300, grad_norm=5.192558288574219, loss=0.9043139815330505 +I0901 00:53:41.607636 140462291654400 logging_writer.py:48] [189400] global_step=189400, grad_norm=4.936967849731445, loss=0.9367093443870544 +I0901 00:54:32.814768 140462207792896 logging_writer.py:48] [189500] global_step=189500, grad_norm=4.999384880065918, loss=0.9523379802703857 +I0901 00:55:25.170565 140462291654400 logging_writer.py:48] [189600] global_step=189600, grad_norm=5.008586883544922, loss=0.9157549738883972 +I0901 00:56:16.354363 140462207792896 logging_writer.py:48] [189700] global_step=189700, grad_norm=5.018721580505371, loss=0.8735097646713257 +I0901 00:57:08.041367 140462291654400 logging_writer.py:48] [189800] global_step=189800, grad_norm=5.122860431671143, loss=1.0168596506118774 +I0901 00:57:59.246551 140462207792896 logging_writer.py:48] [189900] global_step=189900, grad_norm=4.8549885749816895, loss=0.8926045894622803 +I0901 00:58:51.078748 140462291654400 logging_writer.py:48] [190000] global_step=190000, grad_norm=5.514739990234375, loss=0.9789514541625977 +I0901 00:59:42.609136 140462207792896 logging_writer.py:48] [190100] global_step=190100, grad_norm=4.725882530212402, loss=0.8641881346702576 +I0901 01:00:37.537830 140462291654400 logging_writer.py:48] [190200] global_step=190200, grad_norm=5.180016040802002, loss=0.8963220119476318 +I0901 01:01:17.936188 140659750036672 spec.py:333] Evaluating on the training split. +I0901 01:01:26.115476 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 01:01:53.676390 140659750036672 spec.py:363] Evaluating on the test split. +I0901 01:01:54.748594 140659750036672 submission_runner.py:516] Time since start: 56635.74s, Step: 190276, {'train/accuracy': Array(0.9514509, dtype=float32), 'train/loss': Array(0.17048396, dtype=float32), 'validation/accuracy': Array(0.75464, dtype=float32), 'validation/loss': Array(1.0741526, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6272, dtype=float32), 'test/loss': Array(1.8644032, dtype=float32), 'test/num_examples': 10000, 'score': 55939.05248832703, 'total_duration': 56635.74251246452, 'accumulated_submission_time': 55939.05248832703, 'accumulated_eval_time': 690.2438578605652, 'accumulated_logging_time': 4.0174477100372314} +I0901 01:01:55.211954 140462207792896 logging_writer.py:48] [190276] accumulated_eval_time=690.244, accumulated_logging_time=4.01745, accumulated_submission_time=55939.1, global_step=190276, preemption_count=0, score=55939.1, test/accuracy=0.6272000074386597, test/loss=1.864403247833252, test/num_examples=10000, total_duration=56635.7, train/accuracy=0.9514508843421936, train/loss=0.17048396170139313, validation/accuracy=0.7546399831771851, validation/loss=1.0741525888442993, validation/num_examples=50000 +I0901 01:02:02.136723 140462291654400 logging_writer.py:48] [190300] global_step=190300, grad_norm=4.991329669952393, loss=0.9188385605812073 +I0901 01:02:56.961960 140462207792896 logging_writer.py:48] [190400] global_step=190400, grad_norm=4.940360069274902, loss=0.9071121215820312 +I0901 01:03:50.112859 140462291654400 logging_writer.py:48] [190500] global_step=190500, grad_norm=5.436657905578613, loss=0.9951790571212769 +I0901 01:04:44.020709 140462207792896 logging_writer.py:48] [190600] global_step=190600, grad_norm=5.255477428436279, loss=0.9578527212142944 +I0901 01:05:37.877246 140462291654400 logging_writer.py:48] [190700] global_step=190700, grad_norm=4.775386810302734, loss=0.9204561114311218 +I0901 01:06:31.030667 140462207792896 logging_writer.py:48] [190800] global_step=190800, grad_norm=4.937310218811035, loss=0.9485346078872681 +I0901 01:07:22.066493 140462291654400 logging_writer.py:48] [190900] global_step=190900, grad_norm=5.144085884094238, loss=0.929114580154419 +I0901 01:08:14.996941 140462207792896 logging_writer.py:48] [191000] global_step=191000, grad_norm=4.886878490447998, loss=0.875161349773407 +I0901 01:09:09.846013 140462291654400 logging_writer.py:48] [191100] global_step=191100, grad_norm=5.168906211853027, loss=0.9513559341430664 +I0901 01:10:02.066124 140462207792896 logging_writer.py:48] [191200] global_step=191200, grad_norm=5.106042385101318, loss=0.8795433044433594 +I0901 01:10:54.012265 140462291654400 logging_writer.py:48] [191300] global_step=191300, grad_norm=5.0199079513549805, loss=0.9926417469978333 +I0901 01:11:46.462804 140462207792896 logging_writer.py:48] [191400] global_step=191400, grad_norm=5.14337682723999, loss=0.9436733722686768 +I0901 01:12:39.908333 140462291654400 logging_writer.py:48] [191500] global_step=191500, grad_norm=5.4088850021362305, loss=1.0049258470535278 +I0901 01:13:33.430194 140462207792896 logging_writer.py:48] [191600] global_step=191600, grad_norm=5.163417339324951, loss=0.9149528741836548 +I0901 01:14:26.484310 140462291654400 logging_writer.py:48] [191700] global_step=191700, grad_norm=5.281464099884033, loss=1.0535931587219238 +I0901 01:15:17.340895 140462207792896 logging_writer.py:48] [191800] global_step=191800, grad_norm=5.133783340454102, loss=0.8980395793914795 +I0901 01:16:10.319633 140462291654400 logging_writer.py:48] [191900] global_step=191900, grad_norm=5.19895601272583, loss=0.8723468780517578 +I0901 01:17:02.984281 140462207792896 logging_writer.py:48] [192000] global_step=192000, grad_norm=5.112756252288818, loss=0.8882954716682434 +I0901 01:17:53.465596 140462291654400 logging_writer.py:48] [192100] global_step=192100, grad_norm=4.917614936828613, loss=0.8288053870201111 +I0901 01:18:43.993535 140462207792896 logging_writer.py:48] [192200] global_step=192200, grad_norm=5.146633625030518, loss=0.9293472170829773 +I0901 01:19:35.131498 140462291654400 logging_writer.py:48] [192300] global_step=192300, grad_norm=5.035139560699463, loss=0.960735559463501 +I0901 01:20:26.829900 140462207792896 logging_writer.py:48] [192400] global_step=192400, grad_norm=5.466362953186035, loss=0.9237858057022095 +I0901 01:21:19.242294 140462291654400 logging_writer.py:48] [192500] global_step=192500, grad_norm=5.191043853759766, loss=0.8744706511497498 +I0901 01:22:08.774386 140462207792896 logging_writer.py:48] [192600] global_step=192600, grad_norm=5.13855504989624, loss=0.8766710162162781 +I0901 01:23:01.208231 140462291654400 logging_writer.py:48] [192700] global_step=192700, grad_norm=5.148075103759766, loss=0.895211935043335 +I0901 01:23:51.966255 140462207792896 logging_writer.py:48] [192800] global_step=192800, grad_norm=4.98637056350708, loss=0.8951486349105835 +I0901 01:24:43.713422 140462291654400 logging_writer.py:48] [192900] global_step=192900, grad_norm=5.2157368659973145, loss=0.9547462463378906 +I0901 01:25:35.005769 140462207792896 logging_writer.py:48] [193000] global_step=193000, grad_norm=5.3539628982543945, loss=0.9827317595481873 +I0901 01:26:24.715054 140462291654400 logging_writer.py:48] [193100] global_step=193100, grad_norm=5.300951957702637, loss=0.9966809749603271 +I0901 01:27:16.217173 140462207792896 logging_writer.py:48] [193200] global_step=193200, grad_norm=5.291034698486328, loss=0.9528250098228455 +I0901 01:28:07.178827 140462291654400 logging_writer.py:48] [193300] global_step=193300, grad_norm=5.158587455749512, loss=0.9422928690910339 +I0901 01:28:57.695696 140462207792896 logging_writer.py:48] [193400] global_step=193400, grad_norm=5.1155476570129395, loss=0.9477238655090332 +I0901 01:29:49.281405 140462291654400 logging_writer.py:48] [193500] global_step=193500, grad_norm=5.070436954498291, loss=0.8637977838516235 +I0901 01:30:40.766223 140462207792896 logging_writer.py:48] [193600] global_step=193600, grad_norm=5.029070854187012, loss=0.9673832654953003 +I0901 01:31:33.902838 140462291654400 logging_writer.py:48] [193700] global_step=193700, grad_norm=5.620896816253662, loss=0.9647364616394043 +I0901 01:32:25.402030 140462207792896 logging_writer.py:48] [193800] global_step=193800, grad_norm=4.933979034423828, loss=0.8675025701522827 +I0901 01:33:16.513223 140462291654400 logging_writer.py:48] [193900] global_step=193900, grad_norm=5.206149578094482, loss=0.9716039299964905 +I0901 01:34:10.952080 140462207792896 logging_writer.py:48] [194000] global_step=194000, grad_norm=5.099429130554199, loss=0.9420609474182129 +I0901 01:35:02.893828 140462291654400 logging_writer.py:48] [194100] global_step=194100, grad_norm=4.9846906661987305, loss=0.8407999277114868 +I0901 01:35:10.698411 140659750036672 spec.py:333] Evaluating on the training split. +I0901 01:35:19.007531 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 01:35:52.455141 140659750036672 spec.py:363] Evaluating on the test split. +I0901 01:35:53.558786 140659750036672 submission_runner.py:516] Time since start: 58674.52s, Step: 194116, {'train/accuracy': Array(0.95322466, dtype=float32), 'train/loss': Array(0.16543461, dtype=float32), 'validation/accuracy': Array(0.75505996, dtype=float32), 'validation/loss': Array(1.0765518, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62700003, dtype=float32), 'test/loss': Array(1.8665876, dtype=float32), 'test/num_examples': 10000, 'score': 57934.46646571159, 'total_duration': 58674.522223472595, 'accumulated_submission_time': 57934.46646571159, 'accumulated_eval_time': 732.8901488780975, 'accumulated_logging_time': 4.518044471740723} +I0901 01:35:54.058293 140462207792896 logging_writer.py:48] [194116] accumulated_eval_time=732.89, accumulated_logging_time=4.51804, accumulated_submission_time=57934.5, global_step=194116, preemption_count=0, score=57934.5, test/accuracy=0.6270000338554382, test/loss=1.8665876388549805, test/num_examples=10000, total_duration=58674.5, train/accuracy=0.9532246589660645, train/loss=0.16543461382389069, validation/accuracy=0.7550599575042725, validation/loss=1.0765517950057983, validation/num_examples=50000 +I0901 01:36:33.205977 140462291654400 logging_writer.py:48] [194200] global_step=194200, grad_norm=5.122315406799316, loss=0.9044402837753296 +I0901 01:37:26.879209 140462207792896 logging_writer.py:48] [194300] global_step=194300, grad_norm=4.756315231323242, loss=0.9485416412353516 +I0901 01:38:19.267334 140462291654400 logging_writer.py:48] [194400] global_step=194400, grad_norm=5.111224174499512, loss=0.9009151458740234 +I0901 01:39:10.084113 140462207792896 logging_writer.py:48] [194500] global_step=194500, grad_norm=5.268344402313232, loss=0.9104762673377991 +I0901 01:40:00.838107 140462291654400 logging_writer.py:48] [194600] global_step=194600, grad_norm=5.196266174316406, loss=0.955623984336853 +I0901 01:40:53.554857 140462207792896 logging_writer.py:48] [194700] global_step=194700, grad_norm=5.565105438232422, loss=0.951729416847229 +I0901 01:41:43.705689 140462291654400 logging_writer.py:48] [194800] global_step=194800, grad_norm=5.290342807769775, loss=1.037819743156433 +I0901 01:42:37.275873 140462207792896 logging_writer.py:48] [194900] global_step=194900, grad_norm=5.318161964416504, loss=0.9598708152770996 +I0901 01:43:29.775447 140462291654400 logging_writer.py:48] [195000] global_step=195000, grad_norm=5.120434761047363, loss=0.9703190326690674 +I0901 01:44:23.124944 140462207792896 logging_writer.py:48] [195100] global_step=195100, grad_norm=4.755565643310547, loss=0.8529189825057983 +I0901 01:45:16.290702 140462291654400 logging_writer.py:48] [195200] global_step=195200, grad_norm=5.307429313659668, loss=1.0541903972625732 +I0901 01:46:09.056799 140462207792896 logging_writer.py:48] [195300] global_step=195300, grad_norm=5.599088191986084, loss=1.01732337474823 +I0901 01:47:03.593837 140462291654400 logging_writer.py:48] [195400] global_step=195400, grad_norm=4.770899295806885, loss=0.8862137198448181 +I0901 01:47:57.154137 140462207792896 logging_writer.py:48] [195500] global_step=195500, grad_norm=5.278220176696777, loss=0.9788541793823242 +I0901 01:48:49.677451 140462291654400 logging_writer.py:48] [195600] global_step=195600, grad_norm=5.099401950836182, loss=0.8865183591842651 +I0901 01:49:42.260670 140462207792896 logging_writer.py:48] [195700] global_step=195700, grad_norm=4.990214824676514, loss=0.8572313189506531 +I0901 01:50:33.689303 140462291654400 logging_writer.py:48] [195800] global_step=195800, grad_norm=5.367990016937256, loss=0.918639600276947 +I0901 01:51:25.804903 140462207792896 logging_writer.py:48] [195900] global_step=195900, grad_norm=4.694236755371094, loss=0.8653503656387329 +I0901 01:52:17.249797 140462291654400 logging_writer.py:48] [196000] global_step=196000, grad_norm=5.120105743408203, loss=0.9409099221229553 +I0901 01:53:06.800264 140462207792896 logging_writer.py:48] [196100] global_step=196100, grad_norm=4.948042869567871, loss=0.8631553649902344 +I0901 01:53:58.029108 140462291654400 logging_writer.py:48] [196200] global_step=196200, grad_norm=4.8718581199646, loss=0.8913165330886841 +I0901 01:54:50.383693 140462207792896 logging_writer.py:48] [196300] global_step=196300, grad_norm=5.263677597045898, loss=0.9798468351364136 +I0901 01:55:42.350553 140462291654400 logging_writer.py:48] [196400] global_step=196400, grad_norm=5.324559211730957, loss=0.9554418325424194 +I0901 01:56:35.850820 140462207792896 logging_writer.py:48] [196500] global_step=196500, grad_norm=5.284482002258301, loss=1.0349960327148438 +I0901 01:57:27.371001 140462291654400 logging_writer.py:48] [196600] global_step=196600, grad_norm=5.013746738433838, loss=0.8992078304290771 +I0901 01:58:19.733058 140462207792896 logging_writer.py:48] [196700] global_step=196700, grad_norm=5.341068267822266, loss=0.9587164521217346 +I0901 01:59:11.795879 140462291654400 logging_writer.py:48] [196800] global_step=196800, grad_norm=4.944610118865967, loss=0.827064037322998 +I0901 02:00:03.797182 140462207792896 logging_writer.py:48] [196900] global_step=196900, grad_norm=5.506231784820557, loss=0.968206524848938 +I0901 02:00:54.620283 140462291654400 logging_writer.py:48] [197000] global_step=197000, grad_norm=4.726687908172607, loss=0.9002347588539124 +I0901 02:01:47.174383 140462207792896 logging_writer.py:48] [197100] global_step=197100, grad_norm=5.269493103027344, loss=0.9231704473495483 +I0901 02:02:39.396696 140462291654400 logging_writer.py:48] [197200] global_step=197200, grad_norm=5.409403324127197, loss=0.9619237184524536 +I0901 02:03:29.906335 140462207792896 logging_writer.py:48] [197300] global_step=197300, grad_norm=5.2336249351501465, loss=0.9495603442192078 +I0901 02:04:21.181892 140462291654400 logging_writer.py:48] [197400] global_step=197400, grad_norm=4.715895175933838, loss=0.8808165788650513 +I0901 02:05:12.875298 140462207792896 logging_writer.py:48] [197500] global_step=197500, grad_norm=5.377445697784424, loss=0.9640500545501709 +I0901 02:06:05.895000 140462291654400 logging_writer.py:48] [197600] global_step=197600, grad_norm=5.141263008117676, loss=0.8426449298858643 +I0901 02:06:59.704351 140462207792896 logging_writer.py:48] [197700] global_step=197700, grad_norm=4.985975742340088, loss=0.8991923332214355 +I0901 02:07:50.775336 140462291654400 logging_writer.py:48] [197800] global_step=197800, grad_norm=5.1053996086120605, loss=0.9751726984977722 +I0901 02:08:42.613421 140462207792896 logging_writer.py:48] [197900] global_step=197900, grad_norm=5.241578578948975, loss=1.0032023191452026 +I0901 02:09:09.455591 140659750036672 spec.py:333] Evaluating on the training split. +I0901 02:09:17.732891 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 02:09:41.573733 140659750036672 spec.py:363] Evaluating on the test split. +I0901 02:09:42.669802 140659750036672 submission_runner.py:516] Time since start: 60703.65s, Step: 197954, {'train/accuracy': Array(0.95310503, dtype=float32), 'train/loss': Array(0.1635923, dtype=float32), 'validation/accuracy': Array(0.75564, dtype=float32), 'validation/loss': Array(1.0758622, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62600005, dtype=float32), 'test/loss': Array(1.8628919, dtype=float32), 'test/num_examples': 10000, 'score': 59929.77295255661, 'total_duration': 60703.64590597153, 'accumulated_submission_time': 59929.77295255661, 'accumulated_eval_time': 765.902948141098, 'accumulated_logging_time': 5.073758840560913} +I0901 02:09:43.096057 140462291654400 logging_writer.py:48] [197954] accumulated_eval_time=765.903, accumulated_logging_time=5.07376, accumulated_submission_time=59929.8, global_step=197954, preemption_count=0, score=59929.8, test/accuracy=0.6260000467300415, test/loss=1.8628919124603271, test/num_examples=10000, total_duration=60703.6, train/accuracy=0.9531050324440002, train/loss=0.16359229385852814, validation/accuracy=0.7556399703025818, validation/loss=1.075862169265747, validation/num_examples=50000 +I0901 02:10:01.009031 140462207792896 logging_writer.py:48] [198000] global_step=198000, grad_norm=5.733408451080322, loss=1.0850447416305542 +I0901 02:10:51.515936 140462291654400 logging_writer.py:48] [198100] global_step=198100, grad_norm=5.053123950958252, loss=0.9277308583259583 +I0901 02:11:40.552887 140462207792896 logging_writer.py:48] [198200] global_step=198200, grad_norm=5.447939395904541, loss=0.9580576419830322 +I0901 02:12:28.900656 140462291654400 logging_writer.py:48] [198300] global_step=198300, grad_norm=5.692147254943848, loss=1.0252968072891235 +I0901 02:13:18.830354 140462207792896 logging_writer.py:48] [198400] global_step=198400, grad_norm=5.408724308013916, loss=0.9750857353210449 +I0901 02:14:08.247431 140462291654400 logging_writer.py:48] [198500] global_step=198500, grad_norm=5.171483516693115, loss=0.9967667460441589 +I0901 02:14:57.253546 140462207792896 logging_writer.py:48] [198600] global_step=198600, grad_norm=4.909497261047363, loss=0.8407248854637146 +I0901 02:15:45.245699 140462291654400 logging_writer.py:48] [198700] global_step=198700, grad_norm=4.81143856048584, loss=0.8572880029678345 +I0901 02:16:33.104248 140462207792896 logging_writer.py:48] [198800] global_step=198800, grad_norm=4.975239276885986, loss=0.9000000953674316 +I0901 02:17:21.447813 140462291654400 logging_writer.py:48] [198900] global_step=198900, grad_norm=5.156773090362549, loss=0.9078607559204102 +I0901 02:18:12.640591 140462207792896 logging_writer.py:48] [199000] global_step=199000, grad_norm=5.222535610198975, loss=0.9808086156845093 +I0901 02:19:02.269568 140462291654400 logging_writer.py:48] [199100] global_step=199100, grad_norm=5.452895641326904, loss=0.9831409454345703 +I0901 02:19:52.385922 140462207792896 logging_writer.py:48] [199200] global_step=199200, grad_norm=5.115038871765137, loss=0.9376564025878906 +I0901 02:20:40.717133 140462291654400 logging_writer.py:48] [199300] global_step=199300, grad_norm=5.195210933685303, loss=0.9405603408813477 +I0901 02:21:29.074632 140462207792896 logging_writer.py:48] [199400] global_step=199400, grad_norm=4.975637912750244, loss=0.8807753920555115 +I0901 02:22:16.421287 140462291654400 logging_writer.py:48] [199500] global_step=199500, grad_norm=5.0632548332214355, loss=0.8300690650939941 +I0901 02:23:02.945284 140462207792896 logging_writer.py:48] [199600] global_step=199600, grad_norm=5.51336669921875, loss=0.9384492635726929 +I0901 02:23:49.318670 140462291654400 logging_writer.py:48] [199700] global_step=199700, grad_norm=5.092995643615723, loss=0.8486066460609436 +I0901 02:24:37.151591 140462207792896 logging_writer.py:48] [199800] global_step=199800, grad_norm=5.3724541664123535, loss=0.9736224412918091 +I0901 02:25:23.104759 140462291654400 logging_writer.py:48] [199900] global_step=199900, grad_norm=5.003391265869141, loss=0.893470287322998 +I0901 02:26:11.410973 140462207792896 logging_writer.py:48] [200000] global_step=200000, grad_norm=5.176023006439209, loss=0.9010908603668213 +I0901 02:26:59.173644 140462291654400 logging_writer.py:48] [200100] global_step=200100, grad_norm=5.019099712371826, loss=0.8914627432823181 +I0901 02:27:48.658614 140462207792896 logging_writer.py:48] [200200] global_step=200200, grad_norm=5.280834674835205, loss=0.9089116454124451 +I0901 02:28:39.738465 140462291654400 logging_writer.py:48] [200300] global_step=200300, grad_norm=4.843146324157715, loss=0.8675345182418823 +I0901 02:29:29.638105 140462207792896 logging_writer.py:48] [200400] global_step=200400, grad_norm=5.21372127532959, loss=0.8831021785736084 +I0901 02:30:17.217230 140462291654400 logging_writer.py:48] [200500] global_step=200500, grad_norm=4.9101152420043945, loss=0.8199357986450195 +I0901 02:31:04.883280 140462207792896 logging_writer.py:48] [200600] global_step=200600, grad_norm=5.141711235046387, loss=0.948287844657898 +I0901 02:31:53.506973 140462291654400 logging_writer.py:48] [200700] global_step=200700, grad_norm=5.031894683837891, loss=0.9146095514297485 +I0901 02:32:43.306477 140462207792896 logging_writer.py:48] [200800] global_step=200800, grad_norm=5.032165050506592, loss=0.8457980155944824 +I0901 02:33:29.158688 140462291654400 logging_writer.py:48] [200900] global_step=200900, grad_norm=5.334079742431641, loss=1.0012643337249756 +I0901 02:34:14.740139 140462207792896 logging_writer.py:48] [201000] global_step=201000, grad_norm=5.243875503540039, loss=0.9059763550758362 +I0901 02:34:59.046271 140462291654400 logging_writer.py:48] [201100] global_step=201100, grad_norm=4.84406042098999, loss=0.8863714933395386 +I0901 02:35:44.390774 140462207792896 logging_writer.py:48] [201200] global_step=201200, grad_norm=5.301848888397217, loss=0.9486503601074219 +I0901 02:36:28.982264 140462291654400 logging_writer.py:48] [201300] global_step=201300, grad_norm=5.532759189605713, loss=0.9953212738037109 +I0901 02:37:13.890596 140462207792896 logging_writer.py:48] [201400] global_step=201400, grad_norm=5.128269195556641, loss=0.943759560585022 +I0901 02:38:01.795731 140462291654400 logging_writer.py:48] [201500] global_step=201500, grad_norm=5.326333522796631, loss=0.8963664174079895 +I0901 02:38:53.418215 140462207792896 logging_writer.py:48] [201600] global_step=201600, grad_norm=5.203134536743164, loss=0.936785101890564 +I0901 02:39:43.549079 140462291654400 logging_writer.py:48] [201700] global_step=201700, grad_norm=5.103291034698486, loss=0.9194702506065369 +I0901 02:40:34.674233 140462207792896 logging_writer.py:48] [201800] global_step=201800, grad_norm=5.262247562408447, loss=0.9674670696258545 +I0901 02:41:26.527747 140462291654400 logging_writer.py:48] [201900] global_step=201900, grad_norm=5.335480690002441, loss=0.9657902717590332 +I0901 02:42:20.025152 140462207792896 logging_writer.py:48] [202000] global_step=202000, grad_norm=5.107743263244629, loss=0.9109036922454834 +I0901 02:42:58.480935 140659750036672 spec.py:333] Evaluating on the training split. +I0901 02:43:06.750856 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 02:43:26.782853 140659750036672 spec.py:363] Evaluating on the test split. +I0901 02:43:27.883687 140659750036672 submission_runner.py:516] Time since start: 62728.85s, Step: 202078, {'train/accuracy': Array(0.95360327, dtype=float32), 'train/loss': Array(0.1622379, dtype=float32), 'validation/accuracy': Array(0.75534, dtype=float32), 'validation/loss': Array(1.0767275, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62560004, dtype=float32), 'test/loss': Array(1.8653016, dtype=float32), 'test/num_examples': 10000, 'score': 61925.07143211365, 'total_duration': 62728.847462654114, 'accumulated_submission_time': 61925.07143211365, 'accumulated_eval_time': 795.0919539928436, 'accumulated_logging_time': 5.548985481262207} +I0901 02:43:28.354902 140462291654400 logging_writer.py:48] [202078] accumulated_eval_time=795.092, accumulated_logging_time=5.54899, accumulated_submission_time=61925.1, global_step=202078, preemption_count=0, score=61925.1, test/accuracy=0.6256000399589539, test/loss=1.8653016090393066, test/num_examples=10000, total_duration=62728.8, train/accuracy=0.9536032676696777, train/loss=0.16223789751529694, validation/accuracy=0.7553399801254272, validation/loss=1.0767275094985962, validation/num_examples=50000 +I0901 02:43:34.640001 140462207792896 logging_writer.py:48] [202100] global_step=202100, grad_norm=5.063547134399414, loss=0.894709050655365 +I0901 02:44:27.469215 140462291654400 logging_writer.py:48] [202200] global_step=202200, grad_norm=5.362625598907471, loss=0.939994752407074 +I0901 02:45:20.789301 140462207792896 logging_writer.py:48] [202300] global_step=202300, grad_norm=5.088252544403076, loss=0.9433736205101013 +I0901 02:46:14.653619 140462291654400 logging_writer.py:48] [202400] global_step=202400, grad_norm=5.418000221252441, loss=1.0874011516571045 +I0901 02:47:08.224926 140462207792896 logging_writer.py:48] [202500] global_step=202500, grad_norm=4.893282413482666, loss=0.8966118097305298 +I0901 02:48:01.091506 140462291654400 logging_writer.py:48] [202600] global_step=202600, grad_norm=4.878830909729004, loss=0.8828936815261841 +I0901 02:48:55.202977 140462207792896 logging_writer.py:48] [202700] global_step=202700, grad_norm=5.143739223480225, loss=0.924132764339447 +I0901 02:49:45.969709 140462291654400 logging_writer.py:48] [202800] global_step=202800, grad_norm=5.2499189376831055, loss=0.9205634593963623 +I0901 02:50:39.425322 140462207792896 logging_writer.py:48] [202900] global_step=202900, grad_norm=5.3368730545043945, loss=0.9606807231903076 +I0901 02:51:37.862698 140462291654400 logging_writer.py:48] [203000] global_step=203000, grad_norm=5.4535627365112305, loss=1.0593383312225342 +I0901 02:52:33.647561 140462207792896 logging_writer.py:48] [203100] global_step=203100, grad_norm=5.199843883514404, loss=0.9403886795043945 +I0901 02:53:33.606019 140462291654400 logging_writer.py:48] [203200] global_step=203200, grad_norm=5.033539295196533, loss=0.9053245782852173 +I0901 02:54:30.380057 140462207792896 logging_writer.py:48] [203300] global_step=203300, grad_norm=5.0271315574646, loss=0.9347968101501465 +I0901 02:55:25.707437 140462291654400 logging_writer.py:48] [203400] global_step=203400, grad_norm=5.194548606872559, loss=0.9000784754753113 +I0901 02:56:17.559358 140462207792896 logging_writer.py:48] [203500] global_step=203500, grad_norm=4.675726890563965, loss=0.8283615112304688 +I0901 02:57:10.553197 140462291654400 logging_writer.py:48] [203600] global_step=203600, grad_norm=4.967251300811768, loss=0.8491171598434448 +I0901 02:58:02.100975 140462207792896 logging_writer.py:48] [203700] global_step=203700, grad_norm=4.961255073547363, loss=0.8971527814865112 +I0901 02:58:53.765345 140462291654400 logging_writer.py:48] [203800] global_step=203800, grad_norm=4.976996421813965, loss=0.8899999856948853 +I0901 02:59:45.210809 140462207792896 logging_writer.py:48] [203900] global_step=203900, grad_norm=4.8664164543151855, loss=0.8970571756362915 +I0901 03:00:39.685892 140462291654400 logging_writer.py:48] [204000] global_step=204000, grad_norm=4.933059215545654, loss=0.9049407839775085 +I0901 03:01:31.352520 140462207792896 logging_writer.py:48] [204100] global_step=204100, grad_norm=5.3730058670043945, loss=0.9543185830116272 +I0901 03:02:22.883966 140462291654400 logging_writer.py:48] [204200] global_step=204200, grad_norm=5.163533687591553, loss=0.8941898345947266 +I0901 03:03:13.376609 140462207792896 logging_writer.py:48] [204300] global_step=204300, grad_norm=5.605095863342285, loss=0.9884198307991028 +I0901 03:04:07.391359 140462291654400 logging_writer.py:48] [204400] global_step=204400, grad_norm=5.150265693664551, loss=0.9482917785644531 +I0901 03:04:59.515872 140462207792896 logging_writer.py:48] [204500] global_step=204500, grad_norm=5.483210563659668, loss=0.9445587992668152 +I0901 03:05:51.580405 140462291654400 logging_writer.py:48] [204600] global_step=204600, grad_norm=5.356522560119629, loss=1.0227482318878174 +I0901 03:06:41.990092 140462207792896 logging_writer.py:48] [204700] global_step=204700, grad_norm=5.161183834075928, loss=0.8822721838951111 +I0901 03:07:33.812220 140462291654400 logging_writer.py:48] [204800] global_step=204800, grad_norm=5.191273212432861, loss=0.9074015617370605 +I0901 03:08:25.558800 140462207792896 logging_writer.py:48] [204900] global_step=204900, grad_norm=5.120369911193848, loss=0.8773990273475647 +I0901 03:09:18.411076 140462291654400 logging_writer.py:48] [205000] global_step=205000, grad_norm=5.252118110656738, loss=0.9657264947891235 +I0901 03:10:09.056571 140462207792896 logging_writer.py:48] [205100] global_step=205100, grad_norm=5.343626976013184, loss=0.931294858455658 +I0901 03:11:00.668637 140462291654400 logging_writer.py:48] [205200] global_step=205200, grad_norm=5.122676372528076, loss=1.0047038793563843 +I0901 03:11:51.988211 140462207792896 logging_writer.py:48] [205300] global_step=205300, grad_norm=5.112037658691406, loss=0.8695449829101562 +I0901 03:12:44.311822 140462291654400 logging_writer.py:48] [205400] global_step=205400, grad_norm=5.220719814300537, loss=0.9282364845275879 +I0901 03:13:35.358986 140462207792896 logging_writer.py:48] [205500] global_step=205500, grad_norm=5.49167013168335, loss=0.923180341720581 +I0901 03:14:25.800640 140462291654400 logging_writer.py:48] [205600] global_step=205600, grad_norm=5.224966049194336, loss=0.9027285575866699 +I0901 03:15:16.752014 140462207792896 logging_writer.py:48] [205700] global_step=205700, grad_norm=4.929141044616699, loss=0.878108024597168 +I0901 03:16:08.066429 140462291654400 logging_writer.py:48] [205800] global_step=205800, grad_norm=5.327309608459473, loss=0.9108087420463562 +I0901 03:16:44.005623 140659750036672 spec.py:333] Evaluating on the training split. +I0901 03:16:52.598555 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 03:17:22.865527 140659750036672 spec.py:363] Evaluating on the test split. +I0901 03:17:23.956359 140659750036672 submission_runner.py:516] Time since start: 64764.94s, Step: 205873, {'train/accuracy': Array(0.95533717, dtype=float32), 'train/loss': Array(0.15522939, dtype=float32), 'validation/accuracy': Array(0.75549996, dtype=float32), 'validation/loss': Array(1.0767725, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6251, dtype=float32), 'test/loss': Array(1.8661743, dtype=float32), 'test/num_examples': 10000, 'score': 63920.629023075104, 'total_duration': 64764.937913656235, 'accumulated_submission_time': 63920.629023075104, 'accumulated_eval_time': 834.8467264175415, 'accumulated_logging_time': 6.078795433044434} +I0901 03:17:24.392110 140462207792896 logging_writer.py:48] [205873] accumulated_eval_time=834.847, accumulated_logging_time=6.0788, accumulated_submission_time=63920.6, global_step=205873, preemption_count=0, score=63920.6, test/accuracy=0.6251000165939331, test/loss=1.866174340248108, test/num_examples=10000, total_duration=64764.9, train/accuracy=0.9553371667861938, train/loss=0.155229389667511, validation/accuracy=0.7554999589920044, validation/loss=1.0767724514007568, validation/num_examples=50000 +I0901 03:17:32.472672 140462291654400 logging_writer.py:48] [205900] global_step=205900, grad_norm=5.138431072235107, loss=0.9394885301589966 +I0901 03:18:24.891525 140462207792896 logging_writer.py:48] [206000] global_step=206000, grad_norm=5.084817886352539, loss=0.8668659925460815 +I0901 03:19:15.236071 140462291654400 logging_writer.py:48] [206100] global_step=206100, grad_norm=4.837433815002441, loss=0.8696001768112183 +I0901 03:20:06.579502 140462207792896 logging_writer.py:48] [206200] global_step=206200, grad_norm=5.275766849517822, loss=0.939042866230011 +I0901 03:20:56.962096 140462291654400 logging_writer.py:48] [206300] global_step=206300, grad_norm=4.9996209144592285, loss=0.8717149496078491 +I0901 03:21:49.318283 140462207792896 logging_writer.py:48] [206400] global_step=206400, grad_norm=5.188840389251709, loss=0.9823256731033325 +I0901 03:22:41.535969 140462291654400 logging_writer.py:48] [206500] global_step=206500, grad_norm=5.181732654571533, loss=0.9688675999641418 +I0901 03:23:32.132626 140462207792896 logging_writer.py:48] [206600] global_step=206600, grad_norm=4.717846393585205, loss=0.8417767882347107 +I0901 03:24:24.375883 140462291654400 logging_writer.py:48] [206700] global_step=206700, grad_norm=5.298648357391357, loss=1.0614051818847656 +I0901 03:25:14.506935 140462207792896 logging_writer.py:48] [206800] global_step=206800, grad_norm=4.880532741546631, loss=0.9026739597320557 +I0901 03:26:06.788738 140462291654400 logging_writer.py:48] [206900] global_step=206900, grad_norm=4.74763822555542, loss=0.8157472610473633 +I0901 03:26:58.372176 140462207792896 logging_writer.py:48] [207000] global_step=207000, grad_norm=5.40638542175293, loss=0.9624335765838623 +I0901 03:27:48.899054 140462291654400 logging_writer.py:48] [207100] global_step=207100, grad_norm=5.26674747467041, loss=0.950135350227356 +I0901 03:28:39.381477 140462207792896 logging_writer.py:48] [207200] global_step=207200, grad_norm=5.063658714294434, loss=0.9025247693061829 +I0901 03:29:29.595429 140462291654400 logging_writer.py:48] [207300] global_step=207300, grad_norm=4.902191162109375, loss=0.8619416356086731 +I0901 03:30:18.488497 140462207792896 logging_writer.py:48] [207400] global_step=207400, grad_norm=5.5563178062438965, loss=0.9612510204315186 +I0901 03:31:11.090416 140462291654400 logging_writer.py:48] [207500] global_step=207500, grad_norm=5.088079929351807, loss=0.8902285099029541 +I0901 03:32:00.248851 140462207792896 logging_writer.py:48] [207600] global_step=207600, grad_norm=5.168347358703613, loss=0.899899959564209 +I0901 03:32:50.090051 140462291654400 logging_writer.py:48] [207700] global_step=207700, grad_norm=5.195590972900391, loss=0.9421630501747131 +I0901 03:33:41.837281 140462207792896 logging_writer.py:48] [207800] global_step=207800, grad_norm=5.412221908569336, loss=1.0316530466079712 +I0901 03:34:32.411365 140462291654400 logging_writer.py:48] [207900] global_step=207900, grad_norm=4.97196626663208, loss=0.8951390385627747 +I0901 03:35:22.383526 140462207792896 logging_writer.py:48] [208000] global_step=208000, grad_norm=5.143775463104248, loss=0.9214614629745483 +I0901 03:36:12.035170 140462291654400 logging_writer.py:48] [208100] global_step=208100, grad_norm=5.372555255889893, loss=0.9859233498573303 +I0901 03:37:03.326333 140462207792896 logging_writer.py:48] [208200] global_step=208200, grad_norm=5.264011383056641, loss=0.9322772026062012 +I0901 03:37:53.192997 140462291654400 logging_writer.py:48] [208300] global_step=208300, grad_norm=5.067671298980713, loss=0.8775922656059265 +I0901 03:38:40.792531 140462207792896 logging_writer.py:48] [208400] global_step=208400, grad_norm=5.231860160827637, loss=0.918602466583252 +I0901 03:39:29.775679 140462291654400 logging_writer.py:48] [208500] global_step=208500, grad_norm=5.023444652557373, loss=0.9290660619735718 +I0901 03:40:19.091299 140462207792896 logging_writer.py:48] [208600] global_step=208600, grad_norm=5.723453521728516, loss=0.9950141906738281 +I0901 03:41:08.386821 140462291654400 logging_writer.py:48] [208700] global_step=208700, grad_norm=5.120316982269287, loss=0.971717357635498 +I0901 03:41:58.370141 140462207792896 logging_writer.py:48] [208800] global_step=208800, grad_norm=5.315264701843262, loss=0.9530516266822815 +I0901 03:42:48.494664 140462291654400 logging_writer.py:48] [208900] global_step=208900, grad_norm=5.145333290100098, loss=0.9447846412658691 +I0901 03:43:39.967115 140462207792896 logging_writer.py:48] [209000] global_step=209000, grad_norm=5.2433247566223145, loss=0.8558214902877808 +I0901 03:44:29.474658 140462291654400 logging_writer.py:48] [209100] global_step=209100, grad_norm=5.314075946807861, loss=0.9495285749435425 +I0901 03:45:21.450718 140462207792896 logging_writer.py:48] [209200] global_step=209200, grad_norm=5.146735191345215, loss=0.9017874598503113 +I0901 03:46:09.476610 140462291654400 logging_writer.py:48] [209300] global_step=209300, grad_norm=5.461965084075928, loss=0.8903591632843018 +I0901 03:46:58.988608 140462207792896 logging_writer.py:48] [209400] global_step=209400, grad_norm=5.222638130187988, loss=0.9079093933105469 +I0901 03:47:50.368299 140462291654400 logging_writer.py:48] [209500] global_step=209500, grad_norm=5.205080986022949, loss=0.8829348087310791 +I0901 03:48:39.636187 140462207792896 logging_writer.py:48] [209600] global_step=209600, grad_norm=4.99208402633667, loss=0.9260611534118652 +I0901 03:49:28.017786 140462291654400 logging_writer.py:48] [209700] global_step=209700, grad_norm=4.959572792053223, loss=0.9283987879753113 +I0901 03:50:16.407686 140462207792896 logging_writer.py:48] [209800] global_step=209800, grad_norm=5.411559104919434, loss=0.9881188869476318 +I0901 03:50:40.427215 140659750036672 spec.py:333] Evaluating on the training split. +I0901 03:50:48.666615 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 03:51:21.524243 140659750036672 spec.py:363] Evaluating on the test split. +I0901 03:51:22.702107 140659750036672 submission_runner.py:516] Time since start: 66803.60s, Step: 209850, {'train/accuracy': Array(0.9548788, dtype=float32), 'train/loss': Array(0.15816157, dtype=float32), 'validation/accuracy': Array(0.7553, dtype=float32), 'validation/loss': Array(1.0794655, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62710005, dtype=float32), 'test/loss': Array(1.8678958, dtype=float32), 'test/num_examples': 10000, 'score': 65916.38125610352, 'total_duration': 66803.59570097923, 'accumulated_submission_time': 65916.38125610352, 'accumulated_eval_time': 876.8376929759979, 'accumulated_logging_time': 6.746559143066406} +I0901 03:51:23.104138 140462291654400 logging_writer.py:48] [209850] accumulated_eval_time=876.838, accumulated_logging_time=6.74656, accumulated_submission_time=65916.4, global_step=209850, preemption_count=0, score=65916.4, test/accuracy=0.6271000504493713, test/loss=1.8678958415985107, test/num_examples=10000, total_duration=66803.6, train/accuracy=0.9548788070678711, train/loss=0.15816156566143036, validation/accuracy=0.755299985408783, validation/loss=1.0794655084609985, validation/num_examples=50000 +I0901 03:51:42.797831 140462207792896 logging_writer.py:48] [209900] global_step=209900, grad_norm=4.975810527801514, loss=0.8738003373146057 +I0901 03:52:30.831182 140462291654400 logging_writer.py:48] [210000] global_step=210000, grad_norm=5.202803134918213, loss=0.9538751244544983 +I0901 03:53:19.564219 140462207792896 logging_writer.py:48] [210100] global_step=210100, grad_norm=5.218823432922363, loss=0.8986384868621826 +I0901 03:54:10.965544 140462291654400 logging_writer.py:48] [210200] global_step=210200, grad_norm=5.191321849822998, loss=0.9542730450630188 +I0901 03:55:01.645243 140462207792896 logging_writer.py:48] [210300] global_step=210300, grad_norm=5.5616841316223145, loss=0.9296600818634033 +I0901 03:55:51.147323 140462291654400 logging_writer.py:48] [210400] global_step=210400, grad_norm=5.025953769683838, loss=0.8690745234489441 +I0901 03:56:40.085953 140462207792896 logging_writer.py:48] [210500] global_step=210500, grad_norm=5.179275035858154, loss=0.9378422498703003 +I0901 03:57:29.179676 140462291654400 logging_writer.py:48] [210600] global_step=210600, grad_norm=5.0880022048950195, loss=0.9239691495895386 +I0901 03:58:19.329769 140462207792896 logging_writer.py:48] [210700] global_step=210700, grad_norm=5.164870738983154, loss=1.0061347484588623 +I0901 03:59:08.177661 140462291654400 logging_writer.py:48] [210800] global_step=210800, grad_norm=5.154817581176758, loss=0.9579209089279175 +I0901 03:59:56.074132 140462207792896 logging_writer.py:48] [210900] global_step=210900, grad_norm=5.561718940734863, loss=1.0284490585327148 +I0901 04:00:44.169312 140462291654400 logging_writer.py:48] [211000] global_step=211000, grad_norm=5.478799343109131, loss=0.9301892518997192 +I0901 04:01:31.116252 140462207792896 logging_writer.py:48] [211100] global_step=211100, grad_norm=5.2482428550720215, loss=1.0191185474395752 +I0901 04:02:19.191117 140462291654400 logging_writer.py:48] [211200] global_step=211200, grad_norm=5.548893451690674, loss=0.9359729290008545 +I0901 04:03:08.576980 140462207792896 logging_writer.py:48] [211300] global_step=211300, grad_norm=5.073635578155518, loss=0.8680304288864136 +I0901 04:03:58.297357 140462291654400 logging_writer.py:48] [211400] global_step=211400, grad_norm=5.1520915031433105, loss=0.9495339393615723 +I0901 04:04:49.858392 140462207792896 logging_writer.py:48] [211500] global_step=211500, grad_norm=5.212430953979492, loss=0.8735074400901794 +I0901 04:05:40.486253 140462291654400 logging_writer.py:48] [211600] global_step=211600, grad_norm=5.623482704162598, loss=1.1023869514465332 +I0901 04:06:30.377322 140462207792896 logging_writer.py:48] [211700] global_step=211700, grad_norm=5.077763080596924, loss=0.9248204231262207 +I0901 04:07:21.409331 140462291654400 logging_writer.py:48] [211800] global_step=211800, grad_norm=5.128161430358887, loss=0.9439595937728882 +I0901 04:08:09.320055 140462207792896 logging_writer.py:48] [211900] global_step=211900, grad_norm=5.134617328643799, loss=0.952720582485199 +I0901 04:08:58.308761 140462291654400 logging_writer.py:48] [212000] global_step=212000, grad_norm=5.08184814453125, loss=0.9153573513031006 +I0901 04:09:46.972957 140462207792896 logging_writer.py:48] [212100] global_step=212100, grad_norm=5.125270843505859, loss=0.9258460402488708 +I0901 04:10:36.313213 140462291654400 logging_writer.py:48] [212200] global_step=212200, grad_norm=5.423209190368652, loss=0.9479548931121826 +I0901 04:11:23.313781 140462207792896 logging_writer.py:48] [212300] global_step=212300, grad_norm=5.426752090454102, loss=0.9370129108428955 +I0901 04:12:10.962800 140462291654400 logging_writer.py:48] [212400] global_step=212400, grad_norm=5.446508407592773, loss=0.9776602983474731 +I0901 04:13:00.587659 140462207792896 logging_writer.py:48] [212500] global_step=212500, grad_norm=5.247865676879883, loss=0.9407178163528442 +I0901 04:13:49.990008 140462291654400 logging_writer.py:48] [212600] global_step=212600, grad_norm=4.956753253936768, loss=0.8992116451263428 +I0901 04:14:40.888363 140462207792896 logging_writer.py:48] [212700] global_step=212700, grad_norm=4.867729663848877, loss=0.9097065925598145 +I0901 04:15:31.013297 140462291654400 logging_writer.py:48] [212800] global_step=212800, grad_norm=4.921652317047119, loss=0.8248821496963501 +I0901 04:16:19.183791 140462207792896 logging_writer.py:48] [212900] global_step=212900, grad_norm=5.121992588043213, loss=0.9241592288017273 +I0901 04:17:08.635384 140462291654400 logging_writer.py:48] [213000] global_step=213000, grad_norm=5.378377914428711, loss=0.9437297582626343 +I0901 04:17:57.222795 140462207792896 logging_writer.py:48] [213100] global_step=213100, grad_norm=5.437767028808594, loss=1.0068755149841309 +I0901 04:18:45.582759 140462291654400 logging_writer.py:48] [213200] global_step=213200, grad_norm=5.165266513824463, loss=0.974541187286377 +I0901 04:19:32.427398 140462207792896 logging_writer.py:48] [213300] global_step=213300, grad_norm=5.38306188583374, loss=1.0098240375518799 +I0901 04:20:19.549093 140462291654400 logging_writer.py:48] [213400] global_step=213400, grad_norm=4.986531734466553, loss=0.9592106342315674 +I0901 04:21:07.587090 140462207792896 logging_writer.py:48] [213500] global_step=213500, grad_norm=5.183727741241455, loss=0.8461059331893921 +I0901 04:21:52.472969 140462291654400 logging_writer.py:48] [213600] global_step=213600, grad_norm=5.192490100860596, loss=0.9274892807006836 +I0901 04:22:38.355849 140462207792896 logging_writer.py:48] [213700] global_step=213700, grad_norm=5.0443525314331055, loss=0.8849000930786133 +I0901 04:23:24.336490 140462291654400 logging_writer.py:48] [213800] global_step=213800, grad_norm=4.883209228515625, loss=0.8536074161529541 +I0901 04:24:12.378383 140462207792896 logging_writer.py:48] [213900] global_step=213900, grad_norm=5.500051498413086, loss=0.9587920904159546 +I0901 04:24:38.822013 140659750036672 spec.py:333] Evaluating on the training split. +I0901 04:24:47.177166 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 04:25:21.600395 140659750036672 spec.py:363] Evaluating on the test split. +I0901 04:25:22.707769 140659750036672 submission_runner.py:516] Time since start: 68843.67s, Step: 213955, {'train/accuracy': Array(0.9560945, dtype=float32), 'train/loss': Array(0.15828097, dtype=float32), 'validation/accuracy': Array(0.75492, dtype=float32), 'validation/loss': Array(1.0802552, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6265, dtype=float32), 'test/loss': Array(1.8719171, dtype=float32), 'test/num_examples': 10000, 'score': 67912.0241098404, 'total_duration': 68843.66978883743, 'accumulated_submission_time': 67912.0241098404, 'accumulated_eval_time': 920.5079526901245, 'accumulated_logging_time': 7.185899257659912} +I0901 04:25:23.142747 140462291654400 logging_writer.py:48] [213955] accumulated_eval_time=920.508, accumulated_logging_time=7.1859, accumulated_submission_time=67912, global_step=213955, preemption_count=0, score=67912, test/accuracy=0.6265000104904175, test/loss=1.8719171285629272, test/num_examples=10000, total_duration=68843.7, train/accuracy=0.95609450340271, train/loss=0.15828096866607666, validation/accuracy=0.7549200057983398, validation/loss=1.080255150794983, validation/num_examples=50000 +I0901 04:25:39.473812 140462207792896 logging_writer.py:48] [214000] global_step=214000, grad_norm=5.1583781242370605, loss=0.9822828769683838 +I0901 04:26:27.118780 140462291654400 logging_writer.py:48] [214100] global_step=214100, grad_norm=5.411642551422119, loss=0.974001407623291 +I0901 04:27:13.427220 140462207792896 logging_writer.py:48] [214200] global_step=214200, grad_norm=5.3032379150390625, loss=0.9320365190505981 +I0901 04:27:59.172486 140462291654400 logging_writer.py:48] [214300] global_step=214300, grad_norm=4.928860664367676, loss=0.8824174404144287 +I0901 04:28:46.265114 140462207792896 logging_writer.py:48] [214400] global_step=214400, grad_norm=5.251873016357422, loss=0.9090210199356079 +I0901 04:29:32.917682 140462291654400 logging_writer.py:48] [214500] global_step=214500, grad_norm=4.922919750213623, loss=0.7535113096237183 +I0901 04:30:17.890798 140462207792896 logging_writer.py:48] [214600] global_step=214600, grad_norm=5.280426979064941, loss=0.9520424008369446 +I0901 04:31:04.422454 140462291654400 logging_writer.py:48] [214700] global_step=214700, grad_norm=5.138182640075684, loss=0.9079441428184509 +I0901 04:31:51.732976 140462207792896 logging_writer.py:48] [214800] global_step=214800, grad_norm=5.278894424438477, loss=0.9643752574920654 +I0901 04:32:37.418813 140462291654400 logging_writer.py:48] [214900] global_step=214900, grad_norm=5.279441833496094, loss=0.9565223455429077 +I0901 04:33:24.292728 140462207792896 logging_writer.py:48] [215000] global_step=215000, grad_norm=5.0810227394104, loss=0.9058865308761597 +I0901 04:34:12.970263 140462291654400 logging_writer.py:48] [215100] global_step=215100, grad_norm=5.106866359710693, loss=0.8790436387062073 +I0901 04:35:02.749268 140462207792896 logging_writer.py:48] [215200] global_step=215200, grad_norm=5.046652317047119, loss=0.9313181638717651 +I0901 04:35:49.162949 140462291654400 logging_writer.py:48] [215300] global_step=215300, grad_norm=5.205214023590088, loss=1.020447015762329 +I0901 04:36:35.507387 140462207792896 logging_writer.py:48] [215400] global_step=215400, grad_norm=4.986790180206299, loss=0.9134372472763062 +I0901 04:37:21.477267 140462291654400 logging_writer.py:48] [215500] global_step=215500, grad_norm=5.3716912269592285, loss=1.0072996616363525 +I0901 04:38:06.793192 140462207792896 logging_writer.py:48] [215600] global_step=215600, grad_norm=5.423051357269287, loss=0.9340770244598389 +I0901 04:38:53.431885 140462291654400 logging_writer.py:48] [215700] global_step=215700, grad_norm=5.476779460906982, loss=0.9591609239578247 +I0901 04:39:38.353510 140462207792896 logging_writer.py:48] [215800] global_step=215800, grad_norm=4.924777030944824, loss=0.8952077031135559 +I0901 04:40:23.963058 140462291654400 logging_writer.py:48] [215900] global_step=215900, grad_norm=5.789702415466309, loss=1.0134036540985107 +I0901 04:41:10.256492 140462207792896 logging_writer.py:48] [216000] global_step=216000, grad_norm=5.206481456756592, loss=0.8996123671531677 +I0901 04:41:56.772389 140462291654400 logging_writer.py:48] [216100] global_step=216100, grad_norm=5.005434989929199, loss=0.8271104097366333 +I0901 04:42:42.165775 140462207792896 logging_writer.py:48] [216200] global_step=216200, grad_norm=5.3370561599731445, loss=0.9025671482086182 +I0901 04:43:27.918126 140462291654400 logging_writer.py:48] [216300] global_step=216300, grad_norm=5.284339904785156, loss=0.9957187175750732 +I0901 04:44:13.508297 140462207792896 logging_writer.py:48] [216400] global_step=216400, grad_norm=5.274785995483398, loss=0.9208605289459229 +I0901 04:44:59.441037 140462291654400 logging_writer.py:48] [216500] global_step=216500, grad_norm=5.16731595993042, loss=0.8734488487243652 +I0901 04:45:43.170606 140462207792896 logging_writer.py:48] [216600] global_step=216600, grad_norm=5.340208530426025, loss=0.9028455018997192 +I0901 04:46:27.765397 140462291654400 logging_writer.py:48] [216700] global_step=216700, grad_norm=5.1465840339660645, loss=0.902611494064331 +I0901 04:47:12.350570 140462207792896 logging_writer.py:48] [216800] global_step=216800, grad_norm=5.192099571228027, loss=0.9202262759208679 +I0901 04:47:56.530243 140462291654400 logging_writer.py:48] [216900] global_step=216900, grad_norm=5.1995062828063965, loss=0.8770672082901001 +I0901 04:48:40.468916 140462207792896 logging_writer.py:48] [217000] global_step=217000, grad_norm=5.204635143280029, loss=0.922393798828125 +I0901 04:49:21.465444 140462291654400 logging_writer.py:48] [217100] global_step=217100, grad_norm=5.265316009521484, loss=1.0086219310760498 +I0901 04:50:02.473525 140462207792896 logging_writer.py:48] [217200] global_step=217200, grad_norm=4.9912309646606445, loss=0.9071167707443237 +I0901 04:50:43.948445 140462291654400 logging_writer.py:48] [217300] global_step=217300, grad_norm=5.407994270324707, loss=0.9898466467857361 +I0901 04:51:24.991238 140462207792896 logging_writer.py:48] [217400] global_step=217400, grad_norm=5.315845489501953, loss=0.9348857402801514 +I0901 04:52:06.362661 140462291654400 logging_writer.py:48] [217500] global_step=217500, grad_norm=5.3405256271362305, loss=0.9057672023773193 +I0901 04:52:48.386059 140462207792896 logging_writer.py:48] [217600] global_step=217600, grad_norm=5.304472923278809, loss=1.0056856870651245 +I0901 04:53:30.245046 140462291654400 logging_writer.py:48] [217700] global_step=217700, grad_norm=5.162868499755859, loss=0.9535608887672424 +I0901 04:54:12.898977 140462207792896 logging_writer.py:48] [217800] global_step=217800, grad_norm=5.013586044311523, loss=0.8200922012329102 +I0901 04:54:56.060204 140462291654400 logging_writer.py:48] [217900] global_step=217900, grad_norm=5.121224880218506, loss=0.8839071393013 +I0901 04:55:37.723022 140462207792896 logging_writer.py:48] [218000] global_step=218000, grad_norm=5.367900371551514, loss=0.971290111541748 +I0901 04:56:19.222618 140462291654400 logging_writer.py:48] [218100] global_step=218100, grad_norm=5.155570030212402, loss=0.9310786724090576 +I0901 04:56:59.340190 140462207792896 logging_writer.py:48] [218200] global_step=218200, grad_norm=4.991910934448242, loss=0.8853803873062134 +I0901 04:57:40.829395 140462291654400 logging_writer.py:48] [218300] global_step=218300, grad_norm=5.1287970542907715, loss=0.889752984046936 +I0901 04:58:22.555957 140462207792896 logging_writer.py:48] [218400] global_step=218400, grad_norm=5.045790195465088, loss=0.9100313782691956 +I0901 04:58:38.667340 140659750036672 spec.py:333] Evaluating on the training split. +I0901 04:58:46.970493 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 04:59:01.572065 140659750036672 spec.py:363] Evaluating on the test split. +I0901 04:59:02.636217 140659750036672 submission_runner.py:516] Time since start: 70863.64s, Step: 218440, {'train/accuracy': Array(0.95657283, dtype=float32), 'train/loss': Array(0.15558098, dtype=float32), 'validation/accuracy': Array(0.75588, dtype=float32), 'validation/loss': Array(1.08051, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6284, dtype=float32), 'test/loss': Array(1.8729472, dtype=float32), 'test/num_examples': 10000, 'score': 69907.44102859497, 'total_duration': 70863.64268183708, 'accumulated_submission_time': 69907.44102859497, 'accumulated_eval_time': 944.3057725429535, 'accumulated_logging_time': 7.6864330768585205} +I0901 04:59:03.061933 140462291654400 logging_writer.py:48] [218440] accumulated_eval_time=944.306, accumulated_logging_time=7.68643, accumulated_submission_time=69907.4, global_step=218440, preemption_count=0, score=69907.4, test/accuracy=0.6284000277519226, test/loss=1.8729472160339355, test/num_examples=10000, total_duration=70863.6, train/accuracy=0.9565728306770325, train/loss=0.15558098256587982, validation/accuracy=0.7558799982070923, validation/loss=1.0805100202560425, validation/num_examples=50000 +I0901 04:59:23.601460 140462207792896 logging_writer.py:48] [218500] global_step=218500, grad_norm=5.162540435791016, loss=0.8738983273506165 +I0901 05:00:05.191762 140462291654400 logging_writer.py:48] [218600] global_step=218600, grad_norm=4.96341609954834, loss=0.8867851495742798 +I0901 05:00:46.976233 140462207792896 logging_writer.py:48] [218700] global_step=218700, grad_norm=5.533702850341797, loss=0.9417701363563538 +I0901 05:01:30.140523 140462291654400 logging_writer.py:48] [218800] global_step=218800, grad_norm=5.046479225158691, loss=0.8965269327163696 +I0901 05:02:11.360483 140462207792896 logging_writer.py:48] [218900] global_step=218900, grad_norm=5.028689861297607, loss=0.8683226108551025 +I0901 05:02:55.223993 140462291654400 logging_writer.py:48] [219000] global_step=219000, grad_norm=5.559327125549316, loss=1.0115939378738403 +I0901 05:03:36.899252 140462207792896 logging_writer.py:48] [219100] global_step=219100, grad_norm=5.238500118255615, loss=0.8736231923103333 +I0901 05:04:18.007331 140462291654400 logging_writer.py:48] [219200] global_step=219200, grad_norm=5.002896308898926, loss=0.8560653924942017 +I0901 05:05:01.216708 140462207792896 logging_writer.py:48] [219300] global_step=219300, grad_norm=5.293589115142822, loss=0.9241212606430054 +I0901 05:05:43.105331 140462291654400 logging_writer.py:48] [219400] global_step=219400, grad_norm=5.285518646240234, loss=0.9389466047286987 +I0901 05:06:23.400112 140462207792896 logging_writer.py:48] [219500] global_step=219500, grad_norm=5.176750183105469, loss=0.9202163219451904 +I0901 05:07:05.345167 140462291654400 logging_writer.py:48] [219600] global_step=219600, grad_norm=5.148797512054443, loss=0.9449384808540344 +I0901 05:07:47.633664 140462207792896 logging_writer.py:48] [219700] global_step=219700, grad_norm=5.098997592926025, loss=0.9147289395332336 +I0901 05:08:28.783918 140462291654400 logging_writer.py:48] [219800] global_step=219800, grad_norm=5.040084362030029, loss=0.9240447878837585 +I0901 05:09:09.924158 140462207792896 logging_writer.py:48] [219900] global_step=219900, grad_norm=5.295650482177734, loss=0.9965653419494629 +I0901 05:09:51.134065 140462291654400 logging_writer.py:48] [220000] global_step=220000, grad_norm=5.081548690795898, loss=0.875832200050354 +I0901 05:10:34.244324 140462207792896 logging_writer.py:48] [220100] global_step=220100, grad_norm=4.809747695922852, loss=0.8481124639511108 +I0901 05:11:17.986125 140462291654400 logging_writer.py:48] [220200] global_step=220200, grad_norm=5.295973777770996, loss=0.9327491521835327 +I0901 05:11:59.791867 140462207792896 logging_writer.py:48] [220300] global_step=220300, grad_norm=4.762955665588379, loss=0.899517297744751 +I0901 05:12:41.045066 140462291654400 logging_writer.py:48] [220400] global_step=220400, grad_norm=5.1114630699157715, loss=0.9021809101104736 +I0901 05:13:21.333973 140462207792896 logging_writer.py:48] [220500] global_step=220500, grad_norm=5.118753433227539, loss=0.9135053157806396 +I0901 05:14:00.949620 140462291654400 logging_writer.py:48] [220600] global_step=220600, grad_norm=5.103907108306885, loss=0.9097748398780823 +I0901 05:14:39.586796 140462207792896 logging_writer.py:48] [220700] global_step=220700, grad_norm=5.280526161193848, loss=0.9055071473121643 +I0901 05:15:12.594802 140462291654400 logging_writer.py:48] [220800] global_step=220800, grad_norm=5.015108585357666, loss=0.8563700914382935 +I0901 05:15:44.584255 140462207792896 logging_writer.py:48] [220900] global_step=220900, grad_norm=5.028525352478027, loss=0.9307546019554138 +I0901 05:16:16.206203 140462291654400 logging_writer.py:48] [221000] global_step=221000, grad_norm=5.194873332977295, loss=0.9471676349639893 +I0901 05:16:50.315892 140462207792896 logging_writer.py:48] [221100] global_step=221100, grad_norm=5.2057952880859375, loss=0.8555071353912354 +I0901 05:17:24.555025 140462291654400 logging_writer.py:48] [221200] global_step=221200, grad_norm=5.240876197814941, loss=0.9277920126914978 +I0901 05:17:59.664345 140462207792896 logging_writer.py:48] [221300] global_step=221300, grad_norm=5.131241798400879, loss=0.9123184084892273 +I0901 05:18:33.593823 140462291654400 logging_writer.py:48] [221400] global_step=221400, grad_norm=5.162561416625977, loss=0.8571287393569946 +I0901 05:19:08.552078 140462207792896 logging_writer.py:48] [221500] global_step=221500, grad_norm=5.06036901473999, loss=0.89594966173172 +I0901 05:19:42.233757 140462291654400 logging_writer.py:48] [221600] global_step=221600, grad_norm=4.9440388679504395, loss=0.8495545387268066 +I0901 05:20:15.915149 140462207792896 logging_writer.py:48] [221700] global_step=221700, grad_norm=5.301388740539551, loss=0.9402251243591309 +I0901 05:20:50.503466 140462291654400 logging_writer.py:48] [221800] global_step=221800, grad_norm=4.79655647277832, loss=0.8326020240783691 +I0901 05:21:24.587134 140462207792896 logging_writer.py:48] [221900] global_step=221900, grad_norm=5.378048896789551, loss=0.9684708118438721 +I0901 05:21:57.234163 140462291654400 logging_writer.py:48] [222000] global_step=222000, grad_norm=5.089557647705078, loss=0.9247675538063049 +I0901 05:22:28.547178 140462207792896 logging_writer.py:48] [222100] global_step=222100, grad_norm=5.101194381713867, loss=0.867091178894043 +I0901 05:23:00.383990 140462291654400 logging_writer.py:48] [222200] global_step=222200, grad_norm=5.030780792236328, loss=0.9035369157791138 +I0901 05:23:33.071630 140462207792896 logging_writer.py:48] [222300] global_step=222300, grad_norm=5.321121692657471, loss=0.9779195189476013 +I0901 05:24:05.283827 140462291654400 logging_writer.py:48] [222400] global_step=222400, grad_norm=5.405971050262451, loss=0.920830488204956 +I0901 05:24:37.957191 140462207792896 logging_writer.py:48] [222500] global_step=222500, grad_norm=5.372992515563965, loss=0.9580941200256348 +I0901 05:25:10.673765 140462291654400 logging_writer.py:48] [222600] global_step=222600, grad_norm=5.282313346862793, loss=0.8856818079948425 +I0901 05:25:44.085088 140462207792896 logging_writer.py:48] [222700] global_step=222700, grad_norm=4.7988386154174805, loss=0.8005345463752747 +I0901 05:26:16.436264 140462291654400 logging_writer.py:48] [222800] global_step=222800, grad_norm=5.290060997009277, loss=0.9508973956108093 +I0901 05:26:49.736920 140462207792896 logging_writer.py:48] [222900] global_step=222900, grad_norm=5.002452850341797, loss=0.8467998504638672 +I0901 05:27:23.455564 140462291654400 logging_writer.py:48] [223000] global_step=223000, grad_norm=5.560652732849121, loss=0.9224489331245422 +I0901 05:27:55.635149 140462207792896 logging_writer.py:48] [223100] global_step=223100, grad_norm=4.962220668792725, loss=0.893566906452179 +I0901 05:28:28.622892 140462291654400 logging_writer.py:48] [223200] global_step=223200, grad_norm=5.077077865600586, loss=0.8835417628288269 +I0901 05:29:01.260420 140462207792896 logging_writer.py:48] [223300] global_step=223300, grad_norm=4.722568035125732, loss=0.8736380338668823 +I0901 05:29:34.273398 140462291654400 logging_writer.py:48] [223400] global_step=223400, grad_norm=5.070638179779053, loss=0.8940515518188477 +I0901 05:30:07.216280 140462207792896 logging_writer.py:48] [223500] global_step=223500, grad_norm=5.085115909576416, loss=0.8741628527641296 +I0901 05:30:39.959971 140462291654400 logging_writer.py:48] [223600] global_step=223600, grad_norm=5.195613384246826, loss=0.8970745801925659 +I0901 05:31:13.054150 140462207792896 logging_writer.py:48] [223700] global_step=223700, grad_norm=5.04409646987915, loss=0.8703240752220154 +I0901 05:31:46.135004 140462291654400 logging_writer.py:48] [223800] global_step=223800, grad_norm=5.1587066650390625, loss=0.9231534004211426 +I0901 05:32:18.786248 140659750036672 spec.py:333] Evaluating on the training split. +I0901 05:32:26.874207 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 05:33:01.275519 140659750036672 spec.py:363] Evaluating on the test split. +I0901 05:33:02.304513 140659750036672 submission_runner.py:516] Time since start: 72903.30s, Step: 223900, {'train/accuracy': Array(0.9569515, dtype=float32), 'train/loss': Array(0.15209584, dtype=float32), 'validation/accuracy': Array(0.75641996, dtype=float32), 'validation/loss': Array(1.0811065, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6279, dtype=float32), 'test/loss': Array(1.8756953, dtype=float32), 'test/num_examples': 10000, 'score': 71903.08286333084, 'total_duration': 72903.2956624031, 'accumulated_submission_time': 71903.08286333084, 'accumulated_eval_time': 987.6376729011536, 'accumulated_logging_time': 8.14183759689331} +I0901 05:33:02.742921 140462207792896 logging_writer.py:48] [223900] accumulated_eval_time=987.638, accumulated_logging_time=8.14184, accumulated_submission_time=71903.1, global_step=223900, preemption_count=0, score=71903.1, test/accuracy=0.6279000043869019, test/loss=1.8756953477859497, test/num_examples=10000, total_duration=72903.3, train/accuracy=0.9569514989852905, train/loss=0.15209583938121796, validation/accuracy=0.7564199566841125, validation/loss=1.0811065435409546, validation/num_examples=50000 +I0901 05:33:03.117308 140462291654400 logging_writer.py:48] [223900] global_step=223900, grad_norm=5.592931747436523, loss=0.9401077032089233 +I0901 05:33:41.976098 140462207792896 logging_writer.py:48] [224000] global_step=224000, grad_norm=4.979752063751221, loss=0.901873767375946 +I0901 05:34:25.508866 140462291654400 logging_writer.py:48] [224100] global_step=224100, grad_norm=5.2453179359436035, loss=0.8422667980194092 +I0901 05:35:05.751593 140462207792896 logging_writer.py:48] [224200] global_step=224200, grad_norm=5.149021148681641, loss=0.9347405433654785 +I0901 05:35:39.907235 140462291654400 logging_writer.py:48] [224300] global_step=224300, grad_norm=4.876700401306152, loss=0.8228070735931396 +I0901 05:36:14.839515 140462207792896 logging_writer.py:48] [224400] global_step=224400, grad_norm=5.2318572998046875, loss=0.8939640522003174 +I0901 05:36:46.166459 140462291654400 logging_writer.py:48] [224500] global_step=224500, grad_norm=5.367684841156006, loss=0.945205807685852 +I0901 05:37:18.206210 140462207792896 logging_writer.py:48] [224600] global_step=224600, grad_norm=4.9757890701293945, loss=0.8730068206787109 +I0901 05:37:49.698336 140462291654400 logging_writer.py:48] [224700] global_step=224700, grad_norm=5.0797624588012695, loss=0.9025545120239258 +I0901 05:38:21.136781 140462207792896 logging_writer.py:48] [224800] global_step=224800, grad_norm=5.392987251281738, loss=0.8949060440063477 +I0901 05:38:53.909569 140462291654400 logging_writer.py:48] [224900] global_step=224900, grad_norm=4.970251083374023, loss=0.8504171967506409 +I0901 05:39:27.165746 140462207792896 logging_writer.py:48] [225000] global_step=225000, grad_norm=5.259399890899658, loss=0.9436864256858826 +I0901 05:40:00.325066 140462291654400 logging_writer.py:48] [225100] global_step=225100, grad_norm=4.982584476470947, loss=0.8678619861602783 +I0901 05:40:34.533926 140462207792896 logging_writer.py:48] [225200] global_step=225200, grad_norm=5.2541890144348145, loss=0.9355782270431519 +I0901 05:41:07.449899 140462291654400 logging_writer.py:48] [225300] global_step=225300, grad_norm=5.486793518066406, loss=0.982403039932251 +I0901 05:41:40.666162 140462207792896 logging_writer.py:48] [225400] global_step=225400, grad_norm=5.201714992523193, loss=0.9282773733139038 +I0901 05:42:13.175510 140462291654400 logging_writer.py:48] [225500] global_step=225500, grad_norm=5.0024309158325195, loss=0.8954651355743408 +I0901 05:42:46.263078 140462207792896 logging_writer.py:48] [225600] global_step=225600, grad_norm=5.2361369132995605, loss=0.8853433132171631 +I0901 05:43:19.459302 140462291654400 logging_writer.py:48] [225700] global_step=225700, grad_norm=5.160604953765869, loss=0.9501055479049683 +I0901 05:43:52.535014 140462207792896 logging_writer.py:48] [225800] global_step=225800, grad_norm=5.174063205718994, loss=0.8815574645996094 +I0901 05:44:26.187019 140462291654400 logging_writer.py:48] [225900] global_step=225900, grad_norm=4.794394016265869, loss=0.8658007383346558 +I0901 05:44:58.636929 140462207792896 logging_writer.py:48] [226000] global_step=226000, grad_norm=5.072536468505859, loss=0.8909847736358643 +I0901 05:45:31.648150 140462291654400 logging_writer.py:48] [226100] global_step=226100, grad_norm=5.385575294494629, loss=0.9465888738632202 +I0901 05:46:04.642088 140462207792896 logging_writer.py:48] [226200] global_step=226200, grad_norm=5.248011112213135, loss=0.910929799079895 +I0901 05:46:37.754809 140462291654400 logging_writer.py:48] [226300] global_step=226300, grad_norm=4.929739952087402, loss=0.7784014344215393 +I0901 05:47:10.888382 140462207792896 logging_writer.py:48] [226400] global_step=226400, grad_norm=5.4097490310668945, loss=0.9650675654411316 +I0901 05:47:44.078577 140462291654400 logging_writer.py:48] [226500] global_step=226500, grad_norm=5.075380325317383, loss=0.8996926546096802 +I0901 05:48:16.870699 140462207792896 logging_writer.py:48] [226600] global_step=226600, grad_norm=4.82501745223999, loss=0.8080099821090698 +I0901 05:48:49.762482 140462291654400 logging_writer.py:48] [226700] global_step=226700, grad_norm=5.377864360809326, loss=0.9972296953201294 +I0901 05:49:22.241107 140462207792896 logging_writer.py:48] [226800] global_step=226800, grad_norm=5.041151523590088, loss=0.8765994310379028 +I0901 05:49:54.734000 140462291654400 logging_writer.py:48] [226900] global_step=226900, grad_norm=5.392828464508057, loss=0.8786707520484924 +I0901 05:50:27.342955 140462207792896 logging_writer.py:48] [227000] global_step=227000, grad_norm=5.079688549041748, loss=0.9117456674575806 +I0901 05:50:59.859182 140462291654400 logging_writer.py:48] [227100] global_step=227100, grad_norm=5.399343967437744, loss=0.9503396153450012 +I0901 05:51:32.264692 140462207792896 logging_writer.py:48] [227200] global_step=227200, grad_norm=5.661230087280273, loss=1.0111510753631592 +I0901 05:52:04.615071 140462291654400 logging_writer.py:48] [227300] global_step=227300, grad_norm=5.071844100952148, loss=0.8433272838592529 +I0901 05:52:37.204253 140462207792896 logging_writer.py:48] [227400] global_step=227400, grad_norm=5.155460834503174, loss=0.8233242034912109 +I0901 05:53:09.943991 140462291654400 logging_writer.py:48] [227500] global_step=227500, grad_norm=5.202062129974365, loss=0.9467596411705017 +I0901 05:53:42.651454 140462207792896 logging_writer.py:48] [227600] global_step=227600, grad_norm=5.491517543792725, loss=0.9164580702781677 +I0901 05:54:16.095877 140462291654400 logging_writer.py:48] [227700] global_step=227700, grad_norm=5.269688606262207, loss=0.9534658789634705 +I0901 05:54:48.694826 140462207792896 logging_writer.py:48] [227800] global_step=227800, grad_norm=4.682196617126465, loss=0.8097023963928223 +I0901 05:55:21.472756 140462291654400 logging_writer.py:48] [227900] global_step=227900, grad_norm=5.459268569946289, loss=0.9677984118461609 +I0901 05:55:53.719167 140462207792896 logging_writer.py:48] [228000] global_step=228000, grad_norm=5.113245964050293, loss=0.8796414136886597 +I0901 05:56:26.133395 140462291654400 logging_writer.py:48] [228100] global_step=228100, grad_norm=5.528121471405029, loss=0.9921536445617676 +I0901 05:56:58.818026 140462207792896 logging_writer.py:48] [228200] global_step=228200, grad_norm=5.265801906585693, loss=0.9024748802185059 +I0901 05:57:31.880626 140462291654400 logging_writer.py:48] [228300] global_step=228300, grad_norm=5.315360069274902, loss=0.926902711391449 +I0901 05:58:04.037618 140462207792896 logging_writer.py:48] [228400] global_step=228400, grad_norm=5.178040027618408, loss=0.8517076969146729 +I0901 05:58:36.655311 140462291654400 logging_writer.py:48] [228500] global_step=228500, grad_norm=5.3097734451293945, loss=0.9304254651069641 +I0901 05:59:09.479544 140462207792896 logging_writer.py:48] [228600] global_step=228600, grad_norm=4.922999858856201, loss=0.8705898523330688 +I0901 05:59:42.114216 140462291654400 logging_writer.py:48] [228700] global_step=228700, grad_norm=5.105581283569336, loss=0.9199355244636536 +I0901 06:00:14.878482 140462207792896 logging_writer.py:48] [228800] global_step=228800, grad_norm=4.957713603973389, loss=0.8551656007766724 +I0901 06:00:47.394100 140462291654400 logging_writer.py:48] [228900] global_step=228900, grad_norm=5.098817825317383, loss=0.9101102352142334 +I0901 06:01:20.427586 140462207792896 logging_writer.py:48] [229000] global_step=229000, grad_norm=5.401174545288086, loss=1.0052655935287476 +I0901 06:01:53.103759 140462291654400 logging_writer.py:48] [229100] global_step=229100, grad_norm=5.763489246368408, loss=0.9811801910400391 +I0901 06:02:26.026805 140462207792896 logging_writer.py:48] [229200] global_step=229200, grad_norm=5.276676654815674, loss=0.9189068675041199 +I0901 06:02:58.827723 140462291654400 logging_writer.py:48] [229300] global_step=229300, grad_norm=5.153440475463867, loss=0.924785852432251 +I0901 06:03:31.392812 140462207792896 logging_writer.py:48] [229400] global_step=229400, grad_norm=5.136477470397949, loss=0.8872755169868469 +I0901 06:04:04.620372 140462291654400 logging_writer.py:48] [229500] global_step=229500, grad_norm=5.201347827911377, loss=0.925734281539917 +I0901 06:04:36.794684 140462207792896 logging_writer.py:48] [229600] global_step=229600, grad_norm=5.317071914672852, loss=0.9475840330123901 +I0901 06:05:09.355032 140462291654400 logging_writer.py:48] [229700] global_step=229700, grad_norm=5.030836582183838, loss=0.925849437713623 +I0901 06:05:41.875123 140462207792896 logging_writer.py:48] [229800] global_step=229800, grad_norm=4.80920934677124, loss=0.7673079967498779 +I0901 06:06:14.393722 140462291654400 logging_writer.py:48] [229900] global_step=229900, grad_norm=5.174122333526611, loss=0.8652268648147583 +I0901 06:06:18.198080 140659750036672 spec.py:333] Evaluating on the training split. +I0901 06:06:26.502313 140659750036672 spec.py:346] Evaluating on the validation split. +I0901 06:06:38.147287 140659750036672 spec.py:363] Evaluating on the test split. +I0901 06:06:39.236041 140659750036672 submission_runner.py:516] Time since start: 74920.22s, Step: 229913, {'train/accuracy': Array(0.9563337, dtype=float32), 'train/loss': Array(0.1503854, dtype=float32), 'validation/accuracy': Array(0.75696, dtype=float32), 'validation/loss': Array(1.0823314, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6279, dtype=float32), 'test/loss': Array(1.8746536, dtype=float32), 'test/num_examples': 10000, 'score': 73898.40845513344, 'total_duration': 74920.21888542175, 'accumulated_submission_time': 73898.40845513344, 'accumulated_eval_time': 1008.4809603691101, 'accumulated_logging_time': 8.650696277618408} +I0901 06:06:39.726945 140462207792896 logging_writer.py:48] [229913] accumulated_eval_time=1008.48, accumulated_logging_time=8.6507, accumulated_submission_time=73898.4, global_step=229913, preemption_count=0, score=73898.4, test/accuracy=0.6279000043869019, test/loss=1.8746535778045654, test/num_examples=10000, total_duration=74920.2, train/accuracy=0.9563336968421936, train/loss=0.15038539469242096, validation/accuracy=0.7569599747657776, validation/loss=1.0823314189910889, validation/num_examples=50000 +I0901 06:07:03.944028 140462291654400 logging_writer.py:48] [230000] global_step=230000, grad_norm=5.459832668304443, loss=0.9483901262283325 +I0901 06:07:36.465957 140462207792896 logging_writer.py:48] [230100] global_step=230100, grad_norm=5.554044246673584, loss=1.064583659172058 +I0901 06:08:08.314805 140462291654400 logging_writer.py:48] [230200] global_step=230200, grad_norm=5.434643268585205, loss=0.9628562927246094 +I0901 06:08:41.863646 140462207792896 logging_writer.py:48] [230300] global_step=230300, grad_norm=5.760921955108643, loss=0.99224454164505 +I0901 06:09:16.394294 140462291654400 logging_writer.py:48] [230400] global_step=230400, grad_norm=5.066411972045898, loss=0.9287362694740295 +I0901 06:09:57.866452 140462207792896 logging_writer.py:48] [230500] global_step=230500, grad_norm=5.19964075088501, loss=0.8910441994667053 +I0901 06:10:40.252737 140462291654400 logging_writer.py:48] [230600] global_step=230600, grad_norm=4.973382949829102, loss=0.9153250455856323 +I0901 06:11:15.748807 140462207792896 logging_writer.py:48] [230700] global_step=230700, grad_norm=5.397465705871582, loss=0.928959846496582 +I0901 06:11:49.779712 140462291654400 logging_writer.py:48] [230800] global_step=230800, grad_norm=5.235268592834473, loss=0.8436540365219116 +I0901 06:12:24.084454 140462207792896 logging_writer.py:48] [230900] global_step=230900, grad_norm=5.045942306518555, loss=0.8408966064453125 +I0901 06:12:58.217238 140462291654400 logging_writer.py:48] [231000] global_step=231000, grad_norm=5.317002296447754, loss=0.9520221948623657 +I0901 06:13:32.210134 140462207792896 logging_writer.py:48] [231100] global_step=231100, grad_norm=4.971717834472656, loss=0.8423595428466797 +I0901 06:14:05.825866 140462291654400 logging_writer.py:48] [231200] global_step=231200, grad_norm=5.007597923278809, loss=0.876185953617096 +I0901 06:14:40.362250 140462207792896 logging_writer.py:48] [231300] global_step=231300, grad_norm=5.145472049713135, loss=0.883564829826355 +I0901 06:15:14.985759 140462291654400 logging_writer.py:48] [231400] global_step=231400, grad_norm=5.318439960479736, loss=0.9386435151100159 +I0901 06:15:50.658668 140462207792896 logging_writer.py:48] [231500] global_step=231500, grad_norm=5.041165828704834, loss=0.8917028903961182 +I0901 06:16:25.464000 140462291654400 logging_writer.py:48] [231600] global_step=231600, grad_norm=5.321689605712891, loss=0.9708263278007507 +I0901 06:16:59.953503 140462207792896 logging_writer.py:48] [231700] global_step=231700, grad_norm=5.391125202178955, loss=0.9222193360328674 +I0901 06:17:34.186655 140462291654400 logging_writer.py:48] [231800] global_step=231800, grad_norm=5.286707878112793, loss=0.8687229752540588 +I0901 06:18:08.346148 140462207792896 logging_writer.py:48] [231900] global_step=231900, grad_norm=5.61875057220459, loss=1.0255335569381714 +I0901 06:18:42.489746 140462291654400 logging_writer.py:48] [232000] global_step=232000, grad_norm=5.322105407714844, loss=0.9452208280563354 +I0901 06:19:16.206290 140462207792896 logging_writer.py:48] [232100] global_step=232100, grad_norm=5.126627445220947, loss=0.8695262670516968 +I0901 06:19:50.194796 140462291654400 logging_writer.py:48] [232200] global_step=232200, grad_norm=5.071938514709473, loss=0.9554334282875061 +I0901 06:20:23.668804 140462207792896 logging_writer.py:48] [232300] global_step=232300, grad_norm=4.950828552246094, loss=0.8934405446052551 +I0901 06:20:57.357184 140462291654400 logging_writer.py:48] [232400] global_step=232400, grad_norm=5.219566345214844, loss=0.9541730284690857 +I0901 06:21:31.899976 140462207792896 logging_writer.py:48] [232500] global_step=232500, grad_norm=5.185556411743164, loss=0.8340054750442505 +I0901 06:22:05.764685 140462291654400 logging_writer.py:48] [232600] global_step=232600, grad_norm=5.234761714935303, loss=0.8852244019508362 +I0901 06:22:40.875171 140462207792896 logging_writer.py:48] [232700] global_step=232700, grad_norm=4.900444507598877, loss=0.8491372466087341 +I0901 06:23:15.682220 140462291654400 logging_writer.py:48] [232800] global_step=232800, grad_norm=5.341954231262207, loss=0.9503355026245117 +I0901 06:23:49.018763 140462207792896 logging_writer.py:48] [232900] global_step=232900, grad_norm=5.359918594360352, loss=0.9175283908843994 +I0901 06:24:22.808683 140462291654400 logging_writer.py:48] [233000] global_step=233000, grad_norm=5.150016784667969, loss=0.8865371346473694 +I0901 06:24:56.611415 140462207792896 logging_writer.py:48] [233100] global_step=233100, grad_norm=5.055834770202637, loss=0.8365283012390137 +I0901 06:25:30.650639 140462291654400 logging_writer.py:48] [233200] global_step=233200, grad_norm=4.764499664306641, loss=0.8884632587432861 +I0901 06:26:04.297099 140462207792896 logging_writer.py:48] [233300] global_step=233300, grad_norm=5.343118667602539, loss=0.9499721527099609 +I0901 06:26:38.170112 140462291654400 logging_writer.py:48] [233400] global_step=233400, grad_norm=5.140997409820557, loss=0.9434447288513184 +I0901 06:27:11.934696 140462207792896 logging_writer.py:48] [233500] global_step=233500, grad_norm=5.231795787811279, loss=0.8619582056999207 +I0901 06:27:46.066897 140462291654400 logging_writer.py:48] [233600] global_step=233600, grad_norm=5.530186176300049, loss=1.019243836402893 +I0901 06:28:20.607237 140462207792896 logging_writer.py:48] [233700] global_step=233700, grad_norm=4.962613582611084, loss=0.8726866841316223 +I0901 06:28:55.345661 140462291654400 logging_writer.py:48] [233800] global_step=233800, grad_norm=4.918079376220703, loss=0.8924816846847534 +I0901 06:29:30.416050 140462207792896 logging_writer.py:48] [233900] global_step=233900, grad_norm=5.269857406616211, loss=0.9247349500656128 +I0901 06:30:05.587347 140462291654400 logging_writer.py:48] [234000] global_step=234000, grad_norm=5.103051662445068, loss=0.8895589113235474 +I0901 06:30:40.916930 140462207792896 logging_writer.py:48] [234100] global_step=234100, grad_norm=5.452964782714844, loss=0.970308780670166 +I0901 06:31:16.326498 140462291654400 logging_writer.py:48] [234200] global_step=234200, grad_norm=5.057795524597168, loss=0.869415283203125 +I0901 06:31:50.335178 140462207792896 logging_writer.py:48] [234300] global_step=234300, grad_norm=5.216480731964111, loss=0.8028004169464111 +I0901 06:32:24.771865 140462291654400 logging_writer.py:48] [234400] global_step=234400, grad_norm=5.082373142242432, loss=0.874778151512146 +I0901 06:32:58.558581 140462207792896 logging_writer.py:48] [234500] global_step=234500, grad_norm=5.162299633026123, loss=0.9577593803405762 +I0901 06:33:31.805010 140462291654400 logging_writer.py:48] [234600] global_step=234600, grad_norm=5.029625415802002, loss=0.909552812576294 +I0901 06:34:06.029896 140462207792896 logging_writer.py:48] [234700] global_step=234700, grad_norm=5.27858829498291, loss=0.8703659772872925 +I0901 06:34:39.899177 140462291654400 logging_writer.py:48] [234800] global_step=234800, grad_norm=5.384548187255859, loss=1.0069931745529175 +I0901 06:35:14.091834 140462207792896 logging_writer.py:48] [234900] global_step=234900, grad_norm=5.486851215362549, loss=0.9414200782775879 +I0901 06:35:48.529762 140462291654400 logging_writer.py:48] [235000] global_step=235000, grad_norm=5.075716972351074, loss=0.8889106512069702 +I0901 06:36:23.245680 140462207792896 logging_writer.py:48] [235100] global_step=235100, grad_norm=5.547665119171143, loss=0.9955681562423706 +I0901 06:36:59.661284 140462291654400 logging_writer.py:48] [235200] global_step=235200, grad_norm=5.100362777709961, loss=0.8322532773017883 +I0901 06:37:34.610775 140462207792896 logging_writer.py:48] [235300] global_step=235300, grad_norm=5.035505294799805, loss=0.9150928854942322 +I0901 06:38:08.772387 140462291654400 logging_writer.py:48] [235400] global_step=235400, grad_norm=5.3749613761901855, loss=0.9461227655410767 +I0901 06:38:42.734552 140462207792896 logging_writer.py:48] [235500] global_step=235500, grad_norm=5.312295436859131, loss=0.9691227674484253 +I0901 06:39:16.345564 140462291654400 logging_writer.py:48] [235600] global_step=235600, grad_norm=4.978498935699463, loss=0.842284083366394 +I0901 06:39:51.334073 140462207792896 logging_writer.py:48] [235700] global_step=235700, grad_norm=5.330586910247803, loss=0.9574244022369385 +I0901 06:39:55.305958 140462291654400 logging_writer.py:48] [235713] global_step=235713, preemption_count=0, score=75893.8 +I0901 06:39:56.236197 140659750036672 submission_runner.py:857] Final imagenet_resnet score: 75893.77206802368 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/eval_measurements.csv b/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/eval_measurements.csv index 0f7c9a685..b464fa0f5 100644 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/eval_measurements.csv +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/eval_measurements.csv @@ -1,39 +1,39 @@ accumulated_eval_time,accumulated_logging_time,accumulated_submission_time,global_step,preemption_count,score,test/accuracy,test/loss,test/num_examples,total_duration,train/accuracy,train/loss,validation/accuracy,validation/loss,validation/num_examples -88.55480742454529,0.0,60.91907024383545,1,0,60.91907024383545,0.0012,6.9119825,10000,149.47478890419006,0.0018734055,6.911738,0.0012599999,6.9117756,50000 -143.2671549320221,0.0286586284637451,2058.181893348694,3667,0,2058.181893348694,0.029900001,6.208869,10000,2201.5248374938965,0.043088328,5.8607073,0.04014,5.9290338,50000 -181.96476244926453,0.06117844581604,4054.160504579544,8555,0,4054.160504579544,0.013300001,7.0473847,10000,4236.279563426971,0.017498406,6.8538775,0.01714,6.8836646,50000 -289.51229429244995,0.0906753540039062,6050.281397342682,13898,0,6050.281397342682,0.0048,7.8949347,10000,6340.023873090744,0.0054807076,7.792888,0.0055799996,7.8234067,50000 -418.1712100505829,0.1234092712402343,8046.279819965363,19241,0,8046.279819965363,0.0022,8.61545,10000,8464.760076522827,0.0026108099,8.562036,0.00278,8.59723,50000 -517.5279848575592,0.1553406715393066,10042.387595891953,23383,0,10042.387595891953,0.0016000001,9.104018,10000,10560.322937726974,0.0017338967,9.085137,0.00208,9.102534,50000 -661.8034045696259,0.1875710487365722,12038.792675971985,28732,0,12038.792675971985,0.0017000001,9.287293,10000,12701.082419395449,0.001694037,9.291868,0.00156,9.292005,50000 -768.2695868015289,0.2213890552520752,14035.339556455612,33826,0,14035.339556455612,0.0016000001,9.391503,10000,14804.176125764849,0.00129544,9.402161,0.0014,9.399457,50000 -870.2924361228943,0.2536473274230957,16031.727242708206,38401,0,16031.727242708206,0.0014000001,9.444064,10000,16902.664816856384,0.0014947385,9.477643,0.0012599999,9.454873,50000 -980.1608846187592,0.2882544994354248,18027.7465133667,42791,0,18027.7465133667,0.0017000001,9.3738985,10000,19008.6331281662,0.0011957908,9.3953495,0.00122,9.384472,50000 -1123.8164768218994,0.4459917545318603,20023.56044578552,47116,0,20023.56044578552,0.0012,9.251919,10000,21148.31163167953,0.0011160714,9.273086,0.00122,9.26029,50000 -1202.174001455307,0.4815821647644043,22019.91549110413,51274,0,22019.91549110413,0.0011,9.064775,10000,23223.106118679047,0.0010562818,9.120835,0.0011999999,9.070607,50000 -1323.4103920459747,0.5189347267150879,24017.79511475563,55564,0,24017.79511475563,0.0011,8.862023,10000,25342.307918787003,0.0012356505,8.889726,0.00114,8.86706,50000 -1424.8949935436249,0.5556581020355225,26013.837624788284,59800,0,26013.837624788284,0.0012,8.709208,10000,27439.91783833504,0.0011559312,8.759222,0.00118,8.7175255,50000 -1521.804844379425,0.5917441844940186,28011.93083548546,63964,0,28011.93083548546,0.0011,8.567521,10000,29535.00321173668,0.0011160714,8.594898,0.00118,8.576055,50000 -1636.7673811912537,0.634131669998169,30007.89012002945,68219,0,30007.89012002945,0.0013000001,8.452573,10000,31646.01367497444,0.0010961415,8.4948015,0.00116,8.460207,50000 -1711.0713665485382,0.6724264621734619,32003.9418694973,72875,0,32003.9418694973,0.0016000001,8.336133,10000,33716.45394611359,0.0011160714,8.365174,0.00116,8.342442,50000 -1806.4985573291776,0.7217459678649902,33999.91070771217,77093,0,33999.91070771217,0.0011,8.239175,10000,35807.94622397423,0.0014947385,8.261976,0.0012599999,8.243793,50000 -1909.5075569152832,0.7861449718475342,35995.8516535759,81399,0,35995.8516535759,0.0011,8.141392,10000,37907.00846672058,0.0011559312,8.150389,0.00128,8.144362,50000 -2001.8195931911469,0.8286004066467285,37991.905416965485,85799,0,37991.905416965485,0.001,8.047056,10000,39995.46316099167,0.00129544,8.063519,0.00134,8.04747,50000 -2100.913553237915,0.8685033321380615,39987.95483374596,90100,0,39987.95483374596,0.0011,7.9610558,10000,42090.69328832626,0.0013153699,7.9802356,0.0014,7.960504,50000 -2202.9060847759247,0.9071593284606934,41983.89597201347,95129,0,41983.89597201347,0.00090000004,7.8850675,10000,44188.71274614334,0.0010961415,7.89192,0.00134,7.8832655,50000 -2300.74538230896,0.9539778232574464,43981.064588308334,99394,0,43981.064588308334,0.0011,7.8307037,10000,46283.816162109375,0.0009765625,7.833911,0.00114,7.8286967,50000 -2383.4762213230133,1.4857048988342283,45976.930940151215,103708,0,45976.930940151215,0.0011,7.772891,10000,48363.140311956406,0.001694037,7.7662373,0.00136,7.7699013,50000 -2421.472584486008,1.7976596355438232,47972.51910805702,107550,0,47972.51910805702,0.001,7.727155,10000,50397.32648801804,0.0014548788,7.7392006,0.00136,7.724337,50000 -2454.848158359528,2.6185977458953857,49967.55086445808,111343,0,49967.55086445808,0.001,7.677686,10000,52426.83486223221,0.0014150192,7.680137,0.00136,7.674657,50000 -2498.7893834114075,2.9784514904022217,51963.30955648422,114877,0,51963.30955648422,0.001,7.640008,10000,54467.09082698822,0.0013751594,7.636404,0.00144,7.6376934,50000 -2531.648089885712,3.3708062171936035,53959.05523252487,118552,0,53959.05523252487,0.001,7.600594,10000,56496.327038526535,0.0013352997,7.5854464,0.0013799999,7.598124,50000 -2566.183147907257,3.765636682510376,55954.80316567421,122227,0,55954.80316567421,0.0011,7.5624433,10000,58527.34854841232,0.0015744579,7.5687747,0.00144,7.560353,50000 -2639.1051247119904,4.160260915756226,57950.92701792717,125812,0,57950.92701792717,0.0013000001,7.5200286,10000,60597.006568431854,0.0016541772,7.515965,0.0015799999,7.518538,50000 -2666.0562329292297,4.544161796569824,59946.77712726593,129361,0,59946.77712726593,0.0012,7.48373,10000,62620.444516420364,0.0017338967,7.4779243,0.0016399999,7.4817243,50000 -2721.3663465976715,4.906957149505615,61942.77612376213,133006,0,61942.77612376213,0.0013000001,7.444616,10000,64672.361290216446,0.0013153699,7.440227,0.0015799999,7.4432135,50000 -2758.2186217308044,5.240548133850098,63938.894350767136,136687,0,63938.894350767136,0.0013000001,7.4008193,10000,66705.86566662788,0.0017936862,7.3960743,0.00172,7.398652,50000 -2786.5239927768707,5.537599086761475,65934.44663023949,140156,0,65934.44663023949,0.0012,7.357977,10000,68730.21747016907,0.0019132653,7.353606,0.00188,7.3550563,50000 -2837.938948869705,5.867739677429199,67931.0115404129,143098,0,67931.0115404129,0.0013000001,7.324862,10000,70778.71036839485,0.0016342474,7.3302083,0.00194,7.3216434,50000 -2886.9578301906586,6.165061712265015,69928.60928463936,145779,0,69928.60928463936,0.0012,7.291023,10000,72825.79156255722,0.0020328443,7.2840333,0.0020599999,7.286662,50000 -2939.664833545685,6.508132457733154,71924.17374968529,148284,0,71924.17374968529,0.0011,7.262055,10000,74874.6106865406,0.0015943877,7.25432,0.0021,7.256729,50000 -3025.88085770607,6.818323612213135,73921.95043730736,150291,0,73921.95043730736,0.0014000001,7.246727,10000,76959.10509991646,0.0019929847,7.249264,0.00208,7.241174,50000 +63.72952103614807,0.0,53.1759774684906,1,0,53.1759774684906,0.0012,6.9119825,10000,116.90558505058289,0.0018734055,6.9117374,0.0012599999,6.9117756,50000 +85.47269701957703,0.0403082370758056,2049.223411083221,6572,0,2049.223411083221,0.44230002,2.5904195,10000,2134.791506290436,0.6306003,1.5325073,0.57357997,1.8282131,50000 +108.08916926383972,0.1185705661773681,4045.282567501068,13916,0,4045.282567501068,0.53010005,2.1301467,10000,4153.60573720932,0.7621771,0.9435853,0.66736,1.3725033,50000 +129.41019129753113,0.2125744819641113,6041.208413362503,21259,0,6041.208413362503,0.5655,1.971104,10000,6171.006928682327,0.8075972,0.7395382,0.69585997,1.2365702,50000 +150.83281111717224,0.3011040687561035,8037.070692062378,28602,0,8037.070692062378,0.5809,1.8867964,10000,8188.448412895203,0.83296794,0.6281649,0.7123,1.1660333,50000 +172.41977453231812,0.3790128231048584,10033.056078910828,35946,0,10033.056078910828,0.59190005,1.8422476,10000,10206.169721841812,0.85487086,0.53780156,0.72341996,1.1221741,50000 +192.3741476535797,0.4674093723297119,12028.914838790894,43289,0,12028.914838790894,0.60010004,1.833945,10000,12222.131848573685,0.87282765,0.4651624,0.73126,1.0990901,50000 +212.51364135742188,0.5555286407470703,14024.841076850891,50633,0,14024.841076850891,0.6078,1.8267647,10000,14238.34527349472,0.88392854,0.41763052,0.73604,1.089328,50000 +231.5355188846588,0.6126320362091064,16020.97186779976,57978,0,16020.97186779976,0.60950005,1.8237377,10000,16253.614742040634,0.8929368,0.38473663,0.73706,1.0853631,50000 +250.69002890586853,0.701411247253418,18016.97361254692,65322,0,18016.97361254692,0.61170006,1.827934,10000,18268.919723272324,0.9012675,0.35587227,0.74094,1.0815302,50000 +269.40062832832336,0.8057889938354492,20012.83114266396,72666,0,20012.83114266396,0.6153,1.8251016,10000,20283.65297317505,0.9071468,0.33349392,0.74214,1.0764859,50000 +287.4291217327118,0.8914148807525635,22008.97481846809,80011,0,22008.97481846809,0.6151,1.8198733,10000,22297.971310138702,0.9135244,0.30929574,0.74368,1.0724281,50000 +306.2893109321594,0.957449436187744,24004.91706252098,87355,0,24004.91706252098,0.61770004,1.810064,10000,24312.900218486782,0.91770965,0.29734606,0.74516,1.0671043,50000 +323.49813055992126,1.0380325317382812,26000.928428173065,94698,0,26000.928428173065,0.62390006,1.803928,10000,26326.260948896408,0.92386794,0.27556,0.74774,1.0638627,50000 +340.46144580841064,1.1386680603027344,27996.96565246582,102041,0,27996.96565246582,0.62560004,1.8158712,10000,28339.422913074493,0.92610013,0.263924,0.74884,1.0614623,50000 +356.9961848258972,1.2308220863342283,29992.987577676773,109385,0,29992.987577676773,0.6247,1.8259393,10000,30352.131969690323,0.92875075,0.25364253,0.74917996,1.0621685,50000 +373.45445227622986,1.3163821697235107,31988.944032669067,116730,0,31988.944032669067,0.62320006,1.8310074,10000,32364.69255375862,0.9323382,0.23542856,0.75158,1.0637244,50000 +389.9751641750336,1.4011025428771973,33984.88090109825,124073,0,33984.88090109825,0.6204,1.841354,10000,34377.29505729675,0.9353077,0.22499812,0.75262,1.0646015,50000 +406.9744417667389,1.5179433822631836,35980.86967778206,131416,0,35980.86967778206,0.6225,1.8407513,10000,36390.460955142975,0.93769926,0.21849291,0.75236,1.0655642,50000 +424.9661509990692,1.5933632850646973,37976.824387311935,138762,0,37976.824387311935,0.6244,1.843589,10000,38404.54493522644,0.938636,0.20914964,0.75232,1.0671511,50000 +441.8562512397766,1.663456678390503,39972.9033203125,146109,0,39972.9033203125,0.6243,1.8404001,10000,40417.645072460175,0.9416255,0.20841375,0.75446,1.0689094,50000 +460.0173342227936,1.7309489250183103,41968.78911828995,153453,0,41968.78911828995,0.6244,1.8389771,10000,42431.820268154144,0.9456513,0.19136712,0.75391996,1.0708171,50000 +478.41012358665466,1.80169677734375,43964.716457128525,160771,0,43964.716457128525,0.62450004,1.8409966,10000,44446.27172589302,0.94876033,0.18166068,0.75387996,1.07235,50000 +497.6320929527282,1.8793811798095703,45960.70112752914,167832,0,45960.70112752914,0.6258,1.850111,10000,46461.61534452438,0.9493782,0.18224207,0.75308,1.0753248,50000 +536.1628928184509,2.010361909866333,47956.73145079613,172536,0,47956.73145079613,0.62750006,1.8528265,10000,48496.35363531113,0.9483219,0.18042949,0.75468,1.0734031,50000 +573.5259537696838,2.5825936794281006,49952.29267835617,177030,0,49952.29267835617,0.62630004,1.8531942,10000,50530.0918970108,0.94915897,0.1775648,0.7549,1.0738461,50000 +613.3243944644928,3.054151058197021,51947.68952417374,181488,0,51947.68952417374,0.6247,1.8563241,10000,52565.99465203285,0.9496771,0.17702009,0.7542,1.0756431,50000 +653.6150543689728,3.488328456878662,53943.41852045059,186120,0,53943.41852045059,0.62640005,1.8612959,10000,54602.69437932968,0.9524673,0.16976719,0.75492,1.0755513,50000 +690.2438578605652,4.017447710037232,55939.05248832703,190276,0,55939.05248832703,0.6272,1.8644032,10000,56635.74251246452,0.9514509,0.17048396,0.75464,1.0741526,50000 +732.8901488780975,4.518044471740723,57934.46646571159,194116,0,57934.46646571159,0.62700003,1.8665876,10000,58674.5222234726,0.95322466,0.16543461,0.75505996,1.0765518,50000 +765.902948141098,5.073758840560913,59929.77295255661,197954,0,59929.77295255661,0.62600005,1.8628919,10000,60703.64590597153,0.95310503,0.1635923,0.75564,1.0758622,50000 +795.0919539928436,5.548985481262207,61925.07143211365,202078,0,61925.07143211365,0.62560004,1.8653016,10000,62728.847462654114,0.95360327,0.1622379,0.75534,1.0767275,50000 +834.8467264175415,6.078795433044434,63920.629023075104,205873,0,63920.629023075104,0.6251,1.8661743,10000,64764.93791365624,0.95533717,0.15522939,0.75549996,1.0767725,50000 +876.8376929759979,6.746559143066406,65916.38125610352,209850,0,65916.38125610352,0.62710005,1.8678958,10000,66803.59570097923,0.9548788,0.15816157,0.7553,1.0794655,50000 +920.5079526901244,7.185899257659912,67912.0241098404,213955,0,67912.0241098404,0.6265,1.8719171,10000,68843.66978883743,0.9560945,0.15828097,0.75492,1.0802552,50000 +944.3057725429536,7.68643307685852,69907.44102859497,218440,0,69907.44102859497,0.6284,1.8729472,10000,70863.64268183708,0.95657283,0.15558098,0.75588,1.08051,50000 +987.6376729011536,8.14183759689331,71903.08286333084,223900,0,71903.08286333084,0.6279,1.8756953,10000,72903.2956624031,0.9569515,0.15209584,0.75641996,1.0811065,50000 +1008.4809603691101,8.650696277618408,73898.40845513344,229913,0,73898.40845513344,0.6279,1.8746536,10000,74920.21888542175,0.9563337,0.1503854,0.75696,1.0823314,50000 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/events.out.tfevents.1778615207.4ea6dc0cbfa9.13.0.v2 b/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/events.out.tfevents.1778615207.4ea6dc0cbfa9.13.0.v2 deleted file mode 100644 index f87d0423edef07e331d6200ceb2806a5e564deb3..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 352423 zcmbTLuJZT=FE{HQ>IW9WeSB58kC__Dr86@gb<2Q2~h|kN|LBZg(L}4 z5e=dY5yJb}ug6-yb=JCv=l<7y{XYA9uC<4=_jPu|<^KP_N13`sigI!ZCHbC@^;*`q za<*LP(92+%jf3luzEe6Wb#S$DaB+0DwRc?J%i7V|rq@!}h4#H{e*L@4+~v*-U2U9G zg1mS1_?`Cgbh?3JYqzwjUuoUjE45Qnuy=HE8L3oPzOh^f1^Z45y13dn{Ws#*{|pgohjr2h{f;=Y!hO1;xvy@+g8yuP zBl+%{O5Kzixc>S~dRtmsFLk!GUh&_YhTAHKKu&`qh2H&t=gjG+YOd&b;7?9hrOrx9 z|DQvD9+f?q&trWJpKrr{XIL1$D^u(k`X@t6sh3idWtR2}=UKWgbd=p7|1L*ygK8c= zMXi>P49P*-0|xt6(?T!6VzX+uaYVi z`tR~-f68SflpmX=a}AbZ@smI0|Gx|! zUH`L`=ly39{%Ze!b9`1b8wELIqmTF+{N9Y$uNFiqMn3emD=E#WOp_6TvSt{Nu{S(?(NlU329r~Ytw}-e!tiN9(=T~y||W5Dmk?Bv<66l1^q zNfaLmJV!vtZaZbqA-_W!`@KhO-y=a!B~7J9PR=$qi=BRb{Bx}xmpZuqw+S+v zQ=pS`zt``i;^f zqW$}WG@kBLt{A9AGQ1_BuGRJS;%O|TL`IZ(xnO@9qmXtAJX7Z z6++%y9}fXYj0?#SL-rtKZ0yBl0C{^ReuM<^ttkshP)SBefx2T3K#s+daz3PWy<~)R zaT+2AkcZbvh8U89khjZ6=m4a%uir!o@{t3%=aPhw*r6ug0MR=@%J~qF+lrW>yi#o= zK=O`|3^C*@LQ3Kr9|Oqp&+;=QNREPRM=IYW;6`j1yDt|YyX9;ADd$6u?qwj^-if9F znXW`K#E|CtaPaGVe)MvHR6X5fEkWXH$wIQ;Uq#5~F4YYI(zlqD^C3NtcEOAkKjUhE z+^irOVn}a z#1Kn_OzT{6EcunmS62wnQ7NV4Y88>3$v=-|Cvf>me=R^EAZo&-JR(fFoF?vWcgb>fe z2$|VRZ3RG{MR-I@kSY$O^yekqh zf)PMA4l2=>AW@BEA%#BCxDk)^(p~|i^+;0AhXlOy#tic^yJ&zE8j%bkM7O2^T*hw> z4*LX<6WK|lBuIdYEF`uGb3(KmxXlG1flZVDl=C6cE-w+HWmz#0AORCdh7jVZh>$l) z?ZyKn*dk$~1gYjgf@Yq@jktFzzA->_4w7;{pJAl~DBjtREsRi?O*+esC5I}x-kPIP2w>3iaPIgHE zNHZP(r4l5(sVtKOG2R|)}u%{A2NKt3PL>3n!g4}?QtYS2uaXD$dRAVqX9DX zrr9nDQo(`j=^23=k#uEOO@O@3Amw~Wr#%mFXFq(jwi`gs<&X>^M6NeN)+8>&+d--a zN`oXwR&&{oJa2XeAu8sBh5#fsl$7%!jlOlmj7GWU@c@a4AQ?i4?f`@gaVWn89qE-C zds2eLsLDc4IWtSEiMiuffOy>}<$TCw+f@kh8ahKAATCcxh7jU86d}{HzP|-XgNajP zB}kx}ETl#^24Wp}3%|2BGb81ENYI?#2&ta!rV5Zg79>LmNzg;coSR0C0rI@C#Z3v~ z*+Lc)>d1VTxNA_X1szFl)$&g{AELQwIYP!6wpIp+=@gP7gvc2p#Q4*rz5w~EI66~; z6mlRxlo?1|{2zm2$tpPwTgg<>IK~h@E zcI3{gaNLN4(}(u~$nWvLc<7^C7p*G7u8I)cF%Y^s-2X5RxzxAy;pYJpqvD82zsj#J9C9 zWKx~uxDox`+bjggj_ahH59zySC}xy*(z*eVFs2$8cyh;2rbYw+y;RbBl?FMfU1 zz2Ew;rV^~(wT8j>fsU~njp0Gyq9!CrFqTldh51%+$$665@ZV42q=6CBRwU%_<`Nz* z?1Y3o9B<}Hp}s;B!qRYUcozv6~DB=zqH6z1tS#HV^27MMAV`ki-nL+21h9~+YQ zk0um6lS^?$#IzV*4#$NudlD%;E@;_cAJz|8m<01GOI%6Hznf8rs;~bBj+IT9?^l%$ zF&{a{%0?%S;_94hcf1v>aQm~QoPVr*)20_`OAgSk)cjw zCP|QD4y5BtrXw4&2R#DF{uQL04=H+PfFDLP?*SbEGG#5v5JGg>uw&m0 z36kDkwj=TVnU1Wg9fr>Ug7=YfKBQ?UIfPvMaBVO^jvgWzLWspigtQ)3zX3q9o#d<~ zNMr|DNbrXc+=yDM7hZsld|N@v`H+peEihw(WsfL;G+0M6gb>fo2)X}x=zf5--R58~ zLHs((LK4pf<32L>l+Hlf=x)P#i<)mpIUlmomYEXUJI}i&Ksr{E3?U?82SW0-7jFVc4No}_36k4M zwj%{D2l3Oq?U)b-9r4&q%J~raz08<`fnCTPfb88#GK7%)T?nzhHM&1QvTQbaOOS-l zvXJEOp$Hk~?R*y?mmEnsA9AYoM$EXDbfrH)o-HRCLWtZRgnake{Tv_x4p}=TNN5*X z$c}c*wN{I|6XpXX#*LKoApud$Est;Szgq$%$CG3TA?o`O5?0@99zeV+wGT)T@2;|t zWB$zHw{+L;YJfccLCX0MlOxQN%h7379{_T%_P{@zAcW`!BBbfsgFOIpx_fA-1S#V{ zOih`?@Aa>B1_Gq4KPl%!?#pk$rS)}s1Wt*?>5vQ|#Nr@A=3aX04v=B)eIg}DhK6iM z=B{Nr67nSHAwUc_k#aty$&UR9X{K)b20F5G2gwjZJP#wpuf+tMHvX#ID_(*`b(4im z`O3`P`PXcR`$$MLQqG4Y)@g*0>Zdq$q+(9jv}PS`ezNHBg^_fxFtaXy30aZ z4rM;gzE?A+K}XsQC*^#I+g0Yw?iT+LM?zIclMEpwAsitEyIk?1xc#}#_a#U*2l6T^ z2tQ3lzfETVaypKb^C6R;*2j$7mz>8!MsI@7AfaLdW>F-ko*g`_W@+>J(3}W=w3jGPsNBpfb5MpUnN0&ddWhpT$z2O z@_}XpfM{GK<$TCbJ?4wmx-EU10HpE?$q+&;Vi3~lWcg@-_;*vUQTXeQa|H*IY0Vsd zZSEBL0;G38QqG6mJztHFVAtD*tpmu2p(H~H@w|+X-ZS0tgy@>JP+5Xx^_K0(fo!HD z9jtTk5p2a%QqG5TDj9}L>w@N@rvQ2Wl4J-WA@K;w_L%Do9hp@pxv2z+(UOH|+A`lq z3@w`Wg^uiNM9TS)mun{>Wcn=g#Q<5aMlytuglh;{cWh)EfLsWR=_Ek{`^Z9)dNE5Y z$=H7@KuXV&az3Q~TNi{3b}7FIko-#|LkP)FLP(!gD^3HX)I3p3f_V0og}82E&g>Ur zN8z(%Ut3bnhrF1qfshONmgxX#>qs($5V;hDbf3B&Pl&s>JRT}R3ONwx*UaI!mWja| z=tzqjq?`{qkhC5lW?rgC05azT$q+)+(-2bE)ush>}3Ke=A0wmj%WC$U;cMx*y@QdyMxqGFtu>^_eFAM2@dLM4Y!w1(s z03`e-Dd$6e*v`U?EUkBI0J8rs$q+&;?jdAk&{+?FSk1RGlOVnWWFhwtF+2OWNm>s8 z^1_Lf^C6XE#vmlg-%uMMo83r;5aOAIkP%aN^a035^{18+ zQqG60f6UxP>D+FKEkK_1AsIqQ$U}s5J7RqlAcO07x04__+Oi!{tbYVQ&8$_=w*%yY zDkIQ)hfHJD@wA#w!>(VD$>EONyavlcTD@T)aoZ!814JQ> zl=C5N8ZcjV?LN%GakuxGBtr;MFGNVOL0~t4%(%A6SArCCAmj9y!|&_Uj%5IGPb1}g z$nhVp*pYgted@wKGWR~o5JGf|5VE!7(wtBb(=t3?U@p146tNRPgcirE6_(NRZrNvK@&qXF3wv#r-%yI;W9xKIBpY zb2r{$QL7mMQO_b7LP-86gskp4xfyiCqv^uC5+q@`EM%avA1;eSs*@c7Qr?D?^C8L$ znA<^K16tuA`j>l6c$^gqqE7LT7i5N2ENjwgxY+13%oGtg+%_~sW_Nv};wjhOZwQpduY& z0!GOmD{n+HXRp$Ow-&()e|CbD^N*Fq8O$W*`Fq3t;8^+S9LW%VCKk2vGdWs(5NBH_ zxkR^=Ak`d*#YpC$zwgZsi=bx&k3Qn(2-+r zK6jKL`TDXQIX8@fY`NchF+du8Cgps{oz%5>%o*}+1wMkE^pj)=A)bl|aZ>U~2guyA z;XNftl7TGbboox)2%`!Ee09pZN>a{;I3=lKM#$NEHDGBC{XsH>kVeV~i8r}|^A;L` zXS5|q_-I+kgbjOfBd#@e$JavKI7-U-kRjHFn6c=K`A~oupC%bXNJu?|%%62F0y?6( z;mZgK;xk4T@^}cdvn#aoz%LIrJ4iVnVyMTwgy3cVQFnmY?Pi$u#Y5;A?19?iMq^(;nvV&6hMlNNrn)T&=4UNO}o4UNUg-^ zDH0^hP_`qJa+!T(&GkMt0Ag{Ql=C6A)VAZ&QrkNYhjiy>lMEqbk_tiwud0I&GxDFv znM;ruBUwoJS_V?m?j8QZFY_fS=R^FInCE1RXXw3vjzqs98A3>YV}x`b-QWy#>;{cFK2vj&zY5|EHV}*{qs~9jU0A@D3oxbxDR0;?xu&n~a~E14QM= zNe2nyIZhVRd5F*zcA(tm#!?Eds zc{SZ6NFfK(^v(|aG}m;y^$a=^Sxn0LkaZujFhi}Q+70N);CCcL2=P@#h@oQMW#~vB zr3q^#NXmHGj~2Legb?)>2(dT0Sp<++ zyTmOLBw~UrWTPB&6H0DatR{41=~7b8hxBP)h>+5(2`d5O=}t0)kjRz@vHMvv86es6 zHT@)r??hRM8Vm90Wxp37PnAhIAM#>7^YpWWjhQ(>f|`&FA%uOUvYoGH6M)>)9UUY= zK5`%*?lR|qc#oU*07)B5%J~pgGiGO3pO=qg(|J=!h7gj`20JqDvmgFaT>i-QV-h6C zShgcNMa)g8jmKIl0%YSUQqG5%Id8?KWm$d?UsE^b0?80UEZQMNGjD8r=tzW*Vx$C# znu$( znJf!wXtfPL%|l1#Mgru1Z&J>O)Y{4%`P3Df;SZw12ayaR#IplJ?)eSV0my<`Dc2>4 z*A!Vu2Q}u*-feWmXMkkPBjtR^p&@VZ!^m4c8Q;p2XHPPOkVc&lQa5bzPk>}^tCuc8 ziaC(+)y&dbscz8`AjkHSaz5lzq60#*b7rT(QOouq$q+(9IwR!g^iS3RY4~9B0|}CD zBHNL`23v3=47WXu0*Lc7QqG6;-N1Ypvj;rL1<37JBtr<%>WYx?OONINBsV(knFNV6 zm4&zkGbhB!d0I{YDNG>ce8{Oz%rs`VQ`Pv!k)jlmA%rAoAS7;IDE?@0%NeC&3F0?Z z7Sf?M11Y~ZI}afKmq}r>~4)}fx%WEV<2$|F!A?8;$<0DvaLz8zBq>2M+ z5Xvlz8aMWu0Hk6HDd$5}J~D^jgp3+EB^EW4WC$Vong~(+v}rahtsTzyzDkhXX|f%u zvyC~KElW!C1BlvcQqG4s9x%a<^!-wA1?(d(UL->Zaq5YXsKZxU1Egt(<_fQVJ7pvf{;dUU(EzaZZDD{gsAsLNV6`NGNB`lC#tDSkTMRW zq9b!=A7ePC03h1aNjV?Vd-*wpd<$HSFRf@jmt+Vbk^K;I|3i)&K=#a-)lGtA%#`hj zek}%)S1WZ1Kn`1zaz3Owwml9Y>h}oGgN`h-BN;-7?f`_`-sf)ukQ>J?_mv<~vt%Jt zCNXFBH*2>Q0;FjqDd$6s^*dlkRvqiv5FnSMNrn)Tp^cEC{UY$^H)RuB>PV0Pb6JQ& zGIRKSUE!b#kQHX6oDV5l&isNhy3LOWfULG48A6D~AcR=;RDd2T3^}^6ubJTw10Ln%x6P)iIJGgj5Vh$eIRf697`Z<;6q^l0RFvBf1Y5 z$e{Y;p8_PokCgKvn=M!4(yBM8*))Kx4kQ^ui04p*Yzb?FcTxISb~ckBNpoZ&19g~v zi?Gs(LXKFiR#e_kPWsNrn*8Xc$6F+HdUz9r^5LZz(~-=gLA>FJNXR ztS4392gu}gq?`{KzIG0Fq<#a>kI<1xTS$fw5;7biuZNt)Kfh4ku5^(E@v)GFIB)mH zjX1n~VFWiM?DUNCgMt)sI

    >5>y$baAOVTa?h+)+Qnn+5jximP z59s0#kht5VoDcclf%%O@{~|y9V&(OKWC$UXMj^zo(+~%MEHx|JC_!SZWFf(Wm>)!S z>#cVgAO=>XoDZp}zYLd_?Rm{9&=H+QBtr$bMx~&WCh8%sgYcS=incI$q+*13=!h<^_>$yDj!xKmLP>3i24cU62;-f znF#|`KShv;Uy$P2=O&SNWWr-Du7t#=|o776dT#4Wxs}jY(4)I zUz`28n3VG&`_h)+(z@TNItw6)6(mClQ6Gm8ua7NT0_4odO&26cgsm(jcL=kO6!ckv zFKC!Lo|N+;>kaJ@vbk~SEP$Ark_;gvay&xn8VBtK$bz)5@e;&$zAR+VEe2wr7=%x( z&SjBuKBR5zT!eHpYm8&lvAHBe2+^I0ko3B{OaXFx*02-_@{t2E_h5F^c`f(j(h9O6 z<$TD7Ps~dyALyicLPu^aCK*CVhA~3wR}H!V5T6Mf?n#iG1+pD658RBO=9K5%m0)MT zJ)M;EA**AV2R&M~&2eW}wjdcoh{a@txGtNE@BX+mrOp!x61Pwm@@6G-(b&!Fbp}AD zo0D=rq_`H75qo-&0(4~JJdz=VR7^q0hw6JPp(CYdCl*SOkVUeP>900nM^xOkZvaH= zGAZXn4p}k7@L7ea*8!rENHT;FPg8`<^E;0(GIxyL^+tkt*~vmIQ<n90P%z>N>XD%8$HGVb|I`Z}o zDd$5T8!&_HP4lCRp(9z@Btr-ZnU0Xv6C19Dd0Q2^!j?t9ZkeXq|JD5Wy1nOz!mmni z+9q6p`R^u5BuFs-J$>#MY|eN8V?$v6yI6&U{N4PwL(WCqq=Y$JZDIb~rUgm+tNHI4 zGrr<>DGfW;4d%Z;tCOVvY5x231M5(j|DLZw62*5fvPQvVn8 z-+k0?V9??c+bb~teRM1d`ls{XGMhKy+JO-;|J{5tN&Wlz@7>pYasE5)SkJ>S|Gn3Y z@xsK(I*)~$Ru-wT*$mw z13I$x`=Egmq>2OS8OmI$7Pfs94-or0ga4HCAv5YQ@1%S^HJ}6_s~eIGAtZk`LOPtd zg75!6rs%FKL2?~sJ95^HfyC-9eF+fTCZwDXscoE&XMvXoXSkb75N~H$$cqRDGTJK#Z`9s=M9TS)>bDoLBk!B-!PnH~6_5-eM13AYj4w7> z01%61l{ONjj04%ai1`*|J?snqnQ|jfQqG5{+)F@6a?l36zd6vGWC$UVHVA2&n*InN zi|j@?N{|c}*^Z>nV7^nHs5=y&B|mc|<$TDAptA^xvRu3imR9;|k|BiX&PPaeWj{@T z>{Q(7CPAWHWg$k*m=j{BeWAs$k9==T%K4DkJK5c1=}oXY_D zq_@#eg5)og?Z}-&Oh+<5Z?6py?{TD@59y_T9hX+iqEcVzh>9u65JEf`Bjn>6Wix=d zY7_-Ykfi0Zki%z~!|(2$$7cW}z7HwqLz;D|jU5^K@;kn$W5-~UA%rw?M2LNGZGV82 z4;dLILBic+A)1qzaW_jn6}&{Lqfg5DkhWe^5K=UAf*EwA<9L!GgoHRD~NCgLCW6eN%jj+{#r4==tl=C6C=4IkOQnXm-ICNy&Xp$j>B)A~t?W;{!^JLv4qhN@V4NsLXshbOj?SN*n(ud`qg~u zmo7nKR>?wwZ!#Tm%y=0Kkk_wBIUnM@lj+F&?)JB!Bj?|e3?U?c8A68b?twpbz7p~N zfdmO$EelyTof$yfe!)EpAoDMhaz3QeqaN<;UVG}}BUr0=k|BgRxglipsM%*>A2Awh zlqW$v-DM%`HZe=3$TY3>xT7w5Fu4>uBKMq>^C3UR-NKBfshiHi(z;zjGK3J{RR}q`tw%0&ixm}_7 z*E0ZLPkC9%#74~P3wPY#R}XIeXJk@3s|Oi3{;maaqBpilBvuE{SeKBOJE7=LDlmuH@1z#Q_7dnh$67`9zcZ%EpF@3 zq7u??H(2&M)$T4?Le_IEC*+wu|BKG9lVF*+lB(xh`psnK^RsuTD1qhBdMZR{(Or+0 z&Mz-?1B*qWc|XbG<;Af?lr!HLwO15v155fFs-ABd?^up&Y{rADTfx%&6BQz~)b&Ek zyLZ|xz;b-yQ60%r%(Lv%WX_RliKjim(%yxt=Uc3nUqeel&4b6m(q}alBD7>MmX32Q z9KdpLnbK&h{m!vr6IF{48ry5v(cu@6xOZU3W4~H9?71QFD{YM>OXNn5#e5cHX?iL4J6Pm5QuTbxwwcVG<<1x9mV+gA8x0TB(%)c(wzVn#XKrRXsKW<=?liy0?UeY zdppUJ>&@v)?N7{CuWqS7%fQkko2utqcIrMvOI}lx5U@OaN`(k5G2YmhF84+s2TMbZ z3$BtSVKc`P+m+d0nhZAI0v3-xR6XCK@!=R+-Ys?53zqgns1TvWb2C~JM~=GzmY$v3 ztdT6CTR4{ISD5q5YnAlXVENvYs^?q6-+e^OsC&kFU@_IELWCB>Eoj;F`LiBayt=K} zBw4(-ax5K=F#CSDlG>-ha;HC4&$skb8jhBsmQjIVxu-*g2rZ4aqNR4UMH{e847lzq zS;}~pgv-pnUsiL$56OLnF-cGlF3l=Y1Dnw`r@xi|MrZrLp%a~JkMiL%KvCPkkC#};n2FswUREW^BU>jP_#AWpc%c_v12+0z# zontADV=SK*v~dE9%?YZWZ@JT28`qehn=<-t-mzT-v7Qq8j* zU|;b+%E&wpELWyc^?b|T=ggp1lGWjU(3jFVREW@0#aPOfzBdPpx%#H7k|lo!r!U9t zn0k0%dsrG z!mP3WZ_A^=vThJn&$q0xV(usD=p0M~OOh@XBD8q>qQxWgs6X`O%_YrD$r8SkW2w|- zj$`qYu6ux`I+3d9TQrq6V_z~CKN|#=4e3;f&@yQ!TB^s>i_b2OrQ>(z z{gAt_)OG|*Lo2GDZ#k*h3io|=^O=fZd9siS5n9xCp{27@ShJlFwNFGp3vbmWue^ZzW6CZcblJof%7X zKwvKPMLUJ6=UYtknR$>d4dc&)#Vvyh5n94`V_!Zz>=Or;rhC0UOO_ZvjzuquvE0#c zYz7w95UQSUxo*nbAM;Lqy&WvsVN{6F;^c>xvd5ohf~8?lx?IVxd$5829E(X!=KfgF z$VYgtG!3Kb`IcsPmd_zWN8m2$>O<(W4YaiIgb4p zSnLjcN&G<7^DVE&T*Uolx5xa~V7c&(3K3f5_MqkDh7CzDL%y?NlU|bL_PHI~QjqM8e>Z@8%NrDOa7C-9ZAx+V7f$fOj zx5+2lEVESX)`3d-yAk>0TH3f#;Wb9N!Ii7CucV)(X0#7-QTK&IEU~%LYhCoI>8wQ`d_*dakUVfluQG=u-9kqvw^=%fm^!+tyJ1y#?t7>*mS55G)$KQre# zSo$@uAR8gH`0mBsT-I<4rt6PqmtD z0hSRhe~^t3TGaQU#kuCi6I^ZL~UOJHcd#gZjwKc_FXdNWUaw@e(i4lMF{R6XC4^vwuQ(p$5>?N)tXDmO`%7|J z@(Qq|HKsy@7TrL!tlb&(6fCXRJlrW+yn;9ug&1bvpLJwLO|YEZMAh>x^2?dySo<|U zZiB^q2Nfc;)D1#Q=_UQaVEJ~-F-Wo$^DK&^n7(*4*8Bh#|NB%u-x6BEyes6=p>Y9V z+4O`85n3`B%e(^-Bfv7pt1Miyqz7~QadB3K6O zp+ba~fMD#)Q^%Tj!D97rXOv`#Jjk&$KgVo_<~ReKD(GE6)$=VCzRVr@I!nWBz*4V- z3K3c?4x&YMe|ubGWrtcPNEW{kjwN_0a~xA^s5cb)a^)md&$oPOn2q~>|ElX|V6i?= zg$OMgA!upSsl{~Y%lFGKZcCObo~7g&)0aV+Z;pWFaWYlUx9G((PkkoHJ#7t^gSV*= zp{0VcD0b^?0+t2`H$9RpxraD?F+9oKVO*tA-4QH{!>M|{rPd{L?2FgfaV^2JHj)Yv zT4D}iUrL+JT?!Vz^~;MSOTuA}rQ0H=FS}=Kwg$__ZB#wq(z?(O&y~p|K92^=0)Hw* zXz@IZma)@Z@UEPzLH!SsCG-f#vhN|YzXV5K-w&2M##BAuvi>{s*jWC#JWhH&pH77c zErv(XlCUr{7y4o`E&7LK@ebu!zV&C;*vMBuZ-V96T&kXLdC-=5Gt8Q{yWOELlNV4S zLQA7iw8S`ibpgxJl=I4^zwR)W@hsQZF>9=rQRz;w?EXR3^DSKzzv2E;v?B$-i*418 zA{!yJ?7OEO-8rShG2bIFo%l+%}hU}lZoG`KzpEC$^Ml6t=7?p@}UDBI@u-vxcy zJckMqT0)LuU&3a5z=_p{4+7dtmZ)PKOXP9p`-`!5{4uZ;)Za(y`4(%d0_=;@mX&2- z`KC*S2rUbaq2>PL^o`J$ZPqzGBuhXT$5I^6T#q@K48x~BnTx4JfgfpZZ^ubsDUG7)`4+!d(HN4n$oK+S&c#t7 zLW}otv=n6q>;+4W#=~bxmhclCi&`;r9NQ7PKoKlET2b|Ui%T~1&fM(vo|a&V=}3hL zEt5{5MO!uJJy@ExGF~WId`@yKS?s<)Uw32$Sk8B%>iL$oD$IGm=5tpgu=w<#LWCBz zlW3_su|YGiczG7MN|p+qVo%AE6~XDts~gPu7lmxJ77gP0o%a2n%urJO1N@j!QQ#lnPv^bqY zi^{rL_#Jt3#`(RHCGa%IvNDyq9&>cMxEU;89#Qpt%ly5}uLTum_R|1MW&srln)ASNH<$_$TI9~4 z<>jdoC$OA7y6e1TDdbrSDj3U+%J~Igu{cB3^DRBMF(X!*7w_Ky%a4mxh|rS4SZsUx z<2|JrO}fTQmXt_NUz(q0_WgRz8XpCV);Ox3Z}F{Q{vy$aiBE6}$IgQa5n6mBu`hZR z6AGa(Tb%tFvj-}N%=DeS2v`hmm{;R2azQunu^DBZylkV$5Up&015TV8FELv6= z{oDx__3~lalEwEN$MR_~)0dzfR(P*(Mio`hw>(K?{@A+t+d6n}+*EEj*$AOU{Ty1l zZ*e^X7LAWR3nj}(o+W-bV|m!@U3ch9$akuqZ}GA^ipS*Hj~=@~U&hqxMK(fcDP}B2 z%ad}zGW@*Kd&!b>p3@hf!;Iy?*kHW2JRV2Y^DWw=63}vMO!Hx28JI+c2rZH4u`e-o zJ;K1Eo7X_@^{?;yaZwyg(thR|ru2=OKJ;Z=302RxMAebQ^S)D1ljmTmd`E={EpAb0 znXlN&5G>_IQOc4f%csger;Ir3mMO28;QuiWS zcK9n)!Vt<8waS#m-Nh?ZbI;!Cb2G~b+m@7<2RTxtt|8hF2>89Y1 zFdg-I8Wkox9ks2bH7-oQu)hAVF!#@<0{&t;YMeat2LwGW<}85esA2P{q<=acC9{F! zUX(9|>8O@=RP5hRN2OWHs=^zrx&u>Gl{L!~25G4JS^70O(l*VsQ`4y<+Si&Q=T z=y|**^Lyp@F0uHV#r$|GMEJR6FrUkFZD0J>lx6fuU$Ue}bIt@0yD?|zLCNEAa>de= zs^?qeEtoqd{U7A@fF5_AO@#<80nxa_-D~$2=a-+B4VfreB4ap~&K1m+)02RJ`p_4b zPgFhMqCbFn8QdGQHtoSO<|h>*v{=NTCE=6CSvW%%Ih~p=S^O?>EMq^f!V3l4=g=UXb2>@eiH!z6ryrMHd>5n41Zp(S%xqAOS|hL%`JmMWe_?-O(7bm5!(La_8P zpz8US+OL^OulxO%IzV5RPoP4CmI}tQVaAEKV3~NK$0Es+8_Vg-!tsn{qjM#Go4p%P z)$=WdBO2l*Mv8rM1z2|9phAR}m{{z~SD)CmV2Nsf%tf*!T;^FOufgR}Z|D_#x}d5_ z)$=WuQOtWvteV#h0gGloDnw}Uyo?sdVnsZTx$l0tTC#-3aV*c4GIz3?Ee%=&Yb+~} zs^?n@6PW2J*U$Rxz~UTAg$ON%acI$Uzljs8o9r|;N*3=c983R<)!3J+Gt8aea#yOJ zZ!u3kj3G^rC!PdLYj-L{XlZl>EmzOK#osL*N%z?yS;}~pu05Fj#c1Qc)6kdN*;GB> z5?z^xmXCX$EeA{QXHo}8w^WGG5)zMnaafRS2bR2MZ4XP9sH+^yq=EY&>A>D$aVbJuV2=3p78PSx`* zYvKa2FZccGodrvZ1{ET-R56zLuo_>%Vt=oBykyD0#_3CWO=f@DHT&dRusnQB)$=XG z9hf_e#yLIkDe00IREW@$a1HyC-ZK!NlnlMT?51Q%y3Vm^vez(!R{D8^r8JGI=UWcw zRAXOahx$E*zP!Irg$OO)*U?hn!|5CJW#@&P8ImPDkz+aS!1Tq*<;!EROt?zb^DXHP z%zH|#9u7JPmi8%Bh|n@A5iK5?-S>c{=bna-C5ulI$8tlN+4sG7cS{1x{ex6J-(q!> z`7P;4pAYe1ITc2Q2rX(!XgQGThre0${wTAu!P4&s zRnNC1x;Wt)oBc#F94zj&qsc}HEpjPnQ9jw;9s08L>?ak;QpmG3pX|qAZx>kPBmtMQ5 zdcGxiL>7J^7M-SjfxcMpr$U4l-&@$1(v`~ieZMO4SSQI6k;<|7`nlm|tohQa3M@-+ zQ}ujH;e6&Kb-jD!JFx70K!pe`W~pdN7+dch^ksSD;+~SlH;rS7c4S_cA~$1fL$IuC zP1W-)51jq*1F391CLJuUov09@MLi8Iz0QVdfaP)ngMpIeBhOOr<9pnUStAG9gQbHS zRnNDKD`x(R_2tg`U%`^yjtUW4iW$qeKKA$|eooF_UCEM@&gn~pJ99l2Y@ufbmg)kk zo^N@h#GG(l)`aZ_OQY9Rh|m(5j(vH#`$-MxOT$swMv^7&HpilOfZ1PmdG>z`7DHpI zo^R>QkCD7cATTs1Tt=_YPWKI~vS^zGTI0HJ2=2cR7~2FPU?t$)|Dipf7>f zsd~OeEp;S*N6rg9VgQ!ssZ@y2Qui)eUY>ol9V{>GGHoPFG0$@DG-GLDIRXDnWWA+S zJ>PO>6?3j^Vq9}0Sfbsj5TPZ5vDlB9VFs1~b9y*Rmh=oxU&>XOYs-K=4f;Y~au-nb ze9M?M%%4licDr34EZ3Z<5TPX?1N&kUvbi-_G~?H~NtVca9E+n1V`ui59ufQTQ&;OPjlHl`K^}i*6$G)_{kd_fG&z-`!L_ z-_mA730lJU{_F*o69=dep{0Vc#FlN?N) zJUYyLf0_8bLtC&sHlsp>mYDn4mrWT*JAmb!Ra}r{Nyy?@e%LbK_lN7>QUJ>hd#avq zxjmS9P^i7R??|v1FQr0+7SAlSjN9vD1D4s|EyE;B=mUdiu|9<5y{B=j%PO6@7 zvFUpR*I2CL!^dDL*hhs3Ert)!lBBPJ`-@(*bEIVP&gNKb$1@+P(`M~V=*x?1R6XA^ z>`DY$bhS16fW%mMls-ABtsZhrA zOB;=J{H6UZbt*(?33-Tp$-5SH3HFzXdJEGfOH>ZWGPaUAj>S*^iE|fTTJEkl^Ew;6}$#K8Vy6GqkZE$ye@ z!!`C{#LUBB>3N0<5n8k!p+&xBms4Qbs$DNnvQ+ad^?owfF#FzLz+b8)jiBoJmdp;! zwdLb>I)z{{Fr-3+mMX?F>*e$Susl_qS1eibA9MP$>oW69+{@am82WO&302RxoPO97 z_ZKtm^Y{y*@K#ia(30>N`(i(MIZmwJ^gQuSvLrp>SoR%Y`ttIpq6%2Tw@~$bOEZ1u zp3;`c-uUam+_UjafoEjFS9Rh*antfOQ?Fj z<bMm;J-Xvt?RVJlV*1&gKX9#zSb^^DUO752X6stqb5pf5@7sd~O; zq$hL3aI%fxaIjqNMuiA1;m@!y&R6tafMtu^3w6m7lgF_bU18?>JQ_9I1C~dMR6XA^ zzh-Ydjty8hx-VGrRj3f5#VHRhJ!{>~1B;?h*Y1)fFrQ;lTFG4VtL5eBg2lw2s^?qg z1*YQ~+oHY*-{3Yehzb!}2IZqA>!v=wmsBsS>tx9i_aDn&Oy0XFJ;Uj5$ECgSue#lM zMkNU*@7J7ciY=OOro&vAyg&GgO8L9V`%<@b+^8q3PK<=f`_XTyxWAgbuOGo|*}Sr(S)W6j^n7m)qm7P#C%WPn%@9h>adH=RE75n#-_jWFwaq?dC+Hy0PypQim1^%Dze#_crjp68-*<$x}>4@W5z&U!> zp2>XED5-K%grn!^j#NGW=;``%7al!l{y20BK9{fEsSx4kB3FP*Ykq_l{>FE3P@#on zDdbt^t!BQ>{z%${Pq0RWQT2RF_H0eGoPVA*5fTc&yt<{pn zw~%AWoW?v9+-7vKAM|B$0aeepl-*&bJRf{B-vX8eB~*yeqF#s=^E8ubU^)C+d81_c z$g@oIV6M(rciDRbEQ>0sdcMVlec#g6@rH(A>HeJx5n75F%jB$?sbH}fY`;UY=SZraZz;W53on;O#44->OZ-?WL}-b8iG2ybniUL|Gxx3rNS3%) z9LvaD=1W4;_@1d?`Tm5e=UcwLa7D|?@&o2ziF`qY2rX`}(Bk5&hVQ~@bLPh($r4h; zv7Gj2zRhYa$ax8tpcty2Zz(Ijg$Kh)`R;fe(@CI0gcjW*wD_7&z;7Dc&d)h1S-grl zmVoWdaV+9X!hEn)#!>Zri<<%S?%Ujcsr#TW^^>U(p`~syTCQJNcNx~$(wc!$lBJku z84}4n@r}M@_601;^{INk#pf;a-i3DYjq$w;Et^pxLQ4i?iTj$E1D1lAiYt;Oy@b=3 z_|42WKy&~2;b7^KNY(Q#)~d|A3W66O!C!XOOQS-BmVgrM%ZG(eW5E&}HTkAwi7e$< z^evhFWwp`4reHaDg{tRUQs**%h4w~dTYT@rq#IO-&|*=FmT`j>@O|V~3yx(-7QfdV zi{E%=az&xA^L?;XFQMxBmd6d_aNl3H&l>+~^O+S?h|r?(8ZAfL$~}g@MAv)wNU~J% zEIm#4;%4YJm&11yEox5H^DVn;hvDB!&|drU9PImj+fpGyO9f+zvQNXm7BnSm+6&2& zTgK^2XgqVxucYN-27S5ll&a@jg1Rx+{2_Y5(O|juiV6{0V#=^Dv31&?fW9P+3Mi8- z32!)-q1T!B`Au5p9}gCv2~<7b@}*zyNcctd1lKv=z;a+J6(Y2FzClafNncFCGE4Ve zrDO?x%dzD9Gy9A6q>y1?Ni?GB`IhAa_uxYPdF6~hST0SbLWCB>w`kd3UV-0V+AN*+ zQ?huMb1d8BnKgDIJn;os>a3ya`Idpko6(Z7%6$V^$~I6TLQA7^wCtN>5e|LHirriH zt@Msno@Ln?=6bAU^-?reUrcrt z^^`2tJWK7X%)akyT_y*X(+8+}zU6DeX|!}O8jkm21|6Y7gqAACl0EfxK3HbX8#hq0 zIqUB!Fh&f=<^Q1zA7VnQ}@!#lr6)ZLN3XCL6 zcqPXY+mX3POZ>L6OS555})eR(#M3K3cceL>5^g!WIMFZFcoyd{h0 zSB}MDA9Eijy~m+lV3}z|)$=XYIXlpD`A|n~usBbqLWCB%uV~3o4p{&e?KY`9B}*aC z(&RAn%p!HqXnn9e>O$4?EhE*LdrHlHYWso3w-*&6w4^YW{IVTQz|y6l;Q`5#@{QA% zpL)zRq2}SNr(ii!O4ai%&&RaH@0H#cO7QQG>Aj~ygcjd#*q1*2J3au*Rl^0Lk|m;= zV;OdXxgJ{-;@tu)1E)~+d`r67PP|TTV|yKc1zkRa3K3e&s?n1AWbX~I1b$09C0TsG zb1W-WnDhSq8bb}ha$6yt)blNqCNk6N50i9%z#4O`M}-J2>fh1g_~a2jOFA7~KU%VU zU<-ZeroJdavgG{W z^yN$uReOW);65kW=kl*EwWQqI9 zu{0dZJmWhyL-ROT5=y9gzGZEnSJ;=^t5(WGUmV|2AwrAWPqfTXI^PBQ68QDn6Uh?d zS)+S@Trs_+{;NqycQk7Kq0Gw~g(-*c!C zp~Z`_eARDi02X!YLlu&xm}jwS#jLTV=3A1$5@=7=^DTM@nK!t}MNhzA5DjsqLWGuL z#uDipbrCGHuPIhbmh^R;zHDE`T#s4Koc|CkrKwas-y&ziyeZ6fX;>`oFZZbsp=G>0 zuCe1PwekMg(G9+J%75MKi(JpK7_MNh$JScbx&xM?9aKHv(lv{D{!+gpuK`#J_EI52 zOC)14SnMAGeQ6NazKLY<^Ws=0MlHd8e@;KIv0&+NgR19SwqH7oAvr@2j020>9V$d< zX<7sOvc*9a?~j@IuWc<^s(6;K?AwCA$Z48@#ql6j&$nbMFz>xSET4{lukUOa6(Y2_ zF%}Qw=mO}=gOQIqOP1UXoW87^!c41wPdelbmO108dcMUutqS|%G^q#vLiK|w6(Y3c zGM1jBn&5HFrp=h%k|kjy$8y?+xwbs``D<&iJm^Bz^DUj^i_kK=gDd_G)kD3g5TQl4 zCay8d8n$?yoHlmbAjuNCiDQ}Q#2k~y)R?{mEaOe7dcI}9I&-cZ>#-W|!?ZW2LWGu3 z#xkSL`Vd%SH?+(2B#XB<$1-aR^Bh^j;LH!OOxLCA`IZU0n7269$UN%^`~Gr6Dnw|h ztAKrJ*1C%$44-z2elth19R82xFQ%fer5(os2bx*Zks@1~;HmoCI_=Ivh=+DJ#7GTzZM&WgDT4ENjB71nymTdJOa z^qia=j%PJx#X4E=xh(%og$O?vd**XVUy<4f&II=NIy*|1jLn=g!K<0f)wxlV<2Ari z`IxHbTW(}Kpe6D|h%#74yr4pamJG(Cx9iC!uzaiK=_XmCws0)(i}8?an`Te4QN1Z?G4GBz@2=w*5;pTOc< zMb+~yD=+$CU%WL!@Gpfekt-z|A+!WA7F*@#YGCP?ICQIIspeVQbYv_!Po@=u+ygJ5n8GlOLB*i*TCYedeKj^ z%kn2`c!}Y$^8^0&dS^E(L};<7jcd%$b_l*-`NySE!IC9u8^kjSia%pWYRXH)fj z%YyFq*cY`ojqwSVfejTRv}n}9zASz3>OtvV{vw6`jUMl{2*A~JfQ0N zmYMUIUzk-`uT6#HSnN|OL}>A0EJ4Pr@h`UfE|R+x zcaG0Q%Nxb`n$VZI6W?n-ecnlyLZ0QlAMXgFSHy$G~e(E6MVaXYn5Ff}64FrfN^HWN)MD z`IfFVBQc~!r^|RuKD>tt5n5asOOp=I@tu(Yd9}JqmYltuzPw+^Oh*}BKY_o5{8miW z^DUcPnD1hpL*5LB{pH6yDnw|>VJvy~x8S5#%&_KtBuiWX$CCS&xg)=$_QFNbm;5nQ zJ>Sy54|8B|(${)4SSpOE5TQk<9zIyIkM86sIi_HitqycvsIvL0UZo4=&$ z`Icurn4eRoyeTgR%k(!?h|m(kSemA++6w#r&s`e&lErI3$MXF+b4=b-)@C;Jb`GaIkqZ!f@CS?S?Vohj>&mra`BiP>qXV` zEy_n*;Ca9FYRFiyMEX!6Ld*YgcIRq5&pG?7bIwd=xnFFx zKUiGm>=+{~2|FcqsTRZ5lnOMCV&{E?9f_A)T2DQL^H|R8v-s4_fmI}l(vrX|o%iTh z0L%EBccuzUsK3NgV9Dm>Z(layoyAZ0NxaDI!U6m7+1tGwrhC@ zJFv`88*eKt-n%50t##SH9~QU-pSHdFip0w;j(eifa@9HlpCSnOK$0je-pt}2?~ntQ zwVhtA6qW*+rOYmNH*WssVb5XTH!vmfa?6=b>1bKLI(7uqrNkhTL}@8tmJGX9`2Ne! zTBBWrB`rWw7vpDaj73B=#Jh36)kwVD^8O0Dv-o_FCccl{vJOe2v`nakW9*{c=6g_= zd$mSv7M7^p63e-d?0l@nlT`d-dy)}}ms`GdW8dnuH=LXVbxChUk|-@v%yQ>u%SfSAzqStqdk>P_P1mh8{$9HztDaGb|>3?fOC7FTB3{=C5%u#D*X z{gkj|$t>}M*jk_8t*7O{(ycFvms^H6*pB;t+E>R5Fvdm=AxV^$EM_q)o>LJl(VK=` z5tgLAlDafH%5I(P-`^9T>UvDPZ9ekJP za9@wN!cr)+c&%ag#co>Pw*^b)9uhCNg!4s%z#aE)fo0xdl0<27V3z&%{-#hDztcbR zge85yq%Q7{S=Up^)9kuxkiI-d6=QYA}{{5eC%e){-lol;rtc$~RH(XQd z(j%{duw=_DZuQt2Y|o|Z4x*($iI-c>EM!j*Eqf7$Pp5lZkR(crEwcnHuCf^FVzR)i zm9V57lGLT`M|RIbQ!@qU^2P?A9G+0D>~LXm z43SvYy0P81TknQA#x%;2c)4X_z(BM#Zrupqd=y)mBvD$5Rl~ZpxZ>6d8qnS}+B8R4 z*8I=%ht5d-cUQ0pQP01Y7j#B0GbWi7ossi`s^j!*y3Bs{d1GWSx+9_`cZULrF??XJmWc0P57r zb3V+#^G1_IVnqGc8Tmq=0pGuvZ3vx_J`+ij|JWIMcR`D*&>6XTDoOTxXXKmJ_SnyO znGOz3F9)5G9Tt&H|F$#ow+&p7Y}gJuBm3HuwEx^0d1lL4 z?2PPfSFad!MwVDZ68@{sNbM(Uuru=I`AxboH4Bae&aqZ2#Eg?6n?0iju~j*zH(T)ee%;$7UOs!K1bE|fvGZ!gc)|a&_#sK6 z{9mTB|K(%QA^dvd#lR0v!V-5>awNFE(+(%`N&}lOgZ;MtVG=L5w4ZE&Q)bq!sdiv_ z7DkdNEpg0J;a7G1DoshZlr6#%a!g{WR+U|8WG#667A&u}l6bl0yiq3pA?Zi8aYcUb zPLf1vX=wpnZm7L)lT ziPB=JhhuDY~%rT38}aNGvA?vsa90239)^mZ;t&UT!H>mVNoC)Vp(jVA(T} zBvD!-ndQsL;NxIf)Ot^XumpxlET3z#zck}=!>(Y_G$HYFOP|K9L&9+9hd8jbFeOQp zmbx{tF7_SWQo(Z7a@+%9aXKlnOdrO^*!QgW8emysP2%O2@KJMc-~Z8YJ>Dg~Kbs^` zTAY|gFS()*SURtI@IqKJWtQT#*s(J9QM>wJad9T`a!c}y0JMy&?|mHVGR=b|QCc#Y zrM&C5wqPlJa^go}NeGwJWpN#LXR&=u4L`8BN0NBCrCSB|Zjag(x8S}%KZYbxTFmuv zjGa88KLsp3YSj5IETN|)mV*OX=Suo?FL$u4s88bMmbB9B+3P8pN0&ieCN(8Vl$KCt zd9QW78(6yc{;v7??-fSx(-O=0DeOs=?&Ypb21_Y(5-+!eS2w{iHsspn0I<{_Ns=fn zx;3#bYoDw>0hS@BLMsYOfy~msKbyzgw@w-i7PCtvUT!H_p6xF!-M_5_OQTqlL}^*h zEM>P3vjEG1iW)V9CGCu)F10ta=VGJBzMl=2i`7ZI+%nkk3f3jD!I4>DiL67CC@pEs zV&!g77A*V5|P7OQ7H z@J>nA%=`|*;(AVE8DPS?qh9r0IT|eXi%7iOVl#ui|I&Hga96N2v?ocFmNK=lE;nk; z!T0z7*fX=Yuw=!2>5y-5VbhH zz>>V-wXLue$}9n^*gV#?#N}wP)Z0qp<(5n1rl6(P@#X1YG2Tg%C@qD|a%WoE)?k^u z%HBa((l1Kt^4C)~kLi~gxfm?#oJqXgGGnde6p*$ z|3VjGiMb@PIJmGUhz`5_!gt_RN+$6N%VYL6I+IEcL&378%%Y-=P+DS`#Xd_Lzoxxn z#p}(&5*#J5thHnN{@z;ycS2n@ohI>e%cAbA%k$vZ0eH?|{Ub@Dv^1`ZbvYQ(dNz!) z-D4Z>5*GJpiKW~sc6{k}_0tTn+}=#$6m zlfsg6SyGqmVr(9J-Bb4?)MfuZ5-+#xxaxxzkL;-}pe{j2ND`%GWIY^XFYox6LtQpE zDtSp*BCbd*hBw%G@|4_fK42+jOycF1iS^h7l6j{wzWa7sCz3>IiC~sFPp{>IW&1Io zcwzC6kytJlV`~=|dYfSP?cxX$FSoRM>56r+*lDi`b=eq2k|-?(^;um4-`@hu!Y*4M z2#e!YiRA;oerdmEaX+Ywh8c;MTiTVJiIxd12S;So$_L&k&Xj znWgjN(;;zZ+ z@`hPH)yp0OmWf|Uyxg++_;s`_i>iDIET+ZtiZ()N>E8gySVpxb*dek0)cWtj5)v!1 zlsU^RWA1<12bR=JBwlW5R)$?yp4aT=4;Hsrl0<0J2OfmQ!ta)fJWp*ClnC zkjOfTBRv!FUed3wBwlXOF=rj#vD3C1z!*DVMv^Ek516H^qkS2$bg2KMiLgY*ODy$w zvktuM)@|oNT}m`3@p4P7V?NepX!Z0PV7b+nBvD$rHpDTurNcSg_d}QNZ7(c=HzXFV zpKRT-`U*Y#K1?r15-+#B7{kuTW^68rPeIprB}tT)KxVOXf9wr)seB=$yRbMVNGt~y zu`%}j?1}`aOYk)kFSi&S^2fT|ZaX&->N4{dNuso9Hp05hj8Dg>Zl3-y>n|*sGRq`K zc3nAE>uEEvlrG(;C|+(EQ>g@6x+Rb1r^xZ*{C@l%hVqV$C5xUmqz0kH7mV>wc*tJ$Q z(jObds+Yg82D;XE*U>4;r07~3H-f!Xu5I?d<?b`nY(eLtZN$Ygs!#O zE+pNby4DgJu`T;Nu(<(rt(`ngQvIL0)-*;LbcXl&-OVB?)m>}9edWACv7T12cW4+^ zF8c4qi2AK-ZPoXx*!`llcIzRim z2@Fi1Fp}U8U2A4*yfM?ZU$xId*P6vMlIh=et^KxvdbNC4K-XG&n*V@?wB+Y;8 zT6^m4TnMx00RQhR#f%gBzZvJZx1f7GVk^JhyzCC(Sieo;<+JCLQS9t|?EJ=W;D32o zy0B;?l>dt{`(H-Y{ay_$i`PDJ6&CMAi6yQZyV9^VYFi5|R_-KTZfUiOU76cX_Q1Do zYi%P*looGhY1n@?KHq=Ua)!6C6v!-gW7+D@&Rx5TgJrTciI-b+Pu#>KbbRrp`2NY< z*(8b5Qot;u`cKM-x>SqS-zzL>Ns_v}?awT_1KQwx8WK<9<(Ah~7WS=po~)TMV%5-+y|>ai}IXot)guvF?# zk|-@v%(CX=yERakVqTxl2}@A2#A08_o+)|JdBp~>^ztL|a!bA^`!e{%+Z(XU^YUJj zL}@W>f^~^-&^-c{_yeo135)9;iKW2<*5&DUdHQCsj0q+2a!cis?77&Tg~|BrLi-4k zL}_tlmM+mcSHY6@X=sYDWXUZ3_?IaCr)=BO2*eq!dYA*6Oo;-?j5pTrpYJs%}xd+_JU=8>klbHa`Z7qYX)- zv{*F7F}AI}eI8g6&pmr1Ea7)0mb)+6MM7ZjMsuMqH*!h5+_GiEM6An<;KWd{Y;O9z zXd{%CaAr9|$}B7T%*DS{=ZIxps7tCViI-dIB(KFJ6I*Wd0Lwiu zl0<27U>4n|^4Pg@;rZ(Fxqq)Prr(#;rRH!p#%$}v{{YLtED|rbJn>+iUg3APMT2E< zE=i)aq%+IBUk}ruF3AgPRu`6-2NFxd8a5|KkJ7-e1sz{W;^meO&)Q&JuJvr`0+z6K zB#F{u)*Q!JajP@cFtdXR#U7?TU!z@x2&kh?k=>8T@nSBsO2Py z(xTM@>$19-(Nq{?L+k!95teM3rI{r=z7!^p!1pVUYf9qf7At+W@;j#ewAx_NZ$pwO zEw;?^^3a_gV5x4iwV$x0q)O^i;|=R7=u>|0OW5}}&L#13%Z?4~TCBZww~AodvVvP?sm3?=mjZicWpK9N|=$FjQYs@wwKRiJ4=;^mgZe@#Zq`VyY_F3-0OND`%` zSSzf{^NNM|6hXw8@(YC}LuTnah8?m8oH&o~j!Lm2@p4PRS0h}fdv)hRH>gYH*(8b5 zGL=~jOmbU5UA6`I*b7VCQ%PN#7GvjQgHw`YV1LmGCh>C1$G+#VF3mpVo&`(yP?AJx ziDMSG8{Y?l#qVQlCt(RmmssxaVo$1sIETlA<=R&gFSjgsRtD=*-X`cGSXP&4Q?wCE zOLJo!V~0Ad$M=*>?Ox79SiGJ|EH4~bU3zUjbqOp7`jdFMrB?qoXvy&i_z9N3EJzZi z#fw=gT=cyUb%`AqwOv?pWtQ-H?DQk0`L1DLnKqKd%Pm{FpT;pZ_WG)Hu=Jitk|-^? z%#s*d3y+o0$B)?~EDxSb>Qdmsx^Tv?_8tzFuxTV-ZVAe`kCwDUO>phPem+T}v{<&r zF%}kHV;9ur_89x)!V>vHVp*xh=435xtyN&@W=-Pd7UOl^Xj!o-2Jgn@&L&BemPlrK zK4}JiQzgx&`FUXp%#c``{9xB&U8ehCN0H7v5-+#h8^rz)qpII4!5BMeN0KNlb=zQF zmUN85{UxRUj;q4r^ipE^a*o~c9ooL23)Ds5nZ(O2#_>C`F2Tm-hk@nj1(HN*ablKV z?I+?q_PSF+qOfGjECwd*TFiW-6SK@G@p8-gUaiq`V(HGBVCfk|k|-^i%rY^_WgqM> zN2~j!3QNK(NnP^puyg)7`JHg>!nQt%ms_k>3`fhr-eqtNc5_pbL}@W^i(~9~n_PT~ zAY@F_*TNF|T4Gt=ot^W)ie1_W#+acMiI-c()eS<+`<;d#pe|msND`$blv$cjUvV1h zGOB8wJYn(9lvt{SvE$2tQ^QAr#iKTfms^?`Z$V4DI!?{O(!z)&QCf7{VO_Srx-|$a zAKtdo$oqR8ra)#f=+8PcZOy2nc)4YwP6*be{G%7SU@4(Rk|-@{%+k!ROE0kW zG`GxQs*cqEWSVnqGcftOaA0ZRtNpMVa$ZFfkL|JZ?-9JVzTI`G;(BFTR5z^k*sAN%G! zO|JBSfm#0-Hvi`drCf#;P+68@_WyfR}mumkVT!FG3HmPjrnIhC`7Rwtaj9g^4DLI+;B#+Ra5 z;=gs^S;y43g4uIMg*AR+#!3I*jPqLu-uV#r*3B+4``f};zto4s%V*Du`yS!!xoF1p zJosORt{_R2|H~BizvOL>s|A)H13T>(mY7cx%eLlIaWkft=%Wpm!7e0TZW(09&d$eY z4#hX_x?dtml$ID~iQS`%UvT@?Yg(wV1b>!T4t{2JDQ)2V6)YYrNW9!~M04%ekA72xeY?V!E{mMlyKwFw{DiBeoj;Ixxus3)Hax|MXimzH%&yxejuI~40ubM?6y zU`a9}NtBidX0ffHs{xkjRj<7f7XN&S2FRUA}nPS%RfwwZ(bD;`l{ksW^$P{H{&O90Zo@bx6G2@@5))x5waH?eQyt z2}UG|(&ET0JwuHefkmsXM)CZ=SLHHfmTQaHmHF^4ZC-$-N>dUqx46c$FSw1XGaKKu z-@gq>qO@c%OK`(B*sgT%`%+jmFc{?hm6g(@({ z+AbhTl$QQoag5z@yM$lUjYJUb3}ei+OPGoM=-`JnUQ$8Bs)k^g7=Lg2ih+iI-bm9bl_J+g?{V0G3mm zND`%G3A4mIw6g=tlOyGO2+M=-lDd@G$5yfJtge0mi@_-pFSqQS$ZqG`{w(_$EaxIg z5~bwU>*A#PMi7b>@n$=+YOSf4S@f_xAXRD%kxg}u;`@vDYsQ%S?~R0w6FyJkXUN(X6MNPKbt;S5 z`?F{xl$JndnONN623Q_lJ7Xm*PCq4`r{k@&kjwl6apaQCc*6U|lSZT{{Psh~q=&2}`ET(ytwx$6m!hEd)!-SrRX|R5-zw z_7+dOG6^iH(Ikn|GLu=>RJn)eV@azzt`wGpUy`~!FUgLTwpB*?f+f|3#LF#zE$xQI z9dqv)zIP$SlO$1E5}4)0LBmj}i$PMJv#^9ZYW#k^pPt6%F%y%hP zwF?XIohv%eNfM>SxF?P=-GyF{!II+dvPD?D*GMd{QrNXv%!NJUpe{WRka)S}dHNx& zi&vMgi(rg(KSq)$E#AyhK45%9u*@3OCO}vUWR?l#*#6Sn?VdB zQVhR0u~Xx2(MBjOQOsgnxOz0y#i+VgjIabbNi3(vuysz!k@^ZYj9JIwNbZ z%?SsKzb;9lv}7@hh3C~*V0pW2d8V)=IZNvD-i6J{b@Z2AhPr%OL*nI@MVr|ASgV0E zMuVmQ29iW+u`tCk_I>N=eqhlG^~@KRa2JW?SvWggZ}s9Xj6! z<2nDR%Or`?63#5km)FLw=#KN2Ykv8A?ZVepVrg`Q?fWGj7e+x{-YzBaa!bDrtkbKU zMep8Fmj|m!5~W4Y4C^wx{+U>)OPS&owS}cnW*J|box{vKvmd_^F{B)cms{@p)y8#8 ztAfh-)_|3jNfM>SfmsgZ=Hq*>U(8LeDJw{w~B&q~<6W+*s)kau?H%ctVhV0VgS6bz7V0rhG#LF%FyYE5E z;Lv7kpf2l6zb@Jcr6rhIg2N{zLR}J0{pu+!?(Pyx(0FzZV{DMI5-f`@k$AbK-}xPA z@ws*rEp1{+5~W3}FV-bzaol3C=&jslE-cwHOP7Oe-_IFxp$u3KIgxm|_%q9n<6ka7U7lH0SSc)yUJ{FWEPH2Up4sMZV9A<8;^mflruWb?uY2|G zP?znCNfM={Sbwa`_E-CcLlerrZHKlCOZ@*VfB33q_98dz(5bc|v;exJO6?|@6x~su zKIq^nKxMOUzo0v+_Cb>6pSq)l@wTCTm%KuuJId`CN%yDjsP^^PmaQw>Z!mO6ojFBP z{hzv{im$2{4c$?lE|8S!?x^$@Uija)>Y9mrM^Fq&Bu3P4-BH8KGT`8z$MOASb~i|p z|JWT>uCCW(=#E;FOp^WH9d);K2=+9s)4h2P2Ik6#B*7oLqdLq#g_)M$9p(((QH!3E zO#ik!>bDJy*_`DD-BBH0leGWb9kscS4|Ye@ymey~bVogWPZIvC?x^g|`Pd!x?7kyD zYqTtnnSJCOd z#ft?bUTzsPFA>Ms`C2pZ8LaKa?29%+X^CW(itAGRz_M)m=PY3f^pRMGgt13kmINKS z0d*;BOXB606RSMY@@<`Nd#KC!&IC3J_x@^m1xjQ^Au z1Qxq`{fpw|mL3B);Qmsj{ybdyT{w~?QCdQo#Vk{A7u4mkecOh@;_WA~gdSikjHwIr zrh}#KX%a8DthvV8#19Yo@*2k2-uEPl(xN*U>oRNWV0_NccxqK+VJVPV{QI(1Is0;L z@GfcgJQ6Rr*w1;1b$K*iZ$4O_*pVbk%W`J1_uqu8rDhxVng~nUPDx#A)nLzF?`xqO z4|Q4PP2%Mi=luITCrOl+G-gS1i0=YpY*z81{e&gTUt&3agVp6hn?Std z`=%U;ms@OmH$}^{Gp!DSrD7G5L}}?c1jpE@$$CCuS?l?JxUdB6l2{^V%*D;PnOJHe z)Fr`}#LF#T{AOX2UY>iyz;YvyBvD#|nB`=4H@q7+?1AGXVQ~$RSn6xBh0cj4RVRaG z`*RX6w^U7H-M1Gy#}5I^`nM#B(o$w9)}?>gh*?mVVf#&I3rm*Fa_clZPu^d`cLrGc z9VhW}%YYyktjh$a=gYxjaE2sNTIMoK=OwMM^M2p;R6Aiw+AXO|voCDl9~;_M2Q209 zka)S}2w$eWQ2yyfuw*?VNtBi(W~mU~8;|>Sa^|cNmheD{CGr;g$X9A=ay78rDcC%XRy?d zCh>Ah=%P(H#{Any+yzTv97&?I_%cgaiXB?EUjOPREQK;l#X4*i+h|$;4q$ncOycDh zpSkQ>Y)SC*P_W#8M3N{ih0IcF za%MgIYIEO`UwXj);yQyQQCi%Y<#yNO5nwrY?($t>$(C7$hOjl*#zULox@Eh35-+!u z*Sd^<$W5I)SHP0g_)5`6C@tB{QXwwk5?Fq;+@CHiDF-BV>Fmg+psA}GV_p85MdIa_ zz<$AKnV&lm*Dn0`kt9mX$PqZkoVIqxZuQpV@4XY2h=USKmCo$Azxc5ie#Pm*G!iej ztQ}*7mQ%L%H$q*C&nHQgmI!8<_vGGQs7uuG^vZaybSA$Ykdn3&no%#?{yf*LlVpFer$d0gH{JTz66Am zc)4ZUN>{Xa-57WoEVCm?5~anFSxPPHvIpui%{^aRSTbam{+HSDW$ef;RiQ4Md`P_9 zVmh47W3IbGZNQQdK$0je8O&n1@}L1&CZ_K&5SF;ZlDaf(!0x{^&TM1^mhNv!yxek0 ze=yc1BIaRZuo&l%BuY#FQ8>mLI-C9iOWgZ*jfExTh{Vz&g&pSZUyQ#EmQe*HUT$$S zU~8}@+gz4_rAD##MH`{CgfPo>=a3CxX}sfNTVe4EkytXzvRhg<50mhT2=BclUT&${ ziXEf|ADw~U?I}1+k|-_Oqp>bm?3d32%b`c}dk9Oe%yMxHyRJNTY7QPNbzYKqxy8?m zE#%&uwiegNj%Sf1O3M;vX>HKse@@EpeR0ZF2?bRC0Z%;#(%uEF|W?qn@2fyX5l9|v~NqR(vW6xjD`C6IWz<-;r1 zNxc2SSA37}f)tWOX$fSO{q~)3tuOZey?Mgo6e_XY9LDOhGjTk={lw%siI-a*j`hVc zX0bQoB~xZub)22UXl<27>F*b4piI-ct-`jzf zO7%M7_s6-^orxoz#-OBaWKG3yh zzKNvzKXt7QvpehoU2FBWk(B?TYprh1faNfvHu;l8VnqGcwe~Qc0aJ&!SPk#=njb`x z{Ku}fq3h<9fUdQxM@X{YyVkbvJCD6)N9y$62Lsb5j3oF&*IGc~Rm{}m>DMpNwH6ye zGX2}Gwcj={GtuQRbgiAZMAH6q*INB~`>|{7ulF|_Lf6`tSd#Exb**JS(Cjg=e7Uma zN-wstnrpRaar*IQ_!X&vi6p0Tme3lDv-e}&>&DQv*6SWg^WVDGYCY=|4zs8Bb;ILg z#tHr3jPu(&y#hC|v%nokf>U9ve@-Ux^4T*z#RF&04|VM;!;zqCDoLXJUyRxRGA5x0 zKCgVK+>vv_;(bzL*>A_L0GDXoNd`+hT@o+1G}UT_mbyW6@UBmzS|o|m;>|4g%+hd$ zu}6BJYr;|>v!u;ffSZx{%hVbyk#9)6+;UBqJs;E2r+Rg$OVTHjL}@8tmI={y@yh&A z<&jCkk`^whOBV|#wCLoH#%D=K8jyIoWoM-um?S+qx)3Z;h9rs7GGQFW9*3joQ+TyJB`?)c)8`ISX|FYEE!$cYU#*c9(V;ny1jdgvJ!LlsZ zrR2B2SLL!~mPYaHjCOUgRd{wD^pnKPEm0K)qNS?NHhkZbMrqfgjZj*$nB{|!y)Rf= z*w@h!mZS(tUFJ4rcjHWF9V-R8jql@wopamc+{~+nXj}lGGN{@#`l+g(Qj6qBjxivcBcv%P_`b zN(HnPmO`23c~f>bE^M58ZKz9%Hi?&8&Zg_24ASXcA`=hRJLIc-eh<(9{H&R|`ZjN0D{EN45BBub0fBphRL&5ZHg9+N7! z94#!t7bKQIOLjh%v1Y#&SkBfY@p6kr4tt6~Kg$B&9rd9BNusm_Gt2QCyRf6E&|;UB zu()58SgO8aD_IkEc{Kuy&o~k;*oL^wVo;Yyi%7iOQffHsS_}R70G}fGW>1nREh)@W`hJVIV6m^&#z|Nr zq9m4J4R+k`wtWUZ6F=FA#LF!Y8W><*A`719z`ozC6-lDBbee)=tVQJR&S1%R*zYMU z{?QW4u{*3TZHGDF7;|?f@p8+MYwR2*L4O&30b!vBNusp)GfUHYwY0#JJ*U%7VR5`H zv6x@A#=rFZUM;*6U(b%j%Pni#vMx{m_!1kSE~QtIBuYy$E3Avbz&_2uVr8)ZfUso9 zEC}pKbcg-rWkV7#x5PK|#y`YT!vXL3rnewTl$NQ?^5p7@CSVz1SU*fy;;u;Q z(yEaiZbsjJgRq;hR7(;sw@l~7U7EA>9n__N2a-f-iDQ-@z3z4c%jwpyE(uFWjKtFP zEPH2Uzst?Dp)ONQNW9!KyhI%SA>H~~HwBB2DM_NVG`Gev*1VFrIaqRzuDBsAURNcS z55L%Td}S^O%o zwLXWQ3;dxjKOT~Jxuw+s_I|SH>@|bIQt*rC@poTVqGqm^T2mURdy~}^84TGFizJc zmM7g=UAE;8N(aj$PZBS;44usOmkUFZW`kwJc9KMCablLP-M93Iy1X`+r7bL(GK<&r z+4z^nxccEcR{a{0c)4Y3Pj);1>r4B1u#9R+k|-^i%+m5+hY4VD8~wJXuq4Dw>T;wP zo5zMV3cxj`)>BBl-17GHM*KrcZ+dbAEX!t+Bub0c{ z!H)aeUT(qn`aGz(peSB$DfuQ4Ejeuy@V#+KeMu6fC6rmdSM5?0EN+@QZH2`*9N#oL&~%PkR3?4CuxvbFFXMd6)D5~U@LSxk*{Pk<%<>wrHyec)7)=Fc_1#Syt`?bxA2lk|-@bXW$t7e&k3Munb*VXe}&3i4x0- zF|5lo;mGmtFvcGIMdIZa(~0bWn?5{bC|JB)NfMd54+$oUszm|B$jCY zZK&CAj^lZ9&NUJ*w=CYv25Q*ohu6V!^A<^>w3L~Nbun1K8{gmmWAM)v!jdJk%nfE^ ztX#ge9@Hgz42hRpg1?-=G3M;(gU8AxlSvY#WiGS$S4e0KW2|j;kM+WmbX!uF(cz15 zGe+;d>;#sGx+Gq1d3}nVJXEzfe+la1(3m7qT9TNh-xzOPx9rqff3vWJCrd2xE7*Op z!5V&nU^!?);^mh9yI2w}rzZH$`!%K{iPF+;7LKt?cI93{LygtE$ESqF_s$_kL zj!jj$uNy`|r&sCaB$J}k>smuCT#oHq;j=4rdUbXrS^lZh%jB>hZdA($Q};lp*F;y6 z?oXXwh4$Ee^#@fQ%=CZ9C4B) z_(P}Hy~+15)6+MFRiV?%<}AtdZ#%ty+rXoLMfQVEFQX`u_Mbbwf(*j2)9X#l%h%B9 zWgklt{;N(e{~i^w)2qb#src?HpG1;VIZODmSz@BQ>u2cn8h($Y`EQ+Gfe9H8;Yi@` z{;_Dr`K{BddkDL;_@m=fEg0(smr1;Q_RMu>{r;X$+m?hQ!HqnUjVzK|oo+Gc^<}p6E)Vs{;qIkJw;kz6>%!d^EWb7p{&WNtBitW_jK` z`zzGN-$L_^umsjor=(&El6)xNHF z0E=^mX|ck;SLL#07F!o~H!l9<0=(<<@B@jLTb5h0D|5r~b>@QQYd%S$v}7|&>QmGE zU>T5_US3#I9!Tnv`ih;5&YI?kU7n^dNxa;0DBBV1QZ0Y?Hn0qQPm(AtBj@56Gf)1B z&n?~f{J4g&L_CyO1_ZLY)V|RJK@UG8wjVDDLp|nIWi)*Tt zKGdZ_;E0C8;{Qluahk{$p{iGJ77p`R$RQFhw|E}higmfR?^;)=%aju&iPB;)59`wK z{?)l)IndPFSXdlWC6)wjcC5U8yBe-!8TTjga!awM=JXN{b`2w7gUq z-+^~=?++7U$&gv*be)2m5n6f&UYYwoCh>C1;c#{fd|S0(++Tcke->?o(vra}iymoU zSM=to75fWIT$-dV@4VSvQj4jk_AthZEgHfmL<&M z>G-1|)Mb1lJ9}Yy@KjP4t48elrNWSmJg7@GV-hd76g*KYp|Pe^JEe)x6B>H-s@xiW+1+6t@uomL}}@2i(}0G zL`zMuJRN>xv#Z4diPECE5bHA5 z{4Cy$>$R%wAz{gsS<=?C{bk@mot{va8_pzNZu#}KCtBKmsr3OYCY~gT(lV1-hUo3} zgt~k^cQagA5?)B^5_Fhdi&?$3{|hXOVo1E)vT^ASw1jw;&<0CP0!gB@BrwaxDm4ay z{#kF`p{6o5SDm4!*zj~4+N=pH=tX_AaBaE?k zFF&LUOWG?*U8?M4OIy`n4AcP2$wnWG;^me%mMqEr4;Ic~IWvhQQCcP}#xd63WBC-Y zWVLb35|*gf5=(pDQFLbUnfqXI>q_F~mQNMfK>cpscNbXtnvo<*OBA!X6kbaJ%lP_L z3WOynQ({?AmaW6o@1K1EEH9gpc)7*2)-N1mS#S1K1WR^Xl0<1ST!MAcOb;jq7Sm(b zHGcfP4&(YpV)5R<=CMiljn;vsR0N5aTV9y5wZ50L3=P4Oc!?xYT3ngMzUF#Eu(Y;X zQ%P8|WEMM9R+r_M>tp9i@i`=3ZmHOoT?D1xxjG*#Z5ERxN=p{A*q!~<0xX{m4QdKY z(pyPg-i%;fI4vwJ5}_`;T9SCVC8oqM9AoXDlpGJ1)g4F@rNzPy$C!oXVO zSi;{)EQkBC=7tvACrk%R?@$six165J)?rTeJhBcf@6V7VN=rDi#HLojCm}nZUf4!h ze6u8$G7H$6QvTx`JHcXhiNwn-t~#r+E;oXzZ2`;cSdv6((OZgjG0BWw2X%3eEZ<#N z3T2k*JJ`DAm8>4~!BX`aiI-dUZ)GoSnzL*IzMJs&O_D@uabTAELyhqIrCzs)0m730 zUQ(AmCD}a--6`+RfaQE&5-+!ux$zY1QhSYiAk-y(2uY%}q%%u^-L4F%%iI-9MhQ#I z2Z`lcQ?}MO>1G4$yuY!Q#LF$;OR&lC=*wexjqKt^k|-@^%W#ZE=+wlzEURQ~BP_um zC6;rK*xJRU5tr~-sZ)c*%Pp_Fd1GC=PZ+Ql>hh#6Nusm_GfSGrk=js~M~f;f5El1r ziN$Iw>n2>XFYzRdvEdFRUT%53{v=vr+gF_imPt+|iPECA9P45cv;&`pI=iruy|84< zEUnJ5j-sKhlJIF;&m$yWZqZrY8ZD2@oyI4K3Z9T8N{cPGTxy4Na*x<8uELV?Nm3Wv zYHUvaQfy%*sLSNrBwlWrwU}L3);@9yuf>cWk|at?3bW``^$UP8b|JRCx3ENfmRMeH zWv95c@AdNqOWkA=FSnT0bHTcNOti-C+v2GtiPF+(1&*=$_9ZrgMWbZaZej7ykyyUj zv$cy4bDrVa#jc7ZUT%4~UK?AYeidBBIoYKeNusp)Gs{x{8S&6ibGrGyXkkhFpXCo< zqBI<^1)Hkwx1T!(I=!BBA(<4NUg1tTd+ku@0=4`50AIAE#DI{J#d)jPdZ$fr-?19gbkDW!5DF2sY_IM=NVNo|4 zjs*MOw7)GZ88S=h^Xz2w&N%A?u$0Il@p6k}_o;YPOYw~T4mCdeog`6OrZP*Jd)Z&W z@_xXTG+~L$lhkGMVs>ZIYqr}4u;`2+@p8+_1#IQ_UeDW2z|wj=NuspGG0VN~9!_A1 zEH(bEu!Q7GET=u#;$YI{c6Y!Mt4-qNma>_wJF2)vvL#ri>5(K#OY@aD#5o<_otui;0TsM8sm%N<(A6pgK&TO*l`m+OPW-kBvD#&ndOMv*L)ab z2P!3M3Cn{5NnMuIVyg@r+Ev9ni!L=tyxijTGYT#7&ujWZUFy{*Nt6~#2OML|Csi~8 zOUAv@|)f5RSB>Ry+PvT77bnYhgAJqrvq5tB$FgcOWjpimvY{2)4-CU zW8Yd>oW4seg@@P`p!NBlTft&tK;q?=yPC0Bml{9D*?^^?AxWaNI5A7LFWvA8(DC!< zF2a&2vz+5!3i~)^HeNNYE+p}COKBH&cj3;?FCJhC*UTx}2&E;HSyub`;4|@Yla`tZ zOG2TfE>YQRq0{tmb{VKk;!F}Rw+x-k&evBwO^yVM=0cK0X)#}oV@!8@z*HDxyL1K) z7naZ;5=+xTY^WPuUWIq#G!00++_G-=CLCjlxp59)IcP|dC@rDPGVjvi?qJdRd4Hm? zc>k1G`b}rY7pKDgm!U2;WgZm8%PrGxOhAkF4KJL>dTEm+N{jAaSQov@x3L57ZjU9i zgrz`cDfN)u@g10Wel}QQ2atHVWu8|!S{(8duwylQ7)heEEN2!QT^oFkyxE4lCBl;S zOHvnmPqzBgv&|2D&hO?H5-+zD>|lRLxh<{m9YskyNfMQ*F)(lVS`uFclB083*nZ9idg zT`RHVEMqrpG@H8Df&FDpKN2suv0ZO+CA_cMZbL3?iZFU zndRM9cDmlcviM-IoC_rJa?7*~b}cq#sVT1fS{@=vl$O)XQfk>7?C@?>-1vmBB(0Ow zMZY+klXo4Rh;`Xej>O9?4n0R=T^jrd#p6q#$|Q->lF2M)(Fr~<#wJHZMhZ)~lf;tT zhK;cmogd@f>t>fpyxh_$G8ipY-ao^yoi2P0MYyuXI6C{b!vYJ^29B+Z|T8sa-<*BfwJ4@=4a*Lgh zwL9c`7%bm4wid<9Ew@Ygp~c-HRTC^uS|o|m63Q$Nxh-9wF8y8By%&}k7m1~IH8#f5 zE?2@aHf1x3ms@;a2co6k${F}{hrte#L}_`>EFl9@@Tr^eJz9JfmS9(jWrYu0>-+2c z;JQ$k;R8s#+~V1S9rt6;f4L5IsXdG&QCf7@;=Vu8yG$z7C2`I<&0l}7!?O9?nsYSqR!@`LHY36EW)(@IwDe_`AI%CDf<=F>skX3W%PgPl+4|Um z@A)QRiL@i}a!cv?EJ?wdFudbyvx+29T9z}*$MxUxz;ZmXxPh>wxJl|VdJMa>Xg&Q6 zuEQ8VB=K^~^+a~E9lUP?c1SFGMv^EkhnYp^lK)b$RNJ@FNLV5^N-TF@u=V5tpZnvt z2%JMmyxcOj`61l*x6FUO6vo){aFRr6Nn@6Yd#mAhNM@(iXe%uK?h?zA4XiGDy{Zm` zF_!a^#LF!W4zNz*Z~k%CV7dFABvD!_uETx5@~R$qUHM?mqwd1uxJhEsE@0_3Wk|-@bnPp6k!fP3N5FDdqi)egC@n$EG9aMO0H{kscC?MKgm_9UzNJ`~XVY&s@4+(m z8;O@&R@7$q#ZH&=dH|LhB`z0jgwk@4SsY_e;&H!j!=U-X;^if=d@y43SU>0eVPMI( zBJpxdz_5#Wtc-uCbq*{Cc90}WOBpBJ_rJ#cH3sT(IJ^G}Vab(QQkJoG%cE;Gb;07) zbU;zO+)`*gr#F0O%F&@Beo?peM3O{l>B20FQ&;@}OPk{NorLAVW=UOE^kV1y19Cd- z1IxkgBwlWbdhdkum{Wze-JmWXmyje%%Uos|YkUyziKpzEXI>R=NehPKj)QBWeT9TOMOP8$0V0rX&(*a>|@|IZYwV#E5Y4AZy zd;^n48IPiPxn-buX=%qS)tao`0?k2DVK%pg<@^6Ef9P798RCgu zYk3m}#6s8F^a~`DqHFEc+tN5a*ADQofv&YvS4ozC>RK}_?~5CC^q*_L^Jdxl@4JeERI{hzwl0*z1DL)Ti{G?G%?wH7n=Ffy}(-r*OA9=#xm#EAN> zYwbce1}w8NBLUt|cJM7p@*lg_*7VVx4P9$*K9OXx#zhI__YBoEeYwhe$lIh=et^KxvNt^%L1zl@vN=_}>bN;z&&BHeVyVfd27VimN zYge>L!hhAZwq$gD>{{EpGxk2rc>1~|r)rjP?8at^YyGsm{Cy%)dc zP|VWup_p+Jwn>fz>DyVi$M>8*ct*R)ip0xj&*&YuaQ5sGkk}FamqW8i64n1wY&!d2 zZunO_4M&1W&9}Y~mQWvwC1eb{(&#p!{bD#m&#@u#a?7Pb>}qyVkYPEn`1q0}N{cVE zxRx7y3@mZw(zAudd%MJP_M|)hrAPC&;@yRHx+Gq1G54E-)BnP-e0{KVs6~<}E%D6a z`y&nCxceYtZlSOg$SmER*cjVg>f0=+%e2cRUT&%8a~&kOHKO}N3cwPNRlWm?#!~Qw0CK+ z+$}ZMMpz1EmR_yc7%R79O&6$(wKj>DTj~wEf|hzGv+!=5z8*=Uv|MBs>m3>%V5#Nd zF<)5HcT4KBrXCw(-~zXF!=$s}HGu?k^xvX<$V02pIg4@nZGrM@ff`x(`8AA)7tHFqar z3Em^I*cvcPE0?GtU^$*h;^mfomKIo-D31YXNxM&yC@mwI#ixw^9I%v3ujVBz?t3Mc zIT`Hu;&--XAF%9uL*nI@()w;_d0FAG8`Q=16G@`9I5Ug0eV8j)Zur~o6qan6CHW$| ze(Bfk^9-<<93b&>OU1NkwCpN9bt+gI9wSMVmI!94-tlp5usCZxJ0L76`y_Qa`<$(o zcI_E*5G;qaNW9!K-X;|-Y26180E@RSNusp8WtPNFGv0r_ec$;)Dt@&&`@q^rVe#KDv2=aQ#@Nc7 z-6g@YB9z3-E%O7|-G%v??dyZ3R|H9-vOV=Tw8HJ&G1T_H)7mM~`7+SV4IhFWQ^pC&AE2PJj6IG3Fk#$GD5 z9xOdZlX$shQRE4N7+9W{%rDvqrKOr1?)&Ec`+9+8^!dz>!s2yEVp(&KtzG0@ znuG6MNR1)!a*K;DyEuL29))+~=H4Vpl$QR?@}yz!>0t3n3;Zf9xiZUPe|FqY-k0DB zW31{c5-+#J)%V3Q_ReH~6x5~h2a-f-v1gWrQTkKB(z@{qjpBc=DLptWsmqL5HYaQN zPKpDISsM~Bx41^_M$4RhE1bs~bssU91bY;t7s#XmM6?o);Tp8>T=!EL0?z`LnM~D^Vt3}zGhkv zuq?4A@p8-1rMJ+s;cQ*J6F+J;Nuso9Z^V7S>3N?AUo!QCbc#%ao+ST2L455>5LEOXzWlWs(Uy=YRTmaR6BS z%9D7xrFO_((-^=+BzS_^|2Hqj1*tgsZwEY>~Q@nvbhnP;Icdnc24 zxg~NqfBd*b$ueNsHG?EkTDmgJt{E$IV2l~H%(oVnv=frL3=CwJ86g9%gJo7n5-+#h zs>Pll8kF6?3s^?F>|oPdwiNEp^n1h8ZNP19nXgP#l};+fyK{^#LF!%A*1jQ zu^iT|H`L|6IZ2|lXl}xNKdNykejBRaqIGVN4sviI-af9M5B2 zYVWOR3zl{|ZbciRw9I6duT>hnhk5K~#`6GS2|q2dl!#%+%80P;Q^4}`1BsVgGPVbx zrCYY~bg-PxCrOkRKV~Ti?Y9^#9~$Na3ybd=iKUAj>%iNlRX7_g16q-Ixh3?}F0_m) zEQxEdCLKu6u!5~bxQvm~b3yaCI&W{)F?R|a$^Q>n=XXN{-B$J{uGWvc=JpJ#z zux$!-MyAdsS^jfpN_A(X-kICj8F@A%A_(@5p+`v~F`|C!jJ!F80nK;XxI<@TpR**%f9#B0YnDGdlBoWdV%F?`NoX(zJH0CA>AVz{;Byj7Y-N6SzI6_M zz47%=5-+#x%Vb?^wO5|s1D469D;I5q(&Ei5&$@;jg}Rt}tj!h{_wy1mMCU5l(EPAHr+X+jC%(B7yFWiit8?_FA zrQ0eJFSp!(;e|=2pV*ZT7K2kHiPG|!Sx$~vn+%qAExz{?XH03P9RC5v>1BfzMnSK>?~MZ2QC{TEFsYni`PuHzf>Il{WDmq zjwJDN%dE=m!N=QOFX1Zoz6m6W(lVM^^2{|KgT<{&XG>x6x-79+R$z5G?)aktSgM$j zc)2BP6nl46Y@;E^pe`%TNfM>Sm09XO9km`Tl_p)AAuPEvODiL`$`CbR7~WY-oI&E{ zmP)k}ag6QU*snWSt}P%*l$LYMl69+eC$N+_|I1cb9$b;srJM#^W%&Lf5WnW&5<}wU z7QdTQ(6ar8(f{M@PUCvqzQ2KojG1LDD)T&M$dEZx<`5GCbc>PmX&Ksyxih)fL+79{M_FJEGssXBudK&W_f%rwgOmg zY<;+0SR8IhEKT>Z)zTg|Cb;*0+3O@;Zs|UkbrW6;d5?P&4lE!^l$Q0(a-pC`Pp~*` z?{+{~(q)!%z1Y*ZQTaJ%I~4eBwlXWcY+Pn=~t}r{-yIxl0<1qW0qGNYEOW=>=?Ky zR9J$eC6?SZY>YK`9E5unc#k6Sa?9Hk_Qu`MZX0o*#mN&$5~ZcqPMrIB9p@Q>rT?e4 zk;38@Be9%&%TC#m#a;2Kk9Q%7ms?_1vM2Gy&V0mG?C+ZGihhLBGMHKHx{T5WOYxcg z?g&f1%u>dUb>7E3$U6b^#e5-&ms{S)uzeSzMtvF$^Tl{2Nuso@W|p#EwY0I(f(K`c1%Mphx26mXMEI&|t}#iXwDe<^Azwb=7^^fruw041 z*I_bcmLB)nxpI%L9`3i~w}8aUEuKTz{rf$k#ttojmt%Pque`rbK<(4wF)?-~7wfKbB{1rNqBua}Pv%IKg7z395I>`pY z5_(r+nf8*M_ZyG-igRDL5s8;ues0=_$FUI=_u)FbUMrGBX?e;l-}JTdZKwkSR<{xs zk9dh?#TQnWb7QVFh52%01c{ehR$OLp4m(!g`VPieoH%%;-9E@qYT3E7Wmb6Rk zUpgsq&;ziCENcdcHbJjf4o^(5*|wGQaP7( z6HYoh@)cMnd?xX7OCMAATx{x{;kcLQ*6$>V(lVD>dUcG!&ifZDwRZ_i*dvK$bP9Ve z_HFf?bg-BQl6bi#={P&E>$wkhfid>j zjo4+lJ~r+>iI-dIS7dJx`5f}cZuQ9qWs81<(sGAc_M4T(HCWw7i;fG6eWJuN@ix1* z%stx)pAMQ$XC(xR~&=YFe-t!qJDhJBnGDlDlo%c^ng zvan~vGZA2kZaS?fUT!g6eHkskZg=+tOGhh`L}_WyEWX!9yMyK8Lc`0#68%I{mjznv z9(l=)oGW0lJx1c?mTfw00i>>LG(O+YP9jN^mg&s$`i}orXs9{&F!`0R9DVx7POmx{ z*Rb1Y;?>3jq0=iVon%sUdR1)4w(~yLapw)_^lF<&viwu0mwV%*_@hP~?o<*wz2+Bx zRmA&Kr`N8u5_sY$e!h5J==3^Uk)--Rb$V5_v$TUwFAW`%Qr+qG+m~DK#vC39BdStU zl1PlG-#WdHKVZQ0NTDAegB>9h>UPFs(o`6oTMLkKf-#fkD-PObc#=8%myY)7EYcon9+8lQjRW)9Y-< z{O)k{>^pYnYjMO0_}>xdw@$B-_1IP56E9D^I`8R1;^jwA{r2oM|Ges;Ecm}XyGD|z z{x2H4+5e?{XA8UvjG8ncOITc=Nh}Xruy^9OnGbmi^R~`h5-+z5Y{pi9-me+B9W42a zNfMQ$mQI(E<6_#9?Wm^DSF>yI_unSl=*Cz3DOS_lsQgZD_-5Fq6W+|q->0+AV9pC4Yx(rTamWV?-Hejjwki^R^C;PEenR%o0HNevO1xcc` zwA_P-ukqRDcxT?UoM|~>iFhHg6ko zm_5T5EKR~41Pw!ZdY`S_E>%Pk$#*kiC^`91x>5>jex(T`ACa+zgC zjtRa`wm8kDov_3vOX{*^0b4OS*J#dWuvmX0@p8+{!|Z9C_x=urU|H~uBvD!#d*Iw} z`Zf^nW~1DObQ6}~R}xFs4EEHg_MIHu8#&dB#LF#qMhQ4yG)n|_0?TV(l0<1SXO?44 zEGEGivz>RjpRl;4NGwwd*#g1tcE(?zF3pycc)2D025WywY;U|9EHBoOBua}Dv(#?m zj(32ks#}i`mVB8dZycNZUu+lRo-4KWNxaGYvc zm-2V&S_;ea*OI!_eaTj_Km2NmT{uG@lX$sh)rGNWncrYnK3GP)BuSK(EM}=(bNEuQ z1X%>m5|+qRi6!(ZyB=$l9*p}d*PKD(<(6w%?49_9ZR=Tp<<5MPL}}6Ai*w)Ka9k;{ zRQlL?iLm&*kysu)X6OCnQRccZUvA`+c)8`mi4#~C{g6iZ)n<)iU5b8$(lUx!W~@11 z9W0Gvzu5_kW17TL*NxT1`fJ^eV6lEk;^mfwO}C+C&(P-Bg;VtfNusnkFw2&n#c}RG ze!1RBSTbdnkDl!OVs^l+8dyfZC-HJi?F~E7V(@Vao-14CkR(dWS!M|_xmOPAQf+dT z-NF+8R#KNC_U!!9u*q_KQ^jKriI-beeV&Gv*;@7A!+cr1m?TkJ(wXJl*q(T<99X-g zkFbQMODuQlvTMuS#zS!KkE>7O<(8~=ZfIFGKO+t-E1Q!fN=u!6IQN6^_C5i1iP^a= zNLW1HNi1ov*rk-q7|nbbV?8`cyxfvghV8Rhd(D|wVCmvRk|-^v%%V|u?=G->?olsX zSPEqp=Wc9&vNp}E>VW0Y01_{^RB62%>tegb0N*3;HG(8jTI`sm#qAT7!SYa3Cq`J3 z-%IK;Y@!|hjP}###DS%0MG`N!+_Tt;NxIE=kH@iR)kzYiC74;NJukQm7Uv;H6NDuy zLt<$Y$gcUz4-KvXmJBlzFSqQQ$}Cgbt-<}t@+?Ubr6q+~$_#MAH@gm0Hc1v1zf6h6 z{2DvIOe^(911xWAlX$tsWv2($<yso(i}rq;`@Kx6r9fRSf6B@b7UvHV zOPST|VeN#Wskm2ki|-^}ZdtwfBU;vLw?)h6;vb8CgwkTdEWTy)bitw-ZT(ePvSpUp z7ug#`!OadEKwX0Fk$Aaf=~690%KQCgleOU@PZBS;%&0R7>#|zUvL#q*-6u(umWrM@_ovxg;v09(;`|#6i~T2w#e6X9 z@b*5NicfsUX`U#Gms`S3?x4lxZh71f<5zc*L}}^8Eb(hM;dS!Y+f~{NORCKBcsHB- zbJy>VfiY&JRkbKyZmAdGj22`4`fH&sS_4QDrNxF>2I|zoW3t8MM!kh4I!98MeS9y^ z+dbM}1T4ZVab(Q&N#Dsu`_8q z@vF_NN^~xYms`f!kHfmet*MG{-3&1zNtBkZ%(Ae1;tsGlJbt!BSQ5WT>T)lQT_<-K z);<8n*xjnni{j;$8`(lURa3u=~o! znwxO1=u!Phyxiif5s!7rxv{X52UzVe$SdvGht{ohvin zTw4xx**BTQ%PlhtSu4+}(p&7n;yi~WQCi}e<&HzvRRYVG!Mn+$W?T9_y|k7%;Fr!)|775?3xLd z!wpHi+!7dl3CEbJ#k%oOmsu@I5~anESd%(o-2YjTBvD#sGK**LtYokxobe17mf!-3#c~jP+~8ez8@_eZ z{W6J{Te2VTLW@KBU|i>abdw}eT0EF#Zpg21VCnJDIYwCAzDq1e`>^}Unsd^dLtVBf zk$Abq-031(Y+oPT31h7FTarX+xy3A7DtEJi=AdC7N3w+F>;Ei&_;Txk^?R^EOgn30 zHgv6d_>fGBuC<@DisL1~{C7>LkaU0QS{pu` z{j&ENqh~?a+UW?A>i^WWcIfO39q3wn8AVe5hpx3FB?g(mh&mQa5{VJ@Ti05{z6`jw zY|Hx4wbnF&B>9hBYhUKfoDN-Uy`GU|zjv)Y-CrIL8(0S*nek7iK&~W41%t;j#?zmf9qNs{JKXKIC`$0lK%gH zQoiiyd7wXgO8R9)?;$YOzs)7_@}p)zq~JXtN<(# zF-!8r(ertsOqr$H3-)eY@8w5vC2LMy5-+#dePwsEU;JYuVT|cFB1x2%JZ4$y>3SCC%gTge zt%W7tzU1%Sw|9%Nx_q2BAG>gTQc1kr;;wTN>oV=_4txhIFOwuuTAF&}+-12!oTlsBtsuAw%Q~4H2qO@FMmd`0S`-5e@ z%R4h+N!}o-OQo-DzGRm?j!%3$dXRXzrE8xv_z$UPbRSoKD|(Y8N=r7g94q1J36}G( z{H=r~%0Xh$pTSm4#y-;YfHAf^gT%`%ESgvum|(qhIe%f=nVJ@DFZ zk6tV+&YL8b_0QOR=`uOT9qZyj;^mfm*=#=yH>c=%(>O= zg(X{NnQh5BBP}L|-U5rpQW7t>#A>ikuepmS;g>1=SCb@4OBk~Zd{fB{#@MqOOScG1 z!e&Wb7B67eWB1#ZUI><21|(i?>GFvELxweYgVzFLs@LaoCqjU`b6R@p6l$=6#$mUQd_c_1MK^l0<3IJ&bdITa~l;W$;Gbzj_Oc z=N5@&|6tZ|7P#1M8(4n!C-HJiVg|c^33hIv0dv2~D3U~J8OAJ~r#g*+x&$2W<1Z}s zP7+IvQtYYEdw(PRLg>B_5-+z@%VzUMEBn}Iu(*bkBua}tv&{J6fO|y`znywUSW;z{ zYWvuw)Z%$Byr3=%ACq{w%cPf35l0m>ia%NOWu$M*j14GiX>55QkkX8K-i*mNH97yxel3Np)Nzoz&V0_wk+f7fGVD)Hs53e@xKCR$#gQHYQP6TwNrV z7Khn7j8np|>tKohMB?R^%@^2pa+PH|xL0)d0+K{&8OSVlm0csj(y3=sny}=`EbBWn zi$~Xsu3%X`ki^R^HV*6wmfnIwH<&LrBS{ja#gZwLRs8`t#%ef{BuYyFv%E`shi|F<^=zm{>A%;MBDPB`?ccJx zBn}#b-y*n}O5){~el7TQ^6X+~pe_xwND`$biCMnQ9@-V^@>KtqmaurcN-X_)v9*f; zCqukWPHs-(<(B`k|-^eeQ@skA6#`3EH~G#t|=@IJ0zB=2bkr% zqb5GVir!1&l8^(9G^mZi*6ZF3^Nao5}PZEInP-6^R{st>!i44=7o zEgZ+L9wqT|%igbbaf}&utcH87-3ug1lonrR>3({F35>CziB4UGC3u&_5;ul*;H}tQ z-3%;Nvq-$$vgd{w)+K4O^Dr1=Qx=jWN=qWMY%e@894z%Z)-e$lH#dnT&5ylBuyV=y zXJ9etN#f;}ok1>W*?8&bW3Vh7K$0je<&NUqKN}p^5G-}yT^T7X`7%qx{p^@*Q{!nL zu-wie@p8+}1Lx4kT_?Age#8$f zwNpsE+%o?&+dInkd@jDBn)Z$)QCb!-%d`wb7qH}ol%69jk-H@pTTiwQqv2h@H(2go zB=K^~OT8&rmlFQn&~pCEt{-@G1j?b?p|Su-z%xhi^8?|Gs<0yTn%GvnK6l% zTLMS1j#ZPTU-16L+Jq!gT4pm#?We)m<>^y+;Ha>K?vq#w&Dr_ooXuTaAJZ=7RTM9` z)W66!Yqa>Q55769Q;8%|TJ|x^mp1+Jyx%JGQn0Xi?3Y-yoY*a4H7^gm-=Ea%a#6h8 z;^WV*Esvkk!<{L7CzB*fODwZ!>o;2oV{Fxgun1u(lv!T3W$Vdh8h5~MLZj&o@!d}o;eyi@HYF9EdSJjmwh-0e^krmSuW6lXBI%x{iy?Q(NzunWtyuy zAA}CPs1TCs|I~q(YJOx9bl^=0Cn?n(cnhm;K<3TtIs0LDRJcYGi4pZ%2i~Pf27Ks# zr8RWmeT*SV{$mH;;DW3a=)fy|k0krO1226R>zlJKJa!BQ=BLLb!5=#CR&>+F>3V2M z+6(Bw%T6Mh{%r@|Z$GfgN^?`_zzcpu(*AP?Ue~S#*n#&X?@%A;z#H~~B>Yz$cun=p zumi8ZtMM8*N<7acIaNmq`(5lP5n0FeBy`}peJ5%DTL<3oP0fbDnLu+~sc+(lQI-X zL%uZ?masz-%YzK|WN}D|lFh)9KZeB1Eq6z-FH-HP?P3G-rL+}EqO@2tORXWPE5Oo4 z`-GvecpjEm)_bv4x!Rgl@rpLC9*LJ*;_k98Pivnghrv>!2}z=~Y-5(|#@*w=vT57$ zp2A{(L}D?y$;Mbs-+TClonam%UT!J&*B-p0%^h!96D$=Ekt9k>B(vyUYx@-}kKVK% zBrK^ii(M+aYnmUw5btJpd?fL5i`FI9<*D&LuLD@pzmgv@EF&mgnc%n+Z#_ zkEAXOLfMMRfm8ZhVZO}#LgM9?!4ZL2m+0^hE5UN1RfnPn7 zSX_N2mWn3qZg%DJ`LR%!C%Z_z+_GvPdz|lDxNjm@%HJkQlom&3X<@eK16Yi*UMv-s zT$$zS4|d*v{LBT9WA06pisI##rWe@x#b=7yWw1;iN0KNl7nr3?^9oU5(LFuUURV;3 zN$TRP$LgXr@zxQj%b*AnFSm@o&i3{&>$V9y@2fC}*&D~((lXF**KRV7K3mU{j;_g|m2z^?`Q_SxhvEZ)Z@ zmbcby#UyXi-cev_U`pcUmV=)*;M})PuAU2)8D=Dj(lUZsdUy7~>*NxvMjjFt2Y-o0 zzxe|E8BaZ`jRecy4J2M}v3TDPlT7V&$^^*v^I|iPEx~Sti~%ehe&CPGn>V%kz_xx|ptGdlXe{ z(-ilPvP&WHa?7Rl18^boFP|T+z~Wgmv*<@CEhm|!R_iR>&#(T^&UwNT86>f2?qlbd zrETiv!WcW#u}e|B+|sOm9~@&J=WW70BWo-mNtBjkW-%=H1FvDuwoKG0`}aDGPq4(Y zZxy>9YcaqJ@AvaNl6bkL#Kx{zmt9l)#zI}vdXOYaOVt3J`)d}@?FD13UFj|5g~jod z#F8Jt#@M)>u{uzf2~|kE+%jP?+uNgc^K+fSl3R-;QCj*jOSQ5N*co~8m0m4j$&^{P z-(c646HC4*1(x9%BwlWLc9h+}_+%_E1(s2tNfM=HIkOz^dEW`@vMBRf17V3jEvZX` z40cS8X;x|ySjOa$c)8_O@)Mjd0gVo=0!!f!l0<3oW0oE(SK(S;XuaF*ge5dYVwpOV zb$Y$|;dK%$HJg%nxy7vv+c&PH%Y|=Xac)bJC@oK!WkQlZp7%p`PwpWs9-$J;pw+A{ znl8Pqz%u3#iI-bW?qq99Nv%)dw>zpICrOl+@_{(_J<6Nmb+Tp&jX}avD6?!X#a?Ex zXgO{MSeBM2@p8+k$3t+8`Ih^15b9D=n>-GrQxC1-z zOe{&f+@c$I8OPZCmCm+cX*r!FQCd8iWqaM3 zf+gbyiI-biTE4})Y;$+QXM8X2kR(crMi9<@y(ukbgC%PB(2K&7DzjK!Wn)ZR`=c>f z`g@ajxn;BFPP9~L)V>RhvGsl=iPF-ZS%z%#YYUdhQz_BH5`9rp7xVe-{^jgO6I-yH z4JYw(%lwt>xmddsKd@W9dK5{bv`lA~4P{!)088$V)d|89a7kjha)j+0r`fyqbg(=u zy{0H$Zkck1b$E}uyL&R!#kUekqO|O0mLq|l=fL85v`@0IxL%f6v~RI%n5sLLjsVN1 zAQCUPtg~p`XRubevgJxInKotK6q_XzD#znB|FCl;iP93yEIzY(zJunVZQZwPmizmD zKlgu@KkOT~W_AE}PCcE!@jP^`T{a<^6kTgIt$yLuIo9IG2qvq>bgiA5+YnC)9=C13K-b!e%_P&m?OOZo2d<4+)eXAVs&6A{|G8_; zYgmc?Fr&9s_-P7VYpdKz!hhAZ)_trMcC7_gYJ_K^51u5a@+eWjjuLy4vQwdJ&GiUL z^WVDG`c@y0@5Uug^Da8#{MNPhbpbm=UvcP%@A#HnK;q>`PxDr7aD{P8uXYvSOklZ; zB>C6>m$n{Ob>K{3Q~gL)ar8`#l$;5E8nY{>c%MnB$U{^daZ6CT8zvhtF zm?TkJS_b1Zsc|8HJy_z5hu0I9h$|Ayv@dLh@xbImWxAc>b-^0xS(<=g5mSD-G% zMvx>*ixsojbxOuvofG=+Z6Pe)S0$D=!`T=!EZln5 zqO`bj%a14xs7p@$o}Gln;hMzKZYO){Q}OZW4q#a~iNwn-rpp}A@}{q?4%9{4og`6O zt}=^bo0(Xb?p5M@3ro7pvg00mdTqbP@ETYunv!_A#jX_ls`t+Q8g;;uY)_IXExF9{ zMf7C2s5g0D*~6~D0KSYMxCt-vz(Dv6g{mMlGhmOJwo^n$vC-6Bbp7IS79 z-N?}g>T|1@E7yCJbuNM@Gs&u>SA>w7OP0S+*07-ftG&vs^gvc&~+q<(xQJF=YI0$Qdk!) zkG!qI;u9mW=sU1RjpqaW@tFMRIEj~A&b?%FzqaG_6Hu4G9+MbTE z+Fe*2Z%HgiHQCDV_&w9#L0x_}Bk^)e;~S^2F2}V(@IAkm?MM=(#erG;)}~&7F}7~E z;Xz@^lv(l)v9~)O*qdM%&h3LFUTz7B3Pwv-c;O|e%j;t#iPCbGSw7!qUkbZ8daIwX z#NU?GW%EdO9P3v3t`pSdi4}>LTPD@JgqC|}w9bO%!Az1wX-Q|6SUrmaP?vhwvrh?2 zXspCyGKf7T)zy6V6fD~DBwlXmzoHD@zvMK`Y6g~Xk4X}xrA`RW{ZR!bxYwFj->Kok z;t?mYq-e8i%Tuup48hW5D2bO_I%|4kT^0r`aRkeNF(ir7V#+M@YWv{Rh0bPYqlBeU zX3_4)uK6ABn^y&kwiSt&TZXngiIzGpi}CC8w`Yy4L&ka)RefyO?p%aoO?T)?7fMv^EkDa_KYLpcku*wxd0Ei8WV63d%o z%yKZS(IyyU*P4=exn;;T{xr_32VTPjwk1iF7VS`+`CAd|i|&EY9~N7VD4f z1%RauKI0pA1Bx#$ikDkD4qc0Nxp1NE5~xeHawLh;V!|v1bNk?zk9r@t|3z4`WtL5r ztZVJy6W2;mmy5n6UT!J3;SpK}BxTnF%kUtQL}^*cEXB%r2f!HHKd$00VM(|zsZ04b ztlPsZ*Br-Kk0~TxZaH@5ELwu@-NX*O7IR1vrNy6F`er;o1Qxw#^U7%b{d6Jhfy7dC z6FZL8ei4P|{pxQ>yxh{a!+Eqkbc@6e?}r~q5~bxivmE$#tR`3vU5u$BES?Dx%j8RJ z?zi0Thx;tfxJlyW7O!0PZI|&YX5ckU>|K&XX{mSy=YGJ^JR=SNusp$VwT+7L%x7z>Y#}Yg(X#HalOUv z_vie`Z~=?WXc8~Cd^KEymb1Dw8-wMT1xcc`*f2|{J~NJhWq+4xZG|QJk)$s1yu*9U zo%7hS>e7eA%PsdGvm3fW&$J(aWuyDDq938O_%O?e9tkhNVwf0VBrE}sC6)v}#67-xk0`dqeqI7V0!pGfLbBZpbm@2+SAb-Cn7 z;^mg>jZ?8M=3_GO%iz0QNfM=H9xIvfs*LI}NmwGDN-RM$SaZYJgOzYUztWZ@ zUT*2<(G2e^tNqY933V}?PLe1sUd)mj*&qq#%igNTXA6t>Gl^xMK078)(yxVYL9f>( z@p4P5;|r|I0KYCxz_O+xNuspGGmF`YN%*zX_{pCa3yZ^ZiKWq2b~sHxzXIPLbel`! zaPE6Pb;s-Ej(5lYB`oPOOM*YU=HIu%KMUrI-V+ip zw`?+H>o6a4-Pb`~PQ4;Ylomr~DW%mk560MGtt$?~68l0@7uS#Ml%4yRbFbc7EOPErJrIEQ;dg z7QJ=sz@Bcq9oNRSDv~5h%PnT9|K(;&uw>M`>n$w#GD~t-HecSxj>C@Cc1=mV+*0ax zEY@X-f8}%-W9e;35~Zb(SsEKKWo8eSm|rA`($YE%k7GU4nm>oST>P~%L|7tUNi6GUvn3RtQu|F{?r$u+swiG= z88>t=o+}@Z>X`!;=PD$L(lVJ@j{KT!1C}>o&mx4yCq-g$J<8U`i+|XM-?)2umBh;} z-Ah~H7)u+P6%Lkrw@4DDWhb-L_qfy!ERU_5-w+nZ*Ah$1NVWzWHQIXt)TQz=5-+#R zn-GL`x!!c*EwJRRCP|c*C}zn?>(Kx#;~lr&6_!kyrTa{_cG0YYChkf6bq9%;Tl$S; z3+!R-r(Xxl<-H_{(vrt4JHE`F2$ou9KPL)He5#}_(}UTLRCbGR7l*p6txDqM7W?5R zaEyJ|I!A!`@ko<9~*VMQf;spULf&ui(TbHv_wo@ zgwMsQUL{GCmP^bs%GGHWG@ouNUZqO;zt=7IzWrlY^z4A`*nnJl`#w|Xiq6p@nG{{o zo=%Jv)3V?!bVZlbCRzTeEBd#+Jg0opmWX%q8+`$~qV-yklSNwVL&qTl>#fCs z>ECum|Mmk%ZtoHRUD2P%k+lEZ6}|RkS?r3gV14}@bVcv9A_@OhSF}gB+1M2wFk(|v zIC}q_L2{~&685p|C{a1p^D1;jd(R_j{##e{tqgCtS?Uo^7W|0ONBBX+E=UsBURSfbJ; zmM(SK8G2TDC0voWUPI#Lmi5o~q9tmt_g_$By^SP^(o+8dPLtN=<*_relm6-!!s7Q% zVp(s&RvGGj^}@Xg{hN_^xux|2w!)Y<^w&qQ{MC*mQCiHHWzMg9SQouHvpNck^LvS9 z*aUW`adfjOeuehD4T+aq+H7T)lGBbY9t;+Xzep0LWfQY(>t1yY%$NLcCOw5ETW0B# z#Ku@&FV9CXUko${6vfLedM0e|sKSgo3nJzSXygX7sbmhb;EAreCeiTk`IoKB9GzT}L}}3t$GM;X<=922%gN@mt%Sw%gT!*) zk#(&0aIEnXEP;k3UT)b{$li#!)hhTM)TM3@l0<14#w;DeS|x+!*n~0G!eXB#v6R_6 z6aS@A8N+dv;mb4P+%y}#ETEM@1BBua}tvqbgW)fg;WkEg8^mQ6~>%tQ8Sz@u;#OD6|Ui!F_ z)!+h&ms?iMse+ard5+jkSnn!HqO>G2%i_0fw}Zv}W|h0b;+-q8>>S3Pl(?S%uoK4E zuXqwKx15V(Z>iayORENBtnCw$L}{sf5$C>@Wfk1ht7Ta6C&J?JMPl*(%&wDfJZNzb zESIj3c)8`q8&=%J#og|M#VLj)QCj*i%NW03KcOzxcJ1E?OS;VBUzQz{->(V8`M!VEH(nBvD#?nZ^EQC)}qbx5lJGVF~^!u^c_go|48^-x>p!*7_t~ zZfRMUUBkTJabyKpbeobSN=qWM7@zre1IF0C%>0rS{$4G0%a>TzxU=gq>$N?bfWKQtgys1+NnKKG*nHXEr}$v7cn>4- za?9%G9%wNNHy#aRtlL* z#`FHRg^yl>rQ3XxL}|ItEE)y@xDQs7p{=?Ii({e0a(XR0zsxQ;WB_Aq@>mitx7b{D z!Md3A98?i3qpe61rA6~H&i(LjW*eX`)l<&*6_!kyrHm0f?>~I3h5N=uWsrEe#jzYq zvUK4dU9dRhk|at?CuVtcVU9gmD#uJ5E-djsBz37cgw@4PHw5?eI(MJM%Ppo;HsBaL zR<9a<8GPtdl0<2l%`C;Pn4rbt%wS7l3H>Rt%zeey$Kqc2;Qq?~pGdsiGU10SjmDeVn332xg~gH0PY>- zH$4bD@U$Bn7ySsOC6-y1Of)Y8mNUs3Ho{UUvlQ5}p{`@&iyhu?b&D0n%Pr$e&c?dD zy&RA8Wx!aHL}~fSEVB%DwE@fG5@u_JCE32z@6TWER%4gz+Ku16h2z+{@+4kvInaY$ zTOMj_hTn*oS)C+NTG~e9aco-68YRGzKf=mMSfbWREZtVHC-JVipR~ZDJBh^0E&0pX zS7=R7udM|0CEA@NQCg-li&J9P4q#cf>o0d<@mnvkIQC`tm2RsKtp`gFFA^`eWVhIf zV=N?n&}OiB93x4T7B^;T*tQB@TONOQ*GE{KH%Kh!yRt4%gTF3+083ITiI-c#?Sjza z(W+Zju>8s-NtBiw%wqMzLJR6*exgi>uw=_DYxlEb^4+|S9l%od6^WNy?oN$D%gu(} z?|`Loow7wgLTSlomZ_iiMSx}Kp^QjjNpO(VrSfPtU#d?oj@SHsPmp-IW!w)|mnla^ z<8jO=gd|Z~nqR@We`|a5!Cr>M{80<^fO@43UPGr>zo8__f9&*n zmHA>hbb1x0{3se%zju07iEDue@*B6eEro&ka_g_61b^uCx_8?UGetDC7zmwS?RSz) z|F+ZXw;!mNwa*%;9z*y&|6YGpI%^jdp}B>Y#MUa6gzVy9P7R8kxq zC8``JIh98VgRA&2Y?RVYA3D9Ff=HVG*6G!Df_5OB2}~FJ)fGpa=>HvYe(UrabC<2k zIo%1L0AqdYQ4%jddRn%8hKHF~&2Hg)es?dCB+CCwL-v2!mQynkEUwjWHx-tEEfUMO zY3#14R$;qPu#DCx@p8+|YwYU$mENkBP~(KAB#F{8mRW|^44V#?;s@4t5EfS_iRHy7 zb~n36!x;Ph8@(g(a?8$`<5(BVIXYLtQmA8J^dpoOM`lSo^nDIk4qP17Q&@6kmN7rs z)w%zb14E%MriCP4ZW-!T5m$e%mwt|W6P_^~UGyWAmJ7_1+}sY|Sxg>0YN)UzZk5!< z(w24L)e7=j0hVAJ5-+z*{K1|qrn@KOmyTC>lO#$@CbJxEydJ;Wyfd(>xv)eyODre( zyVvfSzPOUrD4oR1EnN+!;uyPV5z!IGm{)^pML$Amsdo*JiIzL_yTTZ|&?aGquz0&j zEQYbHLt?S<5j>7H?@Hq3me;SN(Q>W+duOn;>qC+#EhCs^`Ifuez;b7v%Ti%+*e0>0 z3}xMfXA<{sgE5x0oW#p5{nMwTrQD;kTfy?ijwDf9)-#KxiS1utxv(jAt+1raEF(9w zCnXyy{Iwq}pTbGJ+)^=a7S8=UX{N=&;+sQ~C@rDPVt>5{zU%We@by+JBIsAmJ zWPR@wG7v1ET9bIW#pTu^v@|fkj~zviI*}wwORXrJ`%6m9I{}skm7MuG| zQwz?+7;Dy$#LF!;%d)PurQ_|NfyJ;TNusozWEOpYS2M7z+A-+5ute^XSh{>Kx-6fVWzu5gtkAo9D zz+%*##LF#L2C;VJa;Yvu!7`>jNusp$W0t+uf5w5uhPhsAI9;`HA+_cdxg=%Lt>e*iaqtI`?C_B_n)|t zc)4X%6#M#o)yicLfaUf9l0<1Ke*@>f$EraIFkjZhdQ}#dLYd`iF?KyxJFyM!OWGun z#LF$Nb|F}oIbHS4z>@!xBvD$7nPu0&TJB&OP}fRVSd#Zj>T;zAt4o(t(dA+8mu^Dh zmrP`0VCMD3GUZXIODvHGXh;QDE}R=J{h zxn*4UD`+VZ-8CF6zAZ=+rR6ZQR2Uzw4|SOrm(@X7{Ps&MsWsU;Os-44rcf7$$0S~E zS#qD<@0%Li<9vDhk|a@D5||~uwKpEeChu}I78YkuiDgT^9sWzd#^_80i}L{zFSp#! z%fTc)-#*1_{wcmBiPBQ)CeHo(re+JFF7NM@87wT>GD{S{-w)f>w>4PON|r5(ms?Wx zcHnig*6a`1wbrUPNusoLVU}&fcHaX_?~0*j!jf=6QkR(x+4@+dZp03-nARfka?1)o zc2T!w&*JGYU%u;;BudL%X4yO7Ab#z1jH&))VF~k+S(2EgSn;g!U}?UD#LF$|x7fPn zRsGTUMcvmfB#F}E$t-Jw9>hakx_19yEi9e~C6*6&*c0D}HEeo;W%>#dFSl3@Wa}_p zE|r-Jbr~K&k|-^Am?he3k^xwpqODd6i@mqR;`D{xBVW#2i~Gg)Nhk4gOX?%G&R_CX zM|_K5O^HcGKSF8Ih{n0!w0p{Nunaz3YJ;$($}F2qm?iI6lUq=i)sINL+>%n8Er5J< zsf_EEmvcxGrKLTyjB0OzUD0i4&Dkz2(T5~;@i%7kr9rnniD286SEl`G94KDT$X`4qZBrN%U5i z$OKE08A+nFL^I3M#EG~*Ryux4h_K|!ELT6W>*R@Z-{X84|CGebEt9p}(2`yG1g?!= zPa#Q^mI7wE+q&d!7-NC?WiAOzqK~95Bl@zppvO7v$NS0(9m*EP%PrR*_CZUBKb?og|!s2jDVrklwJx6ZcD&7_>1D2C`xnZXh=l>8aE7y{Ex#e)R&1l&;b_(u)xnVO& zqO>%=g>%12n?3!>U~sX63)vVSSiCNY?3AKhsr-3758VZZ&|15v_0`-z|N!aYBGw<3F=)lXdAej^$c$ypYu|&@n zHn)Nfybse!mVfHNE8XWb{-{2>+aE#)UWfT4-Jd$}${b<8ENjjL?7+LejHLQMb>MAk zb3F$-@ch@1l>eavFL}YpLolLTHsBsZ3LSW@w~98_U*+HCtr5-{OcWCVH|MJbS8{-?GO?# zKYAtxuv6nnZEt+n=i~*FMEQS-VE>mhC&u7i)7&HH`v}YP6OuE*v@@*Vf4lLF2Vhwr zNaE#|wh@!@ka=`*$N6wp8+?`|QChN?#mJ%jP&ljg9{6y$utWw(EFS!ey6YBRn-3P7 zQ6ye&InTGa@9@ZF3|KZ=kR(cr{%xG{EjstZZo-Mjms$vmPoTtdC!CG3K<#FEV2KSO z@p8-8!R#F@ZHu)F!7??1BvD#MF-y8}Cr7A@Q;91xg~jot#M0*^d*T}v^Tq%yqbiVi zxuxtnwm;d3AwO~V;PmPwiPGZ0EY0H9;r=n3Zk1UoESWNk?Mb$h*u~FDajNKR{i6JR?bzmb1)KVfU^qs7pZ1*LA`YA0(;E3w{gPt{7VnNEUT)E-?1h%o zU-U|WC9(%eqO=4vi<#GpV_+#Zf5&xUi3*WejBc>YkN$r3ush1xo5ag4N7qfk)t`sG zroREpgb0#EX-Q$0yM@Pbf3mM}-|h>GU#P?q+@IZuj;%U>9@IrktDq=eZfWhl2kY`? z+fsZtZgdZlL}}5E!@2*n*Cl+nAEFb1eW3w-L!jf=SQkUgXY(I<>{U#5Ex0mA_ZA!p=!7ngdyPRJc(z?zPrs9f_A){JOKZJ7Vuu z!~J4+hma&n%X4NqU$NG3s7vokz7>VVGfZNsRD(Ui3e|J%4|SPRo5ag4KjyF}@h6{6 zY6EqNFd#{kmWp?9?hm?~d=)J7R$kT-7W?xO%P9UNK72u^3t+JfBJpy|(__Q%nCx_Y z7`_wVC!HiwT6!^y^QIK+B<>R&)l67YWtL8D+1$TV{rqdN)Ot_i<(9w!tS)VR+UrAI zs_R`Z`VmTt4YN#IZ2bf3;+or`v#>;8kkqBrP#3m5!q^Q$J$`w2@xxWr-^%pRg2F?GHKmVxa^yxh`m7JEu+_2%79 zu>3M7NtBjH%(7^GGVWDysg2`EVR4O+SSDO!$K>o;VfY^Tej^euw;1$dkMm8pJDmYb zP;Zh%X(@9T=YA=jao9~*Y<{Gbu;j`t%W~MhakZ}0TMKn*w1ULTEd^!Set!0wX9R-f zogGP{v~*>bAXQtlypQer#f#s9tl@c;iFU7wg%c6S33FSkTa zyN>7m-lf)!2a9hDl0<2l$1Hohy5auH`RxN%3QNQ#iRG3C8)FH@H(v$I6&n&Sw|sca zzTozw{dU|ZKHrBVQChs1dVU*(#rv|vGHnw(S9Y<|)d$P4W+Yy2`S!sR z>yl&91Ha(r*^VSpTH={y)Y^p0U}^B=-F9Jdh?H0=cV)-qg9e-Mjfmsb-WSEoEjo^- zXn8jOoh6L1K|@FqrKNa0&i$9oolL<}dHKx!!jdktB+X&R!aMN*gLtyo=(-}9UVmi#d!UT&%1mkrc+ zBMs5g$BHCTT4plK!cCF*=5W7fC(Z~<@KuQ=#f@DjKS=NP1}x*X1B&A17Q+#FIL6k_ z--7$a>I@-Clok(Wv0geO2Fwz%qXBwRmC4msv_RX5H!=ZPM}DvhPw7FSq>FvOHdojjCNT z5-hD&lO#$@A+w~06bu7PXj+4p!ty*yQkRXp+4C3uj~B;4UCQV;EQ*&~GW1xtM|jpx zTA*m?W{ zSn39mBudLpW~toNd?$>tZ=XArtorx;zT-`arQ=!F8QI`-^DwY{Yu&jhUT$&ju>&n; z{chl!Dzm1LBuYyZvwS+R{3uv1uisQfSTbc6+d{UUTz~AS(qJ)nCh>C1#viPWuTjI* zdQcZ*Hra`@CGY!bO*Fy0tCSFS!GnG~HXj>o=Y ziF8}v=>(lCw_cGf|G9JJQP$aB(7E#P9ZC16&Xwz}*e|>7Z#EA0j;i#Dr20Q~u4wIv z-VB{9Hu)r_x^rb=)eZQ+uiMKw3r19JjU`1@6eH@l&Xs5V7%+cp9dqbhNhw8={Kw9f z3biUlLFdZxr^}0e(eIrr8>e@}19?!1H&|B{!v(`a4 zN~E?TIh99=O84@kx6YO7?ZzC2qvwa5?8f4V6aT*>&To51eb~e9 zG^S0Swh+d8VIhf^A3fjvU{@QjpN%XH|Cd@NKNbB5<^QEA`@h^>*$A%!GO99KWHZZ~86TE|#eaXz>B15fE3qW>XDf_N zdk0hm%Vu8^FSk@E!(QgN_w%hISbm)(NtBlQ5Ac|{b<%`juyn8LZ6hpxaT1HJ0qZDg z*WJ1hEVgw?yxg+%2m6PNA7hsXmJ^Lg5~anAS&ZhT-vP@wkNxX}#rclJQtmOk)0kA| z<}CIO6^>6*`e5N=q2CoV+v*_gvZSHq}#D67EXs61kq88W!um z#FebY$4R{067ba$>vGF)!etm^*TP5=rR6=d+&JA3_gb6aUFL+agvCoN<^}9H)};Rc zd|&zLaS|`Ly!eX^RHsra&B0RoG)bbg=qBLY?|SOACX6xfYc65J;(1SEv1!9rvevcT zjl0ut|4HKI7K=dkHM;j>w6M#wamQ^%KSF64#w?4pOXJ<_&ofR@!eW14Vo8i;=gLEs zzB#~r=@&`j<(3f(9C3`z>$=|$>e4rwBvD%InWg9UhyhTS&Wqyi3rniZ;upt0#Lz*% z%^|QD+md*>CAv7I&5#qd!Qznf`K7Q#KakXA_IY-G(JTG| z$5`*mxkd4E%cWi(SeL6)#^L@kjx9(Mr6rYF;&jI0%I~GBS((BTkRY+x*s;ZqdWL7Q zYi&_K5-+!8oT-h+ySRP+s9e9=Y zpErOpWY)DGE1E&Y>eF~ zaWn}meM)H;#mg-gZTH|9OR=koPf6n|k|autEwkh-4Z*JH^DZXUgeCEjq%N6BEAVHe zMwG1$mi^HrUT(Q}&kvKh?tHfx>T>-TNusm_FpJ-WPMx7HpVNOB2us9ciDlPL_VDRo zYL$av*^xuy<(973*kdr)Dy4P6vb=yKQCgCi^C3VrrVRJw6QAA0o z%g;w7UT*Onz7H+lnZ@u6p&ee5Bua}fvpjrL_CD0*#cAEy!V>&UVmXk*x^HWB3h)NY zipwNkZmC%PC0Z&u)sBX`OpYN*l$JzhF)y+7D_ER@pDq;^x91W|>^`>Ecc)}lIj|&L zAn|g`#e=N-cAHx=cCPqbB}tT)a*uKDTNJGA36|O2ci0O{zRY6$iCy!X)lb1|n9C+4 zUT&G3#-3m~m7UiK>f$+!BvD$rGt0R*2YP^I{IE?f!t(ruq%M7)v+J?znaD1XNW9#V z&|ntU<>QHBx?pMQNs=fnnu$2~7uorzgC(wnbEL3j$}B&O*)jR@zN($UqE&&!%Pmvu zvun$xuLgAhOWEoqiPF-ESr!yKUk5C<^Cra!OZ+QIUAC-a*Ou3JSHL|Z9rlrUx#jH> z_V!@Ay_bmQCem*OO^6%@Y=FRw%b!-2~Ck$>U3gt3A*iK4t1HV;a(Ikw|rmd zj$_QtXjeV3j4Dr(C@uS#W!o#KBT$!=x5M8Gi^pq;C9npY`vcnQ;To*}ClW8W7UV>f7o-S`O5RyJ+bi6lo04#aUDuB zDLPkzu6@DbcD28!Cv>ii7(=rB=gyU|j1JGCb47n5N%yDDm0P>nFDrh#-eBll(V0$C z{hvBlCWIf3ht8GF)+FVB=v*lqYKc4D8Z04+#EAN>b7g-J1KO>w5Cfeny0#?Af9zZ- zdq>w6I#-4Vgcgmf-#b@a+VsUq)+Oba6%5Q*p=XK`{GoH@(JfQVWaM{mChWN~D4b;a zx1B4${lGWdtvsP~rSK|A`_G*#JLlKN&XxBG>3-0;vMq)r{8ybTI(wSnbo%~T7yH{P z-6J_wM+y5C>?o0G6q*K|E6*R3H2{ z^6WR%eFSHMhi)WZZn<4J1(VqKHhK+~_yZ(~(sF}YbOJWL2TRahKNDfemRTM(<#)3t z`=h{8?g@#PTVm_7FH;Ql4z&S`UouIewB$3(+Lt{)g2l(?^k`v8NSD;*Mgw-&v~}Yo zN3hJ;PU7X3j0SJ8F0o5GPXo)Y<0OgF()=kNzS2$=YY3L5woN7rOV~S!C9*BE#5-ml z2g`8>5-+z@^?Gxn*;&qga=&+hcctrP2kGL}}T^EOSb^;}fhr4;x!yvCoiLPETTUzeK}FxFWy# zFo~C2mJfGB%aYI4R)J-PKS`prL^6wgXgPc@cF(fLj>3{EvrOe*r21)68as+&_mg4!@2)3$)`V9j9T12EGz*ZB$iG|Z0=_zcC`nK_b?JK zw~RP^6YJ7ySPk5v`twqfL}?kzEW>x3bOX!I@eaYl;+iF~bg^J7CjXDKJ8|o&`~C)= z3?=iNQf4yGLuMhF8w`;tbIF`}N|GcAQAk1|iXwz4Ns1&2A(a#%RFVww+|G0TuGV|m z+kN}~0c%~K*Jtgs&pCUqb54fJS==RUJ)OjhEsq?Zyjk1V=Sr}Ng}meVwPutmi@t^ zzwg)`#ghC%P?shn*rWMv-f5qOF}636#EUKV^#ahc+P@L*_(qjIUh*TPmKWu$t^G!^_~Z#Jr?=SP&!|5w)D`LyaFfJ~Ep;-dW0H%@j>u_bRf>ss3ql4<~!d;LigsU?J2#xD=X zj-p=G7i%e&xQ~LmbQ-e(|I(yaOU8kvZZ8rqwiLh9!1GJ3UfuE8I3F95L~41-EKXf_ z-hjFcGr!PUu>=(eEKA<8E*ygcgK-{9nMUHpmRlEg;U8krZXmvAQEM(qBDH8f!(+c* zM_&W5xLozoS1j(I1eUKc?0E4=>R=3Y85T?8#g>W%Y*~M2;5R()*SJrTNG(H|Wn5n6 zIbd$-b(2vyU;}S3b`pNu-u^W*J>)`c$yA zSROWCvG^7VELSwx?QLeUj&vO6vWQykGMai5FYak{+Vvrop)KV43lfB#~MMGRyweekxEG zlgpW#6-$oD;(v@?^LrW}Xbg3EkVE3dmhRc?i3r^vb@5#|#&xnveuUJrj9ILMeDGY^ zFxJ*xu_SyI)TOKuyB9MYeH_2uII;B7l6bLYKo@VUi_aa`XsFAzt|W=na-3N@Og?l1 z>auvQ@gc<$@=aiQS|DNfN2$3A2=WU3d{J zm)h<*rC2<_3oK5<*=a8FeoPTqhV>-zV#|Rc>>Wk1W$m%c^W7klL~5y+hR6P^m;>0g zw&vdLtBU2j$YOt#?JRCgnh*n)5l>0H*wUkjbrOH_?1`O`=9wgk)Y6AplB^OtL0wWF z?!T>A(u)OkQ6I|cqCU*}2v~B{Nxazd-eV+=u_>RHUV$-oFPkKhS{5?PGc${gU@_KS z^GLBo{Sa7Y>$9EsIk`G@z|x^9i5FY8%rL;~--HAy119A*}c1*&_%a=Frx zOvQ5ir@(U3k=^gBE$9{v7L!pVUTg^)avJM0((2+uuy{F=BvMNfvy85LyB}B8gJy;xCk|a_~xpX}C&!&%>0+u;B z0cC3cy06R?Sth(?YvZj}kH@u(_e}yz;>DJTtX2j?D`=G0P+KzH(#CS}lDAv04Fijr z1xX^cbZ3@cM98C6M|g@u{KMlKvzN1Ad*SaRbVrz0Eb)f zNuT4;Rj_^}$?|Vq1=FsE;g3>PY4;Ag3Tlld>HgGJ@O?J>WyZ;SbD^uiY!XTJKXnz% zIdi}Ux(a^GASwT%tKhM5mzQv)``VC1%82@{s~~?B1CDxJwHb63I4&Vc{<5pU)H${n zbQL7ch%Fgczjqak`Zg2~ved)VV$yGugff(}AvnBRxo_|JqgX z+YfB_tnEGMDtK=|(*Ao_!J*dmu&bcnKD&OYcZ1s}cp;NkRb-ojj%CB7Sz zoU&QMZZexC^xqB`3S9-q$B{ID*H!R-PvIw+J&)hGp`*+=N&n0^zr7Em*8{d<(!=HH z3>fQeRiaDc#k1!cbJkm0z4C%B@V}&WB}ruei;6Y-Uv%F4%>_%hh34%POQ^lTa$v44 zmZ$ZIX1ZW$d6LA7Ev3&5#w6KAokGD9{hTC`TD+Jg)WT;WSQ_4O>8n`0HVQ23uCnLI zoqrlV2Fu=uBwlQ3HhK?QYCq1$ZxR=!k|a_~JhQB*>{bpeJ9o|j)Eqj=y#?4gx#_Gl6^X(Oj+h&2Kc_MqB?BuB3 z_|1Ex=OkWi(H=MzE!A%LH-#}~t+}A&M@TI(%o6%Q^AyzO`II|56-$B0lFH}g1&+5O zpe|+Sl6bKt-O>atK|=~^gQe~gl0<6x$}FSb|1bs1%$S^miY3`eP?xQp+0NoLs|I)n z7;}%ri!EV4*s{KM!g*ZDN_z-l>UZP#6EFB#G2AnOSmIdoKsei)(E|6^oCvz;bF0yPK_|Z-&pt z^@$_#VoTrILvf7lzH!AGEJGfUBvOkTv%K{@gkN#WDe7`Vu{dlMSQhHBd2C;`;X~ng zS*O~yBwlQ}bK^N$viEgv1(tdmB#G1#$t+i!ea7c6S6BOdPqAc+EN<`EJa*m7XDL`3 zwS zi5FYeHr$J2>{7k;`2Ltrhe#5s#hh6Ve;LybES{Ur6e$*WSAoUh0$Y_c+B9()SXx~n z@nTE;I~Q;s3pUs32z7~hN0LY_&dl zC-Gv-uI657d3pLec6%8AAW5W_FlNbcpN$sXPCaWVmXz&+y69D9J6PeTD;R?1%n%YU zwmkMbf`7=Uz-zcmdfbF0ky_p}OG<4EeD=C|;o}C1CBjW$c~YM3V5v-4-4!g`f0B5y zW$H?H@zBkz3_gRkQ|(E~kC0j#Wa6=}ny{k|jIqmmXS7xe#Nipng+jWPJ^8<^< z5fU%9%&BCIb=kP2$y_*Is+}ZBq!v?VX%N-TA1qo$y84R6X{W$aY9YJdKTtUYcNR_B z*Dr||TW*9uMoae=_3^&abSgT;uW^O7GSwVY=b@5w!I2WwH;4kn5vVV9sTPa3jI-28FF3So@Z zs6yh!mZ`@!q2=JzH#jGEszZ`UE!oWSbxe2My>8P_bCO~SaTizy+-3Ls6B{=z28+`W z5-+y+8%Lp~+e#B$hw(HaNu(C7EIjs8ciypvF?OoTKx@U~xm#cg%;iv3wU<`t)IS3C}cm1eT!FBwlQ} zG;%juQU;#*0+uPEB#G3rj#<>sv^@iLv77d4vtmi#BdE*6PVBO9mfv%%%VrxAFScBZ zXA}F(HUYt4S!7F+NG(CklDV!%Ww6Zjd+nuIqC5naCL7qbWmxs2)1WSWPLp`CrA**< ztV`+s&d@SSTZKgyslVsMV9zzHpVm!TJHnPw9O=5Y{`7ZI*A=q z`{~0NtKdbFNG&Uv<^HW*nqaxpd)Gb1lC)1ymz@XLHUG2N*IU3cb_j_VTl%Xt$7}uw z+hG}CDKH^Pq?QxRQkXM%Bv^*N4^34pq5B1vg}3c-cv$Eh%mmAZW*1B1#g@X0>^+O; zW~Chf%aQRUiPVz9ELL$w*nN9=-RU=q#mh@z8GVpl^B-QT)(I>?iSJ%OMM2v^XE% zZVHy5+a!t9(vMlng(s{5OGAsgYMQ_9kuyb>^n7-%eCBVq7A$jTka)3WN&Zo^bm?S% z11$4xND`@KF|%A6)6^L@)%WkzsjXN%z5m$lp*ai8 z?d`GY5OjM~_a#~Wt=mKAIcq^#b|L8ubbEY0Nz(nP+v7pwkNC^VKGk0X-5&MMl2rdw zw?~`)Wj;W+N9HAx@-MnQ-Z>@DgAp}9k|a__)NkD$)w(m_vXgJsq1z)UmL&PhZjV;a z5+^{nN6ks~OUBjj-5!;1n&N~}Ny8%z2B!0^1|74_ zZjax7V5_u#cc9y2!ef&5-@84g^w7m_kBQFZIzYEaS~^MipSnGUULJ>oc2CLwj^F`X|wPoT6WFO!@JqAQzVJhqMnV1eAcaSZ?JUwVLL>z6o@QSve+1_ z{k|@~t6-e^(voJ|w8inKta| zyhHzvt6+?6JVN5dmPLEjvA0ysD-L_HwJ(z-Qi~0DNfN2Wn^~MzRVskGs4QMGU$OWc5m*vVG0SF~ z1UzTgIzi&amYVKHSeJ!nO>t*2_&G@;wIndh%n6@cg5~?zX=@aVgRj7HWeU4GuikTM z2-L;Rio}a8<_>I^w6tZG9@OQ^Vvlmkn5Ba%dF>C7w}#;k7!b$J-|a*tw(^Apsic@cZoXL|kq zcxQeki^Pj9#a_2@P9AYYcPLoYbzhbI2&rWjvzWC>?*o?cX)1n-CFrQY;^WMAN?gJ$ zwV^I|vPrzyGPNx`ztr#kq77KWb4e1Z#gkd~?VZ#dERWwrpHVFC#{?FaZR`$E#bxXU zuv9J}@nTEseDr1buJMyd#Zn})M3!Q^qzmho$LC^O zOeXPS%d~5caEvYLv}-q5T5cyvq?Yf@Qps;Jc6-!Zu69?kq#PI2CEy)fl{=fzaRSsO z$c@B{Ed$E2j-m!T{Of=v@+e6nwX}VM^Vo|>V|?#|)%?>biY3BdU|F@Ft(I1ud>Efw z3OqvM#g-5~Bdp7Z;6_KlqVbd@ky@rQOYb_J@ENT3L$+rt7T*&B%aB2=`_{;C6?RCB z9!uiImd+9NafQ+SrZeu6PFzfqNG&^=<>Tn}cx^d7@oj-(aXKlm6Yb7g3WH}MSI#+tSnOcLz z?=guNTf%RTL(9|h9(XVIqf&CokC0jlndPkhD0~z0fza!f6-z>Zpe`FLu@1bf<=1eR zbY>)p7hAm7PQc@(*-i7lP?rI>ND`?+l_W zaR*Cp8i^NMo}HV8mO<%NUxGz%A4wv$OkkF6r-z1u<)?bJ7K+95l)%z9ip|O6lBxy5 z7+W!j#EUI;TbIKz<~d4#AXxI(k|a`#3$y%Cx5V>)k;%F)isie=63VZ~`aLsi0G1F7 z5-+yow`T7J)qS@A2UuKYkR(#eRc3jep1d6D(rEGK{)#32w4g3No!PZzhkFHh-j7;M z;>DJG@2}$+>#DLp7A*Z7ND`?fmsz}5e(wMlqww4jiX|#YV3`}j#@K~f2KYX|Rz)OU zZ1Ep*7%jISk9`A{$u09reuUK0C?-Iz zn8b@MbKbHWI^BWpx4=?)6iFhrI5JCTsBfJ3DV0}|b zyx7vZ&<8EXw)-YPUFy9hNu-uH%rbU(vEL##u5~-!$J3RJPM&(Zf%at2`M-+?KIf2E#9~)yUlb2otOVhg~ zUTg^(%qEa!zV{}9<@!UCL~0qyEWK>s;JZ9+S9_gSEOzGwma89`B_O8fCa^SjB=KU4 z*>Sd=pJ;v*yKfh{k|a`#9kV>%x92+4MeW0;%ZepaWT`%d9WR}Oj*SOP%hHY|@nTDY zjS61FY(ILZCe$Ue7D*ztoMDzR)oP9gi_>)57{wBEK~R^QJ6PA6R<`D4uxz&>@nXw? zx$Jt(Z$$EWuuQZiNu-ucW;x=hc^ND=eP%vTECHbc%h1K_>5ep$QX9dt$%4d-EqX03 z;yh;d<3L@ooS#jSNG-MB)1?BsLQ}-BwlQp zW6ak1tFDj54!ob`$CvyFsU?6}YB#HJ9yZmt7(A+`SU&z^`NJ1OG$zMmx5w$Aua3~| zkur>Al5~5_?!>x1GNw;&3f&&7%t)4h>-PBV-5zC4-#vqFj~*5z-JiNWtlsD0;AwuP zvMY3ZT%JZ!{ZHK<>XlRvK(~jdHAyM&_DFBzhX4HvqmeaXM5!z$iIfrbTeruSpA2Z& z$FV$gd#J4>N&d3iV~Ms?W9as1HR^H6xca@@BYx)uocy+bYj+O@=8{=YN)r5`+hhE} z8JH=(s!d(!_Ar`9GX2+XkKcaajfx*@K(~j-Qj+%HyFK1dYk=Jz`{s{+3f&%?){un% zsoSH23cF;oJFQmL@c# z*k!@&d9N=cmG{lY_t>oONaDq_XO+Wdc;$5U>IUrlo7#gUk^L_!sqB9l{k2InoC%IE zT+m9f_=E{8YS-BY=i{%Mec`x$wu;1yE!~pVprxqK^afDlOZFs*)Kc{W9wx6_xX%a6 z;t?ZyC>Dn+0!!Of>?+XU?Z!r6X)uk%i!HZav)wqea8rC&L7nv^iPU1iEarQf-3E)R zTf{KMk}a~_-^F&Y)>)+DTQ|?_C-Gv-?s)bhiu%tyRzh6{`jRA4i!HMZtq_TCLEkws z#ayw(T@}RwED|ral@=sK(+)Ws$) zW~O2Z3Kv*hH?Y;8S&i%Bdwu4;BJpC&liIAfbwd+AfTisZl0<4rW|qXrsf)pK#>iu_ zVsVcUSkfxHLP)&Wa<&m$a_!b__;avq$RSCjmR`)#HMf-Z# z*gnM)5h<{2|G;*A^aoFP1$8mi7+w-Dw%jRT6Nr`0Lwt52vldArwfHd0i5Etb!4f|3 z@G-^WdtG43e8D;+4?1?jw>obeN#ezp%?nJhE>4HvTfrE+GnOQgTJAH8%EilGU^y|p z&pE~76eY0ibz%!@<~3LChq`Q?N8-hnqt7R!rEb-8_{?I{Wh9BzQYH_N{ii`?+koXw z>G_e0B}Zg2oyZPb^E2zwvZfM=7h80PUqMUgxZ6{qF6(QNBvMOPW@)kS(N52h>sou%sS7?IQ)V2A7hCr2V4W+izJ-+qOUz`FL~5DCEWKOQTLG4c(wEZ| zOGvc9Qayq_7yHrF)*Z%Jw|XRAY#BYBJ+oMU?`(YA_B%t8L~7Z`ES=qE;q%I4d{=%@ zES@(77OQz|^C|Ie%`#xgI8EZkmPOy$JXZCIlRDJpM<_`mwZt*YQ`s&iiql^D@Aqn@ZxvmLG``XsP*XF+SfvrDDsHA0f50&&PQze#fw*V5u8@ zy}n|JiWOJ}4rg=nnLEXfVA+~R;>DKY1a?u^X6qdXsLR%`B#G2Ajad#^x5uwIbyW3h zqgamL5?Ctvv3oIlb6q#EYU zORmV`=*L#%Yya@O1ePD3BwlQ(HHmc+Z^~Qq4lKP6kt9+}5wk3Awh8aWdTxI^TCpU> z3F7*(#g$nm<(xhYb*VV+#A?N2 zcUNHXsmxBUf98+r2o{aQDZYv&;GV!@ z-;ka66EF620*l8<5-+yg9L5g1)e|GjgJs-Vl0<45%Pi-AR5Jui+O;6OwV4e*8{TJLr)3Qb00EIwWH4bCG4??BD?%66d~?EdSOap`v*ge^jwWtp(5_ z5maVi3GYuG63;x?FSB}*7Y!W}6)KWc|5JxVSlx-<&>``t21zOJkcgUb0-5zq8sN@R zH*Jzg8BxD=NYwCUz`~lT3D6;7*oY+g%MJ;TsGLyfkf>dAP{|+vdxylk$zj?eSSw89F4kk0F`{$at)FN&(CgK66P<=`3-H%@VCgC5?p+iJ-+K&EIuM zSiIeZeX_TlT$?I0PTV8GnV@?FyPGZCX6p=R0{<%{UOao2ntceTnbGs=;&bE^K9VHT z|K%n7Uuvs=!`l6bMDc|3c!M~<8Qa2T%v zT}cwDWhk>$8*hwn5-a|kHBPY6;AY>Ng- zBDJhxmQ(#g@J?glTdNt0CFO~rF0)3n-qMC!FPsC*m(L_#Y}rtaZMQs0>)8V8vP;#v zDzI!xVON1SGw&P*OQSj@UTm4*-T?13)_!c%1S}I9 zkR(z|I6hmj~~B)+!d?6oDoF9-GHX`89Y97MIT?UTmq6&N_+)2K37V%Zr~R ziPTd4Gamb^OQ%|bC3I~3R>k7;m#c>+o=_|ase-yRThCU1%uZX6151}s5-+w~T(bx3a-?bp?8WYqNRmh` z$C<^xp)EdB;(cNC1;rAQCa`$NvN`!%v!VD*$^2_1UThh9+Zrts>~~&+G4?H%B#~O4 zFiV_{7Cuk5seg~_ip4WsU}?g?<$3Lc3O;*%W*CVVTPCNBMN7MsL-?l3hh`*+)Kaks zkNq!EcJE<~WxPCjSFwB-S?q1ts$9QOU2sL-Q751zUTk?fbq8AR-x=@;>at`ANg}oM zVV1+|KjJRw!uX0AiY5KIpe~L>9r0(FuF128<7GoOi5FWm`t`%}%dinG@cGNJxg?3y zvXEKIxJ{Y~b(wX=>b+u#$`Dx2|76#et4}=c2^RI6BwlQJ)XEV5kRPST?}WO1izi8> zmcz{QaJnWw-@i`(+Bd~={Dr{c`G&19_9$M3-GthPBwlP85*Cb>w>7o}fn~~Sl0<4r zVivWA?$w|!o79e0(D}9cEpHTbd+^ z)KcyX9{Y_39Qgp2lko%VD3)B2Wp@X*>$BnJj^1ENPbTqV%jDDkSeLZ6)vycaRys){ zwRC5eX7k4n1j}y2^{o_3QkI}D3#zea7R%1GxdE2#CL~^LNgWx9mX20e^-{O z7Hejym=ksaEV^G>cT+5(F9jA46SlKhqbLxc^Q+&sO-a1i;9c#GGed;ucP^rL5juemB4a!Av<1rFH?DKm&ew2GF0N_A1uEXxA+Xr<_6^mWAz*4=Cb*}7MWnd1LYXKx)Y?-&H7{^#y+q`aISs0>Q@*|`c zm9Kd0FF4|W_sDl!&zr1RGDQ}zC#)_imQ~JxrJE^<7h9$+4#c|5(5{Fb62|jM5~-yl zv-CGSvjL75=W-M0DwdenqPlEm_hJ(*?j8b*P9G94w#<8X9_!NKV902w%N<*iL~5DA zEZw_V_CZUa-wMSN@J3)sInB=dr`kn70L!$?BwlPOa^8b=(X>+A2A0v$B#G3rhgl|_ zbt?syF+O*FQ30!vs0)=^X#a`rk{8Yh!@v1M!9vsf3$BX=HvMPDPKb3 zDYb7|ZI3FJ+(d;1FtQc7m*}VOY3hqkNN1W z=nrG;`}X^16ie7Ufn~WZ+a*1+v&k5!OPQM_UTmp5f_30Ee4%q4#+X(yNg}mOW|q&l z%O`^6Qp%mHipA%>z%nkB9WTDFRcC_bQyz&ITa1RWi@NOo{jhVzqnIR-THKgr$;_~W zVA*^)^|oSh_#m*{s>9YUx|po$1{U?xBwlP;zmdHUL#?rGHq>S41(HN+iDZ`U>(1zd zWsXPP$BHFeWO+D&&B-Zs_3wgZuMdeATkNx(@Oa733dQI91CEmu`YYu!+&OG-Y87h6_8>yDP)KW;1rON@Hwk{=x|qMI_WZ4Oq-J=Qj0UQOdWKl zE^OYmEu~vcvE2K|@`tX1w*y|_&i3Ya1J6QNLAVCVBq7P<69cLB9 zL03VqIVA1BcNM(!X@*?|(`N>!KvzM_1tj5r>MDq~Vym@7+Sy`{{neEQJ#0M8o+%fe)lz1hBGK#_l*;Y^kD7HD0kh|X$0S}nd*&`+ z-@(#ayUzq>&)#_?iR^z-31k1uhHCy!mc*!Lp^X+d#$Q`$=Hwwwqm@N7uQ9^QlW45-+yo zMvcZ(S^nJdVPM(Sg(Q($OqpfDoO4savbRlR6UE|GD6$-8HwpUF9=CJXf(Kd=}JY?sIlU+w*JJD^Qo-Ye~G=()-?J ztjimpEx3EV)}17gTC$nN?28}170qVlk>!ddq)1?CR+g5G+qOlX$TugjiD}!8t!#eES_Hk7ROy|9^2h78&`iSOd;`N%e4hz zX!-2a_#jwfJV_F%WjM2}y+6?zESXhS?NTh?MHas<>|9x4Lsng=%d)y8UTm4ro_(9J zneOapu(WANl1MG4vDesK@-9^aEX%D)yx3y7ejnDw zEZ58iEUAl05~(GKS(=}oF$yeox2psymZ)z6ODA*o9J#CU>m6W8@FekKi*?3kv@H0( zvNu>JA0kPlmJDXuJM{Ddu&kXl>zZOY{#{^cxR-4{9r$z-`~APnCh=m+$R_ML@`kF8 z=fF~aAxR>&)clFZ{`B5y_!jh}0}JC7i%YS<(xg4>koZ|};TY8A)kzXBwuHRN!&L^G z?_~?X^85lxBDEMYOEvS=4WTZ!7o46dmRyl#LRYquH7(Z>-?;lUg2an0ny1+N$<9}L z{t7IPF(ir9vVvJ6*H^>$lz2IQex+EFehBK)sgUiGPA^v(yQ8#xNxayy?y)x>FGaV8 z+y%>!%Or`^a)Ma~Kkxn=>M~%oL4jfk{VA|4+{!v6Mtzxy?~N;1K;p%g#jnnwWw@(h zHLzUWOp-_~Da&^JL# zY|k;+wPv!JB#~OwRE*#iRJk1ou7tXn-wvs)SnM_kEYTZTU0k~Q-iBk}^$m#^TgsS^ z!DD~DQzE`|CA;Rik{=q&HmS*B!MX#f`Q_q+5ION_msF6xunYH4V8<)u)UDb3tV;>DJ8 z<=8sN@?EO;!Ln#PNg}nFGRw*j_wgx$j$0mVFIbCvn5a?+n4>Uz@~>Evdcl zq9v`bF|J|{u1}IkEe_1$@Z2BQl!}MMjZ!QHB1`M)?7lL>xj#OyTr`@*i!Ij>!MEXkV$bxC^2>hiVJLVVv+rK2QXY^n6g2Q3Q^ zzRQF$_9cQOky_3(%h^rZxTbXdVVsp>33C)!x=v@$U*T41&w!}`kfR=01 z3%-CQ?;}YfwcKTvF@EL7!ts(+@%b{v;nVI6#>fN{2gTyxB(TJ^VfQcDXC33fvZ0)5NxaxnuL--~H_E*G z6)dhjNfN20R4F{~&kgp&uLC#JvDv9uvPBmAMr=8#`KlTCzNI_kNxaxHEbbK6rB{Qc z*s|D8O?-+dV^_aFKUTo>;#w{m@;aZ=4SCT|(iDi~3 zgK4Tz7c2d?r5pUZM|RpSu$;NZ)?xbhnz;b#(s3+_7h6u7urKPaPrr%lW4iN65~(GF zS(0K;Z-Gs9Tg#f&6^rXXmOpfRteq2p9S18rZcBx3k4$TlNz&~xuJ(HzZmNc+-JsjU zV=>9{Z`~fhy+h*AquF85?J;f@N%yC2kNCFim*u1_GK6lA5gSOV|Eb&K-rl1xpxYzY ziKLWwd;IneiS0TaKf#C^;6@TDBkH$qk6}F+u(i|EMbPb$;6al7Ww%GgsJHu}+hfDZ+50duZyFH%OYKh$*yJvlw3f&%Ky-C9V)a`M*>;XKS^z3|R!ddl}AIT}3CG1|X zSz@lH|7hs;SQWnsc6A6-&qtfyMhH+gbd0V%#mTR1YWd zV#_*P_T9L(Wu3P}T~eb-5~-yPv$T$!ZVwhqzoiC>#dD{?;?rHn_yx20q z`vqFAJ(!mcmhYv`m;4B+WeT$x^mw%iEZsM~w@@tUy99L^y^pQR&A#o6@Ai1Loy3bR zEl%A;O9lJ5BCu$Bk|a{g3TCNs<>4!^WVJj!N3le?3oKtNuzAdR#CUxEa()*QFSfX@ zW$%!9a6r8eSZ4GlNu-t?%yMvc(*UqsNwZw0SdQ-&SoZ3$F{b_{5?ACuzb5fwOBYSH zyP&?b13nkKBabAJT23%axaIA7U@7&b@fOA6vPWP^Ub+Z}$GciX+rY7JQnN`(yx7vC zC)Y~w>B#~OKGfRx)sVJ~icra+MV#yU*W~Hz$oHv*Jt-#XjD~T6dmJefN zY}A;n(qNgWZCLUnq?QzB3F~wOR~WNASNSWJBo9Gd>MUjFN;TW)Enr!Gn8b@Mqy6t= zaTA();)?vt6C{b$@`+hicCc&=busm7bzZTAdI~I^wddf^_*ydvcYL27CGlcQmI{04 z%9)l01EDTAPLU*1OXbpd-k<$s)Mc=2jhq#!SiJTMEa_9&dEcewO?;;0WE_bXTbgaV zh`XdlyN2TPF%2G&BvMOrW{LJMz66$Tb??M07Q1}{%ZYrp!Z_owRWXdQgN;bM*s?2p zFXKZU#EUI0ROaJB z=Re}(DX^^4B1xo{#mr(g(ha|^6}HT!K(PdP2`p(h*xQ3!U;UsDmdkb|UTk?`&fdG= z=08gnESIZZ^C{Hj@wus`8va_zayuZf1X!?DxwFBu)WK5WEr}OfQtO#w zT?&G-uyf_Owq?nWkXn41C3cf_3$XN_5?e{J6o@R5S!`AA)*2n$iEnO2;>8w|9fQ&G zq>48_-|xGKB#~OKFiUa&uJ~S`YdRgZ6-%V zpS%-cj5TjZl1MENnPvIZ6ud{SH?gOlVhKAauq0};<7K$_(AF@<*3TvJVoRsXWpGY* zY@ezHmWj(q5~<|_v((Q1x*W#XL)&@X6pN3Kz;fDyohxr2RL=)X?{y?zY}sFp?JlU* zas2?65soB@)S_Mn&-?wp>)@9t+m5?1RIxZ55?J;=VAshzVh`$p<;r#vFScBsJ`l&) z-Zl2)z!H6(B#~MgGD~I0b=V=1-s<5v#gZ+u6#KC2WId-xoxu{^g~W?3Iz{KvlImXe zF<3tKCP}209?Wt#QsWMcv3$$OnTjRuu%IrspV@uolBJ#TX(*2+BwlPOUDytfmwV&R zZ2`-@-6V@Vgq>g7RIi8Umr3a)UTi7qdJyY!SuX@T zip;V}5~;<8SqAC)qGeldvYlda_Z3)Lj$rr5QCeH^9Yt@-JT8eBTPA9@#JUveyu!}N zq>3bo)Ut_L=43y{&igV;UbrfjB9Ud^Zq~K7b8QcNKbh%T5-+ywYs)&lj(*vOU(~(5 zi6oI)yqV?p`(nID-Z)3YTd}103F@-)1v_5aHR*$0(M^|=c(LV-^HCgQcfW5w3CByZ z14$yaTws>Bla1!V@zQL8cc5a4I4ZDQ3ufm^m)O1dzNMHIBwlP;t4OOb{oLrc4)vw#o~KRU^yPoI*D~wHpF|el|dw4Y-z6E59{KQq3Hsat6?OG z)bg5Hv<$;=O=*AD#hZ%7>A1kMH;{FEyt&(=37lWHd?4{+i-Gq?v@Gm8*c>b-MI?#T zqFNTu`>w;c-i5mKX`=c-vE+y>Q&m`9zW3UTT{z~wNW9n*ugBKUS083dDl0<6J zW|pfD?eQM@W>t^piY38cP?v~wc3*jDWWFWTrLH!K7h7hgUd6f;ldmU~S|5~-y# zvuOL+oP@e~nVo&FSVB$+ENkwvYyOmqNAP$V;6mcXmW7t=C7Ne)N5z4q?+ubfY8k;S z`g^ua0E_b-`>%?{^Q6GCGk&c_q6F6yd3CW2+V_QsMQA+^k6 z78gI0cd!xN;QivtjegxDfB47phYpGQT5qvKqRK}7#n2()@|k3kbV&FHSB@7ZsPT4FGv6{^i z`EPsefDVcDMC;#;1aW|JgRizl-r4c>)U8v`0lYOPqJ0tJ?~6WCepy}=k`sEb=$5-+yo zwPCkjUh45_P{W$ZR~ zZWuUX!48;5^qY|+Qp*cw$-X`XzZg<~NOe=ik}I+dzs&9c{i|%iFCYvwCGld*$NFs7 zC%W9+7hoAYfh3VyzB7x-^0a)Yi$#-mQxr>5kf1I}yB6T^XlAzu`((#=B=KU)`fJgc zq|mw(-pvl~Ly|}>waVdL(^5Zw?2xe6H?~nMp}_)+PJMQ#arXSob#Uxo7(n91mfLqc z&~l>1(uGi$^TS9IsiiHm6x_|40d?s*anWkU;&nz~**1%<%6;xT`5RcqSd)0MrD&`x zT6(7&VOPPP@+b&0egNu-vk%yRMNIebUavM<306-%bb5*f_y%v%oHTLCQhN}nl-7hB9Xu#?o@ z&9m?wMS2xU5~*b+vzV;Z#b@F#REP>xEHNR1x;#i@7WE51pM%A%7Ks;ID$QYCo*M_3 zISq9&*C9!ymYvM3@ z!7=u9E=eM_oMe{nvpeDU$emAixTjd$&I>H@FW64WOby2zuuQl_;>DIzgV@UN7vF=} zv0COPNg}mGF-yJhz$0+%m+7(etzszN4~!o5%DwKE^K3J&j1b*itom99qit@WE@# zS#3!Ysilxvwm9x+4VJ08iE531t(b&`3M?1>*|Gn!M|3jOW!_j4FShInWvkc^u1&`4 zv2$BV5~-z%8lLxG_w3dLESh)BYbh3=ivr8C7VOcL4 zFpqUhBT1wdJ!a9ppRo%pChgzoDHexI0*l#Mc5PYy;^v`Hm(ThnUTi6S$OgyQvW$^6 zz+z}Xl1MH6nZd9ZlZ zB}t@~CCpN3QXKBa9T|7hM6m>g2`m?5*%-@;D0LVtrz}al*pik%3N67VrA)yx&59(E zT3nbV(8Dtu>hfur*A&I#ennteHh($(jO}B+ZNM^ZBZ(JV!XAvoBtM%k#}4l$p(Kga z;>RqvK4|p@%VoW`^At;w$g&`ioh#3dFbM<8)2<|5Y}s^_eMX`Cy6~M)7c&ErL~6Op zEK8R*!gokG{@l7!v7}rT)aB+X_84fd8Nr9Z61HqxNxax%c$j^Oa?aV&_|C}Pb|i_^ z@`zc!C3}~Lx~#1~WwT<52p3rDrLuV}df<|0V5!i8#EUI;OS5C&R8J4zvFg^IB#~Nj znZ>xd$v&{OZ&k@%vG_&^EED*bDVA2U@B&MX86;k82^h)FmG6sBC%nWrF{Ncv~*7<@nVZ+8P?_b?D`tKf62%sNu-uW%;Mwh zj!)gR-w+g_SaL*`9Cy~aa?-R*FR05hXA&>AJThSS$e;H-SOUlXy`3b9)Y6k#HmtIq z0(H6Uwf(YUNr)8GB{quP-ga~^i`};)E|Yk%Hu|FK8GZcTIMm!q~!Fy zP?t(uFC;1!&nSWApdnjN?t4G4DvYs%?j&ApnU^;W>tZ)+=~%F&c#$MhizBm4o;jr| zSUv|nOH(Z0MV70L`1RQHVfZDNDWN1@Y-zYf7w`9XIxp)97SC%WiPUnCSpqLd)dGv2 zMvZrhCH;n=F8l1*6A=@RWaGM}Ukws3wiN$h9TKZ5UTOrEb=o9})Dp@p=awDCr*2M9 zJoQDfL`4fMQy;P?B98wI$b~U>a~6pgTheuo;26{F8j8oh(E^f0YPrWObI@}7h48Adi~@U=qjjuge3XPu7c4uHAA4QU}?{-CFAP% zt^)m=cGwHh&7hhp49x2fT}l%Cp{t-?Z%543)M>8;bQNelBbokdSHW*TFzMEmhtO3p z>?KM2?_CAc^x9%qK}7cT!_ZYwW1a`@@D=Rb3@tcC#b2fk0=iHZ7k6^5?d`05Lvu8?+nfSka zyL15immLa766yc)hW{@GsrV)_`_XIbDHfL)f#t+Ewp!XC_2DqEXiq2cV$1a1lh6`x zS)&u2)uI=WBvMQ1@_46l?Y-+qsLO?Sr&=kNT#>~(oL!wib6=0oE|?UNc(KJPi9NTJ zHY#cvSaeiQm;4B+r5>}Kt-c+17PsvAJW#PD#R}@u(u7@|d#KICl`In<5-+xFGZ>Cn zPKEXf*jKLWPm)M2U6^Ikw@h3u9iUxcqF6$22`r)I*e0Il z`!)FX;N~_YiPSQiSq!2~aExX2v6-(}?BWEL(*|rOzTEs$cn6qiMB>GkveV1r7_&C> z{Q-4(YD|(yEq2T@`salkP!|)wd#e;nrpRJy&+atzmbb@l59d=PUToQ8SPCs4Y8gj> zCI380BDL&gmIhSsr=biYKBNG-KB z@ERr~=v7a!c-W4)tyseD2`nRPvt6I(ZL8LW<0adk#EUK3+t}MsH+(*aD~!&rB#G40 zj#=u>Uoa6Y=MOwiRxCdE1r~=bZ1=jwx7QEA()J{Y7hBSE+5N+!tbBZLoA(8hL~1c& zmh=a@=fM(uI5JzYI3x-zH51tsWWTvGzE|*DAc+@Sx()Kf^UI+-FK`E|%XyMSYO!RN z*fCqPz~UGk_(QQ|i!2GcY=v=aoJj|$i$*eu7h5(yH^h0Y{~K=K~ds`$K_cc|P0VoHG7t7qGm}Bk^L(-oxxU^5ZdP>QEP}pCpOYa)VhG zz59M2EcFH)9<5l4M3xSd*!_#^xs!O_?`ulp#g=*-XW_A*eS7E}uxyw@l1MG7%;I)n zsU}!nb#69Yv7|f_)TN~!tIPGljp{>P4EvIJvE{3NE40+sj|l4;`>-&8GoG3W4jOaO9ji7<0M{eS^4cTTIv|~SpgRRaFRr7X~8TlbNiHm zy1cP4+NoHao(L>{`fR5pq{7Ysu%yo<@nTC_?>w|vESrk!hDP&A5~;<2SzNWY;x){v zX;1wXOOD9W;0mkDJ44UKVClJl#EUIDyAGp8^IQ&oh4#U6l0<5;V3xP53h>+N^I}e1 zQY;Bi1$B8jo84E|c=vKJSnAy;@nVZn5xb~U`5C_)##pGrIH1Tpx zR4gGW0!u$b)=fBY+xInK>FZD8#g-bEW3etLrhlvrmiHkfiPW-dyq_$-MDHy-e8G>>&5(p-MBadlI7oaeK&4H zpabrqZ4Oy=0tROL_5LLZ{;(To(c2j_H8ZIe47+i;n@Fbrx*PY~54<|0uN&;fz1l|7 z{`+oRxhWlRH*UD)O=sARGuuNF{-@nI$EeGAIPILDwFhRrLkCDs*(_mqmCX_(moKjZ zyK$brB+cLL#sxo8!xj1cr9c1w!hRRcp7TxE3F+gwtBEky6DpE;@$9*>4!e|GTDH;} z_+R{Mk|eVKMJ1g5FZZn8;u}N*4~I8aX3z9g!I@xuWp=gk=*4i{DY+Iv;>DJJJtMKk zGZqiRw>`{0N0LY_$;@&(BN3k^^>Dk;QL#j&2`tUKFiZK7rSJ|g>i~%tTV57VL`&k# zVTMqbdHy7c)RM<6&l&6@e^Q-)gDFS#g_WDywRdH z(->D6-;O0oq!x{ec&8DlwWuOk?y2hyS1c~i1(y1?*wy*B_GufyGRJ_#i!FzHr=cZ1 z`MSUqrBDFMTmgv42{lRj2!Ul82k}I^OFkBCo@ael}DVC%RL0#IuV^h$}L$C0e_~DC4yx5}Ya1!g{^r91f zH_mh|Ng}n3VU}*48sgixk_*l)Q!Jq`1eW?;*|FcFORb+!7h~=7CGlcQxv9=*ajj!H z3&xo05Ryb{na?aoZhRUJb-7#A-%+u6WeP0Q$FY^)He)821xsI5gOYf$<>?XDXg_E5 z>v~|(*C0uxmd(s^EOPe=uvn`QS9bWuvF|t;>DKg0ql0AoAxAp z&d+KINg}oQFw5Pqr6+-9WTK_7V#yR)UZ%3=FcMa|VO>;Ska)3WYR>{3W5XZUS_E}z zkVle8Ef<+ZZAx`duxJn6c15woycE<$_cTd4k1y z9!Vm#+-DZ+IiJ6SB`BlEUBwdcN?!CsuzW&3y?7-P>rka)4Bp+hlRUKHNO>#?G5 zB#G4WmRU~Re4z)H(-kVFDi*hFfyFnEt^9_yGK&JscP$bxwp4dy?_BwI_NqJ7WuzWS zBDIvMgy;P!72g(uWn5752gOn#vb^lcj+gQ`7e5Efm~av=wz!?Vh{sE}oqqGd5*tU7 zNG&?dqPHy##S->LV2R$$t}W|~?w$meg=)u2;>DIY3wGY`v}O(}x=xZv zEgP8S+4kyl!SdRB@c_k=EwXs0Gt17v;KpFN=uhIsmTlvM(bD4Z6MWnD$zYO1YT3su zb7x234wg#boQaAh?wz16<(%2IWk~$ACt#@&O5(+qI$z__Vs#>=K8!J~2$Do1XJB1UWqF^4y1Z{nl1MFC%yPhJ zCO+d^Fu&0@#Zn})v>(onm*D#k=RjTZSCV+KCU$raQk+LK-LcWSg1ulWbWl6bMD$~4whaJv6_d`oTVdnAd}Ql~Or!>my2 zH5e>Q9?l9@ED?DEOMC(AkeFGv1Fk9gq>*^BWozg@ytbU|{&)n`<=`ulL~3czEO(ch zv<1tgxp%S@i*LTba(NrGtUZ-77sgomV?b6N)G~xwo^IY@ z2aD?--4?6s{kk6e_>bieyFM%Q%k~^vL8H8e+LGy(HkON*oO)SzI_&yH^e34lyFPbn zzs6}`_1*Yr*!4Lulw|q0U7xC@)bK}jY8d7RyFNWlNxDDn`iyPOewq4Rqt~$OvvNF1 z^*`+^oVl9F-t`>xMOcPHGdS($h?76xWoN86GFf7tcO>g9@=><6592D?53 z^hu`wy6f}X54>jSdjfWS-t;AD|9#iTY(Zz-^*Q0&sU__CTpmmk{-<4^_C`1Ga4J*% zctx1;zKkR}WwS)YL^exAIYcgmU7wJ#B+cLL`n2r7N(E-m1`j$`Qf3^dkAm6rj5B+a z*nmSzPr~e3A&10^XHT`t)$vT=`aA^hW<7qAB(ncSWg7cmI@g(o-;eoHElyjp?Ao3Iuhj z9mvjVsfF=)XFer_#EUJ5>vhGt%)c8L36|U8B#G4G&Me2ZY<q?fPg8yL=xkIZh;r)RM+5!}j&Vx1+RJ;j%`tq<9z&D86cYS_NvE+&@VV~IfLW zBCxdRN8-hnd6$;p@zS~V{&!$$ZA6ktEfblglz9bns7rgJeUXYK>6@S~su??QIvwHo z0#|=_nv;03<&4pIw3t`XwE|1#ERsZOS;{OI!wrJLqINbkMX`i_7g%cYx1&5YJA&WA zS`tR$#TM^gv+xgT6|)@QD3TUMl1MGC%(6nO3Vw5Cuukd^#o|>guo%X$RqVQN>J-8l zJJFoPi!JHT*lVvcR$bWzb;)T*l1MGbm}SCEBV(vbMT=QATl~5+xBDTmSngxz{U;AK zi@raMv_P^kC~G^uRVhOM--?E`beT_0U3zjS~&`>W` zU8PZ(9{S9(y@}3juy_w5@nVbiH#fA@Omb)sb@3lTl1METs^WRS*1?*1oxF5K#%jgl zwn1R|-j$vAlM>9auY9p1i5FWm&bLKN)S0p;z@oB)B#~O0Fw5OdhR>ibRqvj2RV)P} zi(>^g)X(ZxTMl(;mrCNrmd7`&(Gs(BfHqjn-jXCzOK)b`zbpR(SR#5i^ieFy_JX<$ z-@_hf@3bxd5?F%FNxaySb+s;9%14%Y3l^2dB#G2AmRZj1T3;P3mj{dsRxDu~1(pGK z**tbMFh3A1>r6&vj8k1-ANLuVEKAD_ey3QnMV71lv9_+= zXSsu=aSakLwrE~5#%s$Cfw!N5sktF%c-Gz>BO>V;Og37{wzH|G3+8A`r_&Zm?f&WB{`+D#9cN^7^bZKG;1d)oAwp`jEu?w=Xww=3G#rtAzoEa3@0hf&3d#EWOoylgN0Uo3Bb zb%6gReiTU}{a+H<|FWT6d0fRl{N7Sau>@@vSms`0J0-z$vlqgdz`BUUi!B+R><%z% z`Po#kbX1vE@*|{{9A+6>xd`8l+du!ao?>x#5?IE!Vs$zDWhBl)$4eEJ#EUI<-PrRn z$DXSG0L$_!B#G2gwi@1P__$ug=P+LOZrVq&6p1WPY?$THcH0Y3mqw#Vyx4NukL~(A zdOf@iSk_J=Nu(BCW^pR_1^Y4Il3F2#4^uka%AVvBJEdk&-N zvS+x;Fw2)Dky@;n<>+ed9Iylw*WRjFoLmGJv$=C|c&NX)z8&haQmsx&yx1~sgF7ZM zd8&h7r(aTqB#~O|nMHfjoi$*Y{V2y*vE+y>XVbYHDdAoqk$z}b!;7+_-IEfcqoW@zAWlXI8 z6tMW-BuS)}m&_9ETMgg#aQH9_v7_6I!AoX>mCd(L^E-SfV$DwcmmmPP|RAM<|BQETWIV-<-PTmC5;9)adq43@f=2t1Bn+~*2nye_sb#eoqVu_yh4&l zErHCkbaX1d6>a;i%viX@R* z!kFdYF{d!-mk#|hjbdpRSzK1JwX#&l&XJd0BJpC&0S~r^*_mZb1WRl?Ng}nxGRqEM z?P0L2-Q`iBSZdP+{o*)_{d(=3{{f!E7`J4CIbLk3HoS*pth_cZ1}qV)NfN0ggIUJR zRu2Np8&^_l70Ztq0*gKWUhIjWIk>)LcawOrPufK#qeVccq^GheCasxbQHI1h&kqnkBJDX%bk zs*AD0I9PKp16Gs%LrEfKMD?gJy03gg?mO**JBdR!;TdR7>_ zmnCCe(A#=#JPgc%*DKAJ`!N;9#CI|=(_fCEgQ3Ehag$_v+6rTj1{RLgsG-6b&_&Wd zxx(n+I1MX|{x{BAL50!PLT?s7RfVx|d?VgY_O^kmV3u&`Pjbp;iQ}i(EHTh9b23yI zC)toR&sJf)ygwXIx>U)WPq+1O@1jCEZTi5JhF73pl}6lmzs z!gYyiB}ru0rB@}pE=w}{;@5=YE}ryOEb1(Q#ccxnYuK&A)9b+!(?a6K7N7WX?BgKY z$hlx?F_9!vOB1uq^0rwHtD3()Z=qso7FjkfXTN7&8DYY&ZoYM%#EUI={&d2f(}iuO zFTj%CNRmh`UCc5-^&$S3L2t_bM6u|<6!eS3O7@}#FZGCv&@V&olX$UZNYaaF`OM}p zR)4gr9CIV27Q4atq|sK|nFRgP?d+;nEQh`lSi+;&>B5?*uqLqNdXjjt#b{(ZrxW9? zY@lCO`I974%UEVvxGuOJEIT*Y?N=UX zBCurbB1xo{>CAGbZ_Prm1P<7IU$JzGEYT`9k1d@%V=-9njw115OSr~}mf?9GeZZ3B zMUqG@q0G{6X|*+2Mi>1yY22e-VEt}EzwAB6cIRP1JMfJng~v&}*z!l>3wXbzwXd-Q zi|c8UL~4mDJP$p^7tE?iUZ21|)Q zNg}l*F-!a1Y4|?b30}?bDwga$0*l{OW|por0&XR@(CUUV+8QkFAw$LBsG}pP8W~UTkUDv0p)>o_(-dTC$?e^7Tja*pY03rTq;1A5(mr&vme@jwJD7OKVFe&SQ%Vvzx)<6HSsx zEh>9l_e(k=s^JN$JiM#7VmbQI((?&w-mG+d?#ZeTse*RcMv(-vb{yyJFgZ=}hQ0rS PPf40*+xy>Ey_EPLhq{?) diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/events.out.tfevents.1788167878.1810d6ef6af2.13.0.v2 b/logs/self_tuning/schedule_free_adamw_jax_v2/study_1/imagenet_resnet_jax/trial_1/events.out.tfevents.1788167878.1810d6ef6af2.13.0.v2 new file mode 100644 index 0000000000000000000000000000000000000000..98bd71e7d13f4c7e75f7dc0c235278af1d692c4f GIT binary patch literal 524002 zcmbTIscA!9TMl@LNygfu8ALKH#~GDHX= zNcX4xYc3Gxv>*8jwy~M-GS=;`ve|NW9=H}#K@3ze$ zL8<5Ow8~>Srb_)+C7=H*tw$&2j>>hMUEJM=DJ#mW%5|<&5dOq?bbVzd<$CV6HqJI~ z?qARDd8_-ocAvNR8>O)ue%ER!cTrYybhELuTIAxk@IL{gHuf9>0ZvYvFX{abXmCVr zs*=Xb-vOG+8p`z@on7YII9s`U*t`BG;;;W1B7!~)J!bej;-Kg1X-Ycx%WwX*{VMW3 zw3NFmH}d%FH|cF-YrDkF#@6#cIgJB98$r%7j}LSE{?3`B*K(H9H|W*JbH)9C zhi$IcuR^JD#hKdu|t0c$sN5Is}xcFC#bJ-ALW)*(4r*^ zt?ZZEEOd3Y|I>Ck3zL!XQzcC*+TQPXU~$PcOQmS}ivRrNJ(M+-8+zEgd&vCa6G!bs zV9eahL!tT8l>Wq!@k3&0+81qxc0M)FVZiS#X`>!DODVX?p9pQ`p2|%~1Q%u8ZUmsZ zPU&pj`*+mB+3hWqZnpXp)md3hxvu+PJJ5fIk6!ZRVNky5(T7+VhPONZDgXB{bn*Dt zP`3NmApEQS|H(Oe>ka1EO?chT^!H}GeLXKxNmt`fj-j%Fa@)TK^1>y~HXinNR_;sY zE_8Bt|5u)shtopw#!XO6(i-wRbb|d-8>Knge?t2x>nJy;L;v;f&Jd^gb`5T0JEs8( z!+ytkep(cxv|ayCoW8Q2aw|H{*~QV(X^|r&{#6UV-lHdmZS{XA-oHHiwbJW>e-gz< z0xvw67~8rnH2NK)x})@kQtjb?LV77{DXX};+1oF4{p;tqvUORq$m2gvklCDN@mjaF zeWl!00%9{rO#h1P#3Tfk5JF$wGpzT|r32Zta@@@fsnO^C3|KHX>w_ zg7OA{1do$4#E_i`QLXMh9Ux_`?tdeY;M%f~P66i;vMPSp8Gy`Gm&*B&y_&%YQH^?i z6(C!+qzo}63n4x==Vkz8--Ct<6aPBtR&pTCcAZAZ)|L;;0n+@rRL+Mu<&MCNQO5le z0Aln;$`C`oA!M)OBt3xiXgjS5f#lbb?TEtDQY^RY;m1xMj$zgvK`sAfPoYzr&a(&WrbAEhg_Pt9w8Mq0e1khZ=IAOgm`U4 z$mZ7iD*#fu-G35+Br3^5ZtmZO8}ZT9EejxT_DSV@$oh+GF=PC8o1p-yIwWNXA#ytr zVpO@=6CgkIB4!duh_Wn1(Tv$qGoMY@1xVLHQaK-DKax2YQic%XbpRpF z?~Km`NR`X{xdc+hf&6^A4S&t+g%=F~qU$M@^C2B`Y%t^1Ru@C)$UQG9LkN*eKu8nA zCglLx@xaiDKnm*1cBK1|VBCo4hPrD3a;H!#=R*cmgV%c%f9zhZJfsN3EEMMwbE7eumN}=`{EzX0%kVss@O8TPZ^bk$aAiZacOt z1IWks_C5qs#)0&$yAC(PJ}DH>BSl}Oaz5nglr5MMwLbwnvhbIbA%u9nL&*H>It2i^ zxv3%?!UmM@FxeGK3JhZwPt)Eoc!y zVjBI}N+8K9vJka_4CHOD_W^*IwUWyDke*%v2yw8fHw_?l)T9g{#JIL8T*iO96kq|6 zZ>#Qx5=dwhS%|Iia@>gR`BrrSQtd01^C62Pj$wx1DIL7ky%s2C2q9id2+^&wTmX^&)rNh}PFW;ZqdF=~6i#(%ab!GwjEuF9ygeD=9+=k!yjFXDa=^0VFTU zAf7-no617wSKH%8+-iAlB@C_nvr;)9GGQ%~v7zWn9e`}PBxML8#_bTY;h}G5fLx1` zOD2%$X0nj-WLtzJyJcSnNa8W6oDXR-+#56MS*X_r$hebIh7jV_86j5V`zQjWU*WAY z1QO6(7SgiZ4maY(MDz0iS@K3I=RJ0uV2I+bjaP z*Fv@E#`*{uQNJlZ*K&90bBjPy zTgpNl&Q8UR*jdZ+0zhnBq;fta&3H9tl)f??4UmURr3@j&YXCy@8xO@>-8k2pJOT-8 zB@6ksW+HAxvG>jhfb8*<%J~ramHRQnCwQMHK-#X8GK7$fp$IWry`?b>t*jMy3kbxk zwJc=(({Z>Fr`L_^1CSryQaK+o>AC@C+*V%G6(HL;OBq6loEbto=rv6TNdG&LFA1cG z11WbKg&Waf^bUN_+U>Da&WCiEXN4J?zx(L}B=LomA%qx@L&)9rdOZNr*LOiFfn>Fj z?TFWoVYm^`Z}#{I9r+L_mGdF`T_$11=>Z#T01~@H$`C@lCL^RmKJXAg+_m~u5J+5G zSx7y1rX!1|cg2oG^_9x`kl+X75%O56+zcQ~21*%1NX86=Yz&%$x4IL<>Qxa)a64JZ zovkLg5f8#g?}mL!=PLo?882lBA#yedDIT%g2|AMA^UhBKspLR% zdyT}67{6mkCP0qsOXYmX`@PnfaoOPOBYL`Z^W&0gq8&Bgf*2qdk8 zEM!OHez*}cj=T;ANcSA6oDUiMjk(Jltm^O;Ao+Kt3?U@L9U=K^y5dD+*}=X|2_&MU zEM%&?4nppKy;dI}iSbf7AM#|EE@ljtUwjH6hf|~sAw+HkLgY)f;dvxG;b&_C@llh7 zj9kZTM8}~~y8u#=DwXpg^V>B-h_Jvzax-7(?cTa>E z@A@(xAmi)y|5MI~l@7ZgnC&cO2qAJi5z>8a3SOcF_i?ZwkSGmV$f?l`WR9h( zHFRX*0I8f0(VEE&tusUW`a(xsO{5GVL~S=he9IqGW$r5oc{`d zd^|3d^C7Fsdn3erqsCl-Tt6*k2qD(t2$@s&VLgDf4!&wfAi0{d9ceL`8Cu7-KWz$- z6T78yK19!*xo7|Ns;~qg6C$MyA;jwdLUzjC+XRpv*;^M7NOE^s$i?+-@z;#&R6GD6 zjqgb1d`RYz7MPJRCNvQs4v(Y^Atd4uLW+}y8Uo~}jk6npg!YhyJZjz=HzGICd@4X3 z7D(lM$SfD;eq>f>7bAc)aF;TKkc?=AvK?_OYl0iGx07=vKs*zr zaz12?Ds!!6_t7mMATv^>3?W1=0U@*EkNE&3^5Atp0?F(t3%M895Fz2GPh0}XrZ}mb z4^eFM9Vcf7yZh`1Nc|KkLkLkzLP)#YANvDDZqoKZ0*US=3t1k-oRzh*4;cca*(|A? z4|(O$0e_4i4ei|kVrwU52qDHP2)WZWeIGyuI?M_okN|C2NW>)OqEY?TXPiN2?jn`* zAu*?o5#m4mRZoCqER`~Z5bIQgyqULjGeGa#n4CL^@_f6G*C#ETs87RfN3$c4Y)WW-XS=`4B4~ z22y{F3SP!H@RTxykcRBw>V`H&^a%9t_4 zVw5RB;;*{rUMWKeQM-Qk6Qm%)+B&`N znHX}|Z^zZazn7AbRgdjwNFo2UxK91PDyDr2_?iQ2g&b|9w7*+j_p(+gCVh_x7zc}? zrp%L){$I=MCgvJg!BVnf7b#J^zD{Gwi@`9LNb$X2Ul>VamPt8eB*`qWbIE2#(zf7X z!2^H)>RG1_PE^WYEoJ|&hjV$g-YO1dL{Z-|@z=|f72Tcx1?c9EUT=Dn*@e_;GORkLN~4yD*oDfORM z+133RgUh$vuFtDy@;mRiZ`TN=M&VN4f3?icxJ}nVFd`nuofrtmg~~%xr0}?4eFJ~K z#B|U4u;?;6PD=UT7TVo@t$Z1dl?5Z5v&kVQOHcNozh;IK9xEFR9`hPTcwGyroPVq| z8?XQ~Hr6_fHxEJ6qzvJ2Vtf;SlX(4&SpZ2Z*3bR-v9hl$q|A`Ho*3;u-wz-|E2MHh zq)kXYgtXj~y&amXRV`%*A=bAM;&}A)8-N@vmU}=T!TPe0(jClcc-hsh?*P)XLMrD& z0@}1fh-%F)gj7~Z8A6EHJ%pSHO1}n>E73Qf5J)8l;;zU1n!7p$;zOumNm4l?2Q3NacJ;R(}}Uf%=Dz0P(u}xr#u1`pZHB4>3b)cl1+y{hrl!gz)p{4m| z=vLTAw#}9@gb?Fmgm}c>hy+M!oj`@jf6ceZF_P^_s137^e0mj?0g!tQrE)%G^&nU5b!Flk@OB*@`UNh6Iv0Ko+tiUmkzWU$#e%17!JashkfvZ^_)3 z4)65*AwZ5klrn@6uXhOPv@66EI@0P+&t?P?GEf#0smTnj#+OQ;0p#-|shkfPJBfL| zsl8?KRDe8sA!P_55#Sm@?&WAY4GjCkUbjFh=R@pFn8)w2 z0e-On+4NG%5JC#SAY}8pyUSr{WpCf8Ngx@6Wg(+WnSG=|ug_~>XesQI%K4DTPqpyJ z*qC}d1t3*XQic#B_YEQ63-sLpvdwZ@Zvu%LA`1yMsKyiGjD}NjvB$=_QaK-DVZ-d~ z-;;c<03>X_lp%zu{Xoc>Ql(UzY&fj~k{WFdy(l{mC4OBE~uQeRmr=R-d2XO4W< zM|4L*M}nJ38A6D)0z&2uHNgppZ7VuYCy*Kr#Hiy>gbe9#^bjEDtEF;2WLqTjuI+8a z9vy(#)l&bn2|`GDJ%k*;YHA4`Nmy1khd>I4%XZ|R4KuXnPLO*KkQgtioDb=|Lj#AW z#me1I0GaJCWe6c&N(fOhZIcI(>2as*2_(~07IOYBbNEf_>V>n76H=sdK4d{l<~(Ap zk#rv*ea}l7LWoLzguDycxe6d^d;FXUBzlA_pEk z0g|sRWe6b|jS(`?_~{&gygFIrMIiUgWIGbn?gc_r)&}BxvJM$iIUh1IDGxL1pY$F9 zkUm*bh7e-W1R?kLJgEzium$lO2qblsEaX@4bKHp3kDqZO`wI`LoDbRCk9k_TbgPRA zK!&cCGK7#qRfJf#ZZrTO`nx@Y2qbK@EJR25Jwi4*&i)QVOZk9Q&WGe0Gy6#HgMoPU zd*P^*A%wU#LrB8Wgl7PWYh$*9K)lAtLXNEcgpiy$%B8U9S#d1X9F-SX_C95Y5;Q+R%~K{iJd}JFECtBzrLPYVNS3*5N6bEbMTlMLtg!%zSu2(EA&R}2brwZ!RB=6O z*aj&>2vKW|kU1SjtpO_HG_ z3TLDYAtbR4LcaKg;B4bTd)6cpNbopW$eGW@xDmFChZ{pj!VXI1e8{P}op2v9-;m)B z5RF(VLkMBNciL8dGCupQyKv+w0;%LcsxL5*^dx@`fTS&v%K4C4O3aiw_4U>G2|w4> zQic$c+a5d8ufP@8+HG9Z>^yi?3-*!W8w)ZCB+Wt=5)$&qS~LkKBXL&$`O`9=V_Y@-0udx2EWhs2%KMo4GZ`~?7c z>Lz6fAu3%E;`=2N-yt^oI^ZFJlyD#&>J(u|_SVdH0?3D7QaK;;V^BAQT#Fx%Pf-%; z4g9kSLP$hcgtTpCfQR3d%XJG0BxjOrN3=9^5mKBUzZD?eZ%O5R$l5Q=EU`u1ZXbt^ z6yBFIgb2=NGbjkCn!qMyGeki^NdkZpO75Hjo7jRgSNb6hItL$;OZVMjum^;igy z>!+m*AtXZ+A(LOuJ_`GY#pc*|1QIeu7P9`;Q-rL&>2ebxAyPRX5@5>AAur#jwFw}$ zVN!+=V$lO3ZL(jM0i^#P_euh(;y|KTF#AYe;H(0Gw1|?*`H&HNdtpZ|er%-+kk)Zh zh7eMyg%I6qckx#D&L-pU1X3_nwj=g=W!MqrX5DuJTPUY&zAt{ z5hrB`A#%MDawR(G0YI*`y{$+fQPX81M{=Gaq(IRWr-W|oB9-$YIhssIUS7U~Ur=4u zQ_2uRf^`s*qH-LcYnkcoZ%iQmma>otrFRG^tltXH>`Mkp<$Q>Ik{S*z_cK<@pd-&s zqzoZMtq(#T)HTPi{ZEc@Y(XGp97yOTX42;ugK9k(S|?6O<$Q=mG4p;TPfukUKy=PZ z8A3>+9zyC?D&RZB@^3or2_$!hY)5t|W%lcjP#q*pO>)Y5jHc@;V` zc$SnQgyY-BqU^Y8&ebg+(!=NNf|j;Ecd-n_j$hcK!7ZHdT=kEgfTzrikNph`w10Z94rE)&xbT{Uc0(0(v-VGg@6)0s0 zA#!F2sTorF7$DXAwl5)&A`WD`DRZlV|hRibnS?pG^kwAhSWg*`l=U_)tkKXYD$hwtMIUh2nf|=&@;FKA$vbE`^c0BuU!F>UoMsNAd5{ST!z`EXsjH1&n&>3|;?E`s zAu3Z5;${A&DL@9MA4?*T5)PzuH)d_}-TPO9pd(MSq;fvw)(GaP^?qby{D8G{u9P8! zL`*|SPq&RD0W!ya*=YjFahB~!c_6d1ziqAf0w8iOQaK;;bsY0ttNmVwkubDIE|W5Z z5FJZ|geIhQhQ+>(E7URwq`>*_7N0A3Fu}cL$Ld{=VDY)lHz`Q4`21$jVhoX+aHKCR zJ`bwZ>Ce{u)8g~C`8_c0y>qwEu=re4SxWo6#pgQ55-_RdhSd46_*|!nl=S~ve13k@ z?!~b9{8}q1QM~y4U~)a&vJOAYrol)u?jz-pktDPDe9~lQ$;susIDJ@rzP`Vd@jos; zuUPmb0v4ZlH-8^9~Ykw)cVm67N3XDl=A+o#ph~yA93-yLCO2(a9l{5 zD@6*A3$Dy@;cWN=HCTKe<|L*3Z;Q{HO>I6Oj+M1L|GZ2NF^LOhkCppMnPX*}2hZ?b zva*&`&OcV}p2>WItE)>^7W^hx`bZhV-y~xO9xEfPeYykW@%(E!1QN1H7Gl(-0C&|3 zfuHdUh?m`^az4cC8*`~zVzcTwK(?-wGK3I|SqQ0Uc(V;aY%XuVOCVJoh(!%^g*R*N z#FKE$8GKVJ=R^Dl8sQUXg=_iV08zLvWe6dKvk_wW=0h)l%xOC7(ZA1LF0viTf5|*L zKdU+r-|l~CFO~Bl23wh-)w#b#5kU56NEt$is}({9cMHI8QXW3l?Ky#DxXMD3S~9Pf zG@rWQ448-vshkfn31Kc(@7sR)4v^+qQic#BXM>R7YQyi)kv7dMiU}lYu`I;#^flZ^ zmRX*{FXjUv{cTATnuN1*3eq<_{igNhLjd{7fKa9LO9qrX%Y% zeZwaOUtdY(d`SP=%-@l)^~{!oj_iLYWe6dO_6S+s!x#Vh&4)>oY6v9PUA7}-8O&_@ zXl#CkqLOx`T;1jusJvnm8q!-4pj zT||h!$5wpgQ9Vv7=R=C4$KVN3x%sNr09ij($`C@z7a%0BBC0Jw#w4z9Ng#zwWjj*- zgRWFcpN zF&B+-Z@%LnA5+Vf%K4C16EzVM(fr6#fD}HEGK3Hn7lZ^QMI3{UsLyZRl|Z7G%R*+~ zWOnvWYin*nM=ac>az5nM788Vcw#~r5e)D{Vlp%yfxFY1?b=$)L@pFHnMIZs5vXENG zm}{+;PaACjNWIlkIUjPT7PCfTgVk_+;djeN$`C?y+z@h0KMnuVVEw%KJ_J(Efy~ck z=C(F(rhqeGN}fpNe8_waW~uY%iFe1q(7OCe$`C>_+!3;G&JT6y$eq5+`V+{#6|x

    1^vf;Tmz7Irc#CwVzC4vd+S}i50JTO#>NDax>6SM zB|ig?T2qyCDqv`}zATmVA?15};64(wagQHBZsbTALP+6Kgq$u<_65j$rN$!(By5!| zWSS*&LeyG(zBfS5t4ig3NSj*y5wiKzlLCO)w~;c05ZC1h8NBT3c!0dQ|6nYEc&(O& zya>IF9U1oeCjJfK&_b!44{3dsnG$;KjiVhv`jki+LWtZ7gj~pe6b%roQxTI1q=*B# zG3O*g)~vnJ7&@Y!A(itXCVQD1tMKxseE^buRmu=Tf>$Es%f1;n14e$)>X`(RwMMoh zv3|_U44sZDH2^7HDwXpgCtfjALTh@L;nnZ{HByEUqP7|#jm}M72pw76fB0MiiCZfR zDVojfBg(#|5ddjwB9-$Y=_RvqANhQJE3T1fG)Br0LK4>?WUI=;2Jon}(5Jo=fdqTW z%SxJ0I*nsXCuQp%cw!zANY(Qx2aYgjk{xpEto!}`dS!9RHA^Mu9aM=P@Pv`xj&t%=Uck9?}u&Kd%u|qShiYIAwo+DV;Oz;)djEw zmPIZlmi%>`zJz>BN6V&-_4AqBm(6W3zTE=XA^aw0E z`we`ECBmCykxOItmll3MK7pmj5vrbV(ds%DEpAH6c3|0ij0zE2jJ?s){%wQlU|F6l zx0zUcd^nc-wXa}bzCZJy2bQ_#sCvF-RC}f`+KM%0U>R_k3K3cqeb6%8yXiKtJnx#l zl~_u6mZk?TpkF(2~nosX!?TXnm<*~x3t+k0{aqw_uzTxOHeQsBD7fhq2)q>`~t9aUtxHdSVA^% zEE;>5OZ@H|bl-sGs5e#5w`@0I9Y77>0{d)1lQpL0Q z6{X@B3pciY29{5QsCvHT&gs$EmpyYA;`!y6DHS5Llrxr)!gEc)lHM)%1hEwObNaGh zk2zO9&vl3g%TXPwo^N@jWPz5hz0LYUUo;G<5TPa2AN%s&v)@m!Ja*o7hFCH-ax50N z&tqR|J0CU$i^?*po^Kh|U>sVOeLZpzEct7w5TV6uBU)aoj?)Lr77)%ELSE_^?b|2Yqk{iJ3i_a70skf2s znXh2kc8jX#Tdp>lj+R}uXN?7m*8?gQf+C6jZ1Xp=DkmS`;e=_5;hu6O&&OOK1?sGX6X>OYZQF zB3-bkpQGye7MC6_Xjy&zL@`+EUZz5X7M&opXm>Hk>#+wBI;F%?!?P@V#JrG@zdx}L zSS(_wdcI|b+e)-}ubSNyEV@Zlh|p5SSZ0kj+XH>ET2)m}EQMP*eL41&c_%S1PUizy z+XmW!D`I$nF{9lTDy^?(WyS}eArMZRX}HLzT%o>6z&U$2G(ws9kFOsSr4>6hn)ebGBSeKhoCr!o~Hw6xlWmhfK3(!tXENS6l0QqHsVoz2WVs3}Qz z0gGKCRnND~^x2FSL+`9BV9`mXLWGt=#^N5f_8nMuzWJa^Ecdo^`cgKGxvwm?iNt3Y zJ3dnNe9Oc;JJE9EN9%`RIr@zX5n7_RV_!<2>)^8sE92DG#F84qv7G;T6|XHV&yCH6 zzO)=g)$=W9^tPd;eZQ-1!SdLG3K3ddL(rm>x@Rl&C8?K}8nJ}!;8-+Nm~-X#^5FJh zi8i3>`4)@eOkWajH`@jl*TGbX&|f@S2D7TUy8#IsmAGM1Sa#;*a3{w=DW zZ<*LN1uaJ3lIwz{#S>`)t6tPtDET{6A`STdMZVOe@Nm+rH{d~Z6_5Xw3INGdPxgbfMr`wD+^-D-^1z4^)t-Nb0q@1W9$R8ZEI#`>no4)i72W6%O(DdAacZZh{`W92LGUMzPtRnND)l`BR|-0=b7 zU>WO6g$OOVj73AY1dqu-Qxx2YC1*dUFSUJ{mEx-HX6=PB)-j!`=UaNdC`3!XNm4FY zvae7fLQCj=>`VXZNDJu8O4YpO#F7}!v9u~+-iw`h;A#Vwh_+Nc-_p0vBecAlTVe>7 zo$6GG&|)2qmLscs?FCEptJt-~5)#3&_<1nL;x9| z3@SuuDQ7G(->-RtrBn8>Kw>F~iL$^V5Tn?-?THpva1~x zBD8oNK#TsVUsYfkHRk>vVu?D)u{4Nf=EH27yLUHO>?c$8e9NVIEpUwO9dT|vSZ>dx zLWCBxgJ|hBbnsoU+jzm9W&h?x7W_#IpIY{NJr3KYo53K3(d$=5%|QE$N&{B?%UiTTdK>DNRB`?1ucF z7B|6usg2U^R4V14R+2wg?}KqSyzQIAYS?ZWRNUV!CI2vTKSt?x%~6FFzfZ4FQU9;C z>bMYsUR|Y=5oJ#s{SCq?a;IkoCW@>)*vCU9jk z*zm{k;lBe-EVcJ2efUWQ{&#E2<1O(|+!v@V=nO}RhjmvLtr z_BWtn{gV_!BtS=#|DW|z0!Czi~k9E-)MGk92^s_V5C zES@*1dcH-mWe>EBY_m!qERF9|AwrA)QM80UC@2C;UY8k9h$T9PV<`+{uAJ(a)Vl;0 zRc)%CZ<(W{kCt@(sB2(3u1|#sEfz6o@ePfy4VEu^G>V8NAeLjvxyrmLsiT*P$FXMz zsd~Pp-r1gLnL2dbd$632r9y<3RqONphNXYp^vT%9*-lU)OSY4nJy z=UbKx(?d&rtCL^B684-55n2ivi_6`^E@08OO06K4dvTn;yp?127flPN6<~3nN7eH! zn@fkI=)3tt10n4cSREW^x8jqIZ4W?hfvTeq=U&Io2jAL2r$BZ%UX*PSn($tQs=UeiY z_eIOR*oAUnIl6!f5n2q7p+#+yawo9pU235){jb>`UI`q_BLl{gou}&xmS#7odcMW{ z4AYm(pGJ=W%hfz8L}-yqK+8e@qh?^4e!Zvxu@vzv&UcxoKVKq^AA%+D9#zk`w7O)9 zeYt+)!UeDd6i^{TOBQ1Z`s~mOEWX3in-WV_BBw8F+?iPgP3*hkThd!&sCvG|CzRRu zKUXN)fThV~Dnw`rPQ<>9-FVC&EV;HD+Yn3KagODIJ97iL%H7n9I( zO6wTDjhl9Y3K3e=lF(ujekchnO5;0d5=$k|GP8=Yq|^&}2YpGKM%D8zXSbN6WoP@I zgP||=tf>&8rG&9`Ia_lQEVtudYZFWU2~J-YWim6oeH^T>faTjss-ACo)q=4+xt@qC zB9hKiAwo;y3GB<|cByJ$v7eZtPb_K497}4)!+4$Cs`~>&uw>1o>iL$QC(W=gTW+N; zgT73)r9y<3<;iGyFnd}eSmu9UHIP^$QaF~LkD0gEk)u{W221<8tEGCr#ZYe?TAmzu zjc-bxG@wF+7ULAOgm<1j3oOz4!%c|A=Oo86p%JqZvf#%IoJpK9n5ySnRyvJAi_-%+ z9q7xu5mboKqIeQ5Z!2Rxz_Pme5n!mk4JC9k6_|JUES55>q*rnpozG8WXh=@r|#VEmhCAbci00miNkU6Twn8 zp9&FLtW(kAQMVjdr>~y1W)880oaR_s}?k_xNWLtg?MsCvGo-K?=_Non{2SKYWR zqC$ig&C_VPA=d&|c5S~j!k$>FxR!&Ar6x>sFjx|DsCvF7vF|9fxET+?x7T0pQ6WN0 zIb->*F(VwtSl9SQ&csr1hSQgcWvB4`a^+Qg0$B7?^%JaHqIjI$iez}L+E%ku{MPGA|;i>l{a&Qwi9%f`bE&VZ$ZJ{2Og zc%4Oy!%WSd(3jC=hdhZTDve{gX2Q(Co10*C0xU(Jsd~Pp_hVzU9INb!Gl|1~P$5E# zSsGf@KPA2Z%N+AXUc};mj$>IL&YUaH&#G_)OXn$6J>PQu;V`t+7`9&keR)5N3K3dV z&Y@*`iiZ+d^qvjaKrCfEOYCN5wQWw+YP{zER7}X z`{pyjQaCy%gjkZ(IhHtI=Dt#=;~`wt*>Eyd&$ncLVy=@%?|ym&`jR=53K3f7rK6?a z;U7!EGOr|bFR_GXa4fABGtVvzhL6W<{;3sIJ>POOig~8=?9|;}U};fJg$OM=8E9GY z#dR)NOdriXKrA&pOS37-cz$U-SJMFY{k<=#dcLLgeH|QQZgtyOf@SqvDnw|hVk{lA zQWk>6t;+Byu@qk5^yR5LbN`}y>)8gdG^o{Is^?oq$T9nVnrdB~N&HKR3K3c|E?{4l zHQb8#F9$m*Boa&JMUG`nGv=Lz?f%#XV7b?Ws^?pRM;l^aG&8QOguXcRp+bZf|BGm` zyI+4K^kwe;nm0EJWO6L~x-fql;HDOHsEtMnQ=7Xiz5Gq7yX_bi< z&(r6xKwsp7>@$g_oM*{=7mIy4WjtISER%Lo^?b{(dd&Nmm>A_6uw36yg$ON$jAe4; z6%W92ZA0(t#B%R4r!OB|qS4}gW7>VNdN_qyS?ZOmZXnVh|uDC1ub2h z&BZ_Pu+QvDKCy&diL%ax0v^pn+GHf086JCREW@Gcoi)x zK6mnjzMT0S^o&@%vN)F10nGi2!l>#@uvGn|>iL!szfRbfjf!uMfknUG7HK1d7P&06 z937~M&&DHtr@bbYBA!L>{&DO}d;_~DU^y5})$=VzIbG0lbiyQDo$kAv3K3ee7|WCG zmzP6d+D}t|M=V*{oW8`LVfGjM6#Y72sc%iy^DT$m)X>u3Zfbw%%PmJLL}&@l#=bml zF)#rvQTpFL5lh@Pj^$_*=KPYnWql}E_6(-#`IfaKTA}65{>>I(@fksd2rbjEq2;jo zdz`Vlck;z=#1ed+W3jlJh-1uPc&ABdnN8L6Eic+NN6Wi?mMzd?PlX6AYS+;+Dm@|y z`jR?5Sl;rlcdwN^OY%(S8SKtCU+aR!$A+rsTNXIBLCf;uxj4hy*og`eT1ptpz^g~J zz;Yyhx+1aU=WzORPm_73bi1YJII!G1N!9Z$lgpc+rPGWPJE1Rk&r=~nOJWZ8MZtW_ zbg+bXP;X2uX*W2Q9b=gLN(h`j;y9LBMeuZgms(&kX-9tHb%>f+Bvw=KE5RtBDCZ(mQS6&;1@$$ zzc%kdEIGNHz8IM?*Zfn<+80A#Mt`8{`IdUlifH-p^j$Rc#o;RzBD93&VqXHxlLtUw zrhaO!ODu`EITq(#%(dm4wr_WW#ioX;=Ubl6tb>-?LvPFkOZPf0q>T_-tZ$cF3CHkUxlFr*F<6=nq3Zb-%P)-O+U!TMU`ZN5g$OO0chEBLL-IMW zWbIBJOe|GAOPj{b^;l}F-c!J0e2}W=TVAiIg?;I-n)ebconxsGp{1O$oZQ)Y16W$u zT5n1$1$Q}pG1$O7yBOE11iut^R+*~jTb4cif-CNtzD(=@mbOi)5TPaYF81YDRsU$P zeD7jzPAnPsI2Hvz=9>Rt^S&>jFY4MTM=$@Jw{PR#?b%r>Axgch%RXel%@$9YOm z6k1OrmZ&_A#Ydgl_n(+v_y(4?lM)+xBT#Zi{%-q6WxN^{BD7R7mW3VaS%W34O~QI& zDSX7~%j7-GU(aZ9;}*^xJ)ut3^DXrpnf*npU6X!b@zJ6}gqDm)*q2)=%kX+^b<(np z#FF`#V=;NoT=SdE?NS8$e*Rgio^SD}&7AjF&Q!+vWAT@$5TV8YFm{ba z#1dV=vFz49gvVs%K-cqN@mWLF^DXgj%5aPgoi_@o2S zMJxeNI2Hpx=Glek5d-{cK&#L@jlKJ{cS>C9IFvj+cp+ba~R!`7kKm3&@>@Swi z_x2G>InUxAz+784>SKmqanfE&)$=VqE19pK+{pTc>n!T8p+ba~LdG&+;r8}mvG@@g zMJ)H8a{AKPk$G;}I45ujjIo&t2c&wwrG4mq9Ah6cTP=dVjA}%M2rbc1u`i)bj^%-6 zR?vc2Vo80*vBWKo!TqK3juC!|a@{Vfo^Q!qn~Rn*+lhF7=@?Fh2raJ9&@$`PkQi9d z?0PaUkyy_B%ku99&2BbMi+bDG+AeXkvGx4V3Rv4gQ&45NaJVU~w2w@uk_0R5=igGp zl)$ZeM_{G>imO!0KdrPMV`+eKy&CQx1uO0A-J;_DZl!&Q@eYhq^D0q*mG=GbQ&In~ zmG+I+FWm|&?N>gbBE>81pQJRyZ5!U|$pzRuo)%F-WJJlVwBNkG0z)ob&T9)R?O(s8 zLjK2<_P^>hc?2u%cYmV7WLMg6)%=d@AziRFSCJJDd)$*O8e3sRO~;mwAagPfh+BEJ8#d0mG+0j zsKEbfrM*RuJ-E_-U-KON4%hL6RHpDK;lUgwVn64sf|d4*W2l(_ZKeH?1+#G$&WRJ- zPmv=|SRv==IU#_#+Nho$jo&4`lt$I_kDdnW?&8riPH}&4I1@x?QX#_M#jp^Mp5M2d z;9XPY2CMVL;`N+k$)3u*U07Kw1!t~o+Dg^)Ep9o?H@zAh_A~&?s@+tG&?5I7Eqh)} z{sd>Wi%&H#6H5`#l5~kVLl3pQQUS-H(qgKfZ)w@#7WO4?wmq&;b^Snv2rXHR zFTt{IWzHR9iF?Vhbkt#<{9ZD9ln8w}h)1&@%lcTAn%_ z#?>9&HFiEEmf#|eWpH#9o}wOx2bX~5OmC{5Z@F;p8ulgq>TF!yak?KBBDAO#q2;Sz z;&$jueWf{t#8Sz#7|JpGi&5t^e3GTCO4ai%d5O#zD=iCfgfUjrnhFtGN*IgWOLP2w z|M;XH#l({TiqjVjUuITxBaN60=*!TvR6XA!U;jGxrPaKr%#77bREW@$_zL@CwDSYb zyg!xq=^e48z2;b|3YdN0KwGafSWZ8s>iL%ZewWc=nrNQ_eYx?P3K3eCzebD7h5TPI z#!?l|RuW4@F~>6bGh=ysGW0N5ZmCf9e9Ov`i)h*1C#XJH9=D=Ggcjpsv^-c|hl8O8mc&wyMS;B@8^2KzXRF`uPu24+&n%hqONGVbAQ)pAL#Ys<#kv$N3zSj^L0_uB zq_rTHkhdJms4o@|9vHSTgHTAwrAhTeST2?~89q zl@|N7CzdLnrQ`_nW-&3rF90m1o>V>GvaI`Aw4|JzjWbqftfNANmU6~oIHvpq^yRav zMHgZzDC6{Hb}Vx})~GB>9V|6%sd~O8X!c38d{gR*s|d2xsSu$hwG8{RC1g?sSaSEY z)*_aScN|NTYUb;MPdmKS0n4F-R6XC4tet?CKEX3eVT{d+r9y-juXkwa>(>cas<iL%0hK!|^gT^DUY{;ZSgcg+#Xjy2n{S;VS9(oQT zmNK5D|7_-I=?J%Vu`tGLUQqRXi*7jcw+M#ow;KhPcBNE^(2~zsruypHgXMwKuo1+P zTh8gr}s$)S`o;mApyB z68e#2QB7fNKgONRo+(x*j4Pq56eqe6t1D#qes((Wl(QuLPE5KG}F zPG8LG9l`x&)aQ>8V7W1ns^?qOEt$V{mv=x2+68#nXqBU|FuIy4s?Y)UuQmZ%? z-D>7u?A`gmCeWAR7pQu^Wrh)R-v8Oct^mf^m~1LUXmPDV%jKzG@#|XiwtfvJmauA$ zyzOz+B?2wwWmETnMQI}yBD74eLCgH_AG&}gI-oj5@fpf7eqsCvF7q00^&V?Q51m;+;M!AL4ZXz^hz7yZwsfn`&0 z>SJOl;aM(*GUrO8w17rn@!d_;^DT;dnZ6`>M}7cHSp*d#w3INGV`qCE085DDx);Qf zv!2tJk{`^zUvGLiep#=*8&%J@xJPftzVtY6kN5lSR#G8Ci@7|Gv1l8M31E4?!|V;Q zBzki!A-2pjSTn66XXuN6I91QL6!>pNi(%d!Z|KX4BUFgclE_#(&de?VOV?WJ<-`)= z!?9@gV8+-4?@oGP`JznK^DWsMnK2fn9Nh>kNvc$c(9*0H_GQt}>!x72s#^AiSgLrI zq{w}E%`YD@9aneAwWI3!mJOYlztR0gp$u0J+o@9_Ld$Z-a_!Mb{Hp*H4xacyECs%t zzVzx9juzY4o#xP&K5}q(EP;`LC@r^RM^F z8Gam#*&^mS%*OgN@wuh;HmaU)F`2|%Tb8!n?GIxtYc~}lv>4aMG4}aw&xc?!nPgU< zSfVy?EN}dneLwui$G*^)6SJs#zGcN2QH`ABTz2+PMOTrB*L}+nlEDgq8$C>wY><)D#mfVe; zzEs#S*D$K5m+k!a^_`1B=qHQsEY=%ByZwa zqP8*bUf)fLjR8whBvsG1H15Y7lO1jpH35s$Q7S}e>01}a*w^?B1+cvQF`yT*gl^_o zI-4^0$oN34hOyjg*EJd}avoHO&{D%#rnTrf2`tAmUKkNeVF0HuZYE4$ zoZFqk`UJRHhP3InUDPB{Nr{#p{|( z=u59es-ADzk;?4*%MD5o!T$2<6cr-0lrt97bCHd~;<)$GEMmDA%<0RKdCYxfrsJUN zU^yN|)$=W$S(~vhg*$KHvp(%ODnw|RtcYVw-tlD!SX#_KXiF@qTRE15B*x;G}KeKaIkz` zLe=vvznbhq%Vt||{HkWcYAQr%X{UsJF|RB41T6c@`!6OIuk9Sm4fg)U^7}meqVC0N zs-AC&>Bfw)s9Fu=p)YIXjiikbTD%yG#n&b{Pw8V>?d8N$#Ip=6Vy-P4efo6{`r`hD zs^?oqYBATAtw;9g3YI!QsSu&1h_O7`cl8_eB_-qe8e+)`;q>Lp8|K|>P2;h*!SXem zs^?o=wfu36Jz7(w1(v|uREW@Gri^23d{$3AusmxX=SwVcJ2;l5k&H#P=-33bbfD__ zmc9`|XtAf1{ET-#4(nP*7Mqe#jyF30AdN=$+2wnWv=-(e|X^++ZT7B>iL#V z7uKUiak4MIdmXJog$OMw^|3F#9!B82aks)@+li%;XZfNRfzM%{?y`LieOdR0s^?n@ zRsGO%BHPRm#@L_=Dnw}UU@Rvlq~SYphp75{h$Vj)r!OB_OV^1{@tyd#c&eUnNzY-f zVY*G)GzctTQm7E2C7-cmhpXVci=2A*B8Vj|lw+CREDZb7Tj?}j!*pvx)$=W9~(r-eS1)8so&KsT74ko>POoxocnVghC-tPlWLKy6&uoV;2Ag+0*A@n58$BxEZ~hMavE>~x zN#*d$HLyDUXcJTESN?BTr^{?$C+Bz5VRibfc2w*?uTEdOpeL?QUo~LD3|O6hM4bx! zuU4m5HN1$c)8*9;UxlMYV=XFEc$ApW93`&2d-WYwr#I?D#r$up(~B2hH-Izr=g&&< zmk8!}@p|&% z3|JO>+&MukK6^M8e`n^3w&ngBw7gqQ)$=VnO*f)t?Tc|vVDa{)LWGt&4e?B{@#h`< zX5`KtVP}Y?glF+`VrJldOB*#1jzK^7Q}ujHvI=vxQIz-T4OrG5rb2|4d5k6fhxShB zOHAs53&fJMm(!Oz0nDrdgO}Qe!BT#Ks^?pb+WKN&`o^{H1eUaHDnw|>VJxjKeZGYj z>tR{Mk{HIZq+Mgi*tyO_@NYpcn@`pAEeZXZ{pDI0J$#a7;zorCEjo>Gj5VFv5NCTd zY+3&nv4rg7SoDgSCs|2TtuI4g5<61$d`mMo=FZ&HdN^mBE`ZxrntO@#<8RgC4(Rh7Ch#=2hL zUqCDc`#F87ie;YuygJ_#-;(Ozqw4vVu;A6WzYOh}nGSskEucb#mRXH)jHwK$tOiS; z8S`HfOGY@y;^oiW0jgdeh^wIjZK-;`<((RHbzZ&hSUgzFov9F^C4;f(yiflOmJCJ1 zQeugU;8;>tGf%Pe^hS?^{pD&Cs-ADT=FS|)RtAKw1xs9ODnw}MqJn*S@isLBEd4d) zDu~5Dl4CI#&pc%qZkLZ=Kp3b))$=X?$Jw37_1JCyQ*I`140-Ew0S+yxP{zU@7Go{Y_cYB#YxXX6dx70)CA@9YXWgaiMj-7u-l4vas%(8uW`X!hzu{-XRRhGoPLS33&+m7>PS?nx4^ZsK1iI**_ zX0tcHq48zcfTi$Zl0<7sWR~IE-nRwI#7AD0lqGVXU?~>Cp2a%vuWJiq%rJ?>%a&); zobkS2a_HJIV0rYGB+*(rmclXCcjJRiUa%d{cPnN2C|P2curIztH+ILJUbm-`c-ivh8JjON zywBp-xB>PgiPkcUS(+DpSRLw;Km2M-Wr;f=)Fo;%+p+rDc@KV#D}9B;%a)sM-Eh9d zg-phG7t3Nu60IeUSsvu8Vh@(#qukpoOV~lda_}a*?;C~GwuCX}HkriB7Rv=}zBo6p z@d?IQjaej#*3zsrjW(jtW zZUvV2qev31#fMpp77WKR)?q;PKFZ=8Bv_hVWIKs_+dnjcy7V7O;$@53dbYL<`BE?h zEJhPZ60OC+1nbiKcq3mJW96&n3{sY4$ztox*2yjRp1{+BTx%T^0^-%l#P`7YGuRA&+|Tg(?N!MXnfs;BJqJQ@4&+mY^eoWl#`1skd9pw_U(8ejbUJE!PjsN6UbfR<*!l zv6LjyTI!a;G1jtOb+iPyjhd<~?ned7vc}tRJ?0u_izn=aZz1urrMm@N!<0Fba}6xc zJ4q6)#hqF5-3)66W6a8?<{V|olq|cV*>1uf%LY_}x~%9+;$_Qyb9NoumG4b3SoRMk zNwk(sX373yR~{?@O};EtmiS~APM+SlATN@HDTXuYR!7*l?G;$(XW_BS- zw3Z>JIL4|QPC5jZIepJNDNFb zDvNcgZB*J2Ec^2gSg$NDCj?8_Aa>smG@N=L>N2$`iI*+DQ`t57!`Ly;p)MvSB#GAI z!Yuxd@8^JJubsK4vZP3stU~O0|4faRZ@_ZVp2W+RW=YIa>CvftVA;KdB+*(@m}S$2 ztN5<3ckVaal_e%bsLPEIb{#VfHOJq(7*~kI%a)VpopA17_E<9uEViXc60N0&8IG}$ zLHWI)F0}_;@>iDNlY(W)Rd(NZG1-r6%ax-@ylg4do?VlTt`5cB>YpZ(Bw9-_vwXB^ zfX~S0Yu5!Ti)W}{ai7SZ#j3_-oq;iCyqUzymXK@gU0?Nur%o$ z9Cvv}%o}i2S+XRHk1gv`E7<4z0kE`oBJr}NbMzt{W2y0GS73}iSw)g)Een~&`0^Wk z-XD^?QmC>doD%A?@&h|DD_0F5Nwk)><#CKZb!BlqBUr|5W%tU0*`cMu^7J@~mo2ONv-wi!=Lr1mLHjc#iPmCR0qe4|_igM@ z{V<@@9c4+AEXO*q6REB{r&xofkS&RqEs2$v;TQ{>dkar*tUHk;(ORZ4%dss9d*S|4 z*}Q1Hvc#Sh>f*MCU6WUK8`lRcF*QlNZ0U54JtGe+yyYFt{h5D}Bw9->vjmvBlm*Lx z3Xh*DOGvn2Xyac{iOyXrrrvDtA`>jpPah+_MMv`bPqbuSVTQt2sb`IxjRQQLo zL|zaqxyG_il_9w|kA%8RZ$;u|%fu4wJ-?k*13WS0YX_1)4!p)NN2?iq~y z_qW%;2*GmRYBS#V$Jf1e6f8}Sl6cwjRLBe3l4LE>f0hO+Eg%xe8?{Qk0e07;^?xG_u3La|H0GUBRT zQDylkSxN?O#kx#*7J%Q8P4AF++4AMWT&#=t_SSh}?)%4+BwEWyW?8Z9)={uDv2R#f zS>i4Ub@Bbp*8E~kRTvgPzjN3`s%@0bCWoh~Ga)-s?nj2u$Y%;W2{3#`|4nMeT>A*mWqSe+`s+VtUFjPoFPfHmM~^{R@OfdEPj`xswj(3 zq+l7fn!QsBTmN=1SboNkc-hjbFk4$bsCgVu2X?qml4vdEt6*Ks7j2#jmZnYIYblHK z6~QvN5&Qn)w*89>SlYKC@v`OADhC{6*0sOmiMkcLkR)1*GqYs)d*C&>xW}~m%91Qu zjApab=sLD?{sfjuJ4n22x#G{(Fy1p8y@fHBZx2bLwInmkpyUd3p)Q*ft(q!Jbd*pR zzsbzZG>sV^% zC8xnsFrfx1kqc6Oq&xWovSNdwt?%Y^&M z-N3TMlEllFz=I2Lz8t8&yboB$G$u*3mb^8vE-rb7EYA- z?Z8rFFo~BfKP=gP%OmYBRRYUXTarX;iD8yCW%KWWG4{&BVu`W@-x4gjRr#LE`9Rnu^crTH3P2TM?Ml0<8%UlYfe|A75?dgDO5%w@{rd0VgqUSRK(n%lph z4wmuuBwn^0&BfOIshh83|4X+eB#GAI$t>+_c;h$mTz28E%915n>KAA4lMgxdE&-Mr z_9R}mTztXS$$18>ZVq+HSWJ>=Em_Pmw?qJ*4m@aCfL^W~d=t6@+V%LOEf))K)iXPn3Pg(l!*7uxwNi}&4s>`vc3dlar_ZmvoX zh3@pQ>m-w=J3X@2du*rm$z8KCbf(TH+t|u+6uL(4B7ggrxhY?)2;< zZkQ^jL&Q4hPM`9Ur20Q~r&oVvwjR3EuceZd`tEcKzq0tZ9ZENeg4xmO6G@~-)Ze<( zL(jj(BnN+V`~}_VGqOmM|FJuL>y0<&(4FqOy#1eX_4n@d1D{)BPx(>*5K9=CecVWb zf9OtMon?)gmXZzh@k+wSzg{ee?XooWW%>GgMzwEwv~J$h7I>`wP}bm|J- z>1qBX;eXYgZqw%jPN$C@PmY4CM8N|jr}ipQwiaIT>`PbP2i@smM@gFhtvfyQ*1el> z^*lGrY`41N1pMoYQ`xAZQQm1Y?d{te8RjZ!P$loP25hx)rE(2CH?IF;5-(ppH{@eq zuTNfS{{sGB3VD$v+W(ho?EfXH%+@Dh3E7Yos4TAc1WVv0_B30-amWssx3#=Tylh$i zf_){u`p)M7Sa$l7BwC9rvn(i?I0!6<6OJBLmNdzdYQk3MxjO&q2bPdmBwn_>Z^$ek zTD3?A%gS_;L~BW7mY|9c-N3T?`QlJziH#HLlF7R}hGhlem-sy?Bwn^OHezqGHay-u z1T3vTk|bJ7pV~P0VsH1c1Iy!~9nUFC$bG>w)P{B2z8ZHCJ5}Pxl6cwDJBPjbU1zib zcNHA7BT2ND5N3Jt(jyHlho4uBR2HuXf+giU>$IPk))PDJE2olp*)nMw`%=<)O0YN7 zrQ9cyL~AjvgLN6@{SKeZ`_)Xnt}NM-C3gX~3+IsWTYQ@BS)atqmO?gDa1}VL!6X~7 zWHu#9v=#?uIgzqC4C*rc#IZZdlK4=li)}%6O`dgVHg>994JYxk#V?A@{mIwj&Vj`y zk|fbu5}BpUkmdjCTmJw4ZTS*As-UOwvUp{Qj2A35>aiz{o%I5- zmwXK8OQq^0iPqAwE{?HOo1F<@2|r=;Tv-Ai2^MRQO}M0MWBdT;erPy}mo4k=vp0-$ zoXg_Z>pqbriPjRxEaNAH9|22jM)B9m;`Ugu7$0SI8C}Hy&sI0jCh@YRW*~bb-!o}D z{tn5_JSG185n4-OORUSItJ6P%<%3y5nzDS9EEyfyZjTi;e}9L2W$$VvUbgtIVY_g~ z=8HvLl0<8n#Vqe9&HV|M4(0)$l_l~y5se)YgEC0GY#H8! zjj`X8$1DQNz3(K6*5bn~n@coY0b{I5*vu}&!ka*eBYa@G; z6?SVDUXwlYCH(m#w3hMAVrr6^59+dHVqj@yiGD8B5}Vjo@kyk`D#;h64YhOeG)HQYOi5CB%*HSPXWuE zrzDBi;=(M6Kl zjNMpHl4vc#%u=RDNIF>R%^yEnSv*q&OHvp1U8mcN-p!#d$DK&LZ0T@u5?Us;%!|Kq z7qyxs(OQhHurAZ*Bu@g%yqE@)lqE~D-1TN-%<*HV8DPn|PvT|E%B^f|X&#dN5$e+F z8A+nGEM%7a*;`tJW$u#f>B^FjD%8d5IolOoxn1g0u{8p5?!H7>0^SLhE0*lJa@>Sj1EDV8zmj;_5;cK!ceFeCYChD(FNY-2 zS^}6Q%C0Hi_t&})U9K#yX@Vt>E9;0@^DAZxSay4pc-hkT3)>wvtIkUN4b@KmB#G8y z*aYixuja+qFviTkS6QPhX_BR5VYVK-x#sXau#C=A^-sKPF?_+UV?HlO;m&)DLL`aS zGL2crc67m$DO!*DxKUYR(}lW>f5*ny;P5H!!Sa0yiI*+qhOoMXmL2g9#@L(LB#G7% z%PgOs)~AblfaUpfl0<9q zVwQJ}syjekCaxS7pe)&vIC@uN@|s}Ag2(#|esMqn8mPU2-ttH0QqKjnR~;$V4rnIzF#MmNJT z)^l=F4%FrPk2lAZCGwMCF`43l^JTpA>Hc7`&YSWlUbd8+!PYQW-mKUTmRE&H60IeY zSt^e&h`$e0^T&xWWeLm>EY3C9b*$a?>9|g=oG0W@yln9>W#5q-Z9cFMEQJb@BwC9_ zbF53Ib&w{05FqC_j>;`~jptS!u*#cmgUizn<B-bUhO%ei9gUfKHCQtY(1_a#ZRmg23kF1@?e z#^*}g71Q#J`uBI_OvzHklAV5HZxDvN)#tmBc-iux8yjQa&P3tqjh{D@BwEW{W?BBZ zCO(VZsnpC!S>k^Rb$RT}>N49m^)QUFPqj$AY&jFb)?;qpYvL)i&l`{=T1z~$B(xf} z29^j75bN z?G6^tC)dg-i{CH7@+gPh_dTquCO}@wH4|TZA9W_%b{e}*>$c`UEVR=C@Fn*vjlT?cFDRXaSbB9wc72oLtOKrdU?9V;QhmY$r*ymKbJnZ}g!!)WzcS z?}o|}?40NCd*FXeU5n4i->!RLr+rj5iI*+&2e7@Phx=D*4VLkF4I2Ibe}vXjzb%fj z$=1nuLTH$MbaR!)bD3aC8qDTP*DJq5p)R`|NW5$*Wy`vu44!1H2g?!{l0<9qWR}#K z53fUA3}0_-t1MZPCC?uAu5W2_i6F2X`9R`j%VvM}UCg+O9d@VJ`9_jxEm_Rca#de< zu=Kk(yo<6VEEnqHz)ua>yU=(dSoS_8@v_C;mOWR_xg6*Yb@6yjl4va>+Tj@UJv;v* zSiWwr+Dlm?RtS~~er(PEePTjeu+#}4@v z#(ZV*auqCh%d=el`?VEUh6Vk5N55X@ zqAZDPgt|DMX4kO>aVf!YP44`G#LJfVx!Bq=G3p$C-%t5Ql4vc7%<}23%OI%BsDcC6 zDNE#9!D6_UUB~V{JvSOGOG=V>*>dVWd)JrZZNC}n@}Mk9qP2AFgk!9^$$LD3dU>nz z9?BB9PO#J+!QR1c8J^D$EW0|9c-a!zj(tn28SFg`EVeyK60IeWSxSXOynu%0<32sT zmF3#MEdTI3@YTAG#jCUX`l#;EY5#C9$)xGDKXmUtwg&qQxR45+_AL&REdQy~-h6s1 z{G(FOWo?5_`^XTI?w>mC&1SB{R4wP6Xab$~Dd8m5|Ebe{Xx!MI&}ly>lBCpk+DChp z!@tdNb;?f|QL}H5L~2C+tBKF>*lf9SLyyT}$ZeVm)x1v>3pz9O0a+fMtx{ei{IPb5O8{j2vR z?SJmHcU&<5JMHUL-LU{V?T^ju^=B&nSDp5inij>H)^wVlAFdL27Lc6UtHd>Sl_=V{ zTpV=TuW=-4{is^)&gxc19Ww7}^EKda2PQUcP$9^k6%^ z=GKcH3;!=SCXpoC|Chp@@lLSj;`d0ntA)9T2Pn%&$+Bf5+Z}c9;W^wfdOT6E$nn~=xHktAA6vo1KsMmyKTFTNgCO;0O}&ql$r^`a|Y0wTXIGJ(46tV`l$%heS2 z7W@8%0G7^1Bwn_-2ebL| zw({dJuzV>&l4vakHdvSPu2=D_wa{qz2d%h-eL47@GF*RwTD zcalVF8PW~MSnWE6D!~{tT;TXYS;DsnmKL?xo2=nY%&}8trZb6`Eju=_uM1`ceQUvW zY@{nmqP2uGi`S{FrC{mUpvzZf@!Kj`id1EFiG8-t0>)V3mLy)bEHNLCJBb@#Nyjnv zs69!dwN&bkbxGa2H7{76M;861EH2vwixidY{$Ba#B2A$NJqT zNwgLhW=Tr#iKhj5mVTUP^uKp|q)3*pA6DTSrbLD-?pz5mC-Ji7_!M@3Ne|S}|`QB$>DHl)TWlLmt_WV+M%W^!ytwu`Ut><=bM zw3Y;B*=F#m4$S>2%?{R8mIxogGQytij(T)uYY(tAZcpN6%ex+H(DHW28+;b4+KnX9 zTH5x)F}9}jDtwDQ{@t8L$`Y_suq@mW?EuDw5&aQr4mS#X*P8;0&^G9ec zvCML#aaw(_l_(&2jTe_yQ-BJC9?0*WD^a&)1)?(EM z$5_SooA68g{ALRVDvOuDV7ZmXzF=)~K8s^)SvwLhTk1 zd;U$!LtR!4YBNGvvL#FY;mmUXWOuyppB+WwWy`4*>^)3j$Do;Dxi^U<(OR;ZWmA_a z>%mgfuE02DN!%^e#j6Z^*Eg%qJbZUyXiws0%ap}z%^$n4Xg#omFD6N}meGB2jJ3PD z7CX8xZn!r^St9oc7T@)34b$RN5qzhVv5myb7AqT87mM7Dah+V$k0jArBAF$7XHt8p zOW?j8vy>$;K(K`DV>|FJ4;qN?VWu4=@v`OI4t7nRHm3VPs7tp~B#G8y(GTm=dE%WS zU`hQvW`VM}?G-Em?b)t^;pTY*;F|21L*ivi()EeBhDjK4Cm1ZI`N#hGBeWJbX6Z1r zO((G28UB}}vV4>*sm0lPtdEb;3$WZ6PvT`uXMXb0r{it#ntXa1NusrUWR}*)E|vg` z<*4s2$`ZFvs7t~Yc2d~>?xv5y(rGe@mo2r6uw4ajCMETSx?G=0l4vah`r{aDUH*Ox zuq@bfahM9nEAH@iE3`|)}>PKq7}dr zwSR!OvN#_QEIwvzzL?Z6d;{u|5lrG`i`4_RP7Y|=d?{Ftgpnj#i!-zIF8de$wyoKW zI)2KMELqNVWA84uwTs7hO8pm*c-hj@cotsA&h9nt1D5o~*F5hU zqAcz~f+fO&tu0OW-!Bc8KG7szwsdLB_A2;qnp7R?^5Zs1qO}wsgmp>IYdHt%V(&EY ztg>WEmbl7n--S=Xd$>+M>`LNg%SjJ5_y0QI;{;ghxsxPX%UovZhf~!_EfMeuq5%aW#~wDuPndw%v7*6YD|)7Eiue;Y>M9(und^-?ya%}9~Ufb zGTAe-w{1HAR+fYip)R#@F-usP z!l%HJzblECEh~aq*G<`ieZ9cav@c1bwTu{sV{H5Tn|NxoN428)$Nc*{a>Pl&QnM&K zx%IcJkrPC~9o)uP>fKb8Gz6N{V zKU47Y7_hW1O5$b9@d$Q*$$ihi23kx=60N1$aI8zeRp$6yIWO1#Qp)0bO0c}=U202= zO-_U5s1b>mEpPTZ;j`F|SLN_T-I^sy60OCRS#piJg59=>#pakPOPXYEosa$_~*MzFvgP1+ErGT*f60kHH)+N$-gbq z4?tahRwePWWo8KLAUge58lI+owhl?6we%T*V{HAG19-yDU(bvzlqKY}VA-;cy}KyW z$pJedYrP@yvc;$foBK`;JK-M1Q6ESWttEt6nv9rc1a*1R=w3Z#@j4?|UhoqSPhXw9 z7V1*CHHnuk;fq$|e989M(GJE~nNB2$)?zvm>terU4R%82`YX7xvSdpZqtEO*R;}d_ zJRP`TGZHUbE?KhQPyXFH><3sLwINBg76)cI^(F$>V~foev{IJDvqD|Ue__wa;oHaH z+H!PN5-(f!xX!@3bX}j&1T59+kR)15BD0ufvG*`-uD0%=ERo@YrKBah?^|YG$KQhP zA4}q8OSinNE}t)783&fcha`#C($N;j*l1U)12A9iS{3Z3EP>|)OD`XGvfk(?52nI= zX<|*{WlQ<-Yq2iIsiUxSxN&EaL~99TmiBSi@Z0N;`>}nL#qGRc2{L7C{%&m=WrF47 zGZHUbY@f30*lve+H=!;euSpWErSK@M%R}RSHJ~mHMtKcZmXDHU#Z7iie)aa78(0QT zCh@YR=rp#5`E_B44_M01B1yECS#^R$1aM2zA-llYQe0H0kvPEa#t+ zc-b;?6Z`&Be`3A1V43}jB+**pnB~^Jq)A{Yy2^5bvV=tlmIl4q`>`iK%Jv0|fjfzp zEjh2)80$2v>@~1F+D4LSEzL&b80%I4xd&LPSNb|tS$r-E79ajglbyp`3;@g2S|nb! zSS(}D`wIr0j{(by1|*5r;=?R`4o(~omOYg(&QTWUOM+!vWp)BW%)c6&z>>cXiI**B z_{kK{iZ6BrOUFheiPmB;2J3PmcL(g~{v5S!k+LL97Q<^Ra8Iwn_wo6`GCY>V%a&X3 z**f{SPyRh%$sbRWXf5NJ<*rdy0jNvNyHU={5`9^y%kC!ZUODjh13UpC{}&Q3Tha=# z^;q?f$@uGxKG`IR))LJu*0qqYL@EQ5oMUq5osXG?On0?8_TF?Z%@?E7(%JSo1mVf9@U%qWTj<#!y7H);^bgyqD zlcqbpQFbJDr#sy9iG%L+&A&;O|J0rS?I`O`_qiTk47$@B7pVM)_fOsF_0riJ+A`;V z&xP*vaYae0|5JB*^@}efp*ww~2}!B%PH$*f9{=wRr&s&{BPy^0Nu);9-@4OV&tl!_ zU536~3f<|;s*xoBV|Tjm*9nuMJ3W6LlI-u@=}G+;V^4XHo(+D&z`WjoB>0E!^m$*G zVy0Ki-8(>cxA&qx|Jxt9dcmB_(48LEhNS(^-RY)lCSiBF|D6kA(4F3`GfDVg zb*In0&jyCm_mTPF>OHU*$*H|cEM`}Uv{N&8L3etIK_t!p)}3Ck(bZ#c^}G}i=Ao`Q z?pK7X=j%%R?a#nRc-Gp_!z5n5dOmo-R!+6Dw&IF5_9RK7{eN+1|1Y~ezLo}y_Zeqz zWyzE*jbquyiDR{%jfIti*KIKHSwuw*$t9JYsq95&-Hh3wQ< zM+tSwsLr}>YDcH+21}I#Bwn_Rn8WVSa~|Idfbse*m?Y6!hK$3z+I_3kdQg|8LyY$; zOZZj6l6xEbO8OzJjWbxbj3M!|<;ZzYe45Qy{Vkq>_iGACqP2uGON-vK`+>!1a>8L{ z@rxEL%QvwxR?RVUGgvZbk$Bm%VFc?USa5ZB1*prRg(QjAQfWNaqV+gq{ufxD^dm{M78hn2|8BuFuw-7DeO6gg zB+IC5))BF^q7}X|F}+RVWlPOgZ1?T_&xJ>TCFcQ2qP3(jOZ^81q@TWIBv7r)l}#DNDd@!SeMG z`%==rPI3H7x@aDWmo4*h*mZ2B<%PLmX}pvq(OLqSrAmW>Nno*!zw=31Tw?``Z4`UM zxWW3?U6?OTUyyj&Qf>vijyY9mKLsrJQ%Dl6#c(p#CAHrZ{OU6-&L>M*(j?2z+Uz>k z@Ys`OVCndb#LJevQ(0GMl6!bHu)Kanl4vc{nB_|2%ZH&Zo|h*4R+iX1LR|)%?Zs#H z@WZLI!D1Ri;$@5BXEtAMHM=wjEMM=DBw9->vrHYAQw1zT4mHd__TRfbLhcHd8`*3} zQTet5Ou@1{gv861G#9q!AGF({5m>CkNfNEaY6_0A)juyB1dBmr;Udc7bx*Jutz}>0 zZ6@t*2A1MCNxW<+>cy_fkMivQ1(q#wB#GAI#Vli%Z^Bd0yr$nTr7YQ!Wr!8qfww%j z`Ao39oKNCq%UZsM@jGO52rS-?B#G9N%`6oz&c@$V>0#z?rYwnZLS4eM*m^ACnMn>< z7F8wjvL#|X8)GS#svQJN*E%GL)-u`-$Jo3elN1>hl3 zjKs^98dX>ak%QA6{B5XSr%4j6#bPSfrTwY&o?w|a@>e}&aeF9OhW%ugzZP2J?^_<3 zNaAJ7Su0mu!#ECKKLP5}V+Kj0wYV|M!EOGHdUG5BH*RhO8jjMvCstbvi zEz|0(!@AUel#X-%%{r1qYZ)*N$CzW~r8vf}KO5FTS;8I(mRKuRmlMmL8iOTo3W=93 zZ3eM-7p`Ml;~CyZK9D3@OBl0EzvP4KvBpkSyD5v$W5MF@$WEghcyvNzT z-&Jt6Z&%zMmA#Q9(OSw+$GTiCcRn1(SgD$6eU!!diC~#>g?(9cvNXkO@{;-_Ubd`p zWY3l3?|In5+;7vIB+*)&ndN9seLRusTji^Rl_goSlw9eEYs;=tJMcuREp9lp zb`s|d-GskE!B`FPrKVHOLT%zmxXrhS!`*pYMo(>J-|VKUZ&U2H^^GA(v=*BgIL7?vcEY+O#1EaIEJ05N%jk|Pu`Z806!Zbh$d@Euwsbki zcCI+@`qUN1nCV-RL~99R7RMpp_}haAO4XdEEbh+)%i-(nUYR^&3qFfAsX*do%NSob zye9Wua()w7BC3%jT1)YnSeIE>YhX82hSA44%91Hryoa(~Yqu+>Plmd*GA8k|C9f}g z7VCd{3+^U7SB@mnTIMp#pjTTgVZQWC2wS8q@y~_2Ea&U7+`HG-0?W-;Bwn^m7|GsU z_@5|U1?sXgog~p(;+du1V%uC`+1PlMv$BLI3YOGQY>bt#?obdc*Q`mrYzd#q-d*(F znEyOju68C#w3b%0aEy(qeg7I*YP9IJT3P&F2$qQh*vWdlo%h$S zM|1+qm+mBq*5b!3Mz34-1e|<)vVGop&`}$7VQvb%eUa-XZa_<-#d8#!{U& z;2Os65lN!8iM%u3;9Rjokv4($OS| z)-u8#$C$&7hS=4)W6+6n$`bKLucdBy}RfzwMG_L5+{)) zT1y184Bj}%7wR%^-jYaV2}ltvx!bd6vBaN++JI%MBZ-$Sc}>|E%en9H4lHd~ktAA6 zwYgXqo6;d(U|CqR^9^NjO%*JTr`Wy=qvY=Iz|yfkiI*)Kmu|o@wtDY|Hee~zj3m)o zT$yFu;?g_8a`V;gAzoQx-wJiTqq_CcORumm0=@v>#XAoeVFYemTfu$YFBBwCB^|-m z)a6;qoZ8CboGDlu9AUpY(|c_7RWM&VeI@a-W$7|@9SfZlji&>b{Y{c+Ee4CRE`=PA z>f;rBk{7uZ1+~2`=cuijDfn?C6FXq%XnsSsB4a= zoO_=(Y^E&HUxd0$n7}$AqeFk1LtU~jk$Bl+e~In6T2jQXGFY12AW5{AXlB_st~s8t zbD{G+Yh?-gDp=YiveQn>Ke(_GEcyD7c-gY#5qsXBRA5?1umlYzNwk)_OK^C z=1YB>9i5cL{hMH^numQCyPf?MPd*yZkHpKCYxUW)nC-TDE5Y(@C`qEVxHHRzkmC)& za<$Zy9?Fs_S!%6eo%X|;E_?!Yc~hRm%a${d>{%@H!0fkR>0F&8(ONQ@rNi-T{9WsA z1se2MmiR28E}0$K`(&RRRStl~zXXYwEf0gY;oP5IH-8bZtSC#8Xe~n=aEuMyR<#n; zrB#`qLzN}`yI^rz#4O9xsziY0cO?=pTUJbQN6Y83{nEj*%Yr1)TEdyd`Q6y!V5#UI zHCkExeh8NM-ju#LJewzHIIv|M_JtSe}@WBw9-fvt0UWI20@vP7UTMOH8&< zmwqN}zL@nHx&|!2TuHoa@hr){SfWwz_ zL8pCa7D@L{o%UO1vQGQThaaAYPWx*HU;g}0{hvDRPeruF-+_N`NK)!M?MHjDhUx51vWvX-wkft7kJ)_5^5`5K|le zUnZ0%Nwohj3)%n6ZSOjsV6nViWs9;T{1)y6>9u{)5_G;D?ppJYC-Jf+;=xW_(az5M zbRk&gCz2#uO9HcOtabcYZ{1IA9C1<>=nPx4E153q~XLl=0z%s$oqBr|W z`r+5cbhw^&*+AlDi)|?T>T_bU37+FyX)8&hwe(|_?{^B~uITkkTn;FU>vF+zt}y$8 zb-KVc+#&JBg~ZF2_7SWuo}-Vygt{cTktAA+3$xq^_QIX_c^32tR+co$vLuYH0!szI z!r!~FSVrPyi^FYJmmTx+&H&44SCT|)31yb&d6uTZ82fe6?3A*^t`O={<1HIwZ-1WI z3S(^A4H7R~dX!+Rz?UNfI)i1#f_9iRkWe+?7Ve3MYL~AiyhI7B($XorPE*s3xL@JBdO2KljHhXHa3%Q;J zmQ6KDyllyt!JYu8ZEv{?>SFg7Nuss%V3wbDn?8Ug^!%zD%91Tvp3G)-Dg354?!w6% zP2y!sg&XXx+@g1Z?2XB7l0<89WR^hx`X9macw?Wt%96NBs7t<|tS%9|4u68WT+y1Nusp`Gt2Ql&Ch|Q+?TSClqGVtU|DjNSw_07#vR@*b4a{w z**TwCy7aa#0G8qT&HnrmTFXmj`8J|#ZLn+_`{uc_1iA_qqZ8~rKhH|$*g<4Ajl|2A z`l;+oNkXH^xcl~;JxQXq7%#`UA8dRK-~8Tta^$tLxUCT^bw{zfyy<#!CycQx$s}I3 zINWFJv9SA|_`3@0(?}Anr8BdfxVXw0>N0%H{B&jcC|O+0*c-;!QL(twE7Xp}%a&Y& z*}XDznAai0mAmxfd@R(Z)&deQTav!9 zd*$hlaob_Ocsr3KTFXIZ+0xhtyP#)GG5V=2Ve14-*mCxDq58$PxO3(691<^E>csow zb*%dEvA8?x?P8KdYe`_1HKjicgE7|Q<=tH4|GnG8$4#)bnZfEZuH9Pv65snBiI*)u zZ?Su&U-=>(pe{zANfNE4=n9S2uY%~7`oux-|=M4VVEyfvUk*0mQ2xN#5#vNp4^FFub1ZY`4cZ&Zl$`bA&Sjq>pon9`N9@huU@f9Rqwk#jc-d)uH{dNUd zTCO8Ww3c1WvUtn_+hj6&YA0oJ@f0i@O0&-4R_;DSpf14&NxWDizK{`4Q))4 zXe|?&<db?>{Pe8spezB~1;yNyvodlGcWevGbM8>w_h$8A+nGL^8`upOWpN zF8eoyu2GgW$@0RTb=u$G{5=&cueOtT*r7TT&p@f9W3d^ zo!pfr)<>vIMr-!Gf6aXl{=TKT8HtxI2d}X;zk^>{{9WrEl}Qq<#bPzi{g=I*(GnQd zYn!r!>=Z0-gW23adNQ>G)MeLg5-(eVW-!Zg|8w7=F7^*e60KzzvrI6#jl0&4eXp@g zS-f@$mOeY!&ik9~!g_(FY@V8b;$=&-Fm`|WJ*pV~zGaLNNuss5G0WR0U5mmP>z?{P zKv}XSOZ-aqjc@peC56G#I)lW^mW`I|I<_*&X*1Nt{0B**wVY>`#y)ojgQb6P_#tIU z^cCvT%!55wE-t?22UyZilX%&(JDaWfkCrgTHNV|Ol0<9y$Sk!Rb^8t$i<&EsD@&xG zU>Q_~t@&q>7p}*0uZh0{b$Q&BB+*)`xZ>OoGtb)+EI}`Oo>rDX zf5DOw#nxjj2U?c`i**`_mo4{9e9_|A=*u#&49X-)w3Y$PV(mE}yVNdamyJ*sx7~up zoS$@j#-?vHSWe_O`x7r)E`_mY+Mb$MC!Wt6gflq_9r z*nQv5>D6tpl(8i7vL$UHoBIV9*_Q)Lo5m!G))K}n%Z=u31k1>VK{u5pZjVrxW#8Dd z*!*qNr-5aED-thTHjZZ3Wb53k@g&M>9Z3?cvBMcPB}-mR`)#A!&jMSh8{qUn-09KEZOlJbOPDxM>NF zvGRrm|HR9dvs?G#b?m%{2mZoAY%!8VYjI|l<5k)jgC%x!Qi`%9OO~duSx3Z!ulewl z^S*W@UbgrR+mDu4C&%G4vcVjZL~A+DEXNiWdJUEn-a#LfC3?S5mn?pY)4)B=Kf>Hk z+)d(T%k{dfE{?%w{xDxI9wbS$mSkpWbnsRmu#_q`|EsbD1qv1~N4BvlPSC8#1vqP3W;#kp_gF&}?rHHo)jvG5AJk=96_P}2Y0oT)p#!&rdKNLS>|45r=NI+?^q8JNhRqF7iHkD08VeVd7!ccoDeM6U$FcBqTw+?F!u|zC-Jhy&Weq( z`KwOio-60BB#G8?msz&maV!OOu?ZM9LRnlx1k2*xY&}-c{~Df3)BP}smn|bVu=~FI zp0l_Y?nMYmqP65*k8|JPm>t$-Nr@WclqF5F3`=MG`8yQYQV;5~*o?%>mOU=)z2#?< z!D%qY@>C&7w3a5!(!AR^e5bVg-uo%a5_?jpOYa$M4~a*XcU!PTC6IX8aw?8}OIi?> z6A6|!$s~!^GMQQ4T2^w0F;@KD$yv%05-M2k9%6goYnz4T2Fro3Bwn`EEXVdLY@HTo z1(q5)B#G9tomr-(-of64JDWByP!_LKf+agId#-%gpqvRjY5DAze*|==cT6CeG~MZ^E*!+} zbX(K*w$Ppa@)gPQpSshNvRQY!`{;WK(4F2douvDx?(}Zg*%kZtk!mi`ojx^_r20Q~ zr#A`fvJ-v>zDPDnsqaqj=~4mz?;~s*;&01EwB7TkifTmttvmg_59>~UHRShL=uZD) zLz4WD-RVi6N^$MI2DJ1nLUUk_=oQFUJG_%Cd*nYy`ek3 z_(+oJzwJ){+aKsvD%lab)14=fwEwv~-R8t9>`u?utdI$Gr)N$j3ID6^^!hfvaA55G zzWWqh@e0o+Iki`bbas^}T(rqP=uW@Cn56mNy3>=hj^NJ7#Rf%Is4GskboKP;z;;Ny z3Yw3r^Q{F*ynOXs;>X^|Z&~1jXL&9uL6T_yUvk+0%c)Q2arf=*?hn=|OJbOCCzw5$ zy)E5$C@B%{1oZ}zc-iu)@Bv)W&OiJef3GiW1WBT`)Zc(Br=i2PzJWV*7xz8x$`W~6 zuowihl~Y0)uMn_|@+R@J<;6Hw7r*TRrC_{1^dm{Mma)u|tI>c|uq=+6zD-#I&j^;g z-`SJKK$8u@U^#W0#LJdecaC9QrU#C%0+yZ+NfNEalUaOXpE!Uex_y&f%Hnobu*BQ2 z^J3o@FWLtzH@cH}*)skH`{Fwxpm+{gKJ+I^w3aAl8J&6n&+m8GnRli*>(OR;YC4SFUH>ivC&6q>V5*IGirIr!9jwOC@ z!A_NS4@kUhSu}*rmp9AuEdoo)XC#T%QhOuL{h}xK;tq+n^|u^Xmaubz<%czUlVw*p z7j{86Y);~3%l75$OUd+tCO^UQvn@%YwTxgE`$GrttIxUeV@@lJ&w0V(@rpgo#?Enn z0rRC-4H7R~)i3xZ{5L-quC`9bnM zup~Sr@v^00Bzw!ywa)j#V99z;l4vav%u@D_@ounm`kWJ`EXk6^b~Jm7J;iAl?(h!G zCh@Z6U3vgslS9H=;}hVjyubha5n4+Ivy`4X2;aze*%WhAS)wC^x)gfOo+ZlPYx)N2 z;;@~>%NAcd_5_$S#thFS{^d`SXf4&;aqj1o_#FUsxiG;qPFaF33YPh?Z0>hUtKJkW z1!t0Y+43xdy^(KxqWwJ>WBnJ9BwEWLW*Hrq3x7j(;MCzyl*Rp$U}^r6)g`#^(Mw=? zIhVxC7VEa`I@WH>>l0vk??94hEw0Ss`=vdeu+w!#jhD)jDOuj#VE4+jr#;Ye%ZDX1%qu|d1K|56s0D8XXakIntr7IjL4#UPNx%NEnB z>}$Xi~CoSL~D7?EaMgqI1Y7r_&uzIvIIv97T+kg<}Y$+0PZS?$aDHnyliQ{lie$? z&;OJQEOQHyBwCB92hRQN*Q?lqrT_6|rpn@ZO|V!jU|$v;AJ)d*9~+b|Fc$76)dj(4;u7`Nym^si7m{CcaZL-a?XS zE!NEPdjIzVFvhwZYCBF@67L9g$zH?W^^Gi(c@!+ECrP|)N!rGqU&f5gnF{m8{TxZ6 zwaj9cgLRB>?$;P-Xs0ZZcLhtL8@unP7JV}ZET?Qqylj~^j=i@`n^ig(EDI)*BwEXE zW~mmMfV(_jjfd))L1o z_14_|24iedtDOs!#Vt;-EOKCD?DhG8fl!y2P9$Eoygd|z*Rc*^Yb!%toO+TZT1);d zIQQp2bv_7|=S?O$D$7U75E}9S340(|_RVKIiozOB_5n-N-XvbObaG+uF5bCRJ_(k}gGmyt#fMq0csx4- zmY+daJe0-fpOHO0BQG0xS(eNxW=%9(NSyi{FF1jlgo?97&?J++vnnwud@_ ziI*)I_t>}BCVBjtKwS*_lO$TpcxGAlW$t?zW9Kam zjwnmeW5E*ji+$hkR?g8KELlxRylgSt&vr#yfBhW}mdVy6iPo}}S!%88f+y=WiM|@5 zEbdPP%i3S;xpJ=M`0rqexkKV*%fWfk&z!wL~*Z{#Pb5pe`w;JkBah zreyK_&EAha8PUuMEO#rAc-d0z|8aKbVLf%-{gLvLDPM3`RIk%>t zL>89_kVI-JCzclGoh*^1N3TT+{doBPs$K%|yk%n_vNnDy-vWQH zY<(Rhky^TW;j!QG;+N;hqU_P@y|9G8Vl3(DWWQzjVRhUEJ>vw3=Pe2@qeDaZ@+4Txxso*S=vdDZa>q93GWGUPT;(1HdVp5l%BM!_$7VB*w ziPRE9EZdHc#;2cr9kc(Juw=bv>f$$zSk#x^?0_sc3_v_@xj3DS{kC3H@UNR}nhBCf zEgy+xz@8!Y=o(9ppI+Dazu%E#^BK#AkJK_(xh1mHN&)e_rA&`}ubg3jBo@_0=?O?8 zwWuD&V}I59)+xvmu~Vgyu=p1+mZ>$#3DjG>f)68$<$e&)TV|~z>-^bma^54$)WaZ& z)G~uu$~9LhBFj&ow=IOlv5>Ji^&zJ@Y%)sCM3(3P5YJmw_mVlxt4}@#$l?|Tl1MG~ z#1f$wjK7O*ZV{m(EXABfsXzH9UayCF8M55!0OEN|dL&s-b|3$$9vk#(K=3X9iU#!}IpjQuHnGL|8Wr!|P@Ee-aPJtV~sUW6da3_FlSYMDeVdJ85L zAj{C=qCvt^%~?9{A#<2J6Z>yMmd#&4Ja17CBlBeM<2_d)OPwDeiPW;4SmHLkdW|d^ z`!DMVOYS?SE=N4bJXyh_89rIBpF4=>Ewv5E*iXGXZ!NN99Ro?EmT+P@@XP2Ovdk#i zF;-aO-!qofjpY5sZLIGPR2Rbv5YJm?A0^*k7A{KYi7Xm5?0)|dQcE7Os7+N+LYBh1 z6DJBwU@>FyT1(cGe+}+98d=sg1@XM)rZag~u329A1X)y+K@zE@t0`)f-HxQgCtVRcw&iMmXA*iSvm04 z9APQtEE{i1+kHI#E>^Za$VgZ+ zJ~DM_WKZg{d!9L7gB@fG;(5#cMPyCM*;ffa_bYdSBvMN-v8Z3a5`wO=8l5(+6qcwG z#!|S8JS!J0&~-$08QKHH^OlSD($&BDG`@OWO-ab&)0F#3(ai@hN32 zS=r>h^2L~?W01u+5XAGAmG#Kjf7)R6Q)EdC2T7!s#y)uL>)xA%d*6({)Uy&6yHAW| ztqNJURP63`5?NOM1o6COP5}A-a$wZj{>aj=ZvEeXgw&!-ENcq)3`PyjyAEkv3(Nlh zSpK5ZJ|=55zN7iwO2nP^^^So|l1}?t3s2xq`%vSe4XD%ph9AiCcb)dBOGn~A>e7Ic zuc*_0Z7@jpSDp4V$C0Y7DNS=ho%XR8K&t<#(|$sBhkmHj{%ABvDetuJznOH}Z`RXE zL|4?qSdd6uQGe>R_xV(W|F429M*TpY_WpN4l7HE0-)qCm{;1PlEd?a|bEkdTr*pVJ z{@jA>N$A2{mjM#|MW_9z0bw{({Q>h2p-y|(=OEKR?X>^XA9(cDybRQ7ex27u?-f z2VG;|7lI^GOa0?`wo$8_B0e>sRXdlX!cxvz6s{0U*s2HnkY!635YJmO)5$e<=Zx6~ zWT~qOl1MGXh~-0P|2C*D_9qN|g(dR~Qprp^Ed}wsMQb`)F^Tfk9EL196(EV! zqHqF_{lLuJtH{#3$nKi36mb^yB(f@(w&yn9v3m0pi03U!W|N&Ons>EzkR`MdB#~MM z5=$HV&Um%-;9Y~;!je?U)MaLCGWIX#EHOfL*_jIBd5fzHS;e+I`&kcJ^fN&csbv+h zc-HQ-7}dpWi~2ob39n)-XU37Q*XhGu@b24oyFol}NsA=+L z^2L$J5>r_EP*~i*GZw8#J3PgF6Kr}E)#YXhi03V-Z^+q%+xmZQfGoD(KoY6tKCxI$ zynxRoHLtw-L|7^~OKcx9_G>Ce%tn^O3fF$e^OnkQWDZl+ar=H`X`%#@NG;XGGDC9_ zJ~iO}0EcX0$*N}RqO48EzRShn707Z#2gLK1g>}ge36;7d?2%=>9!Mg!^f`&|V;P$U z3_;geO^50E!V>#~v1INcohn<~4UR#USuP-+x0qy+Isc&s&-uAZM$q%vyne?;^AsB#~O&h-E=ZGki}@ z8~N&!usHr=EGHh56-KpBieAXlZUTtsEwvVrd2(2<5BRrjQ>KC>QcFCstR8%95~|Cm z%#aFUDdsE>T*(~9YGNb&yKw4xAfC5$S0cMThO9k+=gDqwK@zFuE3p_Kb52H<2|=5F z2}`P79St?5PD-`S)>&DNQL10Fc@6d2-;7D;u#K(~-cjTh0pfYfM!LiM%ibSWXzY); z3X(`IJ$&)l-!hJj{Z)HM);0O>ZjZ>Fj3wy|xyH&9wD+THY?LdA=Pli1$#Z{gcE7L4 zqUr^bNGt$o_=Fv; zrXY#b;!G_0d-vZ**I20is}{mi%~^gnfJ?{gLvMecaeO-DrZkD751LwIG`hwVZ;$LDEP)P;#X^x-6wjT+zp3*00f^@<*9Ma>SlXi(;jzEr5lAAn z%p#WE3N!Hzyg3nd`wENm9>$WMPj(e(O^a-V#{T?kAfC5;eMRQ{+vBIEqHFB>ZIDE2 z*+VRM3>Vv=y4dYZA0#ZLoMl=n887+f^|X=Yf)j}6Eu&q@b3gp!L%iGLq#H;gwL}q1 zc%QF>kfmdww~nx6>}Bfmvyfb4GnMz(LzX2MKs;}GyOWIl&L1N1sR5s&KoY6t9kDbF zd$9yr8oXaRR#>7O8Ozv~WDRym)IfZG|LiXyp0_wTlWVMKL@Nh$jjj9vl1MG>{qWc? z=`b}DSIveqOWr$rB{(YFy*FZdP**}`RR}QQdG8S2S#epPJ%M@ZU z>!^^0EUnLcnI3(lXRz-w~x7uuCcT*5YJmybRpx#`1<5bWLbV0B#~O|h^23V zb9ZFf*zw^UVaew#uFJ@Ltl1#%5y;Y_2Z-k_Bi4~??8CC3?~x_9FGwP_L=ej>%l5Iz zvNG_vk+3B0XXkfBAd?#PgP~Tgh5qf!PZDtTc`WNu-v1V$m9+9gQpv z#v88`mJnygVl$TfcKTkQ!KYAN+^2$g-eRv#=9c;?Q)-~P+?xZENG+{T$a($U?aAafM5Ofm#Xq?Qn3X@6z2DzX@txY!9x z%t6M|K#hDCi#bs4j4Z{KAfC5)p9;er-Cn0(EP@AfC5aG$OyHR-q920a@nW1xciq4aBmfPHX%d-6eTfjtNVW z8&j8II>e&nGH5fZOV45u&s+MtkR5m-C0*JgOJ*5JBDDk%OLS(@WmK2Q`!0UM5`LJm zob5=)zP)2Rd~$2%BM{G9?kAG($l2xr_ymMUFF+EhC6icmlgE8Qmhm35gM`K1ow3-) z6N_@2Mqf}}JWhak-jdjdtS2iOR^Z?2JRAU$NG(cd@Yq+qKI#dw+zIUxE-V$CrRx(i z_Q!oVo`5W2mq9#lsV*kpU*fcD;C24ZH$W1pMTb~^?q8OHEcHf}To#rr52h|hvdBES zG}GxBve-=k@w{cp8?rw3ykT<#G+u0`fh1CkIk6n<+if$lL=TO=E-bO0jAg-h^6fQB zcUd!JIW!i;^OnZX$(%nwR|)?rfQ|u3BDMGuOWSCh_sDX+=AL+A@jt>?yyug;^ovc# z>o8SHAfC5u(k9PJ#a20ZCvj*CkVI-pCzj~_@rlS%*=OnlVR7_gEbcm_E?2cHFQU3A z4gvAJ#eF=ntgq=>g03;AQ6P!bQab>T{oqdr8zIYc*EZ?GQp{QQd?st-HhC-Y@4}gS zgLvNJxPZJPFZ zyyZy%d9RG!b}JZNV=L4_5~;fF&pqcrF%&rp0^C? zMDEE22?_Z0`2!gsiPW-;SggZ6b&+LSn~0jL|NHGV@C0MYpGUs&>6K=VLv`7s3gUUo z$4g}YSl7C<4x+j==>d{RE$+lpuk%-Y^3mzdHuZ(Y`6Oexr%2|>SE_t(Bg+Rn5YJo8 zzLT|!r$fU|BTJ)wAc@qHKr9~=V#cGo9N#sjiLjJ%7Ckp|AN%?E=Urq8+z;Y;OT(Gu z8hbE70iP5$>@Y|owNwzx-a13-Bg^4COyFbWj+SP;3{(&qD9)ft@;_6T8 zGNEb|o{yP61xch9jX*s1t-CvWBFm)#neBxo>J(#9&LjKB6x?2BB1`>|AfC4@`byUN zjOOh3Lf6>EaUhA*vXEH5wA?rkSsc7hb`};NKgMFPom^utO^4xM(0O_a#Pb%tALM=i zVAFA%kVPXHB#~NNh{fH13SPS?8*Zv0EOw_E%l)(D`7*iSSY1??pL!skx2*hw)a7b} zQI5#6V-iRrwZsyOcHJ*{f3kh^(fx!apR<_kBWr!?!>Y=Wxgj zdUpm%q?XUbVmL-0ciOioQ5qsFiT+GoE}N3~{ci<3YamPW3J}j*_Uf+V~#PgP{i{u(p{5iW8y2h4i zf+SLlBe8U!Xz>EoWs&ykNy1XjS>8C3@nYCqA75iVM}c_W(xaTbzjXetkAK%Xb38~Q zwOl2Zu@f_fBg+u`K{JIV^DI-BnB(O6qUJRM?}cl10L1f_C=)VI?y*Zl8P(;iJ4hn6 z6cfu&_jJ5h!PWQ2JYk6mWGotY$@_lS4*T)5G9?Ja^OpAw!}0f**vjMmG6`BuQYX0N?0N+zCnzod3|ypvoybV0gac^79gIt^q)iC#Z1=Jt3Y+x-VP*@ zTBZ?8oiPvH&^2b<HNSyOlJ4{yBfM~T`uyJi zd4kpWzkkzLZd$7N<`>BFcirhJw@G(;quuR$qVDu<^~!$p{;E4Y*ONSbimDuIpzd_* z#vs-I)SbTOQEn#cPJh%2q?C83A5rUw|L={Q3(+{st#8KXmyv z>Q3+31tj^G-RZ8e$s?l4%<%8PcOL)}{6%+q@~7)KQ_Z}_ z2BpLVDJ=?}b@Fjf=r|d3acbwcMtSt7NLfz>pvp|}E*PY(2q=yH(dtPp#Y$fhEMcm!< z+z~R{*mijNE_AJLZx7=6yJw;*nFS8g3B%`+59$Jv$o^kyM3VoP_xm*Pxuo`AUTqPU zq;t%Zpy>nhBzX8f9KUHq{{->8It6-XkrOeB_%n=`GE#c}6U z7h$R3EIm4r8Li3LQVUcUMQ0GtTgK;--M0nSUGZ$=^kI-hYS~IG;~L+`mPykOfp@3`B#~Oeh~=%XCf*f2XnVmiVTnD@ zSX9i&&dAwE)9~)7{JkKaw|pNT5^fy>-^LBw>l%|g!&1Ke>h{Q zt4iKx$NI+M-M8bHf_UDtJCuCAwt3}#AB~q#6Ocq|X&!>des0xlyz<+`%Qi?@94{~y z=Lux&uYNReCmJtWr$9V!x%ZH~0Y05@@C&k31cD?|iypCP?K_26e`YQk7cMNtoMn^_ zc~)*WZ;nsd&|d}Od5dy8@~kwvJ@N>$yj%;CNG+R)<@K72_$z5kn?_fJB{hPnOHEH= z@!GKspA|iQH;CsgN6wHH`JQoMzQ~f`1d>QCfy8pk`yL)IN7iLt7naD2j3qIOTw~>0 z+8O8?GqDEoyhY!LblaY|t8oiizSw~zQp+-rJv z0=IgCvJNu-vBp?K_1d%I>YdRA8TvVI^e)tqI#DY3LRU$_8S zx(x#Hyrpn08T-q0a`Dc{b0a_!sbwUw7`&N`&x?JbH9lQfaxXD;nPW$G6?9TvgFCvz zhkz`EjyRbxEWh`dT$;lMHeMWsk>TbIpHud-z;qEVI+d`^!GXaJ=#xITXb6 zmcvHmE9sA6I!lqI%NUSEYDpoM(zbgKqVY1O$fLfn*u^lGotw!$`N$WoNvJM8K7)AP zQsWWXeH(9+F%DT?R)Qo_OO5k*>^ta>#b11{Ew^kUEcu*e>jkn78{fVVd6FvL|%i-Oy*20o_ovBM~IO!rt(|d|faJyRu;(5!j zYVtKMQax)ks>`D9Ac@qnf>^4qn1`ZkEGMjX2Vn`h!C1EcBxe#|9{WoR)kP}`#Pb$! zy7NB8{zojbyu1mLNG+bkqQBL*8d(m%Oz$i#t~VLW4^=X^TsCI&5@abX0`a^>@dCLg zZ&|(P9kLuM1xciqL}EFwYWfUWqN{u~gr%IbdL^=`9{};ZrKKjBTW)=u zwg*|VJwOtvMKc`V$Homz$G=5z@9lsg!V+_fv8-Q8-jS^u72s#(gU=wIxBTc%uCWWw zz2lK(M-@mSwHOi0#qO{1c=3tup^e06GOd_ z<=Y&PL~6NBEQcny#Pj5i``jl9OA%+$t3$^A!3)D;k)@{_i03VpwA;2|@3u%}Y2Xc# zNG;{W61PTc3aU$r*`k@kk`&L>#q$A~Tk1Y&xDi>tZ3OYWCGt5r0pVD`KI@PrZW~A< zwRF3H$9_=rwU>~kxuf=cVF|y(SiF?U2?%e3&G2t@M;-$4yk%J}^4m~HpFetqEDlFO z5~*b_u{8D4u0WPEF}0Qoi+ciNIh;&(dqf|+*9O%^Jpsh?7WEQxPwv-dZbU**VzoTf^2qckO zVu)o>v-UHPMYr11Tv)R1GIa^+K<1X?mv4T7>TByWMNx-5_umH3$K4&xn}H-! z%SU3#tufF9S)3=W+$b!uiHv1d#wL7EzLjo@*ZEHxf_UEI(umZhLhX(d8ZV0%fh1Ck zY6KqpI}~@TBa7MMzFURG{~lurQz7q_?K+fpLv?v*0^)f~og}i>xA5m{ybg2A0wj@I zW)MqpzqE$Pk~q3zm#{eAXDoy3lR3=(k-7N2vT`Db=Pm0D$WE^@Cn7aaU0%!pNu(Bg zVu{cnh}V-%iWBw;OEG70yhgr@jS8$8i0V@FD~RVUmaEBqthx3!cVyZB10<1JE)z@o z$NP^_UEDeyaut@;2TWb^rjRwI4OcgwMV5dbAfC6B-}k_~3G3X8azhqdEs#WNDIykA zYxhcIF={;jh_FN^F_!kb$+L23+wHZGrF9z+&s)mckhQ+1%c81~C8{GxBDJVo#A9Ds zFChw9lGVGP5Eidw#&Xzz{1$ZI>8*PrOJ_|G&s*Z|ley)Vt`)d0?gKy)sbvzeyl8KS zyP#XD{`41?YR#+SaMUCx@?_I>f#%AUmsb_wLm;?+5Uz+D=(E;uSXX3As~s=5>6~ZhL4URi^h7l z2w{nT$XHf9B3;mS=T&u(#m@%B^A`6tblL@yNG*B9V(&M!99c>d7Dfw8 zU@BwjG>$x97Q5?gMV2Y4AfC6p3?<)RT90dqPjI`D36e-HEhF*R?|91A4q5j0?-DC4 z&S{KggBs}`^jxK&hAiR3K|F8qx=Nmvv;3#mM|FuG3zA4JzvFp}O<%H}?4ekOzlpa{0!gHnU}9MvHoXSA#!Q>;&JdQUM~vn5 zTyj!ac-MXT$nw4|i03WIeaYB&>-cO3s!L0CkVI<9B9`3gUGeE!k?zx82#Ze!W3j16 z?#VqWYX3r(%f~@HZ<(~0e19oz6@z~fr05JtBDFNWgvY*>ai}w@i*=+*p0LM7!V>b7v9$9d>oB!n zKW>REcT_<tiF; zrk_F
    !?vG|hlGC8HgIb_iY21%qAGh(r_Z;baT*-=eVfgo84U#|-spSN*v_2Sue?v9j*G5@bd|xn@C(THAx}oLhU}WhZ z4B~mq#TR65*{##&MyM_p5g>`wl1eNJ*L2A@i-fVAgvCCKv8beyZ?7*W`jsM!*+US| zTRugS=YHzy;b)M=_$f#twJ2P{V?U>Gel0X!+O25RO<0OJi*geAF7{A42-ju!S`g1$ zMktc~Fdi2RuOQ3w4IqisGLTr@8Z^Zxg{>X-TvJ$*vYEP+d?wGArVl*pqq@u)3*vdp zXJDTB#~OYh^3(4P5i73p0H(@u(-cuEDigS-h{?wR!>k}ET@Bb-co*# z%*UpldodeX`p*MNq?Y@{vNq%CQdF0QiQ`8LO9f|12_^3@H^1~>iY!CZK|F7%*PB?@ zsa?Qpk}&q| z8&sEuhd~mlrB4*TkBt~MDhXNgTRod1EU~W`%iMG8@%I;p<%{uNg@tM$p0`+yB(t(X%9D->Tb5+-W~y3W(<|#ox(uzhM`% z8OSniHb^41xDks^SsC7U(bCp@p|Ci|_C?aNisc)4v4;(5#PEOH+!R_M?M zS(ZA3BvMN}vGh0U*&cPLk5|_>7M6GavHV4My3Z~%e6_7DOTxdMK3fxHl60pJXng>8 zr`wkwsfoJNR}KVO{;oUywdpAQM;dqR?Nro%a9bx?Qu$8I3gKkZKc(;sNRuyqCMPM;TG@cR$^`|k9(o2PJh`p3f~ z$Dr=?sPiD|*&YTPag zB#~NrMB~{;;msBOQC(6Jx@{De$O6XVGJwnit)IP_g6eYRA&BQKAL@~(+8n=iE0IMf z6C{yZ<`au)cJ(!6$xbe@6&9~T#xni@c{2}PtY?TUjh}*e-tz4}*|lcc?Ihl{_96!) zky@OI#VjQRf4yGj@L-p)RCAVL7szVq{mOHAysUHq@w~;hIeD8s;-*uM>N4g(JN^Fu zKSFA`K`afo^}=19F7G_{i8Q%yn7S19C(q!glWOBPjZ@_yp0~VPKxTp6?p?sY1>LtA zB#~N5iDjXsBfiEKM=Wv`miQvZk~xCRXwNCT=b&qB^ED99ThuGb$rPcz8#F+5(Tf8~ zq!#t7c14*Qo zJ;W0J{f-y1Oj5q;FD#{;<;HdLHLh-r%Lh?i)~kVd-qLd?S;bcJoNR$C8+w8yQcD!E z+=($+h%AGy>^&ze8Sk089Lyy5u^s~sy+@Y4u^^te)ZRiY{SC)wB8x^MNFueoBbJ5} z=dVPTMc=1J2uoBkW6{eX(=yLHPBoFmrS`Pn@w~-!Ch6+Dd%N9HWJzrZl1MG>ui>#D zJTvg8%EjTRQ44~(VmVp5ltRo3`y!g{43p0_B5khhgLXne)$_UKg**?aKi1yd^4|>{{Cs+IAMQsAPd8QcFIu{G1=x7g?6L zPRkIMkW$7{YC^{T=UWx{w>$pxmGyT#Z;8z!&zD@w&HIr>{}f0fwX}}GWB;!8s$s~Y zXV&J0u(*C=EOlFw`It_}0eHNe+z8@%%OC?XA3N_GhQDBWZv#oBmI=f%&1pDZEiLfM zeI+dAoTWgCe7*idVaG`{UMf$4c;1rcNp=(d(SI)f5^r_<-S0m_YOx`f?VHZx-y(2~ zx>6)8nV*@ugx0dbOF6Z+>jWT+|1A*DTV6gT?=O8EHW#C7tp7cbL~02kma+yDTA^#K za)Monu*8%x7NvS*XXLs7^&7~dJ_E$_mfHu(JlXP`t_iYin+K9eEjh%ZJ;qxNS$twA zd=(boa>imYiOk1boV(vdmRt8gJa5rTBF{>_GR;=VG9?Wpky@Hw$74ThervqTGjCX{ zAHrh)g|Rd)B;(~mdT2vrX`%<>c}wC9GWH(~9=!}%9vOfnQp*@(8Ee%Xf5Fnao>R;0 zzq>t(I7?&_dA{rnu6+Pmb|rv#-qOI0yuYMBw#WBmD zB;U5h_vFP38`MUYxTheVx1_uz&zHfIqYRPd;7gE1YEim@$NtV% zkpo;izHft14QTTj#PgPky~rBug26NCq4tgHF%gZI^P@o$sipQ!Joa^q*5lt)c~Ve+w6GL&7LBc>(|+0P zZ*NguUVQ}dyye+y^8Mw|`k{Wv;{FvRky^BgMPFx)HLA}>_fE@kmdTo<5Pqs@+V{Iy?|U}K7Fqy zAa{1kaNK{k8-Ogz zmq0vkS^SiIvnX5QiBIOvi~&icmLy_%I_o9g?NR2cvrt&7Ig4*FSyO5~y)|CDsPYE! zyhX);yo*J4D>#Jea>EZKky?Hd%cY)OucNx0Y_GUnSaR*^YN#o7QmSpX&dTcleM;Z8 z+Y--VHtlYYPt*492I6_k*zfN6+w1-L$u*H>bRUpJYSD_t_pwX^mlMdM7?8GFSmJjw zmZm|Zb2#x$3O>ED<#`a#TYh~YYkl7;u6IS2xXU1k)Uu3NR@L>vyQ6};AGHvcz+H@` zrU|(x*KFnKh{nq!EfCLJmQ|4V$`9vPc1M=LAs~s=;!Z3<(G$O;y7={6wOLr4cQY2V z`((%J$-y=ckR@mUi03V>YLGdT} zY;a+CB(j8e0`a`XI+DC2|M+mA0$Ij)2T7!s3Szn4G;cquOVQ+S_QH~3&(tO2J2?Si ziQyc3WGR^q;(5!|+hiwkL0mY#Cnp+#BvOmUEj;!M^sVseT0?r@*)J?n4veLNDfzx% zYWl_&S%zzac;3>-kv#WJYgpj*MNfXdDW}^*~NG&eJq80fX@4R=3n0-`O?DjGiBX2TZj=1gK zjOvnm4aD=7BOS?iF^|{<*N|m<97rOy#1e~fzl9S~T^85t=PNAvoW*As`TlbFMl|k% zzR?WC^OlDi<>ky>UGOP+!?{`UG}hHIp-xH>VGt9ywhZ(AsSe~DZG z;(3eZNAj#(c2Cs_)n(mskVI;6B$j^l@8RDcvl~C>ny{2}mYe``PhR66rG~Dt+fzY2 zZ&7bZ)?scO4SSC)W9NV*Qp;6h={!!cB^vw1u^n#6?`hq&#{pYg~j&(V>y*Wz9mh#{$Mt;G@lIO zd5e((smtRNx}VTBwrdthBDG8-mLUh<|3GyyZDRLCSnORG%P#s}`T68&{Eoav8N~CJ z!Lx|v`kpNk{-H-*^MxhpAXAt1%gL!{zrru$la5`pKs;}GevG_} z#hcH=r=QHt2T7!sLShLrx`n?bjWzuAPFTWS8OxQ^WUs;$ix0TlcBm1E=Pij* z#z}mlZplHAL~5BxEH>8`;FFJ@40Ws!mI}^d^^`m-<6UgskR^32i03UcO33$@E5|MI z=_k1cAc@qnl~^JTmNYw@h0_=3}!z zrd1+KxCKZewd4}ZadX#=$YS%Ou%WQ{yE7KWzT|zsY=dD}WEocX&F^^L;@Xy2_Gov- z-;uvI0!gHn=6CSef2i4U60!uZJl{fC96cC|&sj2uiCrFyJMIRo1M$2?XFhosGt*y* z-;qyl0!gG6Jz}Xnya9ez4zJ#-A}qz6W%*UoJ-E5X1ia4Qq0YA7@x0~653+x3iIxlA zL(--ZNFudtB9@J3j;uy?(d$26O;}PrnY!42C*$Qtmf=fujYSOy@w}zZL~i7B4Tx684FV{qD7mJ0i>Oz962rl$;=Q{)>5;1Cgc1 zV30&=X_$b=e!XqFc+XXeMZ_Rsspc#u%gA`S)_4m(EvUB*i03U00?8aE-_>j;vMB8W zNu-vM#B!p0oyDjwW=b|X!jgNGsmtu`awOKi03WKN0B-Ii9yQ$ zkVI;+AQrc;`6bA5u#Ns$VTt!VM?=SB)yW`VNyQ+bB-mmJ7dov%hT4G(}cy@hp{|gOzP6pzUmyZXsd#F-tygw+>^U+TwsdolGGg}ky`5B z#bf`+xa_0IQrz5sj5}T80qI zxJ$}8sG)ggs=JY}96et5uR84uTdu`3=EE1Y`=j51*LemqNjmKts_n&{_LseKf1pnL zZm&R=zw5NGx=1?hTMlaZ9(CH+e+$z6Rj2*0Y%*J0F=@9q>a_1s3R3+~o%ZY9-)5su z`^FU@rM%OAW_MNme{V7M?Nf9`t^NfPi7V<)o%Yt|ML5Zv+HH=bPWx8tJ$_#b|FYA5 zW2@MKsM9{%8YKI3r~Sw1=eR%KwOPnXbYZUB0TTSh@4!D>m5no%drq@Qo%WXp_A4G+eh_CFVm+KW2v%MOEt|ESYG_vRs7)AR@0con_H zF_2Ssm#~{e?h;j#Iy6I__Fktznt#`6|1~WC7rJ{MDm}kU+;KAgbI1A9ZjaLI)wC0Np(s)(rpsN67wPYOEpuF9&B2!RN%++itU3SfWla zmU$M$a;)r>H?mB31o6COcVAb$bH%|_c?q)2as^4G79V1{@xc;z58m+o!$MemPBNC8 zjmQ&P?ZR5Tk`)pJ;(5zB9rA>Z@f?89!E(C*l1MEn#40p7WCv`@3m!eZykSjy{= zH;pS>-PWSI+#3SodCQ}wWVJM9id{QYmu|5I@F}m~Dk` z&65W$%s`fpA3;2Cv5X*f@y%I}&-8lm6(o^bJc*_7%rI|c$;ue#CM>R}8OxCe=q(2!08(GS4fh1Ck5wV2iT*hZcuC%la6&BwB#-h2C?2No|Ne}N> z^|J->yk(mg4O!Z0f+SMQZDQ#>!wSzWdmpWNO<0OJi;g3?k2Ox#twPsWXFm|n zTSmPm&liiWZ6+bhv|x}#YAGj{$;q$rZo;!k*|&rxDUhkla!1l_t3A3BpMH|{7{v3I zhx5rfi*a5dc;)v+Hb^41bi0qoez0Q1X>=c3pBR`ZEa5?nW!+D5F6qI`eaE7@c&CAQ z-r{nFeD!&B=AkmG%erSEiPSQeSYkc8c0!gC)eR4Y#XXp@jPWJo<>s4g}AK@zFuBe8T$TZ+Gw=%0M}T3BL37|ZB|E8;$_M|HW;62$YC3+>5xd0zG9 zCc4JfcLYhK7JFjJ3iH=R&RF!ilKa>Ot$q8E#e6!5=Pe^A63ek5)euyd z{_{W*sYN9TkNrBso8=-)z}BM;g~jUvW2rrq>{xAF6@$Ml+U^1IyhXo)d?{&kwLAVw zI?4qkky<7ZOWeZ)_%~F2s#diSmTJxt`jxEp1tl!jMsi4ovKuYCH^8~aWN-L<^;Zus84ug2!l3SDPyf^a*--;~l zj)5dnOCGU2jeZ`4Ec;*G=`JjRk&Gqx4S81fan;#@>XP;j#PgP?2gw?2JEKhe+v!6; zgCtT*%Va$EGp#n}{qW79gItScMWxg=gViRF`CHkVI<9B9`TM z+T&-X*3--S!r~LfSUkUw?=SDaRL)11!MPxwx2%XG&;2_=hW5zPp$H_AS{kR|v0wj{ zIer&wx@q?`VX=#5EM{%UJb7kx06xVjcn65*EkTdTH@-)O&2Sfi&t8y3YSAT@tWgqGxi?;_zBDK_ih{t~OFU#<6s9FYW zw-lCg&hjCS%q=@+RpD+ZxAP#Lx6H91_vE%8Iv68Miz^_B)G~}%rbM;dit4hg!eEQA zWL{_L62F*?m+Z!IDyS|7H7EU!=Phshkndu>cbNV}miCGuiPU07EbD81`-UvH*R-?~ zmY5riWx^zKPj1-zdNs1#u>$eD<-849hiP%Q-hE`*X$z7_EhmWO#`R>po3PqEXOFP> z-efFGCzJ1Dy_dJarv**52JyV5o;I0V=B>W<8CmMufh1B(DzPlPt%u)V*1Lop5ElDb z#`3v28T)=KUE86$%sB$$dCMpI`^iIfd)z`6tCJv!)S{4z$9|m=tp}oOOzrIkcVQ{w zEOQspx#eDko5*tfGl=IcVTokj^3~&BQOMG(3M7$Q1`^AXrOojk600(uW5SYji>b@i z0(yd5c?n+YJ6sLoc}vh8aw65(t%+vH(zf=9-+zSEvWi%&oc~yh>e5ubt)H-j-)1bw z9LTeBymvDG6^)_2Ks;|rTut6zB94a4K$e95Ac@rCMJ%gNjroi$Yhv>Pg~dINvFN6g z=gV!|$;;6-HtZgV=PgI;ld)g^trB;4d`$&Oq?Y@{viL-v7P5?45PV)(DmY8A8@a~f zgKFWrD6IhTyyc`V885p+2c@C93^oHvq?T%8nWXb$B(kg+Zhcu;vf`P#I5i>P#7%6R z^if^*>4A9OGT|awAB%2~iMybiPXbA#mOg3tKK3DS$UbEGRd@7tVTrxNSbBSq-?+P$ zdkUW#pwJY=^A>|Thw*-x8tsw{P+giSgCtVR5@KogY7(x?f`y9l!s4I6SjL?o-z!UA z*1^BksksWo^A@KUWS;CkS_{vU6V`$xQi~h0y#9XYIjW0S+b8#h#qlm6VjMX9`PdB2$<4^!LV3uHXF-js3B6 zKs;}$d!5WJhhEx=zr7Y1fh1B(k90itE$&pULv?vpo-Ith@oQ!G0Um=^nBq zEd%kqWzH31sSG%WPYu|+1|*SMoQY+ByW&b@`4#cBSXin#i<%L6e=)RH$wQX>7!c1} zRIZbEv6|Ob;!|jM#e*bL%MD^F$=QteTsf$pDHE352TWZQ9LQQE)>HIT&k!9{(kVI-xe}u>W=iH$-=o&lIO1tK||9*Q7OlB;GO5}b2!H#D`(KU9b zBZ%iMXDZ3u^48GW{>T#04J46TW)VwLt}EUL8>~^gzOXo_FqWcLWN!I-+qr(ok}v?o z^OiY!WDc{pXZuiO$r}!mNG*GaMYDd!15_7vm6XQ9Qp#Dn+K})2Jv&uhLYAqWK|F8C zd`6y?{cmhbMV7j~KoY4XidbshdUym`x@`AoB`g^anYs)cO!QWpR9t26GmUqPRYxYE6WZCg`S$koLN@XmgmB~J(yg{vNp}IV;vHW*D zZ)w(s)J5S=cz;xvnED`z)Y3i!kNx+J4xUGrACojX3yV)0W6|hIp8Gx3SK&QZf6NB) zyybl&d4GvXNLzs{szxA*)G~!w2A;lvyP!2{eeNkNcIk{II*Y829nsr~f1zz%LlDnf z8Z9RCG3}UwAY@T$36e-HcEn;*_66@7Pdl8XB`o=z<>^E63kN>>>nzbVw(dBH=Pkpp zljlopbYI*#-0KWTBDF*ii_$i;_2?RVR>ys?up~ZW>f*bT)WzlfEc|N?1&u*GZ>cqt zTw`q?UYQ`8qS83|R)6EF38;AsLLN^*(Y>ZnV374ywzj^&pUAP~=6mOUctmW$U` z<9&Tm;UJ0BVnZz3Tot@f6Y$)cw+)5m-jl!VPWQP@y3-fG?{@`tr#mQtOp@;Ob=f;{ zclyp>MNz0beRFe=GQNl`k?OgOKm~Azv@o!{Dq9BTAEEyqwe%~svy<> z)SW)d)iWG*r!VdSQp&s2=M;CsGnTdk4&mKmLH$7@aYg;9JAG-?TbyL%gOL%aJH2c; zNb)bc(|d*0D@NVv{lP>3}ZGxB4K#Uv#JIseZtj@*=H+P z+3oSgd&72g_e{D7;`zJh_1&a5%OTVe!pm zEQw|0%i_nrE1ILa1bc&c-m-Kv*;Npf5Zwn^R{4P>QcKe(c(&0`x z##jQbkyW_>r#n8#qS6M$^OhHOq^ooAp&|JE{*aC!iPSQNSe87mjzAWjTTLv5rHHe1 zr(HLJH^1YqoACP}p0`+^Az!c@(#r0jy6j5_Nu-ty#4_KUh*rugeZ+!GMbTTFM7)t?`+Q@bI{fE|~Q9bIGFbU-|Bc^*qvf6fM7o{cPT^*|D-C6ib#`S(49EKYZA_6Uo67GvqQm#mm% z+^bWGEU~E|p117CA@}4*W0$@{7UxWmL~2obipTzA-G;r9MZfWw1Hw|lS^7ODJ64NV zW{g3WX>C9}ZyEN0e8KAWfSwoI5hRgXbcjXsl64BQ+;5}gE-YEuOkKVVA#>J{5{*yD zVxSJI0GC@3V z8Ej3y`W&n>4nTF8kPDJXEwwZ8*gsLx;1aU*57jy^EXAC~^c;CJU-VuL|Gwqy%OIY& z)NfB#OltO?hTMcYSAW^Al2IVOyVQj-!BPE>MN!$hU>|FZ28=?xH~;|6o}_7 zeS49*>{+yEI;zWzKR^OpgHH{$K2dUyw9dD{TQ^OjTY z1;XO|hOsQ1NPb7L)L;qzZQCZcAfC5$ z>rU>;Md2&(HTG#2NFuc)5Q}EC^A2PgZf;mCETx=f-gfe=+%Pq}3#v=+%^;q)e1AdS z#f(NRd5x|y;~gN0)KWn#Mhh0FAxra99m<3yqll?X{ZrfVH8#y^Dn9)rBpAf=7XJ?9 z`En~@VIySmivUTa7LDh4?2lD6?t?55Pu^4tOVnG&;vGhQhvdi6Pb>s%1eTSnC< zV}IMt3B!?PTMz~%%QYx%ut98>&lJh1rdTC7-jD?kD%;@W38lk)=rk5YJnhHX+Y_ z)29Y=P+cas07;~l&%_ek{y+$_baqg0B`k@>OkL_|kW)71c2mdQP)`nkc;2$WlRWpU zua~Sr7FAD>L~7~s0+0O#f#*^;(5!JH{|q_ zQRXe$Bg_0KkVI;kO)T@i#m+@_X=isqU07T{GM0d3YkW`MeYq>1k9q9_@w{cc8hKXg z-Bz89EE<4}64mACF%ZvN;x?bhJBoH~ zJL`xnuBSl~spTrMv}k66cM=y`kJl2G%o3(9ZS2ULzwP87KanM4Cy3`Qx&6sHfA_^> z@cVvWCy+#HDJGV(?kBFGx_sH)c(Aa%H78ds}jHNIAE*4gBW(l%v3<2@H#m9$yuiT)jjk^b%M1mwz z%S2*%^Vu7BsA_5kt`e3C&hj{pyuYlq23Q`Yqq=nV2l2e+We~Y1XD#}U$9}gEkVI<9B^Kp8 zm26a(J45T(3X6XQW6{`1EG-5OGDDWHdLW*+)PGK{u@e~sPoTP(PX`$sO zAsJbUTc_<57RO4)Qfo80#yq~&9)>KpI)QlJ^7TEr#ytBx#V3S%cLzzN7CmAa71G%n zS@vchbrhCj&JuHd1)lRa(T;eDEZYh}Ja0*v>VoIVZ#p^R_2h{kK@zED6R{LszW5bc z)E=xjC@iT}OkFGs$+L3e*(_gVX_EosdCQThWImQ%WrElGzPtcQq?SNpX+C*x6tc`- zs_7{#k>451*EF&Y(fOy_AcOJRM(szyNipI+*OOQlrc}^^YJ|4y2_;UZK zbzE4ysu_#@CURf4@N7(d{-DB#~MgzQkj{U4qXLRF}o6 zDW`>{nzJkzLawo$R=qo+Yi!3e5YJn5hm-r5dhs`WdShuWNFue2B$i%*%6ROD%=Qcx zmfRmqT^@WV&wZbEz1kwnl_emax3oG+-uIPL3h-Vyz11Lz)M7y_Rs*dTqq>w0F}ffu z@jn?$rV1G^`V-qEAj^U=AfC7QJR*A)OvBCbiBy&PAc@rCM=ZD2ZupKYS10$35|+SU zjAh#la^hk0s5{S)rQsY9&s#QhB_|)nH6H!|jr{{gAc@qHK`cM2e&X+9D=k0W5Ef^< zdVk(K+B4AtKVSNdsKEQj;%z}ZZ^>38YZu<7^XDUr{%(*&YN?xx$Nsj1BRo-ET5Y&` zM_5WZ%iRL9-%@{fFMm{*4ckCGZ+T!#=3@^>)EbK{H61__sbvVUq&MwTiY!eY?@1Dt zjGatf#)pzq13HDqc_B;R91zc2#>JEO$^q9-<4%>l0+2*%Swk$ZRJO(;OZAT_kAx*^ z7h~zo6=a7u98S6Ocq|@gbG~g;gHN^3}HOb7Aq> z%~^@w~-)3c1D>-l~azr73bANFufLCzc$~p?KeTRN}6W!jkB~)TMItBu)D{s4m{`r+g8XkUfmWdOYb+ z)wNZ?>y~HAK|F71(2T4pX_rJUK$hrgkVI!s5D@vH01N=S%S9 zVb!QEmy|&~Z@D*;%*S^7^!kP@3spf9sU?wEZY_wzzi#qs^h<^H|NV|!&RII>lhYhl zwQPjP%Q^!P&s!$`LH5AksTi^YS;}XCBvMNiv1q4Fnvbrr9K#SrVaar4>XNHY_A0Ci zS%ZJotjY+)^Ol=lWIeg=x;s~p$mJAHn05YJmq8IgCfS;P0^IZR4hkVI-RA{PC*QAxNip*n4Z#n*|kOtB*GVl#DD z^+w~R-W3qfTjI6Jd@M8BFB;Y5#to1}YH=l&Ug5J_BTGtk{f@$7zn`&~TaxdUA8&qB zL>A?`iofG|OYK=?J~sGay}QVAr4dLXwcI9_Gb=3diMsI~X~%*iPTa~EM-F?aKBh`r=z`uCCQno%Z@(ep4{PWZ+vP% zK@SkmTh^^4_pySF{`ROYMp__=)Y9!W9{Wn0I^thxavHIzzp#WKU@X_u$@~77Cs*;D z|HB_3p0~83&zHM-yH_Af$~2HfYMDzc%?9u9fyT?b)Dc64#odLm1SXNbq(#{&uh2D? z{|vma6|){;#3ga_MUGzDt)c-)OmX z`Ia$C^=me-pa^cD9Hf+Y z+ABGcPW#q9@8UP&p<_WJaYg;9)85MH9sa*c8+bGeb=qr91WEp7r+wer9^t6de)n{c z?9ZL{uisYU{&-y%?;vzxE}RPz{6(kz@@LgJQ>TyFS5c?^Oe2u#pLW{+=?^qso45gW z+ILz7(*Av?eXpxmai@LXclV-Er+ttaNcfLB?X`CV;+npgA2|Wty-Tb>PU&6Z2f0gF zt7zn-PWufuAkDw)v~PN86Yf;G)+BzqxZ}jSGI!5HZ}QgUm^UR4-90OmK|Fu=w9O)4 zeFBCQ#-RU~^{OC=^#4UQAI~;UG~D2do&+it&U1yu{}5v_U$YXIXW&2w{5vG8YaaL= z&s!YnS!?az>f^JbKi3CIq?Q@PQgF8yKGQ4W#=OPC;^@X$`qn11jg)s|uA{oRB!YO} z(qs{t(RSH37I&v_NCiox7JFjZIWPzBSj`!tzDih%IZN%0WXEdy$4K0%@_96f=PmiO z$rr51<66e(8uQQxNu-v`#1c2vwJEyBY|P)82}|l>rY_xv*LB~2@7MbDwaz~K{Px*v zpLX}aRm7_d#c2v2w9dk07;}417f+ZJZdkh z%lcI&JB6j1v*foVWB*dEdidLe(VIX#Z^_w7o@h7DEX6y$)@%n!q?YZ(qUjLnh%627 zbaoV$f+I{_)-@r!a8Be;!e^}NW`TI#Qo4rx(ok1fJ{HwwT0Tf3wL}ujTjz{j$TIC| zk&Cb-doh+}Gs!b+kC=3AG+z3?1@XM4-Wl=>Q2S&aJ~!@3DM%uFkx@9{TPHoAZ$QcH_Bcd5J6o3&x)i zmNL$wt4nrxr>OcgL6#=PAfC7MNg!_*)-IiGjVvF^KoY6NhFH!HK060lo@{OwDlA#Y zn7Sl2B-dEo?V6jBMP(<5=Pk7wk)8ME!&Unr%XvqTL~02omZ5d);Q3|okmr%Y66?cQ z`Y0d4&z8Z5zS<(oiIE_lw`6T5V}IbWTdR=e!UT{+YRM;-xDUH0pt@`*4Y?#NzP^m5 zvIlt{GuwJ`GP1aR1M$3NULcuY8d}V4fGjP4gCtUm@>@LiU+U`M)zaC@>#hll{c*;! zK$FasXZ6}9A5a|t=_k)Rf+SMQNMhNhAHEsYrDW;Uo5E7aS$+;9 z--ppPUSfhQeb<3_-m=@D?7}hM`xCFoH{1e}NGP!Sn|h_ zQ)!}0PT=n+XG{U{yrt7Y@|N`J-a59(;$;GoNG;jKGBJA--h~s>J{O!|K#Q^4Rnpwy|}qRSjsue-kYS` zwngk-yz+bb8i?mDHhszb5^yjt5!GeLEs#WN8AdGCw_Z3P%ZT42-wI1kAX69Pu4Jxs z*X!+r#>=*PqyEJ6mXj&uKBnSkj!$yQZw!)1Eh~v->hw0hkY!4Iqf%js4`M9KBgi#o zrgQ!*vIO4*@w~;KexuvF?FV&Kmv(nS5~;RdmLFxOR~y^CEx^OS@Vp{mFniJ zXCcdluOObcXiOtxKcV%ab;xq~CrBc-WD<*A;6_bk303y{AuJBTjOF-mvUV{*BW@tF zgN z)COeP<_Y3?OKlTUmqSm+=_AWmUywv<@g|n3&EMef!^Bi%wGtMO(~Kp01$ld|((h$C zvZyOO`V-GvKJ_AFzi(0k{to=;dLW6^a+_Gz8|C0`C?7Z9_QF!hSxVE$`$Y55Rffo7 zcMQbymgV%c)7a8^t5IDnft;vg#WwJR)BDM5=hwo!60$yaI+-KUNJj4URtKoY5CKC$?kwONNO<*h3F3rlb)W6{-LiJ$qkTK2X7vbM8_1A(VQcE(i?A#r&1l6VQZI=nI2a{cEyR5$B?5qB|ua%b@vY`ofYK&eUZ|vNOKMH1B4=N7tB}6^Q38 zQ5VT6v|dTR8<3^VI*>$a>G>Xy{h!Zc@pE$Qoz7E)B|3t!bm>Oc##>Bt#HVRnJOc5& zC8$2BOY9;QH)Jt*3X(`Ivxy}<*%n)t-K(A%^o1LAqhQyntz>s9OE zU7n-Hf+SK)ml8boJ2{8iqiZZ#soNT12|3SL);N*5GUDNr>B#c4If&;i!vYTB^)YqL zo_?q<2JJu+sl}LBW*ZO0&z4DROE(IO(*?%j>qfe6;yqUyAxqR25YJoggp>C#eklu% zB1_CpkVIUl3afS^i*`>C&s*jaun#S+|Vzn}@#x|EVd6=Pf67$-G}tx5+zXx!ndNky;9g z#dPmb+y$*U$Ru1?!ebfBsu7-eysVGN3`KQG(*yCmWy9l>_@1oTw+a6JVy6$1NG+{O z@z}TDa}j?Fy0&|}C}DAnV=Vry$og3JywowsV($audCTo&a*Z_$sv3;yqH_Wyky<7Y z%bFT0kI{Jfailm-Sjsue^pnK$EwCr+t<1H8+V&RKoY6NmRRyy z9G{CU#S^0vg(c@QQRX_?{GJVxNG;*SGReL3 zXk-b>v`Z0|_$!RX{V93>Qan0HA6;YDb3r_BNi!v5zr(4HLy=|iOOQlrc}6U1GkYW; z%RR%1_k<-Np0QLPCj0slPQJ#coz9pI;(3cY-K(I`e&#-8QCkF(NG(l1;<0Zxcgta9 zS--F8BVlp4%2>SWkuKxrz6Wl{^PYq?R0F@mzFdHL|oXoKPk# z9*K-){#UYY`Es^B-VZbV1BmA>YKO=*_Q64|AF9jHuONxkqEv>*{-L+SIw4DjL#uDX zQps7iCy>28N#oibMq|Id3y9|}9X^qHe?pTDPm$%WCrBc-=n%^iyBGNMM#aSiKZPYf ziK)xS1LR#&v1e!ed*5I+i03VjR+9Tz)9kbdsoS^4T4Z^27sT@xO?C2H?Ao4;bI4-<2qckO9uQ0ONgMFqp0~*(nhA?z zGGp=ZB73c;Wpp`(EQ9ZXc-~?*k*rXbSEDtE=Plji$cegj`(DQH_`EuTBvQ*UVp&^b%2s4)*L8CrVew93Ec>$O z;@`+>>qBp#y3E!9@x0~MJ#tST)mx)Ks>@g{kVI-pCze#J?)ZIW%+4`d!cxsy?sZv? zEi+$a;JrP=oj^Qq(KjXgE=Gk1Gv7m#|*FJZ%0kQc?pfi3d@mmFTvgESFY4Ni@MV-azK`U>rRgj8;SoYjgJlOP624Gq+texwsf_+NFW8=FSqn)dmf*%sX;UiSbwWp@dCV{(`1IsL*K)Sd1>9HjZ* zy3^YnpWGeIY6F*7jT3jA&U z|I4JUAc^e%MPVWNf7x9ea2;8$@4IOrEFrfU%bOhX6gXzmm7&OT!X3o(mVvYra`otF z{MM)cQIJGx@gSC@qBVuca>CtphOju@VJy>)$co9Mhikl%W&be{&s$85NH>&+q24!C zm!E+kiPUn7Sl*?M#H(@-TxZP|mNL%L)RA=D9gH<}LY4+cK|F8SIGH?c)OMV32U*Sp zfFx2&C9xbh6jlS(Wy6Oa3xp*rovDkGE!k1D?b`$oWLaJT;(5#JFJ#5UrhdJ3$TH>^ zNFueUf5ErnB;|$pEKlX!vZcZjdzZ1iSVYc^%TlSEjVxncfq33Bzdw0Pdehkv?~E*c z2a-rF=ESnRqZ2-F$>&b&YGLuc$5<|0Br91>TKnPiViU4JJa0)`M0R*PYuEdT>e3+} zB#~NNh~;+3)A8sUYh`p~qp;X#FqQ?k$k|bMiq?9fx=fx0;(5!i5#(E)?vbl;M|Z+} zkVI-pB$nnw_Tcl$5<4%n6P7~G;-XEyIXpSQ8+WJ{tAlvnQsqSMV=nJ6<8P^*8~~C? zEuV>H!Lc>lQC&11tM3+;)J&!>s^b^o=|9QlHU2egRRH37%l%TaL&Cjm6kcJx`4%LR zTGYz%*dI_`umxEb`c>=~mhk(GB`22bD!6rKz+QBX#hnK6yhYuW)TM7hlTc3WHquoO7Zg)d>`A;6C{yZ z9Eqjv;V*cVq5rl6Ucyq&S=y(P)zZVyYA7Sij=dnBx76B4?qlO?kH))j8o7WZQp;sx zX>(c=pRiMHG}BL5avm~uv7AfZ#+@JW6?Z{*n+@W5%b`Bx?X_C?vxVpyb6xrr>US-gG*Z!oi#Ah*<@*CuyTzAEpbX1pwFCd<`Y@AG17!RzRuZAqke}E)XONXy` z?5j-ak9Xd`n^P1fECG)giPKpIAJN`EH)F!uC;|tw5A}-eH{?bTl(3O6~=vwzT%av>0?0>sU@0NzMsFm z5Ls&J87B%$`eUXpRYhdH+!=BJ?<&w558`=C!WMFkN4L9#PgP|o5_7F_FAASvUtu1Nu-w66?p9LQeK32 zM=e%)cUM?Eo-mdn6UhA1|MqcrWbx?);(5yvRq~efoZZyB$nv5mNFudNB9;xGt?|i6 zYd&4b5|&ENGO{)K{Zf^?2#=SsRUn?X*kqI6Vr{SYz#VrjYexO~Bcv8PV!3v%8Ghz3 zoMfLTEcv-iUFt-T-!EUn)y+{|-p&B=yrrZD=~7d$%szzbVmb#Tky_3Yi*e`XC(+pV zvz_o#SQ7IX%j|4&jdj-hy$e}tI)HfIq8Coa{ye?2_&oB12SF04rGQu#_w>h|Di3Zp zFBX>Ie8ytFiqvJw9@X>6a{3;K=Pe1F$aAt?=|?e;Vj9q@hZg&m0JEd%?I=j0^q^6tpeV-H9owQM4m52lLv^pjhsHvSZr%x6qp zX0}+2pJ75D)xtX?k3@oa-r}xIx;jVY4RS%2gjkS7Y6&5hqDBqN(KV*xYgBu~U%Ne` zpEH&(-AOmpr3(gr=o)i93F3K6z#y`oY@RU(ufYzB07;~lTw{Qm$kxJA@Pes} z_IXkl!-o-is4j!UK|F6MxlMKyEe*13gX+>N8YGcg))LEU^*Z>Q_8(KXbQG54myBfu z{T}SE$3K*i<@$db_En~V1OUNt6(x?S_ zXYs|&a6PiD*$v`(%isdCJF3Trr&p1s%R!JtYRMv&{;hoXplhsN$EJOR#i@|7L>(i) z#Z>Dy8;>m7JwZHgsriQd7F%sQ0`I~}(F94P7RB#)?4PlG9*!(Kn&%A^mNL$g-+(;B zXdW`dohn}KKs;|bP!xi9-!2Rpj!$tqtp<`vEklW=?1H8)vYaYAF-%ypUNd#Ml|kND z9*c@qM|COt4&r&s#73l3Wr~v0cXW+4sTus|kC0lH6U$y5s}y7zAGv0fu*AM$EMX?( zU6NX6AnsJLj05q!Wz;=#A9H)@g}I><% zJ?mdK8(9XVfOy_=r9F9H8G7$w6uQRdWPl`6i$WD1`!%0zX@=^O(DSK@uoQBZ+4MPi zbkSFQns!-#5YJoM8_UKU zbq4XgB_)$|-K?xzCl^^ldx9iVi^dOpA5&FQ_eFL2VePwJSmHl0mW9K~T)F8}k9<@Y z{cRwgw-g7EvA;HRDefS;;{cLKEf&P`SuqIjA$g;-c#p6Ilrok<9^~C?O|yo$yW_?< z5YJm`=aPH!Jp(7)(fxf2NFuek5z7*n4tSqZau-b}VR87#Sds(CGyku^$JlbEH;Csg z&r-TCNkz7TdPj=o<4>u0A9zMVzHyG+Db)3halwp-ftXc;2GY zi;VrR-IMV?rDq*M5~-z}Sk`{rh&%25KHfhnEa_!TUFs|%<0W`yvnOc0WK@E9-qNd( zto6+pF>*7i%SMF$kImJ>a?(Ud}b_b$E?EhOPI;DMaa@_G>GRd zCdK5PMbXUWVaRf0B1j^&93Ylrca62klF(zoSz)Q5XG22CPJuynY~_x7;rz z`;-zby6Pf}*HDl|YKbS7YhEMXB8%$Ns%T-!|H9PeeFiy^$~dj80$CQl0r9+L8~u)A z$i>tgWO?!dB#~N5iN&e!n9IoGqMLd}SQ5(_%gS+NtvE<$BA@3H zGmu1Ti6NE-UZdL~%dk16*}{_fjj2oiHS!E|Lp60Qy2czAfq34s;TTy{YOTF1AB~sQ zD?t*erHELTySkVl%X^pPr@|6l$ym;mk$M0A4=X%go_c|J-V$O?o|ESuZTkgTF8hNd zQj6*@JoZnAR%szi%OUQC!s7j%vGlDoAJ3J)T7Ex`>e6c@i03W#j?Q>ZY4GS~O30!( z0VI)H42b2vq4Gv#S@F^Ay|7esmaa}@ep%qS&k|X-R)TomvPYkcedCM%xZ`f4!mK}k zgw(Q~SQ5T%E=Cq}-OitdrJ#zbORNR?TEipb7oU+OM<2xVmc#wYS2XgoMr=oQSvVCW zky;{&W!np7HDvK@@Uc=@l7BFkW8KNxMWN9|{B_1wEkQhQ+4q+0ADg>Y34b~7eLIju zYI#X4{vZ5ekR^QZrQgC5Qq5Q#YmoJ^9Xp@k4pqaEAfC5eUq;4$gX`w_{r=ifu^&Z*jCu6xblB~g=*uNQ{cKV<-i03W-Ysvi5sfD2?y2iS7 z21%qAePRijX@O5Z`mH#%k+76;mXao9-?*l6Fz!(8*964#me=da@0T8a=J>0b9a@7V zQi~0-gevq3MRoZ$tYr&f$@<0ACFCYqw|wmL9`Csd9Sh=l%WWI7&aV}!SQFJ{r~ybK zwS*GOr-ze_kR|a*zN)aq{$?yz&&WR50bTCmJ&H~NAfC43XAIh8Y9buRUn?XXtX8w zMo`(m&cNx^h~X{zQ|JQ1>$+jzE!0E zCFr^$elKPc0Fp>8(}~5iz20ur(ENJz%;Cb4^B>DUdC+FW z(|*-Pkm~=`X@7F#pwXz){)HV#DetuJW!@1pwKW>xkv_!%BobHD-#YED9w(jlTU;OB zMV7*H=}-{3$wvPkl-IW?KgE-#+eGI zzEpicYu1t9IeciLaEyMsIJ-w#MEL7nyo-hzbxRi}OB zpt*t_7wTXgrnR0?uR?-Db~U1DS(zZIy{zGDSQ^S^c4_Ze)DR~cU4Z$C=hal&^q zcTZ&{vUA1J%Ah8?dq(7fc>eC$HJiLC(RbGi*fIs10{3Wp42xuF5#c}x6SvNLkVb0xgW5TFH;NG%S;((2CqzR2P_J!-13 zlyjEVb4hoHX77UwkY&pp5YJnF-6T7`LIMtZpt^jp1WBZp$i;kVI;UA(pC7?)Xjo1Cw#)!Ve~mYzGv*pGPj-3wV}$ActNOA)d3Jb&j3y2geS zJh2iM2S>)TAj1mZn(y5BdvPc9={&M5{REOoEj4Q3UmCYosTd&3 z@AUI)g{6qI%=RNYByv|yzK$#o*&v>`T)9W;qOE>05Ls3~14*P7RbsjJC1548H1559 zv#_M^XX;|}lUO2lU+IJ_^|e4eZ#ipEel>mVXE7OBVuynyQp-SMv1xg6AF}j&G^!)CV_7F8teCV*L-A|`wYbMmLq$}ui1*#2KcSdo7W(T)M7v^LvPo`Cm>9_+HjAs zcpPLbr>e**!&{wb{B~jcEfCLJra6=D^d3sv@V9LZ?|~#z%VJ`AG|Uyxl@kU&IVdcZ zoaJ6RS^2#l^6)LH%PJiZ&s&Nokd>_Uw{Pd6Yb;?bNFudtCzc+?H`btQ?6ghbAz{gP zV(K#c5!qF6u#FZz<;#61i03W+`;fWven&U_HqK-}NFud(5sRK}`*X;$^T@KJ!jkCB zSa#7}o(7xh;FBpj27q|pvOSZmWHsIvj5~*ePlF^>W#km{ zy^D_?v%8}4q7eh)dCRTO-^jB6BS<2({2-R^F9tS6mgr4~qJ<^Xjj7A{V`NPv z!m>0D)y2Uc#PgOu5#$?p_kEw0AxjTOkVI-}Q4>GI42q! zpHN-qi~{kzWq)h3n{b^*Ans5tm;jPUE%S)w(O9?F$g-@?%?x2FILy?g=>^hpr{BvG z-;<{_2JyVba6Q>gSTwa8K4qh<6-Xkr*bs}xvxk1j;ylhNTUe4k7|W||0UiRDAlH5NS+B#~O&iDlFM7zbpzI>YR#u!MLrmeOMKJm&ACI{;l{ zvF0G2w;Y*E=KcEiCHNGlHH$zJsU?(Hx@C97J65MZ>{KW$PDdEasKsPQQT^IW@ec3j z`5>ORRM#i#FhSi1%|>-;_!=aUT5b@_5*=?mSH{{DzY~@+&Qcaac6fh$eFt|BdhP`A zyruitGkE7pgA;xGAZtE9&#V^yJYtrU1K{2 zfOy_wxq@^Lwmg3se{a0uFpxxQ`ARHRcVmtr%U8ecmBJG1%~;yqCp#pPdo(gemhun~ z&s%2HCci63S|;FAX|_dzBvOlVEj;f#U3%(_EavC*e+!H6QN|LOLDu>rKbwRj%hFg7 z&s*Z=lbyspRyEj)uCd;UAc@q{gIGH3pN8L(_VMVdxaqIm9`?r=%cnKuJ{ErR{7Gc_ ziQn`UUt{LKK|F7n zc9!g185_6YHnKcX-1X;=kXjBBOX9JIc!&4S%?6!?#qBs_QC~{d#+RhIZbWslwg>UN zrEO=jcHuJR1Rnd%96=JPC74*crmn{8W43Ku^%RzJ&JxpzjF;(+Hftk`UN;cWTQt9r z-RfI{>UTtS$smK;B(E2%tIC%TaZL*$tIS$)#|&D<^J$9gM=mCpRwfbAUh))dHdrNsp^gc@w}zt9NDq@ zrqA6pWC=6?Nu-vK#Ii#nGZR_9d92Y9mVf}pQgVmP`zq?)9-+Eu_p30L!tY-9b_DUfMSB?;FE=M# znS<(5+#Mv5S~?R;*V@ru$m03A?j&I;;w)e2uITS-xf@Vjmj3|pyk%!CGQW6*YVAUn zHnr;g`6Hy3VZ_q#$1wcNU+nYTNLbQ?n7W+)NY*Wne%*-Ymsi0cp0`wPB0E+?gDdfP z85RMONG&snrS5CnLR6Q~K4(pZCGrGgQ7<9S{Mvq+_@Fl75ywZ@pO5OI zKLR9?T2>N^_RoRyktLw*CJSNl2xcsmGsx*DPo67ZK-XBQ8i?mDBh<(>*5#vuGqTug zfFx4O9%3oI`l1%HbT`pmEG(6rW&Tt$S3WhU#2pd3^FTarnKqNGDaH5ExQ{F;OF$B- z#gAA%DCoo?i*>(xD}^QhBvTh#b+R_zY(#zps>?@h5YJm?tCO{h@O#`hq!O#3#zk+y)RJ(w9-ZH2YnJXK-oBtKn#m*Zfky?t0rToD6 z=E#!pc%i+pIEFBml^00IUF$9xO_8Ot0L1f_+D*uPY*f|m*mdv9dp0`X7BxC>e@BKqiT_OTO5~-yvv2^r(v6quichUTb(s_k;(5#Qt7N=P)Vz>^ESiZRiPSQMSn^J-8iVSR^upUm zSiHj+OW{^BUNSl@$0zH}dk^Ayi^c#lS0)@=V2&(GUqBM6WeKr7`?fm}S%MEP2ojcR z&a%5HdFGF9T1OMrWtj(v=PgH~NnJYj8-Y(h(eMRHq?R4TqCLTC6|#h{>2pR{3c{JX zgv77J@Ay`q^u*6FVKyM1x7gn&YvXA)ekV{}RCa_eXUi<|#=f-FDYgCtVRZDKj5=Z*J|y?l4zs<1enV=QKK zNL@}_?D&A{5?Bu6dCSfg`R|Ac@pcO)T~MJ^6^PvE&DB?+8m)BvThV1#%xd z{CT@GvOLTK@w{c74Ots^U!UxREC*hJBvMODMf?ooKKx25vNW4gbYECv&odU^c+w5E zpcyzjZN*hevz+$ZEdmfO4P z1+pA|0pfYfx_)G?{H0&uj_T677$lKe<`au=kp_O3wC0oX8(}HrEL%R3wTospZ}6$j zzEvQew^aTj*O*20o@vN(qNd58KSFBRLM%J#4%m+BQm}AU2_u$Y_0u*Z%eDQHRl?#H!&rh^k@;oRj1LaT zqIVd?^OjNdiDhSs4&I~q;uuIGwImZu^Cz|O`$~5Wn;M({diPq+SuzaC-k$Zg2k?nh zd4oVaZ+SD4>{Xbh?YRk!mtlG!iPZ9xSWF`IW}>>-^&C}CSaL2gb=k3+tifhX3u%k$ zGIRil=Pg?t$a8XB`{lEcWxz0yL~5xZmW_i);{7nsj2kr;miSo464#paS&WIQjn~Ic zS%7%nl9o#PEW#SMY>g~Kmw+TvOXIqD-hWu+^%K?Q#kj1N!V(b2SbQgwdvbQ}(oLu? zH4H&KZ&~%4%$3WpC%!=zpP3+u)Y6k!KJR)Ji!AeY__h-khs%s5uO*o)Yfdiig)HkG zK|F8qSx(*~H@fck09ih|f+SMQC}OGa-x7auYRN#WuEJ8pSz^Z@#Pds3?X$ti;-UlM zd5dmqvaj#ce(k=<(sL|GBDKsWmYVNs;+~P>F(cH4CH)Fh7xx8Zo&T|IWC^mYo&w@| zOS=!`IXP_dZeL`1W(<-@E!M=+C~U(})C9bJM&1BnIS~Jk-Raho$kn#}*Cr#>oxa5! zWRi5J&*(i9cc;%^)1xiwPG7nRWcj!5^akFfJKaOGLIZWD=dA$g{;@mVo9r-|Yj^w_ z>Q3Kl4O0D|y3>2*X6d2s^z|VirMx@+Y!2y8x9fGLKf0nuUjT{374^67^tdY0oqoDi z%qi5Ju5lS8`5(K}J%1*bq3(3eB#`Xy-RTd_THv1Yc)$D`=)%lR1quG4JKd>vE1c=& z=$W`Xy+sDd^xt-;|LqTK`8InG>Q4Xl7^MC8?)0#&S-3mhrh3~1)SZ6j8A$kFb*CGD zxr1+;m#rJ(b5hjbfSj_sg#9jZmvH&5y90Hnzb*l3{eCW+nxLZtlTjR@AT?C6C{!SzbH77|Cchmb9%^f zGjzI+uy|Z$EH7)5w+j#4@8fex8$^S6-qNxr>FPAs_=r~-7G42Kq?VJ!vi*BbA7mM| zq}>={spKq^X&1rjlsdOiT@J1R@w_GRB(Zp(?ud7Kwb%@jNG;ch#s0`seDSE$ilzmESJ?T=1@fqH>T#YI#g7iiYR#&Xo<@&KU_y zVgh41uTRd2uPj#%M3$lZKs;}WwIG)6{j%|H_3^GCiPTa?EcLuw;3ua}mo}LSOK>7% zd6Y?Z6y;T}9)vB4AfC5;oI$R!jYX=x(UVh~RFFh!sb3HO()iE$pKZ{1X)$u7g|Ij# zF_zL9q}%q=GiUrZu6P27=Pes@NnMUqF2=uR&l`axQcD+N>9D^IJ~3qf=thf$rG&Gj zJe!B7nKxGZ<569d3_v_@S>KZE^6bA|uN$gM?hKGbY8g%}dma?w@4Txc7u4{Qrw*Unhi>szXw?c9t26GmQ}@p6OYPqviPW-}Sh}^mJQ!I99~xyZEY+MPVJaE> z7P{9}kfpf~i03WyCXu&3z8Af5H`MANkVI7n$ zBh&C3UkiH>&s)CrCze$zdoH26jCBM_q?TA>IWcPyep531XO^?DB;R5zhkMS%_p#Y) zFJDA;S)};rPdsnwxSDjRn&`jHLKYikkVI;^Pb}wmMB8SapQ7<%Ge$E=Sjsqysp1NJjb+@_ zx{d0hGX=!+mQJh5%CFkjX-|>mo-s%wwX`FaA-4|qLl(6M-$H~XD~+j3>KO8zZ2kSZ z4YK^q1M$3N;bL-))wfo^hb+5Zfh1DPAYz#jW{SUaHU4$tIbn&t%~)ci$Xn7OeH=d` zOU_3S&s!QOl2y5_n&~RY67dZrky@q_i^1<(M^RnA+;)i(7T-IJ#r*~OPH);|9sHKm zJqyJ1mab38czM{@LJL{E^Fb1+Wht>dx|D5-EcV-%Too4kbjEVljqE0zTh>e^e%Ocec-LBVClJqD zl$(&Zac3_yC_t8|ULc9oa+Fvelx)V|9Den@>b9_?-ev0Yq>Aj2=$x?b6x@pA~&~JxC(8+#!}z-oE(ve$yETa)iY#gRyk3B)dG#b#~x4@pH^UJa4gc zBG1V`w>c-Gx@1{`BvMNuvG}dGEkffZ&c*b(u#|I_V~X?)@4jF0Tc6)vAfC7AtS4iC zd*#ia$TBhjB#~Nv63e+MmYbF&s%J|l66ay zV@L6chxWZd5~-z?5`KpHb<1lMvZR*0D-o9X`;4X2i-q_Z=4QP>{Pwz}0>tx{S6Xg( zyj;F%{|wdT;%|^dYSAPX#b2%Lk>%yxm@mQ-@PM%-eqhp-V(SXxdCT0hWPTaF z?dA_;ndAkMNG%J9#o>1n?(Di1GO>ouU%NeuIE&pB@(goq#$i0~zc~-$d5h6I^7i^? zMT;rOav=^Rky>nt#ca|byrbxubB}t$lAguXMR_xs_ZvUEbR5;Coi~W*Ev}_x9j5e3 z&@W{9?hlekEr*GvK)+lUjh7}JKPd}KSDbFB#~Nf5{p^G-tMR_`y1?QCoGkmMX>`pH9)74(G_%!MYI6%yyfLKGVd>J+-Wkh zc(nsbq?TvI(*2WN8L|wwHtZ@a`Hz{pxcwl{{90C}cz#)G2jY3lHhnTz_8IrQEwZ$7 z07;~lZ^W`?b3eSET)$;&bzw=&VJuF^iRJDPEiY7;@f|=sZ>bwfcHSSl>)#StQoDg9 zQcIHtc-~j*`d$UqWvKtFe!>#`gt6prAonrXlaF1IWm5=<=PmoyNL`+44a4tV8%Ba8 zQcEvlQB|wL?<~fy4jm#aj=7BG+8gq0d3oM(CuCW+8N~CJ=!xX_e(`y2ybC8`J4hn6 zj3$;`cU#=u5u?0LPgqJg%cl?<{5*C;*{TRxvfF@o-g0CdvAEgRz+Gw^yMQE8%N$}E zSZenkU1KdG^~MQHW*$=)wfW>^y$O>Y{E_8G9Ej&F>ix)ltlIw*?ogeZ1d>QC8;E5? zf451<(qW&Xfv`m9GnNTq&d!VbwZHk>?IJ-TYB4*=doc?^YMKwA^{|kT5^bGM*1TSbd9-& z4PPoO$pOmJ)Y0oBDi9wd=kJ`v0Ft#$Az8&)+G zR|`u>0b^;=pX?W`H)j*>?$B2P@w~;T3#rRsgIUXwrLY-DBDE+r#PfbyttFjMT}Jud z-5@MZFBnVLt7Lvzn!7O^)#chr5YJoGM-xl8dad!Pnr0CoiPX}SSbW1>E+Na~$(~z< zrHr#=&OMCh%BL1f@S0M!F^K0a{i;b9!M-z74x_sGnS&%!iw?1@n0X(cve9hn!d=3W z^^&QJ-8ge>QOzET_bKJh1o6D3;Tp2@{%)-+>d3Ok0wj@IOo&BCqyBnSmt+g|{lXIa zim?p+O*)64wt6rNU1I?&Ks;|r=}Oiv#`&JdmQ~guiPW;1SQMKW;?uP@)+~1w7T-d~ zVzi#@yU6Zhf;%C9H30FvWqvfVd`WP^Uj$j$93+ui_7Tg@EeG6CUEa;PcSKn1Uo)0N zTgklt?dT}H?;>*%i03VJ-;n3AIK6>*-XFUXB#~MIh-Ka0q4+m)?L(f&g{6?Q+}TO) z$(ANobI>(5$Opvpmg~BtbNKVfF)Pq`$qNEWq?R~haUbw%F{;a;(H6nNlKO_J%e4jM z8f*9RNHVfqo($r7OH58EUW4t|eDwxo88HJSky;)Qi;KZ=O=Rg%)hkR`!rwBM$5+X3 zF?-A4RAe!$rT!W9r-;@0? zuPo;@LKe>jAfC63)F9(!Zap=7Wa(-Jl1MGJ8{v6B>6#6mUuu2494jp4oMmP?`FfU7 z@zzzy^79*r=PfrnlIO8#r;szq;`$pTky_dlOHS0@&gdE&X6l$AEIGwYU6u?W_c1%C z@CL}zcmas#Ev-k8wTotjeehm5k7Xc<)H0Y@ZW!Od-y5GEXLw6k;@>frwVlX5*j}xJ zMj?ysbr8>6Lb{MOrI_l&_^W1nZ-XRKiy^VRIQJ5N6=2<+)^~*^;5}n`U{2PQ4ELzi zLzet35YJmWSdn%9m976y`x8Wfh1DPPGXt;A!!{NFIyeLa)qUcv+Vsw`Yc9nHSd5d z>zqM6Z@FYf_F9J>n}qAq*%Ks@T8d1Izqplj^hMUX^li6WLywZBCpOJ1Tuk+4LTGM4F2$k};MP+ev_gjNenei>62rEKzhKQiV)KV%7U2JyVbxD$C! z9{;!Z4Saj)A{SRB7FmM+<=ab2vkgQlapyzl|>yybp3 z^3Fn8XZRvim!cCOiPW-?STyRqI*2ScH)g5`O9^MOC?!2sBhGbPgDj(VfOy`r^$2-} z$u+XkMV6AiAc@qnl~@!aW;8*TT|$ntNU_LI~`;o)4T+Wi;mw2xQp@`v|No%V_k$SRHZswW<((>|pU zNcDf}w14NgzBlT$Z`Kl|ly};9dO$kur-jYJ=NgB!1Bt{H^|wy@;fLPi|EtJpji;ba z`(s@}lK-*OzKh|HX{gh_lR8ND_fGq_YVB};{6@=|Z|K5Q?hg|DL#O@87wvJT9py_R z(0Aa?hJZ}}ZKwU;{=mEEUv5C1_WehIwEx~|f9yj(?zCSy;-NF@w9gp_68=}6_T5Xe za81is-_S&Ni4p^lQ+Agad5+vA+#CB_p-%g_86eI7)@lFVb?zKA6EvD;(@)%SyepWy zr?)X#G1-~+3GXIsKO4mJch3f48r>a%oA7i2LG0ZF8mrj7A0jklHiaQ9%7vwo9> zCFDC}3DPF349ikt=AycMhyn4uB}JdCVn4jR2j54wUIR&_7Ik8AJa@|gjs5u}mQNQJ zrz*zs@F`hkn4kS~DzaEl0P(!VU^f{rfjVnrP+g3SKoY5C46!)Z@wOn{1mdx`U&EBOKThQEw#7n6el6e6W?~JFJyce;9k!5@b5YJmoV#w2ZyVA%YWEtHZB#~M+63d17T>PD@tZ}!O2uo}= zV~L?V?;YHt@Q$K`h9I7|`0XMqCZF>a4Uwg@DM%upoi?Q4) zSb(pw)ckoXk>!*Ri03WKmXVH#;cs2=ZuN^NKoY4Xky!3rvb}*Ui@FTkDlCPZWppap ztzOf75nlOSqXFW1%NtX2jm@f4ItSH7RSP7MTAmO~?VoYDBVvqGja|Z$`kSfCKqYcE z;i&5!YooffJ_6!-%g=3O^=IeR`s0zMqaR2jwR|R)bCWk$qq^ugq&W&pxPASy+^cJHI_Z5~-y@6Fl$VaJ2|Qmd92{T!qDLCu3=F zm%NE@+DsMCmG3uzc-}IiDY49MGXk%cdfI^`Qi~d~oZs!LgDhV!&p#q8<($Q_7J1`4 zeW*zws>{A)5YJmWC6Jw7VJgX2(KXg29VC%jbcyA?ZOJ=iiJ#r;xUl5xV(OyWgRIK^ z*qQ$jSyUE*c-~^4MC$Uitg;@m++G2aNG&so#pUV1bI8(Xe#Hr4iQmmwhRq`1>AiG2 z5P#!N%L>Hv787Ig45Pe6cLK6>SqG9xEo+G7zV4i1$g(o>TBxuDI53v6Hsrh3ue|1d zLUp+i2jY3ltkL8IghPhCs*t5s5=bJoI1)=I)!z6dmtdED=Y_>#4`Zph*9Fg&-S5ZY zuA6sWAfC4jnND`$)HrxD9a*mVgCtT*AhFE!`;0px+C-pyTb(-jN$JAk;mBAba>!iy&Y)H+RF}@L zK|F7n~?Qk!4ms zi03U?)#S~hz2}dM$kMbBB#~M=5KH?`V{jMrHXD;%Vaeam)MeOFa*~U~y3KffEOaM` z=Peyiljr0}ZB_goc+LGFiPSQLShm~G!*8$q%>eWQx5td%s4O-gzL2)M7;}>`!{pBfE9md6J6UxlTFvrG*nZ%N~PhH0X@)SC(7d5fF& zX}r^GxzfEp$g*=TNFufP5KCB~QZr?m>@^9k>Y9&!W3^OnQ! z$s6B>ji>ZPmbBX-iPUnJSc+rxERe;{yQIFbc)K!|mRrf1(uZA5@b60RKoHMc?$h7M zqw=@uA}EYWzm`U}MKmWr9=8hf2& z)DBseDPH>XM@TKdiRE75C;VQ_^UDquVJUE9>T)27tS9%1SgnpM2dhCmZ!xnc*O>0* zCipq|U2Xe6e}vT1x*2|k>3@91d323=g&K4amSlIvGG-TfPF7qv6`$ai9tGlgi+f#S zdETII1ggu~c#uSD=}#>F4f<_GmMz0tbrY75LyTpGIawd`JkSS^{md~Sp10V{Bx~bo zXJYa9`cfu?BvQ*HVmY>L#2RF=)_m4mSey<*2QWy1n%ZDGmsVCwSnJ2_n|;>WMc$YQVp#PgP%VDb!es45HZ zSZ!(zl1MF{#FA2xg1GrSo*4wYiz;#Sr3pUb1R7FE%y^hM|b$PN%*wW zi4Gu%)N+(M96^@BT|hiV`m&2hTp0_NWPu4DszN*eZ*I2I$Ac@q{ zn^+bGe8%thSB|c2B`j{o7|S$&@*7!sa^qlRxi%TZ^Oj$CiN(}n@;YP*nE{eWEn|sg zam?*5s4i>gX08#Ia?X;kMCO;RL*fS@%h@F$p116dBhO=BiflI`%kb48iPSQeSgxE4 zyMipMUV3j5mK-0ZE=teL@bCQ*1GnfROQ#YL&s)NR$apb%(gLqtRA5yBcv8rV(EX~ z7{BAY=h)9dSOShS7XM(PQ_h;%-qDuCVy!WuizmfO*fOy`rq?r?5AIraA%MMvao&ZUtmM_FoZ$U~tvfMJX z@)wrK0LEg|iFBtgnP_qt)#aNRi03Uzoyq*t_FxU%sbV)DB#~Mgw!riL)v84NTdc~m z|0!Yd2xKg;)|0V+dGFx%$kJgIi03Uq^gXh6N>zPS7q1NY`=|;(3exXR@c)JZ*#tvYcNIl1MG4#IiK1ViU3ioOHe{EQu!= z%j~V>ncwHYcs%bPI1A!=%bsiGUDAibh$v*~9|Mv|Eo+J8*S-_Z$nq_1R+6v;2Q!xH zJhEjH2lo39tPrh%lctt-}wE;nX#xY+EE~h)RILk#vcl8ktO=DQ@*f7pJpuQ%gCCN zUf3KrRF~JjAfC50F(H+ zM`Y1{1mby1d&5Zb46|_WQ)DU02T7zB#g=&9@9#Qu0MdAc@q{kyut{jNOeanlFmVgr(pNQ>fpZ+S>BujNu-wL#B$|qkNv1FU!2F*+4|ReWT$Y(a(g(LE7SIza6s4C zr6VAox0Jmm>y}M*=JrPxYd?@gYS~RJh5EntAd6*O!-m3A##!`Qk-BW1aC#E5lw1e# zye0Gu887LX=kb2aL$^T^sl}IAnpWxKubY_F%xNYpSrJTKYU+@$HEdbIP zheoNWJH7Tgkm~=`o!;Z8-)+>Le!>Q%ly|59?K|+X6ZYdbt=9G+k+`D%)}4OLl60p} zFnWl;1HXJ9Nb)~+r(1v9bpmy#E4YAUfA3DeUDg%%lz$qi7Jx3y7oH%&KXj)*Na%($ zU5KbmLfz@Vd_boEwmbcAe_-6OK1We^`jJSG_TRhHht_<9yVI4=$FxS>>1X3W!vCr} zJ>K;xzW!5(zQx}@th^3#%I*>)^T=J|^jBMR)Sa$y8>IQ)y3^O1@5E<#U!3|)P26#O z&oOt;$@@sZ#%Jc2$?&&uVB$55UD7+>AFRq@Qve8V? zdfvm{!eSrESO!Froh!BuZSmXdamzqFZ!x$+EWJB>+ak;DH6V%9qRArDQlN%fa+p?7R2)w!$k75F}nV@+sN|dB1j^&v>}$Vs>W)lE;HsV z(iWD~^Gsbjb|u%?zKJdHxjuG_Ks;|LUQbr#9=dAdU26s_K@zED0I~RL1miBXx()h` z6qfJ{jHR?4sms!yH_xNGY;FPKc}r0rvU82qckOUJy(7Hy7~R>nNoq>x89IBsiO8bva$}FMcwA;IDlN$Ta=Sn& z3Rx_Vf_UCie2$El$KE&b*q;&vl1MG%h(+hs2E3badrc=NVX5RSL2Jlv^-R<44Ut8o zA&BQKq4((X*rckts4hA!KoY6Nf><6zo%)38qTgfIVPVO?!qnwuEa_Ca*m^o%`PJ$N z;(5!L6{IeHxoz=T(bYph5~*b~vHVV})eTvc61p4{mc)3*5;T#V=8!kQY6hx{LMn*o zErWWKo%bo@yW^8w5;H&&sl|<0E}AOij)5J-;v*XisVew94ER~JOyzhFwJKjzB(-Fk;mS?+& zW%;6eBT-$f+&~hkr8}{BbqL;pET_C;(}bm(vkYyp5Z}kb4esK3zaRj_^Og=uen;bFOB#shE$+r-uJpdN zWhJuozYmf~EoQ_rFS*rcWZAH7%41eCGSUzXTlPa%vjook>C5rmi6&QmbBU}|HSi_*rnt?Hvh^i zyqoZ8Ly$yjIY2B%8AotuSJa2+g~H-=ld*KKPp+{|h6|LDWpf3H=PjS>lbtJRqffe_ zx=i>5l1MElh~;`O=Ygm$^@>lw7nU;4($kSVC*SSjfZzDe{tV)IOWAI+revUU7?1tu z-$4?oC7xK?3|NVG6s=QO|5;eFZZUO9+)aKXCmuJ)>z01?K|F7X+(35Ue(k#z|3;qH z6eN*a9ubRTt&{i!gfn@1--RVMg|TG0k)4sJF3;_WuCYb0K|F79C=179zg|r5qv#r& z`T-=7T1ts!(#SP&=o+h`siI)_*A5BaRL0V33^~oAox(T#jfm=@AfC5Oc}LdAGz?nd z-^lMqf+SK)-8OjMudUc0pA=?xyFgJ`?9&)alQOc~qe*|8EvPP84?#R{X*!xbTej?T z?m4Q<<~)!@YUxBQ+6yOYqq@v7k8C6?g`6e!AXyuCyXG8jqR8|JP&N3QPDM#xm0O0G{_(91YGyme5iV&s)wIgy6CNsFso* zy2g4}fFx4O3S!YYG3@}d1TJsdMOfU@8OxIWs97I(LOz}b;(5!EIpjX3qH@j~S@zBX zNu-t-VtMR0B^=cy$$I@@VTr%TSo{k}x9$2FcK8{l!8Q=jTlxf(Zm9gtckxNb#~nZt zsU?G0T(z5CN0#r2!*zuvAcL{QDv`Q$divxNy2gH_fq34cJ%_BroE=|z4ArILK1d?9 z6cLN5(~5q`qA^Z!tgtv_GL|DX$*#4Fw+7%7b?Zz8@w_F0p5|a@n&XQs<7a{-QcDe0 zJntXvdVdD8BxV#$5|$#)a_l?VcQIVA-A7cHvpFE1x488tzmfZOj8;IFrUf90)S^l( zyMC+uLKYvhGt-16{XSEddb7zX8xe&cmLtokg&>}{T=OHAnb|{MBFnVpAc@p6kXSq% z7eye;_k{Ik!V>v_v0PIiV}EBaGrTjha~g=}Eq4sbyg$aP8t+zbd>mC>QWPOc!ux&pd%bUKEo(o+?+;k(xoHOau1WFc-IjfV}7O} zp10inMaIe|%ddwZOY7+%iPRECEYG5R@HM&K!$li}CFnV0SyoJ*U#8A=_d}MHWDw6= zMw*jvs5`UIg+U7*#=l$Z@AfC6( z^d@tC{f)EmJB#CsK@zDYhgdRHcB!LdOs)6t9l}z?S^6I#bA2i*Yqz1g4Aleiyk*I7 za-VG1!166R#?BakBvQ){Vp+e%$P-z%>`&P%EQx7MUA`}#j<3nan~m{%v3Nfa&s(yV zk@01?c7J%ami} zJ#x?2P4P;|^FAPn)G~@#22Od7=Pk{a_6-!4O3sp1N$$tuUBi!}y1ZBe;(5!93^MK; zgbvC>bxCmmNu(BQV%gu)3%}z#_p$P{uw-R0bt&CU);v_iH=Knm{xNk!{{ML1va*!i zTTU6`h`*ro=mAI~wb&C&V#3}^bndV5ObHXeB{q|>%sfoylkE@pyofB?he14V(XK=1 z#x?Em3e|A|Ac@rCNh~9p^uoJrJgzu;Sy+N!GL{1|*7)52mV5~JTn(vptSX+jSfBF5 zb4s~)+iyX~Sat)DL~02kmfN2?;$56xsV=x7EN-tDi>p0ZH&pAhnm($_$?hPYw6Nbp*r45ouEq91z$e}Se(J?l8Xy3cSQo>nmP7q7u-Y4*iZu2o9p0{kAN#0lX zI(h=nkM%GGNu-ucVmaMW6~E*2nfNnaSW;edb%`Kz<5v`JE}bn{Ti)D zPepY(IvXUBTD}s?)hpBS*MQ8w-boggs4T`Z(~ztjPPA{Ki;l4ozd<~2x!!?{FY1w2 zc(v`^y0KM1LTXWJhsXW4o2KI39Q5D%rwfZ;He*?_jm#+(bUINRS#&>vc;1rUjNHR0 zS#-tY%joYQiPWM-EJG%_;Qp~ubrxj_i&G9`QMpf^E8mWuh5M7wH3IRxWhH%&yl_ER zyrO$|E09ELF(8(vK{wZ;!F`prU?^%s8?3rk!sQ2oTR(I-Mlv%k_3jy2vs+ z8YGcgb`#4F6??pEz}@!yeh7=l8^&_yAsJt;eem0YEPJ|xc;1p*hs?w1PRnYFEVeoz ziPREAEVCzM)%li03Vt$z-g2VP&`nS-gLP zBvQ+DVwsj)@c>y$n)j$LENS^nU6$V?yKBACt2qH#8ny@VyhU>?8DBnZ?uvVRes={) zq?YHz^37QlkCnFlzBCe+=(mi;`~{ik9}vIj2C_6B3gUT-&v|mbgePmQLv?vG3M7$Q zJ`zhU{~DjrBH;8xW7-JIwt~N1o&GhOoH;%kN9Ll{={~1GCdumb;kEVg>h#d3-@2mJ z>F>fomVa8EZuH9l-^SHcnYdhBrWr1l}gvl?2Ro^cbT`ai8s@3Qt& zEwnm))qRjszB+x->kgQCwf)&#bVS)d0g1#B^~dV;`zPMxB)VB+bJ4fcvtEEC|8aG? zYn{TKXm$GL9FXkK)#;a`d*U_a>*IUZLkH%ELXhBZR;Q0Ts(~}r-D(n!R;TNI0-65p z>hwQ;pq9q9acFhA*LRTipI4_l_Wg)gr*}_I*??B3FRN)%bt?X=)#=ra=Hr^~oa}~I zE^lfCa>}j}G1XM?9ko%#qzhm+yWizEq;M`{_5GWGkIz{5bAOUT|M{Ko?7)IWdBRG zdgOmeIoA9XvS{Y2bP^WlLdG)Co4h6MJAN(hC~_zQ@w}y?5wU#hd1wK$n3sYiQcFi- zQOSHf8d(-Ee$qo&3OGwUO)?`tro092TnV%T@w}zh9MT=N|JbQv$kNCGB#~P5iKXII z_(^2hR_?DWED7(Jx;#? z>Ki{tSYkdfmZpJI@Q~TH!zVnxJPZQyyd|N8Jei-_-V&et-k~6g)RIUnCeKt_Bg^Mb zK8uAV=p$os8AMh?o%QowijJ{gJwZHg*;S8x<4!-QAPCjvULTM|YAGa^@(T-_Aj|dy zn^nT%TEbY=pOG1pI{LqEBTKXAAfC7QjwLIIKL2Q3i7d5Wfh1B(jrRC{sMggs3dnNV zv8TPT6mb^Y1PgqOjU3_l1zFT~f_UDt_w0WBcA@=;8qbg=$^#^kT9k>!P3uY^vRryy z$5B`kKQVRjtxxuiYvS4&pD)b|K|F7BX@BvMOXVtJ`OI~r zED@g>OSKbZd`Y|+*$maiuqlY=Ex%TfYqH@tc-WpciZJv$AL`+fR?c-~T>KxVRD8K&bmzDa{Y5~*cAv8>8mj#qYl?=@$?uvBuE z%MqkQqBLtR?)I?h2jY24Lp!n>>QeNy0#ujSAs~s=;z%s7JNa3m^QBU^m9Ma5l`?gi zW=fu4OxoK0M8{a%P!P{sYN?T~f(uLfZ$g$8Mj(mQ;zKNI$EU79mN={S0m2gdm9Z!_ zB5xO5wgewWmX-@aJa4gFNoM4K^bEO!EDu+JBvQ)-Vwo4}tc5JSzZ3=wOK=%uDO^pS zUzW~ju?bmxDnUGN(VR`LW7+Ffa2HNet#?&FLTY(PEUO(B97LAH?ibGsi`zHG($RA! z9$(@U*VaUq9eqGNZ`ol%?vr2cPs86FJ~j{}ky>(zCF_{OWn{UjwjokjN;r#ed-8T6 z!t-H$WKqxt@w_Fx1)2F>ly9z#j?{WHyd~3vJR=8;{)xwZg))#tYH85{-@{x_ZT}tB#eI3>SYe6!&R7Cpk}l84 zx;^e8%dT}Gp0|8!MUJsuAtCosU79(9BvOkev7B+*h{uvdl3ANu-v2#FF)B8s1$i)30g1u!R0(EYClX9jUTB`s1Bk`kVss zyk&|i*}vcFUWOyGd^iu1NG)fG7$s!QX zTim*k^X2N(&eM=(_bQM?YKbM5x(jS7kfp`LfG@&Q%2_nD$T7C0!F>F@ANLf*^Oj?w z#L{kRViRPUodJ?aEiZ|s%aA)E$fDP6MTM}W{buU&e6I(-CL6i+v_^Hgs0HGA%Z%Zt z@iX$9w;}jC=F%S|ky^@#CEjT`Uft2Mm45Zjf4zy1c2fBBoki>FWUTyLv9&+4G*kug zyhZsn8DErpUcmRsm(@TLsil!B9{0C*?{W^+W%&E*^@PQL3uAfpgv^cGeX=t}mdsKR z&s!2wNmoHYo*`aE5L5}0NGj&)O6-cDE&wlQ_lG$gwPdAPjC--nsg4aD=7>Fdd= z&S7Im-9>eIssWNnEvt!Tt--u~s4j)`^g9bnm@{K}5KJs-tG?j#rTq#J&s$6t$#tw+ zLgzqaaajwJNG)!}lKjQsE3&+^uhm0XytXqI$CG4rho)v8UZtkI6~yxvk0WGf3f;o- zjgZA}7f2$toFEpLC2QIsOT#*kdJ9WAXK}AbRztaNnvU-+4>bbuyyc4)ImYV$va~=J zvsNI9)N+GZuAVaQf-E&g9qKPEnJ!FSn${-sFwTC1N1*FirwJgQx5ScIxiS_lU2=>)w#)E9by1!V;(1HM!KBmcbLUt1eP!}8 zkVI;!-w}`dD{pPatJDrTdEbAYTMdUxm@-Z7-jZxGL0j8~Cyzq!M#d{md3 zb<3)Lgw)cBSd7NnZANw3cJhX)uoQ8Y;ihCxsq=#jys|6$Jc#El-*w3x>~{O!p2*^P z86=Tf1`>-xV(@xo>Ei2RAuNfzn7WKpA>D-ME|(UeW6U-P#PgOqb;&*d&$@>%B1^+i zkVI;kLM(TOW#OH|ZnT;|OIRXyGnT^|WOe%Dk1BX~;Eo+ZJa1Xqnq0@!^RMD@zf2t@ zky@4!i&ndF$5CB|JkeYrEIw|G#o!lt7E}896L)yuNdxh`rO_=i?nh?Z;vIE&XMrSA ziwm(#uy2ohuJ(p~TP7@(oJCcEJR`TNu)@dK@v9)7xAc2Wz6E`==YI^55{7h-K$x z&m?3ynsniyu%vi0b*UXiR;VsbP2G$vlQKX&Z+Y6C^eH`eS#%85<$VrFBDLre%lS>) z4G5Qihe619)9D1ITmqczr@*@`TyjX*qa(J?2_FFIYi;&H z?f_GlnTzQP)tb|1Ba3-=5YJnD+YpP!72D^?qNELyNG%tMQYP3=$^26cr%tJ zW63?tSfztaQC%7h2l2e6hAy$Z`P3y5)g{sxB#~P3h$a8CUqfVx+uZ1pu#|F^7B|Q} z%)Xm8Ymr6E6~yzFdb#9&?B>u~=E%~~10<1JeiBRW&4qYvf=^}IQ(;N-Vd_#co2+*_ zy5Zw_Wa(Hpy(*r!Bo~l-{(y0tijXBy2_%tPT6V_wFabX{;~lBG*qnGFEYXJ;OKcqJ zx4fp>{v@(=SPbHM%lXlyr?>wQ&yL9Qcoj$@wP+DbUX*obbd06-T%Ii~{)ZXM*2Cm^ zKjQ8eyn|a4T@cS(W_2geFSXV-KaDJ*gFq6gWi+u^`z7M**y84T1;XO&%UA;FH|>pI zj>Gr-CMqDFx74pk=3&-OIfPfImvsS2q!t@uX`f!a0M%uzYqO8SQovcX3rX*&n@{zg z=ostg3F3K+%}8<`Yjmf)^z;qJ+jwP#&T;B>AR@WXzL|p zaheO_d5c#G>4Pm=n}@$xYr7OAky`E&OY)D(;m9)0VxWSslyjDmXUO@oub`1OvRKas z@w}yX4RX(K<#rg~Cuc1JNu-un#PYsxzcXmDsj_C5Cc@%#OyTdo(Y;!MyoFWI7(WcH zf?lr=GD%iJha4fRpbrPM$v~^1pA83D{%IBTiw6dHI--Wnw0N`%`sx^v?(bGX4+$c7 z1)&+cx}sIkyC;EE|EE>Z3Fd{FXccrv3y@O23VL~8Rm{A*z+wtIq7KgjiNq20$13QR ziA6Zc#&rW5p;gcw7J?-IaTRn6{U?jjD(KSXAlaX*pu=7J;I+#~+jd%y4$P!AAi>|P zf`g4W&yGX2|C(0}|u)s;sB&?@MS&LHhSuYxW}EXS*$=c(*_gjPW> z-whJ}t5wh*?j^XUQ*V9OMkCRgeITdwDv?jF5-01=I)_$4?>q$3{I^xmQ+7t{p{r-~ zu+$dfij(=jE6yLQpyy8`Z+t`M`{EILoC%2Mub%ZP$*i2W)%9QKe|c>Rl1TrTZ{&aZ z*rVPHWa-x-pq;S9_%oLLZRDvbJ?Aj)yw_F-@w}y@8F@0VXf|g#vMBWiNu-v>UGS4e z#K!v9k)>tt&0T~g=s07^eMIgyzPi3XgswrFdLW*+_`W5xKdaWn4?vdI!$1Ia>aRv1p0{MzAx}+fmi=%;mZDZ5iPU08EUsEB@Ec#V z3Ed_LOGFT3$=XED7prMC$eZ|SAfC6(iXq?b@aZ`CDzfO!0!gHnJ;ai;EHD`zW36m{ zS_q5J3C5!6My|<8N9yiH$C%?{5YJm$%plz!LmcbinXJvLK@zFuB(d~*@WB^Z7Eg$r zD=d|qWn~@GdA}i|5MPr+r-69hGI$hud!2fI5dQAy;8`Gv)N+$p7MRr=gDiIH`&I}` z)=8!={VcZQ;Xlx2rXi|J9~}_STgH4Q$5=u^N4&DDi9SdowWJY?k5xmwvP=2%y7j^m zdy26{P9f)h@~;g|kmd0Y5YJmKb|-HZtLK*E@AgF3w6FRRQp+b|ndS4iCaTNM2nQ!& z2@Yl~E%R(}UGlTu4MY~LTOgjd%nl@DmqRND$9k#`PrkVZF8cOv*Nac-~Syoy;)qEBuZ-y{hj7Nu-uR#Nu9Y8n1%h-BQ6vSW-fm zx{S&vZ++~J4<3r@@<0{D^Ol)~WJdmb`K8*ZE~ec<5~;gvBF_v7ELe5A$SxE>^tT55H}<970s8U*5bOL!5P#nu@-26uS3GXzPb78PPSS7Sg4s>{!@mEVLV?E+Jmm?6ZnS@{co ze(5(A#PgPtVDhHqUZcGAs4lIoK@zE@AF-&^Dqe;xag2I6_kzUQPn>gWCA=a8khCP*T+Odytig&t#(#j0(fqOkafGnUe6q@(D_u=#_K zMY{`#=Pj=7$bIrEV>R3r{i!ENBDE|embvPaaku*Vp{-jBi*p2HDW5^!k2F@izZ_W< zj6pna@ku6g7dQ>vP6SD$mMz4R;Gk51>hj~k+fKq#z*(lxBKMXHmuTR7%N0*S zJa0KtgFGYua`kSGEG;uZ5~<}dv9uc0AP!kJ@2#gPED4vGy7b*fo?o0Be(#Rza>ozE z^Oh$+$UVQ&*f02PoYo1DL~02qmITG8c-ONy%QpjrB`lJ$_;e-be*KRnc&=~oJP^-Y zGA+q7vhTr>W~eStmx3fxO9HW+3!Q>j5AF!I9U&}Uml?}eKQgCOeboDls4mYgfOy`b z)_`1--Ik?{K$de?KoY6t4Y4fqKJXnKW1d&rjTe@3&a&?%ndgsiY*UOZR(C->Z+UZ? z%)#cyWY$C$lSd$l)bfj1I0}>i)F}C%MQfzmL6?M zUDSj7bwG9bun{DYT6z6{?nSj z@ps^ld4eQT%WPt~_OYcIvTR%4ZM(1(ahCH&WX0XMWt-|D%PD&h&s!$Z?~nD^HXgqh zd*cL>NG%(QWkJ?hJnrY$xaJ`&iP2152C9;0WZ&++h9irqB8cZL?)K!lGS|Av(tbPPx$ zwcI0?#s%khA&cevNfE+Q$yr_&lU2gl)Wchvt0l1MGD ziN*Kar%TB4IAqu@VadA5)TN6iIbTY33-SAxiU<(TTQ=vBIi*>1=i&Ea%GW^>simA) zq8HeWN0zx>PaX+NYz$*rX+g#ppVW=`{rPuR#*2Wd^ZS3`tyv zEM4>8{}z^%J4{_ZR-1$0@jdZZT7WDKK7x4Oq8CEWeVsui`2EYcZy<@(vW8f$t()kF zn&}=?FIU|1*PK%M|1AG+Go7u?qUBmPbLOnEwVAX2|N91o-d!@Y98(dq19f>WRREbJ zU7lNS55V8LDNx^|fVw=JHU(M!smt?{&M^E}w{zQiA9Z<-Z3EK%U6<#md8Es8$npnQ zP?x8^DoFKz>hetb79N1QJnhs#N_m%OS|I82OzWAA*Q8}>f<)qo`lHKpN%vy>U-?!H z+lRV5H|v5V|FO%{WZmYWsLRtftD@?-`m@XP+LVE~H}hK5f#K-DOm1CSmEdo>JS|ra z#+mY7+^j%do+DI2rhnVz`Nt0&*ib$Py22WkJg%X67RosnZ2DAZS|yLyhzQXAXV zLB5X^P?zUMEs*eEb$Q-BT@An4T(bXh4!Yu%^aDAiSBcH!D&bUpdl>5SJT(}k`EOmG zr*UuXZ=m&ZsV7=GTzVc@_LX& z`oHWY|4UiVc59JkWBMLdVez}mSnBC7#iv}`KDAHKmlwzwF6(L$E;0U(~YJar>?=h*?faTiXu0Z1aX z#1PBi4VEKOU6#z+J3v_SILqEHWEOjk=MKE5SNtLn&s&yhk~fP_?Vof*bs4b=B#~Ou ziDgts!^_C>F6{hBVTrrX)Wz;QdD3|NB)SKxOYs#D&s#3`A$J?;;gi20%jsJniPZ9$ zSdu+=%|sTZZ{ZV#B{Yt)^v))okvGqVPGla$C0b?1FPuBWOe^|2ts>^&65YJmu7m?Z0{maZpBFntVAc@q{ zg;+8agYZ7Sxsys32umqv$(cf)n(8+Waig*?LH%PjCIxmNu(B6Vj0!! z`CL?&vICvHgvI$0V{xfL-jp0E?0*N<< zaEJGnaNT3VQova>*OSi31ryg?K^Dho5YJmGcadW(dz`KcvdoJGNu-vm#1hxTE(q1- z`!S^uVM%z*)WxUL8sEdTydL9=EUQj~c-}H1oQ#!DF|F~co30l?5~<}Wv6x&QfoB*y zcCx)BEMZR=%S~1CX3@Trw<@xH?Fr&}%Q9bbj0LaCz~8w0*#{(%T8fBe)|0+?-#DWo zwlTuumB?6f29cSpcg?#jL6%|0AfC50P9<;R4^@^1qGN2JDM%u<)b4@DedWMCmFO5- zVXPM~EajXf{tS?u0{BvMO%V$o7mJdf%UH83Mf zSYnbH%eqHq_}o`m-UY89nm!7|^A`Jt-uQXDo73aP$kKKKNFudNB$f_abZa2XkZ^?} zVF`N5STuH#dzfQWr{WpL+0h`LxBRBvx4S|n8X(J#Sdc_&Swt+B*>8U$%L|n=Wy0c` z!dMbaNaxC#l-mjDd>Nw#;(5#P!=x*^`oTN%kmaftNFudtB^EP}LJef8bhD|x^{?5| zBF>VeMV?>oZSDC3S@aKrc-}I%9(jJb7B0Nu(BEVo6@8j>r9TJ&PL%OX4%8 zE*EpiG3KST8sGE3?hE31%LQ9<&+oVJLJMTk84Qw0EfK_GJT9XRs!Mi~e@kJBc+OZx z*C!o#uIGP;pt_vM0P(zK;B@lFx9hCWc*WiRT#!U+c|z?=?ChS2jeuGsWRDXo9#HKSA{oa%D zSUG9bCj(^Zcn-w#mL>*dto+in3tk1S77da}E!xDQQ;=bajuj5?#F8G zSTz+{-kk&Syrp0fc@_(AI^YdD#`Z*lBvQ*9V$mOMI{;a<>y4c&EG3+!PDk=SvGWPv z&&YD535e${0bj@*Y^g$+r z{+JRF&s)?h$#txmMb05)8CedJNG%75W&1s&AE++x-}KrbEKx5R%SK1?eHe$>=6IjQ zt6xAoZwVbuENKrP;5nskKS2_yQC--%`F%OB~;V${z5gs{YAF?CUzOWuopjjxNx{oh?dJa1Xr zo{alLOgCLamPiedL~3c;3y=HJBYRCpmd$PUgb7P%He;DGm+XPJZPK+*s4mYsfq33x zGm?xin-7>(BFm;8Ac@q{lURn2itUOl=A|*$gvBF=vG_kH$C%&vFL;Hj)h`gwTP!We zG3Ffea|^PV)VWplBczs*#PVZe3_kadU%PfsSV}odjnSmb^ZJNeJ5gQg-v#l!WuY!P z#x7|*EkTwo2_T8oVnr-^R|4^h?q#9BlY}KLm#IsSGi23GdC1CD$P#`D#PgPUmE^hd z+pSXkj?e2lNFueYC6>`f2^Q!WGg=*)DJ;=>jO9o;c`tUXixK`7!TC)fp0|8VBCFG% zTy*e9by@8Ul1MG?#FF{)CVr1RU9m@ju=u}WESfiG;d_41Wsa{=UA$U?c-~@}P3Fcm zI$i35EbH5YBvMN-v0SlU_7>H}S9SOoVR6o9EDkToSh=M84*ZPVpa+QOEhE~HZo=`6 z+;R8qQ*Dq$YPm%$bw;>uK$iT3l%K*pD z>SCJ>l1MEth{b&EZd=r}^*m&@!nVKW$6Vh2Z8u?yN^bUx|o6tmc1nvR!e7U+D9hf!J=2j*6n{L9^vkh^km&%RS zpl-r>uRx}M+fDe#4|IJP@*8y%K7Rw!{&P3sX`4p4n^3!E#aPr$xU&c({8!zC{f)cg zw;OM~w93#`!u1QtDZ5HIWs|GK*WEYAqHaQi3XtZ%brU9O4#u+#A3Hp1EUq{S|GVP+ zv1erTQgXL(VCwG1JaMjajSmyWC1bsj7% zneUjo)Eh_YGG}UG5VDMS0P(z~uR1wj8fTW`^X2PykVI;kN-VGRlkwgjQ#TzLEi5tb z8Oy{R^3MwYB*AfC5uvLs!eU1RfJA98DR$d^g{6|SILDD=?7$qo zKvb7&4j`Vl4EajBJqD}^{EaM~w}T{7O9x_6$@q+~$&I3-PYO%cC#EhgPUJ1=&e56^ zk!AEP5YJl_7I@)%?gP^mYaq*mIFLkY89*!@ikI4;x-5L6b5U4gKQor*31lX#ooUDp zWI2}x;(1G$3+Y@@a%#T{S+uf25~*bpu{``X+z?r!pG004mf$aprCT|9OM2$M-7sWv zP6zS4CE*!4UlQB;O+=POxgd$uvY1#FZHdHtc@}<{_&`|PN*PPFF;;k1?yX^N0J6B1 zgLvMux-~gp7Iiajhb&X7O{)44Qp+}CnSJAaZB&;|y3r}ZQo>o*4kUZv`A#Zog)FCr zfOy`Lqf2H>4cb=Xd;Ww`Ac@rCM=VCKr;`1ZOD<&zOUhTKE}cKl#C3UWY55Xas{a7- zyk+JbGVWhEJ~s%}C8uU&)sK)`E)k1gpf}!=*y+&w_relY##jm`lkVHdTF&^M|7ie- z=Pj$ONtdTi@_jrb9~=UbNG*?vrB3$nV04UiFkbOZSp2>*mOP~yxGpQ6&d5S_X<`cE zc}r0u8TadVUgC-B^2`Dxky_po%V9fHMP#|!>~0O`zh+CF${EYlU1UXsxlR2=$TCy~ z#PgO3C&?^!e2=HSk>x`dkVI;!riI7-=9Tt%@2Fc-J}U}K9%q^DL*_2J9eagmu@kI8 zJa4H>-!9Z1+!yZ^J#zs_BDJ(3mM4pcwM2C>eDaw#r8DFAz@0^C}a{3jB z=Pl2S$u&8qKpl5Re#-|*q!t}wX}#zg-VY;j#HB965?aAnHrtVo)vjY>ap%3-I1tZU zE_Wp3e%qZkO6VB7J_RI^T13ILyw*Z?Fz;1WBZpxx})6fen6p9jQG@Usy^xi%vW8T-mNH&JkIR27-9ra(yhZ z9R4wZJR=VWNu(ABVwtgWj59i4Do+eB6qd9~rY_ymh-Fgi7!_oxbsog?mS&p7qPJ-w z-ZxI~GDsq|93+;CAkrNb+-Ax|VTu09SZ-e=mWsKKmdH}p8N~AzjZ8A`s~eg8Khh!RbW36J|HWAL?;!JI+2fOsBFmG;AfC6hyhX;B{=fU-m5^&& zgCtT*9I>o?Jkk?c8V^;TFD%Z#8OzA6q-#y%Qg$h_%p48kdCS=>GC!7IaC#c5OVC7+ zL~6+<7K0w<_?<<&i=``trGT>xYC`V$)$+eSMwS)PAfC5q&~HPHGZ=xtKHnl1B#~Mw zh{b5?3_Pba;F!loVM%am@aOxN)iuai`7G@vUZu9-G>GRdnn%g_vfuMuHB^_n7eNxK zrCD!$52Np#gI7Z>uS|Clmar|1iNNAPLnawB3uUc-~Uokc|5+ZPwyd1lzlU zBvMN+VmX;T06)KI*W2kSEM8j~OFwV2O6{@2Vsm84Yy#qW%Z+T(RS*;$iod4aZ_Tc% zA0f3E5=+RT^poh^PrWhKPgu%1OOzq$zCB=Ef#=6M#({W=MVWN2jB4eYgpRS7Pe2l> zWhSvSh}*p%Sypa}IwdTb+nBnTCe6mrFHgQUsEz94wjIRtmOgaFU6e&49$(t*0ZF8m zb;MGcH6QP>>SyY6QCMP}8O!Hu&e&oSa2c;6*kTFdd5c>6^LT#jtZBtwR2TC(Ac@rC zK`cY>w9iC!>FKigy08RoXDq?jNO#n0ot|@0U3A<*Ja4(=M~<<(9+}IKrHeO6BDI{R z7M(@-x$=Wo!F^$Ibzv+U7LwKJ`};Q?k1U6df_UCyxs!~Q7xp<>B8&S;kVI;^O)P`Q zcfjwGtY-%#3ri7aDN`rse!G?nyphFX35e${%B_fHVXJn9$kK2%NFudl5X)=hwiD4Y zX8-BMOJPae!PMn*SMrR!N^9CRWGSft@w{bG2zg(*$m~cuvfQnaQS~FFmQrG=H8d8l zgzU0qXMwOpxH1;YePl1s`x95)#+F?mp0_ADlT`#y2h@(X*pC{W zMm@EtyZx^@rIPre;Yj!ckA(t&qtBcET216ls5 z1J9rt>A+jLaL+WKyJ1ZnB`Cn!&X}uL&vQ?T1OV%!CB=B-3Z?8>0j`Two z{r4cAw|F^_yA37NOL%Qu%4d*7YMDkX8)mI(hb(&(8nzdf*xih!u#l{a9Y5pdXk;;N z4&r%>UIn?^=v>`B1l8rEGDsq|tfH2$A4Vg~_SzcVg(cXHv8XGPwLWi)ciAFKp%RGa zEq;^845LO_C0<9qxFtv;wd^96@-GT_FC632ZheKtZ4YCKQzWx;rwa>lcU0+55YJnV zUnj@dfbMPZ7j^>cs#X06sU?tD7G+JwZ{rNcb{Z-yC7k7oE}1Rubl|K7s!KI<5YJmW z#*tZu>(Qp#=om}40!gHnXkxkOa0!30{Y_N8F~X9vm#NFqy<}GIQ@80U$a3j6i03US z8|j+(9eeQCf~GwLNu-u%#8NlL-W{DUSx)Dt2uqYZWBFc0o&fXrXYE3k;N2jew^(f< z=Zoe;XD?KjGy6djspSK)l(sni2wAl2Ygr46p9fm?!yt*&qDm}oYC}FEONGJ&J7LLVE!)WRMOvMZ&Bzkp48-#m$D^bx`k3z( zyq0vHGDsq|=n>1}x4-%$i(3mXM`4NE$JE907+Fu2UZ8^akBMjx;(3ePVRDR39~y$k z{lab_iPU0BEGn0}A4Qgh>pt%imeBo-WpX;XhpA__V=$^qg%gP9EmL}uw?0q1JbH(Y zv9qoqiPW-$Sd2Y$HX+M_V?hUm#lwrUTq19MDmAt7HM!Gi5YJnj{m9#e?)P=@+v~;` zKoY6NnOGu{F5>-TUafoRFD#{;GRd3(t^y7|ZG%@Sa|k6G0NG zeSVNSx?xaZ6bIy&21|CuFwt^Rh-u(Ye1%AH?&Pt1Zd6-@KX+-ov}g2#`c- zDIk`((jwg9y;H4ig0MLIFcyU~4tNgbW$(dJ$a0}Ki03Ve?Z`E`N4WAqbc{_I0Fp>8 z)%)OaztU+C{*uc`)2^w)QovafleS=s#hM?*$f7?3#PgP(Kgjz22!(;2kmc7LkVI-} zODy3F=G;Z+%MACs*}{@=h^b4!chY%(EVvJTyRhdli03Un3S@lgTT#0Yva}5ZNu-uO z#Ip919$ux^sjqjDu!J3EEar~np8t&Xd3#isoWUTTw``qD?qM1=h;c#|XG4%gY8gu` z&)=Fz1!T>hHHF6S&KJ<00vhf@uQAj`c|AfC587)YMQGUGp=MsqU@_HwJ-nZ@w;(3cs2|3167V6;+i4z(iiPRECENyq) z#^XyHJN3@O5_FWYWIrTh5Z+W_#JR=`Dd*UIoTs#hvNG(5z7XNh50A!i> z``c7u@d;ooJ|D z)t@aam7GQW_FQ~#`N5_35p?bct^x79Wq}qs_kZ+mHU-ti)Bz-sTC9ns$j)pJvgpm6 zzEoJU0-3sWze?s|63yB?LzV#xK|F7nPQR+@((4uOj_R=zB#~O|i6z7L2kwf#(l2YR zu*3#2mVk!j7;C3i`xUZi8-RG;GHL)BE4LRtx`67EWdxE)EuO@ZawGm8s*8iy>Mg<& ze1frDdr7_#QKG#12D1D*3gUUox!0uAORdJRaAawF3M7$QLWpJ9_w!$oMO%HEo3OZ@ zWGw0n$^6*CirIUS#d^3!RXlI8cuB6w+jNFlAAu$nxRc!8BosJHynaX$Bc9 zch??ggDmsEfq35HvXIOtzi#mT1+u)a=3Vt8q!wFZx%ubnF_p`nar#Z1yo zcvRPZ6tZ-;2l2c`DVLltbKT>|A`q zWR@dK(+MD+x0pJU^F{B{@3+YE)*K{}T7rlr*NGe&hmIGSp{vIJGKe3 zjEMm8yk*Z8GVT|s9A1qqy`n)9spUGc^r|1Gh??nMsP3t?c{<(Z#C-okD- z*s>gTdEV&`GD*5TSEp*=eHOKK>p7q<&u!Wu%RhB_t{hIfJX`5>dxN?>N9cicf7j(1 za)xwyF8kQ<4(jq;XaG|EpSnC_`(!pkU7i<>KuUR+=PgUJ&*Jmy%N5WOReJ(RB#x** zx;#G*`G6naHlA>skGeb`n}H<%vCDHqd>wDp<+;$xz3RC7v&(bIWmDXn*?+`~Aar2H zr+QQ+_?s@zi(|}irku5r?x@T2=}VC5-*$Qa@dMkWrgubLp2yyRwEx`YX|e_-<$YF)}k?`~~EcUL|gmtHi0Ydf!l&r(y+2^WVBWlgDqw zPv)=eEEIq8$XZdDG?r3kiXX5pjcQimeZ)wqzyz!mA#WDa{rs#qsQj0#Z z%q-8XgDfVNJ|l%C;UZHP)uH6vA2TQw-_hQf58`=CH#bt3;eL_$&Pmf2B#~Ms6U*9< zv%VvXrS-fC!V(tFSYB9>8Iy<_sQCmx(1pX+ke#*>I}% zCSi$*WGs%RWM9(O&7XEg$5_B+5YJl{4J2=ULi2mSM3&{ZK@zDYky!LE3^GHOPggIw z2usjq#u9gt?6e=9S=|d+EMh@CZ@JN&tn-_cmamU2zvDp?silxup3LsB09hXQz3nM1 zu2GC-pewm1Pj6ma16kG^fOy_=s}re9-6sw`kR@_7NFue==#R(!4%GtEkY&%I?ni{B zh_iGmCii0pif5=J%g;a%&s+SKli8o7CRVr$XJ9BuBDE+JOLl-7?!Ik1dB|yDNxZ_; zrRdls{5Y30UJdU@K4Auk=PmcF4&d?SPTViNSM>C`Ac@q{msrmIoM(XQ65e&?C1HuU z%2=#Vlkw$CofAcATEh9l#DmjZy zYtmIyl$E@NN6M=omZR2qckOe2C?WZL1~75}VXHPgr8FGnNE* zvVTn7ie`8}ztjv6&szeM$@uctsir@&xaWc-Qp*Km`8ji60kT+C7?cQ0@D0Y|8$f2U z&)2t2M8}wK&0$sXyrr^?-1D3D&e)7BrVT(6spTQDM0ZfFjx6ymnw7%hc9XGWJ|LDU ztGD52v3edLp10`SB=?qYZn@kL~6+;mI{+o?~$eZupaeX#rl5EvRQ*%$4(0*;qYQVTroMSif@4+RJpN395^>I!Ge57!gZNUkxQ>nK8JAp0GIGVJs1D%o1kVI-(Pb_N1^`nuc?4q5iu*AhOb!onmtm+#Zftt{6&+&}bwNCD z8BV`){%vx-JIJCk7$lKeVu{7s@HJjV&~NBkdtoW%EFHFxx=e3#_%pH`7!Bfi%gXNL zp1*pNQFx#D`V&DCspTcHG{}#>hR&DNRtvTZOWJ*=E?NoXp8wd;oA{nzzaxm}EkQrX zbu2_B9bc2P)j<-erHoh#`)tJbV=IUB-6t&3ag1eY0(ocMu-FWDtFJZ!@w}ys?o4q} z(Yy{i#=egSNu-uWdU)KQwQf9qez8_GIVLRr4;ah)<>dKgZ`%w!r{v!f#PgPEqsV)) zk?q}&BTJ9=Ac@qXPAm-*Zni?l*xTK$LWRZoA!9K=YK_nRCyC1jq4TBY0T9nyW*#Dz z9P=bRZ`tSwNFuciBNjieFZf+jYD|qQ!cxFl)~J*5W%0vBeUL@vA&BQK>pzfd^7>=* z8l$?DC4nST%XDJ#R&V$P)n#FR;$2}$h-c~&y@hmm&po7b30a&zf_UE2WCrOb9K2;K z?(~}d4J46TRufC71qqFjrFFNv$-)wrz*s!|$lUnNl`UQ%iexcp{clw&l=@=x7SEqjyhsJm*NKA`TXp|K#{-*rblb0FPOpESO9LfujN@gUXzsXJ=0 zS-o`B9hI02Qp&rdG(AXnl;6c|_)Tk_7a);1qW>+^g!KF8Cf97f9#G@ z4?KPWbw_NHbhjnEps5|Om7D)K7x}!eqb;2_?TX#5r zL|2KAZ$VD!RU(O8B~CulP($5OgFk{a|E)W!V9FkCboH#!b7u{4#mW5N73YucsOoxT zR?hRtDt$B(l=K1d{MGY&6nW#DF{K^uklck+CSn zlC`8S%vPBpOKU3-&s*AlCo@^<>N*kVc>OvTB#~O`>*FVlZ5`C`da|Lm)q4tyYZ7C5 zu1nTo#O>)g6j@?6f_UEYx)*s<;{EPp6tc|R29iiEortBI#bdm7;n^_z0m4$mS!xE7 z^X2+fcl`Fc&R7u7Tgso1wF`cSv~Wk!`^g}Q)H0A-0(K2Y$5?Tb+lIoDn9S6r$5Zkq zK6lC3m%fkwgL~3y%mhsbebwQSVgSE?qrINEWb|d$(dL{3i zv1JU1=PlVZ5(#`a7~L zoC)H2OZZ&!WL|sDHEbC)A0&}lqKM_g+bj!ImnLtVb_h%CbH;L4g^Vx#LW?>gi{}Rr z&s$msk{Ocs3(iqF4<7Cg3a^qq6X|}g9i03T_ zTaaS7jXd0tpbILmZ>GQ+s9@zu+yF2SK7p0`}q zC+EIOz*Kxqwu=Bsq?UHX5?wl>IkH@8b1Yg|Qqq~abjftVGmMc#mGFKTF`Ym>Z@E>S zbb9$tJcQR@n)d`rq!wLbu^-$t6&UgfgLvLjYDCVL zHc#u}{guB|Gq3s)Qp1%6ai!8T1yq*h-Uj}1QH6=5@j#iiP9^R_HAfC6RTah;< z&w8BPfX@A|0U(LgvVd4t4xEkeVHA~b<_e2bCSx&MK;Du@PrTC`oiCPMKs;{=_a);? zRL_j=$kMkLNFud35lg^)0xC-Hb+kVI;UCzeeECm11%iT8#&w|c)eSf#pAfC6>Qz6%}9J@IV z=otGP29iiEKZ#|yzMV47-K|F8qyh!eoCwca{fh+-+KoY6t z6|vl|Y5oRTiZx6Fg(dSXQ zgYmfktk};OSzhO?zb-66g^Z=VoO~Vl+PQo9+a2SJKs;}m^C}$QTV5HVQ3KVbSt&>& zwR9&ITT?ZBKlban#Y16neaBc%HYY1o=T=*aS9gre0`a^>J(oO--AEW7k1X{HKoY5C zII%Q6P#T8nGO}DXRalBROZ{tPb^4@6w)lQ5qDIH6c;0f~hm8Bb7e=*4mSzn=5~;CGkB|mx3PTnjG&L))$>G22DXcZ*kEi*RetCx8ofUY}$e(Qi~n2 zRC`j~0a?aQy7Ea_B8nJG$dIXc+_z}j3-3uBI~~OHmV;Z#z2$|_LHAHy>}G=`Qp+A< z>7p}u6>4_)`h43jVTt{p9a889_qqrJ{%#G6!|xlurO@{#zGLg17nvboET|9#?PoU$b&P#mv=n{!Y^GzxYOn zFf(t0lXUnZOdNu>YFN%FsZ9(GwBS*Cx9Y9TC@oW-AZ6+Foc zOhuNFKS4ZiiS9!>B%0s9z64p`)xJ{oBczs_#Gw{&wSby1B>#v^p05lAAnq!CNXx=y9YvUAgREn$iM$XLENC+qyiBz^Hk zmf9{Lp0^ksBwaXNqRp?My1d;Bl1MF|h~=2WARA<9d0lsqumqPdmV1F@T()lTMPUej(rA|!jkftsY`3Rqx-rRMQc%AEWU$y z-ZER0e9ggZPg6XTwXB9|)sK)`%!s9Sgh4}Oxz^BjrLaVOVJs>0$y3w4Z$XO4^0+04 z=Pid{kn5P6@frM;jimMxXfqBSMvnz<_ zEj}7#eEE3o?O=57fA0m7NG&^vrN%dtOk_E5sp)QEar(+wB0rL@f>S%ywn;@RI ztbI<-m)e9JdXqMyyaR? zGGo%Tt70LtEKLMSq?Rkh@+wuc7OG3F4*4g9C9aIA%iC9EhwA*u1Ndt}y6PaFx9FZB z$5@5pWxRJ(ycS3zwImbEnCl*mkj3Y_ak#LAeq$`X50Z6Do3jV#p}O3yF{Ub>w`jg4 z<9<^0CbLmp4k~~oQpNBa!o|6gqnta(2#PgPMbjGBj z>)|=bvTg@RBDK^SipTxtdD`K~q7v`%NLWfa%eK{)_&TQZ?p!}qmy~B9p0}i)Co?8j zyo2Wc~UZbZ|BC3mDFo@?Zz55@;&!~Cd z9;hNqzc7$QYMDSR$Mxgzei+5GCzc9}{}0Bp{x#XBBz~UBF=WwN3gUT7K`I&dCtI!D zh%7pGAc@qnkXTyRyo6`uCr%2l=JwY!vU4S4iTOtMC|Xc=&_ra(pAO=A%S>~!y5ruj z^wG%DXbwmswQM1lW%ivrq4T9r#TG?jDc~%o8f0DUIen!T$g;KO)~a~kG9sU>i*>5# zkd7?64L}m9}OFNpF>T$kHGJ#PgPhE6MoMIJ)U& zWN8%*l1MG##M1hv>oZiBX`TnwgeB}3V|ngPo+~rDTnk2)lfEFHw>+Fp?)ghyRLhX1 zC=eu(S`vswaqWk4WC`0_)>l}(elwPs4dhvD{@}Fn$TG(V#PgO7ZOOgm_G{PixnIX0 zB#~O)5KG?3ICEra`XY6hu#|I_O*&*AM(b{kI><637sT_HBP+?9#jA(c4MBBLeGigI zEx(B6K=Or|$nvi4$MM3F>D2Jg6?dzalMcMa@n*e{rFRn$&s*xfBgfeKTPrk>rF|Qa zL~3befbU`AHizJ^&{{V5Vks;!TNsP|)yeoC=9#H6-uF7IIf&;iL-olu*=**q$;i@8 z1tgJLdJ{{Ri<3v9x`cK3vQSupwlbF4ndI$tt64wJqq-<-f_UDd*oUm|*NNLP09kVT zfh1CkF|kB6Xyt+|TM|yK5f;~NjOD=^a=z$BtK+>2EFiFtdfdJSfM*Js6G{}BXOYY^A(`iSjzq~EB1$TuNQ0uN6h%oA znotT&2&JMBMU*j=y!W29-qqgQKJLB0KVa|m{e1U%rt|D`9$^XF%UG5*CG(cI;*9YZ zb}qLC@w_F=oP34Wb>Y%%RF|3EKoY6t9I+g1aw-d1F5MV^Mp)cz7)vh;va-wj!R~xi zm%wBY&s&ZckbC~Pwzrxh%khUGiPUn7SX!9AtAi{fTlEVUmU7NgewU1~uYF!))n$9ncWJ^BXUkZ&YLj#F@x-{*$Wq)N z#PgPx2gvij#^^41pZHcIKoY5?(Qtg-?;d`B2eM3vx{xC*0d|b#t*<%0?k9Gzx_~Ti zZ-IE;V!DVtSC&paL*|oHKoY5?GqD`~Tzeg|#4hrGBP;!kNjg`etUKV&l`7UTwNU5EuqhzRKXtBnnUT(w)uY{aqt2CAvq8E)b*?TEl=99M`{AT>#rtUg!)Qe9F$Iali2AK_Wyl87 zxw3Z7MOW0hlDrut`H!6|^^9IUN1ZDPV-hOI)$g4vx71eSev=E67QaFRb4aV(l?ndP zx#Fj}7H7)+cBDJ%Txrn(Wcs(AE5H4~&kHnmqt2BsJwV!j?p(QJJ`{JZ#3fwML!B!c zx**}d>RcH#pY-Y8wsx>VSJhR6K~C9O!g4Y>OPp)?67L_gXcS2E-#S;+OZ4zMKb7EN z72=Fj!ks;1p4;Oqfkv0X_)YxW2oTSoJ)3O3jPEv_HpJjQ*($LhiR^z-m{0zfL1wpW zpew=tvw2Df{+?k>-_KkL;&+jCemX;)zo6qbFAT)<7VV2M|;(e;Z-(Jiu7q4IgLvMe+&&zSvBAsRc0!hQ_dpV<#h+LP z4LfgvEYF8t>LDza2N_F0FLLM9?eeNN$ns?@i03V%Rmc_nE8$EQ@r^Oj@HNsqi?i?|-h(!(Dlky<_yOYA6n zXEes#A5EGrEJ2QpWxy8lX0fklSRk_0%LVbgWnwAWZF}LJ=EsqxWdTScwbU7bpER;; z?qSQ+<6{;Ji=z``S!GC`nmRtXjh_H71%P*xq&s#Kil53@1^IUvRK6C>lky=cN#l>spP-M|dRXr>$ z-p-6=m=by8ThjZ>c4P_90P(!V@H}}kzxZ=uDXNQaE=VG^*b>Xu*@N)DafcMTpAZ)7 zqm0EdkyvV`RmGhv%alMoZ|O3Wba`5sd7C54YGsf_YB@(ZyXAF|y22$D!GSBWKRq8eV~YptvvE-XndOkGy( znvU=J7dBFvhb&`SfOy{0LyytfYEPjJ!76g(=Ek(q#Z%P#2|ML6p>B++4?8aEe4Vi=M zVo+m!4`kW5AH?&P``<}j0wX5`qq>+K0ZF8m8YA&_KmUDQ{FZcU+PF+%`NmnM4kh== z>Ge7-M3(GK5YJoMy(Jw*y#lx4J-t@sfh1Ck2C;;XF~&QG7v8(|QdqLxnYwJfO1eFo z*<5u(mMvKzp0^~blegDahi>Yjy3~6Cl1MFmiRD}^+iB=L=C|@miLk^TV=OPvk^3>H zMon?IdP_49&s)k)l6#m9!7nc$OEq(lL~5B#EZP271|dtW0e8!V#n*$eyw)Xk*%4h^ z3)N-END$9kws?}dXs>F9zeqJ`JV+w77!gZhSX=yD`DSS6+6VuhyRh|SET-qlz2(?_ zt?iKI>{t-bTinaZTT+t|E^(+X`zM1WQp+A%~okrHHdcHzn(p1LqCH zZ?C&0gLvLjzw8)(-nV$Ju@G4j9)ToMiwChxuzk1-SxT!KX$ec(ai%Wi%A{lUNy}yK zs4h8mK|F6sc|`7E2F`nhJBn_ofFx2&7_szFxVHgW;%ZFnE-Yau7|Z0>awjFNFuc)6U(w?m+`&}=T7w-AS`ZPjHTiR*+XL2!kezh zQY#3=^Oje(66%ZQD-xI4<=z+ydNDd#Lt+K_z~=Qz~C`(C@a zfOy_Awg#EEto9{W9a-9(07;~lU&QkDUER@Wj4h3MFjZJ`y_vf7vzd+WVU`q3!6mk z=PijLq|`ugWCh}R zOL{VS-mjUuwFFrO09o~q5Fj;-IuA$;$*TXv0v2veaJFM8^rUL1$8gr z>;8oRod(F#w?9ZCwHzgu*)Ny&Mq{jrhPjKdMEEh50msPR9^((+k42Wm4j`VlBsM44 z7lq~V*2t3310<1J&J#=UNISd_*0tIjP6>3c-}J2jGV`; z&AjR(OXubwiPVxnEH!IiRzP(bnKw6BSS<$*RM@17Eb8MxJa6eVk<9f4Ea-q& zs+3O#Nu-wV#G;#b27i&tdUsHwup|aBbtwuY-M2M5AH~PZ3wsdHTV^aL&o2k`=Uqi( zEdDS^BDE-w#@GFw9%u1>OXbf`J`|RqGmNFy5MpV4Dy%xH%idxT&s#=BlHNG)B6Ww_<|P&CG@uT*&h z^CXZ&Y8gc=_v|)4K}{&#?l=7`EKmQk{GmH)wkBEU5r0ha1nQ3R%mSGt-BA}>x5wR4 zF)`JZPP`J55k`RAw!Z z0m_qTjotl=IDSPgA~TSP&YRJtjA=Oz?;9sLWcMa3)vdCznxo z)WDe_)4%PG`t1kmD4DfE-BB$Tg0%nK9rYq!4|hkktE$-(bw^n*2MPaGca**pxpO;L zcWe!Gme5)Qa>~vUmb1uNV(hR<{n5TYx7UL-|E)V}+J$Px= zeC@Fx&zK~w&B9Mjhkk%~-ZEk_=_tC?uMw`x@~USme}vQ$KrHUv3TC6agr2D1MOduQ zGnTow&jazi#l?zr;cVGvKO9+NmVqQv%MfBI>T8O>6c*xSv_n|3Lz%ivjw07e z?;DHoJ|)H;AfC5qt{}4v{vApbkmZgKNFue&B9@+d{&;^fpVx-_g(dbPW0`xC%=|7( zeTcg}oj!ng-f~@;%*b2yKZf7NUHT4^NG+Z%p3Azj${a zf0ahb8pQLK=>BAkCFC8LhsM~R{UC|da*9|C8?Ana>hko}vM^yO;w+X8NJr7tL$>y) zE{#h-Ja0MZbrjG1o~zmoA1`gbfh1DPHDU?P_1T6j##TPpg(WSVsY~=C(uGr~KN0UC z@x%ed^Oh^0$eQ@$_W~?XT{by`BvMN{u_(6ewgy?I9kEFfmaxl=B_)vDkG%??_8!$G z%@f4)mSZ=FC1d`zuE?U{2a-rF?}){Ar7pg{c&nLZ35#0E-i@hJI%iZQSc>kEdxgegm z1YaRC%i=PVP*j)xN+6!M zcx92U0t4-_v8XPF${>l<(vMgsHBf7VEMxs2RdxLP89DAMW9dJU%qvD8(!gJC-FOMa z^Om#4Gx1wob*5vsl|#|{LVZ≥9aW_A*m63#O0CVAs~L%qdgWErLn;(5!D z6!I+QTurApvQ+O6l1MF{#G+==^a--qw9eBNmh>p5E*>dl9%lEH)rQED+ZDv~md%@k z@pV7Lye9s_&V@c8iPUn5SiUOsh((r%!}3Q8OGGqdSzWjXkFh6x3-SIj9#I;xBRJP^-Yj60HJ z|5%T+&5hh#Nc}sdeXBob?3>xH78P8khHX?6) zySewShsK!QXplr|=|wDCa!kLXx^#~pX)P>4HyFz-BNJSg`(qPbP+fi{gLvLjQ9j^qh5IVV3Ab*{`F2r@}JS2ljuz?~~|-FjQ1&Xsi| zL6(2&Tv@-0bgn#FylgA#Tu~ek(*3D(<@^KExiahFt7oWl<^B|q>VN87NghzOChA;i zI0vMZcdi^%B%Ld1b))c`)*XvMA~B+V>s(2GN;+5SI|h}Z&XxKrL6ZO2xzbUiZd=s3 z@~&R<%5n93=gP?tbKGx|?N+S^8ko9wI#ee3L+46K`3{_^BH17ib*_9(1DXD9=gMzC z(7tp_5b9j%mkHATbLWb7_+;F1sP-7cdM^P7L!sC&s&;2AWwkfU6dOm%j91miPW-?Sd=f%!)xRG z&kt@dEZKLNx`cOJgr5MWxmNPLRiPW-%Sl)S#2}H+> zWy2Z0g(WtLv4onF{V-g2TxDv>z#J1 zH9=TxQy5F|PO?YQ9plOGkfnwXi03Wt-^ebtrV~Ek@A!tC1xciq8^n^A^cK(jcI>7) zS6GTTORH?s?V)n36t5?H)EC6_mVF1wx~0CGU;RLJNgM`}NG(~!k}|IaUQ_b+ptqs0 zq@^--@ohq$00-fVwTqQMLTag}kMD=-R%w!rj{R;$ zu9m{$c8{^tT}pOBR#Weih3fL)3y9|}Evk^t$S<#a*Pt;rwgM!PTG|uKqKlR_k;Tow z{{dks=Pb&D$bPP~p5(PcmdFqg&s%!gk!$6Q>O1i|@&gediPSQbSbpt$-XB@~_GY*X zOYVK9E;)6`@nTwb{syvy9R%^bWoK=2Ki1ak7T%jM&;=xsT4obVgkF8T3;MIZhNp!k z?g3-z6F|O7(_L#;39=j;4&r%B>{D{=D?00SLt|{qIFLkYF(VfHp4u%@T~00X3Kf=s zbjISKLXMZ=hQ_#Cy}1#H=PmP+Nf&>UYKz+;OUgQsL~5}omg>8P-9whiFXLl`#o-}i zss4&|dpuOBj^C2@e+S}u%Yl1jjD7iQ&n;xp`~s3lEk4Aeaz|wXvW#riFiBWSI7`MN z(&<&LL3g~{cHiD0p0_k#M($yp4tu&H%c?;jiPREFEWM-J>tIXX-V9+$f5g{$`_W1$Bd;tXk*r-*Xo)Ig9y<9r%7s zLwE3JWXWCy;(5z=x(`;{@^BYq>1zg(NG<(|CGJxW?($qvk={&L5}z=2i8w>vlHPcI z;S{n|tM#xlp0^~fC-*S>6TNZRC*B5*1sanX=Wgkc)wHOnNW#cV}k!4g?ub#r*A1FrK|xhS&EW zDF^Yq#kM`!TRp6=MwEv249T=3&<74#s=o zw0;fZd5cdbc@rNzr^a?W>Kr*_czG0T^l5kS_+8e?A$uID>^7n(?nRTa~aF39CAOVQ&kCfdyGf{ z@x0~mX)@0r-EhHTWcl(CB#~OG7~t!E$Ar8`s4o6hE4B$s0cX)@HVWgG$)qKfTnYi<@An)`-CMakEzQ<6*9)+XS?C9wQw5{ z&sz%Xk}>A}VA@Somw^XC5~-y(u}mGh+7nssexC0vEFsSs%l<6#{NgZTTqHVPd@Dda zZ>b+j##sJ}AC|~6NpV5tkC0jnh{Yy+9^Q|9qUSAxm>(5YJnl z&^^3&*V3AUEVtK#BvQ*xV(}XFEDTwa&$(U^mh6{IU8>h2t8HsW_c_YW`GMY+o~$`CrJhn2znK70BW-5XAGAon6TD{?FIm zxcgRnBuFB)JSUct>dy61ib`&``9Z-fkifmdXR*tLRJBqZUEpZ=E`lZr=XkePX1_}PqQIur92WJ{kQ&kal z6xqE4nf`4@(QiMnMq3pV)KN6^3rPFV9Yx3c%)=c;y*f5GM;%4?%Rs_^)lro8;xX+Z7?+|KLyIolZ*;&H!JvmFbng3{rI*Q_IgEarGqv(^pGu}b8W5i@UTzJb(6l+@D<4BJJX$(3L=|KS(0`Ulhv7|DrT*WdynsluWqvOIXqh zm@7diZF0`6o%j{6S5}w};(5yg3v##7e?=I6YD!uJl1MEo6Yx`0!@?>QjkPy(Ib{V1Fx+>7WWlK?S!SA zv*=WiJExKDES@4uZ&MJ@Tc&&`Z{yBAINupr;F#vqB*;zTStYE3PX<)-I&ePIcB$5@8g zk#!gaJtpF>Y;-vd;(5!PG%}05`=o6dvOGEml1MFQiKTEuExcYidC0?=!s76ru_)fQ z#J53T3^va}$4mAY5YJn@^hjNL_8f@66aQcmNFue|B$gdDh6kg%tTC^%Ojt@d%fta> zU!VR@Vwz&h9uUu4MtvqTCRI9Jz`Hu{vj<6}mTY2q`*}Hj0vz_Fx~Z_F7c+I~=S#lc zIOA3Ks>t&C1&HS@$LM+NQBVi{nlHrk40lF@j4X~lk`D`u_IVX2mId&qd%jRYviPSQOSf0&^^gx!DT1NtfCGitemjF78z4%N? zA7rWM4B~l9aSl0;y%-eL3t4{l0!gHn^~BOGOBL_8)VAK12w@5O%vgGyAkQz4qHXYf zvPM%7&szp1k$X$Sf&CJZrMM+XBDL%%mTLoZ&Y-$%>aQ6uERJ6oOLg@F_*xljGZoL4 zo)`(@dCTEXWIv4Awbk(M^i>Q%5~;x)-0ETBvQ*GVu>p%t%58MU8lbhmdH}ZvMZmw!8%#5uQsa7nUf%% zx7?mcEYI>6G(>f|e+DFxT8fEf&)pvQ9wud~^H*W<{>E6cr;{)2^cWSt2w4Wq2l2ec z@&p-UzZ^Q?u7YvPK@zE@)+Bu0|DNF)jq0*@adcJZzn}N5zcZGh)yZ?E+oUP8kR{6q z#PgPZRY_fHWmFlC>axTPB#~NL6U+P>!#5*K^P4ISgr$J9bUjD*!fAV?H|~mVm;~Z^ z%X@n=Bmc{FoGP;1d;pS2Edz)p;^iE?Ub%{PdUIh(`oYviC70a8ysXH@onFNQKs;~x z>`yFN+NYh6<>CmCL~5BvERzSs;5}Af-HB>1EFoo#<>Q!b`1*2WKwnoh#uOidc;1pU zj=Uwkx4FwVWLb~{l1MFUh~>TA0(?KV>`J@d!s1-cSf1Y}&-**Z&l!g-j*X_%ji>S9x+e`P#xF^eO6 z;3X#b;T1${YJ((F%VlC|9IjpsS?toU&lZ;03dXXn%QSqv3~ZV{0$HxLyIC2}Te|He z`z{2PAHrXut=|(Qky_G-rBhzm0%V!-^zL$D@%_bEE*>TKV`)#C9z+(O*C3v^EO<>i zy?WSM7Z`!T1rr|uw2UJ8ikEkWu* z_+5^^?Nz+z%FRb0iPTbcGQRE~G3#9i)#dj3i59|A#90jT$oyENvo{tX%bFb^p11U@ zO3ukz6Hfd>$G-YrkVI-xCl=%2iy6q`Zn(luSkiVgb+O$}z7E`MiO(Qp`D_Z}c}vbX zGS{a+&Z!VtT(*EDQi~3;bSQLqfGi4|)Q<{F*dE66X83M=?Dy}O8G&~E5?$3kjz^ic(D_Yv2o)-JZ}kfBJcPX z*fgnyEcK>@zKoY5C7qJw$S{y}|R!cM@geBLSsY{6|x#xe=`|MaW#SYH#b z>U?bml1MFX#8Sh;LmgR~)pfloEOC1oOPVd|^!l0Sfjd@jsepLiGSissM}FQuFJE~w3$nsC!QR_yK?x_0q)$>qy)TUV=-JiOnl1G#7sF#fnK1SV9#}v?x;~Z zAk)9?j{5Bf)*ZHL66%gxH3+2r=kBOItqgH@RK+Ia2-F>=Iua!OSKU!BKIY@I#K(?4 z+2}0MK_BFloh4#R$XTMxtmijSchrN)AkBa4j!IN*gx8X;ZT_r8oN-FHv**jrWY3jO zYJ1bsSl`+X#PerQ(?;ZO!*6EAZS=n=b_Yph|BFI7`Cq#JioyGoSR^Hs3ro5!b0z3F zo7|m`vU~mjE1Ajqa;ys8Q|nm{NFuc~nS!61YL9+qi|#f8H_fYc^zYfy z2s_5|(Ve{Uy}Ie?PGrg33F3K6t1z-pNtN5rPolb{+JGcdOLt@E%#BGe8olWel-2d3kOCs>|KRPMX4E zxu3BdyF_MxCU;#}hAb&9Ks;}m-GHnmb(~)l_sF}n1xciq#l+&)s__eC8U1)b7h!qH zSu|b^!*%i6p_qXzV{<_~Z#iU3>e6VR%K%gtl>(4NYB494^nLNTJ8Ha!xvsDzIxuyq zcZIz5X;Awk{vNr02N2I&MtmoA>E!ho&t(1V4w6VMhl#}~_$~hG*~K>MBZMXB0As0X zO2(L`?HFuH?+W61ONB9cQ{wqm#~0P*k`72BwFDB&^XJ!_pfT2Gaq$FUaXiRa4D!g{ z9?hfG@qSBwt3f<(2`eIXIk_tZA1}|>gCtT*9I>qMh$ui7-8XaR2+L>Aa{4K`C;yf@ zr5GJAM?Zmh-qP~}vAnBm+6>jDMj1#Vwd4?s`V3usyeK|WULh7lyJ*r#tNEY^n^OVtyk)63jHe>$>+ zIf8iJ($a~{{u~T5YKrQz(G4V#T80zL$YpQwtlYYzxz55;z*!U@lU<#=%__4)mJKIB zJa4&ugv>B1kM6q!Sr(iINu-v!#A5Mu?_gAyu%LzB!jg1^sY|b7G9%x*+aO$*R^cF? zx8%PhUkq_r@D9Hvtsf1NNG%(P#V^lA16j;|g$4>sh%;je(j;%MFKYM3Up;erf3Y&2 zxA@qRdw!Q4d3a5{(l?MqYH=Wz^D}j;BTLihd6$L7`6y#Kyr0ZuEjT^|&pX&8fOy`L zFr0LG8bw_HiN=^}8b~6w_z}z51@E(vCHI=k4Pp7lS!Q%4`})|gABsD@)NX-z-g3*8 z+*^L~u}wf0|74IvYKbP60Tb(&Am?{u*_9en%>&{qw>y!JjvZ=cG%X&s}AfC5$zD@RB zxR}2O?^R%t43bDKTEsH=NFTiFW_8YwGGQs=ENfnq{bN2}HF$*T67mYf^OnhKWbVST zn#nv=mx=E|5~*b%u{gbIF$vXW%-zK`UH+bjNjt{W#X5@YQ8d^wR0CN~rGt3hQui*o z=MQr@ho4_GvOyB5Wje8J%5FX#S#J1dD+^1Q2VGPS#14P@z4S8J%S#GPO)+x9KS^JC%NeQ<}wg(e`Lx6E5gER7bH;m$|} z4Uj}?xkoH1b9U}PmfBM^^o1qBi?JwZk?yFk58hyl`CJgsTNXJG%jwtqcA&b%ECorV zmN&#w?ow!jEYk`XOcxdhZ^mNNoZOEkoS%YMb`Aan;(5zf3vy2WF=UP+vNSIPNu-u) z)A4n`_JKUS_wArnCl(4z31>;^N9OvT&a}Ia&SUk~gLvN3If6V_dIYGoM3&5LAc@q{ zf>;)n&FYKll9O7#N?6iQGIen^BwssKzB>w!v8GQzJa5sjPoBl%m6ZaJrTTM_L~7|n zEE=y{<9A7u4-DNPED@&|OVD)kTp4AXrjP0}+ylh(mR0KH`ckKnQv|Zi@BvArmWjj? zz5X=r@J^ZTY9TD1K8$64F1fy(*3SHjEMYG|Ja2KyCzkb!(Mrg2M#Zi2M@TIziRJtL z&-nSJ@Sduzuvq#smPuF1+(qh3Pux+|B@4v!mLbvP-tzqDDR?iOZ7)C)sl}35o((^V z=f_NP4;>bkmz>3^1({D?(V)#mR2RKBAfC7M(#pKel=`i03VjVAc@rSl34WAPbwly zyLl;*!t$B3be%`$#*=Fv>VfKVcR7gXEhaz6eezAW?FGp4c@0P+wfrQOhtu8hv)Glq zhqr_!BY>%kegGL`ie0vLMV54ts2IAkY(v-kVI-}Is@Os^sI2Ni|XRt ztIs`Qi9Ewtp4=jP6CT_Wi_c?Y27!3qqOM8ma{I>d@yPOZBuFB)^dOemuj}Et@%h0% zS;FFdma*I`CHG^su0D1_b;*tZ@w}zw7V`e3!_9N*oFji4pZ%H(_KX=_b6QqEQ!h6HXihlKjVR!k*oiRiJLd14SU&@7;tE zcDA?&u&T4M9vYZ-Vexcenr>XOX3_FZ>z8aaGBx@C&;8wBX^%6lRQn!-x(RPy1)2VB zH{owTP}3y%HR>kRx(U+$b2s7P#%pjl;ZK{j<54%^quU_izv?C&QTheH5tzSj$5M2b zNV^AeO3xBg$yp-0bdW3RCY+E7()_n>!U^XLP0*E~!tK=?amKL@WX_(+m&pv{o?be5 zU2Kr71t$AaH zp8(q!gLvLDF`cYgJgU;U9Mw2#JxC(8Y$KN8uLnIumgh&Je+f%c5L1`TW@J5?dsB_0 z=p3~0IEd#hsw!lklAMBm&d74Z4lwALwkq@!Pl-LqQU$C7xJ(hH6`*G3L51MN?S5ahAgEWL9oc)i2Xf zT}EsJ@w}z#D>BBedmdSi#+c3?kVI;EN-P!zcgG-0HJ1UMg(W+fsY|>cIgfqSRjh+7 zU%Ws(Z@GJtti7JMdL4cOyzdW^NG+wrvZDBbJ+kaQYN;bEu_275c?YsPy}J8iya(R* z6Cj?qTx~|i*tuP{)lprV`-3D>OT(G?NyB@-!vtih@_oWkVe!4dSagq&S?p`J{nb%j z6sk?EjOQ&)R*>1Bhxz{NkflK#kVI-#%we;u&_sicF)y2;TB#~M+5=+1`d)!sf>s-MqVF?RkEa$qA{bTgKd*Wk%OC*TrEk8Sx znXKXaOYyASzBrIXYB@kGwn^>%&={*Trq>2xal6D=dLJam{=3M4m*{x8Fd4-2mPKR9 zb>BC5>l##-N3%f^spT}WEcp3+9WRlpPNXOWb9~@~Y(=T$euf;rM>+QwtE!TQ(NDU4q+zBvQ)UI1^c_SBwZ1mJ-e~{R^>_bSrX0bs17~WMw>WQFkKeoG zD^XpVH2_JZmO;dlo6--j>^in>Mx?N$UuEjDBa3uJui9>f>+;6ZbL zNGQ4s|L|B4i7>jjPGR9i^m*YK&ZC8MJ-jaKiSa#*bhN3a%-1BJVkC0m45{qr2 z#t3x0oH^O%yRbOMGL{4@vggWlQ*FEoS}hR7^Og~p$lS#btFq_F@+K4{ky@(H#@GF} zBXr7qltRZ9UbH67a(XpS|0mSo`FgiCrX|MAuWT~eEl1MGO#Im!> zlR{LNAhR}&g(dO^WBD?l%)xfF>!FA&owkE`-m?Eac^13A;*C49Y_I}Jq?Sp|N76U0%S2#2l2ecsSlawzoHX?R}ekd0!gHnRm75|_yfOx zSrPc7jj&k9F&4K_GN^WOU1ll-l#4L z)4jV2O95wD)ouoU7E2G0h(?yO1t6ZcEVm%fVy>rW;OG5jh9HU5a*SB2PP%>qS={=x z?;|Wp@l0L9-fh8_*26k?LYC!2K|F8yp-T2z^WIjn8`b6R7?4D2xkxOg@z?R(Md8j# zgM}sJ7GrUWBJ(hc&zcz_%RqY&&szd=$n$>bgOvl3rTSrzL~2PQmc$El@jJdG_co)2 z#W{hognVCs>!M$K$x~!;^ab&}rF;;%=eKoJH$;}Y=RgvvC7)OZWmn*DcW8&HP7s!F zoTb-9a?d}c)A)VJG9?DY^Ola~q#Y+0=qJB1}JiK$C9 z6S9+w-J!c3kY%Ydi03Wc$CA2)>dmi*Ec=>+BvOkru^hS=b_>-dW~8Z&u!JQumfQrg zm#6)kymVx#F%iV`7F88;PPREWqaCuWodJ?aE$4`3&is?DkR@(-&Ou>uOJOX>+mY-3 zbERWR$ns%4i03U&OUV2E!NbhQAj`gKAc@p+i&*lNeeiA$k7Em5gr%IbB*!kt$4kM3 z>G*qn_DLX~w^-dF_xvBX zHbiwXR0i?9Wz80HKcdD0`4&u$6PYTiW^1N zM|Bx<7sT_Hz!9V_+d2f{Zx0$j07;~l&cyPn-P9;#nffXDvamSZXDn@Qle#>&?R_3u zZk2<0-ZJ_Mv9yoK$2*0UR!^z?5mL)YVyURH4`267_Wz6#mJ-f#{xq4pII^_?{+8O1 zqadEQY#Bl3`X)?rPex;G;&G5fYMDGn z#PgO`Dx`ESqeg;b8K-#hR3n>yfrdDVK0tkJ+6n|rY`!5{YV zJ-zH8&g6Nlb|~t+*DU~<{%z;|Z$B{k@InRDdEfN|Nc+#7_h|*2ap!#%-{DhG=e_rL zknmr1-hUfab+~>VrP@k0H!NGSX36>u!NILYqqFy>s$rG?qVz1`PtFpW8~wvk=lx(M zkmkR2-n)g)OhRYRJdcFRpVV4OQ>pqI(+wN`e<8)Ivt;JCq<0|x3T+iX5YL}I4Zers zJEuW)ARAGg4RzIE(2~a_nDP?Teqx-EM$*-cn@(=~`PK zUiTWZ9KHjRNG(^1WpTAelab}?I~4!$_&vao4dB#}k?;+RyWe=z0^Vq7kAfC5O(IRy@>g%}> z)um%skVI-JB9`$1%BjebS&%(nSe$biOWe-w__|-G$6il#yuADh;(1Hj5wgc>(ujn! z$kJUQr}9TgEj8xj>;7w%FKv+JTy*CZ!t#x?xcHDxueIJWHIU_X0*L1=8#a=etc-i{ z?U6-44J46TG>BzuOcCCrNNwi$wZf8}$J9kPwKdCR#rWL9qIm_ppO zmgxeLNG(RhvSLjlUO9YcotBNT*uG>evFDcHW8dOJ{$x~_w%b8GZ+YiLEc+&%$8WEd ztw0i~We>4DYv1`08e`WM>FP;Us2uoT%Qy1rEZZ1h1|Z909T3l3wlpK>vEi-OZblZD!61p$vW!?P zbvEGnvHSZEWe7_NXSu(MtWbShBQ+LTv|T_vZ)rJ&yd|~!)e^7nh&ch0NG&^v#bH>f zIjW0#C3u_i zcVV$CW-Nwl$xbesvrgkxosBlARL1j`q>h*Hd8~ej`7vbKx&tJUT5^eHQsKfc$g+29 zhbqVZerNHLvlzT2?_ZjFHgrUFslNxr^A-mKGREdS>bD+Q?%07OQpA79B6E4tMP= zEE%7ex&(}!gzv{9LpJ9ii@OJi=Pd`*kK*&#Wsi0E9iN5|NFudtCYJuk$-AVLix&10 zmdMYH#pgP?=Xb2#56|^=Faq(sWqRm&Y}x+PVH2uL$U2ZjYB@wKmQ!B`p}N#;V=_Qk zyuUD(?mfvo%yX|jcpj#HR}jxz!oA3<8@tF0cz&#k4oD)k1Q1Jkqz;}NcfVpgLRhT7 zGM2=jWY7CY4r6AaF=lcD#PgO#ugKTuA8qWn9o5C-4oD)kTql+(m(F)bb(y~>5Yr#+?h7;}zXI4}y5!V$zJPAZj09{sURwxPT;5%O_${$TG*P)cX2YohK|I-x!N^ z?Ya0o=HwZ63e}}WD2V4RyF zSQhjoJ0P4mF=+;}^lAy>dCS>Cvbv+ykO6oVK~6i6L~3bAEK&PT;r+-bB`-7)Qk$1Q3-8}=)&j)ymZHVvoNTB!s23Vz7u$j)Qp+r2iB>#PJ?03`spx3t4OuKoY5C9kEQiz74O0+%zlGN?3f$8B2LN z*_9^5e5WxQV;>iTc-~^&iPWXE^9(%CKg0ZckxJt6=J~eK0xp zKgHy}K$dNJAfC6_Jt5cq6%UMOBTKhJkVI-pCzgrp#tc9fy?K{>g(d74V~L9K~CM?cvOLVTs$rShiLt_sM-0 z<-9@`-)A76w;ZRt2IxIC4|$3)t7-JQi~O_j2t{%5iL6Da`Jtau(!Q_AL4hEXWHr>u3N^eMs?^aT8?A;~7Y4HY(`u;JiDWg@`t~!e&}yhIQ6Sx) zRzp3qBdej_bUQl-t%hoJ3#9s=RzsystzUpvL-n}_QvQe4P~SRwZWpN@s>6Uc>a_hG| z3I4Des>SrfIMa_qH*cfWP&IylO#gN@)Nem9=<2ql@FD|e_johsIddDhH|=L zyc4a4YF7&+{8y`?h8Zd0)lmK~rmR3`ye$ntPT5&vm<>5g>>dBa53PnWX$sQ(x7ARK z=Y7R9j5Dmu^TZjaggbkN-X>S*si8yap{rV$6Nu-}o}a?WZ0Y^N+tbkhlIRYS$o>}v zPx8O)&wP3kT?v*IwR|Hi>3f+g!I)9D_^Ot^{XM=yyZ!?4yyei%(R$Z{%8t(wQ*Z%Hj}8OxdLtEt{5+Cyh4uv+*oLm6ITe z)S^Z#AvQlBAhN{AnXvfqgdMbIdxO|Rg17sPU3F3Ln=XPWjf$IEpyw75dJdi|c z=|wD0cOSWgEPHE}v=Wved&V+3jf^oRKc7HkS*Q%+d5cm6nf;k$XW)p&*f%whL~0pN zEH0aC9zd2xRn~VB7RUXJCHy<-D#+$2GIby6jC}KU^k-ygqYRQrEz5}| zP4!$cvRt#Q(Mwo9a~6fs#FAds8FzWku?O+IW&T#u9hDLnrh+Wyhd~ml#e!JYFAv50 zEB9-7aDcF6I52g&dYN?LJj>d!3Dsqr8i?mD&)bqG^SZBA1R~4$)*y-0;zBH;nPU&5 zx{NCzOWQmJa196Cf9vcjo0}5Fj>7p5~*c0v0QI`+5pwXuC2o^ zVfn^cPA(^Je9c!)yMZkIx`TM$65f@(y|!v^q>SqFOcx}PS{4$EuS--GvgnU&YAY<+ zN0_<1|Xidq$FL$Z?M+h_=3L?(PRclBDHKG7Kf5PdC2nof|sMP z#5yyU3c726#zDW9s4n%RKs;{=nMjV8;OvdQs4lhRK@zFOkyz?w24*A6w7K=%g~j(M zW6^UbV=OjzZZWcaSq9>H%Zz&D`m$tNSR%4qG6qScmNUfS*!lzB&##-S!%1PWbzv;y zT*-Xzx~v5J#&_IF5YJntP9V>fEzVn|B1`cZkVI;^K`a&;-SKl}%G$SQgr$hH?2RDT zmu|I;aEC0(9dvhw~`T~wEp79fe#(w{qTQ;R_osl}dHj(>lhge_;`s*D0NG&y&K8Ozbld$8rptv$|Y zj9oVc@w{bRDOp9Zt5#|e8e=|NKoY5?6|ppSD8OAf&pYp2F+1j=EQ(_ z-qPEEbVVn+DxX7^>Ioo;)M8C6QHz`6=a=HU14aqUXU@{mkj(Ypp0|YT zB)tmWn@Ub0i(d{%BDEYRmZfTe?NMDy6`D*CmJC0pE((Xq+{JSDAJb7?4k&?m-m*E9 z+{3i&dk@#8i!w+ewS*JPL+jRfu5Xj`(HX)Ld780Ix1NB{$v+)_okNx$J|LdA7`-F! zk!Q}?;e+bZ`T|HIwWJbDy}|A9DuO<{v=<7Cw?AXq{DGX4XB^AMtD&0Pfq34Me2UD& zm{%y^)g9LzK@zE@fLO|o)=oxa>}d->Lt(KFU@Q?m$Tx`U%qdDob?MX|#PgOz+T?kE znMpDJo>KkpAc@pcWd*+Ok5gQZ=adY+KA8wh0cY8mK~{HoUp<9)rBOHo;(5!3kL10W zVZ&DY(HQd&0ZF8m=EP#~#MKMc#kT&QEy9v?hN;WMG;-{RI5i)O>T;!V+?pv2@5MD^;x0FC0Razz`76Tl!5T?<};; z@4iKrvk@SP)M7v^S}FOe$g;U&udT2+2Qrod1#-_{v|%IumRhUYHkI+bWt=lvo!(nX zaXPXLQU*z+mKDSjv1?ccvUr$jISI=*&T{$?x$fUQ_+%ogOPU3U=Pi0`$a&0ADe*e8 z{Imf{q?Vn;lF~pY0a=_Lrn(DDb`VpSgE>3#^`%(RGy++opMrSaVs1nF`dXM&=p&2n zE09ELaV3^TdTtMqCFlN|Q^FE^j>Eku$JRQm#+~=K2Y`6qlC4GVVak7K=b*Y=908I@ zEw_m!=25{rWNEZtJ4{$?gBeTnKBNz})2?s%XpEh30r9*g_!pV$ikQdo*Oi-{_^PqsR78-LUOg${`4Ez>H^4^Q|_AEb~f_HIh;SS&1 | tee -a /logs/imagenet_resnet_jax_05-12-2026-22-28-52.log -2026-05-12 22:29:04.878084: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered -WARNING: All log messages before absl::InitializeLog() is called are written to STDERR -E0000 00:00:1778624945.757940 13 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered -E0000 00:00:1778624945.922716 13 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered -W0000 00:00:1778624948.122061 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -W0000 00:00:1778624948.122103 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -W0000 00:00:1778624948.122106 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -W0000 00:00:1778624948.122109 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. -/usr/local/lib/python3.11/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.) - _C._set_float32_matmul_precision(precision) -/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'repr' attribute with value False was provided to the `Field()` function, which has no effect in the context it was used. 'repr' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. - warnings.warn( -/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'frozen' attribute with value True was provided to the `Field()` function, which has no effect in the context it was used. 'frozen' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. - warnings.warn( -INFO:2026-05-12 22:30:00,162:jax._src.xla_bridge:830: Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory -I0512 22:30:00.162919 140546196993216 xla_bridge.py:830] Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory -I0512 22:30:01.024725 140546196993216 logger_utils.py:84] Creating experiment directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax. -I0512 22:30:05.218404 140546196993216 submission_runner.py:741] Creating directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1. -I0512 22:30:05.609770 140546196993216 submission_runner.py:242] Initializing dataset. -I0512 22:30:06.752776 140546196993216 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 22:30:06.813535 140546196993216 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. -I0512 22:30:07.104005 140546196993216 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. -I0512 22:30:07.403631 140546196993216 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 22:30:08.869365 140546196993216 submission_runner.py:251] Initializing model. -I0512 22:30:33.741610 140546196993216 submission_runner.py:294] Initializing optimizer. -I0512 22:30:35.570825 140546196993216 submission_runner.py:299] Initializing metrics bundle. -I0512 22:30:35.571060 140546196993216 submission_runner.py:321] Initializing checkpoint and logger. -I0512 22:30:35.574183 140546196993216 checkpoints.py:1098] Found no checkpoint files in /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1 with prefix checkpoint_ -I0512 22:30:35.574316 140546196993216 submission_runner.py:344] Saving meta data to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/meta_data_0.json. -I0512 22:30:35.934163 140546196993216 submission_runner.py:348] Saving flags to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/flags_0.json. -I0512 22:30:35.946403 140546196993216 submission_runner.py:359] Starting training loop. -/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. -See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. - warnings.warn("Some donated buffers were not usable:" -I0512 22:31:29.790133 140531426354944 logging_writer.py:48] [0] global_step=0, grad_norm=0.3261333703994751, loss=6.908205986022949 -I0512 22:31:31.350615 140546196993216 spec.py:333] Evaluating on the training split. -I0512 22:31:31.636262 140546196993216 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 22:31:31.641801 140546196993216 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. -I0512 22:31:31.660401 140546196993216 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. -I0512 22:31:31.703538 140546196993216 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 22:31:52.586009 140546196993216 spec.py:346] Evaluating on the validation split. -I0512 22:31:52.592160 140546196993216 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 22:31:52.604061 140546196993216 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. -I0512 22:31:52.606823 140546196993216 reader.py:262] Creating a tf.data.Dataset reading 64 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. -I0512 22:31:52.647400 140546196993216 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split validation, from /data/imagenet/jax/imagenet2012/5.1.0 -I0512 22:32:23.760179 140546196993216 spec.py:363] Evaluating on the test split. -I0512 22:32:23.823080 140546196993216 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 -I0512 22:32:23.895586 140546196993216 reader.py:262] Creating a tf.data.Dataset reading 16 files located in folders: /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0. -I0512 22:32:23.945130 140546196993216 logging_logger.py:49] Constructing tf.data.Dataset imagenet_v2 for split test, from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 -I0512 22:32:29.402005 140546196993216 submission_runner.py:516] Time since start: 113.45s, Step: 1, {'train/accuracy': Array(0.00075733, dtype=float32), 'train/loss': Array(6.9125247, dtype=float32), 'validation/accuracy': Array(0.00068, dtype=float32), 'validation/loss': Array(6.91273, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0008, dtype=float32), 'test/loss': Array(6.912909, dtype=float32), 'test/num_examples': 10000, 'score': 55.403265714645386, 'total_duration': 113.45354461669922, 'accumulated_submission_time': 55.403265714645386, 'accumulated_eval_time': 58.0493426322937, 'accumulated_logging_time': 0} -I0512 22:32:29.420845 140330624063232 logging_writer.py:48] [1] accumulated_eval_time=58.0493, accumulated_logging_time=0, accumulated_submission_time=55.4033, global_step=1, preemption_count=0, score=55.4033, test/accuracy=0.000800000037997961, test/loss=6.912909030914307, test/num_examples=10000, total_duration=113.454, train/accuracy=0.0007573341717943549, train/loss=6.912524700164795, validation/accuracy=0.0006799999973736703, validation/loss=6.9127302169799805, validation/num_examples=50000 -/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. -See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. - warnings.warn("Some donated buffers were not usable:" -I0512 22:33:17.365681 140330171086592 logging_writer.py:48] [100] global_step=100, grad_norm=0.8996247053146362, loss=6.768841743469238 -I0512 22:33:54.943041 140330179479296 logging_writer.py:48] [200] global_step=200, grad_norm=1.6886597871780396, loss=6.4462175369262695 -I0512 22:34:33.023528 140330171086592 logging_writer.py:48] [300] global_step=300, grad_norm=2.236851930618286, loss=6.271664619445801 -I0512 22:35:14.390969 140330179479296 logging_writer.py:48] [400] global_step=400, grad_norm=2.128365993499756, loss=6.224735260009766 -I0512 22:36:02.658727 140330171086592 logging_writer.py:48] [500] global_step=500, grad_norm=2.709414005279541, loss=6.159616947174072 -I0512 22:36:55.749204 140330179479296 logging_writer.py:48] [600] global_step=600, grad_norm=2.2997937202453613, loss=6.0271830558776855 -I0512 22:37:45.489995 140330171086592 logging_writer.py:48] [700] global_step=700, grad_norm=2.8295984268188477, loss=5.930457592010498 -I0512 22:38:34.718007 140330179479296 logging_writer.py:48] [800] global_step=800, grad_norm=4.821854591369629, loss=5.933823585510254 -I0512 22:39:26.136206 140330171086592 logging_writer.py:48] [900] global_step=900, grad_norm=6.284852027893066, loss=5.914339542388916 -I0512 22:40:15.645335 140330179479296 logging_writer.py:48] [1000] global_step=1000, grad_norm=1.5892961025238037, loss=5.952157974243164 -I0512 22:40:53.297286 140330171086592 logging_writer.py:48] [1100] global_step=1100, grad_norm=4.078814506530762, loss=5.8306097984313965 -I0512 22:41:31.346368 140330179479296 logging_writer.py:48] [1200] global_step=1200, grad_norm=3.363912343978882, loss=5.804623603820801 -I0512 22:42:08.867039 140330171086592 logging_writer.py:48] [1300] global_step=1300, grad_norm=2.4896843433380127, loss=5.806685447692871 -I0512 22:42:46.718476 140330179479296 logging_writer.py:48] [1400] global_step=1400, grad_norm=2.7103986740112305, loss=5.822195053100586 -I0512 22:43:24.703319 140330171086592 logging_writer.py:48] [1500] global_step=1500, grad_norm=2.284644603729248, loss=5.713479042053223 -I0512 22:44:02.666852 140330179479296 logging_writer.py:48] [1600] global_step=1600, grad_norm=4.159430027008057, loss=5.635573387145996 -I0512 22:44:40.288602 140330171086592 logging_writer.py:48] [1700] global_step=1700, grad_norm=3.9693167209625244, loss=5.751264572143555 -I0512 22:45:18.289661 140330179479296 logging_writer.py:48] [1800] global_step=1800, grad_norm=2.353191614151001, loss=5.683239936828613 -I0512 22:45:55.821010 140330171086592 logging_writer.py:48] [1900] global_step=1900, grad_norm=5.872997760772705, loss=5.6698713302612305 -I0512 22:46:33.637272 140330179479296 logging_writer.py:48] [2000] global_step=2000, grad_norm=2.685798406600952, loss=5.570158004760742 -I0512 22:47:11.808946 140330171086592 logging_writer.py:48] [2100] global_step=2100, grad_norm=2.9501588344573975, loss=5.531484603881836 -I0512 22:47:49.822270 140330179479296 logging_writer.py:48] [2200] global_step=2200, grad_norm=4.772345066070557, loss=5.671840667724609 -I0512 22:48:27.157534 140330171086592 logging_writer.py:48] [2300] global_step=2300, grad_norm=2.3999035358428955, loss=5.52491569519043 -I0512 22:49:04.924759 140330179479296 logging_writer.py:48] [2400] global_step=2400, grad_norm=5.980453968048096, loss=5.530423164367676 -I0512 22:49:42.147210 140330171086592 logging_writer.py:48] [2500] global_step=2500, grad_norm=5.530206680297852, loss=5.536949634552002 -I0512 22:50:19.858959 140330179479296 logging_writer.py:48] [2600] global_step=2600, grad_norm=2.60951828956604, loss=5.436703681945801 -I0512 22:50:57.539344 140330171086592 logging_writer.py:48] [2700] global_step=2700, grad_norm=2.500755548477173, loss=5.5355753898620605 -I0512 22:51:34.735166 140330179479296 logging_writer.py:48] [2800] global_step=2800, grad_norm=2.1233749389648438, loss=5.372270584106445 -I0512 22:52:12.146931 140330171086592 logging_writer.py:48] [2900] global_step=2900, grad_norm=3.786402702331543, loss=5.477264404296875 -I0512 22:52:49.760658 140330179479296 logging_writer.py:48] [3000] global_step=3000, grad_norm=3.2771642208099365, loss=5.368007183074951 -I0512 22:53:26.779503 140330171086592 logging_writer.py:48] [3100] global_step=3100, grad_norm=3.2992141246795654, loss=5.498304843902588 -I0512 22:54:04.270782 140330179479296 logging_writer.py:48] [3200] global_step=3200, grad_norm=2.212985038757324, loss=5.261931896209717 -I0512 22:54:41.753434 140330171086592 logging_writer.py:48] [3300] global_step=3300, grad_norm=2.050915241241455, loss=5.23733377456665 -I0512 22:55:18.827353 140330179479296 logging_writer.py:48] [3400] global_step=3400, grad_norm=2.3782806396484375, loss=5.3321919441223145 -I0512 22:55:56.051253 140330171086592 logging_writer.py:48] [3500] global_step=3500, grad_norm=2.69576358795166, loss=5.3275556564331055 -I0512 22:56:33.507246 140330179479296 logging_writer.py:48] [3600] global_step=3600, grad_norm=1.549115777015686, loss=5.131993293762207 -I0512 22:57:10.457908 140330171086592 logging_writer.py:48] [3700] global_step=3700, grad_norm=1.996563196182251, loss=5.165701866149902 -I0512 22:57:47.797516 140330179479296 logging_writer.py:48] [3800] global_step=3800, grad_norm=2.4652886390686035, loss=5.186154842376709 -I0512 22:58:25.266681 140330171086592 logging_writer.py:48] [3900] global_step=3900, grad_norm=3.4159767627716064, loss=5.325900077819824 -I0512 22:59:02.168974 140330179479296 logging_writer.py:48] [4000] global_step=4000, grad_norm=2.5840179920196533, loss=5.2093119621276855 -I0512 22:59:39.121025 140330171086592 logging_writer.py:48] [4100] global_step=4100, grad_norm=7.112630367279053, loss=5.181420803070068 -I0512 23:00:16.757699 140330179479296 logging_writer.py:48] [4200] global_step=4200, grad_norm=2.108964443206787, loss=5.031390190124512 -I0512 23:00:53.567319 140330171086592 logging_writer.py:48] [4300] global_step=4300, grad_norm=3.34953236579895, loss=5.31886100769043 -I0512 23:01:30.567174 140330179479296 logging_writer.py:48] [4400] global_step=4400, grad_norm=3.030181407928467, loss=5.145931243896484 -I0512 23:02:07.710606 140330171086592 logging_writer.py:48] [4500] global_step=4500, grad_norm=5.9440598487854, loss=5.051600456237793 -I0512 23:02:44.361296 140330179479296 logging_writer.py:48] [4600] global_step=4600, grad_norm=6.287307262420654, loss=5.062731742858887 -I0512 23:03:21.434231 140330171086592 logging_writer.py:48] [4700] global_step=4700, grad_norm=2.710148334503174, loss=5.037991046905518 -I0512 23:03:58.680780 140330179479296 logging_writer.py:48] [4800] global_step=4800, grad_norm=5.478421688079834, loss=5.159373760223389 -I0512 23:04:35.386637 140330171086592 logging_writer.py:48] [4900] global_step=4900, grad_norm=2.5405635833740234, loss=4.927253246307373 -I0512 23:05:12.512895 140330179479296 logging_writer.py:48] [5000] global_step=5000, grad_norm=2.597332239151001, loss=4.854314804077148 -I0512 23:05:45.472953 140546196993216 spec.py:333] Evaluating on the training split. -I0512 23:05:57.501238 140546196993216 spec.py:346] Evaluating on the validation split. -I0512 23:06:07.353067 140546196993216 spec.py:363] Evaluating on the test split. -I0512 23:06:08.299368 140546196993216 submission_runner.py:516] Time since start: 2132.35s, Step: 5091, {'train/accuracy': Array(0.03683036, dtype=float32), 'train/loss': Array(6.0105863, dtype=float32), 'validation/accuracy': Array(0.03464, dtype=float32), 'validation/loss': Array(6.073887, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0256, dtype=float32), 'test/loss': Array(6.2957478, dtype=float32), 'test/num_examples': 10000, 'score': 2051.405519247055, 'total_duration': 2132.3507702350616, 'accumulated_submission_time': 2051.405519247055, 'accumulated_eval_time': 80.8735921382904, 'accumulated_logging_time': 0.028288841247558594} -I0512 23:06:08.328180 140330632455936 logging_writer.py:48] [5091] accumulated_eval_time=80.8736, accumulated_logging_time=0.0282888, accumulated_submission_time=2051.41, global_step=5091, preemption_count=0, score=2051.41, test/accuracy=0.025600001215934753, test/loss=6.295747756958008, test/num_examples=10000, total_duration=2132.35, train/accuracy=0.0368303582072258, train/loss=6.010586261749268, validation/accuracy=0.03463999927043915, validation/loss=6.073886871337891, validation/num_examples=50000 -I0512 23:06:13.213566 140330640848640 logging_writer.py:48] [5100] global_step=5100, grad_norm=3.053668737411499, loss=4.953907489776611 -I0512 23:06:49.909910 140330632455936 logging_writer.py:48] [5200] global_step=5200, grad_norm=2.6351094245910645, loss=4.91639518737793 -I0512 23:07:27.058860 140330640848640 logging_writer.py:48] [5300] global_step=5300, grad_norm=4.073904991149902, loss=4.984274387359619 -I0512 23:08:04.298047 140330632455936 logging_writer.py:48] [5400] global_step=5400, grad_norm=7.666659832000732, loss=5.010444641113281 -I0512 23:08:41.081388 140330640848640 logging_writer.py:48] [5500] global_step=5500, grad_norm=2.171753168106079, loss=4.946066856384277 -I0512 23:09:18.153785 140330632455936 logging_writer.py:48] [5600] global_step=5600, grad_norm=4.941290855407715, loss=5.053196907043457 -I0512 23:09:55.333907 140330640848640 logging_writer.py:48] [5700] global_step=5700, grad_norm=3.2240264415740967, loss=4.700118064880371 -I0512 23:10:32.387669 140330632455936 logging_writer.py:48] [5800] global_step=5800, grad_norm=5.332955837249756, loss=5.0179338455200195 -I0512 23:11:09.074151 140330640848640 logging_writer.py:48] [5900] global_step=5900, grad_norm=2.567713499069214, loss=4.934077262878418 -I0512 23:11:46.275546 140330632455936 logging_writer.py:48] [6000] global_step=6000, grad_norm=3.471893787384033, loss=4.8420305252075195 -I0512 23:12:22.972817 140330640848640 logging_writer.py:48] [6100] global_step=6100, grad_norm=3.49885892868042, loss=4.718890190124512 -I0512 23:12:59.913679 140330632455936 logging_writer.py:48] [6200] global_step=6200, grad_norm=8.484103202819824, loss=4.963932037353516 -I0512 23:13:37.133269 140330640848640 logging_writer.py:48] [6300] global_step=6300, grad_norm=2.9837536811828613, loss=4.788043975830078 -I0512 23:14:14.268118 140330632455936 logging_writer.py:48] [6400] global_step=6400, grad_norm=3.5700173377990723, loss=4.8348822593688965 -I0512 23:14:50.951000 140330640848640 logging_writer.py:48] [6500] global_step=6500, grad_norm=8.07686710357666, loss=5.0427775382995605 -I0512 23:15:28.172864 140330632455936 logging_writer.py:48] [6600] global_step=6600, grad_norm=5.35474157333374, loss=4.79836368560791 -I0512 23:16:04.903439 140330640848640 logging_writer.py:48] [6700] global_step=6700, grad_norm=8.914992332458496, loss=4.907289505004883 -I0512 23:16:41.900763 140330632455936 logging_writer.py:48] [6800] global_step=6800, grad_norm=5.695804119110107, loss=4.9365129470825195 -I0512 23:17:19.175864 140330640848640 logging_writer.py:48] [6900] global_step=6900, grad_norm=5.530570030212402, loss=4.8735151290893555 -I0512 23:17:55.883324 140330632455936 logging_writer.py:48] [7000] global_step=7000, grad_norm=4.144064903259277, loss=4.715763568878174 -I0512 23:18:32.972416 140330640848640 logging_writer.py:48] [7100] global_step=7100, grad_norm=5.426703929901123, loss=4.743924140930176 -I0512 23:19:10.145425 140330632455936 logging_writer.py:48] [7200] global_step=7200, grad_norm=3.9532980918884277, loss=4.594181537628174 -I0512 23:19:46.861758 140330640848640 logging_writer.py:48] [7300] global_step=7300, grad_norm=5.157228946685791, loss=4.593221664428711 -I0512 23:20:23.868392 140330632455936 logging_writer.py:48] [7400] global_step=7400, grad_norm=4.616650104522705, loss=4.816321849822998 -I0512 23:21:01.183516 140330640848640 logging_writer.py:48] [7500] global_step=7500, grad_norm=2.9376327991485596, loss=4.6537322998046875 -I0512 23:21:37.891778 140330632455936 logging_writer.py:48] [7600] global_step=7600, grad_norm=3.5215342044830322, loss=4.566107273101807 -I0512 23:22:14.913129 140330640848640 logging_writer.py:48] [7700] global_step=7700, grad_norm=2.8087148666381836, loss=4.520059108734131 -I0512 23:22:52.219972 140330632455936 logging_writer.py:48] [7800] global_step=7800, grad_norm=3.9979968070983887, loss=4.738373756408691 -I0512 23:23:28.911386 140330640848640 logging_writer.py:48] [7900] global_step=7900, grad_norm=3.8169260025024414, loss=4.610917568206787 -I0512 23:24:06.002536 140330632455936 logging_writer.py:48] [8000] global_step=8000, grad_norm=6.051599502563477, loss=4.608170509338379 -I0512 23:24:43.194123 140330640848640 logging_writer.py:48] [8100] global_step=8100, grad_norm=5.458561420440674, loss=4.7805938720703125 -I0512 23:25:19.987114 140330632455936 logging_writer.py:48] [8200] global_step=8200, grad_norm=2.866084098815918, loss=4.509045600891113 -I0512 23:25:57.048641 140330640848640 logging_writer.py:48] [8300] global_step=8300, grad_norm=4.061978816986084, loss=4.564599990844727 -I0512 23:26:34.231000 140330632455936 logging_writer.py:48] [8400] global_step=8400, grad_norm=4.84849739074707, loss=4.808914661407471 -I0512 23:27:11.019012 140330640848640 logging_writer.py:48] [8500] global_step=8500, grad_norm=5.6259613037109375, loss=4.656274795532227 -I0512 23:27:48.068397 140330632455936 logging_writer.py:48] [8600] global_step=8600, grad_norm=2.716402530670166, loss=4.465431213378906 -I0512 23:28:25.306610 140330640848640 logging_writer.py:48] [8700] global_step=8700, grad_norm=3.26397442817688, loss=4.493489742279053 -I0512 23:29:02.121843 140330632455936 logging_writer.py:48] [8800] global_step=8800, grad_norm=5.2092437744140625, loss=4.543220520019531 -I0512 23:29:38.903265 140330640848640 logging_writer.py:48] [8900] global_step=8900, grad_norm=7.477419853210449, loss=4.424639701843262 -I0512 23:30:16.375513 140330632455936 logging_writer.py:48] [9000] global_step=9000, grad_norm=7.936310291290283, loss=4.5097761154174805 -I0512 23:30:53.129014 140330640848640 logging_writer.py:48] [9100] global_step=9100, grad_norm=3.1809134483337402, loss=4.536842346191406 -I0512 23:31:30.191936 140330632455936 logging_writer.py:48] [9200] global_step=9200, grad_norm=4.791093826293945, loss=4.713247776031494 -I0512 23:32:07.336721 140330640848640 logging_writer.py:48] [9300] global_step=9300, grad_norm=4.619687080383301, loss=4.56913948059082 -I0512 23:32:44.095336 140330632455936 logging_writer.py:48] [9400] global_step=9400, grad_norm=3.10891056060791, loss=4.452826499938965 -I0512 23:33:21.160023 140330640848640 logging_writer.py:48] [9500] global_step=9500, grad_norm=4.322040557861328, loss=4.496364116668701 -I0512 23:33:58.327754 140330632455936 logging_writer.py:48] [9600] global_step=9600, grad_norm=3.6211421489715576, loss=4.540789604187012 -I0512 23:34:35.179843 140330640848640 logging_writer.py:48] [9700] global_step=9700, grad_norm=5.0317063331604, loss=4.562549114227295 -I0512 23:35:12.223801 140330632455936 logging_writer.py:48] [9800] global_step=9800, grad_norm=3.3404557704925537, loss=4.507644176483154 -I0512 23:35:49.400078 140330640848640 logging_writer.py:48] [9900] global_step=9900, grad_norm=6.4189534187316895, loss=5.337581634521484 -I0512 23:36:26.141850 140330632455936 logging_writer.py:48] [10000] global_step=10000, grad_norm=5.7183918952941895, loss=4.41055965423584 -I0512 23:37:03.368647 140330640848640 logging_writer.py:48] [10100] global_step=10100, grad_norm=4.658958911895752, loss=4.356914520263672 -I0512 23:37:40.560816 140330632455936 logging_writer.py:48] [10200] global_step=10200, grad_norm=5.685758113861084, loss=4.766016006469727 -I0512 23:38:17.271877 140330640848640 logging_writer.py:48] [10300] global_step=10300, grad_norm=3.6148455142974854, loss=4.464681625366211 -I0512 23:38:54.384885 140330632455936 logging_writer.py:48] [10400] global_step=10400, grad_norm=3.386692523956299, loss=4.437700271606445 -I0512 23:39:24.775613 140546196993216 spec.py:333] Evaluating on the training split. -I0512 23:39:37.984788 140546196993216 spec.py:346] Evaluating on the validation split. -I0512 23:39:47.833569 140546196993216 spec.py:363] Evaluating on the test split. -I0512 23:39:48.724383 140546196993216 submission_runner.py:516] Time since start: 4152.78s, Step: 10483, {'train/accuracy': Array(0.01183833, dtype=float32), 'train/loss': Array(7.4746013, dtype=float32), 'validation/accuracy': Array(0.01248, dtype=float32), 'validation/loss': Array(7.483133, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0092, dtype=float32), 'test/loss': Array(7.565653, dtype=float32), 'test/num_examples': 10000, 'score': 4047.7705268859863, 'total_duration': 4152.775046825409, 'accumulated_submission_time': 4047.7705268859863, 'accumulated_eval_time': 104.81943488121033, 'accumulated_logging_time': 0.09688901901245117} -I0512 23:39:48.763230 140330640848640 logging_writer.py:48] [10483] accumulated_eval_time=104.819, accumulated_logging_time=0.096889, accumulated_submission_time=4047.77, global_step=10483, preemption_count=0, score=4047.77, test/accuracy=0.009200000204145908, test/loss=7.565652847290039, test/num_examples=10000, total_duration=4152.78, train/accuracy=0.011838329024612904, train/loss=7.4746012687683105, validation/accuracy=0.012480000033974648, validation/loss=7.483132839202881, validation/num_examples=50000 -I0512 23:39:56.297309 140330632455936 logging_writer.py:48] [10500] global_step=10500, grad_norm=4.742617130279541, loss=4.408299446105957 -I0512 23:40:33.064100 140330640848640 logging_writer.py:48] [10600] global_step=10600, grad_norm=6.10215950012207, loss=4.77248477935791 -I0512 23:41:10.182475 140330632455936 logging_writer.py:48] [10700] global_step=10700, grad_norm=3.5745131969451904, loss=4.301945686340332 -I0512 23:41:47.407188 140330640848640 logging_writer.py:48] [10800] global_step=10800, grad_norm=4.048213481903076, loss=4.50079870223999 -I0512 23:42:24.140448 140330632455936 logging_writer.py:48] [10900] global_step=10900, grad_norm=4.244584560394287, loss=4.384276390075684 -I0512 23:43:00.879302 140330640848640 logging_writer.py:48] [11000] global_step=11000, grad_norm=2.615108013153076, loss=4.353362083435059 -I0512 23:43:38.358006 140330632455936 logging_writer.py:48] [11100] global_step=11100, grad_norm=6.9523844718933105, loss=4.649722099304199 -I0512 23:44:15.130794 140330640848640 logging_writer.py:48] [11200] global_step=11200, grad_norm=4.475399017333984, loss=4.485107421875 -I0512 23:44:52.240787 140330632455936 logging_writer.py:48] [11300] global_step=11300, grad_norm=4.415673732757568, loss=4.444968223571777 -I0512 23:45:29.423520 140330640848640 logging_writer.py:48] [11400] global_step=11400, grad_norm=4.174510955810547, loss=4.252237319946289 -I0512 23:46:06.128505 140330632455936 logging_writer.py:48] [11500] global_step=11500, grad_norm=6.8448662757873535, loss=4.461555480957031 -I0512 23:46:42.855457 140330640848640 logging_writer.py:48] [11600] global_step=11600, grad_norm=3.8904194831848145, loss=4.323244094848633 -I0512 23:47:20.355967 140330632455936 logging_writer.py:48] [11700] global_step=11700, grad_norm=4.4755353927612305, loss=4.250211238861084 -I0512 23:47:57.043379 140330640848640 logging_writer.py:48] [11800] global_step=11800, grad_norm=4.67167329788208, loss=4.293257713317871 -I0512 23:48:34.124796 140330632455936 logging_writer.py:48] [11900] global_step=11900, grad_norm=5.430049419403076, loss=4.391984939575195 -I0512 23:49:11.319457 140330640848640 logging_writer.py:48] [12000] global_step=12000, grad_norm=5.566277503967285, loss=4.557373046875 -I0512 23:49:48.100352 140330632455936 logging_writer.py:48] [12100] global_step=12100, grad_norm=4.70858907699585, loss=4.373384952545166 -I0512 23:50:24.945086 140330640848640 logging_writer.py:48] [12200] global_step=12200, grad_norm=4.117347240447998, loss=4.291561126708984 -I0512 23:51:02.456800 140330632455936 logging_writer.py:48] [12300] global_step=12300, grad_norm=5.649754047393799, loss=4.274422645568848 -I0512 23:51:39.183211 140330640848640 logging_writer.py:48] [12400] global_step=12400, grad_norm=5.82424259185791, loss=4.297607898712158 -I0512 23:52:15.875211 140330632455936 logging_writer.py:48] [12500] global_step=12500, grad_norm=6.6620073318481445, loss=4.369428634643555 -I0512 23:52:53.352773 140330640848640 logging_writer.py:48] [12600] global_step=12600, grad_norm=10.304540634155273, loss=4.356822967529297 -I0512 23:53:30.116830 140330632455936 logging_writer.py:48] [12700] global_step=12700, grad_norm=2.569653272628784, loss=4.292733192443848 -I0512 23:54:06.932348 140330640848640 logging_writer.py:48] [12800] global_step=12800, grad_norm=4.534213542938232, loss=4.364278316497803 -I0512 23:54:44.496871 140330632455936 logging_writer.py:48] [12900] global_step=12900, grad_norm=5.462316513061523, loss=4.389804840087891 -I0512 23:55:21.326714 140330640848640 logging_writer.py:48] [13000] global_step=13000, grad_norm=4.455352783203125, loss=4.215510845184326 -I0512 23:55:58.121049 140330632455936 logging_writer.py:48] [13100] global_step=13100, grad_norm=4.2624969482421875, loss=4.316892623901367 -I0512 23:56:35.567021 140330640848640 logging_writer.py:48] [13200] global_step=13200, grad_norm=4.83694314956665, loss=4.208202362060547 -I0512 23:57:12.377397 140330632455936 logging_writer.py:48] [13300] global_step=13300, grad_norm=4.446932315826416, loss=4.165616035461426 -I0512 23:57:49.193160 140330640848640 logging_writer.py:48] [13400] global_step=13400, grad_norm=5.057380199432373, loss=4.377295017242432 -I0512 23:58:26.737131 140330632455936 logging_writer.py:48] [13500] global_step=13500, grad_norm=4.3487725257873535, loss=4.240519046783447 -I0512 23:59:03.520822 140330640848640 logging_writer.py:48] [13600] global_step=13600, grad_norm=3.4954211711883545, loss=4.212986946105957 -I0512 23:59:40.232921 140330632455936 logging_writer.py:48] [13700] global_step=13700, grad_norm=19.337270736694336, loss=4.730031967163086 -I0513 00:00:17.880152 140330640848640 logging_writer.py:48] [13800] global_step=13800, grad_norm=3.999199628829956, loss=4.161625385284424 -I0513 00:00:54.529289 140330632455936 logging_writer.py:48] [13900] global_step=13900, grad_norm=3.0594756603240967, loss=4.121606826782227 -I0513 00:01:31.233757 140330640848640 logging_writer.py:48] [14000] global_step=14000, grad_norm=13.660304069519043, loss=4.312893867492676 -I0513 00:02:08.814565 140330632455936 logging_writer.py:48] [14100] global_step=14100, grad_norm=4.691737651824951, loss=4.19709587097168 -I0513 00:02:45.554623 140330640848640 logging_writer.py:48] [14200] global_step=14200, grad_norm=6.446077346801758, loss=4.287468433380127 -I0513 00:03:22.266606 140330632455936 logging_writer.py:48] [14300] global_step=14300, grad_norm=5.878997325897217, loss=4.315158367156982 -I0513 00:03:59.837368 140330640848640 logging_writer.py:48] [14400] global_step=14400, grad_norm=5.304042339324951, loss=4.257931709289551 -I0513 00:04:36.549085 140330632455936 logging_writer.py:48] [14500] global_step=14500, grad_norm=3.778127670288086, loss=4.202208518981934 -I0513 00:05:13.230744 140330640848640 logging_writer.py:48] [14600] global_step=14600, grad_norm=7.562634468078613, loss=4.369805335998535 -I0513 00:05:50.828395 140330632455936 logging_writer.py:48] [14700] global_step=14700, grad_norm=4.75720739364624, loss=4.184082508087158 -I0513 00:06:27.512907 140330640848640 logging_writer.py:48] [14800] global_step=14800, grad_norm=3.54793119430542, loss=4.183678150177002 -I0513 00:07:04.286172 140330632455936 logging_writer.py:48] [14900] global_step=14900, grad_norm=5.753579139709473, loss=4.237636566162109 -I0513 00:07:41.779661 140330640848640 logging_writer.py:48] [15000] global_step=15000, grad_norm=7.225032329559326, loss=4.329081058502197 -I0513 00:08:18.544023 140330632455936 logging_writer.py:48] [15100] global_step=15100, grad_norm=4.463640213012695, loss=4.182171821594238 -I0513 00:08:55.250330 140330640848640 logging_writer.py:48] [15200] global_step=15200, grad_norm=5.989433288574219, loss=4.14880895614624 -I0513 00:09:32.728845 140330632455936 logging_writer.py:48] [15300] global_step=15300, grad_norm=9.251641273498535, loss=4.281979084014893 -I0513 00:10:09.376390 140330640848640 logging_writer.py:48] [15400] global_step=15400, grad_norm=6.962396144866943, loss=4.210168361663818 -I0513 00:10:46.106201 140330632455936 logging_writer.py:48] [15500] global_step=15500, grad_norm=5.1545729637146, loss=4.1486616134643555 -I0513 00:11:23.576538 140330640848640 logging_writer.py:48] [15600] global_step=15600, grad_norm=4.016613960266113, loss=4.150571346282959 -I0513 00:12:00.255259 140330632455936 logging_writer.py:48] [15700] global_step=15700, grad_norm=6.127602577209473, loss=4.092342853546143 -I0513 00:12:36.933308 140330640848640 logging_writer.py:48] [15800] global_step=15800, grad_norm=5.064116477966309, loss=4.119712829589844 -I0513 00:13:05.073023 140546196993216 spec.py:333] Evaluating on the training split. -I0513 00:13:18.265767 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 00:13:27.736178 140546196993216 spec.py:363] Evaluating on the test split. -I0513 00:13:28.603624 140546196993216 submission_runner.py:516] Time since start: 6172.65s, Step: 15877, {'train/accuracy': Array(0.00480309, dtype=float32), 'train/loss': Array(8.611359, dtype=float32), 'validation/accuracy': Array(0.00446, dtype=float32), 'validation/loss': Array(8.60309, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0046, dtype=float32), 'test/loss': Array(8.634853, dtype=float32), 'test/num_examples': 10000, 'score': 6044.017327547073, 'total_duration': 6172.6545305252075, 'accumulated_submission_time': 6044.017327547073, 'accumulated_eval_time': 128.34734845161438, 'accumulated_logging_time': 0.15599656105041504} -I0513 00:13:28.641412 140330632455936 logging_writer.py:48] [15877] accumulated_eval_time=128.347, accumulated_logging_time=0.155997, accumulated_submission_time=6044.02, global_step=15877, preemption_count=0, score=6044.02, test/accuracy=0.004600000102072954, test/loss=8.63485336303711, test/num_examples=10000, total_duration=6172.65, train/accuracy=0.004803093150258064, train/loss=8.611358642578125, validation/accuracy=0.004459999967366457, validation/loss=8.603090286254883, validation/num_examples=50000 -I0513 00:13:38.788293 140330640848640 logging_writer.py:48] [15900] global_step=15900, grad_norm=4.5788798332214355, loss=4.093354225158691 -I0513 00:14:15.485550 140330632455936 logging_writer.py:48] [16000] global_step=16000, grad_norm=13.225955963134766, loss=4.051819801330566 -I0513 00:14:52.260981 140330640848640 logging_writer.py:48] [16100] global_step=16100, grad_norm=7.334598064422607, loss=4.472049713134766 -I0513 00:15:29.788726 140330632455936 logging_writer.py:48] [16200] global_step=16200, grad_norm=3.6273224353790283, loss=4.231749534606934 -I0513 00:16:06.550204 140330640848640 logging_writer.py:48] [16300] global_step=16300, grad_norm=8.078614234924316, loss=4.173711776733398 -I0513 00:16:43.272800 140330632455936 logging_writer.py:48] [16400] global_step=16400, grad_norm=5.320896148681641, loss=4.243307113647461 -I0513 00:17:20.800691 140330640848640 logging_writer.py:48] [16500] global_step=16500, grad_norm=10.41242504119873, loss=4.254964351654053 -I0513 00:17:57.516160 140330632455936 logging_writer.py:48] [16600] global_step=16600, grad_norm=10.703930854797363, loss=4.311012268066406 -I0513 00:18:34.275593 140330640848640 logging_writer.py:48] [16700] global_step=16700, grad_norm=3.8985214233398438, loss=4.160813331604004 -I0513 00:19:11.734932 140330632455936 logging_writer.py:48] [16800] global_step=16800, grad_norm=3.5429341793060303, loss=4.151752948760986 -I0513 00:19:48.427730 140330640848640 logging_writer.py:48] [16900] global_step=16900, grad_norm=6.4630255699157715, loss=4.278426170349121 -I0513 00:20:25.107810 140330632455936 logging_writer.py:48] [17000] global_step=17000, grad_norm=4.048737049102783, loss=4.100284576416016 -I0513 00:21:02.740613 140330640848640 logging_writer.py:48] [17100] global_step=17100, grad_norm=6.791495323181152, loss=3.972121238708496 -I0513 00:21:39.477401 140330632455936 logging_writer.py:48] [17200] global_step=17200, grad_norm=3.659362554550171, loss=4.05012321472168 -I0513 00:22:16.169252 140330640848640 logging_writer.py:48] [17300] global_step=17300, grad_norm=6.690219402313232, loss=4.263319492340088 -I0513 00:22:53.500745 140330632455936 logging_writer.py:48] [17400] global_step=17400, grad_norm=4.609320163726807, loss=4.117898464202881 -I0513 00:23:30.451060 140330640848640 logging_writer.py:48] [17500] global_step=17500, grad_norm=7.970854759216309, loss=4.105091571807861 -I0513 00:24:07.158175 140330632455936 logging_writer.py:48] [17600] global_step=17600, grad_norm=2.6216917037963867, loss=4.1468048095703125 -I0513 00:24:44.716186 140330640848640 logging_writer.py:48] [17700] global_step=17700, grad_norm=4.729665279388428, loss=4.056459426879883 -I0513 00:25:21.420770 140330632455936 logging_writer.py:48] [17800] global_step=17800, grad_norm=7.799706935882568, loss=4.2269287109375 -I0513 00:25:58.165429 140330640848640 logging_writer.py:48] [17900] global_step=17900, grad_norm=6.864507675170898, loss=4.107316493988037 -I0513 00:26:35.472539 140330632455936 logging_writer.py:48] [18000] global_step=18000, grad_norm=9.228012084960938, loss=4.1618971824646 -I0513 00:27:12.554463 140330640848640 logging_writer.py:48] [18100] global_step=18100, grad_norm=8.892206192016602, loss=4.076744556427002 -I0513 00:27:49.260925 140330632455936 logging_writer.py:48] [18200] global_step=18200, grad_norm=4.733257293701172, loss=4.035945892333984 -I0513 00:28:26.850257 140330640848640 logging_writer.py:48] [18300] global_step=18300, grad_norm=6.942631721496582, loss=4.379856109619141 -I0513 00:29:03.665925 140330632455936 logging_writer.py:48] [18400] global_step=18400, grad_norm=5.192965030670166, loss=4.246373176574707 -I0513 00:29:40.534008 140330640848640 logging_writer.py:48] [18500] global_step=18500, grad_norm=3.771827220916748, loss=3.9473421573638916 -I0513 00:30:17.820033 140330632455936 logging_writer.py:48] [18600] global_step=18600, grad_norm=6.6334357261657715, loss=4.184978485107422 -I0513 00:30:54.854854 140330640848640 logging_writer.py:48] [18700] global_step=18700, grad_norm=4.732202529907227, loss=4.096646308898926 -I0513 00:31:31.719202 140330632455936 logging_writer.py:48] [18800] global_step=18800, grad_norm=9.549346923828125, loss=4.187075614929199 -I0513 00:32:08.907125 140330640848640 logging_writer.py:48] [18900] global_step=18900, grad_norm=9.255634307861328, loss=4.0793914794921875 -I0513 00:32:45.827051 140330632455936 logging_writer.py:48] [19000] global_step=19000, grad_norm=10.306611061096191, loss=4.305615425109863 -I0513 00:33:22.633076 140330640848640 logging_writer.py:48] [19100] global_step=19100, grad_norm=3.426957845687866, loss=4.068285942077637 -I0513 00:33:59.779063 140330632455936 logging_writer.py:48] [19200] global_step=19200, grad_norm=4.317892551422119, loss=4.030745029449463 -I0513 00:34:36.794956 140330640848640 logging_writer.py:48] [19300] global_step=19300, grad_norm=7.502645015716553, loss=4.164514541625977 -I0513 00:35:13.616872 140330632455936 logging_writer.py:48] [19400] global_step=19400, grad_norm=3.6044952869415283, loss=4.057674884796143 -I0513 00:35:50.763972 140330640848640 logging_writer.py:48] [19500] global_step=19500, grad_norm=4.171695709228516, loss=3.9916868209838867 -I0513 00:36:27.809201 140330632455936 logging_writer.py:48] [19600] global_step=19600, grad_norm=4.710084915161133, loss=4.284005165100098 -I0513 00:37:04.493147 140330640848640 logging_writer.py:48] [19700] global_step=19700, grad_norm=7.094947814941406, loss=4.12871789932251 -I0513 00:37:41.755940 140330632455936 logging_writer.py:48] [19800] global_step=19800, grad_norm=4.996074199676514, loss=4.214055061340332 -I0513 00:38:18.779000 140330640848640 logging_writer.py:48] [19900] global_step=19900, grad_norm=6.2610249519348145, loss=4.188875198364258 -I0513 00:38:55.485845 140330632455936 logging_writer.py:48] [20000] global_step=20000, grad_norm=5.478777885437012, loss=4.156248569488525 -I0513 00:39:32.719083 140330640848640 logging_writer.py:48] [20100] global_step=20100, grad_norm=4.217374801635742, loss=4.1089887619018555 -I0513 00:40:09.388325 140330632455936 logging_writer.py:48] [20200] global_step=20200, grad_norm=5.179275035858154, loss=4.0980143547058105 -I0513 00:40:46.526535 140330640848640 logging_writer.py:48] [20300] global_step=20300, grad_norm=7.884507656097412, loss=4.10603141784668 -I0513 00:41:23.742770 140330632455936 logging_writer.py:48] [20400] global_step=20400, grad_norm=6.454740524291992, loss=4.133495807647705 -I0513 00:42:00.784411 140330640848640 logging_writer.py:48] [20500] global_step=20500, grad_norm=6.616697311401367, loss=4.159370422363281 -I0513 00:42:37.509023 140330632455936 logging_writer.py:48] [20600] global_step=20600, grad_norm=9.413150787353516, loss=4.292880535125732 -I0513 00:43:14.655262 140330640848640 logging_writer.py:48] [20700] global_step=20700, grad_norm=7.022496223449707, loss=4.159736156463623 -I0513 00:43:51.787393 140330632455936 logging_writer.py:48] [20800] global_step=20800, grad_norm=6.854641914367676, loss=3.9935193061828613 -I0513 00:44:28.575999 140330640848640 logging_writer.py:48] [20900] global_step=20900, grad_norm=15.363938331604004, loss=4.427828788757324 -I0513 00:45:05.870188 140330632455936 logging_writer.py:48] [21000] global_step=21000, grad_norm=4.507478713989258, loss=4.078317165374756 -I0513 00:45:42.915359 140330640848640 logging_writer.py:48] [21100] global_step=21100, grad_norm=172.121826171875, loss=4.6842169761657715 -I0513 00:46:19.654111 140330632455936 logging_writer.py:48] [21200] global_step=21200, grad_norm=8.730419158935547, loss=4.150398254394531 -I0513 00:46:44.736973 140546196993216 spec.py:333] Evaluating on the training split. -I0513 00:46:59.632646 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 00:47:09.342656 140546196993216 spec.py:363] Evaluating on the test split. -I0513 00:47:10.234863 140546196993216 submission_runner.py:516] Time since start: 8194.29s, Step: 21268, {'train/accuracy': Array(0.00229193, dtype=float32), 'train/loss': Array(9.054386, dtype=float32), 'validation/accuracy': Array(0.00214, dtype=float32), 'validation/loss': Array(9.063879, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0025, dtype=float32), 'test/loss': Array(9.078246, dtype=float32), 'test/num_examples': 10000, 'score': 8040.030968904495, 'total_duration': 8194.285801887512, 'accumulated_submission_time': 8040.030968904495, 'accumulated_eval_time': 153.84260439872742, 'accumulated_logging_time': 0.23279166221618652} -I0513 00:47:10.268338 140330640848640 logging_writer.py:48] [21268] accumulated_eval_time=153.843, accumulated_logging_time=0.232792, accumulated_submission_time=8040.03, global_step=21268, preemption_count=0, score=8040.03, test/accuracy=0.002500000176951289, test/loss=9.078246116638184, test/num_examples=10000, total_duration=8194.29, train/accuracy=0.002291932236403227, train/loss=9.054386138916016, validation/accuracy=0.002139999996870756, validation/loss=9.063879013061523, validation/num_examples=50000 -I0513 00:47:23.349791 140330632455936 logging_writer.py:48] [21300] global_step=21300, grad_norm=4.054925441741943, loss=3.9688007831573486 -I0513 00:48:00.377999 140330640848640 logging_writer.py:48] [21400] global_step=21400, grad_norm=5.245344161987305, loss=3.9229979515075684 -I0513 00:48:37.150371 140330632455936 logging_writer.py:48] [21500] global_step=21500, grad_norm=3.6690187454223633, loss=3.979999542236328 -I0513 00:49:14.596124 140330640848640 logging_writer.py:48] [21600] global_step=21600, grad_norm=4.3650078773498535, loss=4.003007888793945 -I0513 00:49:51.375471 140330632455936 logging_writer.py:48] [21700] global_step=21700, grad_norm=8.622546195983887, loss=4.0196051597595215 -I0513 00:50:28.068899 140330640848640 logging_writer.py:48] [21800] global_step=21800, grad_norm=6.2972612380981445, loss=4.07935905456543 -I0513 00:51:05.652889 140330632455936 logging_writer.py:48] [21900] global_step=21900, grad_norm=4.256289482116699, loss=4.009181499481201 -I0513 00:51:42.342063 140330640848640 logging_writer.py:48] [22000] global_step=22000, grad_norm=5.0948405265808105, loss=4.071035385131836 -I0513 00:52:19.104618 140330632455936 logging_writer.py:48] [22100] global_step=22100, grad_norm=5.62742280960083, loss=4.087027072906494 -I0513 00:52:56.256392 140330640848640 logging_writer.py:48] [22200] global_step=22200, grad_norm=9.705975532531738, loss=3.964193820953369 -I0513 00:53:33.190306 140330632455936 logging_writer.py:48] [22300] global_step=22300, grad_norm=11.566600799560547, loss=3.8987789154052734 -I0513 00:54:09.962926 140330640848640 logging_writer.py:48] [22400] global_step=22400, grad_norm=14.853710174560547, loss=4.249449729919434 -I0513 00:54:47.500360 140330632455936 logging_writer.py:48] [22500] global_step=22500, grad_norm=5.387282371520996, loss=3.9543471336364746 -I0513 00:55:24.207715 140330640848640 logging_writer.py:48] [22600] global_step=22600, grad_norm=3.6187617778778076, loss=4.0621256828308105 -I0513 00:56:00.945128 140330632455936 logging_writer.py:48] [22700] global_step=22700, grad_norm=5.334948539733887, loss=4.110098838806152 -I0513 00:56:38.169515 140330640848640 logging_writer.py:48] [22800] global_step=22800, grad_norm=5.688113212585449, loss=4.121650218963623 -I0513 00:57:15.293376 140330632455936 logging_writer.py:48] [22900] global_step=22900, grad_norm=5.4531025886535645, loss=4.049832820892334 -I0513 00:57:52.022062 140330640848640 logging_writer.py:48] [23000] global_step=23000, grad_norm=10.37078857421875, loss=3.953493356704712 -I0513 00:58:29.542658 140330632455936 logging_writer.py:48] [23100] global_step=23100, grad_norm=6.2461042404174805, loss=4.078622817993164 -I0513 00:59:06.316485 140330640848640 logging_writer.py:48] [23200] global_step=23200, grad_norm=2.473088502883911, loss=3.98260498046875 -I0513 00:59:43.040689 140330632455936 logging_writer.py:48] [23300] global_step=23300, grad_norm=4.933840751647949, loss=4.062488555908203 -I0513 01:00:20.222128 140330640848640 logging_writer.py:48] [23400] global_step=23400, grad_norm=4.4797587394714355, loss=3.8812568187713623 -I0513 01:00:57.303169 140330632455936 logging_writer.py:48] [23500] global_step=23500, grad_norm=8.269773483276367, loss=4.0552978515625 -I0513 01:01:34.058375 140330640848640 logging_writer.py:48] [23600] global_step=23600, grad_norm=4.281411647796631, loss=4.004667282104492 -I0513 01:02:11.275937 140330632455936 logging_writer.py:48] [23700] global_step=23700, grad_norm=4.901803016662598, loss=3.9208598136901855 -I0513 01:02:48.310311 140330640848640 logging_writer.py:48] [23800] global_step=23800, grad_norm=3.2840914726257324, loss=3.897930145263672 -I0513 01:03:25.144976 140330632455936 logging_writer.py:48] [23900] global_step=23900, grad_norm=5.047675132751465, loss=4.082335948944092 -I0513 01:04:02.322563 140330640848640 logging_writer.py:48] [24000] global_step=24000, grad_norm=7.715060234069824, loss=4.089296340942383 -I0513 01:04:39.444877 140330632455936 logging_writer.py:48] [24100] global_step=24100, grad_norm=5.444097995758057, loss=4.100450038909912 -I0513 01:05:16.143984 140330640848640 logging_writer.py:48] [24200] global_step=24200, grad_norm=12.715811729431152, loss=4.072267055511475 -I0513 01:05:53.355273 140330632455936 logging_writer.py:48] [24300] global_step=24300, grad_norm=7.617428779602051, loss=3.9664363861083984 -I0513 01:06:30.341941 140330640848640 logging_writer.py:48] [24400] global_step=24400, grad_norm=34.86245346069336, loss=4.144719123840332 -I0513 01:07:07.127916 140330632455936 logging_writer.py:48] [24500] global_step=24500, grad_norm=3.3269736766815186, loss=3.8677778244018555 -I0513 01:07:44.360513 140330640848640 logging_writer.py:48] [24600] global_step=24600, grad_norm=6.338802814483643, loss=3.9111123085021973 -I0513 01:08:21.467101 140330632455936 logging_writer.py:48] [24700] global_step=24700, grad_norm=4.701180458068848, loss=3.908792018890381 -I0513 01:08:58.108979 140330640848640 logging_writer.py:48] [24800] global_step=24800, grad_norm=4.329361438751221, loss=3.8776957988739014 -I0513 01:09:35.330323 140330632455936 logging_writer.py:48] [24900] global_step=24900, grad_norm=6.150138854980469, loss=3.9630978107452393 -I0513 01:10:12.062304 140330640848640 logging_writer.py:48] [25000] global_step=25000, grad_norm=5.302036285400391, loss=3.8640472888946533 -I0513 01:10:49.061878 140330632455936 logging_writer.py:48] [25100] global_step=25100, grad_norm=12.331409454345703, loss=4.263438701629639 -I0513 01:11:26.268259 140330640848640 logging_writer.py:48] [25200] global_step=25200, grad_norm=6.835732460021973, loss=4.014175891876221 -I0513 01:12:03.411643 140330632455936 logging_writer.py:48] [25300] global_step=25300, grad_norm=8.959190368652344, loss=3.943959951400757 -I0513 01:12:40.162852 140330640848640 logging_writer.py:48] [25400] global_step=25400, grad_norm=7.754523754119873, loss=3.9830777645111084 -I0513 01:13:17.423863 140330632455936 logging_writer.py:48] [25500] global_step=25500, grad_norm=2.975907564163208, loss=3.884788990020752 -I0513 01:13:54.485334 140330640848640 logging_writer.py:48] [25600] global_step=25600, grad_norm=5.631503105163574, loss=4.134166240692139 -I0513 01:14:31.265782 140330632455936 logging_writer.py:48] [25700] global_step=25700, grad_norm=5.440842628479004, loss=3.9994335174560547 -I0513 01:15:08.477118 140330640848640 logging_writer.py:48] [25800] global_step=25800, grad_norm=9.098928451538086, loss=3.7892770767211914 -I0513 01:15:45.512006 140330632455936 logging_writer.py:48] [25900] global_step=25900, grad_norm=5.686288833618164, loss=3.944399356842041 -I0513 01:16:22.312386 140330640848640 logging_writer.py:48] [26000] global_step=26000, grad_norm=12.45071029663086, loss=3.830690383911133 -I0513 01:16:59.510936 140330632455936 logging_writer.py:48] [26100] global_step=26100, grad_norm=6.919436454772949, loss=4.021775245666504 -I0513 01:17:36.558934 140330640848640 logging_writer.py:48] [26200] global_step=26200, grad_norm=5.154388904571533, loss=3.9351558685302734 -I0513 01:18:13.375610 140330632455936 logging_writer.py:48] [26300] global_step=26300, grad_norm=3.630779266357422, loss=3.873992443084717 -I0513 01:18:50.613182 140330640848640 logging_writer.py:48] [26400] global_step=26400, grad_norm=10.270404815673828, loss=4.105071067810059 -I0513 01:19:27.540822 140330632455936 logging_writer.py:48] [26500] global_step=26500, grad_norm=4.810779094696045, loss=3.8887929916381836 -I0513 01:20:04.263509 140330640848640 logging_writer.py:48] [26600] global_step=26600, grad_norm=5.127201557159424, loss=3.8464903831481934 -I0513 01:20:26.345214 140546196993216 spec.py:333] Evaluating on the training split. -I0513 01:20:39.242441 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 01:20:48.657478 140546196993216 spec.py:363] Evaluating on the test split. -I0513 01:20:49.543316 140546196993216 submission_runner.py:516] Time since start: 10213.60s, Step: 26661, {'train/accuracy': Array(0.00147481, dtype=float32), 'train/loss': Array(9.214938, dtype=float32), 'validation/accuracy': Array(0.00156, dtype=float32), 'validation/loss': Array(9.203479, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(9.203677, dtype=float32), 'test/num_examples': 10000, 'score': 10036.045359134674, 'total_duration': 10213.59507727623, 'accumulated_submission_time': 10036.045359134674, 'accumulated_eval_time': 177.0388777256012, 'accumulated_logging_time': 0.28616857528686523} -I0513 01:20:49.581535 140330632455936 logging_writer.py:48] [26661] accumulated_eval_time=177.039, accumulated_logging_time=0.286169, accumulated_submission_time=10036, global_step=26661, preemption_count=0, score=10036, test/accuracy=0.001500000013038516, test/loss=9.2036771774292, test/num_examples=10000, total_duration=10213.6, train/accuracy=0.001474808668717742, train/loss=9.214938163757324, validation/accuracy=0.001560000004246831, validation/loss=9.203478813171387, validation/num_examples=50000 -I0513 01:21:05.631721 140330640848640 logging_writer.py:48] [26700] global_step=26700, grad_norm=5.959621906280518, loss=4.092477798461914 -I0513 01:21:42.675909 140330632455936 logging_writer.py:48] [26800] global_step=26800, grad_norm=5.150821208953857, loss=3.9019198417663574 -I0513 01:22:19.401638 140330640848640 logging_writer.py:48] [26900] global_step=26900, grad_norm=6.179085731506348, loss=3.983262300491333 -I0513 01:22:56.637435 140330632455936 logging_writer.py:48] [27000] global_step=27000, grad_norm=5.658907413482666, loss=3.9820308685302734 -I0513 01:23:33.631447 140330640848640 logging_writer.py:48] [27100] global_step=27100, grad_norm=8.318817138671875, loss=4.019503593444824 -I0513 01:24:10.397252 140330632455936 logging_writer.py:48] [27200] global_step=27200, grad_norm=6.774874687194824, loss=4.122509002685547 -I0513 01:24:47.672466 140330640848640 logging_writer.py:48] [27300] global_step=27300, grad_norm=4.2716569900512695, loss=3.9145846366882324 -I0513 01:25:24.764015 140330632455936 logging_writer.py:48] [27400] global_step=27400, grad_norm=7.361724853515625, loss=4.194630146026611 -I0513 01:26:01.525138 140330640848640 logging_writer.py:48] [27500] global_step=27500, grad_norm=5.072947978973389, loss=3.8473644256591797 -I0513 01:26:38.728860 140330632455936 logging_writer.py:48] [27600] global_step=27600, grad_norm=4.327601909637451, loss=3.9479808807373047 -I0513 01:27:15.505667 140330640848640 logging_writer.py:48] [27700] global_step=27700, grad_norm=4.496596813201904, loss=4.020790100097656 -I0513 01:27:52.467814 140330632455936 logging_writer.py:48] [27800] global_step=27800, grad_norm=3.777169704437256, loss=3.898707866668701 -I0513 01:28:29.818157 140330640848640 logging_writer.py:48] [27900] global_step=27900, grad_norm=4.4086103439331055, loss=3.791557550430298 -I0513 01:29:06.946410 140330632455936 logging_writer.py:48] [28000] global_step=28000, grad_norm=8.001785278320312, loss=3.8547556400299072 -I0513 01:29:43.667924 140330640848640 logging_writer.py:48] [28100] global_step=28100, grad_norm=4.09814453125, loss=4.063959121704102 -I0513 01:30:20.867007 140330632455936 logging_writer.py:48] [28200] global_step=28200, grad_norm=5.188048839569092, loss=3.9182615280151367 -I0513 01:30:57.580360 140330640848640 logging_writer.py:48] [28300] global_step=28300, grad_norm=6.277266025543213, loss=3.912248134613037 -I0513 01:31:34.627861 140330632455936 logging_writer.py:48] [28400] global_step=28400, grad_norm=5.167067050933838, loss=3.8868422508239746 -I0513 01:32:11.884458 140330640848640 logging_writer.py:48] [28500] global_step=28500, grad_norm=5.705763339996338, loss=3.7813003063201904 -I0513 01:32:48.648631 140330632455936 logging_writer.py:48] [28600] global_step=28600, grad_norm=10.778225898742676, loss=4.391670227050781 -I0513 01:33:25.681180 140330640848640 logging_writer.py:48] [28700] global_step=28700, grad_norm=9.745854377746582, loss=4.0737481117248535 -I0513 01:34:02.886937 140330632455936 logging_writer.py:48] [28800] global_step=28800, grad_norm=5.042022705078125, loss=3.9458041191101074 -I0513 01:34:39.557746 140330640848640 logging_writer.py:48] [28900] global_step=28900, grad_norm=5.434535503387451, loss=4.0171380043029785 -I0513 01:35:16.749714 140330632455936 logging_writer.py:48] [29000] global_step=29000, grad_norm=5.016842842102051, loss=3.826582670211792 -I0513 01:35:54.002449 140330640848640 logging_writer.py:48] [29100] global_step=29100, grad_norm=11.051190376281738, loss=3.8809945583343506 -I0513 01:36:30.668571 140330632455936 logging_writer.py:48] [29200] global_step=29200, grad_norm=4.100366592407227, loss=3.8374457359313965 -I0513 01:37:07.674517 140330640848640 logging_writer.py:48] [29300] global_step=29300, grad_norm=4.071099281311035, loss=3.835803270339966 -I0513 01:37:44.849326 140330632455936 logging_writer.py:48] [29400] global_step=29400, grad_norm=3.488067626953125, loss=3.837186813354492 -I0513 01:38:21.551241 140330640848640 logging_writer.py:48] [29500] global_step=29500, grad_norm=9.52908706665039, loss=4.1671247482299805 -I0513 01:38:58.726649 140330632455936 logging_writer.py:48] [29600] global_step=29600, grad_norm=7.605472087860107, loss=4.119682312011719 -I0513 01:39:35.913287 140330640848640 logging_writer.py:48] [29700] global_step=29700, grad_norm=4.77837610244751, loss=4.027796268463135 -I0513 01:40:12.594791 140330632455936 logging_writer.py:48] [29800] global_step=29800, grad_norm=3.3929550647735596, loss=3.9394233226776123 -I0513 01:40:49.653905 140330640848640 logging_writer.py:48] [29900] global_step=29900, grad_norm=3.6783909797668457, loss=3.9557762145996094 -I0513 01:41:26.911340 140330632455936 logging_writer.py:48] [30000] global_step=30000, grad_norm=3.3452789783477783, loss=3.8883354663848877 -I0513 01:42:03.701367 140330640848640 logging_writer.py:48] [30100] global_step=30100, grad_norm=6.906122207641602, loss=3.9582860469818115 -I0513 01:42:40.747853 140330632455936 logging_writer.py:48] [30200] global_step=30200, grad_norm=4.670402526855469, loss=3.855644464492798 -I0513 01:43:17.964844 140330640848640 logging_writer.py:48] [30300] global_step=30300, grad_norm=5.804624557495117, loss=3.8618998527526855 -I0513 01:43:54.779825 140330632455936 logging_writer.py:48] [30400] global_step=30400, grad_norm=5.695254802703857, loss=3.794842481613159 -I0513 01:44:31.799379 140330640848640 logging_writer.py:48] [30500] global_step=30500, grad_norm=6.7588348388671875, loss=3.8469481468200684 -I0513 01:45:09.096083 140330632455936 logging_writer.py:48] [30600] global_step=30600, grad_norm=8.10305404663086, loss=4.009036064147949 -I0513 01:45:45.835278 140330640848640 logging_writer.py:48] [30700] global_step=30700, grad_norm=4.64174222946167, loss=3.8836963176727295 -I0513 01:46:22.883608 140330632455936 logging_writer.py:48] [30800] global_step=30800, grad_norm=6.360194206237793, loss=3.9035303592681885 -I0513 01:47:00.105978 140330640848640 logging_writer.py:48] [30900] global_step=30900, grad_norm=13.155695915222168, loss=3.9695382118225098 -I0513 01:47:36.785331 140330632455936 logging_writer.py:48] [31000] global_step=31000, grad_norm=6.139472007751465, loss=3.9184348583221436 -I0513 01:48:13.857661 140330640848640 logging_writer.py:48] [31100] global_step=31100, grad_norm=6.638227939605713, loss=3.9322586059570312 -I0513 01:48:51.071362 140330632455936 logging_writer.py:48] [31200] global_step=31200, grad_norm=5.0092620849609375, loss=3.8185033798217773 -I0513 01:49:27.802181 140330640848640 logging_writer.py:48] [31300] global_step=31300, grad_norm=7.4982099533081055, loss=3.880725145339966 -I0513 01:50:04.892716 140330632455936 logging_writer.py:48] [31400] global_step=31400, grad_norm=5.573868751525879, loss=3.795883893966675 -I0513 01:50:42.039892 140330640848640 logging_writer.py:48] [31500] global_step=31500, grad_norm=3.59294056892395, loss=3.8185696601867676 -I0513 01:51:18.772951 140330632455936 logging_writer.py:48] [31600] global_step=31600, grad_norm=10.701888084411621, loss=3.8235368728637695 -I0513 01:51:55.796439 140330640848640 logging_writer.py:48] [31700] global_step=31700, grad_norm=4.719196319580078, loss=3.9521374702453613 -I0513 01:52:32.969866 140330632455936 logging_writer.py:48] [31800] global_step=31800, grad_norm=135.08714294433594, loss=4.245108604431152 -I0513 01:53:09.664474 140330640848640 logging_writer.py:48] [31900] global_step=31900, grad_norm=8.264283180236816, loss=3.883502960205078 -I0513 01:53:46.748141 140330632455936 logging_writer.py:48] [32000] global_step=32000, grad_norm=5.289907932281494, loss=3.9542880058288574 -I0513 01:54:06.178173 140546196993216 spec.py:333] Evaluating on the training split. -I0513 01:54:18.778635 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 01:54:28.881000 140546196993216 spec.py:363] Evaluating on the test split. -I0513 01:54:29.756125 140546196993216 submission_runner.py:516] Time since start: 12233.81s, Step: 32053, {'train/accuracy': Array(0.00143495, dtype=float32), 'train/loss': Array(8.892026, dtype=float32), 'validation/accuracy': Array(0.00172, dtype=float32), 'validation/loss': Array(8.847008, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0014, dtype=float32), 'test/loss': Array(8.848667, dtype=float32), 'test/num_examples': 10000, 'score': 12032.575147390366, 'total_duration': 12233.807253837585, 'accumulated_submission_time': 12032.575147390366, 'accumulated_eval_time': 200.61437606811523, 'accumulated_logging_time': 0.3487226963043213} -I0513 01:54:29.794233 140330640848640 logging_writer.py:48] [32053] accumulated_eval_time=200.614, accumulated_logging_time=0.348723, accumulated_submission_time=12032.6, global_step=32053, preemption_count=0, score=12032.6, test/accuracy=0.0014000000664964318, test/loss=8.84866714477539, test/num_examples=10000, total_duration=12233.8, train/accuracy=0.001434948993846774, train/loss=8.8920259475708, validation/accuracy=0.00171999994199723, validation/loss=8.847007751464844, validation/num_examples=50000 -I0513 01:54:48.129840 140330632455936 logging_writer.py:48] [32100] global_step=32100, grad_norm=7.591610908508301, loss=3.8695662021636963 -I0513 01:55:24.827170 140330640848640 logging_writer.py:48] [32200] global_step=32200, grad_norm=6.857138156890869, loss=3.9086103439331055 -I0513 01:56:01.854206 140330632455936 logging_writer.py:48] [32300] global_step=32300, grad_norm=5.195123195648193, loss=3.973390817642212 -I0513 01:56:38.989147 140330640848640 logging_writer.py:48] [32400] global_step=32400, grad_norm=4.338387966156006, loss=3.8967554569244385 -I0513 01:57:15.766377 140330632455936 logging_writer.py:48] [32500] global_step=32500, grad_norm=5.395411014556885, loss=3.8643743991851807 -I0513 01:57:52.828648 140330640848640 logging_writer.py:48] [32600] global_step=32600, grad_norm=10.146530151367188, loss=3.927581787109375 -I0513 01:58:30.020979 140330632455936 logging_writer.py:48] [32700] global_step=32700, grad_norm=4.221452236175537, loss=3.789698600769043 -I0513 01:59:06.730550 140330640848640 logging_writer.py:48] [32800] global_step=32800, grad_norm=3.669297218322754, loss=3.7686641216278076 -I0513 01:59:43.730797 140330632455936 logging_writer.py:48] [32900] global_step=32900, grad_norm=33.26420974731445, loss=3.8806402683258057 -I0513 02:00:20.885241 140330640848640 logging_writer.py:48] [33000] global_step=33000, grad_norm=3.8690147399902344, loss=3.7900948524475098 -I0513 02:00:57.624024 140330632455936 logging_writer.py:48] [33100] global_step=33100, grad_norm=5.126633644104004, loss=3.7427845001220703 -I0513 02:01:34.698094 140330640848640 logging_writer.py:48] [33200] global_step=33200, grad_norm=3.991867780685425, loss=3.7234888076782227 -I0513 02:02:11.823081 140330632455936 logging_writer.py:48] [33300] global_step=33300, grad_norm=3.6903114318847656, loss=3.8621559143066406 -I0513 02:02:48.964571 140330640848640 logging_writer.py:48] [33400] global_step=33400, grad_norm=5.7082037925720215, loss=3.8487424850463867 -I0513 02:03:25.856041 140330632455936 logging_writer.py:48] [33500] global_step=33500, grad_norm=8.163433074951172, loss=3.9448986053466797 -I0513 02:04:03.488718 140330640848640 logging_writer.py:48] [33600] global_step=33600, grad_norm=6.869625568389893, loss=3.9603540897369385 -I0513 02:04:40.369583 140330632455936 logging_writer.py:48] [33700] global_step=33700, grad_norm=7.287048816680908, loss=3.9410109519958496 -I0513 02:05:17.629226 140330640848640 logging_writer.py:48] [33800] global_step=33800, grad_norm=3.119192600250244, loss=3.786210536956787 -I0513 02:05:55.092685 140330632455936 logging_writer.py:48] [33900] global_step=33900, grad_norm=2.610368490219116, loss=3.7794127464294434 -I0513 02:06:31.967429 140330640848640 logging_writer.py:48] [34000] global_step=34000, grad_norm=7.537170886993408, loss=3.93512225151062 -I0513 02:07:08.900459 140330632455936 logging_writer.py:48] [34100] global_step=34100, grad_norm=4.395468235015869, loss=3.6783266067504883 -I0513 02:07:46.525833 140330640848640 logging_writer.py:48] [34200] global_step=34200, grad_norm=4.27325439453125, loss=3.968123435974121 -I0513 02:08:23.400884 140330632455936 logging_writer.py:48] [34300] global_step=34300, grad_norm=5.303800582885742, loss=3.870522975921631 -I0513 02:09:00.707727 140330640848640 logging_writer.py:48] [34400] global_step=34400, grad_norm=3.087153673171997, loss=3.6702046394348145 -I0513 02:09:38.150902 140330632455936 logging_writer.py:48] [34500] global_step=34500, grad_norm=4.371456146240234, loss=3.738798141479492 -I0513 02:10:15.083504 140330640848640 logging_writer.py:48] [34600] global_step=34600, grad_norm=4.3779802322387695, loss=3.897418260574341 -I0513 02:10:51.938863 140330632455936 logging_writer.py:48] [34700] global_step=34700, grad_norm=6.71708345413208, loss=3.919844150543213 -I0513 02:11:29.629474 140330640848640 logging_writer.py:48] [34800] global_step=34800, grad_norm=5.078333854675293, loss=3.7783455848693848 -I0513 02:12:06.531444 140330632455936 logging_writer.py:48] [34900] global_step=34900, grad_norm=6.995848655700684, loss=3.9151248931884766 -I0513 02:12:43.376316 140330640848640 logging_writer.py:48] [35000] global_step=35000, grad_norm=6.569180488586426, loss=3.9535412788391113 -I0513 02:13:21.021699 140330632455936 logging_writer.py:48] [35100] global_step=35100, grad_norm=5.553155422210693, loss=3.7505807876586914 -I0513 02:13:57.901180 140330640848640 logging_writer.py:48] [35200] global_step=35200, grad_norm=4.052000999450684, loss=3.797654867172241 -I0513 02:14:34.727900 140330632455936 logging_writer.py:48] [35300] global_step=35300, grad_norm=8.910616874694824, loss=3.8613529205322266 -I0513 02:15:12.508833 140330640848640 logging_writer.py:48] [35400] global_step=35400, grad_norm=5.354469299316406, loss=3.718710422515869 -I0513 02:15:49.399018 140330632455936 logging_writer.py:48] [35500] global_step=35500, grad_norm=5.268887042999268, loss=3.851135015487671 -I0513 02:16:26.244835 140330640848640 logging_writer.py:48] [35600] global_step=35600, grad_norm=6.603230953216553, loss=3.7419064044952393 -I0513 02:17:03.898806 140330632455936 logging_writer.py:48] [35700] global_step=35700, grad_norm=8.008224487304688, loss=3.9110960960388184 -I0513 02:17:40.818940 140330640848640 logging_writer.py:48] [35800] global_step=35800, grad_norm=5.127634525299072, loss=3.8945794105529785 -I0513 02:18:17.740148 140330632455936 logging_writer.py:48] [35900] global_step=35900, grad_norm=4.422190189361572, loss=3.902697801589966 -I0513 02:18:55.388470 140330640848640 logging_writer.py:48] [36000] global_step=36000, grad_norm=6.402425289154053, loss=3.970759868621826 -I0513 02:19:32.228447 140330632455936 logging_writer.py:48] [36100] global_step=36100, grad_norm=5.630108833312988, loss=3.815640449523926 -I0513 02:20:09.194781 140330640848640 logging_writer.py:48] [36200] global_step=36200, grad_norm=3.0079281330108643, loss=3.899364471435547 -I0513 02:20:46.610500 140330632455936 logging_writer.py:48] [36300] global_step=36300, grad_norm=10.671001434326172, loss=3.840017557144165 -I0513 02:21:23.768990 140330640848640 logging_writer.py:48] [36400] global_step=36400, grad_norm=3.9214088916778564, loss=3.7437782287597656 -I0513 02:22:00.553886 140330632455936 logging_writer.py:48] [36500] global_step=36500, grad_norm=6.950275421142578, loss=3.8418235778808594 -I0513 02:22:38.253174 140330640848640 logging_writer.py:48] [36600] global_step=36600, grad_norm=7.770075798034668, loss=3.9480373859405518 -I0513 02:23:15.032374 140330632455936 logging_writer.py:48] [36700] global_step=36700, grad_norm=4.585089683532715, loss=3.8245913982391357 -I0513 02:23:51.938949 140330640848640 logging_writer.py:48] [36800] global_step=36800, grad_norm=6.174434185028076, loss=3.887665033340454 -I0513 02:24:29.584218 140330632455936 logging_writer.py:48] [36900] global_step=36900, grad_norm=6.223568439483643, loss=3.817554473876953 -I0513 02:25:06.407203 140330640848640 logging_writer.py:48] [37000] global_step=37000, grad_norm=5.6397552490234375, loss=3.755741596221924 -I0513 02:25:43.280507 140330632455936 logging_writer.py:48] [37100] global_step=37100, grad_norm=5.175654411315918, loss=3.8452529907226562 -I0513 02:26:20.926700 140330640848640 logging_writer.py:48] [37200] global_step=37200, grad_norm=5.509775161743164, loss=3.7919764518737793 -I0513 02:26:57.753969 140330632455936 logging_writer.py:48] [37300] global_step=37300, grad_norm=5.152783393859863, loss=3.7999765872955322 -I0513 02:27:34.722686 140330640848640 logging_writer.py:48] [37400] global_step=37400, grad_norm=5.322890758514404, loss=3.828697443008423 -I0513 02:27:45.840445 140546196993216 spec.py:333] Evaluating on the training split. -I0513 02:27:58.327883 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 02:28:07.809023 140546196993216 spec.py:363] Evaluating on the test split. -I0513 02:28:08.696988 140546196993216 submission_runner.py:516] Time since start: 14252.75s, Step: 37430, {'train/accuracy': Array(0.0015346, dtype=float32), 'train/loss': Array(8.616239, dtype=float32), 'validation/accuracy': Array(0.00152, dtype=float32), 'validation/loss': Array(8.605657, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(8.609212, dtype=float32), 'test/num_examples': 10000, 'score': 14028.56815481186, 'total_duration': 14252.748258113861, 'accumulated_submission_time': 14028.56815481186, 'accumulated_eval_time': 223.46860146522522, 'accumulated_logging_time': 0.39777278900146484} -I0513 02:28:08.741512 140330632455936 logging_writer.py:48] [37430] accumulated_eval_time=223.469, accumulated_logging_time=0.397773, accumulated_submission_time=14028.6, global_step=37430, preemption_count=0, score=14028.6, test/accuracy=0.001500000013038516, test/loss=8.609211921691895, test/num_examples=10000, total_duration=14252.7, train/accuracy=0.0015345981810241938, train/loss=8.616238594055176, validation/accuracy=0.0015199999324977398, validation/loss=8.605656623840332, validation/num_examples=50000 -I0513 02:28:36.331962 140330640848640 logging_writer.py:48] [37500] global_step=37500, grad_norm=5.6014604568481445, loss=3.831350326538086 -I0513 02:29:13.083176 140330632455936 logging_writer.py:48] [37600] global_step=37600, grad_norm=4.705032825469971, loss=3.84816312789917 -I0513 02:29:49.894394 140330640848640 logging_writer.py:48] [37700] global_step=37700, grad_norm=6.8924689292907715, loss=3.6450440883636475 -I0513 02:30:27.534926 140330632455936 logging_writer.py:48] [37800] global_step=37800, grad_norm=4.984471797943115, loss=3.717393398284912 -I0513 02:31:04.423434 140330640848640 logging_writer.py:48] [37900] global_step=37900, grad_norm=7.033219814300537, loss=3.9073634147644043 -I0513 02:31:41.212486 140330632455936 logging_writer.py:48] [38000] global_step=38000, grad_norm=6.5689921379089355, loss=3.983156681060791 -I0513 02:32:18.815451 140330640848640 logging_writer.py:48] [38100] global_step=38100, grad_norm=5.034747123718262, loss=3.7493090629577637 -I0513 02:32:55.565728 140330632455936 logging_writer.py:48] [38200] global_step=38200, grad_norm=6.7669758796691895, loss=3.8459019660949707 -I0513 02:33:32.367430 140330640848640 logging_writer.py:48] [38300] global_step=38300, grad_norm=3.7803938388824463, loss=3.710801124572754 -I0513 02:34:09.911254 140330632455936 logging_writer.py:48] [38400] global_step=38400, grad_norm=4.603743076324463, loss=3.7370963096618652 -I0513 02:34:46.691868 140330640848640 logging_writer.py:48] [38500] global_step=38500, grad_norm=3.151135206222534, loss=3.5548014640808105 -I0513 02:35:23.519416 140330632455936 logging_writer.py:48] [38600] global_step=38600, grad_norm=4.93752908706665, loss=3.9110605716705322 -I0513 02:36:01.202512 140330640848640 logging_writer.py:48] [38700] global_step=38700, grad_norm=5.062730312347412, loss=3.7136924266815186 -I0513 02:36:38.003779 140330632455936 logging_writer.py:48] [38800] global_step=38800, grad_norm=3.458925724029541, loss=3.739170551300049 -I0513 02:37:14.856435 140330640848640 logging_writer.py:48] [38900] global_step=38900, grad_norm=6.55255126953125, loss=3.8391783237457275 -I0513 02:37:52.415570 140330632455936 logging_writer.py:48] [39000] global_step=39000, grad_norm=4.789771556854248, loss=3.7391197681427 -I0513 02:38:29.184994 140330640848640 logging_writer.py:48] [39100] global_step=39100, grad_norm=4.056324481964111, loss=3.7620627880096436 -I0513 02:39:06.049758 140330632455936 logging_writer.py:48] [39200] global_step=39200, grad_norm=14.808853149414062, loss=3.654914140701294 -I0513 02:39:43.698987 140330640848640 logging_writer.py:48] [39300] global_step=39300, grad_norm=3.6112406253814697, loss=3.8304412364959717 -I0513 02:40:20.470168 140330632455936 logging_writer.py:48] [39400] global_step=39400, grad_norm=23.42954444885254, loss=3.827974319458008 -I0513 02:40:57.312403 140330640848640 logging_writer.py:48] [39500] global_step=39500, grad_norm=3.535881519317627, loss=3.617460250854492 -I0513 02:41:34.968295 140330632455936 logging_writer.py:48] [39600] global_step=39600, grad_norm=4.3819499015808105, loss=3.746717929840088 -I0513 02:42:11.746450 140330640848640 logging_writer.py:48] [39700] global_step=39700, grad_norm=3.687037229537964, loss=3.753664970397949 -I0513 02:42:48.487557 140330632455936 logging_writer.py:48] [39800] global_step=39800, grad_norm=14.768377304077148, loss=3.8182902336120605 -I0513 02:43:26.093868 140330640848640 logging_writer.py:48] [39900] global_step=39900, grad_norm=4.571193218231201, loss=3.784554958343506 -I0513 02:44:02.864657 140330632455936 logging_writer.py:48] [40000] global_step=40000, grad_norm=5.682659149169922, loss=3.836113452911377 -I0513 02:44:39.687206 140330640848640 logging_writer.py:48] [40100] global_step=40100, grad_norm=2.5467467308044434, loss=3.7267065048217773 -I0513 02:45:17.221132 140330632455936 logging_writer.py:48] [40200] global_step=40200, grad_norm=4.475897312164307, loss=3.7318711280822754 -I0513 02:45:54.142063 140330640848640 logging_writer.py:48] [40300] global_step=40300, grad_norm=27.178619384765625, loss=3.948974609375 -I0513 02:46:30.908896 140330632455936 logging_writer.py:48] [40400] global_step=40400, grad_norm=5.570227146148682, loss=3.945172071456909 -I0513 02:47:08.603105 140330640848640 logging_writer.py:48] [40500] global_step=40500, grad_norm=5.410062313079834, loss=3.85543155670166 -I0513 02:47:45.412245 140330632455936 logging_writer.py:48] [40600] global_step=40600, grad_norm=8.718293190002441, loss=3.6932644844055176 -I0513 02:48:22.272116 140330640848640 logging_writer.py:48] [40700] global_step=40700, grad_norm=5.832205772399902, loss=3.726072311401367 -I0513 02:48:59.885483 140330632455936 logging_writer.py:48] [40800] global_step=40800, grad_norm=4.978761196136475, loss=3.8989200592041016 -I0513 02:49:36.710705 140330640848640 logging_writer.py:48] [40900] global_step=40900, grad_norm=5.143380641937256, loss=3.9319138526916504 -I0513 02:50:13.528226 140330632455936 logging_writer.py:48] [41000] global_step=41000, grad_norm=3.3733279705047607, loss=3.7544047832489014 -I0513 02:50:51.172260 140330640848640 logging_writer.py:48] [41100] global_step=41100, grad_norm=4.638525009155273, loss=3.7489850521087646 -I0513 02:51:28.060606 140330632455936 logging_writer.py:48] [41200] global_step=41200, grad_norm=3.9721176624298096, loss=3.681501865386963 -I0513 02:52:04.917001 140330640848640 logging_writer.py:48] [41300] global_step=41300, grad_norm=4.114132881164551, loss=3.754934310913086 -I0513 02:52:42.448063 140330632455936 logging_writer.py:48] [41400] global_step=41400, grad_norm=3.9547038078308105, loss=3.680047035217285 -I0513 02:53:19.237518 140330640848640 logging_writer.py:48] [41500] global_step=41500, grad_norm=3.8075296878814697, loss=3.8150434494018555 -I0513 02:53:56.099756 140330632455936 logging_writer.py:48] [41600] global_step=41600, grad_norm=5.353863716125488, loss=3.7212414741516113 -I0513 02:54:33.709290 140330640848640 logging_writer.py:48] [41700] global_step=41700, grad_norm=8.78135871887207, loss=4.0569987297058105 -I0513 02:55:10.534381 140330632455936 logging_writer.py:48] [41800] global_step=41800, grad_norm=3.9678258895874023, loss=3.9165444374084473 -I0513 02:55:47.375991 140330640848640 logging_writer.py:48] [41900] global_step=41900, grad_norm=5.200408935546875, loss=3.8966636657714844 -I0513 02:56:25.078623 140330632455936 logging_writer.py:48] [42000] global_step=42000, grad_norm=5.744967460632324, loss=3.7625110149383545 -I0513 02:57:01.852958 140330640848640 logging_writer.py:48] [42100] global_step=42100, grad_norm=4.745189189910889, loss=3.6860382556915283 -I0513 02:57:39.030864 140330632455936 logging_writer.py:48] [42200] global_step=42200, grad_norm=4.544768333435059, loss=3.729149341583252 -I0513 02:58:16.298225 140330640848640 logging_writer.py:48] [42300] global_step=42300, grad_norm=4.93120813369751, loss=3.7591326236724854 -I0513 02:58:53.159935 140330632455936 logging_writer.py:48] [42400] global_step=42400, grad_norm=3.7140719890594482, loss=3.5845320224761963 -I0513 02:59:30.001179 140330640848640 logging_writer.py:48] [42500] global_step=42500, grad_norm=4.775661945343018, loss=3.661674737930298 -I0513 03:00:07.553593 140330632455936 logging_writer.py:48] [42600] global_step=42600, grad_norm=3.607151508331299, loss=3.5559768676757812 -I0513 03:00:44.448067 140330640848640 logging_writer.py:48] [42700] global_step=42700, grad_norm=6.350673198699951, loss=3.9236867427825928 -I0513 03:01:21.631183 140330632455936 logging_writer.py:48] [42800] global_step=42800, grad_norm=4.080056190490723, loss=3.7406718730926514 -I0513 03:01:24.883943 140546196993216 spec.py:333] Evaluating on the training split. -I0513 03:01:36.689038 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 03:01:46.161758 140546196993216 spec.py:363] Evaluating on the test split. -I0513 03:01:47.040333 140546196993216 submission_runner.py:516] Time since start: 16271.09s, Step: 42810, {'train/accuracy': Array(0.00161432, dtype=float32), 'train/loss': Array(8.523104, dtype=float32), 'validation/accuracy': Array(0.00164, dtype=float32), 'validation/loss': Array(8.457493, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(8.466099, dtype=float32), 'test/num_examples': 10000, 'score': 16024.628784179688, 'total_duration': 16271.091526985168, 'accumulated_submission_time': 16024.628784179688, 'accumulated_eval_time': 245.62259697914124, 'accumulated_logging_time': 0.48175764083862305} -I0513 03:01:47.076057 140330640848640 logging_writer.py:48] [42810] accumulated_eval_time=245.623, accumulated_logging_time=0.481758, accumulated_submission_time=16024.6, global_step=42810, preemption_count=0, score=16024.6, test/accuracy=0.0013000001199543476, test/loss=8.46609878540039, test/num_examples=10000, total_duration=16271.1, train/accuracy=0.0016143175307661295, train/loss=8.523103713989258, validation/accuracy=0.0016399999149143696, validation/loss=8.45749282836914, validation/num_examples=50000 -I0513 03:02:22.211360 140330632455936 logging_writer.py:48] [42900] global_step=42900, grad_norm=3.2333590984344482, loss=3.7912869453430176 -I0513 03:02:58.969739 140330640848640 logging_writer.py:48] [43000] global_step=43000, grad_norm=8.144944190979004, loss=3.779796600341797 -I0513 03:03:35.760586 140330632455936 logging_writer.py:48] [43100] global_step=43100, grad_norm=6.860063076019287, loss=3.8046460151672363 -I0513 03:04:13.354977 140330640848640 logging_writer.py:48] [43200] global_step=43200, grad_norm=5.361058235168457, loss=3.7727441787719727 -I0513 03:04:50.187433 140330632455936 logging_writer.py:48] [43300] global_step=43300, grad_norm=7.4865403175354, loss=3.7987003326416016 -I0513 03:05:27.084983 140330640848640 logging_writer.py:48] [43400] global_step=43400, grad_norm=4.070934295654297, loss=3.7418737411499023 -I0513 03:06:04.735471 140330632455936 logging_writer.py:48] [43500] global_step=43500, grad_norm=5.784102916717529, loss=3.8518972396850586 -I0513 03:06:41.529753 140330640848640 logging_writer.py:48] [43600] global_step=43600, grad_norm=6.771197319030762, loss=3.88135027885437 -I0513 03:07:18.446572 140330632455936 logging_writer.py:48] [43700] global_step=43700, grad_norm=4.172581195831299, loss=3.6968984603881836 -I0513 03:07:56.069529 140330640848640 logging_writer.py:48] [43800] global_step=43800, grad_norm=6.011769771575928, loss=3.7050116062164307 -I0513 03:08:32.842399 140330632455936 logging_writer.py:48] [43900] global_step=43900, grad_norm=6.12124490737915, loss=3.7749640941619873 -I0513 03:09:09.652272 140330640848640 logging_writer.py:48] [44000] global_step=44000, grad_norm=2.862607479095459, loss=3.7536752223968506 -I0513 03:09:47.273900 140330632455936 logging_writer.py:48] [44100] global_step=44100, grad_norm=3.684569835662842, loss=3.8123459815979004 -I0513 03:10:24.324623 140330640848640 logging_writer.py:48] [44200] global_step=44200, grad_norm=4.879319667816162, loss=3.7568652629852295 -I0513 03:11:01.240334 140330632455936 logging_writer.py:48] [44300] global_step=44300, grad_norm=6.3242902755737305, loss=3.867445230484009 -I0513 03:11:38.854880 140330640848640 logging_writer.py:48] [44400] global_step=44400, grad_norm=3.4310786724090576, loss=3.7691636085510254 -I0513 03:12:15.680207 140330632455936 logging_writer.py:48] [44500] global_step=44500, grad_norm=6.1671223640441895, loss=3.7819504737854004 -I0513 03:12:52.466582 140330640848640 logging_writer.py:48] [44600] global_step=44600, grad_norm=6.7096405029296875, loss=3.6179826259613037 -I0513 03:13:29.991282 140330632455936 logging_writer.py:48] [44700] global_step=44700, grad_norm=4.375880718231201, loss=3.77047061920166 -I0513 03:14:06.862046 140330640848640 logging_writer.py:48] [44800] global_step=44800, grad_norm=6.274622440338135, loss=3.7718989849090576 -I0513 03:14:43.651967 140330632455936 logging_writer.py:48] [44900] global_step=44900, grad_norm=5.318879127502441, loss=3.911259174346924 -I0513 03:15:21.384747 140330640848640 logging_writer.py:48] [45000] global_step=45000, grad_norm=4.7372870445251465, loss=3.693368434906006 -I0513 03:15:58.153243 140330632455936 logging_writer.py:48] [45100] global_step=45100, grad_norm=5.210596084594727, loss=3.843363046646118 -I0513 03:16:34.922440 140330640848640 logging_writer.py:48] [45200] global_step=45200, grad_norm=5.880410194396973, loss=3.6363155841827393 -I0513 03:17:12.291059 140330632455936 logging_writer.py:48] [45300] global_step=45300, grad_norm=4.370427131652832, loss=3.7357852458953857 -I0513 03:17:49.395169 140330640848640 logging_writer.py:48] [45400] global_step=45400, grad_norm=3.740309000015259, loss=3.7522785663604736 -I0513 03:18:26.242725 140330632455936 logging_writer.py:48] [45500] global_step=45500, grad_norm=7.485471248626709, loss=3.802645683288574 -I0513 03:19:03.823009 140330640848640 logging_writer.py:48] [45600] global_step=45600, grad_norm=6.412841320037842, loss=3.799297332763672 -I0513 03:19:40.603659 140330632455936 logging_writer.py:48] [45700] global_step=45700, grad_norm=5.777948379516602, loss=3.8500192165374756 -I0513 03:20:17.438914 140330640848640 logging_writer.py:48] [45800] global_step=45800, grad_norm=3.458433151245117, loss=3.6915454864501953 -I0513 03:20:54.776211 140330632455936 logging_writer.py:48] [45900] global_step=45900, grad_norm=4.392711162567139, loss=3.7696352005004883 -I0513 03:21:31.955363 140330640848640 logging_writer.py:48] [46000] global_step=46000, grad_norm=3.6281676292419434, loss=3.6314585208892822 -I0513 03:22:08.832176 140330632455936 logging_writer.py:48] [46100] global_step=46100, grad_norm=4.875340461730957, loss=3.9160802364349365 -I0513 03:22:46.001650 140330640848640 logging_writer.py:48] [46200] global_step=46200, grad_norm=8.391310691833496, loss=3.898752450942993 -I0513 03:23:22.990599 140330632455936 logging_writer.py:48] [46300] global_step=46300, grad_norm=7.641426086425781, loss=3.8653035163879395 -I0513 03:23:59.816308 140330640848640 logging_writer.py:48] [46400] global_step=46400, grad_norm=2.728656053543091, loss=3.704878807067871 -I0513 03:24:37.104410 140330632455936 logging_writer.py:48] [46500] global_step=46500, grad_norm=3.2073116302490234, loss=3.570653200149536 -I0513 03:25:14.211648 140330640848640 logging_writer.py:48] [46600] global_step=46600, grad_norm=7.368955612182617, loss=3.9729738235473633 -I0513 03:25:51.043834 140330632455936 logging_writer.py:48] [46700] global_step=46700, grad_norm=3.8383431434631348, loss=3.663578510284424 -I0513 03:26:28.315768 140330640848640 logging_writer.py:48] [46800] global_step=46800, grad_norm=4.545426368713379, loss=3.7427725791931152 -I0513 03:27:05.452882 140330632455936 logging_writer.py:48] [46900] global_step=46900, grad_norm=4.4640679359436035, loss=3.812859058380127 -I0513 03:27:42.304315 140330640848640 logging_writer.py:48] [47000] global_step=47000, grad_norm=4.4735636711120605, loss=3.772693157196045 -I0513 03:28:19.604373 140330632455936 logging_writer.py:48] [47100] global_step=47100, grad_norm=4.298505783081055, loss=3.76556396484375 -I0513 03:28:56.752434 140330640848640 logging_writer.py:48] [47200] global_step=47200, grad_norm=3.9281539916992188, loss=3.7812399864196777 -I0513 03:29:33.537243 140330632455936 logging_writer.py:48] [47300] global_step=47300, grad_norm=4.241396903991699, loss=3.6908483505249023 -I0513 03:30:10.852756 140330640848640 logging_writer.py:48] [47400] global_step=47400, grad_norm=4.854413986206055, loss=3.729196548461914 -I0513 03:30:48.005179 140330632455936 logging_writer.py:48] [47500] global_step=47500, grad_norm=9.734720230102539, loss=3.8846559524536133 -I0513 03:31:24.900112 140330640848640 logging_writer.py:48] [47600] global_step=47600, grad_norm=6.39498233795166, loss=3.775496482849121 -I0513 03:32:02.216276 140330632455936 logging_writer.py:48] [47700] global_step=47700, grad_norm=2.965461492538452, loss=3.777952194213867 -I0513 03:32:39.375244 140330640848640 logging_writer.py:48] [47800] global_step=47800, grad_norm=4.183248043060303, loss=3.7936105728149414 -I0513 03:33:16.168187 140330632455936 logging_writer.py:48] [47900] global_step=47900, grad_norm=4.523298740386963, loss=3.857405185699463 -I0513 03:33:53.407140 140330640848640 logging_writer.py:48] [48000] global_step=48000, grad_norm=5.1494832038879395, loss=3.505331039428711 -I0513 03:34:30.612990 140330632455936 logging_writer.py:48] [48100] global_step=48100, grad_norm=6.391189098358154, loss=3.8798933029174805 -I0513 03:35:03.549628 140546196993216 spec.py:333] Evaluating on the training split. -I0513 03:35:15.130419 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 03:35:24.632349 140546196993216 spec.py:363] Evaluating on the test split. -I0513 03:35:25.513559 140546196993216 submission_runner.py:516] Time since start: 18289.56s, Step: 48190, {'train/accuracy': Array(0.00177376, dtype=float32), 'train/loss': Array(8.355035, dtype=float32), 'validation/accuracy': Array(0.00138, dtype=float32), 'validation/loss': Array(8.319946, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(8.331606, dtype=float32), 'test/num_examples': 10000, 'score': 18021.00621008873, 'total_duration': 18289.564914941788, 'accumulated_submission_time': 18021.00621008873, 'accumulated_eval_time': 267.5842914581299, 'accumulated_logging_time': 0.5713906288146973} -I0513 03:35:25.561599 140330640848640 logging_writer.py:48] [48190] accumulated_eval_time=267.584, accumulated_logging_time=0.571391, accumulated_submission_time=18021, global_step=48190, preemption_count=0, score=18021, test/accuracy=0.001500000013038516, test/loss=8.331605911254883, test/num_examples=10000, total_duration=18289.6, train/accuracy=0.0017737563466653228, train/loss=8.355034828186035, validation/accuracy=0.0013799999142065644, validation/loss=8.3199462890625, validation/num_examples=50000 -I0513 03:35:30.551396 140330632455936 logging_writer.py:48] [48200] global_step=48200, grad_norm=3.9567606449127197, loss=3.7370500564575195 -I0513 03:36:08.046290 140330640848640 logging_writer.py:48] [48300] global_step=48300, grad_norm=4.739984512329102, loss=3.9334893226623535 -I0513 03:36:45.237485 140330632455936 logging_writer.py:48] [48400] global_step=48400, grad_norm=4.1335248947143555, loss=3.669086456298828 -I0513 03:37:22.141043 140330640848640 logging_writer.py:48] [48500] global_step=48500, grad_norm=4.190154552459717, loss=3.6990604400634766 -I0513 03:37:59.561124 140330632455936 logging_writer.py:48] [48600] global_step=48600, grad_norm=5.445068836212158, loss=3.5764153003692627 -I0513 03:38:36.748731 140330640848640 logging_writer.py:48] [48700] global_step=48700, grad_norm=5.5197601318359375, loss=3.7048001289367676 -I0513 03:39:13.680372 140330632455936 logging_writer.py:48] [48800] global_step=48800, grad_norm=4.202643871307373, loss=3.655001640319824 -I0513 03:39:51.066021 140330640848640 logging_writer.py:48] [48900] global_step=48900, grad_norm=4.585452079772949, loss=3.668020009994507 -I0513 03:40:28.256146 140330632455936 logging_writer.py:48] [49000] global_step=49000, grad_norm=5.417252063751221, loss=3.8134169578552246 -I0513 03:41:05.075256 140330640848640 logging_writer.py:48] [49100] global_step=49100, grad_norm=3.867770195007324, loss=3.7248406410217285 -I0513 03:41:42.628315 140330632455936 logging_writer.py:48] [49200] global_step=49200, grad_norm=6.149771213531494, loss=3.9360222816467285 -I0513 03:42:19.429530 140330640848640 logging_writer.py:48] [49300] global_step=49300, grad_norm=3.756166934967041, loss=3.633617639541626 -I0513 03:42:56.284980 140330632455936 logging_writer.py:48] [49400] global_step=49400, grad_norm=3.7987117767333984, loss=3.779092788696289 -I0513 03:43:33.565469 140330640848640 logging_writer.py:48] [49500] global_step=49500, grad_norm=3.8745131492614746, loss=3.572293758392334 -I0513 03:44:10.605837 140330632455936 logging_writer.py:48] [49600] global_step=49600, grad_norm=4.035453796386719, loss=3.5715646743774414 -I0513 03:44:47.530743 140330640848640 logging_writer.py:48] [49700] global_step=49700, grad_norm=7.4084954261779785, loss=3.8669395446777344 -I0513 03:45:25.189486 140330632455936 logging_writer.py:48] [49800] global_step=49800, grad_norm=6.317180156707764, loss=3.9559760093688965 -I0513 03:46:01.945509 140330640848640 logging_writer.py:48] [49900] global_step=49900, grad_norm=4.68111515045166, loss=3.795253038406372 -I0513 03:46:38.803411 140330632455936 logging_writer.py:48] [50000] global_step=50000, grad_norm=4.701210021972656, loss=3.756073236465454 -I0513 03:47:16.081220 140330640848640 logging_writer.py:48] [50100] global_step=50100, grad_norm=5.136722564697266, loss=3.6229777336120605 -I0513 03:47:53.203854 140330632455936 logging_writer.py:48] [50200] global_step=50200, grad_norm=3.3418500423431396, loss=3.620741844177246 -I0513 03:48:30.041186 140330640848640 logging_writer.py:48] [50300] global_step=50300, grad_norm=6.2351250648498535, loss=3.8704543113708496 -I0513 03:49:07.409808 140330632455936 logging_writer.py:48] [50400] global_step=50400, grad_norm=4.1189422607421875, loss=3.7074592113494873 -I0513 03:49:44.575235 140330640848640 logging_writer.py:48] [50500] global_step=50500, grad_norm=4.613983631134033, loss=3.64998197555542 -I0513 03:50:21.398972 140330632455936 logging_writer.py:48] [50600] global_step=50600, grad_norm=3.270390510559082, loss=3.683006763458252 -I0513 03:50:58.734962 140330640848640 logging_writer.py:48] [50700] global_step=50700, grad_norm=3.6553051471710205, loss=3.5095953941345215 -I0513 03:51:35.906364 140330632455936 logging_writer.py:48] [50800] global_step=50800, grad_norm=3.1986117362976074, loss=3.647040843963623 -I0513 03:52:12.686429 140330640848640 logging_writer.py:48] [50900] global_step=50900, grad_norm=2.684603214263916, loss=3.5843653678894043 -I0513 03:52:49.965212 140330632455936 logging_writer.py:48] [51000] global_step=51000, grad_norm=3.2130091190338135, loss=3.6980884075164795 -I0513 03:53:27.077286 140330640848640 logging_writer.py:48] [51100] global_step=51100, grad_norm=5.278060436248779, loss=3.7652196884155273 -I0513 03:54:03.936634 140330632455936 logging_writer.py:48] [51200] global_step=51200, grad_norm=6.74578332901001, loss=3.774393081665039 -I0513 03:54:41.211245 140330640848640 logging_writer.py:48] [51300] global_step=51300, grad_norm=3.914358615875244, loss=3.7409496307373047 -I0513 03:55:18.414293 140330632455936 logging_writer.py:48] [51400] global_step=51400, grad_norm=4.522525787353516, loss=3.6831398010253906 -I0513 03:55:55.269948 140330640848640 logging_writer.py:48] [51500] global_step=51500, grad_norm=5.541073322296143, loss=3.637594699859619 -I0513 03:56:32.509567 140330632455936 logging_writer.py:48] [51600] global_step=51600, grad_norm=4.519454002380371, loss=3.737182378768921 -I0513 03:57:09.572961 140330640848640 logging_writer.py:48] [51700] global_step=51700, grad_norm=5.549561977386475, loss=3.8758459091186523 -I0513 03:57:46.411134 140330632455936 logging_writer.py:48] [51800] global_step=51800, grad_norm=5.52198600769043, loss=3.7139415740966797 -I0513 03:58:23.748087 140330640848640 logging_writer.py:48] [51900] global_step=51900, grad_norm=11.282449722290039, loss=3.696845054626465 -I0513 03:59:00.951739 140330632455936 logging_writer.py:48] [52000] global_step=52000, grad_norm=4.92664909362793, loss=3.6357481479644775 -I0513 03:59:37.781002 140330640848640 logging_writer.py:48] [52100] global_step=52100, grad_norm=4.020849704742432, loss=3.6913790702819824 -I0513 04:00:15.054325 140330632455936 logging_writer.py:48] [52200] global_step=52200, grad_norm=4.351099967956543, loss=3.800809860229492 -I0513 04:00:51.872304 140330640848640 logging_writer.py:48] [52300] global_step=52300, grad_norm=3.7141847610473633, loss=3.800837993621826 -I0513 04:01:28.992910 140330632455936 logging_writer.py:48] [52400] global_step=52400, grad_norm=5.326633930206299, loss=3.6528477668762207 -I0513 04:02:06.396903 140330640848640 logging_writer.py:48] [52500] global_step=52500, grad_norm=4.7055768966674805, loss=3.8540375232696533 -I0513 04:02:43.586855 140330632455936 logging_writer.py:48] [52600] global_step=52600, grad_norm=5.361658573150635, loss=3.684241771697998 -I0513 04:03:20.387706 140330640848640 logging_writer.py:48] [52700] global_step=52700, grad_norm=4.863687515258789, loss=3.9323952198028564 -I0513 04:03:57.654536 140330632455936 logging_writer.py:48] [52800] global_step=52800, grad_norm=6.670470237731934, loss=3.746936798095703 -I0513 04:04:34.840779 140330640848640 logging_writer.py:48] [52900] global_step=52900, grad_norm=4.015689849853516, loss=3.6574666500091553 -I0513 04:05:11.657033 140330632455936 logging_writer.py:48] [53000] global_step=53000, grad_norm=9.742965698242188, loss=4.1271514892578125 -I0513 04:05:48.906156 140330640848640 logging_writer.py:48] [53100] global_step=53100, grad_norm=5.422811985015869, loss=3.6224780082702637 -I0513 04:06:26.019781 140330632455936 logging_writer.py:48] [53200] global_step=53200, grad_norm=5.2998528480529785, loss=3.6691436767578125 -I0513 04:07:02.914588 140330640848640 logging_writer.py:48] [53300] global_step=53300, grad_norm=3.746198892593384, loss=3.79569411277771 -I0513 04:07:40.173532 140330632455936 logging_writer.py:48] [53400] global_step=53400, grad_norm=10.378280639648438, loss=4.048086166381836 -I0513 04:08:17.381978 140330640848640 logging_writer.py:48] [53500] global_step=53500, grad_norm=4.191989421844482, loss=3.8442282676696777 -I0513 04:08:41.688872 140546196993216 spec.py:333] Evaluating on the training split. -I0513 04:08:53.219667 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 04:09:02.503143 140546196993216 spec.py:363] Evaluating on the test split. -I0513 04:09:03.374750 140546196993216 submission_runner.py:516] Time since start: 20307.43s, Step: 53567, {'train/accuracy': Array(0.00157446, dtype=float32), 'train/loss': Array(8.254096, dtype=float32), 'validation/accuracy': Array(0.00138, dtype=float32), 'validation/loss': Array(8.238354, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(8.248967, dtype=float32), 'test/num_examples': 10000, 'score': 20017.06850886345, 'total_duration': 20307.42605495453, 'accumulated_submission_time': 20017.06850886345, 'accumulated_eval_time': 289.2678849697113, 'accumulated_logging_time': 0.6422865390777588} -I0513 04:09:03.401330 140330632455936 logging_writer.py:48] [53567] accumulated_eval_time=289.268, accumulated_logging_time=0.642287, accumulated_submission_time=20017.1, global_step=53567, preemption_count=0, score=20017.1, test/accuracy=0.0010999999940395355, test/loss=8.248967170715332, test/num_examples=10000, total_duration=20307.4, train/accuracy=0.0015744578558951616, train/loss=8.254096031188965, validation/accuracy=0.0013799999142065644, validation/loss=8.238353729248047, validation/num_examples=50000 -I0513 04:09:17.220808 140330640848640 logging_writer.py:48] [53600] global_step=53600, grad_norm=5.102597236633301, loss=3.6819746494293213 -I0513 04:09:54.544542 140330632455936 logging_writer.py:48] [53700] global_step=53700, grad_norm=4.544042110443115, loss=3.7306621074676514 -I0513 04:10:31.608341 140330640848640 logging_writer.py:48] [53800] global_step=53800, grad_norm=4.305638313293457, loss=3.761448860168457 -I0513 04:11:08.562282 140330632455936 logging_writer.py:48] [53900] global_step=53900, grad_norm=4.134336471557617, loss=3.7539780139923096 -I0513 04:11:45.844879 140330640848640 logging_writer.py:48] [54000] global_step=54000, grad_norm=3.4224352836608887, loss=3.724240779876709 -I0513 04:12:23.044984 140330632455936 logging_writer.py:48] [54100] global_step=54100, grad_norm=4.330402374267578, loss=3.82969331741333 -I0513 04:12:59.893333 140330640848640 logging_writer.py:48] [54200] global_step=54200, grad_norm=4.171808242797852, loss=3.5228168964385986 -I0513 04:13:37.220546 140330632455936 logging_writer.py:48] [54300] global_step=54300, grad_norm=4.583864212036133, loss=3.90634822845459 -I0513 04:14:14.373433 140330640848640 logging_writer.py:48] [54400] global_step=54400, grad_norm=3.7564191818237305, loss=3.7491283416748047 -I0513 04:14:51.187837 140330632455936 logging_writer.py:48] [54500] global_step=54500, grad_norm=3.2643561363220215, loss=3.6584184169769287 -I0513 04:15:28.518941 140330640848640 logging_writer.py:48] [54600] global_step=54600, grad_norm=4.0951151847839355, loss=3.531507968902588 -I0513 04:16:05.722969 140330632455936 logging_writer.py:48] [54700] global_step=54700, grad_norm=3.1383891105651855, loss=3.7429168224334717 -I0513 04:16:42.547140 140330640848640 logging_writer.py:48] [54800] global_step=54800, grad_norm=3.5899171829223633, loss=3.6408021450042725 -I0513 04:17:19.900204 140330632455936 logging_writer.py:48] [54900] global_step=54900, grad_norm=5.700765609741211, loss=3.7326455116271973 -I0513 04:17:56.746216 140330640848640 logging_writer.py:48] [55000] global_step=55000, grad_norm=5.843623161315918, loss=3.7705183029174805 -I0513 04:18:33.821521 140330632455936 logging_writer.py:48] [55100] global_step=55100, grad_norm=5.732938289642334, loss=3.7339067459106445 -I0513 04:19:11.216326 140330640848640 logging_writer.py:48] [55200] global_step=55200, grad_norm=3.5471031665802, loss=3.709216356277466 -I0513 04:19:48.379262 140330632455936 logging_writer.py:48] [55300] global_step=55300, grad_norm=5.172370910644531, loss=3.7323317527770996 -I0513 04:20:25.210943 140330640848640 logging_writer.py:48] [55400] global_step=55400, grad_norm=4.1132612228393555, loss=3.7847139835357666 -I0513 04:21:02.474568 140330632455936 logging_writer.py:48] [55500] global_step=55500, grad_norm=3.9758899211883545, loss=3.570065498352051 -I0513 04:21:39.730902 140330640848640 logging_writer.py:48] [55600] global_step=55600, grad_norm=3.0925424098968506, loss=3.6374568939208984 -I0513 04:22:16.517256 140330632455936 logging_writer.py:48] [55700] global_step=55700, grad_norm=2.855377674102783, loss=3.6340649127960205 -I0513 04:22:53.786422 140330640848640 logging_writer.py:48] [55800] global_step=55800, grad_norm=3.619464159011841, loss=3.6808505058288574 -I0513 04:23:30.927027 140330632455936 logging_writer.py:48] [55900] global_step=55900, grad_norm=4.6797685623168945, loss=3.570249080657959 -I0513 04:24:07.812807 140330640848640 logging_writer.py:48] [56000] global_step=56000, grad_norm=8.178003311157227, loss=3.9456992149353027 -I0513 04:24:45.147276 140330632455936 logging_writer.py:48] [56100] global_step=56100, grad_norm=5.856185436248779, loss=3.767313241958618 -I0513 04:25:22.316986 140330640848640 logging_writer.py:48] [56200] global_step=56200, grad_norm=2.754542350769043, loss=3.6635448932647705 -I0513 04:25:59.106558 140330632455936 logging_writer.py:48] [56300] global_step=56300, grad_norm=6.551484107971191, loss=3.731510639190674 -I0513 04:26:36.360156 140330640848640 logging_writer.py:48] [56400] global_step=56400, grad_norm=3.697321891784668, loss=3.7253241539001465 -I0513 04:27:13.552223 140330632455936 logging_writer.py:48] [56500] global_step=56500, grad_norm=4.882788181304932, loss=3.8524649143218994 -I0513 04:27:50.391391 140330640848640 logging_writer.py:48] [56600] global_step=56600, grad_norm=2.703216075897217, loss=3.7038207054138184 -I0513 04:28:27.710891 140330632455936 logging_writer.py:48] [56700] global_step=56700, grad_norm=4.222611904144287, loss=3.8228204250335693 -I0513 04:29:04.925201 140330640848640 logging_writer.py:48] [56800] global_step=56800, grad_norm=5.585664749145508, loss=3.6046032905578613 -I0513 04:29:41.817612 140330632455936 logging_writer.py:48] [56900] global_step=56900, grad_norm=4.493036270141602, loss=3.7248306274414062 -I0513 04:30:19.235647 140330640848640 logging_writer.py:48] [57000] global_step=57000, grad_norm=5.184736251831055, loss=3.749220609664917 -I0513 04:30:55.983618 140330632455936 logging_writer.py:48] [57100] global_step=57100, grad_norm=6.630568504333496, loss=3.6776583194732666 -I0513 04:31:33.000259 140330640848640 logging_writer.py:48] [57200] global_step=57200, grad_norm=4.798628330230713, loss=3.8181347846984863 -I0513 04:32:10.353596 140330632455936 logging_writer.py:48] [57300] global_step=57300, grad_norm=4.777194499969482, loss=3.802016258239746 -I0513 04:32:47.436214 140330640848640 logging_writer.py:48] [57400] global_step=57400, grad_norm=5.665223121643066, loss=3.563699245452881 -I0513 04:33:24.421829 140330632455936 logging_writer.py:48] [57500] global_step=57500, grad_norm=3.253300428390503, loss=3.6566176414489746 -I0513 04:34:01.709592 140330640848640 logging_writer.py:48] [57600] global_step=57600, grad_norm=7.712425231933594, loss=3.7218313217163086 -I0513 04:34:38.935384 140330632455936 logging_writer.py:48] [57700] global_step=57700, grad_norm=3.7068686485290527, loss=3.5647127628326416 -I0513 04:35:15.809533 140330640848640 logging_writer.py:48] [57800] global_step=57800, grad_norm=4.308841228485107, loss=3.698747158050537 -I0513 04:35:53.122376 140330632455936 logging_writer.py:48] [57900] global_step=57900, grad_norm=4.616729259490967, loss=3.628946304321289 -I0513 04:36:30.291605 140330640848640 logging_writer.py:48] [58000] global_step=58000, grad_norm=3.591322898864746, loss=3.809053897857666 -I0513 04:37:07.182934 140330632455936 logging_writer.py:48] [58100] global_step=58100, grad_norm=3.6830928325653076, loss=3.7053158283233643 -I0513 04:37:44.465506 140330640848640 logging_writer.py:48] [58200] global_step=58200, grad_norm=2.9360179901123047, loss=3.67476749420166 -I0513 04:38:21.622424 140330632455936 logging_writer.py:48] [58300] global_step=58300, grad_norm=2.849902868270874, loss=3.657604217529297 -I0513 04:38:58.470394 140330640848640 logging_writer.py:48] [58400] global_step=58400, grad_norm=5.002469062805176, loss=3.6125569343566895 -I0513 04:39:35.745040 140330632455936 logging_writer.py:48] [58500] global_step=58500, grad_norm=7.891635894775391, loss=3.922863245010376 -I0513 04:40:12.837265 140330640848640 logging_writer.py:48] [58600] global_step=58600, grad_norm=5.153986930847168, loss=3.777348518371582 -I0513 04:40:49.679895 140330632455936 logging_writer.py:48] [58700] global_step=58700, grad_norm=4.58257532119751, loss=3.721977472305298 -I0513 04:41:26.980742 140330640848640 logging_writer.py:48] [58800] global_step=58800, grad_norm=3.539006471633911, loss=3.4685003757476807 -I0513 04:42:04.151350 140330632455936 logging_writer.py:48] [58900] global_step=58900, grad_norm=4.233903884887695, loss=3.6904335021972656 -I0513 04:42:19.554321 140546196993216 spec.py:333] Evaluating on the training split. -I0513 04:42:30.726631 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 04:42:40.257771 140546196993216 spec.py:363] Evaluating on the test split. -I0513 04:42:41.132006 140546196993216 submission_runner.py:516] Time since start: 22325.18s, Step: 58943, {'train/accuracy': Array(0.001136, dtype=float32), 'train/loss': Array(8.292785, dtype=float32), 'validation/accuracy': Array(0.00138, dtype=float32), 'validation/loss': Array(8.236893, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0014, dtype=float32), 'test/loss': Array(8.245119, dtype=float32), 'test/num_examples': 10000, 'score': 22013.158291101456, 'total_duration': 22325.183495998383, 'accumulated_submission_time': 22013.158291101456, 'accumulated_eval_time': 310.8434724807739, 'accumulated_logging_time': 0.6898555755615234} -I0513 04:42:41.164509 140330640848640 logging_writer.py:48] [58943] accumulated_eval_time=310.843, accumulated_logging_time=0.689856, accumulated_submission_time=22013.2, global_step=58943, preemption_count=0, score=22013.2, test/accuracy=0.0014000000664964318, test/loss=8.245119094848633, test/num_examples=10000, total_duration=22325.2, train/accuracy=0.0011360011994838715, train/loss=8.292784690856934, validation/accuracy=0.0013799999142065644, validation/loss=8.236892700195312, validation/num_examples=50000 -I0513 04:43:03.677715 140330632455936 logging_writer.py:48] [59000] global_step=59000, grad_norm=3.9962210655212402, loss=3.6594250202178955 -I0513 04:43:40.916000 140330640848640 logging_writer.py:48] [59100] global_step=59100, grad_norm=6.34234094619751, loss=3.790186882019043 -I0513 04:44:17.767460 140330632455936 logging_writer.py:48] [59200] global_step=59200, grad_norm=2.921161413192749, loss=3.706528902053833 -I0513 04:44:54.840973 140330640848640 logging_writer.py:48] [59300] global_step=59300, grad_norm=3.9754884243011475, loss=3.794590950012207 -I0513 04:45:32.205066 140330632455936 logging_writer.py:48] [59400] global_step=59400, grad_norm=8.345151901245117, loss=3.889280080795288 -I0513 04:46:09.314100 140330640848640 logging_writer.py:48] [59500] global_step=59500, grad_norm=3.616891622543335, loss=3.603909492492676 -I0513 04:46:46.186025 140330632455936 logging_writer.py:48] [59600] global_step=59600, grad_norm=3.834829330444336, loss=3.648516893386841 -I0513 04:47:23.546964 140330640848640 logging_writer.py:48] [59700] global_step=59700, grad_norm=4.2452192306518555, loss=3.7583670616149902 -I0513 04:48:00.399090 140330632455936 logging_writer.py:48] [59800] global_step=59800, grad_norm=4.28662109375, loss=3.661334991455078 -I0513 04:48:37.554302 140330640848640 logging_writer.py:48] [59900] global_step=59900, grad_norm=5.903714179992676, loss=3.6392507553100586 -I0513 04:49:14.896625 140330632455936 logging_writer.py:48] [60000] global_step=60000, grad_norm=3.795125961303711, loss=3.7105050086975098 -I0513 04:49:52.062797 140330640848640 logging_writer.py:48] [60100] global_step=60100, grad_norm=3.206115245819092, loss=3.6317548751831055 -I0513 04:50:28.977545 140330632455936 logging_writer.py:48] [60200] global_step=60200, grad_norm=3.8096795082092285, loss=3.5454235076904297 -I0513 04:51:06.328133 140330640848640 logging_writer.py:48] [60300] global_step=60300, grad_norm=4.213667392730713, loss=3.5361809730529785 -I0513 04:51:43.520880 140330632455936 logging_writer.py:48] [60400] global_step=60400, grad_norm=5.514969825744629, loss=3.6808218955993652 -I0513 04:52:20.365998 140330640848640 logging_writer.py:48] [60500] global_step=60500, grad_norm=3.8992791175842285, loss=3.685945510864258 -I0513 04:52:57.720964 140330632455936 logging_writer.py:48] [60600] global_step=60600, grad_norm=5.116889953613281, loss=3.714916706085205 -I0513 04:53:34.483519 140330640848640 logging_writer.py:48] [60700] global_step=60700, grad_norm=5.63176965713501, loss=3.757978677749634 -I0513 04:54:11.537407 140330632455936 logging_writer.py:48] [60800] global_step=60800, grad_norm=4.037842750549316, loss=3.8505618572235107 -I0513 04:54:48.783064 140330640848640 logging_writer.py:48] [60900] global_step=60900, grad_norm=5.8288726806640625, loss=3.7450296878814697 -I0513 04:55:26.079939 140330632455936 logging_writer.py:48] [61000] global_step=61000, grad_norm=5.518537521362305, loss=3.857708215713501 -I0513 04:56:02.917908 140330640848640 logging_writer.py:48] [61100] global_step=61100, grad_norm=3.3579602241516113, loss=3.672649621963501 -I0513 04:56:40.219864 140330632455936 logging_writer.py:48] [61200] global_step=61200, grad_norm=4.966516971588135, loss=3.656342029571533 -I0513 04:57:17.121182 140330640848640 logging_writer.py:48] [61300] global_step=61300, grad_norm=6.544070243835449, loss=3.824009656906128 -I0513 04:57:54.258549 140330632455936 logging_writer.py:48] [61400] global_step=61400, grad_norm=3.394970178604126, loss=3.6574292182922363 -I0513 04:58:31.508949 140330640848640 logging_writer.py:48] [61500] global_step=61500, grad_norm=2.8325977325439453, loss=3.8428854942321777 -I0513 04:59:08.727870 140330632455936 logging_writer.py:48] [61600] global_step=61600, grad_norm=7.366558074951172, loss=3.6677353382110596 -I0513 04:59:45.566901 140330640848640 logging_writer.py:48] [61700] global_step=61700, grad_norm=5.059511661529541, loss=3.7454073429107666 -I0513 05:00:22.895248 140330632455936 logging_writer.py:48] [61800] global_step=61800, grad_norm=5.064797401428223, loss=3.74307918548584 -I0513 05:00:59.700336 140330640848640 logging_writer.py:48] [61900] global_step=61900, grad_norm=4.390219211578369, loss=3.7563982009887695 -I0513 05:01:36.859862 140330632455936 logging_writer.py:48] [62000] global_step=62000, grad_norm=6.871188640594482, loss=3.8388831615448 -I0513 05:02:14.129849 140330640848640 logging_writer.py:48] [62100] global_step=62100, grad_norm=3.092611312866211, loss=3.625126600265503 -I0513 05:02:50.935907 140330632455936 logging_writer.py:48] [62200] global_step=62200, grad_norm=4.685038089752197, loss=3.693298101425171 -I0513 05:03:27.963410 140330640848640 logging_writer.py:48] [62300] global_step=62300, grad_norm=3.8061482906341553, loss=3.8138036727905273 -I0513 05:04:05.218840 140330632455936 logging_writer.py:48] [62400] global_step=62400, grad_norm=3.711824417114258, loss=3.782397747039795 -I0513 05:04:42.135198 140330640848640 logging_writer.py:48] [62500] global_step=62500, grad_norm=4.312089443206787, loss=3.7387189865112305 -I0513 05:05:19.313779 140330632455936 logging_writer.py:48] [62600] global_step=62600, grad_norm=3.674647808074951, loss=3.5901455879211426 -I0513 05:05:56.649011 140330640848640 logging_writer.py:48] [62700] global_step=62700, grad_norm=4.344249725341797, loss=3.6378209590911865 -I0513 05:06:33.506982 140330632455936 logging_writer.py:48] [62800] global_step=62800, grad_norm=5.202394962310791, loss=3.7120611667633057 -I0513 05:07:10.767966 140330640848640 logging_writer.py:48] [62900] global_step=62900, grad_norm=3.2613353729248047, loss=3.600369930267334 -I0513 05:07:48.022511 140330632455936 logging_writer.py:48] [63000] global_step=63000, grad_norm=4.592917442321777, loss=3.7763655185699463 -I0513 05:08:24.889235 140330640848640 logging_writer.py:48] [63100] global_step=63100, grad_norm=6.5342278480529785, loss=3.857440710067749 -I0513 05:09:02.142275 140330632455936 logging_writer.py:48] [63200] global_step=63200, grad_norm=6.051773548126221, loss=3.724705219268799 -I0513 05:09:39.422002 140330640848640 logging_writer.py:48] [63300] global_step=63300, grad_norm=3.800131320953369, loss=3.71567964553833 -I0513 05:10:16.256952 140330632455936 logging_writer.py:48] [63400] global_step=63400, grad_norm=4.399792671203613, loss=3.7067313194274902 -I0513 05:10:53.171141 140330640848640 logging_writer.py:48] [63500] global_step=63500, grad_norm=4.888543128967285, loss=3.560816764831543 -I0513 05:11:30.800881 140330632455936 logging_writer.py:48] [63600] global_step=63600, grad_norm=4.580099582672119, loss=3.741103172302246 -I0513 05:12:07.601401 140330640848640 logging_writer.py:48] [63700] global_step=63700, grad_norm=3.951918363571167, loss=3.831528425216675 -I0513 05:12:44.731739 140330632455936 logging_writer.py:48] [63800] global_step=63800, grad_norm=3.582120180130005, loss=3.6953396797180176 -I0513 05:13:22.026169 140330640848640 logging_writer.py:48] [63900] global_step=63900, grad_norm=6.237988471984863, loss=3.56007719039917 -I0513 05:13:58.861344 140330632455936 logging_writer.py:48] [64000] global_step=64000, grad_norm=5.788256645202637, loss=3.7112364768981934 -I0513 05:14:36.153913 140330640848640 logging_writer.py:48] [64100] global_step=64100, grad_norm=3.3939170837402344, loss=3.634544849395752 -I0513 05:15:13.437421 140330632455936 logging_writer.py:48] [64200] global_step=64200, grad_norm=6.212123870849609, loss=3.938603401184082 -I0513 05:15:50.290005 140330640848640 logging_writer.py:48] [64300] global_step=64300, grad_norm=4.902135848999023, loss=3.960193157196045 -I0513 05:15:57.775956 140546196993216 spec.py:333] Evaluating on the training split. -I0513 05:16:08.643454 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 05:16:18.721480 140546196993216 spec.py:363] Evaluating on the test split. -I0513 05:16:19.613256 140546196993216 submission_runner.py:516] Time since start: 24343.66s, Step: 64321, {'train/accuracy': Array(0.00137516, dtype=float32), 'train/loss': Array(8.30454, dtype=float32), 'validation/accuracy': Array(0.00134, dtype=float32), 'validation/loss': Array(8.2691765, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(8.277351, dtype=float32), 'test/num_examples': 10000, 'score': 24009.709911584854, 'total_duration': 24343.664563417435, 'accumulated_submission_time': 24009.709911584854, 'accumulated_eval_time': 332.6784863471985, 'accumulated_logging_time': 0.7396748065948486} -I0513 05:16:19.671374 140330632455936 logging_writer.py:48] [64321] accumulated_eval_time=332.678, accumulated_logging_time=0.739675, accumulated_submission_time=24009.7, global_step=64321, preemption_count=0, score=24009.7, test/accuracy=0.0012000000569969416, test/loss=8.277351379394531, test/num_examples=10000, total_duration=24343.7, train/accuracy=0.0013751593651250005, train/loss=8.304539680480957, validation/accuracy=0.001339999958872795, validation/loss=8.269176483154297, validation/num_examples=50000 -I0513 05:16:50.348241 140330640848640 logging_writer.py:48] [64400] global_step=64400, grad_norm=3.328418493270874, loss=3.5913028717041016 -I0513 05:17:27.621559 140330632455936 logging_writer.py:48] [64500] global_step=64500, grad_norm=3.3737330436706543, loss=3.7671031951904297 -I0513 05:18:04.415192 140330640848640 logging_writer.py:48] [64600] global_step=64600, grad_norm=4.026939392089844, loss=3.9272353649139404 -I0513 05:18:41.609545 140330632455936 logging_writer.py:48] [64700] global_step=64700, grad_norm=3.6824326515197754, loss=3.588958263397217 -I0513 05:19:18.864534 140330640848640 logging_writer.py:48] [64800] global_step=64800, grad_norm=5.389061450958252, loss=3.9841387271881104 -I0513 05:19:55.742230 140330632455936 logging_writer.py:48] [64900] global_step=64900, grad_norm=5.435727119445801, loss=3.790348529815674 -I0513 05:20:32.867039 140330640848640 logging_writer.py:48] [65000] global_step=65000, grad_norm=4.872454643249512, loss=3.6676723957061768 -I0513 05:21:10.216480 140330632455936 logging_writer.py:48] [65100] global_step=65100, grad_norm=6.05443811416626, loss=3.6745290756225586 -I0513 05:21:47.096602 140330640848640 logging_writer.py:48] [65200] global_step=65200, grad_norm=4.00006628036499, loss=3.7813355922698975 -I0513 05:22:24.295623 140330632455936 logging_writer.py:48] [65300] global_step=65300, grad_norm=5.103696346282959, loss=3.661020278930664 -I0513 05:23:01.573174 140330640848640 logging_writer.py:48] [65400] global_step=65400, grad_norm=5.413623332977295, loss=3.700960397720337 -I0513 05:23:38.449756 140330632455936 logging_writer.py:48] [65500] global_step=65500, grad_norm=6.212788105010986, loss=3.6808199882507324 -I0513 05:24:15.535648 140330640848640 logging_writer.py:48] [65600] global_step=65600, grad_norm=4.302270889282227, loss=3.6404733657836914 -I0513 05:24:52.838535 140330632455936 logging_writer.py:48] [65700] global_step=65700, grad_norm=5.258028507232666, loss=3.6237094402313232 -I0513 05:25:29.691320 140330640848640 logging_writer.py:48] [65800] global_step=65800, grad_norm=6.012365341186523, loss=3.9236745834350586 -I0513 05:26:06.938814 140330632455936 logging_writer.py:48] [65900] global_step=65900, grad_norm=4.4949951171875, loss=3.894881010055542 -I0513 05:26:44.343569 140330640848640 logging_writer.py:48] [66000] global_step=66000, grad_norm=5.0843400955200195, loss=3.799992561340332 -I0513 05:27:21.254062 140330632455936 logging_writer.py:48] [66100] global_step=66100, grad_norm=4.503166675567627, loss=3.7028727531433105 -I0513 05:27:58.476150 140330640848640 logging_writer.py:48] [66200] global_step=66200, grad_norm=3.7355682849884033, loss=3.806025981903076 -I0513 05:28:35.756435 140330632455936 logging_writer.py:48] [66300] global_step=66300, grad_norm=5.225771427154541, loss=3.7539877891540527 -I0513 05:29:12.719005 140330640848640 logging_writer.py:48] [66400] global_step=66400, grad_norm=4.777586936950684, loss=3.785007953643799 -I0513 05:29:49.941034 140330632455936 logging_writer.py:48] [66500] global_step=66500, grad_norm=7.373021602630615, loss=3.9139275550842285 -I0513 05:30:27.310071 140330640848640 logging_writer.py:48] [66600] global_step=66600, grad_norm=4.196310997009277, loss=3.640261650085449 -I0513 05:31:04.086789 140330632455936 logging_writer.py:48] [66700] global_step=66700, grad_norm=4.037747383117676, loss=3.764148235321045 -I0513 05:31:41.276967 140330640848640 logging_writer.py:48] [66800] global_step=66800, grad_norm=3.2618649005889893, loss=3.7893576622009277 -I0513 05:32:18.646261 140330632455936 logging_writer.py:48] [66900] global_step=66900, grad_norm=4.287854194641113, loss=3.729135274887085 -I0513 05:32:55.432167 140330640848640 logging_writer.py:48] [67000] global_step=67000, grad_norm=5.2318549156188965, loss=3.9797253608703613 -I0513 05:33:32.543678 140330632455936 logging_writer.py:48] [67100] global_step=67100, grad_norm=5.75941276550293, loss=3.6006603240966797 -I0513 05:34:09.839936 140330640848640 logging_writer.py:48] [67200] global_step=67200, grad_norm=3.8676271438598633, loss=3.7739062309265137 -I0513 05:34:46.618402 140330632455936 logging_writer.py:48] [67300] global_step=67300, grad_norm=4.317509651184082, loss=3.5438597202301025 -I0513 05:35:23.738952 140330640848640 logging_writer.py:48] [67400] global_step=67400, grad_norm=5.739094257354736, loss=3.747023105621338 -I0513 05:36:01.000904 140330632455936 logging_writer.py:48] [67500] global_step=67500, grad_norm=5.143769264221191, loss=3.5731215476989746 -I0513 05:36:37.833046 140330640848640 logging_writer.py:48] [67600] global_step=67600, grad_norm=4.146783351898193, loss=3.633589267730713 -I0513 05:37:15.000513 140330632455936 logging_writer.py:48] [67700] global_step=67700, grad_norm=4.8056440353393555, loss=3.744607448577881 -I0513 05:37:52.334500 140330640848640 logging_writer.py:48] [67800] global_step=67800, grad_norm=7.355105400085449, loss=3.819505214691162 -I0513 05:38:29.135733 140330632455936 logging_writer.py:48] [67900] global_step=67900, grad_norm=3.635033130645752, loss=3.7038607597351074 -I0513 05:39:06.444096 140330640848640 logging_writer.py:48] [68000] global_step=68000, grad_norm=3.7977473735809326, loss=3.721081256866455 -I0513 05:39:43.744916 140330632455936 logging_writer.py:48] [68100] global_step=68100, grad_norm=4.364473819732666, loss=3.6158082485198975 -I0513 05:40:20.540118 140330640848640 logging_writer.py:48] [68200] global_step=68200, grad_norm=3.610705614089966, loss=3.777418613433838 -I0513 05:40:57.664581 140330632455936 logging_writer.py:48] [68300] global_step=68300, grad_norm=5.445450782775879, loss=3.783783435821533 -I0513 05:41:34.899272 140330640848640 logging_writer.py:48] [68400] global_step=68400, grad_norm=6.025855541229248, loss=3.797513961791992 -I0513 05:42:11.816357 140330632455936 logging_writer.py:48] [68500] global_step=68500, grad_norm=4.596244812011719, loss=3.6736764907836914 -I0513 05:42:48.915960 140330640848640 logging_writer.py:48] [68600] global_step=68600, grad_norm=2.983428955078125, loss=3.6206512451171875 -I0513 05:43:26.195641 140330632455936 logging_writer.py:48] [68700] global_step=68700, grad_norm=4.111405372619629, loss=3.585360050201416 -I0513 05:44:02.941992 140330640848640 logging_writer.py:48] [68800] global_step=68800, grad_norm=4.649420261383057, loss=3.654691219329834 -I0513 05:44:40.037757 140330632455936 logging_writer.py:48] [68900] global_step=68900, grad_norm=4.231954574584961, loss=3.765204429626465 -I0513 05:45:17.243742 140330640848640 logging_writer.py:48] [69000] global_step=69000, grad_norm=3.7322397232055664, loss=3.6137373447418213 -I0513 05:45:54.105058 140330632455936 logging_writer.py:48] [69100] global_step=69100, grad_norm=4.546603202819824, loss=3.8685598373413086 -I0513 05:46:31.276028 140330640848640 logging_writer.py:48] [69200] global_step=69200, grad_norm=3.325200080871582, loss=3.6184258460998535 -I0513 05:47:08.665156 140330632455936 logging_writer.py:48] [69300] global_step=69300, grad_norm=5.571754455566406, loss=3.70341157913208 -I0513 05:47:45.565453 140330640848640 logging_writer.py:48] [69400] global_step=69400, grad_norm=4.3447265625, loss=3.7390713691711426 -I0513 05:48:22.696301 140330632455936 logging_writer.py:48] [69500] global_step=69500, grad_norm=3.962700128555298, loss=3.697201728820801 -I0513 05:48:59.965692 140330640848640 logging_writer.py:48] [69600] global_step=69600, grad_norm=3.548571825027466, loss=3.907412528991699 -I0513 05:49:35.620773 140546196993216 spec.py:333] Evaluating on the training split. -I0513 05:49:46.700690 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 05:49:56.338049 140546196993216 spec.py:363] Evaluating on the test split. -I0513 05:49:57.224725 140546196993216 submission_runner.py:516] Time since start: 26361.28s, Step: 69699, {'train/accuracy': Array(0.00165418, dtype=float32), 'train/loss': Array(8.343106, dtype=float32), 'validation/accuracy': Array(0.00138, dtype=float32), 'validation/loss': Array(8.300176, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(8.307022, dtype=float32), 'test/num_examples': 10000, 'score': 26005.58596253395, 'total_duration': 26361.276525497437, 'accumulated_submission_time': 26005.58596253395, 'accumulated_eval_time': 354.2806496620178, 'accumulated_logging_time': 0.8288896083831787} -I0513 05:49:57.265342 140330632455936 logging_writer.py:48] [69699] accumulated_eval_time=354.281, accumulated_logging_time=0.82889, accumulated_submission_time=26005.6, global_step=69699, preemption_count=0, score=26005.6, test/accuracy=0.0013000001199543476, test/loss=8.307022094726562, test/num_examples=10000, total_duration=26361.3, train/accuracy=0.0016541772056370974, train/loss=8.343106269836426, validation/accuracy=0.0013799999142065644, validation/loss=8.300175666809082, validation/num_examples=50000 -I0513 05:49:58.818293 140330640848640 logging_writer.py:48] [69700] global_step=69700, grad_norm=3.076359748840332, loss=3.634094715118408 -I0513 05:50:36.500882 140330632455936 logging_writer.py:48] [69800] global_step=69800, grad_norm=3.2726645469665527, loss=3.7183995246887207 -I0513 05:51:13.729590 140330640848640 logging_writer.py:48] [69900] global_step=69900, grad_norm=3.726896286010742, loss=3.7068419456481934 -I0513 05:51:50.552533 140330632455936 logging_writer.py:48] [70000] global_step=70000, grad_norm=4.3550920486450195, loss=3.622380256652832 -I0513 05:52:27.727488 140330640848640 logging_writer.py:48] [70100] global_step=70100, grad_norm=4.119032382965088, loss=3.723910093307495 -I0513 05:53:04.991724 140330632455936 logging_writer.py:48] [70200] global_step=70200, grad_norm=4.109454154968262, loss=3.7333221435546875 -I0513 05:53:41.721277 140330640848640 logging_writer.py:48] [70300] global_step=70300, grad_norm=3.976731300354004, loss=3.605494260787964 -I0513 05:54:18.829263 140330632455936 logging_writer.py:48] [70400] global_step=70400, grad_norm=3.7855007648468018, loss=3.624716281890869 -I0513 05:54:56.118472 140330640848640 logging_writer.py:48] [70500] global_step=70500, grad_norm=4.40328311920166, loss=3.5916829109191895 -I0513 05:55:32.845602 140330632455936 logging_writer.py:48] [70600] global_step=70600, grad_norm=4.0885725021362305, loss=3.612351179122925 -I0513 05:56:10.013060 140330640848640 logging_writer.py:48] [70700] global_step=70700, grad_norm=3.9812023639678955, loss=3.6567602157592773 -I0513 05:56:47.274737 140330632455936 logging_writer.py:48] [70800] global_step=70800, grad_norm=4.877323150634766, loss=3.7366747856140137 -I0513 05:57:24.118319 140330640848640 logging_writer.py:48] [70900] global_step=70900, grad_norm=5.0617780685424805, loss=3.642244815826416 -I0513 05:58:00.971282 140330632455936 logging_writer.py:48] [71000] global_step=71000, grad_norm=5.0544257164001465, loss=3.7712371349334717 -I0513 05:58:38.588810 140330640848640 logging_writer.py:48] [71100] global_step=71100, grad_norm=4.626042366027832, loss=3.6438050270080566 -I0513 05:59:15.333581 140330632455936 logging_writer.py:48] [71200] global_step=71200, grad_norm=3.695007085800171, loss=3.66652512550354 -I0513 05:59:52.504387 140330640848640 logging_writer.py:48] [71300] global_step=71300, grad_norm=4.119859218597412, loss=3.7348079681396484 -I0513 06:00:29.814728 140330632455936 logging_writer.py:48] [71400] global_step=71400, grad_norm=5.016263961791992, loss=3.833533763885498 -I0513 06:01:06.703899 140330640848640 logging_writer.py:48] [71500] global_step=71500, grad_norm=3.450361967086792, loss=3.7055654525756836 -I0513 06:01:43.876431 140330632455936 logging_writer.py:48] [71600] global_step=71600, grad_norm=4.172699451446533, loss=3.6913747787475586 -I0513 06:02:21.157962 140330640848640 logging_writer.py:48] [71700] global_step=71700, grad_norm=3.688565731048584, loss=3.792475461959839 -I0513 06:02:58.044377 140330632455936 logging_writer.py:48] [71800] global_step=71800, grad_norm=5.224074840545654, loss=3.685199737548828 -I0513 06:03:35.116878 140330640848640 logging_writer.py:48] [71900] global_step=71900, grad_norm=4.763123035430908, loss=3.730048179626465 -I0513 06:04:12.398281 140330632455936 logging_writer.py:48] [72000] global_step=72000, grad_norm=4.546492576599121, loss=3.6381759643554688 -I0513 06:04:49.203570 140330640848640 logging_writer.py:48] [72100] global_step=72100, grad_norm=4.336789131164551, loss=3.6343278884887695 -I0513 06:05:26.278238 140330632455936 logging_writer.py:48] [72200] global_step=72200, grad_norm=5.337066173553467, loss=3.7263004779815674 -I0513 06:06:03.548561 140330640848640 logging_writer.py:48] [72300] global_step=72300, grad_norm=4.312495231628418, loss=3.770547866821289 -I0513 06:06:40.357288 140330632455936 logging_writer.py:48] [72400] global_step=72400, grad_norm=5.234017372131348, loss=3.701587438583374 -I0513 06:07:17.387105 140330640848640 logging_writer.py:48] [72500] global_step=72500, grad_norm=16.160306930541992, loss=3.644972324371338 -I0513 06:07:54.693932 140330632455936 logging_writer.py:48] [72600] global_step=72600, grad_norm=4.123131275177002, loss=3.782938003540039 -I0513 06:08:31.486361 140330640848640 logging_writer.py:48] [72700] global_step=72700, grad_norm=3.7927169799804688, loss=3.542992353439331 -I0513 06:09:08.693989 140330632455936 logging_writer.py:48] [72800] global_step=72800, grad_norm=3.610224485397339, loss=3.7499983310699463 -I0513 06:09:45.987571 140330640848640 logging_writer.py:48] [72900] global_step=72900, grad_norm=5.934783458709717, loss=3.85444974899292 -I0513 06:10:22.836856 140330632455936 logging_writer.py:48] [73000] global_step=73000, grad_norm=3.9955101013183594, loss=3.685392379760742 -I0513 06:10:59.621087 140330640848640 logging_writer.py:48] [73100] global_step=73100, grad_norm=3.9729502201080322, loss=3.7176408767700195 -I0513 06:11:37.190286 140330632455936 logging_writer.py:48] [73200] global_step=73200, grad_norm=5.499223232269287, loss=3.7231640815734863 -I0513 06:12:13.981347 140330640848640 logging_writer.py:48] [73300] global_step=73300, grad_norm=3.7130520343780518, loss=3.6006546020507812 -I0513 06:12:51.208038 140330632455936 logging_writer.py:48] [73400] global_step=73400, grad_norm=3.9206700325012207, loss=3.571289300918579 -I0513 06:13:28.575235 140330640848640 logging_writer.py:48] [73500] global_step=73500, grad_norm=6.6406073570251465, loss=3.690612554550171 -I0513 06:14:05.322846 140330632455936 logging_writer.py:48] [73600] global_step=73600, grad_norm=3.4652507305145264, loss=3.667346477508545 -I0513 06:14:42.440015 140330640848640 logging_writer.py:48] [73700] global_step=73700, grad_norm=5.465666770935059, loss=3.787923812866211 -I0513 06:15:19.713603 140330632455936 logging_writer.py:48] [73800] global_step=73800, grad_norm=5.20517635345459, loss=3.8075289726257324 -I0513 06:15:56.614305 140330640848640 logging_writer.py:48] [73900] global_step=73900, grad_norm=4.6203413009643555, loss=3.700029134750366 -I0513 06:16:33.847981 140330632455936 logging_writer.py:48] [74000] global_step=74000, grad_norm=16.703052520751953, loss=3.889380931854248 -I0513 06:17:11.201663 140330640848640 logging_writer.py:48] [74100] global_step=74100, grad_norm=3.9956490993499756, loss=3.722299337387085 -I0513 06:17:48.001679 140330632455936 logging_writer.py:48] [74200] global_step=74200, grad_norm=5.20900821685791, loss=3.7081942558288574 -I0513 06:18:25.218096 140330640848640 logging_writer.py:48] [74300] global_step=74300, grad_norm=3.275273323059082, loss=3.5741190910339355 -I0513 06:19:02.531837 140330632455936 logging_writer.py:48] [74400] global_step=74400, grad_norm=4.05195951461792, loss=3.8073205947875977 -I0513 06:19:39.322543 140330640848640 logging_writer.py:48] [74500] global_step=74500, grad_norm=4.077786922454834, loss=3.6988322734832764 -I0513 06:20:16.349592 140330632455936 logging_writer.py:48] [74600] global_step=74600, grad_norm=6.819699764251709, loss=4.015129089355469 -I0513 06:20:53.650506 140330640848640 logging_writer.py:48] [74700] global_step=74700, grad_norm=5.710982322692871, loss=3.827796697616577 -I0513 06:21:30.412498 140330632455936 logging_writer.py:48] [74800] global_step=74800, grad_norm=10.868199348449707, loss=3.9117016792297363 -I0513 06:22:07.590058 140330640848640 logging_writer.py:48] [74900] global_step=74900, grad_norm=4.0015106201171875, loss=3.7430801391601562 -I0513 06:22:44.875174 140330632455936 logging_writer.py:48] [75000] global_step=75000, grad_norm=4.1827850341796875, loss=3.820199966430664 -I0513 06:23:13.573406 140546196993216 spec.py:333] Evaluating on the training split. -I0513 06:23:24.144600 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 06:23:33.688733 140546196993216 spec.py:363] Evaluating on the test split. -I0513 06:23:34.549701 140546196993216 submission_runner.py:516] Time since start: 28378.60s, Step: 75079, {'train/accuracy': Array(0.00127551, dtype=float32), 'train/loss': Array(8.378688, dtype=float32), 'validation/accuracy': Array(0.0015, dtype=float32), 'validation/loss': Array(8.342849, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(8.351302, dtype=float32), 'test/num_examples': 10000, 'score': 28001.827844142914, 'total_duration': 28378.600981235504, 'accumulated_submission_time': 28001.827844142914, 'accumulated_eval_time': 375.2546329498291, 'accumulated_logging_time': 0.8931126594543457} -I0513 06:23:34.580048 140330640848640 logging_writer.py:48] [75079] accumulated_eval_time=375.255, accumulated_logging_time=0.893113, accumulated_submission_time=28001.8, global_step=75079, preemption_count=0, score=28001.8, test/accuracy=0.001500000013038516, test/loss=8.351302146911621, test/num_examples=10000, total_duration=28378.6, train/accuracy=0.0012755101779475808, train/loss=8.378687858581543, validation/accuracy=0.001500000013038516, validation/loss=8.342848777770996, validation/num_examples=50000 -I0513 06:23:43.959537 140330632455936 logging_writer.py:48] [75100] global_step=75100, grad_norm=4.624239444732666, loss=3.877131462097168 -I0513 06:24:20.745367 140330640848640 logging_writer.py:48] [75200] global_step=75200, grad_norm=4.678016662597656, loss=3.7693421840667725 -I0513 06:24:58.383081 140330632455936 logging_writer.py:48] [75300] global_step=75300, grad_norm=4.211870193481445, loss=3.63447904586792 -I0513 06:25:35.173362 140330640848640 logging_writer.py:48] [75400] global_step=75400, grad_norm=8.498342514038086, loss=3.8081367015838623 -I0513 06:26:12.355038 140330632455936 logging_writer.py:48] [75500] global_step=75500, grad_norm=5.1561150550842285, loss=3.833014488220215 -I0513 06:26:49.688500 140330640848640 logging_writer.py:48] [75600] global_step=75600, grad_norm=4.77055025100708, loss=3.8850786685943604 -I0513 06:27:26.492383 140330632455936 logging_writer.py:48] [75700] global_step=75700, grad_norm=6.722166061401367, loss=3.856078624725342 -I0513 06:28:03.343230 140330640848640 logging_writer.py:48] [75800] global_step=75800, grad_norm=13.116930961608887, loss=4.0729827880859375 -I0513 06:28:40.993906 140330632455936 logging_writer.py:48] [75900] global_step=75900, grad_norm=4.106633186340332, loss=3.726559638977051 -I0513 06:29:17.783189 140330640848640 logging_writer.py:48] [76000] global_step=76000, grad_norm=10.040434837341309, loss=3.7526583671569824 -I0513 06:29:54.869766 140330632455936 logging_writer.py:48] [76100] global_step=76100, grad_norm=4.963212013244629, loss=3.661153793334961 -I0513 06:30:32.188730 140330640848640 logging_writer.py:48] [76200] global_step=76200, grad_norm=4.213085174560547, loss=3.820439577102661 -I0513 06:31:09.111206 140330632455936 logging_writer.py:48] [76300] global_step=76300, grad_norm=4.40770959854126, loss=3.7450599670410156 -I0513 06:31:46.332546 140330640848640 logging_writer.py:48] [76400] global_step=76400, grad_norm=4.204817771911621, loss=3.765798568725586 -I0513 06:32:23.670019 140330632455936 logging_writer.py:48] [76500] global_step=76500, grad_norm=4.79732608795166, loss=3.6525580883026123 -I0513 06:33:00.382611 140330640848640 logging_writer.py:48] [76600] global_step=76600, grad_norm=9.651665687561035, loss=3.915093183517456 -I0513 06:33:37.188878 140330632455936 logging_writer.py:48] [76700] global_step=76700, grad_norm=4.236350059509277, loss=3.6699442863464355 -I0513 06:34:14.797305 140330640848640 logging_writer.py:48] [76800] global_step=76800, grad_norm=8.72778606414795, loss=3.886427164077759 -I0513 06:34:51.578584 140330632455936 logging_writer.py:48] [76900] global_step=76900, grad_norm=4.427398681640625, loss=3.682887554168701 -I0513 06:35:28.759915 140330640848640 logging_writer.py:48] [77000] global_step=77000, grad_norm=3.7011256217956543, loss=3.792154312133789 -I0513 06:36:06.050823 140330632455936 logging_writer.py:48] [77100] global_step=77100, grad_norm=3.880734443664551, loss=3.7487175464630127 -I0513 06:36:42.823055 140330640848640 logging_writer.py:48] [77200] global_step=77200, grad_norm=4.657238960266113, loss=3.7088682651519775 -I0513 06:37:19.712205 140330632455936 logging_writer.py:48] [77300] global_step=77300, grad_norm=4.5682220458984375, loss=3.746717691421509 -I0513 06:37:57.269086 140330640848640 logging_writer.py:48] [77400] global_step=77400, grad_norm=3.5019302368164062, loss=3.7731432914733887 -I0513 06:38:34.035832 140330632455936 logging_writer.py:48] [77500] global_step=77500, grad_norm=5.8318963050842285, loss=3.6897573471069336 -I0513 06:39:11.275570 140330640848640 logging_writer.py:48] [77600] global_step=77600, grad_norm=5.976806640625, loss=3.6930665969848633 -I0513 06:39:48.476716 140330632455936 logging_writer.py:48] [77700] global_step=77700, grad_norm=5.286202907562256, loss=3.583258867263794 -I0513 06:40:25.353207 140330640848640 logging_writer.py:48] [77800] global_step=77800, grad_norm=6.991676330566406, loss=3.71274995803833 -I0513 06:41:02.216609 140330632455936 logging_writer.py:48] [77900] global_step=77900, grad_norm=7.630866050720215, loss=3.9102892875671387 -I0513 06:41:39.846865 140330640848640 logging_writer.py:48] [78000] global_step=78000, grad_norm=3.68612003326416, loss=3.6625661849975586 -I0513 06:42:16.682425 140330632455936 logging_writer.py:48] [78100] global_step=78100, grad_norm=3.644824743270874, loss=3.7135496139526367 -I0513 06:42:53.540818 140330640848640 logging_writer.py:48] [78200] global_step=78200, grad_norm=4.2014594078063965, loss=3.589132308959961 -I0513 06:43:31.038579 140330632455936 logging_writer.py:48] [78300] global_step=78300, grad_norm=6.301097869873047, loss=3.8720932006835938 -I0513 06:44:07.891976 140330640848640 logging_writer.py:48] [78400] global_step=78400, grad_norm=3.63702654838562, loss=3.70827579498291 -I0513 06:44:44.747225 140330632455936 logging_writer.py:48] [78500] global_step=78500, grad_norm=5.1371684074401855, loss=3.804873466491699 -I0513 06:45:22.413647 140330640848640 logging_writer.py:48] [78600] global_step=78600, grad_norm=4.721362113952637, loss=3.682798385620117 -I0513 06:45:59.214525 140330632455936 logging_writer.py:48] [78700] global_step=78700, grad_norm=4.5497026443481445, loss=3.818438768386841 -I0513 06:46:36.051218 140330640848640 logging_writer.py:48] [78800] global_step=78800, grad_norm=6.933742046356201, loss=3.7381796836853027 -I0513 06:47:13.575660 140330632455936 logging_writer.py:48] [78900] global_step=78900, grad_norm=2.6900875568389893, loss=3.526486873626709 -I0513 06:47:50.396089 140330640848640 logging_writer.py:48] [79000] global_step=79000, grad_norm=5.184777736663818, loss=3.7332544326782227 -I0513 06:48:27.271049 140330632455936 logging_writer.py:48] [79100] global_step=79100, grad_norm=5.100890159606934, loss=3.758392810821533 -I0513 06:49:04.909993 140330640848640 logging_writer.py:48] [79200] global_step=79200, grad_norm=4.841084003448486, loss=3.730787754058838 -I0513 06:49:41.763351 140330632455936 logging_writer.py:48] [79300] global_step=79300, grad_norm=8.546918869018555, loss=3.6919631958007812 -I0513 06:50:18.627419 140330640848640 logging_writer.py:48] [79400] global_step=79400, grad_norm=22.877117156982422, loss=3.955033779144287 -I0513 06:50:55.955269 140330632455936 logging_writer.py:48] [79500] global_step=79500, grad_norm=3.3079047203063965, loss=3.67149019241333 -I0513 06:51:33.157552 140330640848640 logging_writer.py:48] [79600] global_step=79600, grad_norm=3.279519557952881, loss=3.663867950439453 -I0513 06:52:09.972828 140330632455936 logging_writer.py:48] [79700] global_step=79700, grad_norm=3.969151020050049, loss=3.712500810623169 -I0513 06:52:47.590421 140330640848640 logging_writer.py:48] [79800] global_step=79800, grad_norm=4.833071231842041, loss=3.743773937225342 -I0513 06:53:24.463082 140330632455936 logging_writer.py:48] [79900] global_step=79900, grad_norm=4.423922061920166, loss=3.646617889404297 -I0513 06:54:01.200556 140330640848640 logging_writer.py:48] [80000] global_step=80000, grad_norm=28.221586227416992, loss=3.837369680404663 -I0513 06:54:38.972104 140330632455936 logging_writer.py:48] [80100] global_step=80100, grad_norm=3.881962537765503, loss=3.6733217239379883 -I0513 06:55:15.811893 140330640848640 logging_writer.py:48] [80200] global_step=80200, grad_norm=4.293656349182129, loss=3.657726526260376 -I0513 06:55:52.568244 140330632455936 logging_writer.py:48] [80300] global_step=80300, grad_norm=4.427382469177246, loss=3.7685670852661133 -I0513 06:56:30.235173 140330640848640 logging_writer.py:48] [80400] global_step=80400, grad_norm=4.206446170806885, loss=3.6105470657348633 -I0513 06:56:51.158092 140546196993216 spec.py:333] Evaluating on the training split. -I0513 06:57:02.355762 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 06:57:12.012284 140546196993216 spec.py:363] Evaluating on the test split. -I0513 06:57:12.883615 140546196993216 submission_runner.py:516] Time since start: 30396.94s, Step: 80458, {'train/accuracy': Array(0.00125558, dtype=float32), 'train/loss': Array(8.448164, dtype=float32), 'validation/accuracy': Array(0.00148, dtype=float32), 'validation/loss': Array(8.408929, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0017, dtype=float32), 'test/loss': Array(8.418403, dtype=float32), 'test/num_examples': 10000, 'score': 29998.31781888008, 'total_duration': 30396.9351439476, 'accumulated_submission_time': 29998.31781888008, 'accumulated_eval_time': 396.9780945777893, 'accumulated_logging_time': 0.9690923690795898} -I0513 06:57:12.937535 140330632455936 logging_writer.py:48] [80458] accumulated_eval_time=396.978, accumulated_logging_time=0.969092, accumulated_submission_time=29998.3, global_step=80458, preemption_count=0, score=29998.3, test/accuracy=0.0017000001389533281, test/loss=8.418402671813965, test/num_examples=10000, total_duration=30396.9, train/accuracy=0.0012555803405120969, train/loss=8.448163986206055, validation/accuracy=0.0014799999771639705, validation/loss=8.408928871154785, validation/num_examples=50000 -I0513 06:57:29.909366 140330640848640 logging_writer.py:48] [80500] global_step=80500, grad_norm=4.3604817390441895, loss=3.670694351196289 -I0513 06:58:06.634826 140330632455936 logging_writer.py:48] [80600] global_step=80600, grad_norm=3.7497291564941406, loss=3.545121192932129 -I0513 06:58:44.218338 140330640848640 logging_writer.py:48] [80700] global_step=80700, grad_norm=4.407201290130615, loss=3.612947940826416 -I0513 06:59:20.941333 140330632455936 logging_writer.py:48] [80800] global_step=80800, grad_norm=9.745924949645996, loss=4.0865631103515625 -I0513 06:59:57.798352 140330640848640 logging_writer.py:48] [80900] global_step=80900, grad_norm=5.294980525970459, loss=3.8920769691467285 -I0513 07:00:35.361910 140330632455936 logging_writer.py:48] [81000] global_step=81000, grad_norm=5.342458724975586, loss=3.899242401123047 -I0513 07:01:12.180294 140330640848640 logging_writer.py:48] [81100] global_step=81100, grad_norm=4.088430404663086, loss=3.840555191040039 -I0513 07:01:48.970706 140330632455936 logging_writer.py:48] [81200] global_step=81200, grad_norm=7.487679958343506, loss=3.7788116931915283 -I0513 07:02:26.581030 140330640848640 logging_writer.py:48] [81300] global_step=81300, grad_norm=3.333434820175171, loss=3.6949148178100586 -I0513 07:03:03.402619 140330632455936 logging_writer.py:48] [81400] global_step=81400, grad_norm=4.138752460479736, loss=3.7388055324554443 -I0513 07:03:40.232343 140330640848640 logging_writer.py:48] [81500] global_step=81500, grad_norm=3.3276383876800537, loss=3.6580677032470703 -I0513 07:04:17.719280 140330632455936 logging_writer.py:48] [81600] global_step=81600, grad_norm=5.316971778869629, loss=3.781006097793579 -I0513 07:04:54.587272 140330640848640 logging_writer.py:48] [81700] global_step=81700, grad_norm=4.367125511169434, loss=3.6154561042785645 -I0513 07:05:31.383226 140330632455936 logging_writer.py:48] [81800] global_step=81800, grad_norm=4.582855224609375, loss=3.661938190460205 -I0513 07:06:09.039717 140330640848640 logging_writer.py:48] [81900] global_step=81900, grad_norm=4.655252933502197, loss=3.764108180999756 -I0513 07:06:45.854231 140330632455936 logging_writer.py:48] [82000] global_step=82000, grad_norm=4.685189723968506, loss=3.7448973655700684 -I0513 07:07:22.694694 140330640848640 logging_writer.py:48] [82100] global_step=82100, grad_norm=3.67099666595459, loss=3.661623477935791 -I0513 07:08:00.340162 140330632455936 logging_writer.py:48] [82200] global_step=82200, grad_norm=6.234794616699219, loss=3.8002195358276367 -I0513 07:08:37.078202 140330640848640 logging_writer.py:48] [82300] global_step=82300, grad_norm=13.550300598144531, loss=3.6217920780181885 -I0513 07:09:13.932484 140330632455936 logging_writer.py:48] [82400] global_step=82400, grad_norm=3.8587498664855957, loss=3.631495475769043 -I0513 07:09:51.604167 140330640848640 logging_writer.py:48] [82500] global_step=82500, grad_norm=4.253410339355469, loss=3.7629470825195312 -I0513 07:10:28.438066 140330632455936 logging_writer.py:48] [82600] global_step=82600, grad_norm=5.122408866882324, loss=3.8258845806121826 -I0513 07:11:05.224337 140330640848640 logging_writer.py:48] [82700] global_step=82700, grad_norm=3.8801050186157227, loss=3.8083062171936035 -I0513 07:11:42.835816 140330632455936 logging_writer.py:48] [82800] global_step=82800, grad_norm=6.0307745933532715, loss=3.8386478424072266 -I0513 07:12:19.607496 140330640848640 logging_writer.py:48] [82900] global_step=82900, grad_norm=4.631067276000977, loss=3.6410317420959473 -I0513 07:12:56.459422 140330632455936 logging_writer.py:48] [83000] global_step=83000, grad_norm=13.350249290466309, loss=3.6552670001983643 -I0513 07:13:34.039330 140330640848640 logging_writer.py:48] [83100] global_step=83100, grad_norm=4.3594584465026855, loss=3.636547565460205 -I0513 07:14:10.824080 140330632455936 logging_writer.py:48] [83200] global_step=83200, grad_norm=5.215039253234863, loss=3.7967262268066406 -I0513 07:14:48.017091 140330640848640 logging_writer.py:48] [83300] global_step=83300, grad_norm=3.7474353313446045, loss=3.715390920639038 -I0513 07:15:25.363099 140330632455936 logging_writer.py:48] [83400] global_step=83400, grad_norm=3.8922643661499023, loss=3.657003402709961 -I0513 07:16:02.207293 140330640848640 logging_writer.py:48] [83500] global_step=83500, grad_norm=3.5506114959716797, loss=3.7588706016540527 -I0513 07:16:39.034344 140330632455936 logging_writer.py:48] [83600] global_step=83600, grad_norm=8.76722526550293, loss=3.622953176498413 -I0513 07:17:16.641891 140330640848640 logging_writer.py:48] [83700] global_step=83700, grad_norm=4.536018371582031, loss=3.7456276416778564 -I0513 07:17:53.518926 140330632455936 logging_writer.py:48] [83800] global_step=83800, grad_norm=4.077093124389648, loss=3.6660897731781006 -I0513 07:18:30.686109 140330640848640 logging_writer.py:48] [83900] global_step=83900, grad_norm=3.4527814388275146, loss=3.6512081623077393 -I0513 07:19:07.958862 140330632455936 logging_writer.py:48] [84000] global_step=84000, grad_norm=5.360208511352539, loss=3.68607759475708 -I0513 07:19:44.726428 140330640848640 logging_writer.py:48] [84100] global_step=84100, grad_norm=7.207712173461914, loss=3.952117919921875 -I0513 07:20:21.620165 140330632455936 logging_writer.py:48] [84200] global_step=84200, grad_norm=3.9100406169891357, loss=3.711930751800537 -I0513 07:20:59.206102 140330640848640 logging_writer.py:48] [84300] global_step=84300, grad_norm=4.449293613433838, loss=3.6870944499969482 -I0513 07:21:36.021001 140330632455936 logging_writer.py:48] [84400] global_step=84400, grad_norm=8.033103942871094, loss=3.8072690963745117 -I0513 07:22:12.872160 140330640848640 logging_writer.py:48] [84500] global_step=84500, grad_norm=3.5776264667510986, loss=3.8489365577697754 -I0513 07:22:50.496583 140330632455936 logging_writer.py:48] [84600] global_step=84600, grad_norm=3.480429172515869, loss=3.672394275665283 -I0513 07:23:27.285877 140330640848640 logging_writer.py:48] [84700] global_step=84700, grad_norm=6.809543609619141, loss=3.8988161087036133 -I0513 07:24:04.157198 140330632455936 logging_writer.py:48] [84800] global_step=84800, grad_norm=4.181275367736816, loss=3.734114170074463 -I0513 07:24:41.737306 140330640848640 logging_writer.py:48] [84900] global_step=84900, grad_norm=4.142294883728027, loss=3.6755125522613525 -I0513 07:25:18.505292 140330632455936 logging_writer.py:48] [85000] global_step=85000, grad_norm=4.370267868041992, loss=3.7085046768188477 -I0513 07:25:55.406044 140330640848640 logging_writer.py:48] [85100] global_step=85100, grad_norm=8.095829963684082, loss=3.684710741043091 -I0513 07:26:32.970919 140330632455936 logging_writer.py:48] [85200] global_step=85200, grad_norm=5.95587682723999, loss=3.6833243370056152 -I0513 07:27:09.742795 140330640848640 logging_writer.py:48] [85300] global_step=85300, grad_norm=4.294651508331299, loss=3.8861446380615234 -I0513 07:27:46.510770 140330632455936 logging_writer.py:48] [85400] global_step=85400, grad_norm=4.476201057434082, loss=3.730527639389038 -I0513 07:28:24.084104 140330640848640 logging_writer.py:48] [85500] global_step=85500, grad_norm=3.46254301071167, loss=3.8077375888824463 -I0513 07:29:00.895246 140330632455936 logging_writer.py:48] [85600] global_step=85600, grad_norm=3.1786367893218994, loss=3.727991819381714 -I0513 07:29:37.743256 140330640848640 logging_writer.py:48] [85700] global_step=85700, grad_norm=5.984392166137695, loss=3.6953201293945312 -I0513 07:30:15.205221 140330632455936 logging_writer.py:48] [85800] global_step=85800, grad_norm=4.9365410804748535, loss=3.9158365726470947 -I0513 07:30:29.564141 140546196993216 spec.py:333] Evaluating on the training split. -I0513 07:30:39.419680 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 07:30:49.455254 140546196993216 spec.py:363] Evaluating on the test split. -I0513 07:30:50.347849 140546196993216 submission_runner.py:516] Time since start: 32414.40s, Step: 85840, {'train/accuracy': Array(0.00139509, dtype=float32), 'train/loss': Array(8.513527, dtype=float32), 'validation/accuracy': Array(0.00146, dtype=float32), 'validation/loss': Array(8.475016, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0019, dtype=float32), 'test/loss': Array(8.485265, dtype=float32), 'test/num_examples': 10000, 'score': 31994.878091335297, 'total_duration': 32414.399368047714, 'accumulated_submission_time': 31994.878091335297, 'accumulated_eval_time': 417.75972962379456, 'accumulated_logging_time': 1.046638011932373} -I0513 07:30:50.401725 140330640848640 logging_writer.py:48] [85840] accumulated_eval_time=417.76, accumulated_logging_time=1.04664, accumulated_submission_time=31994.9, global_step=85840, preemption_count=0, score=31994.9, test/accuracy=0.0019000000320374966, test/loss=8.485264778137207, test/num_examples=10000, total_duration=32414.4, train/accuracy=0.0013950892025604844, train/loss=8.513526916503906, validation/accuracy=0.001459999941289425, validation/loss=8.475015640258789, validation/num_examples=50000 -I0513 07:31:14.036357 140330632455936 logging_writer.py:48] [85900] global_step=85900, grad_norm=3.808500051498413, loss=3.817429304122925 -I0513 07:31:50.839610 140330640848640 logging_writer.py:48] [86000] global_step=86000, grad_norm=5.710396766662598, loss=3.711862325668335 -I0513 07:32:28.540114 140330632455936 logging_writer.py:48] [86100] global_step=86100, grad_norm=4.4302754402160645, loss=3.6620054244995117 -I0513 07:33:05.243593 140330640848640 logging_writer.py:48] [86200] global_step=86200, grad_norm=5.166123390197754, loss=3.793114185333252 -I0513 07:33:42.089411 140330632455936 logging_writer.py:48] [86300] global_step=86300, grad_norm=4.067062854766846, loss=3.704857587814331 -I0513 07:34:19.405549 140330640848640 logging_writer.py:48] [86400] global_step=86400, grad_norm=10.743439674377441, loss=3.735095739364624 -I0513 07:34:56.439215 140330632455936 logging_writer.py:48] [86500] global_step=86500, grad_norm=2.9152019023895264, loss=3.627150535583496 -I0513 07:35:33.332906 140330640848640 logging_writer.py:48] [86600] global_step=86600, grad_norm=4.370521545410156, loss=3.6843042373657227 -I0513 07:36:10.945784 140330632455936 logging_writer.py:48] [86700] global_step=86700, grad_norm=10.071300506591797, loss=3.874650001525879 -I0513 07:36:47.805037 140330640848640 logging_writer.py:48] [86800] global_step=86800, grad_norm=3.346658229827881, loss=3.7135398387908936 -I0513 07:37:24.611358 140330632455936 logging_writer.py:48] [86900] global_step=86900, grad_norm=4.72181510925293, loss=3.79429292678833 -I0513 07:38:01.833200 140330640848640 logging_writer.py:48] [87000] global_step=87000, grad_norm=4.197606563568115, loss=3.649181365966797 -I0513 07:38:38.947530 140330632455936 logging_writer.py:48] [87100] global_step=87100, grad_norm=3.2356841564178467, loss=3.6637003421783447 -I0513 07:39:15.755295 140330640848640 logging_writer.py:48] [87200] global_step=87200, grad_norm=5.8492913246154785, loss=3.664050579071045 -I0513 07:39:52.945518 140330632455936 logging_writer.py:48] [87300] global_step=87300, grad_norm=2.7033638954162598, loss=3.8005833625793457 -I0513 07:40:30.069493 140330640848640 logging_writer.py:48] [87400] global_step=87400, grad_norm=3.140760898590088, loss=3.634202718734741 -I0513 07:41:06.885066 140330632455936 logging_writer.py:48] [87500] global_step=87500, grad_norm=3.3822546005249023, loss=3.568477153778076 -I0513 07:41:44.141892 140330640848640 logging_writer.py:48] [87600] global_step=87600, grad_norm=10.273377418518066, loss=3.8324248790740967 -I0513 07:42:21.280413 140330632455936 logging_writer.py:48] [87700] global_step=87700, grad_norm=4.6839213371276855, loss=3.6868467330932617 -I0513 07:42:58.063105 140330640848640 logging_writer.py:48] [87800] global_step=87800, grad_norm=5.802631855010986, loss=3.7671499252319336 -I0513 07:43:35.338426 140330632455936 logging_writer.py:48] [87900] global_step=87900, grad_norm=4.292171955108643, loss=3.8844566345214844 -I0513 07:44:12.376651 140330640848640 logging_writer.py:48] [88000] global_step=88000, grad_norm=5.062900066375732, loss=3.851940631866455 -I0513 07:44:49.231216 140330632455936 logging_writer.py:48] [88100] global_step=88100, grad_norm=17.83724021911621, loss=3.820922374725342 -I0513 07:45:26.524157 140330640848640 logging_writer.py:48] [88200] global_step=88200, grad_norm=4.969198226928711, loss=3.730607509613037 -I0513 07:46:03.664189 140330632455936 logging_writer.py:48] [88300] global_step=88300, grad_norm=4.503081798553467, loss=3.7727484703063965 -I0513 07:46:40.448812 140330640848640 logging_writer.py:48] [88400] global_step=88400, grad_norm=4.101708889007568, loss=3.696124792098999 -I0513 07:47:17.674875 140330632455936 logging_writer.py:48] [88500] global_step=88500, grad_norm=5.928832530975342, loss=3.7562146186828613 -I0513 07:47:54.747257 140330640848640 logging_writer.py:48] [88600] global_step=88600, grad_norm=3.8102591037750244, loss=3.642665147781372 -I0513 07:48:31.646864 140330632455936 logging_writer.py:48] [88700] global_step=88700, grad_norm=6.553823471069336, loss=3.850250482559204 -I0513 07:49:08.956309 140330640848640 logging_writer.py:48] [88800] global_step=88800, grad_norm=3.7759695053100586, loss=3.6644880771636963 -I0513 07:49:46.128979 140330632455936 logging_writer.py:48] [88900] global_step=88900, grad_norm=3.8416059017181396, loss=3.79587984085083 -I0513 07:50:22.948765 140330640848640 logging_writer.py:48] [89000] global_step=89000, grad_norm=5.151310920715332, loss=3.729548454284668 -I0513 07:51:00.188062 140330632455936 logging_writer.py:48] [89100] global_step=89100, grad_norm=6.675684928894043, loss=3.879709482192993 -I0513 07:51:36.967173 140330640848640 logging_writer.py:48] [89200] global_step=89200, grad_norm=9.659250259399414, loss=4.125170707702637 -I0513 07:52:14.012100 140330632455936 logging_writer.py:48] [89300] global_step=89300, grad_norm=9.926822662353516, loss=4.122852325439453 -I0513 07:52:51.276093 140330640848640 logging_writer.py:48] [89400] global_step=89400, grad_norm=3.5175087451934814, loss=3.7726688385009766 -I0513 07:53:28.370355 140330632455936 logging_writer.py:48] [89500] global_step=89500, grad_norm=5.905864715576172, loss=3.701249122619629 -I0513 07:54:05.127431 140330640848640 logging_writer.py:48] [89600] global_step=89600, grad_norm=5.382782459259033, loss=3.789618968963623 -I0513 07:54:42.493421 140330632455936 logging_writer.py:48] [89700] global_step=89700, grad_norm=5.071943283081055, loss=3.682936191558838 -I0513 07:55:19.571375 140330640848640 logging_writer.py:48] [89800] global_step=89800, grad_norm=4.976498126983643, loss=3.8719992637634277 -I0513 07:55:56.346637 140330632455936 logging_writer.py:48] [89900] global_step=89900, grad_norm=5.642221927642822, loss=3.6912548542022705 -I0513 07:56:33.685874 140330640848640 logging_writer.py:48] [90000] global_step=90000, grad_norm=3.9541499614715576, loss=3.6538870334625244 -I0513 07:57:10.926332 140330632455936 logging_writer.py:48] [90100] global_step=90100, grad_norm=14.245800971984863, loss=3.664566993713379 -I0513 07:57:47.731969 140330640848640 logging_writer.py:48] [90200] global_step=90200, grad_norm=3.9937429428100586, loss=3.655139923095703 -I0513 07:58:25.029311 140330632455936 logging_writer.py:48] [90300] global_step=90300, grad_norm=3.3769164085388184, loss=3.8674795627593994 -I0513 07:59:02.215175 140330640848640 logging_writer.py:48] [90400] global_step=90400, grad_norm=4.024641036987305, loss=3.829083204269409 -I0513 07:59:39.105012 140330632455936 logging_writer.py:48] [90500] global_step=90500, grad_norm=4.1054301261901855, loss=3.836758852005005 -I0513 08:00:16.457448 140330640848640 logging_writer.py:48] [90600] global_step=90600, grad_norm=5.207498073577881, loss=3.7125067710876465 -I0513 08:00:53.616231 140330632455936 logging_writer.py:48] [90700] global_step=90700, grad_norm=4.796743869781494, loss=3.7784547805786133 -I0513 08:01:30.439885 140330640848640 logging_writer.py:48] [90800] global_step=90800, grad_norm=6.21027946472168, loss=3.60261869430542 -I0513 08:02:07.799706 140330632455936 logging_writer.py:48] [90900] global_step=90900, grad_norm=4.541248798370361, loss=3.674804210662842 -I0513 08:02:44.937130 140330640848640 logging_writer.py:48] [91000] global_step=91000, grad_norm=3.656707763671875, loss=3.6694111824035645 -I0513 08:03:21.788177 140330632455936 logging_writer.py:48] [91100] global_step=91100, grad_norm=3.0911290645599365, loss=3.651654005050659 -I0513 08:03:59.071532 140330640848640 logging_writer.py:48] [91200] global_step=91200, grad_norm=5.0620646476745605, loss=3.7096142768859863 -I0513 08:04:06.720214 140546196993216 spec.py:333] Evaluating on the training split. -I0513 08:04:17.179907 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 08:04:27.440743 140546196993216 spec.py:363] Evaluating on the test split. -I0513 08:04:28.325505 140546196993216 submission_runner.py:516] Time since start: 34432.38s, Step: 91222, {'train/accuracy': Array(0.00145488, dtype=float32), 'train/loss': Array(8.590439, dtype=float32), 'validation/accuracy': Array(0.00144, dtype=float32), 'validation/loss': Array(8.561899, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0017, dtype=float32), 'test/loss': Array(8.573786, dtype=float32), 'test/num_examples': 10000, 'score': 33991.13047719002, 'total_duration': 34432.37569451332, 'accumulated_submission_time': 33991.13047719002, 'accumulated_eval_time': 439.36161637306213, 'accumulated_logging_time': 1.124169111251831} -I0513 08:04:28.360752 140330632455936 logging_writer.py:48] [91222] accumulated_eval_time=439.362, accumulated_logging_time=1.12417, accumulated_submission_time=33991.1, global_step=91222, preemption_count=0, score=33991.1, test/accuracy=0.0017000001389533281, test/loss=8.573785781860352, test/num_examples=10000, total_duration=34432.4, train/accuracy=0.001454878831282258, train/loss=8.590438842773438, validation/accuracy=0.0014400000218302011, validation/loss=8.561899185180664, validation/num_examples=50000 -I0513 08:04:58.222579 140330640848640 logging_writer.py:48] [91300] global_step=91300, grad_norm=2.547637939453125, loss=3.7277615070343018 -I0513 08:05:35.302780 140330632455936 logging_writer.py:48] [91400] global_step=91400, grad_norm=4.451051712036133, loss=3.7435569763183594 -I0513 08:06:12.561839 140330640848640 logging_writer.py:48] [91500] global_step=91500, grad_norm=3.2788939476013184, loss=3.7302980422973633 -I0513 08:06:49.768332 140330632455936 logging_writer.py:48] [91600] global_step=91600, grad_norm=3.72701358795166, loss=3.7888941764831543 -I0513 08:07:26.697923 140330640848640 logging_writer.py:48] [91700] global_step=91700, grad_norm=4.7237348556518555, loss=3.8680853843688965 -I0513 08:08:04.076333 140330632455936 logging_writer.py:48] [91800] global_step=91800, grad_norm=5.584935665130615, loss=3.7523326873779297 -I0513 08:08:40.864342 140330640848640 logging_writer.py:48] [91900] global_step=91900, grad_norm=6.965238571166992, loss=3.829820394515991 -I0513 08:09:17.990829 140330632455936 logging_writer.py:48] [92000] global_step=92000, grad_norm=3.290675401687622, loss=3.83518123626709 -I0513 08:09:55.276668 140330640848640 logging_writer.py:48] [92100] global_step=92100, grad_norm=4.310474872589111, loss=3.876110553741455 -I0513 08:10:32.353444 140330632455936 logging_writer.py:48] [92200] global_step=92200, grad_norm=4.570540428161621, loss=3.6848669052124023 -I0513 08:11:09.174777 140330640848640 logging_writer.py:48] [92300] global_step=92300, grad_norm=5.217610836029053, loss=3.8911352157592773 -I0513 08:11:46.400915 140330632455936 logging_writer.py:48] [92400] global_step=92400, grad_norm=3.680069923400879, loss=3.6798558235168457 -I0513 08:12:23.604820 140330640848640 logging_writer.py:48] [92500] global_step=92500, grad_norm=5.909483432769775, loss=3.7619094848632812 -I0513 08:13:00.431791 140330632455936 logging_writer.py:48] [92600] global_step=92600, grad_norm=10.775208473205566, loss=3.735401153564453 -I0513 08:13:37.703605 140330640848640 logging_writer.py:48] [92700] global_step=92700, grad_norm=3.2637531757354736, loss=3.680539846420288 -I0513 08:14:14.552265 140330632455936 logging_writer.py:48] [92800] global_step=92800, grad_norm=4.919007778167725, loss=3.851997137069702 -I0513 08:14:51.597404 140330640848640 logging_writer.py:48] [92900] global_step=92900, grad_norm=32.958885192871094, loss=3.6928515434265137 -I0513 08:15:28.891688 140330632455936 logging_writer.py:48] [93000] global_step=93000, grad_norm=4.020893573760986, loss=3.813184976577759 -I0513 08:16:06.056349 140330640848640 logging_writer.py:48] [93100] global_step=93100, grad_norm=4.284795761108398, loss=3.804291009902954 -I0513 08:16:42.886839 140330632455936 logging_writer.py:48] [93200] global_step=93200, grad_norm=11.706283569335938, loss=3.7293760776519775 -I0513 08:17:20.205594 140330640848640 logging_writer.py:48] [93300] global_step=93300, grad_norm=3.4159865379333496, loss=3.648869037628174 -I0513 08:17:57.009018 140330632455936 logging_writer.py:48] [93400] global_step=93400, grad_norm=4.351746082305908, loss=3.7468314170837402 -I0513 08:18:34.056361 140330640848640 logging_writer.py:48] [93500] global_step=93500, grad_norm=4.4123454093933105, loss=3.940256357192993 -I0513 08:19:11.342744 140330632455936 logging_writer.py:48] [93600] global_step=93600, grad_norm=3.144557237625122, loss=3.7008910179138184 -I0513 08:19:48.096064 140330640848640 logging_writer.py:48] [93700] global_step=93700, grad_norm=4.346450328826904, loss=3.745946168899536 -I0513 08:20:25.215074 140330632455936 logging_writer.py:48] [93800] global_step=93800, grad_norm=2.414959192276001, loss=3.771294116973877 -I0513 08:21:02.505450 140330640848640 logging_writer.py:48] [93900] global_step=93900, grad_norm=4.031586170196533, loss=3.758655309677124 -I0513 08:21:39.223051 140330632455936 logging_writer.py:48] [94000] global_step=94000, grad_norm=3.6518430709838867, loss=3.7202234268188477 -I0513 08:22:16.330329 140330640848640 logging_writer.py:48] [94100] global_step=94100, grad_norm=5.830986976623535, loss=4.001197338104248 -I0513 08:22:53.648804 140330632455936 logging_writer.py:48] [94200] global_step=94200, grad_norm=8.6563138961792, loss=3.7309765815734863 -I0513 08:23:30.355443 140330640848640 logging_writer.py:48] [94300] global_step=94300, grad_norm=3.555253505706787, loss=3.7525475025177 -I0513 08:24:07.323934 140330632455936 logging_writer.py:48] [94400] global_step=94400, grad_norm=9.551026344299316, loss=3.858856201171875 -I0513 08:24:44.631502 140330640848640 logging_writer.py:48] [94500] global_step=94500, grad_norm=4.011026382446289, loss=3.712745428085327 -I0513 08:25:21.581644 140330632455936 logging_writer.py:48] [94600] global_step=94600, grad_norm=4.665242671966553, loss=3.855640172958374 -I0513 08:25:58.758017 140330640848640 logging_writer.py:48] [94700] global_step=94700, grad_norm=7.016716003417969, loss=3.9069907665252686 -I0513 08:26:36.070134 140330632455936 logging_writer.py:48] [94800] global_step=94800, grad_norm=7.016180992126465, loss=3.7113358974456787 -I0513 08:27:12.985982 140330640848640 logging_writer.py:48] [94900] global_step=94900, grad_norm=8.3674955368042, loss=3.9818115234375 -I0513 08:27:50.148982 140330632455936 logging_writer.py:48] [95000] global_step=95000, grad_norm=4.58731746673584, loss=3.57319974899292 -I0513 08:28:27.491508 140330640848640 logging_writer.py:48] [95100] global_step=95100, grad_norm=3.1140894889831543, loss=3.6465017795562744 -I0513 08:29:04.317138 140330632455936 logging_writer.py:48] [95200] global_step=95200, grad_norm=3.72571063041687, loss=3.6965134143829346 -I0513 08:29:41.483265 140330640848640 logging_writer.py:48] [95300] global_step=95300, grad_norm=2.9257566928863525, loss=3.798872232437134 -I0513 08:30:18.713860 140330632455936 logging_writer.py:48] [95400] global_step=95400, grad_norm=4.151865482330322, loss=3.8605289459228516 -I0513 08:30:55.542095 140330640848640 logging_writer.py:48] [95500] global_step=95500, grad_norm=3.662628173828125, loss=3.734103202819824 -I0513 08:31:32.318748 140330632455936 logging_writer.py:48] [95600] global_step=95600, grad_norm=4.438009738922119, loss=3.785069704055786 -I0513 08:32:09.865225 140330640848640 logging_writer.py:48] [95700] global_step=95700, grad_norm=4.739010334014893, loss=3.829413890838623 -I0513 08:32:46.710963 140330632455936 logging_writer.py:48] [95800] global_step=95800, grad_norm=7.633068084716797, loss=4.066880226135254 -I0513 08:33:23.866908 140330640848640 logging_writer.py:48] [95900] global_step=95900, grad_norm=5.583685874938965, loss=3.8612060546875 -I0513 08:34:01.157684 140330632455936 logging_writer.py:48] [96000] global_step=96000, grad_norm=9.669242858886719, loss=3.9009947776794434 -I0513 08:34:37.948996 140330640848640 logging_writer.py:48] [96100] global_step=96100, grad_norm=5.557234764099121, loss=3.8858442306518555 -I0513 08:35:15.122040 140330632455936 logging_writer.py:48] [96200] global_step=96200, grad_norm=3.537205934524536, loss=3.743440866470337 -I0513 08:35:52.377493 140330640848640 logging_writer.py:48] [96300] global_step=96300, grad_norm=6.2380571365356445, loss=3.91030216217041 -I0513 08:36:29.148234 140330632455936 logging_writer.py:48] [96400] global_step=96400, grad_norm=5.585464954376221, loss=3.844693660736084 -I0513 08:37:06.333168 140330640848640 logging_writer.py:48] [96500] global_step=96500, grad_norm=3.815178632736206, loss=3.783318042755127 -I0513 08:37:43.685847 140330632455936 logging_writer.py:48] [96600] global_step=96600, grad_norm=7.198297023773193, loss=3.8589224815368652 -I0513 08:37:44.714356 140546196993216 spec.py:333] Evaluating on the training split. -I0513 08:37:55.643681 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 08:38:05.712289 140546196993216 spec.py:363] Evaluating on the test split. -I0513 08:38:06.593746 140546196993216 submission_runner.py:516] Time since start: 36450.65s, Step: 96604, {'train/accuracy': Array(0.00137516, dtype=float32), 'train/loss': Array(8.670943, dtype=float32), 'validation/accuracy': Array(0.00134, dtype=float32), 'validation/loss': Array(8.643026, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0016, dtype=float32), 'test/loss': Array(8.656847, dtype=float32), 'test/num_examples': 10000, 'score': 35987.429832696915, 'total_duration': 36450.645072460175, 'accumulated_submission_time': 35987.429832696915, 'accumulated_eval_time': 461.238737821579, 'accumulated_logging_time': 1.1705234050750732} -I0513 08:38:06.635002 140330640848640 logging_writer.py:48] [96604] accumulated_eval_time=461.239, accumulated_logging_time=1.17052, accumulated_submission_time=35987.4, global_step=96604, preemption_count=0, score=35987.4, test/accuracy=0.001600000075995922, test/loss=8.65684700012207, test/num_examples=10000, total_duration=36450.6, train/accuracy=0.0013751593651250005, train/loss=8.670943260192871, validation/accuracy=0.001339999958872795, validation/loss=8.643026351928711, validation/num_examples=50000 -I0513 08:38:43.240749 140330632455936 logging_writer.py:48] [96700] global_step=96700, grad_norm=10.021149635314941, loss=4.164205074310303 -I0513 08:39:20.383203 140330640848640 logging_writer.py:48] [96800] global_step=96800, grad_norm=4.1400532722473145, loss=3.875054121017456 -I0513 08:39:57.701235 140330632455936 logging_writer.py:48] [96900] global_step=96900, grad_norm=6.066524028778076, loss=3.7650132179260254 -I0513 08:40:34.535938 140330640848640 logging_writer.py:48] [97000] global_step=97000, grad_norm=22.117839813232422, loss=3.8674890995025635 -I0513 08:41:11.757058 140330632455936 logging_writer.py:48] [97100] global_step=97100, grad_norm=3.9644839763641357, loss=3.7056915760040283 -I0513 08:41:49.054766 140330640848640 logging_writer.py:48] [97200] global_step=97200, grad_norm=11.715841293334961, loss=3.734234094619751 -I0513 08:42:25.867850 140330632455936 logging_writer.py:48] [97300] global_step=97300, grad_norm=9.084395408630371, loss=3.59783935546875 -I0513 08:43:02.972992 140330640848640 logging_writer.py:48] [97400] global_step=97400, grad_norm=2.992642879486084, loss=3.6115639209747314 -I0513 08:43:40.242797 140330632455936 logging_writer.py:48] [97500] global_step=97500, grad_norm=4.084071159362793, loss=3.876377582550049 -I0513 08:44:17.025499 140330640848640 logging_writer.py:48] [97600] global_step=97600, grad_norm=3.447087049484253, loss=3.6784024238586426 -I0513 08:44:54.070517 140330632455936 logging_writer.py:48] [97700] global_step=97700, grad_norm=2.899346113204956, loss=3.7545361518859863 -I0513 08:45:31.464104 140330640848640 logging_writer.py:48] [97800] global_step=97800, grad_norm=5.158788204193115, loss=3.7822744846343994 -I0513 08:46:08.215013 140330632455936 logging_writer.py:48] [97900] global_step=97900, grad_norm=2.6765317916870117, loss=3.6032869815826416 -I0513 08:46:45.343226 140330640848640 logging_writer.py:48] [98000] global_step=98000, grad_norm=19.483348846435547, loss=3.768779754638672 -I0513 08:47:22.651153 140330632455936 logging_writer.py:48] [98100] global_step=98100, grad_norm=4.160740375518799, loss=3.7848563194274902 -I0513 08:47:59.445968 140330640848640 logging_writer.py:48] [98200] global_step=98200, grad_norm=5.433201789855957, loss=3.80072283744812 -I0513 08:48:36.621837 140330632455936 logging_writer.py:48] [98300] global_step=98300, grad_norm=6.115579605102539, loss=3.640069007873535 -I0513 08:49:13.963984 140330640848640 logging_writer.py:48] [98400] global_step=98400, grad_norm=20.02985954284668, loss=3.7205514907836914 -I0513 08:49:50.769139 140330632455936 logging_writer.py:48] [98500] global_step=98500, grad_norm=3.7253966331481934, loss=3.7649407386779785 -I0513 08:50:27.925536 140330640848640 logging_writer.py:48] [98600] global_step=98600, grad_norm=14.66468334197998, loss=3.750187397003174 -I0513 08:51:05.237031 140330632455936 logging_writer.py:48] [98700] global_step=98700, grad_norm=4.192646503448486, loss=3.674121856689453 -I0513 08:51:42.032039 140330640848640 logging_writer.py:48] [98800] global_step=98800, grad_norm=5.4085283279418945, loss=3.803826332092285 -I0513 08:52:19.209302 140330632455936 logging_writer.py:48] [98900] global_step=98900, grad_norm=5.3144426345825195, loss=3.753774881362915 -I0513 08:52:56.495500 140330640848640 logging_writer.py:48] [99000] global_step=99000, grad_norm=5.616574764251709, loss=3.930765151977539 -I0513 08:53:33.433012 140330632455936 logging_writer.py:48] [99100] global_step=99100, grad_norm=2.150057554244995, loss=3.6343374252319336 -I0513 08:54:10.633023 140330640848640 logging_writer.py:48] [99200] global_step=99200, grad_norm=4.622518062591553, loss=3.790588855743408 -I0513 08:54:47.987453 140330632455936 logging_writer.py:48] [99300] global_step=99300, grad_norm=9.94797420501709, loss=3.825784206390381 -I0513 08:55:24.760805 140330640848640 logging_writer.py:48] [99400] global_step=99400, grad_norm=29.408967971801758, loss=3.872987747192383 -I0513 08:56:01.978032 140330632455936 logging_writer.py:48] [99500] global_step=99500, grad_norm=3.762756824493408, loss=3.977792739868164 -I0513 08:56:39.269338 140330640848640 logging_writer.py:48] [99600] global_step=99600, grad_norm=6.849031448364258, loss=3.8683109283447266 -I0513 08:57:16.111933 140330632455936 logging_writer.py:48] [99700] global_step=99700, grad_norm=2.9914586544036865, loss=3.7217633724212646 -I0513 08:57:53.219966 140330640848640 logging_writer.py:48] [99800] global_step=99800, grad_norm=3.6575682163238525, loss=3.940422773361206 -I0513 08:58:30.585156 140330632455936 logging_writer.py:48] [99900] global_step=99900, grad_norm=3.7437429428100586, loss=3.8181705474853516 -I0513 08:59:07.420946 140330640848640 logging_writer.py:48] [100000] global_step=100000, grad_norm=5.213537216186523, loss=3.7580344676971436 -I0513 08:59:44.609855 140330632455936 logging_writer.py:48] [100100] global_step=100100, grad_norm=5.768113613128662, loss=3.817657947540283 -I0513 09:00:21.954509 140330640848640 logging_writer.py:48] [100200] global_step=100200, grad_norm=3.1379637718200684, loss=3.766251564025879 -I0513 09:00:58.774679 140330632455936 logging_writer.py:48] [100300] global_step=100300, grad_norm=3.5252156257629395, loss=3.702547073364258 -I0513 09:01:35.959953 140330640848640 logging_writer.py:48] [100400] global_step=100400, grad_norm=5.090540409088135, loss=3.6574957370758057 -I0513 09:02:13.309986 140330632455936 logging_writer.py:48] [100500] global_step=100500, grad_norm=2.606562852859497, loss=3.6551356315612793 -I0513 09:02:50.106312 140330640848640 logging_writer.py:48] [100600] global_step=100600, grad_norm=6.8358612060546875, loss=3.848422050476074 -I0513 09:03:27.249743 140330632455936 logging_writer.py:48] [100700] global_step=100700, grad_norm=4.024737358093262, loss=3.770709753036499 -I0513 09:04:04.605352 140330640848640 logging_writer.py:48] [100800] global_step=100800, grad_norm=3.2217373847961426, loss=3.7488934993743896 -I0513 09:04:41.432550 140330632455936 logging_writer.py:48] [100900] global_step=100900, grad_norm=3.2105562686920166, loss=3.8217697143554688 -I0513 09:05:18.605648 140330640848640 logging_writer.py:48] [101000] global_step=101000, grad_norm=3.19221830368042, loss=3.730053424835205 -I0513 09:05:55.924958 140330632455936 logging_writer.py:48] [101100] global_step=101100, grad_norm=4.2420454025268555, loss=3.748396873474121 -I0513 09:06:32.685063 140330640848640 logging_writer.py:48] [101200] global_step=101200, grad_norm=3.3248651027679443, loss=3.761288642883301 -I0513 09:07:09.952535 140330632455936 logging_writer.py:48] [101300] global_step=101300, grad_norm=7.927113056182861, loss=3.803004741668701 -I0513 09:07:47.210090 140330640848640 logging_writer.py:48] [101400] global_step=101400, grad_norm=4.214240074157715, loss=3.6951756477355957 -I0513 09:08:24.050827 140330632455936 logging_writer.py:48] [101500] global_step=101500, grad_norm=5.226306915283203, loss=3.6563315391540527 -I0513 09:09:01.340651 140330640848640 logging_writer.py:48] [101600] global_step=101600, grad_norm=4.479387283325195, loss=3.6535587310791016 -I0513 09:09:38.636371 140330632455936 logging_writer.py:48] [101700] global_step=101700, grad_norm=3.873558759689331, loss=3.8200595378875732 -I0513 09:10:15.405613 140330640848640 logging_writer.py:48] [101800] global_step=101800, grad_norm=8.283503532409668, loss=3.765413284301758 -I0513 09:10:52.454773 140330632455936 logging_writer.py:48] [101900] global_step=101900, grad_norm=3.9880592823028564, loss=3.7171432971954346 -I0513 09:11:22.987013 140546196993216 spec.py:333] Evaluating on the training split. -I0513 09:11:34.036427 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 09:11:43.888634 140546196993216 spec.py:363] Evaluating on the test split. -I0513 09:11:44.765812 140546196993216 submission_runner.py:516] Time since start: 38468.82s, Step: 101983, {'train/accuracy': Array(0.00131537, dtype=float32), 'train/loss': Array(8.769457, dtype=float32), 'validation/accuracy': Array(0.00138, dtype=float32), 'validation/loss': Array(8.745812, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0016, dtype=float32), 'test/loss': Array(8.761139, dtype=float32), 'test/num_examples': 10000, 'score': 37983.7277469635, 'total_duration': 38468.81688165665, 'accumulated_submission_time': 37983.7277469635, 'accumulated_eval_time': 483.01501297950745, 'accumulated_logging_time': 1.223264217376709} -I0513 09:11:44.804268 140330640848640 logging_writer.py:48] [101983] accumulated_eval_time=483.015, accumulated_logging_time=1.22326, accumulated_submission_time=37983.7, global_step=101983, preemption_count=0, score=37983.7, test/accuracy=0.001600000075995922, test/loss=8.761138916015625, test/num_examples=10000, total_duration=38468.8, train/accuracy=0.0013153698528185487, train/loss=8.76945686340332, validation/accuracy=0.0013799999142065644, validation/loss=8.74581241607666, validation/num_examples=50000 -I0513 09:11:52.402562 140330632455936 logging_writer.py:48] [102000] global_step=102000, grad_norm=3.1575124263763428, loss=3.6576619148254395 -I0513 09:12:29.195251 140330640848640 logging_writer.py:48] [102100] global_step=102100, grad_norm=4.6281538009643555, loss=3.748225212097168 -I0513 09:13:06.301305 140330632455936 logging_writer.py:48] [102200] global_step=102200, grad_norm=2.7923035621643066, loss=3.787606716156006 -I0513 09:13:43.600332 140330640848640 logging_writer.py:48] [102300] global_step=102300, grad_norm=6.834512710571289, loss=4.025428771972656 -I0513 09:14:20.412166 140330632455936 logging_writer.py:48] [102400] global_step=102400, grad_norm=5.147302150726318, loss=3.82944393157959 -I0513 09:14:57.458414 140330640848640 logging_writer.py:48] [102500] global_step=102500, grad_norm=3.5125365257263184, loss=3.9068799018859863 -I0513 09:15:34.770300 140330632455936 logging_writer.py:48] [102600] global_step=102600, grad_norm=4.619770050048828, loss=3.7374935150146484 -I0513 09:16:11.678741 140330640848640 logging_writer.py:48] [102700] global_step=102700, grad_norm=3.0486178398132324, loss=3.7172420024871826 -I0513 09:16:48.900154 140330632455936 logging_writer.py:48] [102800] global_step=102800, grad_norm=4.0725812911987305, loss=3.8248727321624756 -I0513 09:17:26.174721 140330640848640 logging_writer.py:48] [102900] global_step=102900, grad_norm=5.263497829437256, loss=3.8242504596710205 -I0513 09:18:03.014113 140330632455936 logging_writer.py:48] [103000] global_step=103000, grad_norm=4.610568046569824, loss=3.871041774749756 -I0513 09:18:39.887651 140330640848640 logging_writer.py:48] [103100] global_step=103100, grad_norm=4.167436122894287, loss=3.850247383117676 -I0513 09:19:17.542887 140330632455936 logging_writer.py:48] [103200] global_step=103200, grad_norm=7.477087020874023, loss=3.87408447265625 -I0513 09:19:54.351459 140330640848640 logging_writer.py:48] [103300] global_step=103300, grad_norm=4.102344989776611, loss=3.740440845489502 -I0513 09:20:31.487922 140330632455936 logging_writer.py:48] [103400] global_step=103400, grad_norm=11.069900512695312, loss=3.7372560501098633 -I0513 09:21:08.770216 140330640848640 logging_writer.py:48] [103500] global_step=103500, grad_norm=4.9087443351745605, loss=3.787257671356201 -I0513 09:21:45.538155 140330632455936 logging_writer.py:48] [103600] global_step=103600, grad_norm=3.2853171825408936, loss=3.6824209690093994 -I0513 09:22:22.695398 140330640848640 logging_writer.py:48] [103700] global_step=103700, grad_norm=6.016867160797119, loss=3.995199680328369 -I0513 09:22:59.943579 140330632455936 logging_writer.py:48] [103800] global_step=103800, grad_norm=4.7697529792785645, loss=3.890556812286377 -I0513 09:23:36.758546 140330640848640 logging_writer.py:48] [103900] global_step=103900, grad_norm=7.359390735626221, loss=3.772035598754883 -I0513 09:24:13.997348 140330632455936 logging_writer.py:48] [104000] global_step=104000, grad_norm=4.55443000793457, loss=3.846228837966919 -I0513 09:24:51.350125 140330640848640 logging_writer.py:48] [104100] global_step=104100, grad_norm=4.29936408996582, loss=3.9009737968444824 -I0513 09:25:28.092969 140330632455936 logging_writer.py:48] [104200] global_step=104200, grad_norm=3.7769522666931152, loss=3.8696813583374023 -I0513 09:26:05.273176 140330640848640 logging_writer.py:48] [104300] global_step=104300, grad_norm=5.869290828704834, loss=3.740556001663208 -I0513 09:26:42.537182 140330632455936 logging_writer.py:48] [104400] global_step=104400, grad_norm=3.3916916847229004, loss=3.6460437774658203 -I0513 09:27:19.353105 140330640848640 logging_writer.py:48] [104500] global_step=104500, grad_norm=2.8064730167388916, loss=3.7474963665008545 -I0513 09:27:56.399655 140330632455936 logging_writer.py:48] [104600] global_step=104600, grad_norm=3.9873008728027344, loss=3.7899160385131836 -I0513 09:28:33.763944 140330640848640 logging_writer.py:48] [104700] global_step=104700, grad_norm=5.763408184051514, loss=3.755129337310791 -I0513 09:29:10.632251 140330632455936 logging_writer.py:48] [104800] global_step=104800, grad_norm=4.924015998840332, loss=3.692098617553711 -I0513 09:29:47.917005 140330640848640 logging_writer.py:48] [104900] global_step=104900, grad_norm=3.200927495956421, loss=3.756887435913086 -I0513 09:30:25.229944 140330632455936 logging_writer.py:48] [105000] global_step=105000, grad_norm=4.498651027679443, loss=3.8319554328918457 -I0513 09:31:01.966533 140330640848640 logging_writer.py:48] [105100] global_step=105100, grad_norm=3.713254451751709, loss=3.716759204864502 -I0513 09:31:38.834967 140330632455936 logging_writer.py:48] [105200] global_step=105200, grad_norm=3.3872196674346924, loss=3.732008934020996 -I0513 09:32:16.483524 140330640848640 logging_writer.py:48] [105300] global_step=105300, grad_norm=7.569216251373291, loss=3.971698760986328 -I0513 09:32:53.257987 140330632455936 logging_writer.py:48] [105400] global_step=105400, grad_norm=5.086342811584473, loss=3.839622735977173 -I0513 09:33:30.390529 140330640848640 logging_writer.py:48] [105500] global_step=105500, grad_norm=7.8396806716918945, loss=3.8229987621307373 -I0513 09:34:07.628909 140330632455936 logging_writer.py:48] [105600] global_step=105600, grad_norm=5.020621299743652, loss=3.6654438972473145 -I0513 09:34:44.560686 140330640848640 logging_writer.py:48] [105700] global_step=105700, grad_norm=5.56778621673584, loss=3.784489870071411 -I0513 09:35:21.734053 140330632455936 logging_writer.py:48] [105800] global_step=105800, grad_norm=5.003288269042969, loss=3.872894763946533 -I0513 09:35:58.968318 140330640848640 logging_writer.py:48] [105900] global_step=105900, grad_norm=6.976188659667969, loss=3.8329274654388428 -I0513 09:36:35.788533 140330632455936 logging_writer.py:48] [106000] global_step=106000, grad_norm=2.743333578109741, loss=3.7737574577331543 -I0513 09:37:13.035808 140330640848640 logging_writer.py:48] [106100] global_step=106100, grad_norm=3.813394784927368, loss=3.721386432647705 -I0513 09:37:50.385686 140330632455936 logging_writer.py:48] [106200] global_step=106200, grad_norm=18.875595092773438, loss=3.73864483833313 -I0513 09:38:27.243727 140330640848640 logging_writer.py:48] [106300] global_step=106300, grad_norm=7.865479469299316, loss=3.9503183364868164 -I0513 09:39:04.417283 140330632455936 logging_writer.py:48] [106400] global_step=106400, grad_norm=7.27689790725708, loss=3.7091264724731445 -I0513 09:39:41.742119 140330640848640 logging_writer.py:48] [106500] global_step=106500, grad_norm=3.848013162612915, loss=3.629521131515503 -I0513 09:40:18.650873 140330632455936 logging_writer.py:48] [106600] global_step=106600, grad_norm=5.056128978729248, loss=3.7617225646972656 -I0513 09:40:55.807839 140330640848640 logging_writer.py:48] [106700] global_step=106700, grad_norm=6.690022945404053, loss=3.903895378112793 -I0513 09:41:33.114153 140330632455936 logging_writer.py:48] [106800] global_step=106800, grad_norm=4.426621913909912, loss=3.8032302856445312 -I0513 09:42:09.941118 140330640848640 logging_writer.py:48] [106900] global_step=106900, grad_norm=4.592653751373291, loss=3.744769811630249 -I0513 09:42:47.119687 140330632455936 logging_writer.py:48] [107000] global_step=107000, grad_norm=9.12771987915039, loss=3.9798007011413574 -I0513 09:43:24.382512 140330640848640 logging_writer.py:48] [107100] global_step=107100, grad_norm=4.271956443786621, loss=3.756532669067383 -I0513 09:44:01.118960 140330632455936 logging_writer.py:48] [107200] global_step=107200, grad_norm=4.971668720245361, loss=3.6858863830566406 -I0513 09:44:37.873035 140330640848640 logging_writer.py:48] [107300] global_step=107300, grad_norm=3.9434478282928467, loss=3.899979829788208 -I0513 09:45:00.939816 140546196993216 spec.py:333] Evaluating on the training split. -I0513 09:45:11.021727 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 09:45:20.758059 140546196993216 spec.py:363] Evaluating on the test split. -I0513 09:45:21.630462 140546196993216 submission_runner.py:516] Time since start: 40485.68s, Step: 107362, {'train/accuracy': Array(0.00085698, dtype=float32), 'train/loss': Array(8.899464, dtype=float32), 'validation/accuracy': Array(0.00134, dtype=float32), 'validation/loss': Array(8.870979, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0014, dtype=float32), 'test/loss': Array(8.886436, dtype=float32), 'test/num_examples': 10000, 'score': 39979.80666279793, 'total_duration': 40485.68164682388, 'accumulated_submission_time': 39979.80666279793, 'accumulated_eval_time': 503.7032513618469, 'accumulated_logging_time': 1.2757625579833984} -I0513 09:45:21.668572 140330632455936 logging_writer.py:48] [107362] accumulated_eval_time=503.703, accumulated_logging_time=1.27576, accumulated_submission_time=39979.8, global_step=107362, preemption_count=0, score=39979.8, test/accuracy=0.0014000000664964318, test/loss=8.886436462402344, test/num_examples=10000, total_duration=40485.7, train/accuracy=0.0008569834171794355, train/loss=8.899463653564453, validation/accuracy=0.001339999958872795, validation/loss=8.870979309082031, validation/num_examples=50000 -I0513 09:45:36.991364 140330640848640 logging_writer.py:48] [107400] global_step=107400, grad_norm=4.839107036590576, loss=3.8362488746643066 -I0513 09:46:13.723860 140330632455936 logging_writer.py:48] [107500] global_step=107500, grad_norm=9.340968132019043, loss=4.177791595458984 -I0513 09:46:50.888641 140330640848640 logging_writer.py:48] [107600] global_step=107600, grad_norm=3.2588741779327393, loss=3.652313709259033 -I0513 09:47:28.177669 140330632455936 logging_writer.py:48] [107700] global_step=107700, grad_norm=10.93560791015625, loss=3.834002733230591 -I0513 09:48:04.992438 140330640848640 logging_writer.py:48] [107800] global_step=107800, grad_norm=4.255331516265869, loss=3.81552791595459 -I0513 09:48:41.768803 140330632455936 logging_writer.py:48] [107900] global_step=107900, grad_norm=3.230623722076416, loss=3.6731112003326416 -I0513 09:49:19.344572 140330640848640 logging_writer.py:48] [108000] global_step=108000, grad_norm=2.6548826694488525, loss=3.8348121643066406 -I0513 09:49:56.241385 140330632455936 logging_writer.py:48] [108100] global_step=108100, grad_norm=3.2797813415527344, loss=3.7285099029541016 -I0513 09:50:33.441344 140330640848640 logging_writer.py:48] [108200] global_step=108200, grad_norm=3.3155040740966797, loss=3.8000576496124268 -I0513 09:51:10.671200 140330632455936 logging_writer.py:48] [108300] global_step=108300, grad_norm=5.638856887817383, loss=3.792868137359619 -I0513 09:51:47.427429 140330640848640 logging_writer.py:48] [108400] global_step=108400, grad_norm=3.7285563945770264, loss=3.6927735805511475 -I0513 09:52:24.539072 140330632455936 logging_writer.py:48] [108500] global_step=108500, grad_norm=4.639146327972412, loss=3.6407289505004883 -I0513 09:53:01.821181 140330640848640 logging_writer.py:48] [108600] global_step=108600, grad_norm=4.036068916320801, loss=3.6203818321228027 -I0513 09:53:38.597003 140330632455936 logging_writer.py:48] [108700] global_step=108700, grad_norm=5.32140588760376, loss=3.7284646034240723 -I0513 09:54:15.402677 140330640848640 logging_writer.py:48] [108800] global_step=108800, grad_norm=3.8763961791992188, loss=3.7474918365478516 -I0513 09:54:52.887694 140330632455936 logging_writer.py:48] [108900] global_step=108900, grad_norm=4.126560211181641, loss=3.759810447692871 -I0513 09:55:29.818485 140330640848640 logging_writer.py:48] [109000] global_step=109000, grad_norm=2.948451042175293, loss=3.7390713691711426 -I0513 09:56:07.010682 140330632455936 logging_writer.py:48] [109100] global_step=109100, grad_norm=7.121793746948242, loss=3.83284854888916 -I0513 09:56:44.283341 140330640848640 logging_writer.py:48] [109200] global_step=109200, grad_norm=14.449772834777832, loss=3.778683662414551 -I0513 09:57:21.103878 140330632455936 logging_writer.py:48] [109300] global_step=109300, grad_norm=6.579891204833984, loss=3.9989449977874756 -I0513 09:57:57.864829 140330640848640 logging_writer.py:48] [109400] global_step=109400, grad_norm=3.9652912616729736, loss=3.8671813011169434 -I0513 09:58:35.368349 140330632455936 logging_writer.py:48] [109500] global_step=109500, grad_norm=7.098851203918457, loss=3.701186180114746 -I0513 09:59:12.202916 140330640848640 logging_writer.py:48] [109600] global_step=109600, grad_norm=4.696351051330566, loss=3.837880849838257 -I0513 09:59:49.144586 140330632455936 logging_writer.py:48] [109700] global_step=109700, grad_norm=4.209230422973633, loss=3.719440221786499 -I0513 10:00:26.903461 140330640848640 logging_writer.py:48] [109800] global_step=109800, grad_norm=3.586735963821411, loss=3.680482864379883 -I0513 10:01:03.802339 140330632455936 logging_writer.py:48] [109900] global_step=109900, grad_norm=4.035359859466553, loss=3.656175374984741 -I0513 10:01:40.623566 140330640848640 logging_writer.py:48] [110000] global_step=110000, grad_norm=3.2376575469970703, loss=3.6879425048828125 -I0513 10:02:18.278385 140330632455936 logging_writer.py:48] [110100] global_step=110100, grad_norm=4.124032497406006, loss=3.8797430992126465 -I0513 10:02:55.039005 140330640848640 logging_writer.py:48] [110200] global_step=110200, grad_norm=6.419761657714844, loss=3.7658538818359375 -I0513 10:03:31.844223 140330632455936 logging_writer.py:48] [110300] global_step=110300, grad_norm=3.57970929145813, loss=3.5959458351135254 -I0513 10:04:09.408477 140330640848640 logging_writer.py:48] [110400] global_step=110400, grad_norm=2.766726016998291, loss=3.663450002670288 -I0513 10:04:46.341873 140330632455936 logging_writer.py:48] [110500] global_step=110500, grad_norm=3.310948133468628, loss=3.8067126274108887 -I0513 10:05:23.125495 140330640848640 logging_writer.py:48] [110600] global_step=110600, grad_norm=4.097723960876465, loss=3.761744976043701 -I0513 10:06:00.832879 140330632455936 logging_writer.py:48] [110700] global_step=110700, grad_norm=3.327580213546753, loss=3.7649402618408203 -I0513 10:06:37.620700 140330640848640 logging_writer.py:48] [110800] global_step=110800, grad_norm=4.843514442443848, loss=3.7639338970184326 -I0513 10:07:14.490847 140330632455936 logging_writer.py:48] [110900] global_step=110900, grad_norm=5.922109603881836, loss=3.923407554626465 -I0513 10:07:52.072910 140330640848640 logging_writer.py:48] [111000] global_step=111000, grad_norm=4.884547710418701, loss=3.7729790210723877 -I0513 10:08:28.978979 140330632455936 logging_writer.py:48] [111100] global_step=111100, grad_norm=4.47864294052124, loss=3.74979305267334 -I0513 10:09:05.809935 140330640848640 logging_writer.py:48] [111200] global_step=111200, grad_norm=4.743607044219971, loss=3.6786394119262695 -I0513 10:09:43.436581 140330632455936 logging_writer.py:48] [111300] global_step=111300, grad_norm=4.3449177742004395, loss=3.840507984161377 -I0513 10:10:20.233322 140330640848640 logging_writer.py:48] [111400] global_step=111400, grad_norm=6.915902614593506, loss=3.726304531097412 -I0513 10:10:57.088024 140330632455936 logging_writer.py:48] [111500] global_step=111500, grad_norm=4.796620845794678, loss=3.772803783416748 -I0513 10:11:34.449476 140330640848640 logging_writer.py:48] [111600] global_step=111600, grad_norm=3.722310781478882, loss=3.675476551055908 -I0513 10:12:11.526386 140330632455936 logging_writer.py:48] [111700] global_step=111700, grad_norm=4.905661106109619, loss=3.6850318908691406 -I0513 10:12:48.329455 140330640848640 logging_writer.py:48] [111800] global_step=111800, grad_norm=5.364709854125977, loss=3.905210018157959 -I0513 10:13:25.940517 140330632455936 logging_writer.py:48] [111900] global_step=111900, grad_norm=3.858893632888794, loss=3.7621896266937256 -I0513 10:14:02.734628 140330640848640 logging_writer.py:48] [112000] global_step=112000, grad_norm=3.233548164367676, loss=3.683565616607666 -I0513 10:14:39.520479 140330632455936 logging_writer.py:48] [112100] global_step=112100, grad_norm=6.745646953582764, loss=3.9613497257232666 -I0513 10:15:17.143850 140330640848640 logging_writer.py:48] [112200] global_step=112200, grad_norm=4.042329788208008, loss=3.736609935760498 -I0513 10:15:53.919418 140330632455936 logging_writer.py:48] [112300] global_step=112300, grad_norm=4.882387638092041, loss=3.6574184894561768 -I0513 10:16:30.789811 140330640848640 logging_writer.py:48] [112400] global_step=112400, grad_norm=3.0430808067321777, loss=3.8217811584472656 -I0513 10:17:08.489862 140330632455936 logging_writer.py:48] [112500] global_step=112500, grad_norm=3.4230740070343018, loss=3.7725226879119873 -I0513 10:17:45.312949 140330640848640 logging_writer.py:48] [112600] global_step=112600, grad_norm=3.0485925674438477, loss=3.7058446407318115 -I0513 10:18:22.115490 140330632455936 logging_writer.py:48] [112700] global_step=112700, grad_norm=2.699436664581299, loss=3.8776583671569824 -I0513 10:18:38.560521 140546196993216 spec.py:333] Evaluating on the training split. -I0513 10:18:48.988152 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 10:18:58.832479 140546196993216 spec.py:363] Evaluating on the test split. -I0513 10:18:59.710526 140546196993216 submission_runner.py:516] Time since start: 42503.76s, Step: 112744, {'train/accuracy': Array(0.00121572, dtype=float32), 'train/loss': Array(9.044412, dtype=float32), 'validation/accuracy': Array(0.00132, dtype=float32), 'validation/loss': Array(9.01385, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(9.029619, dtype=float32), 'test/num_examples': 10000, 'score': 41976.645936489105, 'total_duration': 42503.76131486893, 'accumulated_submission_time': 41976.645936489105, 'accumulated_eval_time': 524.8504509925842, 'accumulated_logging_time': 1.324159860610962} -I0513 10:18:59.754186 140330640848640 logging_writer.py:48] [112744] accumulated_eval_time=524.85, accumulated_logging_time=1.32416, accumulated_submission_time=41976.6, global_step=112744, preemption_count=0, score=41976.6, test/accuracy=0.0013000001199543476, test/loss=9.029619216918945, test/num_examples=10000, total_duration=42503.8, train/accuracy=0.001215720665641129, train/loss=9.044411659240723, validation/accuracy=0.0013199999229982495, validation/loss=9.013850212097168, validation/num_examples=50000 -I0513 10:19:21.719910 140330632455936 logging_writer.py:48] [112800] global_step=112800, grad_norm=3.1572721004486084, loss=3.7129812240600586 -I0513 10:19:58.484150 140330640848640 logging_writer.py:48] [112900] global_step=112900, grad_norm=3.7611427307128906, loss=3.6595230102539062 -I0513 10:20:35.386028 140330632455936 logging_writer.py:48] [113000] global_step=113000, grad_norm=6.650253772735596, loss=3.831244707107544 -I0513 10:21:12.921564 140330640848640 logging_writer.py:48] [113100] global_step=113100, grad_norm=3.552168607711792, loss=3.8325486183166504 -I0513 10:21:49.773540 140330632455936 logging_writer.py:48] [113200] global_step=113200, grad_norm=3.8724706172943115, loss=3.845125198364258 -I0513 10:22:26.603818 140330640848640 logging_writer.py:48] [113300] global_step=113300, grad_norm=4.374044895172119, loss=3.7740821838378906 -I0513 10:23:04.234903 140330632455936 logging_writer.py:48] [113400] global_step=113400, grad_norm=8.170890808105469, loss=3.6765048503875732 -I0513 10:23:40.997685 140330640848640 logging_writer.py:48] [113500] global_step=113500, grad_norm=3.776390790939331, loss=3.6576132774353027 -I0513 10:24:17.852753 140330632455936 logging_writer.py:48] [113600] global_step=113600, grad_norm=3.042482852935791, loss=3.8300459384918213 -I0513 10:24:55.514088 140330640848640 logging_writer.py:48] [113700] global_step=113700, grad_norm=3.753460168838501, loss=3.6690378189086914 -I0513 10:25:32.313829 140330632455936 logging_writer.py:48] [113800] global_step=113800, grad_norm=11.987959861755371, loss=3.762814998626709 -I0513 10:26:09.127612 140330640848640 logging_writer.py:48] [113900] global_step=113900, grad_norm=2.735116958618164, loss=3.7645063400268555 -I0513 10:26:46.821554 140330632455936 logging_writer.py:48] [114000] global_step=114000, grad_norm=5.075464248657227, loss=3.872195243835449 -I0513 10:27:23.737650 140330640848640 logging_writer.py:48] [114100] global_step=114100, grad_norm=3.009256601333618, loss=3.7166271209716797 -I0513 10:28:00.582046 140330632455936 logging_writer.py:48] [114200] global_step=114200, grad_norm=10.224396705627441, loss=3.824002265930176 -I0513 10:28:37.822889 140330640848640 logging_writer.py:48] [114300] global_step=114300, grad_norm=4.792410373687744, loss=3.755627155303955 -I0513 10:29:14.831362 140330632455936 logging_writer.py:48] [114400] global_step=114400, grad_norm=2.961219072341919, loss=3.8031558990478516 -I0513 10:29:51.669186 140330640848640 logging_writer.py:48] [114500] global_step=114500, grad_norm=6.412239074707031, loss=3.9888124465942383 -I0513 10:30:29.297888 140330632455936 logging_writer.py:48] [114600] global_step=114600, grad_norm=3.6869993209838867, loss=3.7691545486450195 -I0513 10:31:06.101426 140330640848640 logging_writer.py:48] [114700] global_step=114700, grad_norm=3.788454532623291, loss=3.8960134983062744 -I0513 10:31:42.892914 140330632455936 logging_writer.py:48] [114800] global_step=114800, grad_norm=23.122779846191406, loss=3.997596263885498 -I0513 10:32:20.536829 140330640848640 logging_writer.py:48] [114900] global_step=114900, grad_norm=2.61480975151062, loss=3.7914047241210938 -I0513 10:32:57.306405 140330632455936 logging_writer.py:48] [115000] global_step=115000, grad_norm=6.205552577972412, loss=3.887047290802002 -I0513 10:33:34.076665 140330640848640 logging_writer.py:48] [115100] global_step=115100, grad_norm=21.316307067871094, loss=3.8755717277526855 -I0513 10:34:11.773184 140330632455936 logging_writer.py:48] [115200] global_step=115200, grad_norm=8.036630630493164, loss=4.068155765533447 -I0513 10:34:48.600145 140330640848640 logging_writer.py:48] [115300] global_step=115300, grad_norm=13.400392532348633, loss=3.944715976715088 -I0513 10:35:25.371913 140330632455936 logging_writer.py:48] [115400] global_step=115400, grad_norm=3.169494390487671, loss=3.9061529636383057 -I0513 10:36:03.076839 140330640848640 logging_writer.py:48] [115500] global_step=115500, grad_norm=3.6735963821411133, loss=3.9165167808532715 -I0513 10:36:39.902977 140330632455936 logging_writer.py:48] [115600] global_step=115600, grad_norm=3.603346347808838, loss=3.8713717460632324 -I0513 10:37:16.792084 140330640848640 logging_writer.py:48] [115700] global_step=115700, grad_norm=5.875718116760254, loss=4.0980730056762695 -I0513 10:37:54.363796 140330632455936 logging_writer.py:48] [115800] global_step=115800, grad_norm=6.849591255187988, loss=3.860513925552368 -I0513 10:38:31.126850 140330640848640 logging_writer.py:48] [115900] global_step=115900, grad_norm=3.280613899230957, loss=3.734471321105957 -I0513 10:39:07.955845 140330632455936 logging_writer.py:48] [116000] global_step=116000, grad_norm=5.065332889556885, loss=3.841580867767334 -I0513 10:39:45.561964 140330640848640 logging_writer.py:48] [116100] global_step=116100, grad_norm=6.075076103210449, loss=3.894672393798828 -I0513 10:40:22.284171 140330632455936 logging_writer.py:48] [116200] global_step=116200, grad_norm=5.439451694488525, loss=3.894758701324463 -I0513 10:40:59.063956 140330640848640 logging_writer.py:48] [116300] global_step=116300, grad_norm=2.5137221813201904, loss=3.7099814414978027 -I0513 10:41:36.402631 140330632455936 logging_writer.py:48] [116400] global_step=116400, grad_norm=8.748764991760254, loss=3.8077516555786133 -I0513 10:42:13.471211 140330640848640 logging_writer.py:48] [116500] global_step=116500, grad_norm=4.77944803237915, loss=3.891127347946167 -I0513 10:42:50.316388 140330632455936 logging_writer.py:48] [116600] global_step=116600, grad_norm=6.297405242919922, loss=3.9183783531188965 -I0513 10:43:27.933332 140330640848640 logging_writer.py:48] [116700] global_step=116700, grad_norm=22.318897247314453, loss=3.889739513397217 -I0513 10:44:04.710437 140330632455936 logging_writer.py:48] [116800] global_step=116800, grad_norm=4.018735408782959, loss=3.729644298553467 -I0513 10:44:41.482468 140330640848640 logging_writer.py:48] [116900] global_step=116900, grad_norm=3.1749613285064697, loss=3.7554755210876465 -I0513 10:45:19.097835 140330632455936 logging_writer.py:48] [117000] global_step=117000, grad_norm=8.57232666015625, loss=3.6103603839874268 -I0513 10:45:55.987238 140330640848640 logging_writer.py:48] [117100] global_step=117100, grad_norm=6.926043510437012, loss=3.8333919048309326 -I0513 10:46:32.806890 140330632455936 logging_writer.py:48] [117200] global_step=117200, grad_norm=3.331742525100708, loss=3.7946834564208984 -I0513 10:47:10.575207 140330640848640 logging_writer.py:48] [117300] global_step=117300, grad_norm=3.7748963832855225, loss=3.8654439449310303 -I0513 10:47:47.485069 140330632455936 logging_writer.py:48] [117400] global_step=117400, grad_norm=3.702160120010376, loss=3.773488998413086 -I0513 10:48:24.331663 140330640848640 logging_writer.py:48] [117500] global_step=117500, grad_norm=3.601355791091919, loss=3.7998149394989014 -I0513 10:49:01.956835 140330632455936 logging_writer.py:48] [117600] global_step=117600, grad_norm=6.136844158172607, loss=3.8516488075256348 -I0513 10:49:38.716644 140330640848640 logging_writer.py:48] [117700] global_step=117700, grad_norm=4.409923076629639, loss=3.7458817958831787 -I0513 10:50:15.492601 140330632455936 logging_writer.py:48] [117800] global_step=117800, grad_norm=3.477771520614624, loss=3.833772659301758 -I0513 10:50:52.999397 140330640848640 logging_writer.py:48] [117900] global_step=117900, grad_norm=6.718493938446045, loss=3.694401979446411 -I0513 10:51:29.823935 140330632455936 logging_writer.py:48] [118000] global_step=118000, grad_norm=3.283419370651245, loss=3.8344197273254395 -I0513 10:52:06.650034 140330640848640 logging_writer.py:48] [118100] global_step=118100, grad_norm=4.201540946960449, loss=3.7724103927612305 -I0513 10:52:16.141688 140546196993216 spec.py:333] Evaluating on the training split. -I0513 10:52:26.799729 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 10:52:37.459921 140546196993216 spec.py:363] Evaluating on the test split. -I0513 10:52:38.339806 140546196993216 submission_runner.py:516] Time since start: 44522.39s, Step: 118126, {'train/accuracy': Array(0.00131537, dtype=float32), 'train/loss': Array(9.158456, dtype=float32), 'validation/accuracy': Array(0.0013, dtype=float32), 'validation/loss': Array(9.12994, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(9.145188, dtype=float32), 'test/num_examples': 10000, 'score': 43972.96816420555, 'total_duration': 44522.39081430435, 'accumulated_submission_time': 43972.96816420555, 'accumulated_eval_time': 547.0459926128387, 'accumulated_logging_time': 1.3902931213378906} -I0513 10:52:38.379231 140330632455936 logging_writer.py:48] [118126] accumulated_eval_time=547.046, accumulated_logging_time=1.39029, accumulated_submission_time=43973, global_step=118126, preemption_count=0, score=43973, test/accuracy=0.0012000000569969416, test/loss=9.145188331604004, test/num_examples=10000, total_duration=44522.4, train/accuracy=0.0013153698528185487, train/loss=9.158455848693848, validation/accuracy=0.0013000000035390258, validation/loss=9.129940032958984, validation/num_examples=50000 -I0513 10:53:07.146807 140330640848640 logging_writer.py:48] [118200] global_step=118200, grad_norm=3.522374153137207, loss=3.8941869735717773 -I0513 10:53:43.997274 140330632455936 logging_writer.py:48] [118300] global_step=118300, grad_norm=3.4716122150421143, loss=3.675401210784912 -I0513 10:54:20.801444 140330640848640 logging_writer.py:48] [118400] global_step=118400, grad_norm=4.1320319175720215, loss=3.9381072521209717 -I0513 10:54:58.360369 140330632455936 logging_writer.py:48] [118500] global_step=118500, grad_norm=5.380265235900879, loss=3.8550586700439453 -I0513 10:55:35.169420 140330640848640 logging_writer.py:48] [118600] global_step=118600, grad_norm=5.332126617431641, loss=3.739299774169922 -I0513 10:56:11.982759 140330632455936 logging_writer.py:48] [118700] global_step=118700, grad_norm=4.553676605224609, loss=3.8869376182556152 -I0513 10:56:49.624538 140330640848640 logging_writer.py:48] [118800] global_step=118800, grad_norm=5.244413375854492, loss=3.7666993141174316 -I0513 10:57:27.556351 140330632455936 logging_writer.py:48] [118900] global_step=118900, grad_norm=3.67285418510437, loss=3.8157129287719727 -I0513 10:58:04.537482 140330640848640 logging_writer.py:48] [119000] global_step=119000, grad_norm=4.8743720054626465, loss=3.8619141578674316 -I0513 10:58:41.803629 140330632455936 logging_writer.py:48] [119100] global_step=119100, grad_norm=11.357101440429688, loss=3.748810291290283 -I0513 10:59:18.986193 140330640848640 logging_writer.py:48] [119200] global_step=119200, grad_norm=20.027647018432617, loss=3.907317876815796 -I0513 10:59:55.802936 140330632455936 logging_writer.py:48] [119300] global_step=119300, grad_norm=6.830605506896973, loss=3.9416399002075195 -I0513 11:00:33.502220 140330640848640 logging_writer.py:48] [119400] global_step=119400, grad_norm=3.236597776412964, loss=3.795792818069458 -I0513 11:01:10.235175 140330632455936 logging_writer.py:48] [119500] global_step=119500, grad_norm=8.095415115356445, loss=3.793891668319702 -I0513 11:01:47.078099 140330640848640 logging_writer.py:48] [119600] global_step=119600, grad_norm=5.320235729217529, loss=3.8365440368652344 -I0513 11:02:24.302553 140330632455936 logging_writer.py:48] [119700] global_step=119700, grad_norm=34.19307327270508, loss=4.090809345245361 -I0513 11:03:01.549707 140330640848640 logging_writer.py:48] [119800] global_step=119800, grad_norm=8.464475631713867, loss=4.021947860717773 -I0513 11:03:38.344255 140330632455936 logging_writer.py:48] [119900] global_step=119900, grad_norm=4.135895252227783, loss=3.9543302059173584 -I0513 11:04:15.623381 140330640848640 logging_writer.py:48] [120000] global_step=120000, grad_norm=6.860583305358887, loss=3.89605450630188 -I0513 11:04:52.736787 140330632455936 logging_writer.py:48] [120100] global_step=120100, grad_norm=5.27660608291626, loss=3.8676669597625732 -I0513 11:05:29.521014 140330640848640 logging_writer.py:48] [120200] global_step=120200, grad_norm=7.493173122406006, loss=3.7432568073272705 -I0513 11:06:06.791218 140330632455936 logging_writer.py:48] [120300] global_step=120300, grad_norm=3.5400197505950928, loss=3.8375542163848877 -I0513 11:06:43.974797 140330640848640 logging_writer.py:48] [120400] global_step=120400, grad_norm=3.9103856086730957, loss=3.867438316345215 -I0513 11:07:20.943874 140330632455936 logging_writer.py:48] [120500] global_step=120500, grad_norm=3.791984796524048, loss=3.7902512550354004 -I0513 11:07:58.229241 140330640848640 logging_writer.py:48] [120600] global_step=120600, grad_norm=5.063652992248535, loss=3.6972975730895996 -I0513 11:08:35.309916 140330632455936 logging_writer.py:48] [120700] global_step=120700, grad_norm=11.456694602966309, loss=3.879082441329956 -I0513 11:09:12.101370 140330640848640 logging_writer.py:48] [120800] global_step=120800, grad_norm=4.384346008300781, loss=3.902266263961792 -I0513 11:09:49.384792 140330632455936 logging_writer.py:48] [120900] global_step=120900, grad_norm=3.039144515991211, loss=3.6180803775787354 -I0513 11:10:26.539558 140330640848640 logging_writer.py:48] [121000] global_step=121000, grad_norm=6.0294880867004395, loss=3.8754377365112305 -I0513 11:11:03.338581 140330632455936 logging_writer.py:48] [121100] global_step=121100, grad_norm=4.942533016204834, loss=3.799370050430298 -I0513 11:11:40.638107 140330640848640 logging_writer.py:48] [121200] global_step=121200, grad_norm=19.82294464111328, loss=3.930974245071411 -I0513 11:12:17.814066 140330632455936 logging_writer.py:48] [121300] global_step=121300, grad_norm=3.5402655601501465, loss=3.815023899078369 -I0513 11:12:54.605027 140330640848640 logging_writer.py:48] [121400] global_step=121400, grad_norm=4.359009742736816, loss=3.8298397064208984 -I0513 11:13:32.033444 140330632455936 logging_writer.py:48] [121500] global_step=121500, grad_norm=5.436450481414795, loss=3.904906749725342 -I0513 11:14:09.129997 140330640848640 logging_writer.py:48] [121600] global_step=121600, grad_norm=5.000278472900391, loss=3.8602163791656494 -I0513 11:14:45.976410 140330632455936 logging_writer.py:48] [121700] global_step=121700, grad_norm=5.150325775146484, loss=3.838139057159424 -I0513 11:15:23.261044 140330640848640 logging_writer.py:48] [121800] global_step=121800, grad_norm=4.970336437225342, loss=3.723294734954834 -I0513 11:16:00.350655 140330632455936 logging_writer.py:48] [121900] global_step=121900, grad_norm=3.7039198875427246, loss=3.6724140644073486 -I0513 11:16:37.100829 140330640848640 logging_writer.py:48] [122000] global_step=122000, grad_norm=7.048591136932373, loss=3.9133963584899902 -I0513 11:17:14.417391 140330632455936 logging_writer.py:48] [122100] global_step=122100, grad_norm=4.06885290145874, loss=3.9169888496398926 -I0513 11:17:51.661990 140330640848640 logging_writer.py:48] [122200] global_step=122200, grad_norm=3.9676520824432373, loss=3.7793960571289062 -I0513 11:18:28.583798 140330632455936 logging_writer.py:48] [122300] global_step=122300, grad_norm=10.353906631469727, loss=3.877218723297119 -I0513 11:19:06.293838 140330640848640 logging_writer.py:48] [122400] global_step=122400, grad_norm=3.6035284996032715, loss=3.727926731109619 -I0513 11:19:42.992716 140330632455936 logging_writer.py:48] [122500] global_step=122500, grad_norm=3.9310503005981445, loss=3.9201197624206543 -I0513 11:20:19.825220 140330640848640 logging_writer.py:48] [122600] global_step=122600, grad_norm=8.407293319702148, loss=3.9246768951416016 -I0513 11:20:57.050248 140330632455936 logging_writer.py:48] [122700] global_step=122700, grad_norm=13.110547065734863, loss=3.9743263721466064 -I0513 11:21:34.088619 140330640848640 logging_writer.py:48] [122800] global_step=122800, grad_norm=4.8760576248168945, loss=3.835310220718384 -I0513 11:22:10.897205 140330632455936 logging_writer.py:48] [122900] global_step=122900, grad_norm=3.495871067047119, loss=3.8266701698303223 -I0513 11:22:48.513327 140330640848640 logging_writer.py:48] [123000] global_step=123000, grad_norm=7.492894649505615, loss=3.9273524284362793 -I0513 11:23:25.309556 140330632455936 logging_writer.py:48] [123100] global_step=123100, grad_norm=4.180524826049805, loss=3.8668951988220215 -I0513 11:24:02.150306 140330640848640 logging_writer.py:48] [123200] global_step=123200, grad_norm=3.4188320636749268, loss=3.7293169498443604 -I0513 11:24:39.410036 140330632455936 logging_writer.py:48] [123300] global_step=123300, grad_norm=5.213826656341553, loss=3.889155864715576 -I0513 11:25:16.491666 140330640848640 logging_writer.py:48] [123400] global_step=123400, grad_norm=3.447411298751831, loss=3.6259474754333496 -I0513 11:25:53.266087 140330632455936 logging_writer.py:48] [123500] global_step=123500, grad_norm=5.185448169708252, loss=3.8281807899475098 -I0513 11:25:54.992833 140546196993216 spec.py:333] Evaluating on the training split. -I0513 11:26:05.453265 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 11:26:17.325201 140546196993216 spec.py:363] Evaluating on the test split. -I0513 11:26:18.222793 140546196993216 submission_runner.py:516] Time since start: 46542.27s, Step: 123505, {'train/accuracy': Array(0.00127551, dtype=float32), 'train/loss': Array(9.268085, dtype=float32), 'validation/accuracy': Array(0.00128, dtype=float32), 'validation/loss': Array(9.239881, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(9.25315, dtype=float32), 'test/num_examples': 10000, 'score': 45969.52896118164, 'total_duration': 46542.27374839783, 'accumulated_submission_time': 45969.52896118164, 'accumulated_eval_time': 570.2733144760132, 'accumulated_logging_time': 1.4398610591888428} -I0513 11:26:18.267545 140330640848640 logging_writer.py:48] [123505] accumulated_eval_time=570.273, accumulated_logging_time=1.43986, accumulated_submission_time=45969.5, global_step=123505, preemption_count=0, score=45969.5, test/accuracy=0.0012000000569969416, test/loss=9.25314998626709, test/num_examples=10000, total_duration=46542.3, train/accuracy=0.0012755101779475808, train/loss=9.268084526062012, validation/accuracy=0.0012799999676644802, validation/loss=9.239880561828613, validation/num_examples=50000 -I0513 11:26:54.339227 140330632455936 logging_writer.py:48] [123600] global_step=123600, grad_norm=3.7744901180267334, loss=3.7466297149658203 -I0513 11:27:31.543891 140330640848640 logging_writer.py:48] [123700] global_step=123700, grad_norm=4.574336051940918, loss=3.7731404304504395 -I0513 11:28:08.317600 140330632455936 logging_writer.py:48] [123800] global_step=123800, grad_norm=3.63686466217041, loss=3.8211827278137207 -I0513 11:28:45.672948 140330640848640 logging_writer.py:48] [123900] global_step=123900, grad_norm=4.785826206207275, loss=3.718329429626465 -I0513 11:29:24.180264 140330632455936 logging_writer.py:48] [124000] global_step=124000, grad_norm=5.114258289337158, loss=3.777095317840576 -I0513 11:30:02.960008 140330640848640 logging_writer.py:48] [124100] global_step=124100, grad_norm=3.923779249191284, loss=3.844639778137207 -I0513 11:30:40.238026 140330632455936 logging_writer.py:48] [124200] global_step=124200, grad_norm=9.965909004211426, loss=3.7087082862854004 -I0513 11:31:17.420613 140330640848640 logging_writer.py:48] [124300] global_step=124300, grad_norm=11.48297119140625, loss=3.7525687217712402 -I0513 11:31:57.269210 140330632455936 logging_writer.py:48] [124400] global_step=124400, grad_norm=4.877635478973389, loss=3.781496047973633 -I0513 11:32:44.958256 140330640848640 logging_writer.py:48] [124500] global_step=124500, grad_norm=3.410552740097046, loss=3.6849234104156494 -I0513 11:33:22.121691 140330632455936 logging_writer.py:48] [124600] global_step=124600, grad_norm=5.58048152923584, loss=3.8295559883117676 -I0513 11:33:59.064776 140330640848640 logging_writer.py:48] [124700] global_step=124700, grad_norm=11.496071815490723, loss=3.919949769973755 -I0513 11:34:36.440830 140330632455936 logging_writer.py:48] [124800] global_step=124800, grad_norm=2.904376745223999, loss=3.8523471355438232 -I0513 11:35:13.663345 140330640848640 logging_writer.py:48] [124900] global_step=124900, grad_norm=5.701645374298096, loss=3.9726290702819824 -I0513 11:35:50.598834 140330632455936 logging_writer.py:48] [125000] global_step=125000, grad_norm=4.52528190612793, loss=3.873971939086914 -I0513 11:36:27.821570 140330640848640 logging_writer.py:48] [125100] global_step=125100, grad_norm=4.994039535522461, loss=3.8486783504486084 -I0513 11:37:05.006266 140330632455936 logging_writer.py:48] [125200] global_step=125200, grad_norm=6.059115409851074, loss=3.715644598007202 -I0513 11:37:44.851013 140330640848640 logging_writer.py:48] [125300] global_step=125300, grad_norm=3.8805112838745117, loss=3.8669333457946777 -I0513 11:38:22.414009 140330632455936 logging_writer.py:48] [125400] global_step=125400, grad_norm=5.980544090270996, loss=3.883272886276245 -I0513 11:39:03.741675 140330640848640 logging_writer.py:48] [125500] global_step=125500, grad_norm=33.823524475097656, loss=4.041839599609375 -I0513 11:39:40.743987 140330632455936 logging_writer.py:48] [125600] global_step=125600, grad_norm=5.169351577758789, loss=3.7640058994293213 -I0513 11:40:23.872363 140330640848640 logging_writer.py:48] [125700] global_step=125700, grad_norm=4.455743312835693, loss=3.855510711669922 -I0513 11:41:06.442835 140330632455936 logging_writer.py:48] [125800] global_step=125800, grad_norm=2.555471897125244, loss=3.8115363121032715 -I0513 11:41:43.242084 140330640848640 logging_writer.py:48] [125900] global_step=125900, grad_norm=3.4355766773223877, loss=3.996633529663086 -I0513 11:42:20.768317 140330632455936 logging_writer.py:48] [126000] global_step=126000, grad_norm=2.897357225418091, loss=3.7518439292907715 -I0513 11:42:57.565342 140330640848640 logging_writer.py:48] [126100] global_step=126100, grad_norm=4.03682804107666, loss=3.7267470359802246 -I0513 11:43:34.828198 140330632455936 logging_writer.py:48] [126200] global_step=126200, grad_norm=4.334954261779785, loss=3.7973718643188477 -I0513 11:44:18.696019 140330640848640 logging_writer.py:48] [126300] global_step=126300, grad_norm=3.0264647006988525, loss=3.9016478061676025 -I0513 11:44:55.833591 140330632455936 logging_writer.py:48] [126400] global_step=126400, grad_norm=3.5046937465667725, loss=3.843338966369629 -I0513 11:45:32.694546 140330640848640 logging_writer.py:48] [126500] global_step=126500, grad_norm=6.399008750915527, loss=4.018481254577637 -I0513 11:46:10.022829 140330632455936 logging_writer.py:48] [126600] global_step=126600, grad_norm=24.8284912109375, loss=3.960603952407837 -I0513 11:46:51.265254 140330640848640 logging_writer.py:48] [126700] global_step=126700, grad_norm=6.760029315948486, loss=3.922492504119873 -I0513 11:47:35.885837 140330632455936 logging_writer.py:48] [126800] global_step=126800, grad_norm=5.609131336212158, loss=3.8782248497009277 -I0513 11:48:19.600806 140330640848640 logging_writer.py:48] [126900] global_step=126900, grad_norm=4.110615253448486, loss=3.86014461517334 -I0513 11:48:56.732511 140330632455936 logging_writer.py:48] [127000] global_step=127000, grad_norm=4.831293106079102, loss=3.8351736068725586 -I0513 11:49:33.557694 140330640848640 logging_writer.py:48] [127100] global_step=127100, grad_norm=5.194890022277832, loss=3.894216299057007 -I0513 11:50:10.928852 140330632455936 logging_writer.py:48] [127200] global_step=127200, grad_norm=5.850169658660889, loss=3.7962632179260254 -I0513 11:50:47.998753 140330640848640 logging_writer.py:48] [127300] global_step=127300, grad_norm=5.937769889831543, loss=3.9155008792877197 -I0513 11:51:24.829274 140330632455936 logging_writer.py:48] [127400] global_step=127400, grad_norm=4.625711917877197, loss=3.8318629264831543 -I0513 11:52:02.617600 140330640848640 logging_writer.py:48] [127500] global_step=127500, grad_norm=3.6980538368225098, loss=3.9674839973449707 -I0513 11:52:39.758904 140330632455936 logging_writer.py:48] [127600] global_step=127600, grad_norm=3.826399564743042, loss=3.9507975578308105 -I0513 11:53:16.668846 140330640848640 logging_writer.py:48] [127700] global_step=127700, grad_norm=4.245905876159668, loss=3.8231201171875 -I0513 11:53:55.348686 140330632455936 logging_writer.py:48] [127800] global_step=127800, grad_norm=4.689609527587891, loss=3.8320798873901367 -I0513 11:54:36.632074 140330640848640 logging_writer.py:48] [127900] global_step=127900, grad_norm=2.725513458251953, loss=3.7265548706054688 -I0513 11:55:15.756856 140330632455936 logging_writer.py:48] [128000] global_step=128000, grad_norm=3.4792730808258057, loss=3.9693121910095215 -I0513 11:56:08.381075 140330640848640 logging_writer.py:48] [128100] global_step=128100, grad_norm=3.9251673221588135, loss=3.876957893371582 -I0513 11:56:46.698498 140330632455936 logging_writer.py:48] [128200] global_step=128200, grad_norm=5.95255708694458, loss=3.7428219318389893 -I0513 11:57:24.015993 140330640848640 logging_writer.py:48] [128300] global_step=128300, grad_norm=4.997304916381836, loss=3.7520558834075928 -I0513 11:58:01.427269 140330632455936 logging_writer.py:48] [128400] global_step=128400, grad_norm=15.75358772277832, loss=3.946855306625366 -I0513 11:58:38.543294 140330640848640 logging_writer.py:48] [128500] global_step=128500, grad_norm=4.140839099884033, loss=3.8287105560302734 -I0513 11:59:15.344341 140330632455936 logging_writer.py:48] [128600] global_step=128600, grad_norm=5.913165092468262, loss=3.6912713050842285 -I0513 11:59:35.031781 140546196993216 spec.py:333] Evaluating on the training split. -I0513 11:59:45.039284 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 12:00:12.517902 140546196993216 spec.py:363] Evaluating on the test split. -I0513 12:00:13.576918 140546196993216 submission_runner.py:516] Time since start: 48577.46s, Step: 128650, {'train/accuracy': Array(0.00143495, dtype=float32), 'train/loss': Array(9.360594, dtype=float32), 'validation/accuracy': Array(0.0013, dtype=float32), 'validation/loss': Array(9.334594, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0011, dtype=float32), 'test/loss': Array(9.347084, dtype=float32), 'test/num_examples': 10000, 'score': 47966.240132808685, 'total_duration': 48577.46176099777, 'accumulated_submission_time': 47966.240132808685, 'accumulated_eval_time': 608.6497070789337, 'accumulated_logging_time': 1.4954533576965332} -I0513 12:00:13.727325 140330640848640 logging_writer.py:48] [128650] accumulated_eval_time=608.65, accumulated_logging_time=1.49545, accumulated_submission_time=47966.2, global_step=128650, preemption_count=0, score=47966.2, test/accuracy=0.0010999999940395355, test/loss=9.347084045410156, test/num_examples=10000, total_duration=48577.5, train/accuracy=0.001434948993846774, train/loss=9.360593795776367, validation/accuracy=0.0013000000035390258, validation/loss=9.334593772888184, validation/num_examples=50000 -I0513 12:00:34.189648 140330632455936 logging_writer.py:48] [128700] global_step=128700, grad_norm=4.136106491088867, loss=3.7888736724853516 -I0513 12:01:11.413332 140330640848640 logging_writer.py:48] [128800] global_step=128800, grad_norm=4.131703853607178, loss=3.7710371017456055 -I0513 12:01:48.256663 140330632455936 logging_writer.py:48] [128900] global_step=128900, grad_norm=10.327168464660645, loss=3.8868606090545654 -I0513 12:02:25.586171 140330640848640 logging_writer.py:48] [129000] global_step=129000, grad_norm=5.01239538192749, loss=3.8645691871643066 -I0513 12:03:07.205469 140330632455936 logging_writer.py:48] [129100] global_step=129100, grad_norm=6.1088643074035645, loss=3.962691307067871 -I0513 12:03:47.478212 140330640848640 logging_writer.py:48] [129200] global_step=129200, grad_norm=6.001586437225342, loss=3.859652519226074 -I0513 12:04:26.778283 140330632455936 logging_writer.py:48] [129300] global_step=129300, grad_norm=8.474433898925781, loss=3.8456106185913086 -I0513 12:05:04.569812 140330640848640 logging_writer.py:48] [129400] global_step=129400, grad_norm=6.763920783996582, loss=3.890622615814209 -I0513 12:05:42.134544 140330632455936 logging_writer.py:48] [129500] global_step=129500, grad_norm=7.88284158706665, loss=3.8710665702819824 -I0513 12:06:19.872750 140330640848640 logging_writer.py:48] [129600] global_step=129600, grad_norm=4.612990856170654, loss=3.7496066093444824 -I0513 12:06:57.032233 140330632455936 logging_writer.py:48] [129700] global_step=129700, grad_norm=7.094852924346924, loss=3.9525527954101562 -I0513 12:07:33.887070 140330640848640 logging_writer.py:48] [129800] global_step=129800, grad_norm=2.6987664699554443, loss=3.7641921043395996 -I0513 12:08:11.758899 140330632455936 logging_writer.py:48] [129900] global_step=129900, grad_norm=6.644770622253418, loss=3.780395984649658 -I0513 12:08:52.100362 140330640848640 logging_writer.py:48] [130000] global_step=130000, grad_norm=4.265841960906982, loss=3.8424835205078125 -I0513 12:09:32.023282 140330632455936 logging_writer.py:48] [130100] global_step=130100, grad_norm=4.4824113845825195, loss=3.8164000511169434 -I0513 12:10:11.340602 140330640848640 logging_writer.py:48] [130200] global_step=130200, grad_norm=3.8789942264556885, loss=3.931309700012207 -I0513 12:10:54.749695 140330632455936 logging_writer.py:48] [130300] global_step=130300, grad_norm=15.35117244720459, loss=3.8777480125427246 -I0513 12:11:31.974325 140330640848640 logging_writer.py:48] [130400] global_step=130400, grad_norm=5.872207164764404, loss=3.8162217140197754 -I0513 12:12:11.699770 140330632455936 logging_writer.py:48] [130500] global_step=130500, grad_norm=5.085325717926025, loss=3.743034601211548 -I0513 12:12:49.799985 140330640848640 logging_writer.py:48] [130600] global_step=130600, grad_norm=5.332602024078369, loss=3.738469123840332 -I0513 12:13:32.876139 140330632455936 logging_writer.py:48] [130700] global_step=130700, grad_norm=6.817777156829834, loss=3.8348751068115234 -I0513 12:14:17.552556 140330640848640 logging_writer.py:48] [130800] global_step=130800, grad_norm=3.1836419105529785, loss=3.8912534713745117 -I0513 12:14:54.366579 140330632455936 logging_writer.py:48] [130900] global_step=130900, grad_norm=14.383554458618164, loss=3.752012252807617 -I0513 12:15:32.320606 140330640848640 logging_writer.py:48] [131000] global_step=131000, grad_norm=8.499100685119629, loss=4.028433799743652 -I0513 12:16:09.734260 140330632455936 logging_writer.py:48] [131100] global_step=131100, grad_norm=5.4911580085754395, loss=3.8884406089782715 -I0513 12:16:46.895199 140330640848640 logging_writer.py:48] [131200] global_step=131200, grad_norm=7.538158893585205, loss=3.84987735748291 -I0513 12:17:24.234026 140330632455936 logging_writer.py:48] [131300] global_step=131300, grad_norm=4.607253551483154, loss=3.824941873550415 -I0513 12:18:01.884022 140330640848640 logging_writer.py:48] [131400] global_step=131400, grad_norm=7.248748779296875, loss=4.164377212524414 -I0513 12:18:44.000414 140330632455936 logging_writer.py:48] [131500] global_step=131500, grad_norm=3.100877046585083, loss=3.640526294708252 -I0513 12:19:26.197418 140330640848640 logging_writer.py:48] [131600] global_step=131600, grad_norm=3.633017063140869, loss=3.8352198600769043 -I0513 12:20:06.771520 140330632455936 logging_writer.py:48] [131700] global_step=131700, grad_norm=3.6362364292144775, loss=3.897200107574463 -I0513 12:20:45.360540 140330640848640 logging_writer.py:48] [131800] global_step=131800, grad_norm=5.719751358032227, loss=3.790762424468994 -I0513 12:21:22.456847 140330632455936 logging_writer.py:48] [131900] global_step=131900, grad_norm=4.5348992347717285, loss=3.768251895904541 -I0513 12:21:59.821352 140330640848640 logging_writer.py:48] [132000] global_step=132000, grad_norm=3.7257754802703857, loss=3.7793033123016357 -I0513 12:22:43.052333 140330632455936 logging_writer.py:48] [132100] global_step=132100, grad_norm=24.015209197998047, loss=3.9003732204437256 -I0513 12:23:23.974376 140330640848640 logging_writer.py:48] [132200] global_step=132200, grad_norm=8.483253479003906, loss=3.8734607696533203 -I0513 12:24:01.731597 140330632455936 logging_writer.py:48] [132300] global_step=132300, grad_norm=3.9109675884246826, loss=3.8937816619873047 -I0513 12:24:39.661170 140330640848640 logging_writer.py:48] [132400] global_step=132400, grad_norm=8.857048988342285, loss=4.1150031089782715 -I0513 12:25:25.002822 140330632455936 logging_writer.py:48] [132500] global_step=132500, grad_norm=17.306520462036133, loss=3.973799228668213 -I0513 12:26:10.939861 140330640848640 logging_writer.py:48] [132600] global_step=132600, grad_norm=4.500816345214844, loss=4.027688026428223 -I0513 12:26:48.767019 140330632455936 logging_writer.py:48] [132700] global_step=132700, grad_norm=4.966627597808838, loss=3.7501673698425293 -I0513 12:27:28.073875 140330640848640 logging_writer.py:48] [132800] global_step=132800, grad_norm=3.8770272731781006, loss=3.8106906414031982 -I0513 12:28:05.910175 140330632455936 logging_writer.py:48] [132900] global_step=132900, grad_norm=7.126029968261719, loss=3.9820058345794678 -I0513 12:28:43.026096 140330640848640 logging_writer.py:48] [133000] global_step=133000, grad_norm=4.1196794509887695, loss=3.7295024394989014 -I0513 12:29:25.674215 140330632455936 logging_writer.py:48] [133100] global_step=133100, grad_norm=5.309162139892578, loss=3.8222262859344482 -I0513 12:30:03.094771 140330640848640 logging_writer.py:48] [133200] global_step=133200, grad_norm=3.51045560836792, loss=3.794701099395752 -I0513 12:30:41.287210 140330632455936 logging_writer.py:48] [133300] global_step=133300, grad_norm=4.7314958572387695, loss=3.868478775024414 -I0513 12:31:22.964922 140330640848640 logging_writer.py:48] [133400] global_step=133400, grad_norm=3.721165657043457, loss=3.9165468215942383 -I0513 12:32:01.167807 140330632455936 logging_writer.py:48] [133500] global_step=133500, grad_norm=5.4945173263549805, loss=3.8485302925109863 -I0513 12:32:42.730925 140330640848640 logging_writer.py:48] [133600] global_step=133600, grad_norm=4.174462795257568, loss=3.798111915588379 -I0513 12:33:20.184518 140330632455936 logging_writer.py:48] [133700] global_step=133700, grad_norm=7.936197757720947, loss=3.879735231399536 -I0513 12:33:29.653286 140546196993216 spec.py:333] Evaluating on the training split. -I0513 12:33:36.478708 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 12:34:06.881850 140546196993216 spec.py:363] Evaluating on the test split. -I0513 12:34:08.193190 140546196993216 submission_runner.py:516] Time since start: 50611.83s, Step: 133726, {'train/accuracy': Array(0.00121572, dtype=float32), 'train/loss': Array(9.420927, dtype=float32), 'validation/accuracy': Array(0.00136, dtype=float32), 'validation/loss': Array(9.384761, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(9.396325, dtype=float32), 'test/num_examples': 10000, 'score': 49962.06882119179, 'total_duration': 50611.829721450806, 'accumulated_submission_time': 49962.06882119179, 'accumulated_eval_time': 646.7725491523743, 'accumulated_logging_time': 1.700319766998291} -I0513 12:34:08.789715 140330640848640 logging_writer.py:48] [133726] accumulated_eval_time=646.773, accumulated_logging_time=1.70032, accumulated_submission_time=49962.1, global_step=133726, preemption_count=0, score=49962.1, test/accuracy=0.0012000000569969416, test/loss=9.39632511138916, test/num_examples=10000, total_duration=50611.8, train/accuracy=0.001215720665641129, train/loss=9.420927047729492, validation/accuracy=0.0013599999947473407, validation/loss=9.384760856628418, validation/num_examples=50000 -I0513 12:34:37.040524 140330632455936 logging_writer.py:48] [133800] global_step=133800, grad_norm=6.585117340087891, loss=3.989114761352539 -I0513 12:35:16.344574 140330640848640 logging_writer.py:48] [133900] global_step=133900, grad_norm=3.3092920780181885, loss=3.9502782821655273 -I0513 12:35:59.631837 140330632455936 logging_writer.py:48] [134000] global_step=134000, grad_norm=4.126384258270264, loss=3.8912813663482666 -I0513 12:36:41.879941 140330640848640 logging_writer.py:48] [134100] global_step=134100, grad_norm=3.070361614227295, loss=3.93333101272583 -I0513 12:37:23.455664 140330632455936 logging_writer.py:48] [134200] global_step=134200, grad_norm=5.707635402679443, loss=3.8480613231658936 -I0513 12:38:04.190408 140330640848640 logging_writer.py:48] [134300] global_step=134300, grad_norm=4.212728500366211, loss=3.8783822059631348 -I0513 12:38:41.416890 140330632455936 logging_writer.py:48] [134400] global_step=134400, grad_norm=3.9787826538085938, loss=3.983729839324951 -I0513 12:39:18.144142 140330640848640 logging_writer.py:48] [134500] global_step=134500, grad_norm=5.955033302307129, loss=3.8338687419891357 -I0513 12:39:55.248986 140330632455936 logging_writer.py:48] [134600] global_step=134600, grad_norm=8.219697952270508, loss=4.014591693878174 -I0513 12:40:35.586695 140330640848640 logging_writer.py:48] [134700] global_step=134700, grad_norm=10.423572540283203, loss=3.6400797367095947 -I0513 12:41:15.936653 140330632455936 logging_writer.py:48] [134800] global_step=134800, grad_norm=4.22722053527832, loss=3.862760543823242 -I0513 12:41:55.091465 140330640848640 logging_writer.py:48] [134900] global_step=134900, grad_norm=4.825908184051514, loss=3.897738218307495 -I0513 12:42:37.230227 140330632455936 logging_writer.py:48] [135000] global_step=135000, grad_norm=3.4322166442871094, loss=3.9354820251464844 -I0513 12:43:26.713407 140330640848640 logging_writer.py:48] [135100] global_step=135100, grad_norm=10.541150093078613, loss=3.708101749420166 -I0513 12:44:04.213622 140330632455936 logging_writer.py:48] [135200] global_step=135200, grad_norm=3.7590482234954834, loss=3.751068115234375 -I0513 12:44:43.164794 140330640848640 logging_writer.py:48] [135300] global_step=135300, grad_norm=5.547554016113281, loss=3.7563552856445312 -I0513 12:45:21.717346 140330632455936 logging_writer.py:48] [135400] global_step=135400, grad_norm=5.192934989929199, loss=3.931824207305908 -I0513 12:45:58.817228 140330640848640 logging_writer.py:48] [135500] global_step=135500, grad_norm=3.7250821590423584, loss=3.6914708614349365 -I0513 12:46:38.243577 140330632455936 logging_writer.py:48] [135600] global_step=135600, grad_norm=2.45643949508667, loss=3.7507126331329346 -I0513 12:47:22.257232 140330640848640 logging_writer.py:48] [135700] global_step=135700, grad_norm=3.1030454635620117, loss=3.7164573669433594 -I0513 12:48:00.729744 140330632455936 logging_writer.py:48] [135800] global_step=135800, grad_norm=5.200261116027832, loss=3.901614189147949 -I0513 12:48:39.074952 140330640848640 logging_writer.py:48] [135900] global_step=135900, grad_norm=3.749889373779297, loss=3.907069444656372 -I0513 12:49:16.481811 140330632455936 logging_writer.py:48] [136000] global_step=136000, grad_norm=91.96007537841797, loss=3.942603826522827 -I0513 12:49:56.786890 140330640848640 logging_writer.py:48] [136100] global_step=136100, grad_norm=6.819379806518555, loss=3.8811378479003906 -I0513 12:50:38.898564 140330632455936 logging_writer.py:48] [136200] global_step=136200, grad_norm=5.027085781097412, loss=3.6630566120147705 -I0513 12:51:16.052030 140330640848640 logging_writer.py:48] [136300] global_step=136300, grad_norm=6.209240436553955, loss=3.8189926147460938 -I0513 12:51:55.556422 140330632455936 logging_writer.py:48] [136400] global_step=136400, grad_norm=4.024073123931885, loss=3.7784829139709473 -I0513 12:52:36.256445 140330640848640 logging_writer.py:48] [136500] global_step=136500, grad_norm=3.3244071006774902, loss=3.8160572052001953 -I0513 12:53:18.564718 140330632455936 logging_writer.py:48] [136600] global_step=136600, grad_norm=8.178828239440918, loss=4.015040874481201 -I0513 12:54:03.987629 140330640848640 logging_writer.py:48] [136700] global_step=136700, grad_norm=4.094839572906494, loss=3.8216214179992676 -I0513 12:54:48.957778 140330632455936 logging_writer.py:48] [136800] global_step=136800, grad_norm=5.650028705596924, loss=3.766555070877075 -I0513 12:55:30.232160 140330640848640 logging_writer.py:48] [136900] global_step=136900, grad_norm=5.209512710571289, loss=3.8094542026519775 -I0513 12:56:10.004900 140330632455936 logging_writer.py:48] [137000] global_step=137000, grad_norm=3.257946729660034, loss=3.613023281097412 -I0513 12:56:57.237430 140330640848640 logging_writer.py:48] [137100] global_step=137100, grad_norm=5.818464756011963, loss=3.8766913414001465 -I0513 12:57:41.877903 140330632455936 logging_writer.py:48] [137200] global_step=137200, grad_norm=4.040082931518555, loss=3.761336088180542 -I0513 12:58:26.094701 140330640848640 logging_writer.py:48] [137300] global_step=137300, grad_norm=4.883850574493408, loss=3.7759060859680176 -I0513 12:59:05.102057 140330632455936 logging_writer.py:48] [137400] global_step=137400, grad_norm=3.477041244506836, loss=3.761631965637207 -I0513 12:59:44.210086 140330640848640 logging_writer.py:48] [137500] global_step=137500, grad_norm=5.116903781890869, loss=3.780167579650879 -I0513 13:00:28.739907 140330632455936 logging_writer.py:48] [137600] global_step=137600, grad_norm=3.8256075382232666, loss=3.8200767040252686 -I0513 13:01:15.916288 140330640848640 logging_writer.py:48] [137700] global_step=137700, grad_norm=3.937955379486084, loss=3.8451313972473145 -I0513 13:01:55.914520 140330632455936 logging_writer.py:48] [137800] global_step=137800, grad_norm=3.9868881702423096, loss=3.7340757846832275 -I0513 13:02:33.356860 140330640848640 logging_writer.py:48] [137900] global_step=137900, grad_norm=4.596782207489014, loss=3.9094347953796387 -I0513 13:03:10.918118 140330632455936 logging_writer.py:48] [138000] global_step=138000, grad_norm=2.9868075847625732, loss=3.840625286102295 -I0513 13:04:07.748569 140330640848640 logging_writer.py:48] [138100] global_step=138100, grad_norm=2.9404895305633545, loss=3.7078092098236084 -I0513 13:04:53.061163 140330632455936 logging_writer.py:48] [138200] global_step=138200, grad_norm=7.182351112365723, loss=3.8980658054351807 -I0513 13:05:36.914340 140330640848640 logging_writer.py:48] [138300] global_step=138300, grad_norm=11.149739265441895, loss=3.744079351425171 -I0513 13:06:21.805096 140330632455936 logging_writer.py:48] [138400] global_step=138400, grad_norm=6.542133331298828, loss=3.906740665435791 -I0513 13:07:20.140668 140330640848640 logging_writer.py:48] [138500] global_step=138500, grad_norm=9.344233512878418, loss=4.246111869812012 -I0513 13:07:24.006447 140546196993216 spec.py:333] Evaluating on the training split. -I0513 13:07:30.545465 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 13:07:53.293397 140546196993216 spec.py:363] Evaluating on the test split. -I0513 13:07:54.441582 140546196993216 submission_runner.py:516] Time since start: 52638.24s, Step: 138508, {'train/accuracy': Array(0.00127551, dtype=float32), 'train/loss': Array(9.419978, dtype=float32), 'validation/accuracy': Array(0.00136, dtype=float32), 'validation/loss': Array(9.398717, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0013, dtype=float32), 'test/loss': Array(9.4097805, dtype=float32), 'test/num_examples': 10000, 'score': 51957.18427658081, 'total_duration': 52638.24155449867, 'accumulated_submission_time': 51957.18427658081, 'accumulated_eval_time': 676.9540636539459, 'accumulated_logging_time': 2.3545475006103516} -I0513 13:07:54.864463 140330632455936 logging_writer.py:48] [138508] accumulated_eval_time=676.954, accumulated_logging_time=2.35455, accumulated_submission_time=51957.2, global_step=138508, preemption_count=0, score=51957.2, test/accuracy=0.0013000001199543476, test/loss=9.409780502319336, test/num_examples=10000, total_duration=52638.2, train/accuracy=0.0012755101779475808, train/loss=9.419978141784668, validation/accuracy=0.0013599999947473407, validation/loss=9.398716926574707, validation/num_examples=50000 -I0513 13:08:29.338152 140330640848640 logging_writer.py:48] [138600] global_step=138600, grad_norm=3.6632020473480225, loss=3.890957832336426 -I0513 13:09:10.850072 140330632455936 logging_writer.py:48] [138700] global_step=138700, grad_norm=64.22468566894531, loss=4.100459575653076 -I0513 13:09:56.533874 140330640848640 logging_writer.py:48] [138800] global_step=138800, grad_norm=3.2570676803588867, loss=3.785815477371216 -I0513 13:10:55.153367 140330632455936 logging_writer.py:48] [138900] global_step=138900, grad_norm=4.353621959686279, loss=3.785520076751709 -I0513 13:11:32.475528 140330640848640 logging_writer.py:48] [139000] global_step=139000, grad_norm=5.263772964477539, loss=4.095412731170654 -I0513 13:12:09.737462 140330632455936 logging_writer.py:48] [139100] global_step=139100, grad_norm=5.298800468444824, loss=3.9184815883636475 -I0513 13:12:47.821369 140330640848640 logging_writer.py:48] [139200] global_step=139200, grad_norm=3.988530158996582, loss=3.8498177528381348 -I0513 13:13:25.762560 140330632455936 logging_writer.py:48] [139300] global_step=139300, grad_norm=76.36328887939453, loss=4.218533992767334 -I0513 13:14:08.939174 140330640848640 logging_writer.py:48] [139400] global_step=139400, grad_norm=4.249964237213135, loss=3.8577637672424316 -I0513 13:14:50.589458 140330632455936 logging_writer.py:48] [139500] global_step=139500, grad_norm=3.457367181777954, loss=3.750880241394043 -I0513 13:15:40.553478 140330640848640 logging_writer.py:48] [139600] global_step=139600, grad_norm=7.177361011505127, loss=3.8358511924743652 -I0513 13:16:18.315171 140330632455936 logging_writer.py:48] [139700] global_step=139700, grad_norm=4.927772521972656, loss=3.818434715270996 -I0513 13:16:57.847064 140330640848640 logging_writer.py:48] [139800] global_step=139800, grad_norm=7.20550537109375, loss=3.826725482940674 -I0513 13:17:35.504189 140330632455936 logging_writer.py:48] [139900] global_step=139900, grad_norm=12.950654983520508, loss=4.053020477294922 -I0513 13:18:12.658271 140330640848640 logging_writer.py:48] [140000] global_step=140000, grad_norm=4.73526668548584, loss=3.823206901550293 -I0513 13:18:50.128906 140330632455936 logging_writer.py:48] [140100] global_step=140100, grad_norm=5.064901828765869, loss=3.886415719985962 -I0513 13:19:30.604791 140330640848640 logging_writer.py:48] [140200] global_step=140200, grad_norm=5.458664894104004, loss=3.9222054481506348 -I0513 13:20:10.943483 140330632455936 logging_writer.py:48] [140300] global_step=140300, grad_norm=3.3594112396240234, loss=3.720768928527832 -I0513 13:20:56.766402 140330640848640 logging_writer.py:48] [140400] global_step=140400, grad_norm=3.6445326805114746, loss=3.8913087844848633 -I0513 13:21:40.300580 140330632455936 logging_writer.py:48] [140500] global_step=140500, grad_norm=4.131740570068359, loss=3.764554977416992 -I0513 13:22:22.455273 140330640848640 logging_writer.py:48] [140600] global_step=140600, grad_norm=6.49498176574707, loss=3.872887134552002 -I0513 13:23:00.564013 140330632455936 logging_writer.py:48] [140700] global_step=140700, grad_norm=4.390547275543213, loss=3.812690258026123 -I0513 13:23:37.276911 140330640848640 logging_writer.py:48] [140800] global_step=140800, grad_norm=7.276374340057373, loss=3.9773364067077637 -I0513 13:24:18.177953 140330632455936 logging_writer.py:48] [140900] global_step=140900, grad_norm=4.154356002807617, loss=3.805471658706665 -I0513 13:25:01.702927 140330640848640 logging_writer.py:48] [141000] global_step=141000, grad_norm=24.050682067871094, loss=4.037720203399658 -I0513 13:25:43.466901 140330632455936 logging_writer.py:48] [141100] global_step=141100, grad_norm=4.127982139587402, loss=3.7962687015533447 -I0513 13:26:24.077126 140330640848640 logging_writer.py:48] [141200] global_step=141200, grad_norm=4.220571994781494, loss=3.7542223930358887 -I0513 13:27:09.907799 140330632455936 logging_writer.py:48] [141300] global_step=141300, grad_norm=3.7134833335876465, loss=3.754962205886841 -I0513 13:27:56.713405 140330640848640 logging_writer.py:48] [141400] global_step=141400, grad_norm=3.170220375061035, loss=3.7270150184631348 -I0513 13:28:38.706776 140330632455936 logging_writer.py:48] [141500] global_step=141500, grad_norm=5.334285259246826, loss=3.8579397201538086 -I0513 13:29:25.765289 140330640848640 logging_writer.py:48] [141600] global_step=141600, grad_norm=4.222395896911621, loss=3.8799962997436523 -I0513 13:30:03.333922 140330632455936 logging_writer.py:48] [141700] global_step=141700, grad_norm=6.0724382400512695, loss=3.8960862159729004 -I0513 13:30:41.269894 140330640848640 logging_writer.py:48] [141800] global_step=141800, grad_norm=6.571535110473633, loss=3.892942428588867 -I0513 13:31:24.775068 140330632455936 logging_writer.py:48] [141900] global_step=141900, grad_norm=4.063754081726074, loss=3.775085687637329 -I0513 13:32:09.259777 140330640848640 logging_writer.py:48] [142000] global_step=142000, grad_norm=4.381030559539795, loss=3.809974431991577 -I0513 13:32:53.963173 140330632455936 logging_writer.py:48] [142100] global_step=142100, grad_norm=5.392723083496094, loss=3.800435781478882 -I0513 13:33:41.046147 140330640848640 logging_writer.py:48] [142200] global_step=142200, grad_norm=8.136300086975098, loss=3.8270134925842285 -I0513 13:34:18.710286 140330632455936 logging_writer.py:48] [142300] global_step=142300, grad_norm=4.029725074768066, loss=3.784245014190674 -I0513 13:35:01.410122 140330640848640 logging_writer.py:48] [142400] global_step=142400, grad_norm=6.9994964599609375, loss=3.8289875984191895 -I0513 13:35:44.904802 140330632455936 logging_writer.py:48] [142500] global_step=142500, grad_norm=7.58408784866333, loss=3.78753662109375 -I0513 13:36:30.543677 140330640848640 logging_writer.py:48] [142600] global_step=142600, grad_norm=7.731447219848633, loss=3.9578492641448975 -I0513 13:37:19.181820 140330632455936 logging_writer.py:48] [142700] global_step=142700, grad_norm=3.374532699584961, loss=3.6836023330688477 -I0513 13:38:09.017460 140330640848640 logging_writer.py:48] [142800] global_step=142800, grad_norm=3.4365313053131104, loss=3.8794684410095215 -I0513 13:38:59.571108 140330632455936 logging_writer.py:48] [142900] global_step=142900, grad_norm=8.509371757507324, loss=3.891690254211426 -I0513 13:39:43.344325 140330640848640 logging_writer.py:48] [143000] global_step=143000, grad_norm=4.133147716522217, loss=3.9186899662017822 -I0513 13:40:31.799865 140330632455936 logging_writer.py:48] [143100] global_step=143100, grad_norm=7.039206504821777, loss=3.7184042930603027 -I0513 13:41:10.383868 140546196993216 spec.py:333] Evaluating on the training split. -I0513 13:41:18.875229 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 13:41:49.715568 140546196993216 spec.py:363] Evaluating on the test split. -I0513 13:41:50.812548 140546196993216 submission_runner.py:516] Time since start: 54674.66s, Step: 143180, {'train/accuracy': Array(0.00131537, dtype=float32), 'train/loss': Array(9.403687, dtype=float32), 'validation/accuracy': Array(0.0014, dtype=float32), 'validation/loss': Array(9.406859, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0012, dtype=float32), 'test/loss': Array(9.415645, dtype=float32), 'test/num_examples': 10000, 'score': 53952.60812115669, 'total_duration': 54674.6584277153, 'accumulated_submission_time': 53952.60812115669, 'accumulated_eval_time': 717.1750359535217, 'accumulated_logging_time': 2.830092191696167} -I0513 13:41:51.290949 140330640848640 logging_writer.py:48] [143180] accumulated_eval_time=717.175, accumulated_logging_time=2.83009, accumulated_submission_time=53952.6, global_step=143180, preemption_count=0, score=53952.6, test/accuracy=0.0012000000569969416, test/loss=9.415644645690918, test/num_examples=10000, total_duration=54674.7, train/accuracy=0.0013153698528185487, train/loss=9.4036865234375, validation/accuracy=0.00139999995008111, validation/loss=9.406859397888184, validation/num_examples=50000 -I0513 13:41:59.560266 140330632455936 logging_writer.py:48] [143200] global_step=143200, grad_norm=8.417012214660645, loss=3.800234317779541 -I0513 13:42:42.826157 140330640848640 logging_writer.py:48] [143300] global_step=143300, grad_norm=3.885958194732666, loss=3.7243056297302246 -I0513 13:43:29.851758 140330632455936 logging_writer.py:48] [143400] global_step=143400, grad_norm=3.806203603744507, loss=3.7753543853759766 -I0513 13:44:18.125216 140330640848640 logging_writer.py:48] [143500] global_step=143500, grad_norm=5.661909580230713, loss=3.8199589252471924 -I0513 13:45:04.483465 140330632455936 logging_writer.py:48] [143600] global_step=143600, grad_norm=17.891427993774414, loss=3.992310047149658 -I0513 13:45:51.739003 140330640848640 logging_writer.py:48] [143700] global_step=143700, grad_norm=2.8821465969085693, loss=3.7749252319335938 -I0513 13:46:38.904789 140330632455936 logging_writer.py:48] [143800] global_step=143800, grad_norm=3.601325035095215, loss=3.826688528060913 -I0513 13:47:26.675939 140330640848640 logging_writer.py:48] [143900] global_step=143900, grad_norm=3.353801965713501, loss=3.835735321044922 -I0513 13:48:16.777686 140330632455936 logging_writer.py:48] [144000] global_step=144000, grad_norm=5.24884033203125, loss=3.8046536445617676 -I0513 13:49:07.816229 140330640848640 logging_writer.py:48] [144100] global_step=144100, grad_norm=67.02204895019531, loss=3.989696502685547 -I0513 13:49:55.011918 140330632455936 logging_writer.py:48] [144200] global_step=144200, grad_norm=3.6371889114379883, loss=3.693881034851074 -I0513 13:50:42.482844 140330640848640 logging_writer.py:48] [144300] global_step=144300, grad_norm=8.611148834228516, loss=3.89901065826416 -I0513 13:51:32.266165 140330632455936 logging_writer.py:48] [144400] global_step=144400, grad_norm=5.865719318389893, loss=3.885667562484741 -I0513 13:52:20.820917 140330640848640 logging_writer.py:48] [144500] global_step=144500, grad_norm=3.796079158782959, loss=3.8399200439453125 -I0513 13:53:10.111679 140330632455936 logging_writer.py:48] [144600] global_step=144600, grad_norm=5.471855163574219, loss=3.882789373397827 -I0513 13:53:58.593020 140330640848640 logging_writer.py:48] [144700] global_step=144700, grad_norm=6.017971038818359, loss=3.9730420112609863 -I0513 13:54:49.658340 140330632455936 logging_writer.py:48] [144800] global_step=144800, grad_norm=3.723093271255493, loss=3.7869057655334473 -I0513 13:55:36.127456 140330640848640 logging_writer.py:48] [144900] global_step=144900, grad_norm=4.171130180358887, loss=3.919508457183838 -I0513 13:56:21.930718 140330632455936 logging_writer.py:48] [145000] global_step=145000, grad_norm=26.1159725189209, loss=3.954209327697754 -I0513 13:56:59.753403 140330640848640 logging_writer.py:48] [145100] global_step=145100, grad_norm=2.748377561569214, loss=3.7512102127075195 -I0513 13:57:42.877909 140330632455936 logging_writer.py:48] [145200] global_step=145200, grad_norm=5.9460835456848145, loss=3.8501486778259277 -I0513 13:58:26.588683 140330640848640 logging_writer.py:48] [145300] global_step=145300, grad_norm=3.850977897644043, loss=3.7350804805755615 -I0513 13:59:13.829001 140330632455936 logging_writer.py:48] [145400] global_step=145400, grad_norm=6.275508403778076, loss=4.062191963195801 -I0513 14:00:03.101190 140330640848640 logging_writer.py:48] [145500] global_step=145500, grad_norm=4.9364519119262695, loss=3.907970428466797 -I0513 14:00:52.042449 140330632455936 logging_writer.py:48] [145600] global_step=145600, grad_norm=3.0653879642486572, loss=3.729074001312256 -I0513 14:01:40.944610 140330640848640 logging_writer.py:48] [145700] global_step=145700, grad_norm=5.951106071472168, loss=3.743091583251953 -I0513 14:02:30.697243 140330632455936 logging_writer.py:48] [145800] global_step=145800, grad_norm=24.099552154541016, loss=3.8994054794311523 -I0513 14:03:29.423979 140330640848640 logging_writer.py:48] [145900] global_step=145900, grad_norm=11.07242488861084, loss=3.857088088989258 -I0513 14:04:13.824651 140330632455936 logging_writer.py:48] [146000] global_step=146000, grad_norm=3.9223759174346924, loss=3.8250789642333984 -I0513 14:04:59.069419 140330640848640 logging_writer.py:48] [146100] global_step=146100, grad_norm=3.4354655742645264, loss=3.7345051765441895 -I0513 14:05:43.021342 140330632455936 logging_writer.py:48] [146200] global_step=146200, grad_norm=4.444423675537109, loss=3.8440661430358887 -I0513 14:06:38.142812 140330640848640 logging_writer.py:48] [146300] global_step=146300, grad_norm=7.004632472991943, loss=3.9814000129699707 -I0513 14:07:17.417031 140330632455936 logging_writer.py:48] [146400] global_step=146400, grad_norm=5.335745811462402, loss=3.762026786804199 -I0513 14:08:14.983199 140330640848640 logging_writer.py:48] [146500] global_step=146500, grad_norm=6.760083198547363, loss=3.9517574310302734 -I0513 14:08:57.275699 140330632455936 logging_writer.py:48] [146600] global_step=146600, grad_norm=4.293241024017334, loss=3.7198445796966553 -I0513 14:09:42.293479 140330640848640 logging_writer.py:48] [146700] global_step=146700, grad_norm=6.031266689300537, loss=3.7503433227539062 -I0513 14:10:21.083077 140330632455936 logging_writer.py:48] [146800] global_step=146800, grad_norm=3.2752747535705566, loss=3.7268857955932617 -I0513 14:11:07.059512 140330640848640 logging_writer.py:48] [146900] global_step=146900, grad_norm=5.1298112869262695, loss=3.792056083679199 -I0513 14:11:45.561752 140330632455936 logging_writer.py:48] [147000] global_step=147000, grad_norm=4.258084774017334, loss=3.9766101837158203 -I0513 14:12:23.294600 140330640848640 logging_writer.py:48] [147100] global_step=147100, grad_norm=5.023778438568115, loss=3.796319007873535 -I0513 14:13:01.445697 140330632455936 logging_writer.py:48] [147200] global_step=147200, grad_norm=3.267125129699707, loss=3.8037283420562744 -I0513 14:13:39.583991 140330640848640 logging_writer.py:48] [147300] global_step=147300, grad_norm=3.988088846206665, loss=3.877974510192871 -I0513 14:14:17.329105 140330632455936 logging_writer.py:48] [147400] global_step=147400, grad_norm=3.7063822746276855, loss=3.7941246032714844 -I0513 14:14:57.127991 140330640848640 logging_writer.py:48] [147500] global_step=147500, grad_norm=4.9225029945373535, loss=3.781665325164795 -I0513 14:15:07.868850 140546196993216 spec.py:333] Evaluating on the training split. -I0513 14:15:16.147010 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 14:16:06.399239 140546196993216 spec.py:363] Evaluating on the test split. -I0513 14:16:07.521577 140546196993216 submission_runner.py:516] Time since start: 56731.34s, Step: 147523, {'train/accuracy': Array(0.00159439, dtype=float32), 'train/loss': Array(9.445664, dtype=float32), 'validation/accuracy': Array(0.00138, dtype=float32), 'validation/loss': Array(9.394714, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0014, dtype=float32), 'test/loss': Array(9.402309, dtype=float32), 'test/num_examples': 10000, 'score': 55949.09929037094, 'total_duration': 56731.34460759163, 'accumulated_submission_time': 55949.09929037094, 'accumulated_eval_time': 776.5972204208374, 'accumulated_logging_time': 3.3548340797424316} -I0513 14:16:07.901608 140330632455936 logging_writer.py:48] [147523] accumulated_eval_time=776.597, accumulated_logging_time=3.35483, accumulated_submission_time=55949.1, global_step=147523, preemption_count=0, score=55949.1, test/accuracy=0.0014000000664964318, test/loss=9.40230941772461, test/num_examples=10000, total_duration=56731.3, train/accuracy=0.0015943876933306456, train/loss=9.445664405822754, validation/accuracy=0.0013799999142065644, validation/loss=9.39471435546875, validation/num_examples=50000 -I0513 14:16:39.192274 140330640848640 logging_writer.py:48] [147600] global_step=147600, grad_norm=3.171635389328003, loss=3.737839698791504 -I0513 14:17:21.865343 140330632455936 logging_writer.py:48] [147700] global_step=147700, grad_norm=3.304255485534668, loss=3.720695972442627 -I0513 14:18:02.579068 140330640848640 logging_writer.py:48] [147800] global_step=147800, grad_norm=4.225671291351318, loss=3.766728401184082 -I0513 14:18:41.788146 140330632455936 logging_writer.py:48] [147900] global_step=147900, grad_norm=4.739205360412598, loss=3.937917470932007 -I0513 14:19:19.448270 140330640848640 logging_writer.py:48] [148000] global_step=148000, grad_norm=3.5724470615386963, loss=3.9006009101867676 -I0513 14:19:57.637033 140330632455936 logging_writer.py:48] [148100] global_step=148100, grad_norm=9.288153648376465, loss=3.8983397483825684 -I0513 14:20:37.064834 140330640848640 logging_writer.py:48] [148200] global_step=148200, grad_norm=10.898133277893066, loss=4.0100250244140625 -I0513 14:21:21.302808 140330632455936 logging_writer.py:48] [148300] global_step=148300, grad_norm=15.055841445922852, loss=3.7654237747192383 -I0513 14:22:07.174324 140330640848640 logging_writer.py:48] [148400] global_step=148400, grad_norm=6.096031188964844, loss=3.790715217590332 -I0513 14:22:58.239964 140330632455936 logging_writer.py:48] [148500] global_step=148500, grad_norm=4.69727087020874, loss=3.900144577026367 -I0513 14:23:43.552012 140330640848640 logging_writer.py:48] [148600] global_step=148600, grad_norm=6.91024112701416, loss=3.9030861854553223 -I0513 14:24:28.476331 140330632455936 logging_writer.py:48] [148700] global_step=148700, grad_norm=21.462139129638672, loss=3.753276824951172 -I0513 14:25:08.701843 140330640848640 logging_writer.py:48] [148800] global_step=148800, grad_norm=7.779849529266357, loss=3.877763271331787 -I0513 14:25:48.326626 140330632455936 logging_writer.py:48] [148900] global_step=148900, grad_norm=4.2619404792785645, loss=3.889697313308716 -I0513 14:26:31.741247 140330640848640 logging_writer.py:48] [149000] global_step=149000, grad_norm=4.098154067993164, loss=3.7394490242004395 -I0513 14:27:26.394732 140330632455936 logging_writer.py:48] [149100] global_step=149100, grad_norm=2.8748207092285156, loss=3.9043445587158203 -I0513 14:28:11.735489 140330640848640 logging_writer.py:48] [149200] global_step=149200, grad_norm=8.268768310546875, loss=4.014427661895752 -I0513 14:29:01.718910 140330632455936 logging_writer.py:48] [149300] global_step=149300, grad_norm=5.459614276885986, loss=3.8668227195739746 -I0513 14:29:50.122786 140330640848640 logging_writer.py:48] [149400] global_step=149400, grad_norm=4.121537685394287, loss=4.012465476989746 -I0513 14:30:28.141892 140330632455936 logging_writer.py:48] [149500] global_step=149500, grad_norm=3.0758721828460693, loss=3.7618629932403564 -I0513 14:31:05.307974 140330640848640 logging_writer.py:48] [149600] global_step=149600, grad_norm=5.506356239318848, loss=3.799036979675293 -I0513 14:31:42.710665 140330632455936 logging_writer.py:48] [149700] global_step=149700, grad_norm=5.030240058898926, loss=3.799504518508911 -I0513 14:32:20.592167 140330640848640 logging_writer.py:48] [149800] global_step=149800, grad_norm=4.217313289642334, loss=3.857665777206421 -I0513 14:32:57.754242 140330632455936 logging_writer.py:48] [149900] global_step=149900, grad_norm=5.636961936950684, loss=3.7762184143066406 -I0513 14:33:34.938003 140330640848640 logging_writer.py:48] [150000] global_step=150000, grad_norm=4.91845703125, loss=3.735990047454834 -I0513 14:34:14.820819 140330632455936 logging_writer.py:48] [150100] global_step=150100, grad_norm=11.616698265075684, loss=3.7973947525024414 -I0513 14:34:57.668208 140330640848640 logging_writer.py:48] [150200] global_step=150200, grad_norm=4.347864151000977, loss=4.036248207092285 -I0513 14:35:35.122761 140330632455936 logging_writer.py:48] [150300] global_step=150300, grad_norm=6.676313400268555, loss=3.7735044956207275 -I0513 14:36:12.736408 140330640848640 logging_writer.py:48] [150400] global_step=150400, grad_norm=5.4279255867004395, loss=3.9672842025756836 -I0513 14:36:49.745759 140330632455936 logging_writer.py:48] [150500] global_step=150500, grad_norm=4.634213924407959, loss=3.857707977294922 -I0513 14:37:28.032158 140330640848640 logging_writer.py:48] [150600] global_step=150600, grad_norm=15.116458892822266, loss=4.028441429138184 -I0513 14:38:04.815494 140330632455936 logging_writer.py:48] [150700] global_step=150700, grad_norm=4.18612003326416, loss=3.8675360679626465 -I0513 14:38:42.006574 140330640848640 logging_writer.py:48] [150800] global_step=150800, grad_norm=7.285975933074951, loss=3.7961573600769043 -I0513 14:39:19.211353 140330632455936 logging_writer.py:48] [150900] global_step=150900, grad_norm=3.558384656906128, loss=3.762108325958252 -I0513 14:39:55.952964 140330640848640 logging_writer.py:48] [151000] global_step=151000, grad_norm=3.3244009017944336, loss=3.797433614730835 -I0513 14:40:33.029269 140330632455936 logging_writer.py:48] [151100] global_step=151100, grad_norm=6.050889015197754, loss=3.8820176124572754 -I0513 14:41:10.354563 140330640848640 logging_writer.py:48] [151200] global_step=151200, grad_norm=13.137903213500977, loss=4.185361385345459 -I0513 14:41:55.396671 140330632455936 logging_writer.py:48] [151300] global_step=151300, grad_norm=2.7950596809387207, loss=3.8225584030151367 -I0513 14:42:35.206148 140330640848640 logging_writer.py:48] [151400] global_step=151400, grad_norm=3.147218942642212, loss=3.8597285747528076 -I0513 14:43:13.514359 140330632455936 logging_writer.py:48] [151500] global_step=151500, grad_norm=3.586172342300415, loss=3.699145793914795 -I0513 14:43:51.328198 140330640848640 logging_writer.py:48] [151600] global_step=151600, grad_norm=8.565522193908691, loss=3.8736214637756348 -I0513 14:44:28.125187 140330632455936 logging_writer.py:48] [151700] global_step=151700, grad_norm=4.320493221282959, loss=3.6521337032318115 -I0513 14:45:12.882889 140330640848640 logging_writer.py:48] [151800] global_step=151800, grad_norm=7.360283851623535, loss=3.822566032409668 -I0513 14:45:58.423081 140330632455936 logging_writer.py:48] [151900] global_step=151900, grad_norm=6.559606552124023, loss=3.7674622535705566 -I0513 14:46:45.450056 140330640848640 logging_writer.py:48] [152000] global_step=152000, grad_norm=4.849315643310547, loss=3.754808187484741 -I0513 14:47:31.600962 140330632455936 logging_writer.py:48] [152100] global_step=152100, grad_norm=7.494271278381348, loss=3.948124408721924 -I0513 14:48:18.607301 140330640848640 logging_writer.py:48] [152200] global_step=152200, grad_norm=4.505961894989014, loss=3.6923234462738037 -I0513 14:49:01.882078 140330632455936 logging_writer.py:48] [152300] global_step=152300, grad_norm=3.938826560974121, loss=3.816234588623047 -I0513 14:49:23.416209 140546196993216 spec.py:333] Evaluating on the training split. -I0513 14:49:30.993670 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 14:50:03.578682 140546196993216 spec.py:363] Evaluating on the test split. -I0513 14:50:04.726084 140546196993216 submission_runner.py:516] Time since start: 58768.53s, Step: 152348, {'train/accuracy': Array(0.00145488, dtype=float32), 'train/loss': Array(9.38589, dtype=float32), 'validation/accuracy': Array(0.00148, dtype=float32), 'validation/loss': Array(9.364571, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0014, dtype=float32), 'test/loss': Array(9.372682, dtype=float32), 'test/num_examples': 10000, 'score': 57944.53552174568, 'total_duration': 58768.525552749634, 'accumulated_submission_time': 57944.53552174568, 'accumulated_eval_time': 817.6529867649078, 'accumulated_logging_time': 3.770930528640747} -I0513 14:50:05.104979 140330640848640 logging_writer.py:48] [152348] accumulated_eval_time=817.653, accumulated_logging_time=3.77093, accumulated_submission_time=57944.5, global_step=152348, preemption_count=0, score=57944.5, test/accuracy=0.0014000000664964318, test/loss=9.372681617736816, test/num_examples=10000, total_duration=58768.5, train/accuracy=0.001454878831282258, train/loss=9.385890007019043, validation/accuracy=0.0014799999771639705, validation/loss=9.364570617675781, validation/num_examples=50000 -I0513 14:50:30.691588 140330632455936 logging_writer.py:48] [152400] global_step=152400, grad_norm=17.208049774169922, loss=3.8556580543518066 -I0513 14:51:09.954208 140330640848640 logging_writer.py:48] [152500] global_step=152500, grad_norm=6.974955081939697, loss=3.951392889022827 -I0513 14:51:49.023472 140330632455936 logging_writer.py:48] [152600] global_step=152600, grad_norm=5.717930793762207, loss=3.780928134918213 -I0513 14:52:32.792656 140330640848640 logging_writer.py:48] [152700] global_step=152700, grad_norm=5.258434295654297, loss=3.8981168270111084 -I0513 14:53:18.297731 140330632455936 logging_writer.py:48] [152800] global_step=152800, grad_norm=3.8020989894866943, loss=3.919856309890747 -I0513 14:54:02.969590 140330640848640 logging_writer.py:48] [152900] global_step=152900, grad_norm=3.505695104598999, loss=3.8322505950927734 -I0513 14:54:46.716646 140330632455936 logging_writer.py:48] [153000] global_step=153000, grad_norm=2.8507699966430664, loss=3.8038218021392822 -I0513 14:55:31.379460 140330640848640 logging_writer.py:48] [153100] global_step=153100, grad_norm=11.638980865478516, loss=4.108363151550293 -I0513 14:56:16.718481 140330632455936 logging_writer.py:48] [153200] global_step=153200, grad_norm=4.421864986419678, loss=3.7954037189483643 -I0513 14:57:02.455754 140330640848640 logging_writer.py:48] [153300] global_step=153300, grad_norm=2.532761573791504, loss=3.729057788848877 -I0513 14:57:49.129212 140330632455936 logging_writer.py:48] [153400] global_step=153400, grad_norm=4.07662296295166, loss=3.9194631576538086 -I0513 14:58:37.481167 140330640848640 logging_writer.py:48] [153500] global_step=153500, grad_norm=2.5437397956848145, loss=3.667017936706543 -I0513 14:59:24.653720 140330632455936 logging_writer.py:48] [153600] global_step=153600, grad_norm=2.8428187370300293, loss=3.821521282196045 -I0513 15:00:11.656915 140330640848640 logging_writer.py:48] [153700] global_step=153700, grad_norm=5.318944454193115, loss=3.8194093704223633 -I0513 15:00:58.306108 140330632455936 logging_writer.py:48] [153800] global_step=153800, grad_norm=5.86020040512085, loss=3.8340673446655273 -I0513 15:01:46.906480 140330640848640 logging_writer.py:48] [153900] global_step=153900, grad_norm=3.5451645851135254, loss=3.749239921569824 -I0513 15:02:35.628386 140330632455936 logging_writer.py:48] [154000] global_step=154000, grad_norm=3.9360411167144775, loss=3.766315460205078 -I0513 15:03:22.194244 140330640848640 logging_writer.py:48] [154100] global_step=154100, grad_norm=4.932788848876953, loss=3.8913042545318604 -I0513 15:04:08.790114 140330632455936 logging_writer.py:48] [154200] global_step=154200, grad_norm=5.093003749847412, loss=3.922196388244629 -I0513 15:04:55.419370 140330640848640 logging_writer.py:48] [154300] global_step=154300, grad_norm=5.601552486419678, loss=3.8808422088623047 -I0513 15:05:42.303330 140330632455936 logging_writer.py:48] [154400] global_step=154400, grad_norm=11.102788925170898, loss=3.9678430557250977 -I0513 15:06:29.641497 140330640848640 logging_writer.py:48] [154500] global_step=154500, grad_norm=4.572290897369385, loss=3.9171812534332275 -I0513 15:07:16.940922 140330632455936 logging_writer.py:48] [154600] global_step=154600, grad_norm=6.4344048500061035, loss=4.035294532775879 -I0513 15:08:04.261616 140330640848640 logging_writer.py:48] [154700] global_step=154700, grad_norm=2.342451333999634, loss=3.754913330078125 -I0513 15:08:53.991575 140330632455936 logging_writer.py:48] [154800] global_step=154800, grad_norm=6.845343589782715, loss=3.866798162460327 -I0513 15:09:37.654467 140330640848640 logging_writer.py:48] [154900] global_step=154900, grad_norm=6.397772312164307, loss=3.787294387817383 -I0513 15:10:24.271569 140330632455936 logging_writer.py:48] [155000] global_step=155000, grad_norm=2.34704327583313, loss=3.793665885925293 -I0513 15:11:12.089468 140330640848640 logging_writer.py:48] [155100] global_step=155100, grad_norm=3.2869186401367188, loss=3.8164806365966797 -I0513 15:12:00.904147 140330632455936 logging_writer.py:48] [155200] global_step=155200, grad_norm=2.823338270187378, loss=3.7084624767303467 -I0513 15:12:45.684814 140330640848640 logging_writer.py:48] [155300] global_step=155300, grad_norm=4.066617965698242, loss=3.617995500564575 -I0513 15:13:30.615384 140330632455936 logging_writer.py:48] [155400] global_step=155400, grad_norm=5.534333229064941, loss=3.885864019393921 -I0513 15:14:17.842694 140330640848640 logging_writer.py:48] [155500] global_step=155500, grad_norm=4.004809379577637, loss=3.8666129112243652 -I0513 15:15:06.203131 140330632455936 logging_writer.py:48] [155600] global_step=155600, grad_norm=4.071564674377441, loss=3.8133769035339355 -I0513 15:15:55.731601 140330640848640 logging_writer.py:48] [155700] global_step=155700, grad_norm=3.95855712890625, loss=3.803419351577759 -I0513 15:16:40.696298 140330632455936 logging_writer.py:48] [155800] global_step=155800, grad_norm=7.641286373138428, loss=3.85568904876709 -I0513 15:17:31.652808 140330640848640 logging_writer.py:48] [155900] global_step=155900, grad_norm=3.2434892654418945, loss=3.77272367477417 -I0513 15:18:17.221478 140330632455936 logging_writer.py:48] [156000] global_step=156000, grad_norm=2.732757568359375, loss=3.760586738586426 -I0513 15:19:05.886517 140330640848640 logging_writer.py:48] [156100] global_step=156100, grad_norm=4.404449939727783, loss=3.8449583053588867 -I0513 15:19:59.223678 140330632455936 logging_writer.py:48] [156200] global_step=156200, grad_norm=5.548603534698486, loss=3.809950590133667 -I0513 15:20:45.025786 140330640848640 logging_writer.py:48] [156300] global_step=156300, grad_norm=4.079488277435303, loss=3.874774694442749 -I0513 15:21:30.200598 140330632455936 logging_writer.py:48] [156400] global_step=156400, grad_norm=7.245952606201172, loss=3.7785863876342773 -I0513 15:22:16.691085 140330640848640 logging_writer.py:48] [156500] global_step=156500, grad_norm=4.056859493255615, loss=3.811737060546875 -I0513 15:23:03.974549 140330632455936 logging_writer.py:48] [156600] global_step=156600, grad_norm=37.37944412231445, loss=3.8493261337280273 -I0513 15:23:20.548763 140546196993216 spec.py:333] Evaluating on the training split. -I0513 15:23:29.157151 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 15:24:01.572967 140546196993216 spec.py:363] Evaluating on the test split. -I0513 15:24:02.667693 140546196993216 submission_runner.py:516] Time since start: 60806.51s, Step: 156638, {'train/accuracy': Array(0.00151467, dtype=float32), 'train/loss': Array(9.388604, dtype=float32), 'validation/accuracy': Array(0.00148, dtype=float32), 'validation/loss': Array(9.365797, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(9.373345, dtype=float32), 'test/num_examples': 10000, 'score': 59939.896312236786, 'total_duration': 60806.513526916504, 'accumulated_submission_time': 59939.896312236786, 'accumulated_eval_time': 859.5641634464264, 'accumulated_logging_time': 4.192249059677124} -I0513 15:24:03.107502 140330640848640 logging_writer.py:48] [156638] accumulated_eval_time=859.564, accumulated_logging_time=4.19225, accumulated_submission_time=59939.9, global_step=156638, preemption_count=0, score=59939.9, test/accuracy=0.001500000013038516, test/loss=9.373345375061035, test/num_examples=10000, total_duration=60806.5, train/accuracy=0.0015146683435887098, train/loss=9.388604164123535, validation/accuracy=0.0014799999771639705, validation/loss=9.36579704284668, validation/num_examples=50000 -I0513 15:24:31.326719 140330632455936 logging_writer.py:48] [156700] global_step=156700, grad_norm=5.616528034210205, loss=3.7639949321746826 -I0513 15:25:13.364111 140330640848640 logging_writer.py:48] [156800] global_step=156800, grad_norm=2.974623918533325, loss=3.7518389225006104 -I0513 15:25:56.968416 140330632455936 logging_writer.py:48] [156900] global_step=156900, grad_norm=4.03665828704834, loss=3.8593993186950684 -I0513 15:26:41.457400 140330640848640 logging_writer.py:48] [157000] global_step=157000, grad_norm=3.9090735912323, loss=3.743299961090088 -I0513 15:27:25.318411 140330632455936 logging_writer.py:48] [157100] global_step=157100, grad_norm=16.009733200073242, loss=3.856306552886963 -I0513 15:28:11.465682 140330640848640 logging_writer.py:48] [157200] global_step=157200, grad_norm=9.098628997802734, loss=3.911116361618042 -I0513 15:28:55.512621 140330632455936 logging_writer.py:48] [157300] global_step=157300, grad_norm=7.167022228240967, loss=3.9568848609924316 -I0513 15:29:44.788228 140330640848640 logging_writer.py:48] [157400] global_step=157400, grad_norm=4.793281078338623, loss=4.0036940574646 -I0513 15:30:29.824859 140330632455936 logging_writer.py:48] [157500] global_step=157500, grad_norm=3.5251970291137695, loss=3.728787422180176 -I0513 15:31:15.365667 140330640848640 logging_writer.py:48] [157600] global_step=157600, grad_norm=3.3508903980255127, loss=3.677532911300659 -I0513 15:32:00.685241 140330632455936 logging_writer.py:48] [157700] global_step=157700, grad_norm=6.518189907073975, loss=3.8434534072875977 -I0513 15:32:48.272272 140330640848640 logging_writer.py:48] [157800] global_step=157800, grad_norm=4.421384811401367, loss=3.8205161094665527 -I0513 15:33:37.377413 140330632455936 logging_writer.py:48] [157900] global_step=157900, grad_norm=5.97616720199585, loss=3.857008934020996 -I0513 15:34:24.696318 140330640848640 logging_writer.py:48] [158000] global_step=158000, grad_norm=4.109250068664551, loss=3.764469623565674 -I0513 15:35:15.402590 140330632455936 logging_writer.py:48] [158100] global_step=158100, grad_norm=6.2408318519592285, loss=3.930741310119629 -I0513 15:36:04.266419 140330640848640 logging_writer.py:48] [158200] global_step=158200, grad_norm=4.5723466873168945, loss=3.706730842590332 -I0513 15:36:53.195552 140330632455936 logging_writer.py:48] [158300] global_step=158300, grad_norm=4.800233840942383, loss=3.834646701812744 -I0513 15:37:41.781814 140330640848640 logging_writer.py:48] [158400] global_step=158400, grad_norm=2.468862295150757, loss=3.8526554107666016 -I0513 15:38:29.333536 140330632455936 logging_writer.py:48] [158500] global_step=158500, grad_norm=7.5310797691345215, loss=3.9493961334228516 -I0513 15:39:17.076654 140330640848640 logging_writer.py:48] [158600] global_step=158600, grad_norm=5.399686813354492, loss=4.023742198944092 -I0513 15:40:04.553236 140330632455936 logging_writer.py:48] [158700] global_step=158700, grad_norm=4.4904656410217285, loss=3.919135808944702 -I0513 15:40:52.372441 140330640848640 logging_writer.py:48] [158800] global_step=158800, grad_norm=10.111827850341797, loss=3.810091257095337 -I0513 15:41:42.149687 140330632455936 logging_writer.py:48] [158900] global_step=158900, grad_norm=12.60224437713623, loss=3.848933696746826 -I0513 15:42:36.165048 140330640848640 logging_writer.py:48] [159000] global_step=159000, grad_norm=9.391827583312988, loss=4.004151344299316 -I0513 15:43:23.102290 140330632455936 logging_writer.py:48] [159100] global_step=159100, grad_norm=3.2428464889526367, loss=3.8558120727539062 -I0513 15:44:13.861261 140330640848640 logging_writer.py:48] [159200] global_step=159200, grad_norm=4.290992736816406, loss=3.842972755432129 -I0513 15:45:02.152390 140330632455936 logging_writer.py:48] [159300] global_step=159300, grad_norm=4.977232933044434, loss=3.802751064300537 -I0513 15:45:48.816347 140330640848640 logging_writer.py:48] [159400] global_step=159400, grad_norm=7.917830467224121, loss=3.6169185638427734 -I0513 15:46:39.600788 140330632455936 logging_writer.py:48] [159500] global_step=159500, grad_norm=6.192450046539307, loss=3.889695644378662 -I0513 15:47:26.229257 140330640848640 logging_writer.py:48] [159600] global_step=159600, grad_norm=10.224847793579102, loss=4.01845121383667 -I0513 15:48:13.155579 140330632455936 logging_writer.py:48] [159700] global_step=159700, grad_norm=6.392644882202148, loss=3.8596911430358887 -I0513 15:49:01.683457 140330640848640 logging_writer.py:48] [159800] global_step=159800, grad_norm=4.641001224517822, loss=3.9912939071655273 -I0513 15:49:50.250688 140330632455936 logging_writer.py:48] [159900] global_step=159900, grad_norm=5.837399959564209, loss=3.7533555030822754 -I0513 15:50:38.261828 140330640848640 logging_writer.py:48] [160000] global_step=160000, grad_norm=4.218524932861328, loss=3.7982335090637207 -I0513 15:51:26.007630 140330632455936 logging_writer.py:48] [160100] global_step=160100, grad_norm=2.6148219108581543, loss=3.7390379905700684 -I0513 15:52:18.549523 140330640848640 logging_writer.py:48] [160200] global_step=160200, grad_norm=3.596529245376587, loss=3.8617138862609863 -I0513 15:53:07.518945 140330632455936 logging_writer.py:48] [160300] global_step=160300, grad_norm=3.6457738876342773, loss=3.5917720794677734 -I0513 15:53:55.398031 140330640848640 logging_writer.py:48] [160400] global_step=160400, grad_norm=2.6256206035614014, loss=3.769011974334717 -I0513 15:54:43.241677 140330632455936 logging_writer.py:48] [160500] global_step=160500, grad_norm=14.687573432922363, loss=3.9166553020477295 -I0513 15:55:29.161185 140330640848640 logging_writer.py:48] [160600] global_step=160600, grad_norm=8.888832092285156, loss=3.624258279800415 -I0513 15:56:13.596119 140330632455936 logging_writer.py:48] [160700] global_step=160700, grad_norm=7.009731769561768, loss=3.967411756515503 -I0513 15:57:01.889801 140330640848640 logging_writer.py:48] [160800] global_step=160800, grad_norm=6.100178241729736, loss=3.9146838188171387 -I0513 15:57:18.761381 140546196993216 spec.py:333] Evaluating on the training split. -I0513 15:57:27.227093 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 15:57:59.377460 140546196993216 spec.py:363] Evaluating on the test split. -I0513 15:58:00.517997 140546196993216 submission_runner.py:516] Time since start: 62844.32s, Step: 160837, {'train/accuracy': Array(0.00139509, dtype=float32), 'train/loss': Array(9.382287, dtype=float32), 'validation/accuracy': Array(0.00154, dtype=float32), 'validation/loss': Array(9.348522, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(9.357764, dtype=float32), 'test/num_examples': 10000, 'score': 61935.455322265625, 'total_duration': 62844.32478642464, 'accumulated_submission_time': 61935.455322265625, 'accumulated_eval_time': 901.0739798545837, 'accumulated_logging_time': 4.68764328956604} -I0513 15:58:00.909642 140330632455936 logging_writer.py:48] [160837] accumulated_eval_time=901.074, accumulated_logging_time=4.68764, accumulated_submission_time=61935.5, global_step=160837, preemption_count=0, score=61935.5, test/accuracy=0.001500000013038516, test/loss=9.35776424407959, test/num_examples=10000, total_duration=62844.3, train/accuracy=0.0013950892025604844, train/loss=9.38228702545166, validation/accuracy=0.0015399999683722854, validation/loss=9.348522186279297, validation/num_examples=50000 -I0513 15:58:28.801178 140330640848640 logging_writer.py:48] [160900] global_step=160900, grad_norm=9.48507308959961, loss=3.8564271926879883 -I0513 15:59:13.454044 140330632455936 logging_writer.py:48] [161000] global_step=161000, grad_norm=5.414852142333984, loss=3.816882848739624 -I0513 16:00:01.301773 140330640848640 logging_writer.py:48] [161100] global_step=161100, grad_norm=32.928565979003906, loss=3.9093034267425537 -I0513 16:00:46.247533 140330632455936 logging_writer.py:48] [161200] global_step=161200, grad_norm=12.961155891418457, loss=3.788252353668213 -I0513 16:01:29.745965 140330640848640 logging_writer.py:48] [161300] global_step=161300, grad_norm=9.303775787353516, loss=3.701697826385498 -I0513 16:02:18.215026 140330632455936 logging_writer.py:48] [161400] global_step=161400, grad_norm=5.413610935211182, loss=3.755455493927002 -I0513 16:03:06.997489 140330640848640 logging_writer.py:48] [161500] global_step=161500, grad_norm=4.036447525024414, loss=3.7915122509002686 -I0513 16:03:54.363811 140330632455936 logging_writer.py:48] [161600] global_step=161600, grad_norm=5.795438289642334, loss=3.893468141555786 -I0513 16:04:42.350757 140330640848640 logging_writer.py:48] [161700] global_step=161700, grad_norm=3.7775707244873047, loss=3.9295907020568848 -I0513 16:05:28.856761 140330632455936 logging_writer.py:48] [161800] global_step=161800, grad_norm=8.024208068847656, loss=3.6629934310913086 -I0513 16:06:17.004428 140330640848640 logging_writer.py:48] [161900] global_step=161900, grad_norm=4.136982440948486, loss=3.7662277221679688 -I0513 16:07:05.743881 140330632455936 logging_writer.py:48] [162000] global_step=162000, grad_norm=5.7180328369140625, loss=3.860940456390381 -I0513 16:07:52.347765 140330640848640 logging_writer.py:48] [162100] global_step=162100, grad_norm=4.380660533905029, loss=3.7654688358306885 -I0513 16:08:39.564491 140330632455936 logging_writer.py:48] [162200] global_step=162200, grad_norm=2.7619121074676514, loss=3.8143720626831055 -I0513 16:09:49.084339 140330640848640 logging_writer.py:48] [162300] global_step=162300, grad_norm=6.58899450302124, loss=3.784156322479248 -I0513 16:10:33.032044 140330632455936 logging_writer.py:48] [162400] global_step=162400, grad_norm=3.136727809906006, loss=3.8828845024108887 -I0513 16:11:20.113165 140330640848640 logging_writer.py:48] [162500] global_step=162500, grad_norm=5.043025016784668, loss=3.833749771118164 -I0513 16:12:08.751455 140330632455936 logging_writer.py:48] [162600] global_step=162600, grad_norm=3.942906379699707, loss=3.6909422874450684 -I0513 16:12:58.079251 140330640848640 logging_writer.py:48] [162700] global_step=162700, grad_norm=6.2426910400390625, loss=3.737183094024658 -I0513 16:13:43.985899 140330632455936 logging_writer.py:48] [162800] global_step=162800, grad_norm=3.559417724609375, loss=3.8462023735046387 -I0513 16:14:33.978353 140330640848640 logging_writer.py:48] [162900] global_step=162900, grad_norm=3.352123737335205, loss=3.9334826469421387 -I0513 16:15:19.418357 140330632455936 logging_writer.py:48] [163000] global_step=163000, grad_norm=4.699305534362793, loss=4.022701263427734 -I0513 16:16:05.382137 140330640848640 logging_writer.py:48] [163100] global_step=163100, grad_norm=3.197995662689209, loss=3.821601152420044 -I0513 16:16:52.976105 140330632455936 logging_writer.py:48] [163200] global_step=163200, grad_norm=3.1136677265167236, loss=3.7315146923065186 -I0513 16:17:43.961818 140330640848640 logging_writer.py:48] [163300] global_step=163300, grad_norm=2.816382884979248, loss=3.6603634357452393 -I0513 16:18:37.145843 140330632455936 logging_writer.py:48] [163400] global_step=163400, grad_norm=3.992751121520996, loss=3.7463507652282715 -I0513 16:19:27.284739 140330640848640 logging_writer.py:48] [163500] global_step=163500, grad_norm=3.8040173053741455, loss=3.660216808319092 -I0513 16:20:13.030622 140330632455936 logging_writer.py:48] [163600] global_step=163600, grad_norm=7.340603351593018, loss=3.706406593322754 -I0513 16:21:00.280879 140330640848640 logging_writer.py:48] [163700] global_step=163700, grad_norm=14.617626190185547, loss=3.79779314994812 -I0513 16:21:48.207735 140330632455936 logging_writer.py:48] [163800] global_step=163800, grad_norm=2.075350284576416, loss=3.8294224739074707 -I0513 16:22:36.502352 140330640848640 logging_writer.py:48] [163900] global_step=163900, grad_norm=5.589822769165039, loss=4.0049147605896 -I0513 16:23:23.630176 140330632455936 logging_writer.py:48] [164000] global_step=164000, grad_norm=6.152740955352783, loss=3.8283979892730713 -I0513 16:24:23.625846 140330640848640 logging_writer.py:48] [164100] global_step=164100, grad_norm=2.5785505771636963, loss=3.6663684844970703 -I0513 16:25:07.583775 140330632455936 logging_writer.py:48] [164200] global_step=164200, grad_norm=6.494200229644775, loss=3.816770076751709 -I0513 16:25:52.837793 140330640848640 logging_writer.py:48] [164300] global_step=164300, grad_norm=3.73013973236084, loss=3.8629953861236572 -I0513 16:26:42.506839 140330632455936 logging_writer.py:48] [164400] global_step=164400, grad_norm=4.371313095092773, loss=3.780320644378662 -I0513 16:27:28.668999 140330640848640 logging_writer.py:48] [164500] global_step=164500, grad_norm=13.651944160461426, loss=3.9233407974243164 -I0513 16:28:16.746765 140330632455936 logging_writer.py:48] [164600] global_step=164600, grad_norm=6.033596992492676, loss=3.802554130554199 -I0513 16:29:04.581151 140330640848640 logging_writer.py:48] [164700] global_step=164700, grad_norm=7.616768836975098, loss=3.828150510787964 -I0513 16:29:50.225115 140330632455936 logging_writer.py:48] [164800] global_step=164800, grad_norm=6.935061454772949, loss=3.866687774658203 -I0513 16:30:35.205215 140330640848640 logging_writer.py:48] [164900] global_step=164900, grad_norm=4.721319198608398, loss=3.7539567947387695 -I0513 16:31:16.526071 140546196993216 spec.py:333] Evaluating on the training split. -I0513 16:31:25.269692 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 16:31:56.667938 140546196993216 spec.py:363] Evaluating on the test split. -I0513 16:31:57.819261 140546196993216 submission_runner.py:516] Time since start: 64881.60s, Step: 164985, {'train/accuracy': Array(0.00159439, dtype=float32), 'train/loss': Array(9.390181, dtype=float32), 'validation/accuracy': Array(0.00154, dtype=float32), 'validation/loss': Array(9.356932, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(9.366373, dtype=float32), 'test/num_examples': 10000, 'score': 63930.971826553345, 'total_duration': 64881.60176515579, 'accumulated_submission_time': 63930.971826553345, 'accumulated_eval_time': 942.0960912704468, 'accumulated_logging_time': 5.1383795738220215} -I0513 16:31:58.257967 140330632455936 logging_writer.py:48] [164985] accumulated_eval_time=942.096, accumulated_logging_time=5.13838, accumulated_submission_time=63931, global_step=164985, preemption_count=0, score=63931, test/accuracy=0.001500000013038516, test/loss=9.366373062133789, test/num_examples=10000, total_duration=64881.6, train/accuracy=0.0015943876933306456, train/loss=9.390180587768555, validation/accuracy=0.0015399999683722854, validation/loss=9.356931686401367, validation/num_examples=50000 -I0513 16:32:05.054936 140330640848640 logging_writer.py:48] [165000] global_step=165000, grad_norm=2.8817663192749023, loss=3.868962049484253 -I0513 16:32:50.835650 140330632455936 logging_writer.py:48] [165100] global_step=165100, grad_norm=3.5316550731658936, loss=3.6790289878845215 -I0513 16:33:37.899616 140330640848640 logging_writer.py:48] [165200] global_step=165200, grad_norm=5.255014419555664, loss=3.769261121749878 -I0513 16:34:28.560075 140330632455936 logging_writer.py:48] [165300] global_step=165300, grad_norm=8.49795913696289, loss=3.8375730514526367 -I0513 16:35:14.421175 140330640848640 logging_writer.py:48] [165400] global_step=165400, grad_norm=4.741466999053955, loss=3.8573174476623535 -I0513 16:36:00.771173 140330632455936 logging_writer.py:48] [165500] global_step=165500, grad_norm=8.722345352172852, loss=3.8135201930999756 -I0513 16:36:48.248277 140330640848640 logging_writer.py:48] [165600] global_step=165600, grad_norm=3.1732826232910156, loss=3.905071973800659 -I0513 16:37:36.238858 140330640848640 logging_writer.py:48] [165700] global_step=165700, grad_norm=4.945898532867432, loss=3.7662312984466553 -I0513 16:38:24.139210 140330632455936 logging_writer.py:48] [165800] global_step=165800, grad_norm=3.9893229007720947, loss=3.8304808139801025 -I0513 16:39:14.194977 140330640848640 logging_writer.py:48] [165900] global_step=165900, grad_norm=3.918534755706787, loss=3.7746078968048096 -I0513 16:40:12.319454 140330632455936 logging_writer.py:48] [166000] global_step=166000, grad_norm=3.4116222858428955, loss=3.7802181243896484 -I0513 16:41:06.425552 140330640848640 logging_writer.py:48] [166100] global_step=166100, grad_norm=3.8921239376068115, loss=3.7741050720214844 -I0513 16:41:53.985817 140330632455936 logging_writer.py:48] [166200] global_step=166200, grad_norm=3.8589890003204346, loss=3.7895212173461914 -I0513 16:42:43.567908 140330640848640 logging_writer.py:48] [166300] global_step=166300, grad_norm=6.338927745819092, loss=4.3401079177856445 -I0513 16:43:33.025405 140330632455936 logging_writer.py:48] [166400] global_step=166400, grad_norm=4.938511371612549, loss=3.896489381790161 -I0513 16:44:20.887402 140330640848640 logging_writer.py:48] [166500] global_step=166500, grad_norm=4.4624199867248535, loss=3.828437566757202 -I0513 16:45:26.171337 140330632455936 logging_writer.py:48] [166600] global_step=166600, grad_norm=4.206891059875488, loss=3.8421831130981445 -I0513 16:46:14.110895 140330640848640 logging_writer.py:48] [166700] global_step=166700, grad_norm=5.591477394104004, loss=3.8133087158203125 -I0513 16:46:59.337939 140330632455936 logging_writer.py:48] [166800] global_step=166800, grad_norm=3.0470263957977295, loss=3.8175251483917236 -I0513 16:47:41.395593 140330640848640 logging_writer.py:48] [166900] global_step=166900, grad_norm=3.4779295921325684, loss=3.7050747871398926 -I0513 16:48:26.422862 140330632455936 logging_writer.py:48] [167000] global_step=167000, grad_norm=5.627996444702148, loss=3.878571033477783 -I0513 16:49:13.686027 140330640848640 logging_writer.py:48] [167100] global_step=167100, grad_norm=23.44816780090332, loss=4.057926177978516 -I0513 16:50:00.132146 140330632455936 logging_writer.py:48] [167200] global_step=167200, grad_norm=9.976771354675293, loss=3.8266303539276123 -I0513 16:50:48.870432 140330640848640 logging_writer.py:48] [167300] global_step=167300, grad_norm=4.949581623077393, loss=3.8194499015808105 -I0513 16:51:37.494477 140330632455936 logging_writer.py:48] [167400] global_step=167400, grad_norm=4.2471795082092285, loss=3.745184898376465 -I0513 16:52:26.329995 140330640848640 logging_writer.py:48] [167500] global_step=167500, grad_norm=4.182222366333008, loss=3.7951602935791016 -I0513 16:53:13.796108 140330632455936 logging_writer.py:48] [167600] global_step=167600, grad_norm=6.384552001953125, loss=3.8764638900756836 -I0513 16:54:05.570096 140330640848640 logging_writer.py:48] [167700] global_step=167700, grad_norm=4.230560779571533, loss=3.8991079330444336 -I0513 16:54:55.102388 140330632455936 logging_writer.py:48] [167800] global_step=167800, grad_norm=4.318760395050049, loss=3.870553970336914 -I0513 16:55:43.860182 140330640848640 logging_writer.py:48] [167900] global_step=167900, grad_norm=3.667536497116089, loss=3.68312931060791 -I0513 16:56:35.468173 140330632455936 logging_writer.py:48] [168000] global_step=168000, grad_norm=6.562756061553955, loss=3.906938076019287 -I0513 16:57:23.316387 140330640848640 logging_writer.py:48] [168100] global_step=168100, grad_norm=2.53609561920166, loss=3.7194299697875977 -I0513 16:58:11.642863 140330632455936 logging_writer.py:48] [168200] global_step=168200, grad_norm=4.120571613311768, loss=3.752197504043579 -I0513 16:59:01.132735 140330640848640 logging_writer.py:48] [168300] global_step=168300, grad_norm=4.154031753540039, loss=3.7317943572998047 -I0513 16:59:50.105552 140330632455936 logging_writer.py:48] [168400] global_step=168400, grad_norm=2.8844056129455566, loss=3.753460645675659 -I0513 17:00:58.550048 140330640848640 logging_writer.py:48] [168500] global_step=168500, grad_norm=4.902185916900635, loss=3.674804210662842 -I0513 17:01:44.049768 140330632455936 logging_writer.py:48] [168600] global_step=168600, grad_norm=3.8304624557495117, loss=3.9786550998687744 -I0513 17:02:32.105732 140330640848640 logging_writer.py:48] [168700] global_step=168700, grad_norm=4.645920276641846, loss=3.654566526412964 -I0513 17:03:19.013349 140330632455936 logging_writer.py:48] [168800] global_step=168800, grad_norm=3.026641607284546, loss=3.8108716011047363 -I0513 17:04:09.094961 140330640848640 logging_writer.py:48] [168900] global_step=168900, grad_norm=4.193042278289795, loss=3.8335256576538086 -I0513 17:04:55.512871 140330632455936 logging_writer.py:48] [169000] global_step=169000, grad_norm=6.090036392211914, loss=3.770517587661743 -I0513 17:05:13.631514 140546196993216 spec.py:333] Evaluating on the training split. -I0513 17:05:23.630346 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 17:05:55.364922 140546196993216 spec.py:363] Evaluating on the test split. -I0513 17:05:56.382477 140546196993216 submission_runner.py:516] Time since start: 66920.26s, Step: 169039, {'train/accuracy': Array(0.0015346, dtype=float32), 'train/loss': Array(9.38863, dtype=float32), 'validation/accuracy': Array(0.0016, dtype=float32), 'validation/loss': Array(9.360721, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(9.372903, dtype=float32), 'test/num_examples': 10000, 'score': 65926.25481128693, 'total_duration': 66920.25790929794, 'accumulated_submission_time': 65926.25481128693, 'accumulated_eval_time': 984.6689186096191, 'accumulated_logging_time': 5.627253293991089} -I0513 17:05:56.797101 140330640848640 logging_writer.py:48] [169039] accumulated_eval_time=984.669, accumulated_logging_time=5.62725, accumulated_submission_time=65926.3, global_step=169039, preemption_count=0, score=65926.3, test/accuracy=0.001500000013038516, test/loss=9.372902870178223, test/num_examples=10000, total_duration=66920.3, train/accuracy=0.0015345981810241938, train/loss=9.388629913330078, validation/accuracy=0.0015999999595806003, validation/loss=9.36072063446045, validation/num_examples=50000 -I0513 17:06:24.497544 140330632455936 logging_writer.py:48] [169100] global_step=169100, grad_norm=4.401185989379883, loss=3.813405990600586 -I0513 17:07:10.292026 140330640848640 logging_writer.py:48] [169200] global_step=169200, grad_norm=4.476151466369629, loss=3.6735217571258545 -I0513 17:07:57.117961 140330632455936 logging_writer.py:48] [169300] global_step=169300, grad_norm=6.976400852203369, loss=3.9594013690948486 -I0513 17:08:45.469748 140330640848640 logging_writer.py:48] [169400] global_step=169400, grad_norm=4.407276153564453, loss=3.8219943046569824 -I0513 17:09:31.209799 140330632455936 logging_writer.py:48] [169500] global_step=169500, grad_norm=5.184383392333984, loss=4.018228054046631 -I0513 17:10:15.948413 140330640848640 logging_writer.py:48] [169600] global_step=169600, grad_norm=8.0792236328125, loss=3.8805811405181885 -I0513 17:11:01.668047 140330632455936 logging_writer.py:48] [169700] global_step=169700, grad_norm=3.4439404010772705, loss=3.764495849609375 -I0513 17:11:50.592837 140330640848640 logging_writer.py:48] [169800] global_step=169800, grad_norm=2.455962657928467, loss=3.634181022644043 -I0513 17:12:38.814867 140330632455936 logging_writer.py:48] [169900] global_step=169900, grad_norm=7.106099605560303, loss=3.7142786979675293 -I0513 17:13:27.171226 140330640848640 logging_writer.py:48] [170000] global_step=170000, grad_norm=3.2568485736846924, loss=3.6986215114593506 -I0513 17:14:36.702842 140330632455936 logging_writer.py:48] [170100] global_step=170100, grad_norm=5.084025859832764, loss=3.7134788036346436 -I0513 17:15:20.961330 140330640848640 logging_writer.py:48] [170200] global_step=170200, grad_norm=5.328367233276367, loss=3.821471691131592 -I0513 17:16:07.851045 140330632455936 logging_writer.py:48] [170300] global_step=170300, grad_norm=2.5380024909973145, loss=3.871640682220459 -I0513 17:16:57.991275 140330640848640 logging_writer.py:48] [170400] global_step=170400, grad_norm=3.5903141498565674, loss=3.864966869354248 -I0513 17:17:48.371958 140330632455936 logging_writer.py:48] [170500] global_step=170500, grad_norm=3.5469069480895996, loss=3.763190984725952 -I0513 17:18:34.594945 140330640848640 logging_writer.py:48] [170600] global_step=170600, grad_norm=5.476774215698242, loss=3.7902414798736572 -I0513 17:19:30.419986 140330632455936 logging_writer.py:48] [170700] global_step=170700, grad_norm=8.321218490600586, loss=3.716876745223999 -I0513 17:20:30.556852 140330640848640 logging_writer.py:48] [170800] global_step=170800, grad_norm=6.068613052368164, loss=3.768695116043091 -I0513 17:21:14.647802 140330632455936 logging_writer.py:48] [170900] global_step=170900, grad_norm=4.212215900421143, loss=3.899695873260498 -I0513 17:22:05.338914 140330640848640 logging_writer.py:48] [171000] global_step=171000, grad_norm=4.788431644439697, loss=3.7263882160186768 -I0513 17:22:54.065543 140330632455936 logging_writer.py:48] [171100] global_step=171100, grad_norm=3.493884325027466, loss=3.75062894821167 -I0513 17:23:43.778722 140330640848640 logging_writer.py:48] [171200] global_step=171200, grad_norm=2.898311138153076, loss=3.7881851196289062 -I0513 17:24:32.406650 140330632455936 logging_writer.py:48] [171300] global_step=171300, grad_norm=4.453056812286377, loss=3.7739064693450928 -I0513 17:25:23.255803 140330640848640 logging_writer.py:48] [171400] global_step=171400, grad_norm=3.987097978591919, loss=3.8540358543395996 -I0513 17:26:23.490485 140330632455936 logging_writer.py:48] [171500] global_step=171500, grad_norm=6.175843715667725, loss=3.6884186267852783 -I0513 17:27:09.472493 140330640848640 logging_writer.py:48] [171600] global_step=171600, grad_norm=13.986028671264648, loss=3.8641233444213867 -I0513 17:27:56.603164 140330632455936 logging_writer.py:48] [171700] global_step=171700, grad_norm=5.179997444152832, loss=3.8561058044433594 -I0513 17:28:43.938885 140330640848640 logging_writer.py:48] [171800] global_step=171800, grad_norm=4.156984329223633, loss=3.7561306953430176 -I0513 17:29:32.940533 140330632455936 logging_writer.py:48] [171900] global_step=171900, grad_norm=6.451966762542725, loss=3.836153507232666 -I0513 17:30:18.321105 140330640848640 logging_writer.py:48] [172000] global_step=172000, grad_norm=4.898999214172363, loss=3.766544818878174 -I0513 17:31:03.232771 140330632455936 logging_writer.py:48] [172100] global_step=172100, grad_norm=22.53387451171875, loss=3.852262258529663 -I0513 17:31:47.697299 140330640848640 logging_writer.py:48] [172200] global_step=172200, grad_norm=4.209978103637695, loss=3.8126909732818604 -I0513 17:32:31.459942 140330632455936 logging_writer.py:48] [172300] global_step=172300, grad_norm=2.4928014278411865, loss=3.727382183074951 -I0513 17:33:17.370033 140330640848640 logging_writer.py:48] [172400] global_step=172400, grad_norm=4.214897155761719, loss=3.8088326454162598 -I0513 17:34:06.104796 140330632455936 logging_writer.py:48] [172500] global_step=172500, grad_norm=3.80338978767395, loss=3.894094944000244 -I0513 17:34:53.992582 140330640848640 logging_writer.py:48] [172600] global_step=172600, grad_norm=3.8589463233947754, loss=3.788013458251953 -I0513 17:35:42.597764 140330632455936 logging_writer.py:48] [172700] global_step=172700, grad_norm=2.750561475753784, loss=3.8571224212646484 -I0513 17:36:29.608413 140330640848640 logging_writer.py:48] [172800] global_step=172800, grad_norm=6.754169464111328, loss=3.806910276412964 -I0513 17:37:18.694686 140330632455936 logging_writer.py:48] [172900] global_step=172900, grad_norm=2.9733850955963135, loss=3.787644863128662 -I0513 17:38:06.108795 140330640848640 logging_writer.py:48] [173000] global_step=173000, grad_norm=4.425631999969482, loss=3.8298556804656982 -I0513 17:38:54.111892 140330632455936 logging_writer.py:48] [173100] global_step=173100, grad_norm=3.751965284347534, loss=3.796031951904297 -I0513 17:39:13.094996 140546196993216 spec.py:333] Evaluating on the training split. -I0513 17:39:22.700820 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 17:39:53.032614 140546196993216 spec.py:363] Evaluating on the test split. -I0513 17:39:54.243906 140546196993216 submission_runner.py:516] Time since start: 68957.96s, Step: 173140, {'train/accuracy': Array(0.00123565, dtype=float32), 'train/loss': Array(9.44169, dtype=float32), 'validation/accuracy': Array(0.00156, dtype=float32), 'validation/loss': Array(9.386361, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(9.398773, dtype=float32), 'test/num_examples': 10000, 'score': 67922.46532917023, 'total_duration': 68957.95928025246, 'accumulated_submission_time': 67922.46532917023, 'accumulated_eval_time': 1025.4796116352081, 'accumulated_logging_time': 6.089812755584717} -I0513 17:39:54.599251 140330640848640 logging_writer.py:48] [173140] accumulated_eval_time=1025.48, accumulated_logging_time=6.08981, accumulated_submission_time=67922.5, global_step=173140, preemption_count=0, score=67922.5, test/accuracy=0.001500000013038516, test/loss=9.398773193359375, test/num_examples=10000, total_duration=68958, train/accuracy=0.001235650503076613, train/loss=9.441690444946289, validation/accuracy=0.001560000004246831, validation/loss=9.386361122131348, validation/num_examples=50000 -I0513 17:40:22.316167 140330632455936 logging_writer.py:48] [173200] global_step=173200, grad_norm=9.121914863586426, loss=3.8934316635131836 -I0513 17:41:08.039661 140330640848640 logging_writer.py:48] [173300] global_step=173300, grad_norm=4.1409454345703125, loss=3.7607810497283936 -I0513 17:41:53.451153 140330632455936 logging_writer.py:48] [173400] global_step=173400, grad_norm=5.6834025382995605, loss=3.7004265785217285 -I0513 17:42:38.305438 140330640848640 logging_writer.py:48] [173500] global_step=173500, grad_norm=5.464582443237305, loss=3.9960522651672363 -I0513 17:43:24.016302 140330632455936 logging_writer.py:48] [173600] global_step=173600, grad_norm=12.505391120910645, loss=3.799708604812622 -I0513 17:44:10.970787 140330640848640 logging_writer.py:48] [173700] global_step=173700, grad_norm=7.038081645965576, loss=3.844940662384033 -I0513 17:44:57.223985 140330632455936 logging_writer.py:48] [173800] global_step=173800, grad_norm=4.546360492706299, loss=3.8073580265045166 -I0513 17:45:43.153514 140330640848640 logging_writer.py:48] [173900] global_step=173900, grad_norm=63.09172439575195, loss=3.951075553894043 -I0513 17:46:32.219717 140330632455936 logging_writer.py:48] [174000] global_step=174000, grad_norm=5.939052581787109, loss=3.9618782997131348 -I0513 17:47:21.757565 140330640848640 logging_writer.py:48] [174100] global_step=174100, grad_norm=2.531963586807251, loss=3.83882212638855 -I0513 17:48:10.086665 140330632455936 logging_writer.py:48] [174200] global_step=174200, grad_norm=6.4054694175720215, loss=3.754302978515625 -I0513 17:48:58.339079 140330640848640 logging_writer.py:48] [174300] global_step=174300, grad_norm=5.541139125823975, loss=3.9109914302825928 -I0513 17:49:45.874229 140330632455936 logging_writer.py:48] [174400] global_step=174400, grad_norm=3.91288685798645, loss=3.704026222229004 -I0513 17:50:32.886604 140330640848640 logging_writer.py:48] [174500] global_step=174500, grad_norm=3.4187371730804443, loss=3.847479820251465 -I0513 17:51:21.738321 140330632455936 logging_writer.py:48] [174600] global_step=174600, grad_norm=3.7105233669281006, loss=3.7777907848358154 -I0513 17:52:06.852152 140330640848640 logging_writer.py:48] [174700] global_step=174700, grad_norm=3.03397798538208, loss=3.7939565181732178 -I0513 17:52:53.625320 140330632455936 logging_writer.py:48] [174800] global_step=174800, grad_norm=50.762062072753906, loss=4.060712814331055 -I0513 17:53:42.770452 140330640848640 logging_writer.py:48] [174900] global_step=174900, grad_norm=3.599175214767456, loss=3.659116506576538 -I0513 17:54:30.355761 140330632455936 logging_writer.py:48] [175000] global_step=175000, grad_norm=10.934308052062988, loss=3.9261436462402344 -I0513 17:55:17.621280 140330640848640 logging_writer.py:48] [175100] global_step=175100, grad_norm=3.1374034881591797, loss=3.842270851135254 -I0513 17:56:07.652477 140330632455936 logging_writer.py:48] [175200] global_step=175200, grad_norm=10.226550102233887, loss=3.753103494644165 -I0513 17:56:55.234270 140330640848640 logging_writer.py:48] [175300] global_step=175300, grad_norm=13.348221778869629, loss=3.8291468620300293 -I0513 17:57:43.028816 140330632455936 logging_writer.py:48] [175400] global_step=175400, grad_norm=7.144501686096191, loss=3.990003824234009 -I0513 17:58:30.413166 140330640848640 logging_writer.py:48] [175500] global_step=175500, grad_norm=7.457942008972168, loss=3.821526527404785 -I0513 17:59:17.959178 140330632455936 logging_writer.py:48] [175600] global_step=175600, grad_norm=4.315648078918457, loss=3.8346190452575684 -I0513 18:00:05.544751 140330640848640 logging_writer.py:48] [175700] global_step=175700, grad_norm=3.597656726837158, loss=3.7941794395446777 -I0513 18:00:52.212580 140330632455936 logging_writer.py:48] [175800] global_step=175800, grad_norm=6.780549049377441, loss=3.9405641555786133 -I0513 18:01:40.470204 140330640848640 logging_writer.py:48] [175900] global_step=175900, grad_norm=4.5283331871032715, loss=3.8013558387756348 -I0513 18:02:26.052589 140330632455936 logging_writer.py:48] [176000] global_step=176000, grad_norm=4.923598766326904, loss=3.771397590637207 -I0513 18:03:14.330182 140330640848640 logging_writer.py:48] [176100] global_step=176100, grad_norm=6.9161272048950195, loss=3.831939697265625 -I0513 18:04:01.573351 140330632455936 logging_writer.py:48] [176200] global_step=176200, grad_norm=12.30150032043457, loss=3.7492990493774414 -I0513 18:04:47.441688 140330640848640 logging_writer.py:48] [176300] global_step=176300, grad_norm=4.6831583976745605, loss=3.7265751361846924 -I0513 18:05:36.266370 140330632455936 logging_writer.py:48] [176400] global_step=176400, grad_norm=4.77200174331665, loss=3.848798990249634 -I0513 18:06:23.618226 140330640848640 logging_writer.py:48] [176500] global_step=176500, grad_norm=3.4377663135528564, loss=3.822870969772339 -I0513 18:07:10.733516 140330632455936 logging_writer.py:48] [176600] global_step=176600, grad_norm=4.711678504943848, loss=3.853363037109375 -I0513 18:08:03.571703 140330640848640 logging_writer.py:48] [176700] global_step=176700, grad_norm=17.058340072631836, loss=3.7871062755584717 -I0513 18:09:04.148707 140330632455936 logging_writer.py:48] [176800] global_step=176800, grad_norm=3.234631299972534, loss=3.8516926765441895 -I0513 18:09:49.879626 140330640848640 logging_writer.py:48] [176900] global_step=176900, grad_norm=3.7220077514648438, loss=3.817111015319824 -I0513 18:10:35.074086 140330632455936 logging_writer.py:48] [177000] global_step=177000, grad_norm=3.3779304027557373, loss=3.6953158378601074 -I0513 18:11:22.318048 140330640848640 logging_writer.py:48] [177100] global_step=177100, grad_norm=3.7908895015716553, loss=3.8298091888427734 -I0513 18:12:09.433949 140330632455936 logging_writer.py:48] [177200] global_step=177200, grad_norm=2.120241165161133, loss=3.6911661624908447 -I0513 18:12:54.631733 140330640848640 logging_writer.py:48] [177300] global_step=177300, grad_norm=3.0206151008605957, loss=3.8426513671875 -I0513 18:13:10.493425 140546196993216 spec.py:333] Evaluating on the training split. -I0513 18:13:18.541137 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 18:13:50.719417 140546196993216 spec.py:363] Evaluating on the test split. -I0513 18:13:51.868622 140546196993216 submission_runner.py:516] Time since start: 70995.67s, Step: 177337, {'train/accuracy': Array(0.00149474, dtype=float32), 'train/loss': Array(9.390521, dtype=float32), 'validation/accuracy': Array(0.0017, dtype=float32), 'validation/loss': Array(9.368401, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(9.381394, dtype=float32), 'test/num_examples': 10000, 'score': 69918.275775671, 'total_duration': 70995.66643381119, 'accumulated_submission_time': 69918.275775671, 'accumulated_eval_time': 1066.5990326404572, 'accumulated_logging_time': 6.489320278167725} -I0513 18:13:52.250320 140330632455936 logging_writer.py:48] [177337] accumulated_eval_time=1066.6, accumulated_logging_time=6.48932, accumulated_submission_time=69918.3, global_step=177337, preemption_count=0, score=69918.3, test/accuracy=0.001500000013038516, test/loss=9.381394386291504, test/num_examples=10000, total_duration=70995.7, train/accuracy=0.001494738506153226, train/loss=9.390521049499512, validation/accuracy=0.0016999999061226845, validation/loss=9.368400573730469, validation/num_examples=50000 -I0513 18:14:16.553914 140330640848640 logging_writer.py:48] [177400] global_step=177400, grad_norm=3.4462966918945312, loss=3.8149242401123047 -I0513 18:15:01.796920 140330632455936 logging_writer.py:48] [177500] global_step=177500, grad_norm=2.801884651184082, loss=3.6784846782684326 -I0513 18:15:46.928255 140330640848640 logging_writer.py:48] [177600] global_step=177600, grad_norm=3.7320804595947266, loss=3.831608533859253 -I0513 18:16:36.485517 140330632455936 logging_writer.py:48] [177700] global_step=177700, grad_norm=2.8259358406066895, loss=3.7889299392700195 -I0513 18:17:24.258096 140330640848640 logging_writer.py:48] [177800] global_step=177800, grad_norm=10.393410682678223, loss=3.693122148513794 -I0513 18:18:13.589771 140330632455936 logging_writer.py:48] [177900] global_step=177900, grad_norm=4.267276287078857, loss=3.8128867149353027 -I0513 18:19:02.127458 140330640848640 logging_writer.py:48] [178000] global_step=178000, grad_norm=10.409967422485352, loss=3.914790391921997 -I0513 18:19:50.303731 140330632455936 logging_writer.py:48] [178100] global_step=178100, grad_norm=3.486240863800049, loss=3.795346736907959 -I0513 18:20:40.521890 140330640848640 logging_writer.py:48] [178200] global_step=178200, grad_norm=6.665155410766602, loss=3.825798749923706 -I0513 18:21:29.743016 140330632455936 logging_writer.py:48] [178300] global_step=178300, grad_norm=3.076569080352783, loss=3.759843349456787 -I0513 18:22:16.858437 140330640848640 logging_writer.py:48] [178400] global_step=178400, grad_norm=11.86605453491211, loss=3.6389055252075195 -I0513 18:23:05.551725 140330632455936 logging_writer.py:48] [178500] global_step=178500, grad_norm=7.75984001159668, loss=4.072914123535156 -I0513 18:23:54.676443 140330640848640 logging_writer.py:48] [178600] global_step=178600, grad_norm=26.591285705566406, loss=3.924654483795166 -I0513 18:24:46.031514 140330632455936 logging_writer.py:48] [178700] global_step=178700, grad_norm=5.452311038970947, loss=3.7823545932769775 -I0513 18:25:35.292399 140330640848640 logging_writer.py:48] [178800] global_step=178800, grad_norm=3.829418420791626, loss=3.821162223815918 -I0513 18:26:25.678038 140330632455936 logging_writer.py:48] [178900] global_step=178900, grad_norm=3.404630184173584, loss=3.6761302947998047 -I0513 18:27:13.882343 140330632455936 logging_writer.py:48] [179000] global_step=179000, grad_norm=5.052337646484375, loss=3.953644275665283 -I0513 18:28:04.863869 140330640848640 logging_writer.py:48] [179100] global_step=179100, grad_norm=5.920718669891357, loss=3.918936014175415 -I0513 18:28:54.373010 140330632455936 logging_writer.py:48] [179200] global_step=179200, grad_norm=6.689146995544434, loss=3.79925537109375 -I0513 18:29:42.003672 140330640848640 logging_writer.py:48] [179300] global_step=179300, grad_norm=4.677877426147461, loss=3.741456985473633 -I0513 18:30:31.740455 140330632455936 logging_writer.py:48] [179400] global_step=179400, grad_norm=4.424162864685059, loss=3.9351089000701904 -I0513 18:31:21.234042 140330640848640 logging_writer.py:48] [179500] global_step=179500, grad_norm=5.863819122314453, loss=3.7922301292419434 -I0513 18:32:09.161639 140330632455936 logging_writer.py:48] [179600] global_step=179600, grad_norm=6.657369613647461, loss=3.837399482727051 -I0513 18:32:56.114272 140330640848640 logging_writer.py:48] [179700] global_step=179700, grad_norm=5.544375896453857, loss=3.7151975631713867 -I0513 18:33:45.081086 140330632455936 logging_writer.py:48] [179800] global_step=179800, grad_norm=9.340442657470703, loss=3.933922290802002 -I0513 18:34:35.643649 140330640848640 logging_writer.py:48] [179900] global_step=179900, grad_norm=5.60516357421875, loss=3.7072689533233643 -I0513 18:35:27.241586 140330632455936 logging_writer.py:48] [180000] global_step=180000, grad_norm=9.3582181930542, loss=3.827061414718628 -I0513 18:36:23.380666 140330640848640 logging_writer.py:48] [180100] global_step=180100, grad_norm=3.045610189437866, loss=3.6531810760498047 -I0513 18:37:12.437749 140330632455936 logging_writer.py:48] [180200] global_step=180200, grad_norm=3.6660714149475098, loss=3.716721296310425 -I0513 18:38:07.603085 140330640848640 logging_writer.py:48] [180300] global_step=180300, grad_norm=6.116247177124023, loss=3.832632541656494 -I0513 18:39:05.366819 140330632455936 logging_writer.py:48] [180400] global_step=180400, grad_norm=22.15658950805664, loss=3.7939486503601074 -I0513 18:39:56.631591 140330640848640 logging_writer.py:48] [180500] global_step=180500, grad_norm=4.3683319091796875, loss=3.6504878997802734 -I0513 18:40:56.715330 140330632455936 logging_writer.py:48] [180600] global_step=180600, grad_norm=4.008203029632568, loss=3.7099881172180176 -I0513 18:41:41.136671 140330640848640 logging_writer.py:48] [180700] global_step=180700, grad_norm=7.203024387359619, loss=3.8491287231445312 -I0513 18:42:26.337896 140330632455936 logging_writer.py:48] [180800] global_step=180800, grad_norm=6.107968330383301, loss=3.678436517715454 -I0513 18:43:13.202454 140330640848640 logging_writer.py:48] [180900] global_step=180900, grad_norm=11.55007266998291, loss=3.731461763381958 -I0513 18:43:59.107377 140330632455936 logging_writer.py:48] [181000] global_step=181000, grad_norm=3.6507937908172607, loss=3.6979310512542725 -I0513 18:44:46.485055 140330640848640 logging_writer.py:48] [181100] global_step=181100, grad_norm=5.409557819366455, loss=3.858147621154785 -I0513 18:45:34.430448 140330632455936 logging_writer.py:48] [181200] global_step=181200, grad_norm=4.303968906402588, loss=3.8596127033233643 -I0513 18:46:22.373203 140330640848640 logging_writer.py:48] [181300] global_step=181300, grad_norm=8.622206687927246, loss=3.724776268005371 -I0513 18:47:07.833595 140546196993216 spec.py:333] Evaluating on the training split. -I0513 18:47:16.565487 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 18:47:48.915986 140546196993216 spec.py:363] Evaluating on the test split. -I0513 18:47:50.052252 140546196993216 submission_runner.py:516] Time since start: 73033.81s, Step: 181398, {'train/accuracy': Array(0.0013353, dtype=float32), 'train/loss': Array(9.425164, dtype=float32), 'validation/accuracy': Array(0.0017, dtype=float32), 'validation/loss': Array(9.390179, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0015, dtype=float32), 'test/loss': Array(9.403217, dtype=float32), 'test/num_examples': 10000, 'score': 71913.75967979431, 'total_duration': 73033.81380438805, 'accumulated_submission_time': 71913.75967979431, 'accumulated_eval_time': 1108.525652885437, 'accumulated_logging_time': 6.931232213973999} -I0513 18:47:50.396345 140330632455936 logging_writer.py:48] [181398] accumulated_eval_time=1108.53, accumulated_logging_time=6.93123, accumulated_submission_time=71913.8, global_step=181398, preemption_count=0, score=71913.8, test/accuracy=0.001500000013038516, test/loss=9.403217315673828, test/num_examples=10000, total_duration=73033.8, train/accuracy=0.0013352996902540326, train/loss=9.425164222717285, validation/accuracy=0.0016999999061226845, validation/loss=9.390178680419922, validation/num_examples=50000 -I0513 18:47:51.772410 140330640848640 logging_writer.py:48] [181400] global_step=181400, grad_norm=4.451914310455322, loss=3.8184468746185303 -I0513 18:48:36.625463 140330632455936 logging_writer.py:48] [181500] global_step=181500, grad_norm=6.602558135986328, loss=3.7349209785461426 -I0513 18:49:21.223402 140330640848640 logging_writer.py:48] [181600] global_step=181600, grad_norm=19.92081069946289, loss=3.7950711250305176 -I0513 18:50:08.627836 140330632455936 logging_writer.py:48] [181700] global_step=181700, grad_norm=3.1505606174468994, loss=3.7838587760925293 -I0513 18:50:57.861531 140330640848640 logging_writer.py:48] [181800] global_step=181800, grad_norm=2.2375082969665527, loss=3.668553590774536 -I0513 18:51:47.374923 140330632455936 logging_writer.py:48] [181900] global_step=181900, grad_norm=5.150576591491699, loss=3.8722801208496094 -I0513 18:52:35.935404 140330640848640 logging_writer.py:48] [182000] global_step=182000, grad_norm=4.788939476013184, loss=3.6704773902893066 -I0513 18:53:23.261204 140330632455936 logging_writer.py:48] [182100] global_step=182100, grad_norm=3.4288129806518555, loss=3.7717957496643066 -I0513 18:54:10.100581 140330640848640 logging_writer.py:48] [182200] global_step=182200, grad_norm=4.052312850952148, loss=3.940524101257324 -I0513 18:54:57.402179 140330632455936 logging_writer.py:48] [182300] global_step=182300, grad_norm=4.859977722167969, loss=3.7823495864868164 -I0513 18:55:45.159395 140330640848640 logging_writer.py:48] [182400] global_step=182400, grad_norm=7.318413734436035, loss=3.8677406311035156 -I0513 18:56:29.458030 140330632455936 logging_writer.py:48] [182500] global_step=182500, grad_norm=3.247648239135742, loss=3.7229154109954834 -I0513 18:57:14.231504 140330640848640 logging_writer.py:48] [182600] global_step=182600, grad_norm=3.1097772121429443, loss=3.7150039672851562 -I0513 18:57:58.667768 140330632455936 logging_writer.py:48] [182700] global_step=182700, grad_norm=5.128710746765137, loss=3.8141982555389404 -I0513 18:58:45.156695 140330640848640 logging_writer.py:48] [182800] global_step=182800, grad_norm=9.308672904968262, loss=3.7156949043273926 -I0513 18:59:33.184405 140330632455936 logging_writer.py:48] [182900] global_step=182900, grad_norm=3.914994478225708, loss=3.848404884338379 -I0513 19:00:25.343823 140330640848640 logging_writer.py:48] [183000] global_step=183000, grad_norm=4.728837013244629, loss=3.8135569095611572 -I0513 19:01:27.295558 140330632455936 logging_writer.py:48] [183100] global_step=183100, grad_norm=3.8737833499908447, loss=3.7999496459960938 -I0513 19:02:10.655515 140330640848640 logging_writer.py:48] [183200] global_step=183200, grad_norm=3.3406007289886475, loss=3.8564138412475586 -I0513 19:02:57.608671 140330632455936 logging_writer.py:48] [183300] global_step=183300, grad_norm=4.869384765625, loss=3.827547073364258 -I0513 19:03:55.409813 140330640848640 logging_writer.py:48] [183400] global_step=183400, grad_norm=3.006131172180176, loss=3.6441726684570312 -I0513 19:04:40.152919 140330632455936 logging_writer.py:48] [183500] global_step=183500, grad_norm=3.0475783348083496, loss=3.669548273086548 -I0513 19:05:24.594895 140330640848640 logging_writer.py:48] [183600] global_step=183600, grad_norm=4.890719890594482, loss=3.9064273834228516 -I0513 19:06:13.542520 140330632455936 logging_writer.py:48] [183700] global_step=183700, grad_norm=5.7796854972839355, loss=3.900893211364746 -I0513 19:07:16.374402 140330640848640 logging_writer.py:48] [183800] global_step=183800, grad_norm=5.312910079956055, loss=3.8191723823547363 -I0513 19:08:01.466617 140330632455936 logging_writer.py:48] [183900] global_step=183900, grad_norm=2.2104780673980713, loss=3.58917236328125 -I0513 19:08:47.103137 140330640848640 logging_writer.py:48] [184000] global_step=184000, grad_norm=3.0383903980255127, loss=3.790130138397217 -I0513 19:09:31.849222 140330632455936 logging_writer.py:48] [184100] global_step=184100, grad_norm=7.502665996551514, loss=3.777385711669922 -I0513 19:10:19.281360 140330640848640 logging_writer.py:48] [184200] global_step=184200, grad_norm=5.701760768890381, loss=4.097142696380615 -I0513 19:11:20.857137 140330632455936 logging_writer.py:48] [184300] global_step=184300, grad_norm=3.3446156978607178, loss=3.818999767303467 -I0513 19:12:12.228069 140330640848640 logging_writer.py:48] [184400] global_step=184400, grad_norm=6.162267684936523, loss=3.8467674255371094 -I0513 19:12:56.692720 140330632455936 logging_writer.py:48] [184500] global_step=184500, grad_norm=3.423668146133423, loss=3.6349925994873047 -I0513 19:13:41.040556 140330640848640 logging_writer.py:48] [184600] global_step=184600, grad_norm=3.74397873878479, loss=3.759003162384033 -I0513 19:14:27.480320 140330632455936 logging_writer.py:48] [184700] global_step=184700, grad_norm=4.541653633117676, loss=3.8524115085601807 -I0513 19:15:12.655530 140330640848640 logging_writer.py:48] [184800] global_step=184800, grad_norm=20.175745010375977, loss=3.9131007194519043 -I0513 19:16:03.228490 140330632455936 logging_writer.py:48] [184900] global_step=184900, grad_norm=7.603712558746338, loss=3.8262009620666504 -I0513 19:16:47.524857 140330640848640 logging_writer.py:48] [185000] global_step=185000, grad_norm=3.2753520011901855, loss=3.7667479515075684 -I0513 19:17:32.593462 140330632455936 logging_writer.py:48] [185100] global_step=185100, grad_norm=8.906726837158203, loss=3.793959617614746 -I0513 19:18:17.868408 140330640848640 logging_writer.py:48] [185200] global_step=185200, grad_norm=9.632233619689941, loss=3.8845887184143066 -I0513 19:19:03.586950 140330632455936 logging_writer.py:48] [185300] global_step=185300, grad_norm=1.8322484493255615, loss=3.678419589996338 -I0513 19:19:50.964303 140330640848640 logging_writer.py:48] [185400] global_step=185400, grad_norm=4.365950584411621, loss=3.886353015899658 -I0513 19:20:35.857127 140330632455936 logging_writer.py:48] [185500] global_step=185500, grad_norm=3.660421371459961, loss=3.734647035598755 -I0513 19:21:05.786114 140546196993216 spec.py:333] Evaluating on the training split. -I0513 19:21:13.902265 140546196993216 spec.py:346] Evaluating on the validation split. -I0513 19:21:42.769754 140546196993216 spec.py:363] Evaluating on the test split. -I0513 19:21:43.757784 140546196993216 submission_runner.py:516] Time since start: 75067.67s, Step: 185561, {'train/accuracy': Array(0.00147481, dtype=float32), 'train/loss': Array(9.436946, dtype=float32), 'validation/accuracy': Array(0.00168, dtype=float32), 'validation/loss': Array(9.422385, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0017, dtype=float32), 'test/loss': Array(9.43556, dtype=float32), 'test/num_examples': 10000, 'score': 73909.05753827095, 'total_duration': 75067.66927337646, 'accumulated_submission_time': 73909.05753827095, 'accumulated_eval_time': 1146.3552420139313, 'accumulated_logging_time': 7.326283931732178} -I0513 19:21:44.092549 140330640848640 logging_writer.py:48] [185561] accumulated_eval_time=1146.36, accumulated_logging_time=7.32628, accumulated_submission_time=73909.1, global_step=185561, preemption_count=0, score=73909.1, test/accuracy=0.0017000001389533281, test/loss=9.43556022644043, test/num_examples=10000, total_duration=75067.7, train/accuracy=0.001474808668717742, train/loss=9.436945915222168, validation/accuracy=0.0016799999866634607, validation/loss=9.422385215759277, validation/num_examples=50000 -I0513 19:22:01.085949 140330632455936 logging_writer.py:48] [185600] global_step=185600, grad_norm=5.798754692077637, loss=3.761122226715088 -I0513 19:22:45.741697 140330640848640 logging_writer.py:48] [185700] global_step=185700, grad_norm=3.8410375118255615, loss=3.820404529571533 -I0513 19:23:33.778144 140330632455936 logging_writer.py:48] [185800] global_step=185800, grad_norm=4.26568603515625, loss=3.661353588104248 -I0513 19:24:42.441950 140330640848640 logging_writer.py:48] [185900] global_step=185900, grad_norm=4.474493026733398, loss=3.796161651611328 -I0513 19:25:27.986922 140330632455936 logging_writer.py:48] [186000] global_step=186000, grad_norm=11.974518775939941, loss=3.864711046218872 -I0513 19:26:12.465941 140330640848640 logging_writer.py:48] [186100] global_step=186100, grad_norm=15.058660507202148, loss=3.8051199913024902 -I0513 19:26:59.666205 140330632455936 logging_writer.py:48] [186200] global_step=186200, grad_norm=3.4640109539031982, loss=3.796034336090088 -I0513 19:27:48.464459 140330640848640 logging_writer.py:48] [186300] global_step=186300, grad_norm=3.7255899906158447, loss=3.7767324447631836 -I0513 19:28:35.266659 140330632455936 logging_writer.py:48] [186400] global_step=186400, grad_norm=2.6054844856262207, loss=3.8534555435180664 -I0513 19:29:22.384791 140330640848640 logging_writer.py:48] [186500] global_step=186500, grad_norm=2.7246367931365967, loss=3.8030779361724854 -I0513 19:30:08.986281 140330632455936 logging_writer.py:48] [186600] global_step=186600, grad_norm=17.85517692565918, loss=3.8306803703308105 -I0513 19:30:55.407007 140330640848640 logging_writer.py:48] [186700] global_step=186700, grad_norm=4.526782512664795, loss=3.67397403717041 -I0513 19:31:42.908051 140330632455936 logging_writer.py:48] [186800] global_step=186800, grad_norm=3.969761371612549, loss=3.7109460830688477 -I0513 19:32:45.373576 140330640848640 logging_writer.py:48] [186900] global_step=186900, grad_norm=4.81820821762085, loss=3.600381851196289 -I0513 19:33:28.690441 140330632455936 logging_writer.py:48] [187000] global_step=187000, grad_norm=6.288241386413574, loss=3.702303647994995 -I0513 19:34:12.609128 140330640848640 logging_writer.py:48] [187100] global_step=187100, grad_norm=60.914363861083984, loss=3.9375295639038086 -I0513 19:34:58.200030 140330632455936 logging_writer.py:48] [187200] global_step=187200, grad_norm=3.9198246002197266, loss=3.828490972518921 -I0513 19:36:02.921286 140330640848640 logging_writer.py:48] [187300] global_step=187300, grad_norm=3.72769832611084, loss=3.8319478034973145 -I0513 19:36:47.535980 140330632455936 logging_writer.py:48] [187400] global_step=187400, grad_norm=11.482843399047852, loss=3.72928524017334 -I0513 19:37:33.032524 140330640848640 logging_writer.py:48] [187500] global_step=187500, grad_norm=4.434323787689209, loss=3.8703768253326416 -I0513 19:38:19.644137 140330632455936 logging_writer.py:48] [187600] global_step=187600, grad_norm=5.75684118270874, loss=3.7401342391967773 -I0513 19:39:06.571174 140330640848640 logging_writer.py:48] [187700] global_step=187700, grad_norm=4.182270526885986, loss=3.714305877685547 -I0513 19:40:11.546821 140330632455936 logging_writer.py:48] [187800] global_step=187800, grad_norm=6.96590518951416, loss=3.777968168258667 -I0513 19:40:59.365110 140330640848640 logging_writer.py:48] [187900] global_step=187900, grad_norm=4.1128249168396, loss=3.7807397842407227 -I0513 19:41:44.548927 140330632455936 logging_writer.py:48] [188000] global_step=188000, grad_norm=6.200401306152344, loss=3.9138073921203613 -I0513 19:42:30.562572 140330640848640 logging_writer.py:48] [188100] global_step=188100, grad_norm=5.002246856689453, loss=3.769448757171631 -I0513 19:43:14.475728 140330632455936 logging_writer.py:48] [188200] global_step=188200, grad_norm=3.162055015563965, loss=3.7219972610473633 -I0513 19:43:59.029791 140330640848640 logging_writer.py:48] [188300] global_step=188300, grad_norm=4.309389591217041, loss=3.864147663116455 -I0513 19:44:44.367692 140330632455936 logging_writer.py:48] [188400] global_step=188400, grad_norm=3.952334403991699, loss=3.8093295097351074 -I0513 19:45:28.872071 140330640848640 logging_writer.py:48] [188500] global_step=188500, grad_norm=3.5791008472442627, loss=3.6218738555908203 -I0513 19:46:12.964443 140330632455936 logging_writer.py:48] [188600] global_step=188600, grad_norm=13.27508544921875, loss=3.7232189178466797 -I0513 19:46:59.005333 140330640848640 logging_writer.py:48] [188700] global_step=188700, grad_norm=2.8682544231414795, loss=3.60642409324646 -I0513 19:47:43.952053 140330632455936 logging_writer.py:48] [188800] global_step=188800, grad_norm=4.375033378601074, loss=3.7182931900024414 -I0513 19:48:29.781509 140330640848640 logging_writer.py:48] [188900] global_step=188900, grad_norm=8.633360862731934, loss=3.781147003173828 -I0513 19:49:18.449068 140330632455936 logging_writer.py:48] [189000] global_step=189000, grad_norm=5.240957736968994, loss=3.7180373668670654 -I0513 19:50:14.773686 140330640848640 logging_writer.py:48] [189100] global_step=189100, grad_norm=21.551528930664062, loss=3.8371636867523193 -I0513 19:51:07.464441 140330632455936 logging_writer.py:48] [189200] global_step=189200, grad_norm=13.451132774353027, loss=3.8281867504119873 -I0513 19:51:50.055149 140330640848640 logging_writer.py:48] [189300] global_step=189300, grad_norm=4.374985218048096, loss=3.7372753620147705 -I0513 19:52:35.831780 140330632455936 logging_writer.py:48] [189400] global_step=189400, grad_norm=13.766833305358887, loss=3.852825164794922 -I0513 19:53:20.713396 140330640848640 logging_writer.py:48] [189500] global_step=189500, grad_norm=11.595244407653809, loss=3.8546676635742188 -I0513 19:54:06.471761 140330632455936 logging_writer.py:48] [189600] global_step=189600, grad_norm=3.7334601879119873, loss=3.7801430225372314 -I0513 19:54:51.413628 140330640848640 logging_writer.py:48] [189700] global_step=189700, grad_norm=3.7160935401916504, loss=3.740877151489258 -I0513 19:55:00.425554 140330632455936 logging_writer.py:48] [189721] global_step=189721, preemption_count=0, score=75905.2 -I0513 19:55:01.207609 140546196993216 submission_runner.py:857] Final imagenet_resnet score: 75905.17727708817 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-09-17-19.log b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-09-17-19.log new file mode 100644 index 000000000..beab0f02d --- /dev/null +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/imagenet_resnet_jax_08-31-2026-09-17-19.log @@ -0,0 +1,2564 @@ +python submission_runner.py --framework=jax --workload=imagenet_resnet --submission_path=submissions_algorithms/submissions/self_tuning/schedule_free_adamw_jax_v2_bn_fix/submission.py --data_dir=/data/imagenet/jax --experiment_dir=/experiment_runs --experiment_name=submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_2 --overwrite=True --save_checkpoints=False --rng_seed=-939886995 --imagenet_v2_data_dir=/data/imagenet/jax --tuning_ruleset=self 2>&1 | tee -a /logs/imagenet_resnet_jax_08-31-2026-09-17-19.log +2026-08-31 09:17:20.321782: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:467] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered +WARNING: All log messages before absl::InitializeLog() is called are written to STDERR +E0000 00:00:1788167840.344362 13 cuda_dnn.cc:8579] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered +E0000 00:00:1788167840.351810 13 cuda_blas.cc:1407] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered +W0000 00:00:1788167840.369739 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +W0000 00:00:1788167840.369764 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +W0000 00:00:1788167840.369766 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +W0000 00:00:1788167840.369768 13 computation_placer.cc:177] computation placer already registered. Please check linkage and avoid linking the same target more than once. +/usr/local/lib/python3.11/site-packages/torch/__init__.py:1617: UserWarning: Please use the new API settings to control TF32 behavior, such as torch.backends.cudnn.conv.fp32_precision = 'tf32' or torch.backends.cuda.matmul.fp32_precision = 'ieee'. Old settings, e.g, torch.backends.cuda.matmul.allow_tf32 = True, torch.backends.cudnn.allow_tf32 = True, allowTF32CuDNN() and allowTF32CuBLAS() will be deprecated after Pytorch 2.9. Please see https://pytorch.org/docs/main/notes/cuda.html#tensorfloat-32-tf32-on-ampere-and-later-devices (Triggered internally at /pytorch/aten/src/ATen/Context.cpp:80.) + _C._set_float32_matmul_precision(precision) +/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'repr' attribute with value False was provided to the `Field()` function, which has no effect in the context it was used. 'repr' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. + warnings.warn( +/usr/local/lib/python3.11/site-packages/pydantic/_internal/_generate_schema.py:2249: UnsupportedFieldAttributeWarning: The 'frozen' attribute with value True was provided to the `Field()` function, which has no effect in the context it was used. 'frozen' is field-specific metadata, and can only be attached to a model field using `Annotated` metadata or by assignment. This may have happened because an `Annotated` type alias using the `type` statement was used, or if the `Field()` function was attached to a single member of a union type. + warnings.warn( +INFO:2026-08-31 09:17:30,316:jax._src.xla_bridge:830: Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory +I0831 09:17:30.316229 140117123622080 xla_bridge.py:830] Unable to initialize backend 'tpu': INTERNAL: Failed to open libtpu.so: libtpu.so: cannot open shared object file: No such file or directory +I0831 09:17:31.017008 140117123622080 logger_utils.py:84] Creating experiment directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_2/imagenet_resnet_jax. +I0831 09:17:31.569508 140117123622080 submission_runner.py:741] Creating directory at /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_2/imagenet_resnet_jax/trial_1. +I0831 09:17:31.835416 140117123622080 submission_runner.py:242] Initializing dataset. +I0831 09:17:32.366775 140117123622080 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:17:32.415141 140117123622080 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. +I0831 09:17:32.695226 140117123622080 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. +I0831 09:17:32.772963 140117123622080 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:17:33.654872 140117123622080 submission_runner.py:251] Initializing model. +I0831 09:17:56.796554 140117123622080 submission_runner.py:294] Initializing optimizer. +I0831 09:17:58.368351 140117123622080 submission_runner.py:299] Initializing metrics bundle. +I0831 09:17:58.368559 140117123622080 submission_runner.py:321] Initializing checkpoint and logger. +I0831 09:17:58.374812 140117123622080 checkpoints.py:1098] Found no checkpoint files in /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_2/imagenet_resnet_jax/trial_1 with prefix checkpoint_ +I0831 09:17:58.374938 140117123622080 submission_runner.py:344] Saving meta data to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_2/imagenet_resnet_jax/trial_1/meta_data_0.json. +I0831 09:17:58.643120 140117123622080 submission_runner.py:348] Saving flags to /experiment_runs/submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_2/imagenet_resnet_jax/trial_1/flags_0.json. +I0831 09:17:58.658167 140117123622080 submission_runner.py:359] Starting training loop. +/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. +See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. + warnings.warn("Some donated buffers were not usable:" +I0831 09:18:50.374566 139958705452800 logging_writer.py:48] [0] global_step=0, grad_norm=0.671302318572998, loss=6.922396659851074 +I0831 09:18:51.571356 140117123622080 spec.py:333] Evaluating on the training split. +I0831 09:18:51.840241 140117123622080 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:18:51.845649 140117123622080 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. +I0831 09:18:51.864247 140117123622080 reader.py:262] Creating a tf.data.Dataset reading 1024 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. +I0831 09:18:51.904102 140117123622080 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split train, from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:19:17.482455 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 09:19:17.486814 140117123622080 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:19:17.506456 140117123622080 dataset_info.py:793] For 'imagenet2012/5.1.0': fields info.[splits, supervised_keys] differ on disk and in the code. Keeping the one from code. +I0831 09:19:17.509088 140117123622080 reader.py:262] Creating a tf.data.Dataset reading 64 files located in folders: /data/imagenet/jax/imagenet2012/5.1.0. +I0831 09:19:17.544004 140117123622080 logging_logger.py:49] Constructing tf.data.Dataset imagenet2012 for split validation, from /data/imagenet/jax/imagenet2012/5.1.0 +I0831 09:19:49.577220 140117123622080 spec.py:363] Evaluating on the test split. +I0831 09:19:49.713175 140117123622080 dataset_info.py:707] Load dataset info from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 +I0831 09:19:49.783481 140117123622080 reader.py:262] Creating a tf.data.Dataset reading 16 files located in folders: /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0. +I0831 09:19:49.828219 140117123622080 logging_logger.py:49] Constructing tf.data.Dataset imagenet_v2 for split test, from /data/imagenet/jax/imagenet_v2/matched-frequency/3.0.0 +I0831 09:19:55.844435 140117123622080 submission_runner.py:516] Time since start: 117.18s, Step: 1, {'train/accuracy': Array(0.00075733, dtype=float32), 'train/loss': Array(6.9125247, dtype=float32), 'validation/accuracy': Array(0.00068, dtype=float32), 'validation/loss': Array(6.91273, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.0008, dtype=float32), 'test/loss': Array(6.9129095, dtype=float32), 'test/num_examples': 10000, 'score': 52.91309928894043, 'total_duration': 117.18459677696228, 'accumulated_submission_time': 52.91309928894043, 'accumulated_eval_time': 64.27141284942627, 'accumulated_logging_time': 0} +I0831 09:19:55.874442 139921964652288 logging_writer.py:48] [1] accumulated_eval_time=64.2714, accumulated_logging_time=0, accumulated_submission_time=52.9131, global_step=1, preemption_count=0, score=52.9131, test/accuracy=0.000800000037997961, test/loss=6.912909507751465, test/num_examples=10000, total_duration=117.185, train/accuracy=0.0007573341717943549, train/loss=6.912524700164795, validation/accuracy=0.0006799999973736703, validation/loss=6.9127302169799805, validation/num_examples=50000 +/usr/local/lib/python3.11/site-packages/jax/_src/interpreters/mlir.py:1268: UserWarning: Some donated buffers were not usable: float32[64], float32[64], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[256], float32[256], float32[1,1,64,64], float32[3,3,64,64], float32[1,1,64,256], float32[1,1,64,256], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[2048], float32[2048], float32[1,1,1024,512], float32[3,3,512,512], float32[1,1,512,2048], float32[1,1,1024,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[512], float32[512], float32[512], float32[512], float32[2048], float32[2048], float32[1,1,2048,512], float32[3,3,512,512], float32[1,1,512,2048], float32[64], float32[64], float32[64], float32[64], float32[256], float32[256], float32[1,1,256,64], float32[3,3,64,64], float32[1,1,64,256], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[512], float32[512], float32[1,1,256,128], float32[3,3,128,128], float32[1,1,128,512], float32[1,1,256,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[128], float32[128], float32[128], float32[128], float32[512], float32[512], float32[1,1,512,128], float32[3,3,128,128], float32[1,1,128,512], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1024], float32[1024], float32[1,1,512,256], float32[3,3,256,256], float32[1,1,256,1024], float32[1,1,512,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[256], float32[256], float32[256], float32[256], float32[1024], float32[1024], float32[1,1,1024,256], float32[3,3,256,256], float32[1,1,256,1024], float32[7,7,3,64], float32[1000], float32[2048,1000]. +See an explanation at https://docs.jax.dev/en/latest/faq.html#buffer-donation. + warnings.warn("Some donated buffers were not usable:" +I0831 09:20:37.044055 139921713002240 logging_writer.py:48] [100] global_step=100, grad_norm=0.6732674837112427, loss=6.805178642272949 +I0831 09:21:22.362460 139921931081472 logging_writer.py:48] [200] global_step=200, grad_norm=0.8279271721839905, loss=6.496425628662109 +I0831 09:22:07.204123 139921713002240 logging_writer.py:48] [300] global_step=300, grad_norm=0.98941570520401, loss=6.196466445922852 +I0831 09:22:51.327261 139921931081472 logging_writer.py:48] [400] global_step=400, grad_norm=1.8994367122650146, loss=5.963691711425781 +I0831 09:23:36.293325 139921713002240 logging_writer.py:48] [500] global_step=500, grad_norm=4.463675022125244, loss=5.802135467529297 +I0831 09:24:21.424843 139921931081472 logging_writer.py:48] [600] global_step=600, grad_norm=3.4486489295959473, loss=5.549911022186279 +I0831 09:25:05.476954 139921713002240 logging_writer.py:48] [700] global_step=700, grad_norm=4.136732578277588, loss=5.357512474060059 +I0831 09:25:49.082906 139921931081472 logging_writer.py:48] [800] global_step=800, grad_norm=3.4241135120391846, loss=5.240438938140869 +I0831 09:26:33.715470 139921713002240 logging_writer.py:48] [900] global_step=900, grad_norm=6.021355628967285, loss=4.997196197509766 +I0831 09:27:17.499777 139921931081472 logging_writer.py:48] [1000] global_step=1000, grad_norm=5.664543151855469, loss=5.166516304016113 +I0831 09:28:01.349140 139921713002240 logging_writer.py:48] [1100] global_step=1100, grad_norm=3.5806498527526855, loss=4.8967132568359375 +I0831 09:28:46.059575 139921931081472 logging_writer.py:48] [1200] global_step=1200, grad_norm=4.215710639953613, loss=4.7692060470581055 +I0831 09:29:18.409837 139921713002240 logging_writer.py:48] [1300] global_step=1300, grad_norm=4.731967926025391, loss=4.741401672363281 +I0831 09:29:45.553718 139921931081472 logging_writer.py:48] [1400] global_step=1400, grad_norm=6.556045055389404, loss=4.54466438293457 +I0831 09:30:12.788432 139921713002240 logging_writer.py:48] [1500] global_step=1500, grad_norm=8.186121940612793, loss=4.488224506378174 +I0831 09:30:39.924591 139921931081472 logging_writer.py:48] [1600] global_step=1600, grad_norm=6.157769203186035, loss=4.4683990478515625 +I0831 09:31:07.106417 139921713002240 logging_writer.py:48] [1700] global_step=1700, grad_norm=4.845212459564209, loss=4.409547328948975 +I0831 09:31:34.335842 139921931081472 logging_writer.py:48] [1800] global_step=1800, grad_norm=3.7392683029174805, loss=4.2734832763671875 +I0831 09:32:01.451253 139921713002240 logging_writer.py:48] [1900] global_step=1900, grad_norm=5.120582103729248, loss=4.172300338745117 +I0831 09:32:28.608979 139921931081472 logging_writer.py:48] [2000] global_step=2000, grad_norm=5.783570289611816, loss=3.9605648517608643 +I0831 09:32:55.820533 139921713002240 logging_writer.py:48] [2100] global_step=2100, grad_norm=3.6723592281341553, loss=3.865442991256714 +I0831 09:33:23.153527 139921931081472 logging_writer.py:48] [2200] global_step=2200, grad_norm=4.436460494995117, loss=3.8933777809143066 +I0831 09:33:50.270193 139921713002240 logging_writer.py:48] [2300] global_step=2300, grad_norm=3.628775119781494, loss=3.8791348934173584 +I0831 09:34:17.463001 139921931081472 logging_writer.py:48] [2400] global_step=2400, grad_norm=3.223262310028076, loss=3.6756186485290527 +I0831 09:34:44.608311 139921713002240 logging_writer.py:48] [2500] global_step=2500, grad_norm=3.4139716625213623, loss=3.656585454940796 +I0831 09:35:11.731884 139921931081472 logging_writer.py:48] [2600] global_step=2600, grad_norm=3.741330146789551, loss=3.519340991973877 +I0831 09:35:38.952089 139921713002240 logging_writer.py:48] [2700] global_step=2700, grad_norm=3.8499300479888916, loss=3.7066566944122314 +I0831 09:36:06.085273 139921931081472 logging_writer.py:48] [2800] global_step=2800, grad_norm=3.781147003173828, loss=3.416112184524536 +I0831 09:36:33.196752 139921713002240 logging_writer.py:48] [2900] global_step=2900, grad_norm=4.345715045928955, loss=3.342191696166992 +I0831 09:37:00.389617 139921931081472 logging_writer.py:48] [3000] global_step=3000, grad_norm=3.900597095489502, loss=3.3178553581237793 +I0831 09:37:27.508085 139921713002240 logging_writer.py:48] [3100] global_step=3100, grad_norm=3.4220376014709473, loss=3.2843198776245117 +I0831 09:37:54.871908 139921931081472 logging_writer.py:48] [3200] global_step=3200, grad_norm=3.2616593837738037, loss=3.1931347846984863 +I0831 09:38:22.081495 139921713002240 logging_writer.py:48] [3300] global_step=3300, grad_norm=3.641723155975342, loss=3.092127799987793 +I0831 09:38:49.190782 139921931081472 logging_writer.py:48] [3400] global_step=3400, grad_norm=4.614095211029053, loss=3.1763710975646973 +I0831 09:39:16.359427 139921713002240 logging_writer.py:48] [3500] global_step=3500, grad_norm=3.8950982093811035, loss=3.0589523315429688 +I0831 09:39:43.578144 139921931081472 logging_writer.py:48] [3600] global_step=3600, grad_norm=3.406949520111084, loss=2.9844846725463867 +I0831 09:40:10.733139 139921713002240 logging_writer.py:48] [3700] global_step=3700, grad_norm=3.46993088722229, loss=3.021303176879883 +I0831 09:40:37.874263 139921931081472 logging_writer.py:48] [3800] global_step=3800, grad_norm=3.9887073040008545, loss=2.92179536819458 +I0831 09:41:05.088846 139921713002240 logging_writer.py:48] [3900] global_step=3900, grad_norm=2.6450412273406982, loss=2.973175048828125 +I0831 09:41:32.244333 139921931081472 logging_writer.py:48] [4000] global_step=4000, grad_norm=3.2164812088012695, loss=2.9449028968811035 +I0831 09:41:59.404278 139921713002240 logging_writer.py:48] [4100] global_step=4100, grad_norm=3.143095016479492, loss=2.845137119293213 +I0831 09:42:26.596357 139921931081472 logging_writer.py:48] [4200] global_step=4200, grad_norm=2.5667765140533447, loss=2.575690269470215 +I0831 09:42:53.992221 139921713002240 logging_writer.py:48] [4300] global_step=4300, grad_norm=3.828885793685913, loss=2.8851840496063232 +I0831 09:43:21.156685 139921931081472 logging_writer.py:48] [4400] global_step=4400, grad_norm=2.493292808532715, loss=2.6563308238983154 +I0831 09:43:48.395231 139921713002240 logging_writer.py:48] [4500] global_step=4500, grad_norm=3.52335262298584, loss=2.6862270832061768 +I0831 09:44:15.575482 139921931081472 logging_writer.py:48] [4600] global_step=4600, grad_norm=2.3761024475097656, loss=2.6628615856170654 +I0831 09:44:42.705690 139921713002240 logging_writer.py:48] [4700] global_step=4700, grad_norm=3.711188554763794, loss=2.5667147636413574 +I0831 09:45:09.894652 139921931081472 logging_writer.py:48] [4800] global_step=4800, grad_norm=2.7228426933288574, loss=2.598125696182251 +I0831 09:45:37.031065 139921713002240 logging_writer.py:48] [4900] global_step=4900, grad_norm=2.3265883922576904, loss=2.572826385498047 +I0831 09:46:04.159627 139921931081472 logging_writer.py:48] [5000] global_step=5000, grad_norm=3.056170701980591, loss=2.4604644775390625 +I0831 09:46:31.361546 139921713002240 logging_writer.py:48] [5100] global_step=5100, grad_norm=4.208132266998291, loss=2.518136501312256 +I0831 09:46:58.488577 139921931081472 logging_writer.py:48] [5200] global_step=5200, grad_norm=2.7506370544433594, loss=2.480219602584839 +I0831 09:47:25.668330 139921713002240 logging_writer.py:48] [5300] global_step=5300, grad_norm=3.078336238861084, loss=2.5320334434509277 +I0831 09:47:53.109953 139921931081472 logging_writer.py:48] [5400] global_step=5400, grad_norm=2.381141424179077, loss=2.354475259780884 +I0831 09:48:20.273151 139921713002240 logging_writer.py:48] [5500] global_step=5500, grad_norm=1.9294296503067017, loss=2.4226274490356445 +I0831 09:48:47.438530 139921931081472 logging_writer.py:48] [5600] global_step=5600, grad_norm=2.0576999187469482, loss=2.3770675659179688 +I0831 09:49:14.666751 139921713002240 logging_writer.py:48] [5700] global_step=5700, grad_norm=2.644810199737549, loss=2.323413372039795 +I0831 09:49:41.797448 139921931081472 logging_writer.py:48] [5800] global_step=5800, grad_norm=2.4552206993103027, loss=2.509725570678711 +I0831 09:50:08.916086 139921713002240 logging_writer.py:48] [5900] global_step=5900, grad_norm=3.1033997535705566, loss=2.396461009979248 +I0831 09:50:36.111771 139921931081472 logging_writer.py:48] [6000] global_step=6000, grad_norm=2.5513463020324707, loss=2.127607583999634 +I0831 09:51:03.260863 139921713002240 logging_writer.py:48] [6100] global_step=6100, grad_norm=2.3323240280151367, loss=2.1998984813690186 +I0831 09:51:30.392834 139921931081472 logging_writer.py:48] [6200] global_step=6200, grad_norm=2.1485862731933594, loss=2.1842122077941895 +I0831 09:51:57.600767 139921713002240 logging_writer.py:48] [6300] global_step=6300, grad_norm=2.212850332260132, loss=2.197127342224121 +I0831 09:52:24.743527 139921931081472 logging_writer.py:48] [6400] global_step=6400, grad_norm=2.2067325115203857, loss=2.1780166625976562 +I0831 09:52:52.138629 139921713002240 logging_writer.py:48] [6500] global_step=6500, grad_norm=2.62827205657959, loss=2.2342987060546875 +I0831 09:53:11.858306 140117123622080 spec.py:333] Evaluating on the training split. +I0831 09:53:22.911187 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 09:53:33.210484 140117123622080 spec.py:363] Evaluating on the test split. +I0831 09:54:12.737363 140117123622080 submission_runner.py:516] Time since start: 2174.08s, Step: 6574, {'train/accuracy': Array(0.63345027, dtype=float32), 'train/loss': Array(1.5319011, dtype=float32), 'validation/accuracy': Array(0.57534, dtype=float32), 'validation/loss': Array(1.8242104, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.44320002, dtype=float32), 'test/loss': Array(2.5707448, dtype=float32), 'test/num_examples': 10000, 'score': 2048.840027332306, 'total_duration': 2174.077612400055, 'accumulated_submission_time': 2048.840027332306, 'accumulated_eval_time': 125.14888834953308, 'accumulated_logging_time': 0.038361549377441406} +I0831 09:54:12.777824 139921973044992 logging_writer.py:48] [6574] accumulated_eval_time=125.149, accumulated_logging_time=0.0383615, accumulated_submission_time=2048.84, global_step=6574, preemption_count=0, score=2048.84, test/accuracy=0.443200021982193, test/loss=2.570744752883911, test/num_examples=10000, total_duration=2174.08, train/accuracy=0.6334502696990967, train/loss=1.5319011211395264, validation/accuracy=0.5753399729728699, validation/loss=1.8242104053497314, validation/num_examples=50000 +I0831 09:54:20.115710 139921981437696 logging_writer.py:48] [6600] global_step=6600, grad_norm=3.4276084899902344, loss=2.143886089324951 +I0831 09:54:47.183719 139921973044992 logging_writer.py:48] [6700] global_step=6700, grad_norm=2.9612326622009277, loss=2.1813905239105225 +I0831 09:55:14.277994 139921981437696 logging_writer.py:48] [6800] global_step=6800, grad_norm=2.567972421646118, loss=2.1731934547424316 +I0831 09:55:41.477204 139921973044992 logging_writer.py:48] [6900] global_step=6900, grad_norm=2.43359112739563, loss=2.189258575439453 +I0831 09:56:08.631566 139921981437696 logging_writer.py:48] [7000] global_step=7000, grad_norm=2.3358943462371826, loss=2.1699442863464355 +I0831 09:56:35.768024 139921973044992 logging_writer.py:48] [7100] global_step=7100, grad_norm=2.1723597049713135, loss=2.035247325897217 +I0831 09:57:02.979728 139921981437696 logging_writer.py:48] [7200] global_step=7200, grad_norm=2.4455323219299316, loss=2.0193636417388916 +I0831 09:57:30.094182 139921973044992 logging_writer.py:48] [7300] global_step=7300, grad_norm=2.4453837871551514, loss=1.9815784692764282 +I0831 09:57:57.268525 139921981437696 logging_writer.py:48] [7400] global_step=7400, grad_norm=2.3310165405273438, loss=2.0499401092529297 +I0831 09:58:24.491646 139921973044992 logging_writer.py:48] [7500] global_step=7500, grad_norm=2.156442642211914, loss=2.0957813262939453 +I0831 09:58:51.887776 139921981437696 logging_writer.py:48] [7600] global_step=7600, grad_norm=2.736370801925659, loss=2.0305700302124023 +I0831 09:59:19.058232 139921973044992 logging_writer.py:48] [7700] global_step=7700, grad_norm=2.27716326713562, loss=1.9784483909606934 +I0831 09:59:46.274121 139921981437696 logging_writer.py:48] [7800] global_step=7800, grad_norm=3.3548831939697266, loss=2.1549479961395264 +I0831 10:00:13.441508 139921973044992 logging_writer.py:48] [7900] global_step=7900, grad_norm=2.4959888458251953, loss=2.093942642211914 +I0831 10:00:40.570418 139921981437696 logging_writer.py:48] [8000] global_step=8000, grad_norm=2.060521364212036, loss=1.9716020822525024 +I0831 10:01:07.767853 139921973044992 logging_writer.py:48] [8100] global_step=8100, grad_norm=2.0269997119903564, loss=1.975365400314331 +I0831 10:01:34.883898 139921981437696 logging_writer.py:48] [8200] global_step=8200, grad_norm=2.659266233444214, loss=1.9484673738479614 +I0831 10:02:02.016165 139921973044992 logging_writer.py:48] [8300] global_step=8300, grad_norm=2.3306727409362793, loss=1.9319210052490234 +I0831 10:02:29.204666 139921981437696 logging_writer.py:48] [8400] global_step=8400, grad_norm=2.3693923950195312, loss=2.103003978729248 +I0831 10:02:56.354918 139921973044992 logging_writer.py:48] [8500] global_step=8500, grad_norm=2.8095836639404297, loss=2.0188708305358887 +I0831 10:03:23.712198 139921981437696 logging_writer.py:48] [8600] global_step=8600, grad_norm=2.207477331161499, loss=1.8949849605560303 +I0831 10:03:50.908318 139921973044992 logging_writer.py:48] [8700] global_step=8700, grad_norm=2.711432933807373, loss=1.7700984477996826 +I0831 10:04:18.059498 139921981437696 logging_writer.py:48] [8800] global_step=8800, grad_norm=2.2592148780822754, loss=1.962151288986206 +I0831 10:04:45.224476 139921973044992 logging_writer.py:48] [8900] global_step=8900, grad_norm=2.608541488647461, loss=1.7523128986358643 +I0831 10:05:12.417922 139921981437696 logging_writer.py:48] [9000] global_step=9000, grad_norm=2.1136386394500732, loss=1.8664042949676514 +I0831 10:05:39.566835 139921973044992 logging_writer.py:48] [9100] global_step=9100, grad_norm=2.277000665664673, loss=1.8920762538909912 +I0831 10:06:06.705202 139921981437696 logging_writer.py:48] [9200] global_step=9200, grad_norm=2.6492016315460205, loss=1.9573493003845215 +I0831 10:06:33.907359 139921973044992 logging_writer.py:48] [9300] global_step=9300, grad_norm=3.013233184814453, loss=1.932753324508667 +I0831 10:07:01.055327 139921981437696 logging_writer.py:48] [9400] global_step=9400, grad_norm=2.2009568214416504, loss=1.8772724866867065 +I0831 10:07:28.204766 139921973044992 logging_writer.py:48] [9500] global_step=9500, grad_norm=2.3411340713500977, loss=1.9010428190231323 +I0831 10:07:55.409616 139921981437696 logging_writer.py:48] [9600] global_step=9600, grad_norm=2.5960745811462402, loss=1.868543267250061 +I0831 10:08:22.803319 139921973044992 logging_writer.py:48] [9700] global_step=9700, grad_norm=2.203566789627075, loss=1.82309889793396 +I0831 10:08:49.944330 139921981437696 logging_writer.py:48] [9800] global_step=9800, grad_norm=2.3515594005584717, loss=1.8992986679077148 +I0831 10:09:17.122526 139921973044992 logging_writer.py:48] [9900] global_step=9900, grad_norm=1.9531469345092773, loss=1.8034528493881226 +I0831 10:09:44.244627 139921981437696 logging_writer.py:48] [10000] global_step=10000, grad_norm=2.011896848678589, loss=1.7611663341522217 +I0831 10:10:11.351901 139921973044992 logging_writer.py:48] [10100] global_step=10100, grad_norm=2.066847562789917, loss=1.7491786479949951 +I0831 10:10:38.555243 139921981437696 logging_writer.py:48] [10200] global_step=10200, grad_norm=2.490779161453247, loss=1.8243978023529053 +I0831 10:11:05.695623 139921973044992 logging_writer.py:48] [10300] global_step=10300, grad_norm=2.2087104320526123, loss=1.867042064666748 +I0831 10:11:32.842186 139921981437696 logging_writer.py:48] [10400] global_step=10400, grad_norm=2.306541919708252, loss=1.968435525894165 +I0831 10:12:00.054238 139921973044992 logging_writer.py:48] [10500] global_step=10500, grad_norm=2.5482168197631836, loss=1.8215970993041992 +I0831 10:12:27.199740 139921981437696 logging_writer.py:48] [10600] global_step=10600, grad_norm=2.6003804206848145, loss=1.826767086982727 +I0831 10:12:54.347085 139921973044992 logging_writer.py:48] [10700] global_step=10700, grad_norm=2.502636432647705, loss=1.753539800643921 +I0831 10:13:21.791948 139921981437696 logging_writer.py:48] [10800] global_step=10800, grad_norm=2.380722761154175, loss=1.9236633777618408 +I0831 10:13:48.889847 139921973044992 logging_writer.py:48] [10900] global_step=10900, grad_norm=2.4224681854248047, loss=1.7430059909820557 +I0831 10:14:16.011931 139921981437696 logging_writer.py:48] [11000] global_step=11000, grad_norm=2.4657652378082275, loss=1.8093527555465698 +I0831 10:14:43.211802 139921973044992 logging_writer.py:48] [11100] global_step=11100, grad_norm=2.211857557296753, loss=1.8176498413085938 +I0831 10:15:10.346898 139921981437696 logging_writer.py:48] [11200] global_step=11200, grad_norm=2.6209747791290283, loss=1.8674273490905762 +I0831 10:15:37.499604 139921973044992 logging_writer.py:48] [11300] global_step=11300, grad_norm=2.86953067779541, loss=1.8229273557662964 +I0831 10:16:04.676340 139921981437696 logging_writer.py:48] [11400] global_step=11400, grad_norm=3.2597477436065674, loss=1.728245496749878 +I0831 10:16:31.802548 139921973044992 logging_writer.py:48] [11500] global_step=11500, grad_norm=2.1995089054107666, loss=1.683197021484375 +I0831 10:16:58.941525 139921981437696 logging_writer.py:48] [11600] global_step=11600, grad_norm=2.242316961288452, loss=1.715667724609375 +I0831 10:17:26.106333 139921973044992 logging_writer.py:48] [11700] global_step=11700, grad_norm=2.2985260486602783, loss=1.674835443496704 +I0831 10:17:53.257793 139921981437696 logging_writer.py:48] [11800] global_step=11800, grad_norm=1.9165884256362915, loss=1.6611478328704834 +I0831 10:18:20.613840 139921973044992 logging_writer.py:48] [11900] global_step=11900, grad_norm=2.277613401412964, loss=1.7896418571472168 +I0831 10:18:47.820008 139921981437696 logging_writer.py:48] [12000] global_step=12000, grad_norm=2.513701915740967, loss=1.7087757587432861 +I0831 10:19:14.991380 139921973044992 logging_writer.py:48] [12100] global_step=12100, grad_norm=2.635773181915283, loss=1.7630031108856201 +I0831 10:19:42.110202 139921981437696 logging_writer.py:48] [12200] global_step=12200, grad_norm=2.350038528442383, loss=1.6332464218139648 +I0831 10:20:09.319403 139921973044992 logging_writer.py:48] [12300] global_step=12300, grad_norm=2.189542770385742, loss=1.587756872177124 +I0831 10:20:36.423456 139921981437696 logging_writer.py:48] [12400] global_step=12400, grad_norm=2.141578435897827, loss=1.6569583415985107 +I0831 10:21:03.536365 139921973044992 logging_writer.py:48] [12500] global_step=12500, grad_norm=2.6314451694488525, loss=1.5640114545822144 +I0831 10:21:30.710562 139921981437696 logging_writer.py:48] [12600] global_step=12600, grad_norm=2.172481060028076, loss=1.6280877590179443 +I0831 10:21:57.855106 139921973044992 logging_writer.py:48] [12700] global_step=12700, grad_norm=2.6087169647216797, loss=1.762603521347046 +I0831 10:22:24.983995 139921981437696 logging_writer.py:48] [12800] global_step=12800, grad_norm=2.3205695152282715, loss=1.6876832246780396 +I0831 10:22:52.395025 139921973044992 logging_writer.py:48] [12900] global_step=12900, grad_norm=2.7624080181121826, loss=1.8627490997314453 +I0831 10:23:19.479101 139921981437696 logging_writer.py:48] [13000] global_step=13000, grad_norm=2.7997493743896484, loss=1.5938355922698975 +I0831 10:23:46.587140 139921973044992 logging_writer.py:48] [13100] global_step=13100, grad_norm=2.597593307495117, loss=1.662946105003357 +I0831 10:24:13.799708 139921981437696 logging_writer.py:48] [13200] global_step=13200, grad_norm=2.4437668323516846, loss=1.58249032497406 +I0831 10:24:40.892146 139921973044992 logging_writer.py:48] [13300] global_step=13300, grad_norm=2.2696990966796875, loss=1.5391333103179932 +I0831 10:25:08.025183 139921981437696 logging_writer.py:48] [13400] global_step=13400, grad_norm=2.3729562759399414, loss=1.6627873182296753 +I0831 10:25:35.229311 139921973044992 logging_writer.py:48] [13500] global_step=13500, grad_norm=2.497978448867798, loss=1.6674211025238037 +I0831 10:26:02.353099 139921981437696 logging_writer.py:48] [13600] global_step=13600, grad_norm=2.3331096172332764, loss=1.582601547241211 +I0831 10:26:29.539808 139921973044992 logging_writer.py:48] [13700] global_step=13700, grad_norm=2.6142168045043945, loss=1.6689107418060303 +I0831 10:26:56.744802 139921981437696 logging_writer.py:48] [13800] global_step=13800, grad_norm=2.6323695182800293, loss=1.6393215656280518 +I0831 10:27:23.866545 139921973044992 logging_writer.py:48] [13900] global_step=13900, grad_norm=2.3569791316986084, loss=1.660409688949585 +I0831 10:27:28.905031 140117123622080 spec.py:333] Evaluating on the training split. +I0831 10:27:40.108533 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 10:27:50.312516 140117123622080 spec.py:363] Evaluating on the test split. +I0831 10:27:51.188020 140117123622080 submission_runner.py:516] Time since start: 4192.53s, Step: 13920, {'train/accuracy': Array(0.7597656, dtype=float32), 'train/loss': Array(0.9458039, dtype=float32), 'validation/accuracy': Array(0.66716, dtype=float32), 'validation/loss': Array(1.3747368, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.5343, dtype=float32), 'test/loss': Array(2.1329305, dtype=float32), 'test/num_examples': 10000, 'score': 4044.9069986343384, 'total_duration': 4192.527692556381, 'accumulated_submission_time': 4044.9069986343384, 'accumulated_eval_time': 147.4297194480896, 'accumulated_logging_time': 0.0868222713470459} +I0831 10:27:51.227185 139921981437696 logging_writer.py:48] [13920] accumulated_eval_time=147.43, accumulated_logging_time=0.0868223, accumulated_submission_time=4044.91, global_step=13920, preemption_count=0, score=4044.91, test/accuracy=0.5343000292778015, test/loss=2.1329305171966553, test/num_examples=10000, total_duration=4192.53, train/accuracy=0.759765625, train/loss=0.9458038806915283, validation/accuracy=0.6671599745750427, validation/loss=1.3747367858886719, validation/num_examples=50000 +I0831 10:28:13.470738 139921973044992 logging_writer.py:48] [14000] global_step=14000, grad_norm=3.0350115299224854, loss=1.696923851966858 +I0831 10:28:40.659779 139921981437696 logging_writer.py:48] [14100] global_step=14100, grad_norm=2.6212310791015625, loss=1.553265929222107 +I0831 10:29:07.792446 139921973044992 logging_writer.py:48] [14200] global_step=14200, grad_norm=2.9743869304656982, loss=1.6971187591552734 +I0831 10:29:34.911542 139921981437696 logging_writer.py:48] [14300] global_step=14300, grad_norm=2.7357542514801025, loss=1.702121615409851 +I0831 10:30:02.115375 139921973044992 logging_writer.py:48] [14400] global_step=14400, grad_norm=2.7801671028137207, loss=1.6691523790359497 +I0831 10:30:29.238617 139921981437696 logging_writer.py:48] [14500] global_step=14500, grad_norm=2.149615526199341, loss=1.6176680326461792 +I0831 10:30:56.376146 139921973044992 logging_writer.py:48] [14600] global_step=14600, grad_norm=2.585505247116089, loss=1.609571933746338 +I0831 10:31:23.596660 139921981437696 logging_writer.py:48] [14700] global_step=14700, grad_norm=2.2697601318359375, loss=1.6139100790023804 +I0831 10:31:50.710155 139921973044992 logging_writer.py:48] [14800] global_step=14800, grad_norm=2.8866593837738037, loss=1.7605444192886353 +I0831 10:32:17.819612 139921981437696 logging_writer.py:48] [14900] global_step=14900, grad_norm=2.4191031455993652, loss=1.6256027221679688 +I0831 10:32:45.014321 139921973044992 logging_writer.py:48] [15000] global_step=15000, grad_norm=2.7074241638183594, loss=1.6248866319656372 +I0831 10:33:12.362349 139921981437696 logging_writer.py:48] [15100] global_step=15100, grad_norm=2.6054131984710693, loss=1.653914213180542 +I0831 10:33:39.486848 139921973044992 logging_writer.py:48] [15200] global_step=15200, grad_norm=2.5788371562957764, loss=1.5797102451324463 +I0831 10:34:06.722464 139921981437696 logging_writer.py:48] [15300] global_step=15300, grad_norm=2.3922674655914307, loss=1.5725655555725098 +I0831 10:34:33.856937 139921973044992 logging_writer.py:48] [15400] global_step=15400, grad_norm=2.6645755767822266, loss=1.5964326858520508 +I0831 10:35:00.954557 139921981437696 logging_writer.py:48] [15500] global_step=15500, grad_norm=2.7354321479797363, loss=1.6281343698501587 +I0831 10:35:28.149845 139921973044992 logging_writer.py:48] [15600] global_step=15600, grad_norm=2.836534261703491, loss=1.5315821170806885 +I0831 10:35:55.268275 139921981437696 logging_writer.py:48] [15700] global_step=15700, grad_norm=2.9557156562805176, loss=1.4942699670791626 +I0831 10:36:22.437966 139921973044992 logging_writer.py:48] [15800] global_step=15800, grad_norm=2.622300148010254, loss=1.50852632522583 +I0831 10:36:49.638053 139921981437696 logging_writer.py:48] [15900] global_step=15900, grad_norm=2.7346372604370117, loss=1.4732731580734253 +I0831 10:37:16.798240 139921973044992 logging_writer.py:48] [16000] global_step=16000, grad_norm=2.64225697517395, loss=1.4495433568954468 +I0831 10:37:44.129991 139921981437696 logging_writer.py:48] [16100] global_step=16100, grad_norm=2.8667867183685303, loss=1.6103379726409912 +I0831 10:38:11.330109 139921973044992 logging_writer.py:48] [16200] global_step=16200, grad_norm=3.1004464626312256, loss=1.6541290283203125 +I0831 10:38:38.463275 139921981437696 logging_writer.py:48] [16300] global_step=16300, grad_norm=2.8889665603637695, loss=1.4718906879425049 +I0831 10:39:05.612018 139921973044992 logging_writer.py:48] [16400] global_step=16400, grad_norm=3.2106285095214844, loss=1.5856781005859375 +I0831 10:39:32.815634 139921981437696 logging_writer.py:48] [16500] global_step=16500, grad_norm=2.707829713821411, loss=1.473958969116211 +I0831 10:39:59.952388 139921973044992 logging_writer.py:48] [16600] global_step=16600, grad_norm=2.9253182411193848, loss=1.5339279174804688 +I0831 10:40:27.098172 139921981437696 logging_writer.py:48] [16700] global_step=16700, grad_norm=2.774484395980835, loss=1.53962242603302 +I0831 10:40:54.319318 139921973044992 logging_writer.py:48] [16800] global_step=16800, grad_norm=3.0922129154205322, loss=1.5274388790130615 +I0831 10:41:21.442504 139921981437696 logging_writer.py:48] [16900] global_step=16900, grad_norm=2.9338717460632324, loss=1.475264072418213 +I0831 10:41:48.596605 139921973044992 logging_writer.py:48] [17000] global_step=17000, grad_norm=3.4707348346710205, loss=1.4933245182037354 +I0831 10:42:15.768720 139921981437696 logging_writer.py:48] [17100] global_step=17100, grad_norm=3.425891637802124, loss=1.4522218704223633 +I0831 10:42:43.079227 139921973044992 logging_writer.py:48] [17200] global_step=17200, grad_norm=3.340009927749634, loss=1.503456711769104 +I0831 10:43:10.231001 139921981437696 logging_writer.py:48] [17300] global_step=17300, grad_norm=3.1978204250335693, loss=1.5897223949432373 +I0831 10:43:37.431075 139921973044992 logging_writer.py:48] [17400] global_step=17400, grad_norm=3.486799716949463, loss=1.5765739679336548 +I0831 10:44:04.550753 139921981437696 logging_writer.py:48] [17500] global_step=17500, grad_norm=3.2303504943847656, loss=1.4816519021987915 +I0831 10:44:31.680552 139921973044992 logging_writer.py:48] [17600] global_step=17600, grad_norm=3.287923812866211, loss=1.5123013257980347 +I0831 10:44:58.872562 139921981437696 logging_writer.py:48] [17700] global_step=17700, grad_norm=3.3260560035705566, loss=1.4286954402923584 +I0831 10:45:26.010697 139921973044992 logging_writer.py:48] [17800] global_step=17800, grad_norm=3.5685641765594482, loss=1.4889247417449951 +I0831 10:45:53.153273 139921981437696 logging_writer.py:48] [17900] global_step=17900, grad_norm=3.491772413253784, loss=1.5587143898010254 +I0831 10:46:20.322552 139921973044992 logging_writer.py:48] [18000] global_step=18000, grad_norm=3.584510087966919, loss=1.5186439752578735 +I0831 10:46:47.450887 139921981437696 logging_writer.py:48] [18100] global_step=18100, grad_norm=4.201836585998535, loss=1.4934589862823486 +I0831 10:47:14.563385 139921973044992 logging_writer.py:48] [18200] global_step=18200, grad_norm=3.9445481300354004, loss=1.4944424629211426 +I0831 10:47:41.985266 139921981437696 logging_writer.py:48] [18300] global_step=18300, grad_norm=4.000143051147461, loss=1.5370042324066162 +I0831 10:48:09.114527 139921973044992 logging_writer.py:48] [18400] global_step=18400, grad_norm=4.293737411499023, loss=1.554142713546753 +I0831 10:48:36.238921 139921981437696 logging_writer.py:48] [18500] global_step=18500, grad_norm=3.875096321105957, loss=1.4579441547393799 +I0831 10:49:03.452561 139921973044992 logging_writer.py:48] [18600] global_step=18600, grad_norm=4.548758506774902, loss=1.395484447479248 +I0831 10:49:30.585047 139921981437696 logging_writer.py:48] [18700] global_step=18700, grad_norm=4.045810699462891, loss=1.5189390182495117 +I0831 10:49:57.741370 139921973044992 logging_writer.py:48] [18800] global_step=18800, grad_norm=4.105224132537842, loss=1.4922289848327637 +I0831 10:50:24.926796 139921981437696 logging_writer.py:48] [18900] global_step=18900, grad_norm=4.743765830993652, loss=1.4784990549087524 +I0831 10:50:52.076395 139921973044992 logging_writer.py:48] [19000] global_step=19000, grad_norm=4.483284950256348, loss=1.5229483842849731 +I0831 10:51:19.173301 139921981437696 logging_writer.py:48] [19100] global_step=19100, grad_norm=4.762443542480469, loss=1.5833650827407837 +I0831 10:51:46.350666 139921973044992 logging_writer.py:48] [19200] global_step=19200, grad_norm=5.052002429962158, loss=1.5220892429351807 +I0831 10:52:13.697761 139921981437696 logging_writer.py:48] [19300] global_step=19300, grad_norm=5.090843677520752, loss=1.4739071130752563 +I0831 10:52:40.810135 139921973044992 logging_writer.py:48] [19400] global_step=19400, grad_norm=5.638278007507324, loss=1.5797584056854248 +I0831 10:53:08.015201 139921981437696 logging_writer.py:48] [19500] global_step=19500, grad_norm=5.05310583114624, loss=1.4385422468185425 +I0831 10:53:35.158515 139921973044992 logging_writer.py:48] [19600] global_step=19600, grad_norm=5.3574442863464355, loss=1.5606913566589355 +I0831 10:54:02.290697 139921981437696 logging_writer.py:48] [19700] global_step=19700, grad_norm=5.573651313781738, loss=1.5450961589813232 +I0831 10:54:29.470190 139921973044992 logging_writer.py:48] [19800] global_step=19800, grad_norm=6.741726398468018, loss=1.6309282779693604 +I0831 10:54:56.599427 139921981437696 logging_writer.py:48] [19900] global_step=19900, grad_norm=6.430349349975586, loss=1.6098532676696777 +I0831 10:55:23.729465 139921973044992 logging_writer.py:48] [20000] global_step=20000, grad_norm=6.62655782699585, loss=1.5774004459381104 +I0831 10:55:50.908279 139921981437696 logging_writer.py:48] [20100] global_step=20100, grad_norm=6.508082389831543, loss=1.5472185611724854 +I0831 10:56:18.033620 139921973044992 logging_writer.py:48] [20200] global_step=20200, grad_norm=7.016639232635498, loss=1.4600238800048828 +I0831 10:56:45.147241 139921981437696 logging_writer.py:48] [20300] global_step=20300, grad_norm=6.589377403259277, loss=1.41517972946167 +I0831 10:57:12.638421 139921973044992 logging_writer.py:48] [20400] global_step=20400, grad_norm=7.282048225402832, loss=1.5558980703353882 +I0831 10:57:39.785723 139921981437696 logging_writer.py:48] [20500] global_step=20500, grad_norm=7.552374839782715, loss=1.5217782258987427 +I0831 10:58:06.895072 139921973044992 logging_writer.py:48] [20600] global_step=20600, grad_norm=7.863426208496094, loss=1.5843675136566162 +I0831 10:58:34.120080 139921981437696 logging_writer.py:48] [20700] global_step=20700, grad_norm=7.922863006591797, loss=1.4261691570281982 +I0831 10:59:01.242961 139921973044992 logging_writer.py:48] [20800] global_step=20800, grad_norm=8.0282564163208, loss=1.429382562637329 +I0831 10:59:28.383879 139921981437696 logging_writer.py:48] [20900] global_step=20900, grad_norm=10.486676216125488, loss=1.5300984382629395 +I0831 10:59:55.590959 139921973044992 logging_writer.py:48] [21000] global_step=21000, grad_norm=8.677633285522461, loss=1.4619032144546509 +I0831 11:00:22.762533 139921981437696 logging_writer.py:48] [21100] global_step=21100, grad_norm=8.904156684875488, loss=1.5087047815322876 +I0831 11:00:49.861669 139921973044992 logging_writer.py:48] [21200] global_step=21200, grad_norm=9.725543975830078, loss=1.4908292293548584 +I0831 11:01:07.372734 140117123622080 spec.py:333] Evaluating on the training split. +I0831 11:01:18.697727 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 11:01:28.883568 140117123622080 spec.py:363] Evaluating on the test split. +I0831 11:01:29.764864 140117123622080 submission_runner.py:516] Time since start: 6211.10s, Step: 21266, {'train/accuracy': Array(0.8064812, dtype=float32), 'train/loss': Array(0.74569297, dtype=float32), 'validation/accuracy': Array(0.6971, dtype=float32), 'validation/loss': Array(1.2396864, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.5623, dtype=float32), 'test/loss': Array(1.972755, dtype=float32), 'test/num_examples': 10000, 'score': 6040.991535186768, 'total_duration': 6211.1035261154175, 'accumulated_submission_time': 6040.991535186768, 'accumulated_eval_time': 169.818678855896, 'accumulated_logging_time': 0.13496828079223633} +I0831 11:01:29.806053 139921981437696 logging_writer.py:48] [21266] accumulated_eval_time=169.819, accumulated_logging_time=0.134968, accumulated_submission_time=6040.99, global_step=21266, preemption_count=0, score=6040.99, test/accuracy=0.5623000264167786, test/loss=1.972754955291748, test/num_examples=10000, total_duration=6211.1, train/accuracy=0.8064811825752258, train/loss=0.7456929683685303, validation/accuracy=0.6970999836921692, validation/loss=1.239686369895935, validation/num_examples=50000 +I0831 11:01:39.455677 139921973044992 logging_writer.py:48] [21300] global_step=21300, grad_norm=9.406684875488281, loss=1.5405936241149902 +I0831 11:02:06.547469 139921981437696 logging_writer.py:48] [21400] global_step=21400, grad_norm=9.00126838684082, loss=1.322901964187622 +I0831 11:02:33.874349 139921973044992 logging_writer.py:48] [21500] global_step=21500, grad_norm=10.26239013671875, loss=1.4281567335128784 +I0831 11:03:01.060523 139921981437696 logging_writer.py:48] [21600] global_step=21600, grad_norm=10.113791465759277, loss=1.4933806657791138 +I0831 11:03:28.198747 139921973044992 logging_writer.py:48] [21700] global_step=21700, grad_norm=11.149961471557617, loss=1.4745469093322754 +I0831 11:03:55.315768 139921981437696 logging_writer.py:48] [21800] global_step=21800, grad_norm=10.801688194274902, loss=1.4619765281677246 +I0831 11:04:22.512084 139921973044992 logging_writer.py:48] [21900] global_step=21900, grad_norm=11.78383731842041, loss=1.5049852132797241 +I0831 11:04:49.684288 139921981437696 logging_writer.py:48] [22000] global_step=22000, grad_norm=11.140398025512695, loss=1.4829232692718506 +I0831 11:05:16.797026 139921973044992 logging_writer.py:48] [22100] global_step=22100, grad_norm=11.593482971191406, loss=1.5430198907852173 +I0831 11:05:43.976268 139921981437696 logging_writer.py:48] [22200] global_step=22200, grad_norm=11.26374340057373, loss=1.463257074356079 +I0831 11:06:11.109401 139921973044992 logging_writer.py:48] [22300] global_step=22300, grad_norm=12.335799217224121, loss=1.4707428216934204 +I0831 11:06:38.295112 139921981437696 logging_writer.py:48] [22400] global_step=22400, grad_norm=11.759468078613281, loss=1.5352859497070312 +I0831 11:07:05.701172 139921973044992 logging_writer.py:48] [22500] global_step=22500, grad_norm=11.995282173156738, loss=1.5268123149871826 +I0831 11:07:32.829876 139921981437696 logging_writer.py:48] [22600] global_step=22600, grad_norm=12.468128204345703, loss=1.4467936754226685 +I0831 11:07:59.950872 139921973044992 logging_writer.py:48] [22700] global_step=22700, grad_norm=12.801198959350586, loss=1.6025171279907227 +I0831 11:08:27.163942 139921981437696 logging_writer.py:48] [22800] global_step=22800, grad_norm=12.915553092956543, loss=1.5500178337097168 +I0831 11:08:54.286881 139921973044992 logging_writer.py:48] [22900] global_step=22900, grad_norm=12.996624946594238, loss=1.4899919033050537 +I0831 11:09:21.390110 139921981437696 logging_writer.py:48] [23000] global_step=23000, grad_norm=12.457157135009766, loss=1.459533452987671 +I0831 11:09:48.583037 139921973044992 logging_writer.py:48] [23100] global_step=23100, grad_norm=13.232903480529785, loss=1.5153093338012695 +I0831 11:10:15.727595 139921981437696 logging_writer.py:48] [23200] global_step=23200, grad_norm=12.763252258300781, loss=1.5098117589950562 +I0831 11:10:42.899512 139921973044992 logging_writer.py:48] [23300] global_step=23300, grad_norm=12.972905158996582, loss=1.630241870880127 +I0831 11:11:10.099871 139921981437696 logging_writer.py:48] [23400] global_step=23400, grad_norm=12.698519706726074, loss=1.4136146306991577 +I0831 11:11:37.200750 139921973044992 logging_writer.py:48] [23500] global_step=23500, grad_norm=13.72935676574707, loss=1.546349287033081 +I0831 11:12:04.553127 139921981437696 logging_writer.py:48] [23600] global_step=23600, grad_norm=13.272187232971191, loss=1.528562307357788 +I0831 11:12:31.765105 139921973044992 logging_writer.py:48] [23700] global_step=23700, grad_norm=12.644031524658203, loss=1.4883387088775635 +I0831 11:12:58.881719 139921981437696 logging_writer.py:48] [23800] global_step=23800, grad_norm=13.084260940551758, loss=1.4790527820587158 +I0831 11:13:26.005377 139921973044992 logging_writer.py:48] [23900] global_step=23900, grad_norm=12.575275421142578, loss=1.5619521141052246 +I0831 11:13:53.243292 139921981437696 logging_writer.py:48] [24000] global_step=24000, grad_norm=12.304837226867676, loss=1.390584945678711 +I0831 11:14:20.387007 139921973044992 logging_writer.py:48] [24100] global_step=24100, grad_norm=12.258387565612793, loss=1.5254509449005127 +I0831 11:14:47.533431 139921981437696 logging_writer.py:48] [24200] global_step=24200, grad_norm=11.880969047546387, loss=1.4036471843719482 +I0831 11:15:14.713719 139921973044992 logging_writer.py:48] [24300] global_step=24300, grad_norm=12.103780746459961, loss=1.4992352724075317 +I0831 11:15:41.842554 139921981437696 logging_writer.py:48] [24400] global_step=24400, grad_norm=11.928203582763672, loss=1.3931465148925781 +I0831 11:16:08.956238 139921973044992 logging_writer.py:48] [24500] global_step=24500, grad_norm=11.705646514892578, loss=1.469421148300171 +I0831 11:16:36.151454 139921981437696 logging_writer.py:48] [24600] global_step=24600, grad_norm=12.138449668884277, loss=1.5468780994415283 +I0831 11:17:03.512737 139921973044992 logging_writer.py:48] [24700] global_step=24700, grad_norm=11.193843841552734, loss=1.4395384788513184 +I0831 11:17:30.660895 139921981437696 logging_writer.py:48] [24800] global_step=24800, grad_norm=11.531634330749512, loss=1.4053001403808594 +I0831 11:17:57.868263 139921973044992 logging_writer.py:48] [24900] global_step=24900, grad_norm=11.036750793457031, loss=1.417051076889038 +I0831 11:18:25.011195 139921981437696 logging_writer.py:48] [25000] global_step=25000, grad_norm=10.839705467224121, loss=1.4573945999145508 +I0831 11:18:52.121787 139921973044992 logging_writer.py:48] [25100] global_step=25100, grad_norm=11.357416152954102, loss=1.4634182453155518 +I0831 11:19:19.318676 139921981437696 logging_writer.py:48] [25200] global_step=25200, grad_norm=10.925622940063477, loss=1.5204107761383057 +I0831 11:19:46.422720 139921973044992 logging_writer.py:48] [25300] global_step=25300, grad_norm=10.379297256469727, loss=1.4389749765396118 +I0831 11:20:13.531986 139921981437696 logging_writer.py:48] [25400] global_step=25400, grad_norm=10.401284217834473, loss=1.4607157707214355 +I0831 11:20:40.718140 139921973044992 logging_writer.py:48] [25500] global_step=25500, grad_norm=9.391441345214844, loss=1.309553623199463 +I0831 11:21:07.869062 139921981437696 logging_writer.py:48] [25600] global_step=25600, grad_norm=10.705259323120117, loss=1.5459126234054565 +I0831 11:21:35.227403 139921973044992 logging_writer.py:48] [25700] global_step=25700, grad_norm=9.696977615356445, loss=1.3967657089233398 +I0831 11:22:02.404239 139921981437696 logging_writer.py:48] [25800] global_step=25800, grad_norm=9.816316604614258, loss=1.4359928369522095 +I0831 11:22:29.538554 139921973044992 logging_writer.py:48] [25900] global_step=25900, grad_norm=9.650331497192383, loss=1.4513461589813232 +I0831 11:22:56.645617 139921981437696 logging_writer.py:48] [26000] global_step=26000, grad_norm=9.117806434631348, loss=1.3358750343322754 +I0831 11:23:23.823630 139921973044992 logging_writer.py:48] [26100] global_step=26100, grad_norm=9.917698860168457, loss=1.4318798780441284 +I0831 11:23:50.952708 139921981437696 logging_writer.py:48] [26200] global_step=26200, grad_norm=8.983199119567871, loss=1.296919345855713 +I0831 11:24:18.047026 139921973044992 logging_writer.py:48] [26300] global_step=26300, grad_norm=8.571432113647461, loss=1.3564395904541016 +I0831 11:24:45.221361 139921981437696 logging_writer.py:48] [26400] global_step=26400, grad_norm=9.130450248718262, loss=1.4363667964935303 +I0831 11:25:12.365995 139921973044992 logging_writer.py:48] [26500] global_step=26500, grad_norm=8.884771347045898, loss=1.4640765190124512 +I0831 11:25:39.485832 139921981437696 logging_writer.py:48] [26600] global_step=26600, grad_norm=8.978346824645996, loss=1.4533213376998901 +I0831 11:26:06.658720 139921973044992 logging_writer.py:48] [26700] global_step=26700, grad_norm=8.780391693115234, loss=1.454515814781189 +I0831 11:26:34.004344 139921981437696 logging_writer.py:48] [26800] global_step=26800, grad_norm=8.460187911987305, loss=1.4112629890441895 +I0831 11:27:01.132078 139921973044992 logging_writer.py:48] [26900] global_step=26900, grad_norm=8.300875663757324, loss=1.5080320835113525 +I0831 11:27:28.308481 139921981437696 logging_writer.py:48] [27000] global_step=27000, grad_norm=8.671027183532715, loss=1.4209853410720825 +I0831 11:27:55.420516 139921973044992 logging_writer.py:48] [27100] global_step=27100, grad_norm=8.315540313720703, loss=1.48972487449646 +I0831 11:28:22.526274 139921981437696 logging_writer.py:48] [27200] global_step=27200, grad_norm=7.979002475738525, loss=1.3534343242645264 +I0831 11:28:49.730326 139921973044992 logging_writer.py:48] [27300] global_step=27300, grad_norm=7.8381547927856445, loss=1.4412386417388916 +I0831 11:29:16.873912 139921981437696 logging_writer.py:48] [27400] global_step=27400, grad_norm=8.53399658203125, loss=1.4924297332763672 +I0831 11:29:44.017818 139921973044992 logging_writer.py:48] [27500] global_step=27500, grad_norm=8.105305671691895, loss=1.3942785263061523 +I0831 11:30:11.222815 139921981437696 logging_writer.py:48] [27600] global_step=27600, grad_norm=7.361893653869629, loss=1.3881618976593018 +I0831 11:30:38.361924 139921973044992 logging_writer.py:48] [27700] global_step=27700, grad_norm=7.707334041595459, loss=1.4619085788726807 +I0831 11:31:05.486790 139921981437696 logging_writer.py:48] [27800] global_step=27800, grad_norm=7.432123184204102, loss=1.388923168182373 +I0831 11:31:32.894568 139921973044992 logging_writer.py:48] [27900] global_step=27900, grad_norm=7.237215518951416, loss=1.347717523574829 +I0831 11:32:00.025931 139921981437696 logging_writer.py:48] [28000] global_step=28000, grad_norm=7.277935981750488, loss=1.3317420482635498 +I0831 11:32:27.175334 139921973044992 logging_writer.py:48] [28100] global_step=28100, grad_norm=7.511354446411133, loss=1.5032631158828735 +I0831 11:32:54.365677 139921981437696 logging_writer.py:48] [28200] global_step=28200, grad_norm=6.93440055847168, loss=1.4020024538040161 +I0831 11:33:21.486584 139921973044992 logging_writer.py:48] [28300] global_step=28300, grad_norm=6.751073360443115, loss=1.2943435907363892 +I0831 11:33:48.635211 139921981437696 logging_writer.py:48] [28400] global_step=28400, grad_norm=6.576148509979248, loss=1.3428833484649658 +I0831 11:34:15.818835 139921973044992 logging_writer.py:48] [28500] global_step=28500, grad_norm=6.855224132537842, loss=1.3868749141693115 +I0831 11:34:42.955046 139921981437696 logging_writer.py:48] [28600] global_step=28600, grad_norm=7.0012617111206055, loss=1.3856619596481323 +I0831 11:34:45.835970 140117123622080 spec.py:333] Evaluating on the training split. +I0831 11:34:56.947783 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 11:35:06.899092 140117123622080 spec.py:363] Evaluating on the test split. +I0831 11:35:07.781029 140117123622080 submission_runner.py:516] Time since start: 8229.12s, Step: 28612, {'train/accuracy': Array(0.83091515, dtype=float32), 'train/loss': Array(0.63762105, dtype=float32), 'validation/accuracy': Array(0.71488, dtype=float32), 'validation/loss': Array(1.1683154, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.58250004, dtype=float32), 'test/loss': Array(1.8919845, dtype=float32), 'test/num_examples': 10000, 'score': 8036.958885192871, 'total_duration': 8229.120635509491, 'accumulated_submission_time': 8036.958885192871, 'accumulated_eval_time': 191.76151204109192, 'accumulated_logging_time': 0.1867382526397705} +I0831 11:35:07.834580 139921973044992 logging_writer.py:48] [28612] accumulated_eval_time=191.762, accumulated_logging_time=0.186738, accumulated_submission_time=8036.96, global_step=28612, preemption_count=0, score=8036.96, test/accuracy=0.5825000405311584, test/loss=1.891984462738037, test/num_examples=10000, total_duration=8229.12, train/accuracy=0.8309151530265808, train/loss=0.6376210451126099, validation/accuracy=0.7148799896240234, validation/loss=1.1683154106140137, validation/num_examples=50000 +I0831 11:35:32.088268 139921981437696 logging_writer.py:48] [28700] global_step=28700, grad_norm=7.163887023925781, loss=1.442373275756836 +I0831 11:35:59.244001 139921973044992 logging_writer.py:48] [28800] global_step=28800, grad_norm=6.726470947265625, loss=1.3722470998764038 +I0831 11:36:26.578151 139921981437696 logging_writer.py:48] [28900] global_step=28900, grad_norm=6.451501846313477, loss=1.3578377962112427 +I0831 11:36:53.716747 139921973044992 logging_writer.py:48] [29000] global_step=29000, grad_norm=6.5456414222717285, loss=1.469851016998291 +I0831 11:37:20.885295 139921981437696 logging_writer.py:48] [29100] global_step=29100, grad_norm=6.104044437408447, loss=1.2386316061019897 +I0831 11:37:48.006497 139921973044992 logging_writer.py:48] [29200] global_step=29200, grad_norm=6.364472389221191, loss=1.3190021514892578 +I0831 11:38:15.095359 139921981437696 logging_writer.py:48] [29300] global_step=29300, grad_norm=5.965795040130615, loss=1.275193214416504 +I0831 11:38:42.270210 139921973044992 logging_writer.py:48] [29400] global_step=29400, grad_norm=6.086014270782471, loss=1.3182551860809326 +I0831 11:39:09.389003 139921981437696 logging_writer.py:48] [29500] global_step=29500, grad_norm=6.659902095794678, loss=1.3723340034484863 +I0831 11:39:36.516043 139921973044992 logging_writer.py:48] [29600] global_step=29600, grad_norm=5.902381420135498, loss=1.3997684717178345 +I0831 11:40:03.704919 139921981437696 logging_writer.py:48] [29700] global_step=29700, grad_norm=6.285914897918701, loss=1.5150458812713623 +I0831 11:40:30.814225 139921973044992 logging_writer.py:48] [29800] global_step=29800, grad_norm=6.31815242767334, loss=1.452488660812378 +I0831 11:40:57.929125 139921981437696 logging_writer.py:48] [29900] global_step=29900, grad_norm=6.77439546585083, loss=1.360316514968872 +I0831 11:41:25.306886 139921973044992 logging_writer.py:48] [30000] global_step=30000, grad_norm=6.0320892333984375, loss=1.4418996572494507 +I0831 11:41:52.401064 139921981437696 logging_writer.py:48] [30100] global_step=30100, grad_norm=5.522887706756592, loss=1.3230851888656616 +I0831 11:42:19.538672 139921973044992 logging_writer.py:48] [30200] global_step=30200, grad_norm=6.063093185424805, loss=1.3351341485977173 +I0831 11:42:46.718826 139921981437696 logging_writer.py:48] [30300] global_step=30300, grad_norm=5.768405914306641, loss=1.3954832553863525 +I0831 11:43:13.850247 139921973044992 logging_writer.py:48] [30400] global_step=30400, grad_norm=5.789111137390137, loss=1.3224598169326782 +I0831 11:43:40.976785 139921981437696 logging_writer.py:48] [30500] global_step=30500, grad_norm=5.717803478240967, loss=1.3184634447097778 +I0831 11:44:08.137033 139921973044992 logging_writer.py:48] [30600] global_step=30600, grad_norm=5.868815898895264, loss=1.3037062883377075 +I0831 11:44:35.245319 139921981437696 logging_writer.py:48] [30700] global_step=30700, grad_norm=5.802486896514893, loss=1.408795952796936 +I0831 11:45:02.378905 139921973044992 logging_writer.py:48] [30800] global_step=30800, grad_norm=5.763869762420654, loss=1.4135808944702148 +I0831 11:45:29.562753 139921981437696 logging_writer.py:48] [30900] global_step=30900, grad_norm=5.700441837310791, loss=1.3622815608978271 +I0831 11:45:56.702739 139921973044992 logging_writer.py:48] [31000] global_step=31000, grad_norm=5.450464248657227, loss=1.3133822679519653 +I0831 11:46:24.034278 139921981437696 logging_writer.py:48] [31100] global_step=31100, grad_norm=5.201324939727783, loss=1.30841064453125 +I0831 11:46:51.227692 139921973044992 logging_writer.py:48] [31200] global_step=31200, grad_norm=5.693900108337402, loss=1.4155871868133545 +I0831 11:47:18.358658 139921981437696 logging_writer.py:48] [31300] global_step=31300, grad_norm=5.006798267364502, loss=1.310665249824524 +I0831 11:47:45.489391 139921973044992 logging_writer.py:48] [31400] global_step=31400, grad_norm=5.285375595092773, loss=1.262117624282837 +I0831 11:48:12.656674 139921981437696 logging_writer.py:48] [31500] global_step=31500, grad_norm=5.336551189422607, loss=1.4012649059295654 +I0831 11:48:39.779335 139921973044992 logging_writer.py:48] [31600] global_step=31600, grad_norm=5.116517066955566, loss=1.3187211751937866 +I0831 11:49:06.906994 139921981437696 logging_writer.py:48] [31700] global_step=31700, grad_norm=5.501377105712891, loss=1.4875959157943726 +I0831 11:49:34.099945 139921973044992 logging_writer.py:48] [31800] global_step=31800, grad_norm=5.645822525024414, loss=1.3838729858398438 +I0831 11:50:01.223986 139921981437696 logging_writer.py:48] [31900] global_step=31900, grad_norm=5.158445835113525, loss=1.3052215576171875 +I0831 11:50:28.369321 139921973044992 logging_writer.py:48] [32000] global_step=32000, grad_norm=5.2547125816345215, loss=1.373795747756958 +I0831 11:50:55.764033 139921981437696 logging_writer.py:48] [32100] global_step=32100, grad_norm=5.0896759033203125, loss=1.260233759880066 +I0831 11:51:22.886819 139921973044992 logging_writer.py:48] [32200] global_step=32200, grad_norm=5.141696929931641, loss=1.3746885061264038 +I0831 11:51:50.047598 139921981437696 logging_writer.py:48] [32300] global_step=32300, grad_norm=5.075890064239502, loss=1.3319129943847656 +I0831 11:52:17.204963 139921973044992 logging_writer.py:48] [32400] global_step=32400, grad_norm=5.1060590744018555, loss=1.3676745891571045 +I0831 11:52:44.363832 139921981437696 logging_writer.py:48] [32500] global_step=32500, grad_norm=4.707171440124512, loss=1.2621331214904785 +I0831 11:53:11.520153 139921973044992 logging_writer.py:48] [32600] global_step=32600, grad_norm=5.100059986114502, loss=1.291191577911377 +I0831 11:53:38.693369 139921981437696 logging_writer.py:48] [32700] global_step=32700, grad_norm=4.842025279998779, loss=1.3518191576004028 +I0831 11:54:05.814719 139921973044992 logging_writer.py:48] [32800] global_step=32800, grad_norm=4.840071678161621, loss=1.338102102279663 +I0831 11:54:32.941047 139921981437696 logging_writer.py:48] [32900] global_step=32900, grad_norm=4.893854141235352, loss=1.348619818687439 +I0831 11:55:00.142399 139921973044992 logging_writer.py:48] [33000] global_step=33000, grad_norm=5.056823253631592, loss=1.3219120502471924 +I0831 11:55:27.290831 139921981437696 logging_writer.py:48] [33100] global_step=33100, grad_norm=5.262635231018066, loss=1.2984070777893066 +I0831 11:55:54.637288 139921973044992 logging_writer.py:48] [33200] global_step=33200, grad_norm=4.661323070526123, loss=1.2864489555358887 +I0831 11:56:21.837561 139921981437696 logging_writer.py:48] [33300] global_step=33300, grad_norm=5.243496894836426, loss=1.2617566585540771 +I0831 11:56:48.982925 139921973044992 logging_writer.py:48] [33400] global_step=33400, grad_norm=4.586352825164795, loss=1.325636386871338 +I0831 11:57:16.130385 139921981437696 logging_writer.py:48] [33500] global_step=33500, grad_norm=4.599881649017334, loss=1.3848494291305542 +I0831 11:57:43.343936 139921973044992 logging_writer.py:48] [33600] global_step=33600, grad_norm=4.798940181732178, loss=1.3619968891143799 +I0831 11:58:10.477178 139921981437696 logging_writer.py:48] [33700] global_step=33700, grad_norm=4.520998001098633, loss=1.1590206623077393 +I0831 11:58:37.604510 139921973044992 logging_writer.py:48] [33800] global_step=33800, grad_norm=4.625613212585449, loss=1.3215343952178955 +I0831 11:59:04.788025 139921981437696 logging_writer.py:48] [33900] global_step=33900, grad_norm=4.527818202972412, loss=1.3259789943695068 +I0831 11:59:31.912271 139921973044992 logging_writer.py:48] [34000] global_step=34000, grad_norm=4.7602362632751465, loss=1.3617404699325562 +I0831 11:59:59.050849 139921981437696 logging_writer.py:48] [34100] global_step=34100, grad_norm=4.549473285675049, loss=1.2068984508514404 +I0831 12:00:26.228141 139921973044992 logging_writer.py:48] [34200] global_step=34200, grad_norm=4.735305309295654, loss=1.3046237230300903 +I0831 12:00:53.570113 139921981437696 logging_writer.py:48] [34300] global_step=34300, grad_norm=4.552788257598877, loss=1.298144817352295 +I0831 12:01:20.825428 139921973044992 logging_writer.py:48] [34400] global_step=34400, grad_norm=4.192521095275879, loss=1.185105800628662 +I0831 12:01:48.003092 139921981437696 logging_writer.py:48] [34500] global_step=34500, grad_norm=4.558001518249512, loss=1.294264554977417 +I0831 12:02:15.138071 139921973044992 logging_writer.py:48] [34600] global_step=34600, grad_norm=4.519845008850098, loss=1.317403793334961 +I0831 12:02:42.265488 139921981437696 logging_writer.py:48] [34700] global_step=34700, grad_norm=4.4508442878723145, loss=1.3302569389343262 +I0831 12:03:09.464908 139921973044992 logging_writer.py:48] [34800] global_step=34800, grad_norm=4.205676078796387, loss=1.169825792312622 +I0831 12:03:36.586245 139921981437696 logging_writer.py:48] [34900] global_step=34900, grad_norm=4.58042573928833, loss=1.3509093523025513 +I0831 12:04:03.720161 139921973044992 logging_writer.py:48] [35000] global_step=35000, grad_norm=4.436227321624756, loss=1.3542262315750122 +I0831 12:04:30.905965 139921981437696 logging_writer.py:48] [35100] global_step=35100, grad_norm=4.1063432693481445, loss=1.2658731937408447 +I0831 12:04:58.044490 139921973044992 logging_writer.py:48] [35200] global_step=35200, grad_norm=4.472425937652588, loss=1.3692381381988525 +I0831 12:05:25.190547 139921981437696 logging_writer.py:48] [35300] global_step=35300, grad_norm=4.153969764709473, loss=1.198420763015747 +I0831 12:05:52.630917 139921973044992 logging_writer.py:48] [35400] global_step=35400, grad_norm=4.284669876098633, loss=1.3462679386138916 +I0831 12:06:19.776799 139921981437696 logging_writer.py:48] [35500] global_step=35500, grad_norm=4.3600897789001465, loss=1.2508689165115356 +I0831 12:06:46.936126 139921973044992 logging_writer.py:48] [35600] global_step=35600, grad_norm=4.424654006958008, loss=1.2479652166366577 +I0831 12:07:14.143855 139921981437696 logging_writer.py:48] [35700] global_step=35700, grad_norm=4.1384501457214355, loss=1.300716757774353 +I0831 12:07:41.258309 139921973044992 logging_writer.py:48] [35800] global_step=35800, grad_norm=4.190006732940674, loss=1.2789454460144043 +I0831 12:08:08.390031 139921981437696 logging_writer.py:48] [35900] global_step=35900, grad_norm=4.565182209014893, loss=1.3138452768325806 +I0831 12:08:24.018189 140117123622080 spec.py:333] Evaluating on the training split. +I0831 12:08:34.997783 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 12:08:45.301233 140117123622080 spec.py:363] Evaluating on the test split. +I0831 12:08:46.169492 140117123622080 submission_runner.py:516] Time since start: 10247.51s, Step: 35959, {'train/accuracy': Array(0.85227996, dtype=float32), 'train/loss': Array(0.538871, dtype=float32), 'validation/accuracy': Array(0.72279996, dtype=float32), 'validation/loss': Array(1.1272918, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.59010005, dtype=float32), 'test/loss': Array(1.8540716, dtype=float32), 'test/num_examples': 10000, 'score': 10032.998893737793, 'total_duration': 10247.509325027466, 'accumulated_submission_time': 10032.998893737793, 'accumulated_eval_time': 213.9108157157898, 'accumulated_logging_time': 0.3050100803375244} +I0831 12:08:46.234897 139921973044992 logging_writer.py:48] [35959] accumulated_eval_time=213.911, accumulated_logging_time=0.30501, accumulated_submission_time=10033, global_step=35959, preemption_count=0, score=10033, test/accuracy=0.5901000499725342, test/loss=1.8540716171264648, test/num_examples=10000, total_duration=10247.5, train/accuracy=0.8522799611091614, train/loss=0.5388709902763367, validation/accuracy=0.7227999567985535, validation/loss=1.1272917985916138, validation/num_examples=50000 +I0831 12:08:57.778723 139921981437696 logging_writer.py:48] [36000] global_step=36000, grad_norm=4.20889949798584, loss=1.295737862586975 +I0831 12:09:24.892527 139921973044992 logging_writer.py:48] [36100] global_step=36100, grad_norm=4.147944450378418, loss=1.2280350923538208 +I0831 12:09:52.022734 139921981437696 logging_writer.py:48] [36200] global_step=36200, grad_norm=4.187687873840332, loss=1.2583427429199219 +I0831 12:10:19.187777 139921973044992 logging_writer.py:48] [36300] global_step=36300, grad_norm=4.332421779632568, loss=1.2669442892074585 +I0831 12:10:46.543876 139921981437696 logging_writer.py:48] [36400] global_step=36400, grad_norm=4.346745491027832, loss=1.2571616172790527 +I0831 12:11:13.656511 139921973044992 logging_writer.py:48] [36500] global_step=36500, grad_norm=4.4550089836120605, loss=1.2742817401885986 +I0831 12:11:40.882476 139921981437696 logging_writer.py:48] [36600] global_step=36600, grad_norm=4.033005237579346, loss=1.3168771266937256 +I0831 12:12:08.007813 139921973044992 logging_writer.py:48] [36700] global_step=36700, grad_norm=4.167398929595947, loss=1.387932538986206 +I0831 12:12:35.146059 139921981437696 logging_writer.py:48] [36800] global_step=36800, grad_norm=4.182551383972168, loss=1.3181161880493164 +I0831 12:13:02.359377 139921973044992 logging_writer.py:48] [36900] global_step=36900, grad_norm=4.306579113006592, loss=1.3641237020492554 +I0831 12:13:29.522636 139921981437696 logging_writer.py:48] [37000] global_step=37000, grad_norm=3.765108823776245, loss=1.2041950225830078 +I0831 12:13:56.644252 139921973044992 logging_writer.py:48] [37100] global_step=37100, grad_norm=4.1901726722717285, loss=1.2142888307571411 +I0831 12:14:23.856267 139921981437696 logging_writer.py:48] [37200] global_step=37200, grad_norm=4.191021919250488, loss=1.2689383029937744 +I0831 12:14:50.968385 139921973044992 logging_writer.py:48] [37300] global_step=37300, grad_norm=4.128292083740234, loss=1.342078685760498 +I0831 12:15:18.144078 139921981437696 logging_writer.py:48] [37400] global_step=37400, grad_norm=4.099293231964111, loss=1.2561657428741455 +I0831 12:15:45.545095 139921973044992 logging_writer.py:48] [37500] global_step=37500, grad_norm=4.128584384918213, loss=1.328758716583252 +I0831 12:16:12.666006 139921981437696 logging_writer.py:48] [37600] global_step=37600, grad_norm=4.063676834106445, loss=1.2648398876190186 +I0831 12:16:39.801412 139921973044992 logging_writer.py:48] [37700] global_step=37700, grad_norm=4.030208587646484, loss=1.2339768409729004 +I0831 12:17:07.012846 139921981437696 logging_writer.py:48] [37800] global_step=37800, grad_norm=4.2506489753723145, loss=1.291438102722168 +I0831 12:17:34.121759 139921973044992 logging_writer.py:48] [37900] global_step=37900, grad_norm=4.355396747589111, loss=1.2763267755508423 +I0831 12:18:01.247632 139921981437696 logging_writer.py:48] [38000] global_step=38000, grad_norm=4.476174354553223, loss=1.3179142475128174 +I0831 12:18:28.445848 139921973044992 logging_writer.py:48] [38100] global_step=38100, grad_norm=4.205198764801025, loss=1.309321641921997 +I0831 12:18:55.612597 139921981437696 logging_writer.py:48] [38200] global_step=38200, grad_norm=4.412575721740723, loss=1.2881683111190796 +I0831 12:19:22.733898 139921973044992 logging_writer.py:48] [38300] global_step=38300, grad_norm=4.18325662612915, loss=1.308577537536621 +I0831 12:19:49.962475 139921981437696 logging_writer.py:48] [38400] global_step=38400, grad_norm=4.0725531578063965, loss=1.297827124595642 +I0831 12:20:17.074592 139921973044992 logging_writer.py:48] [38500] global_step=38500, grad_norm=3.93291974067688, loss=1.1758286952972412 +I0831 12:20:44.445775 139921981437696 logging_writer.py:48] [38600] global_step=38600, grad_norm=4.050321578979492, loss=1.2662596702575684 +I0831 12:21:11.624705 139921973044992 logging_writer.py:48] [38700] global_step=38700, grad_norm=3.8490729331970215, loss=1.1775622367858887 +I0831 12:21:38.748711 139921981437696 logging_writer.py:48] [38800] global_step=38800, grad_norm=4.278644561767578, loss=1.2504998445510864 +I0831 12:22:05.884104 139921973044992 logging_writer.py:48] [38900] global_step=38900, grad_norm=4.142755508422852, loss=1.2898879051208496 +I0831 12:22:33.126429 139921981437696 logging_writer.py:48] [39000] global_step=39000, grad_norm=4.014084339141846, loss=1.1732808351516724 +I0831 12:23:00.261519 139921973044992 logging_writer.py:48] [39100] global_step=39100, grad_norm=4.159514904022217, loss=1.3284045457839966 +I0831 12:23:27.359034 139921981437696 logging_writer.py:48] [39200] global_step=39200, grad_norm=4.13672399520874, loss=1.2364813089370728 +I0831 12:23:54.569314 139921973044992 logging_writer.py:48] [39300] global_step=39300, grad_norm=3.880845308303833, loss=1.237086534500122 +I0831 12:24:21.707346 139921981437696 logging_writer.py:48] [39400] global_step=39400, grad_norm=3.8092620372772217, loss=1.1478583812713623 +I0831 12:24:48.834065 139921973044992 logging_writer.py:48] [39500] global_step=39500, grad_norm=4.213837623596191, loss=1.2179760932922363 +I0831 12:25:16.244582 139921981437696 logging_writer.py:48] [39600] global_step=39600, grad_norm=4.01539945602417, loss=1.2585365772247314 +I0831 12:25:43.344376 139921973044992 logging_writer.py:48] [39700] global_step=39700, grad_norm=4.005892276763916, loss=1.247807502746582 +I0831 12:26:10.477813 139921981437696 logging_writer.py:48] [39800] global_step=39800, grad_norm=3.6523702144622803, loss=1.1494145393371582 +I0831 12:26:37.674543 139921973044992 logging_writer.py:48] [39900] global_step=39900, grad_norm=3.8997113704681396, loss=1.2695133686065674 +I0831 12:27:04.799939 139921981437696 logging_writer.py:48] [40000] global_step=40000, grad_norm=4.055899143218994, loss=1.2207326889038086 +I0831 12:27:31.908724 139921973044992 logging_writer.py:48] [40100] global_step=40100, grad_norm=4.05573034286499, loss=1.2420049905776978 +I0831 12:27:59.101491 139921981437696 logging_writer.py:48] [40200] global_step=40200, grad_norm=3.7824559211730957, loss=1.177244782447815 +I0831 12:28:26.225884 139921973044992 logging_writer.py:48] [40300] global_step=40300, grad_norm=3.9004335403442383, loss=1.2059837579727173 +I0831 12:28:53.354022 139921981437696 logging_writer.py:48] [40400] global_step=40400, grad_norm=3.835263252258301, loss=1.2575255632400513 +I0831 12:29:20.542331 139921973044992 logging_writer.py:48] [40500] global_step=40500, grad_norm=4.093811511993408, loss=1.2450588941574097 +I0831 12:29:47.652086 139921981437696 logging_writer.py:48] [40600] global_step=40600, grad_norm=3.881366729736328, loss=1.204058289527893 +I0831 12:30:14.969296 139921973044992 logging_writer.py:48] [40700] global_step=40700, grad_norm=3.8351757526397705, loss=1.2053323984146118 +I0831 12:30:42.145714 139921981437696 logging_writer.py:48] [40800] global_step=40800, grad_norm=4.054604530334473, loss=1.3762013912200928 +I0831 12:31:09.285847 139921973044992 logging_writer.py:48] [40900] global_step=40900, grad_norm=3.9939475059509277, loss=1.2447409629821777 +I0831 12:31:36.398717 139921981437696 logging_writer.py:48] [41000] global_step=41000, grad_norm=3.879377603530884, loss=1.2561140060424805 +I0831 12:32:03.610169 139921973044992 logging_writer.py:48] [41100] global_step=41100, grad_norm=3.6541547775268555, loss=1.1486544609069824 +I0831 12:32:30.743958 139921981437696 logging_writer.py:48] [41200] global_step=41200, grad_norm=3.65122127532959, loss=1.1559855937957764 +I0831 12:32:57.879310 139921973044992 logging_writer.py:48] [41300] global_step=41300, grad_norm=3.960081100463867, loss=1.1922838687896729 +I0831 12:33:25.068385 139921981437696 logging_writer.py:48] [41400] global_step=41400, grad_norm=3.803696870803833, loss=1.1998052597045898 +I0831 12:33:52.223909 139921973044992 logging_writer.py:48] [41500] global_step=41500, grad_norm=3.946402072906494, loss=1.264918565750122 +I0831 12:34:19.361503 139921981437696 logging_writer.py:48] [41600] global_step=41600, grad_norm=3.9684009552001953, loss=1.2477939128875732 +I0831 12:34:46.585584 139921973044992 logging_writer.py:48] [41700] global_step=41700, grad_norm=3.964876890182495, loss=1.2460286617279053 +I0831 12:35:13.892904 139921981437696 logging_writer.py:48] [41800] global_step=41800, grad_norm=3.8166441917419434, loss=1.3088868856430054 +I0831 12:35:41.002326 139921973044992 logging_writer.py:48] [41900] global_step=41900, grad_norm=3.9835963249206543, loss=1.2928441762924194 +I0831 12:36:08.192360 139921981437696 logging_writer.py:48] [42000] global_step=42000, grad_norm=4.047582626342773, loss=1.2919238805770874 +I0831 12:36:35.314109 139921973044992 logging_writer.py:48] [42100] global_step=42100, grad_norm=3.8659348487854004, loss=1.191026210784912 +I0831 12:37:02.488047 139921981437696 logging_writer.py:48] [42200] global_step=42200, grad_norm=3.6807844638824463, loss=1.2429823875427246 +I0831 12:37:29.674840 139921973044992 logging_writer.py:48] [42300] global_step=42300, grad_norm=3.909884452819824, loss=1.183009386062622 +I0831 12:37:56.799801 139921981437696 logging_writer.py:48] [42400] global_step=42400, grad_norm=3.9391720294952393, loss=1.2722432613372803 +I0831 12:38:23.975216 139921973044992 logging_writer.py:48] [42500] global_step=42500, grad_norm=3.724644422531128, loss=1.1566801071166992 +I0831 12:38:51.171405 139921981437696 logging_writer.py:48] [42600] global_step=42600, grad_norm=3.6974968910217285, loss=1.2084400653839111 +I0831 12:39:18.320669 139921973044992 logging_writer.py:48] [42700] global_step=42700, grad_norm=3.6626546382904053, loss=1.2575581073760986 +I0831 12:39:45.448506 139921981437696 logging_writer.py:48] [42800] global_step=42800, grad_norm=3.8989973068237305, loss=1.2413718700408936 +I0831 12:40:12.847841 139921973044992 logging_writer.py:48] [42900] global_step=42900, grad_norm=4.058259963989258, loss=1.3022229671478271 +I0831 12:40:39.937497 139921981437696 logging_writer.py:48] [43000] global_step=43000, grad_norm=4.105683326721191, loss=1.2063617706298828 +I0831 12:41:07.059417 139921973044992 logging_writer.py:48] [43100] global_step=43100, grad_norm=3.785878896713257, loss=1.2155444622039795 +I0831 12:41:34.256969 139921981437696 logging_writer.py:48] [43200] global_step=43200, grad_norm=4.103187561035156, loss=1.2614989280700684 +I0831 12:42:01.383259 139921973044992 logging_writer.py:48] [43300] global_step=43300, grad_norm=3.7373056411743164, loss=1.2178308963775635 +I0831 12:42:02.320527 140117123622080 spec.py:333] Evaluating on the training split. +I0831 12:42:12.295601 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 12:42:21.768241 140117123622080 spec.py:363] Evaluating on the test split. +I0831 12:42:22.640969 140117123622080 submission_runner.py:516] Time since start: 12263.98s, Step: 43305, {'train/accuracy': Array(0.87549824, dtype=float32), 'train/loss': Array(0.461255, dtype=float32), 'validation/accuracy': Array(0.72856, dtype=float32), 'validation/loss': Array(1.110315, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6014, dtype=float32), 'test/loss': Array(1.8353652, dtype=float32), 'test/num_examples': 10000, 'score': 12029.003738880157, 'total_duration': 12263.979522943497, 'accumulated_submission_time': 12029.003738880157, 'accumulated_eval_time': 234.2279772758484, 'accumulated_logging_time': 0.3990471363067627} +I0831 12:42:22.708001 139921981437696 logging_writer.py:48] [43305] accumulated_eval_time=234.228, accumulated_logging_time=0.399047, accumulated_submission_time=12029, global_step=43305, preemption_count=0, score=12029, test/accuracy=0.6014000177383423, test/loss=1.8353651762008667, test/num_examples=10000, total_duration=12264, train/accuracy=0.8754982352256775, train/loss=0.4612550139427185, validation/accuracy=0.7285599708557129, validation/loss=1.110314965248108, validation/num_examples=50000 +I0831 12:42:48.833587 139921973044992 logging_writer.py:48] [43400] global_step=43400, grad_norm=3.758774995803833, loss=1.2211493253707886 +I0831 12:43:15.995599 139921981437696 logging_writer.py:48] [43500] global_step=43500, grad_norm=3.8383688926696777, loss=1.3141534328460693 +I0831 12:43:43.099672 139921973044992 logging_writer.py:48] [43600] global_step=43600, grad_norm=3.840034008026123, loss=1.201387643814087 +I0831 12:44:10.233377 139921981437696 logging_writer.py:48] [43700] global_step=43700, grad_norm=3.802267551422119, loss=1.2862813472747803 +I0831 12:44:37.442019 139921973044992 logging_writer.py:48] [43800] global_step=43800, grad_norm=3.905487537384033, loss=1.139348030090332 +I0831 12:45:04.842673 139921981437696 logging_writer.py:48] [43900] global_step=43900, grad_norm=3.9180703163146973, loss=1.2099111080169678 +I0831 12:45:31.959470 139921973044992 logging_writer.py:48] [44000] global_step=44000, grad_norm=3.7666146755218506, loss=1.1831305027008057 +I0831 12:45:59.160781 139921981437696 logging_writer.py:48] [44100] global_step=44100, grad_norm=3.934126853942871, loss=1.237207293510437 +I0831 12:46:26.269600 139921973044992 logging_writer.py:48] [44200] global_step=44200, grad_norm=3.684337615966797, loss=1.208390712738037 +I0831 12:46:53.412480 139921981437696 logging_writer.py:48] [44300] global_step=44300, grad_norm=3.996690034866333, loss=1.2787346839904785 +I0831 12:47:20.624670 139921973044992 logging_writer.py:48] [44400] global_step=44400, grad_norm=3.924163818359375, loss=1.243706226348877 +I0831 12:47:47.727102 139921981437696 logging_writer.py:48] [44500] global_step=44500, grad_norm=3.6839964389801025, loss=1.1668802499771118 +I0831 12:48:14.820322 139921973044992 logging_writer.py:48] [44600] global_step=44600, grad_norm=3.646383047103882, loss=1.1201465129852295 +I0831 12:48:42.010682 139921981437696 logging_writer.py:48] [44700] global_step=44700, grad_norm=3.77172589302063, loss=1.2500965595245361 +I0831 12:49:09.154320 139921973044992 logging_writer.py:48] [44800] global_step=44800, grad_norm=3.558877468109131, loss=1.1874282360076904 +I0831 12:49:36.296138 139921981437696 logging_writer.py:48] [44900] global_step=44900, grad_norm=3.690375804901123, loss=1.208626389503479 +I0831 12:50:03.695808 139921973044992 logging_writer.py:48] [45000] global_step=45000, grad_norm=3.8533034324645996, loss=1.1994409561157227 +I0831 12:50:30.848026 139921981437696 logging_writer.py:48] [45100] global_step=45100, grad_norm=3.742692708969116, loss=1.2166876792907715 +I0831 12:50:57.975621 139921973044992 logging_writer.py:48] [45200] global_step=45200, grad_norm=3.845475196838379, loss=1.0834128856658936 +I0831 12:51:25.169368 139921981437696 logging_writer.py:48] [45300] global_step=45300, grad_norm=3.651200294494629, loss=1.0973831415176392 +I0831 12:51:52.290460 139921973044992 logging_writer.py:48] [45400] global_step=45400, grad_norm=4.042835235595703, loss=1.264158010482788 +I0831 12:52:19.453541 139921981437696 logging_writer.py:48] [45500] global_step=45500, grad_norm=3.895289897918701, loss=1.2336699962615967 +I0831 12:52:46.672436 139921973044992 logging_writer.py:48] [45600] global_step=45600, grad_norm=3.9634361267089844, loss=1.265818476676941 +I0831 12:53:13.817578 139921981437696 logging_writer.py:48] [45700] global_step=45700, grad_norm=3.6631569862365723, loss=1.2126045227050781 +I0831 12:53:40.932258 139921973044992 logging_writer.py:48] [45800] global_step=45800, grad_norm=3.7708585262298584, loss=1.1736377477645874 +I0831 12:54:08.122269 139921981437696 logging_writer.py:48] [45900] global_step=45900, grad_norm=4.161580562591553, loss=1.298155665397644 +I0831 12:54:35.476387 139921973044992 logging_writer.py:48] [46000] global_step=46000, grad_norm=3.975205421447754, loss=1.2058007717132568 +I0831 12:55:02.574496 139921981437696 logging_writer.py:48] [46100] global_step=46100, grad_norm=3.816837787628174, loss=1.2613496780395508 +I0831 12:55:29.772361 139921973044992 logging_writer.py:48] [46200] global_step=46200, grad_norm=3.848348379135132, loss=1.2571221590042114 +I0831 12:55:56.913572 139921981437696 logging_writer.py:48] [46300] global_step=46300, grad_norm=3.633871078491211, loss=1.2187614440917969 +I0831 12:56:24.063732 139921973044992 logging_writer.py:48] [46400] global_step=46400, grad_norm=3.8177499771118164, loss=1.139207363128662 +I0831 12:56:51.259900 139921981437696 logging_writer.py:48] [46500] global_step=46500, grad_norm=3.8533263206481934, loss=1.1855034828186035 +I0831 12:57:18.388047 139921973044992 logging_writer.py:48] [46600] global_step=46600, grad_norm=3.7138874530792236, loss=1.1475058794021606 +I0831 12:57:45.497115 139921981437696 logging_writer.py:48] [46700] global_step=46700, grad_norm=3.8574652671813965, loss=1.2164347171783447 +I0831 12:58:12.669606 139921973044992 logging_writer.py:48] [46800] global_step=46800, grad_norm=3.855808734893799, loss=1.216722846031189 +I0831 12:58:39.814184 139921981437696 logging_writer.py:48] [46900] global_step=46900, grad_norm=3.8240184783935547, loss=1.1806303262710571 +I0831 12:59:06.914835 139921973044992 logging_writer.py:48] [47000] global_step=47000, grad_norm=3.751664400100708, loss=1.19265878200531 +I0831 12:59:34.306400 139921981437696 logging_writer.py:48] [47100] global_step=47100, grad_norm=3.5360891819000244, loss=1.1235671043395996 +I0831 13:00:01.445533 139921973044992 logging_writer.py:48] [47200] global_step=47200, grad_norm=3.7130393981933594, loss=1.231561303138733 +I0831 13:00:28.555427 139921981437696 logging_writer.py:48] [47300] global_step=47300, grad_norm=3.892523765563965, loss=1.1290243864059448 +I0831 13:00:55.719560 139921973044992 logging_writer.py:48] [47400] global_step=47400, grad_norm=3.7779324054718018, loss=1.1525757312774658 +I0831 13:01:22.852639 139921981437696 logging_writer.py:48] [47500] global_step=47500, grad_norm=3.9430336952209473, loss=1.1644970178604126 +I0831 13:01:49.985494 139921973044992 logging_writer.py:48] [47600] global_step=47600, grad_norm=3.727900505065918, loss=1.159172773361206 +I0831 13:02:17.196195 139921981437696 logging_writer.py:48] [47700] global_step=47700, grad_norm=4.063667297363281, loss=1.2984592914581299 +I0831 13:02:44.309406 139921973044992 logging_writer.py:48] [47800] global_step=47800, grad_norm=3.70924711227417, loss=1.1685614585876465 +I0831 13:03:11.462556 139921981437696 logging_writer.py:48] [47900] global_step=47900, grad_norm=3.7909657955169678, loss=1.1385629177093506 +I0831 13:03:38.679863 139921973044992 logging_writer.py:48] [48000] global_step=48000, grad_norm=3.665463447570801, loss=1.1005070209503174 +I0831 13:04:06.070457 139921981437696 logging_writer.py:48] [48100] global_step=48100, grad_norm=3.959243059158325, loss=1.270626187324524 +I0831 13:04:33.181965 139921973044992 logging_writer.py:48] [48200] global_step=48200, grad_norm=3.6838219165802, loss=1.2227424383163452 +I0831 13:05:00.351938 139921981437696 logging_writer.py:48] [48300] global_step=48300, grad_norm=3.951263904571533, loss=1.2738168239593506 +I0831 13:05:27.453322 139921973044992 logging_writer.py:48] [48400] global_step=48400, grad_norm=3.831524610519409, loss=1.2564189434051514 +I0831 13:05:54.564762 139921981437696 logging_writer.py:48] [48500] global_step=48500, grad_norm=3.6905150413513184, loss=1.2040053606033325 +I0831 13:06:21.730543 139921973044992 logging_writer.py:48] [48600] global_step=48600, grad_norm=3.7551534175872803, loss=1.1426451206207275 +I0831 13:06:48.844764 139921981437696 logging_writer.py:48] [48700] global_step=48700, grad_norm=3.548980236053467, loss=1.1523898839950562 +I0831 13:07:15.991983 139921973044992 logging_writer.py:48] [48800] global_step=48800, grad_norm=4.064699649810791, loss=1.182087779045105 +I0831 13:07:43.184684 139921981437696 logging_writer.py:48] [48900] global_step=48900, grad_norm=3.819570541381836, loss=1.2134461402893066 +I0831 13:08:10.336171 139921973044992 logging_writer.py:48] [49000] global_step=49000, grad_norm=3.8589844703674316, loss=1.2861207723617554 +I0831 13:08:37.467447 139921981437696 logging_writer.py:48] [49100] global_step=49100, grad_norm=4.076883792877197, loss=1.101996660232544 +I0831 13:09:04.880295 139921973044992 logging_writer.py:48] [49200] global_step=49200, grad_norm=3.5611021518707275, loss=1.1566298007965088 +I0831 13:09:32.013833 139921981437696 logging_writer.py:48] [49300] global_step=49300, grad_norm=4.0794453620910645, loss=1.1880989074707031 +I0831 13:09:59.163721 139921973044992 logging_writer.py:48] [49400] global_step=49400, grad_norm=3.722257137298584, loss=1.284245252609253 +I0831 13:10:26.361722 139921981437696 logging_writer.py:48] [49500] global_step=49500, grad_norm=3.6199259757995605, loss=1.2544176578521729 +I0831 13:10:53.463009 139921973044992 logging_writer.py:48] [49600] global_step=49600, grad_norm=3.6106789112091064, loss=1.1426619291305542 +I0831 13:11:20.594751 139921981437696 logging_writer.py:48] [49700] global_step=49700, grad_norm=3.700551748275757, loss=1.1931248903274536 +I0831 13:11:47.799355 139921973044992 logging_writer.py:48] [49800] global_step=49800, grad_norm=4.162051200866699, loss=1.2766393423080444 +I0831 13:12:14.951845 139921981437696 logging_writer.py:48] [49900] global_step=49900, grad_norm=3.8201911449432373, loss=1.1742186546325684 +I0831 13:12:42.072209 139921973044992 logging_writer.py:48] [50000] global_step=50000, grad_norm=4.033286094665527, loss=1.196784496307373 +I0831 13:13:09.268753 139921981437696 logging_writer.py:48] [50100] global_step=50100, grad_norm=3.7190630435943604, loss=1.1439299583435059 +I0831 13:13:36.579819 139921973044992 logging_writer.py:48] [50200] global_step=50200, grad_norm=3.9788906574249268, loss=1.150423526763916 +I0831 13:14:03.690792 139921981437696 logging_writer.py:48] [50300] global_step=50300, grad_norm=3.800527811050415, loss=1.2322545051574707 +I0831 13:14:30.911051 139921973044992 logging_writer.py:48] [50400] global_step=50400, grad_norm=3.844547748565674, loss=1.2171235084533691 +I0831 13:14:58.005284 139921981437696 logging_writer.py:48] [50500] global_step=50500, grad_norm=3.5295474529266357, loss=1.1034694910049438 +I0831 13:15:25.151910 139921973044992 logging_writer.py:48] [50600] global_step=50600, grad_norm=3.8341891765594482, loss=1.1428818702697754 +I0831 13:15:38.871582 140117123622080 spec.py:333] Evaluating on the training split. +I0831 13:15:48.259579 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 13:15:57.915292 140117123622080 spec.py:363] Evaluating on the test split. +I0831 13:15:58.812527 140117123622080 submission_runner.py:516] Time since start: 14280.15s, Step: 50652, {'train/accuracy': Array(0.8825733, dtype=float32), 'train/loss': Array(0.42156616, dtype=float32), 'validation/accuracy': Array(0.73304, dtype=float32), 'validation/loss': Array(1.099628, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6049, dtype=float32), 'test/loss': Array(1.8242571, dtype=float32), 'test/num_examples': 10000, 'score': 14025.104043245316, 'total_duration': 14280.152136325836, 'accumulated_submission_time': 14025.104043245316, 'accumulated_eval_time': 254.16670107841492, 'accumulated_logging_time': 0.4768648147583008} +I0831 13:15:58.865587 139921981437696 logging_writer.py:48] [50652] accumulated_eval_time=254.167, accumulated_logging_time=0.476865, accumulated_submission_time=14025.1, global_step=50652, preemption_count=0, score=14025.1, test/accuracy=0.6049000024795532, test/loss=1.8242571353912354, test/num_examples=10000, total_duration=14280.2, train/accuracy=0.8825733065605164, train/loss=0.4215661585330963, validation/accuracy=0.7330399751663208, validation/loss=1.09962797164917, validation/num_examples=50000 +I0831 13:16:12.395057 139921973044992 logging_writer.py:48] [50700] global_step=50700, grad_norm=3.8564460277557373, loss=1.1265738010406494 +I0831 13:16:39.475228 139921981437696 logging_writer.py:48] [50800] global_step=50800, grad_norm=3.7209386825561523, loss=1.2232613563537598 +I0831 13:17:06.580659 139921973044992 logging_writer.py:48] [50900] global_step=50900, grad_norm=3.9475173950195312, loss=1.1787421703338623 +I0831 13:17:33.765849 139921981437696 logging_writer.py:48] [51000] global_step=51000, grad_norm=3.7816567420959473, loss=1.1190156936645508 +I0831 13:18:00.870675 139921973044992 logging_writer.py:48] [51100] global_step=51100, grad_norm=4.438498020172119, loss=1.2693188190460205 +I0831 13:18:28.007091 139921981437696 logging_writer.py:48] [51200] global_step=51200, grad_norm=3.6849558353424072, loss=1.1518340110778809 +I0831 13:18:55.396752 139921973044992 logging_writer.py:48] [51300] global_step=51300, grad_norm=3.8855135440826416, loss=1.174688696861267 +I0831 13:19:22.520631 139921981437696 logging_writer.py:48] [51400] global_step=51400, grad_norm=3.557337999343872, loss=1.111802101135254 +I0831 13:19:49.632914 139921973044992 logging_writer.py:48] [51500] global_step=51500, grad_norm=3.8223540782928467, loss=1.2042450904846191 +I0831 13:20:16.825700 139921981437696 logging_writer.py:48] [51600] global_step=51600, grad_norm=3.8021438121795654, loss=1.1663926839828491 +I0831 13:20:43.935636 139921973044992 logging_writer.py:48] [51700] global_step=51700, grad_norm=3.75398850440979, loss=1.1431139707565308 +I0831 13:21:11.071697 139921981437696 logging_writer.py:48] [51800] global_step=51800, grad_norm=3.6831185817718506, loss=1.171515941619873 +I0831 13:21:38.245535 139921973044992 logging_writer.py:48] [51900] global_step=51900, grad_norm=3.65390682220459, loss=1.1556968688964844 +I0831 13:22:05.361962 139921981437696 logging_writer.py:48] [52000] global_step=52000, grad_norm=4.12195348739624, loss=1.0937387943267822 +I0831 13:22:32.500060 139921973044992 logging_writer.py:48] [52100] global_step=52100, grad_norm=4.095597743988037, loss=1.1905462741851807 +I0831 13:22:59.714375 139921981437696 logging_writer.py:48] [52200] global_step=52200, grad_norm=3.925208568572998, loss=1.1882750988006592 +I0831 13:23:27.095614 139921973044992 logging_writer.py:48] [52300] global_step=52300, grad_norm=3.871145248413086, loss=1.2405598163604736 +I0831 13:23:54.235255 139921981437696 logging_writer.py:48] [52400] global_step=52400, grad_norm=3.66726016998291, loss=1.0790393352508545 +I0831 13:24:21.417216 139921973044992 logging_writer.py:48] [52500] global_step=52500, grad_norm=4.031847953796387, loss=1.198161005973816 +I0831 13:24:48.525769 139921981437696 logging_writer.py:48] [52600] global_step=52600, grad_norm=3.805248260498047, loss=1.1680519580841064 +I0831 13:25:15.650002 139921973044992 logging_writer.py:48] [52700] global_step=52700, grad_norm=4.176838397979736, loss=1.2107280492782593 +I0831 13:25:42.839518 139921981437696 logging_writer.py:48] [52800] global_step=52800, grad_norm=3.856381416320801, loss=1.222642421722412 +I0831 13:26:09.973017 139921973044992 logging_writer.py:48] [52900] global_step=52900, grad_norm=3.828699827194214, loss=1.1992026567459106 +I0831 13:26:37.114919 139921981437696 logging_writer.py:48] [53000] global_step=53000, grad_norm=4.061285972595215, loss=1.2435791492462158 +I0831 13:27:04.310390 139921973044992 logging_writer.py:48] [53100] global_step=53100, grad_norm=3.774291753768921, loss=1.0716408491134644 +I0831 13:27:31.425753 139921981437696 logging_writer.py:48] [53200] global_step=53200, grad_norm=3.7737669944763184, loss=1.104125738143921 +I0831 13:27:58.547856 139921973044992 logging_writer.py:48] [53300] global_step=53300, grad_norm=4.0174336433410645, loss=1.2508738040924072 +I0831 13:28:25.968504 139921981437696 logging_writer.py:48] [53400] global_step=53400, grad_norm=3.9714999198913574, loss=1.1899487972259521 +I0831 13:28:53.086027 139921973044992 logging_writer.py:48] [53500] global_step=53500, grad_norm=4.311335563659668, loss=1.2024060487747192 +I0831 13:29:20.214003 139921981437696 logging_writer.py:48] [53600] global_step=53600, grad_norm=4.308870792388916, loss=1.2274600267410278 +I0831 13:29:47.389422 139921973044992 logging_writer.py:48] [53700] global_step=53700, grad_norm=3.9699792861938477, loss=1.2339712381362915 +I0831 13:30:14.493676 139921981437696 logging_writer.py:48] [53800] global_step=53800, grad_norm=4.078004360198975, loss=1.1912181377410889 +I0831 13:30:41.614540 139921973044992 logging_writer.py:48] [53900] global_step=53900, grad_norm=3.831033945083618, loss=1.2666468620300293 +I0831 13:31:08.788411 139921981437696 logging_writer.py:48] [54000] global_step=54000, grad_norm=4.032386302947998, loss=1.1796027421951294 +I0831 13:31:35.899281 139921973044992 logging_writer.py:48] [54100] global_step=54100, grad_norm=4.352496147155762, loss=1.2586601972579956 +I0831 13:32:03.024032 139921981437696 logging_writer.py:48] [54200] global_step=54200, grad_norm=3.6723556518554688, loss=1.1189687252044678 +I0831 13:32:30.202784 139921973044992 logging_writer.py:48] [54300] global_step=54300, grad_norm=4.552175045013428, loss=1.2961983680725098 +I0831 13:32:57.522696 139921981437696 logging_writer.py:48] [54400] global_step=54400, grad_norm=4.091862678527832, loss=1.2088381052017212 +I0831 13:33:24.626670 139921973044992 logging_writer.py:48] [54500] global_step=54500, grad_norm=3.8744678497314453, loss=1.1262266635894775 +I0831 13:33:51.799132 139921981437696 logging_writer.py:48] [54600] global_step=54600, grad_norm=3.652543544769287, loss=1.0134422779083252 +I0831 13:34:18.937994 139921973044992 logging_writer.py:48] [54700] global_step=54700, grad_norm=3.895554542541504, loss=1.1288979053497314 +I0831 13:34:46.040562 139921981437696 logging_writer.py:48] [54800] global_step=54800, grad_norm=3.6525681018829346, loss=1.0719738006591797 +I0831 13:35:13.231813 139921973044992 logging_writer.py:48] [54900] global_step=54900, grad_norm=3.7272417545318604, loss=1.0911667346954346 +I0831 13:35:40.318574 139921981437696 logging_writer.py:48] [55000] global_step=55000, grad_norm=3.862983465194702, loss=1.1250967979431152 +I0831 13:36:07.450498 139921973044992 logging_writer.py:48] [55100] global_step=55100, grad_norm=3.9354248046875, loss=1.2094066143035889 +I0831 13:36:34.623650 139921981437696 logging_writer.py:48] [55200] global_step=55200, grad_norm=4.147250175476074, loss=1.2374930381774902 +I0831 13:37:01.729660 139921973044992 logging_writer.py:48] [55300] global_step=55300, grad_norm=3.853046178817749, loss=1.0989837646484375 +I0831 13:37:28.859025 139921981437696 logging_writer.py:48] [55400] global_step=55400, grad_norm=3.86596941947937, loss=1.1312689781188965 +I0831 13:37:56.255115 139921973044992 logging_writer.py:48] [55500] global_step=55500, grad_norm=4.146842002868652, loss=1.1317919492721558 +I0831 13:38:23.377129 139921981437696 logging_writer.py:48] [55600] global_step=55600, grad_norm=4.116367340087891, loss=1.2117164134979248 +I0831 13:38:50.502391 139921973044992 logging_writer.py:48] [55700] global_step=55700, grad_norm=3.8638272285461426, loss=1.112353801727295 +I0831 13:39:17.695424 139921981437696 logging_writer.py:48] [55800] global_step=55800, grad_norm=3.8218631744384766, loss=1.1377207040786743 +I0831 13:39:44.832883 139921973044992 logging_writer.py:48] [55900] global_step=55900, grad_norm=3.7054080963134766, loss=1.0731494426727295 +I0831 13:40:11.950986 139921981437696 logging_writer.py:48] [56000] global_step=56000, grad_norm=3.8027780055999756, loss=1.1962592601776123 +I0831 13:40:39.131671 139921973044992 logging_writer.py:48] [56100] global_step=56100, grad_norm=4.01143741607666, loss=1.169816017150879 +I0831 13:41:06.261523 139921981437696 logging_writer.py:48] [56200] global_step=56200, grad_norm=3.991499900817871, loss=1.1629681587219238 +I0831 13:41:33.429948 139921973044992 logging_writer.py:48] [56300] global_step=56300, grad_norm=4.351630687713623, loss=1.1437630653381348 +I0831 13:42:00.596826 139921981437696 logging_writer.py:48] [56400] global_step=56400, grad_norm=4.037383079528809, loss=1.1680290699005127 +I0831 13:42:27.946726 139921973044992 logging_writer.py:48] [56500] global_step=56500, grad_norm=4.216047763824463, loss=1.3193590641021729 +I0831 13:42:55.105044 139921981437696 logging_writer.py:48] [56600] global_step=56600, grad_norm=4.0099592208862305, loss=1.1605302095413208 +I0831 13:43:22.311755 139921973044992 logging_writer.py:48] [56700] global_step=56700, grad_norm=4.074021816253662, loss=1.114471435546875 +I0831 13:43:49.426563 139921981437696 logging_writer.py:48] [56800] global_step=56800, grad_norm=3.766449451446533, loss=1.0946992635726929 +I0831 13:44:16.536736 139921973044992 logging_writer.py:48] [56900] global_step=56900, grad_norm=3.957066774368286, loss=1.093938946723938 +I0831 13:44:43.729047 139921981437696 logging_writer.py:48] [57000] global_step=57000, grad_norm=4.007513046264648, loss=1.160604476928711 +I0831 13:45:10.837894 139921973044992 logging_writer.py:48] [57100] global_step=57100, grad_norm=3.911201000213623, loss=1.143389105796814 +I0831 13:45:37.971544 139921981437696 logging_writer.py:48] [57200] global_step=57200, grad_norm=3.995591402053833, loss=1.2456777095794678 +I0831 13:46:05.146398 139921973044992 logging_writer.py:48] [57300] global_step=57300, grad_norm=3.9539754390716553, loss=1.1241140365600586 +I0831 13:46:32.292541 139921981437696 logging_writer.py:48] [57400] global_step=57400, grad_norm=4.0560383796691895, loss=0.9987987279891968 +I0831 13:46:59.432651 139921973044992 logging_writer.py:48] [57500] global_step=57500, grad_norm=4.183343887329102, loss=1.16025710105896 +I0831 13:47:26.849992 139921981437696 logging_writer.py:48] [57600] global_step=57600, grad_norm=4.007669448852539, loss=1.183336615562439 +I0831 13:47:53.975753 139921973044992 logging_writer.py:48] [57700] global_step=57700, grad_norm=4.057479381561279, loss=1.108009696006775 +I0831 13:48:21.090221 139921981437696 logging_writer.py:48] [57800] global_step=57800, grad_norm=4.2681989669799805, loss=1.153936505317688 +I0831 13:48:48.284666 139921973044992 logging_writer.py:48] [57900] global_step=57900, grad_norm=4.091436862945557, loss=1.1722476482391357 +I0831 13:49:14.990505 140117123622080 spec.py:333] Evaluating on the training split. +I0831 13:49:24.660010 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 13:49:34.919996 140117123622080 spec.py:363] Evaluating on the test split. +I0831 13:49:35.805817 140117123622080 submission_runner.py:516] Time since start: 16297.15s, Step: 58000, {'train/accuracy': Array(0.89333546, dtype=float32), 'train/loss': Array(0.38592175, dtype=float32), 'validation/accuracy': Array(0.73793995, dtype=float32), 'validation/loss': Array(1.0922809, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.60840005, dtype=float32), 'test/loss': Array(1.827331, dtype=float32), 'test/num_examples': 10000, 'score': 16021.156431436539, 'total_duration': 16297.145352840424, 'accumulated_submission_time': 16021.156431436539, 'accumulated_eval_time': 274.97972774505615, 'accumulated_logging_time': 0.5500438213348389} +I0831 13:49:35.859057 139921981437696 logging_writer.py:48] [58000] accumulated_eval_time=274.98, accumulated_logging_time=0.550044, accumulated_submission_time=16021.2, global_step=58000, preemption_count=0, score=16021.2, test/accuracy=0.6084000468254089, test/loss=1.8273309469223022, test/num_examples=10000, total_duration=16297.1, train/accuracy=0.8933354616165161, train/loss=0.38592174649238586, validation/accuracy=0.7379399538040161, validation/loss=1.0922808647155762, validation/num_examples=50000 +I0831 13:49:36.219742 139921973044992 logging_writer.py:48] [58000] global_step=58000, grad_norm=4.266199111938477, loss=1.2664750814437866 +I0831 13:50:03.479547 139921981437696 logging_writer.py:48] [58100] global_step=58100, grad_norm=4.2715277671813965, loss=1.17368483543396 +I0831 13:50:30.667026 139921973044992 logging_writer.py:48] [58200] global_step=58200, grad_norm=3.909823417663574, loss=1.1469874382019043 +I0831 13:50:57.759113 139921981437696 logging_writer.py:48] [58300] global_step=58300, grad_norm=4.2364325523376465, loss=1.204797625541687 +I0831 13:51:24.885847 139921973044992 logging_writer.py:48] [58400] global_step=58400, grad_norm=3.8811357021331787, loss=1.1175302267074585 +I0831 13:51:52.089239 139921981437696 logging_writer.py:48] [58500] global_step=58500, grad_norm=4.3930253982543945, loss=1.2825934886932373 +I0831 13:52:19.239369 139921973044992 logging_writer.py:48] [58600] global_step=58600, grad_norm=4.021306037902832, loss=1.0993365049362183 +I0831 13:52:46.567324 139921981437696 logging_writer.py:48] [58700] global_step=58700, grad_norm=4.363784313201904, loss=1.1538273096084595 +I0831 13:53:13.743472 139921973044992 logging_writer.py:48] [58800] global_step=58800, grad_norm=3.8454103469848633, loss=1.0427528619766235 +I0831 13:53:40.855246 139921981437696 logging_writer.py:48] [58900] global_step=58900, grad_norm=3.9080584049224854, loss=1.1944273710250854 +I0831 13:54:07.970517 139921973044992 logging_writer.py:48] [59000] global_step=59000, grad_norm=4.290788650512695, loss=1.132047414779663 +I0831 13:54:35.158033 139921981437696 logging_writer.py:48] [59100] global_step=59100, grad_norm=3.9867589473724365, loss=1.1571305990219116 +I0831 13:55:02.303617 139921973044992 logging_writer.py:48] [59200] global_step=59200, grad_norm=4.425419807434082, loss=1.2238481044769287 +I0831 13:55:29.446578 139921981437696 logging_writer.py:48] [59300] global_step=59300, grad_norm=4.166377544403076, loss=1.1755133867263794 +I0831 13:55:56.623919 139921973044992 logging_writer.py:48] [59400] global_step=59400, grad_norm=4.1101884841918945, loss=1.1590778827667236 +I0831 13:56:23.740624 139921981437696 logging_writer.py:48] [59500] global_step=59500, grad_norm=4.17479133605957, loss=1.1312954425811768 +I0831 13:56:50.851554 139921973044992 logging_writer.py:48] [59600] global_step=59600, grad_norm=4.131350517272949, loss=1.1840250492095947 +I0831 13:57:18.233172 139921981437696 logging_writer.py:48] [59700] global_step=59700, grad_norm=4.198848724365234, loss=1.1663825511932373 +I0831 13:57:45.352008 139921973044992 logging_writer.py:48] [59800] global_step=59800, grad_norm=4.288026809692383, loss=1.1486767530441284 +I0831 13:58:12.450733 139921981437696 logging_writer.py:48] [59900] global_step=59900, grad_norm=4.131298542022705, loss=1.1336679458618164 +I0831 13:58:39.627044 139921973044992 logging_writer.py:48] [60000] global_step=60000, grad_norm=4.259805202484131, loss=1.1808662414550781 +I0831 13:59:06.734422 139921981437696 logging_writer.py:48] [60100] global_step=60100, grad_norm=4.359585285186768, loss=1.1758352518081665 +I0831 13:59:33.886660 139921973044992 logging_writer.py:48] [60200] global_step=60200, grad_norm=4.0811357498168945, loss=1.0540581941604614 +I0831 14:00:01.089081 139921981437696 logging_writer.py:48] [60300] global_step=60300, grad_norm=4.340694427490234, loss=1.1885871887207031 +I0831 14:00:28.189343 139921973044992 logging_writer.py:48] [60400] global_step=60400, grad_norm=4.150360584259033, loss=1.1544097661972046 +I0831 14:00:55.326776 139921981437696 logging_writer.py:48] [60500] global_step=60500, grad_norm=4.077099323272705, loss=1.0865249633789062 +I0831 14:01:22.524581 139921973044992 logging_writer.py:48] [60600] global_step=60600, grad_norm=4.1834635734558105, loss=1.1093826293945312 +I0831 14:01:49.649631 139921981437696 logging_writer.py:48] [60700] global_step=60700, grad_norm=4.333500862121582, loss=1.2967233657836914 +I0831 14:02:16.976417 139921973044992 logging_writer.py:48] [60800] global_step=60800, grad_norm=4.377627372741699, loss=1.1511332988739014 +I0831 14:02:44.163438 139921981437696 logging_writer.py:48] [60900] global_step=60900, grad_norm=4.620772361755371, loss=1.2068122625350952 +I0831 14:03:11.276202 139921973044992 logging_writer.py:48] [61000] global_step=61000, grad_norm=4.27354621887207, loss=1.0745434761047363 +I0831 14:03:38.403067 139921981437696 logging_writer.py:48] [61100] global_step=61100, grad_norm=4.186177730560303, loss=1.1857340335845947 +I0831 14:04:05.593792 139921973044992 logging_writer.py:48] [61200] global_step=61200, grad_norm=4.260356426239014, loss=1.129649043083191 +I0831 14:04:32.723909 139921981437696 logging_writer.py:48] [61300] global_step=61300, grad_norm=4.143117427825928, loss=1.1200348138809204 +I0831 14:04:59.859618 139921973044992 logging_writer.py:48] [61400] global_step=61400, grad_norm=4.191991806030273, loss=1.1138744354248047 +I0831 14:05:27.054837 139921981437696 logging_writer.py:48] [61500] global_step=61500, grad_norm=4.13618278503418, loss=1.2025253772735596 +I0831 14:05:54.163546 139921973044992 logging_writer.py:48] [61600] global_step=61600, grad_norm=4.447849273681641, loss=1.2000651359558105 +I0831 14:06:21.295289 139921981437696 logging_writer.py:48] [61700] global_step=61700, grad_norm=4.166731834411621, loss=1.0973763465881348 +I0831 14:06:48.498056 139921973044992 logging_writer.py:48] [61800] global_step=61800, grad_norm=3.85280179977417, loss=1.0914725065231323 +I0831 14:07:15.842453 139921981437696 logging_writer.py:48] [61900] global_step=61900, grad_norm=4.2000226974487305, loss=1.1129132509231567 +I0831 14:07:42.953731 139921973044992 logging_writer.py:48] [62000] global_step=62000, grad_norm=3.9603779315948486, loss=1.140796184539795 +I0831 14:08:10.129324 139921981437696 logging_writer.py:48] [62100] global_step=62100, grad_norm=4.152520179748535, loss=1.1188960075378418 +I0831 14:08:37.257449 139921973044992 logging_writer.py:48] [62200] global_step=62200, grad_norm=4.323079586029053, loss=1.0994720458984375 +I0831 14:09:04.380950 139921981437696 logging_writer.py:48] [62300] global_step=62300, grad_norm=4.439029216766357, loss=1.1521847248077393 +I0831 14:09:31.564558 139921973044992 logging_writer.py:48] [62400] global_step=62400, grad_norm=4.405654430389404, loss=1.1520469188690186 +I0831 14:09:58.676613 139921981437696 logging_writer.py:48] [62500] global_step=62500, grad_norm=4.458680629730225, loss=1.0874254703521729 +I0831 14:10:25.818032 139921973044992 logging_writer.py:48] [62600] global_step=62600, grad_norm=4.125410556793213, loss=1.0835649967193604 +I0831 14:10:53.001259 139921981437696 logging_writer.py:48] [62700] global_step=62700, grad_norm=4.187085151672363, loss=1.0691039562225342 +I0831 14:11:20.138988 139921973044992 logging_writer.py:48] [62800] global_step=62800, grad_norm=4.339331150054932, loss=1.1084336042404175 +I0831 14:11:47.466447 139921981437696 logging_writer.py:48] [62900] global_step=62900, grad_norm=4.255492687225342, loss=1.088625431060791 +I0831 14:12:14.641345 139921973044992 logging_writer.py:48] [63000] global_step=63000, grad_norm=4.6627397537231445, loss=1.2226793766021729 +I0831 14:12:41.792482 139921981437696 logging_writer.py:48] [63100] global_step=63100, grad_norm=4.649829864501953, loss=1.2214763164520264 +I0831 14:13:08.929298 139921973044992 logging_writer.py:48] [63200] global_step=63200, grad_norm=4.228373050689697, loss=1.148026704788208 +I0831 14:13:36.112559 139921981437696 logging_writer.py:48] [63300] global_step=63300, grad_norm=4.240910053253174, loss=1.116142988204956 +I0831 14:14:03.215472 139921973044992 logging_writer.py:48] [63400] global_step=63400, grad_norm=4.505836009979248, loss=1.2003531455993652 +I0831 14:14:30.348933 139921981437696 logging_writer.py:48] [63500] global_step=63500, grad_norm=4.397470951080322, loss=1.0809308290481567 +I0831 14:14:57.531858 139921973044992 logging_writer.py:48] [63600] global_step=63600, grad_norm=4.388909339904785, loss=1.0458390712738037 +I0831 14:15:24.707447 139921981437696 logging_writer.py:48] [63700] global_step=63700, grad_norm=4.265614032745361, loss=1.1787378787994385 +I0831 14:15:51.806914 139921973044992 logging_writer.py:48] [63800] global_step=63800, grad_norm=4.398782730102539, loss=1.0236880779266357 +I0831 14:16:19.011554 139921981437696 logging_writer.py:48] [63900] global_step=63900, grad_norm=4.145705699920654, loss=1.053657054901123 +I0831 14:16:46.375256 139921973044992 logging_writer.py:48] [64000] global_step=64000, grad_norm=4.44474983215332, loss=1.1091941595077515 +I0831 14:17:13.477711 139921981437696 logging_writer.py:48] [64100] global_step=64100, grad_norm=4.503633975982666, loss=1.125067114830017 +I0831 14:17:40.637193 139921973044992 logging_writer.py:48] [64200] global_step=64200, grad_norm=4.771666049957275, loss=1.1823153495788574 +I0831 14:18:07.743129 139921981437696 logging_writer.py:48] [64300] global_step=64300, grad_norm=4.398667812347412, loss=1.171494960784912 +I0831 14:18:34.906331 139921973044992 logging_writer.py:48] [64400] global_step=64400, grad_norm=4.5516862869262695, loss=1.1718559265136719 +I0831 14:19:02.090162 139921981437696 logging_writer.py:48] [64500] global_step=64500, grad_norm=4.655465602874756, loss=1.137829303741455 +I0831 14:19:29.214288 139921973044992 logging_writer.py:48] [64600] global_step=64600, grad_norm=4.5468926429748535, loss=1.139796257019043 +I0831 14:19:56.321996 139921981437696 logging_writer.py:48] [64700] global_step=64700, grad_norm=4.432785511016846, loss=1.09183931350708 +I0831 14:20:23.540030 139921973044992 logging_writer.py:48] [64800] global_step=64800, grad_norm=5.000415325164795, loss=1.2811094522476196 +I0831 14:20:50.662891 139921981437696 logging_writer.py:48] [64900] global_step=64900, grad_norm=4.546041011810303, loss=1.1815170049667358 +I0831 14:21:18.029473 139921973044992 logging_writer.py:48] [65000] global_step=65000, grad_norm=4.373263359069824, loss=1.0954612493515015 +I0831 14:21:45.187669 139921981437696 logging_writer.py:48] [65100] global_step=65100, grad_norm=4.645698070526123, loss=1.074838399887085 +I0831 14:22:12.318854 139921973044992 logging_writer.py:48] [65200] global_step=65200, grad_norm=4.889631748199463, loss=1.2568674087524414 +I0831 14:22:39.441518 139921981437696 logging_writer.py:48] [65300] global_step=65300, grad_norm=4.395275115966797, loss=1.1298604011535645 +I0831 14:22:51.855942 140117123622080 spec.py:333] Evaluating on the training split. +I0831 14:23:00.602661 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 14:23:10.278431 140117123622080 spec.py:363] Evaluating on the test split. +I0831 14:23:11.169956 140117123622080 submission_runner.py:516] Time since start: 18312.51s, Step: 65347, {'train/accuracy': Array(0.902264, dtype=float32), 'train/loss': Array(0.35404882, dtype=float32), 'validation/accuracy': Array(0.74158, dtype=float32), 'validation/loss': Array(1.082871, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.60940003, dtype=float32), 'test/loss': Array(1.8199519, dtype=float32), 'test/num_examples': 10000, 'score': 18017.071206331253, 'total_duration': 18312.510173797607, 'accumulated_submission_time': 18017.071206331253, 'accumulated_eval_time': 294.29212617874146, 'accumulated_logging_time': 0.6333343982696533} +I0831 14:23:11.238999 139921973044992 logging_writer.py:48] [65347] accumulated_eval_time=294.292, accumulated_logging_time=0.633334, accumulated_submission_time=18017.1, global_step=65347, preemption_count=0, score=18017.1, test/accuracy=0.6094000339508057, test/loss=1.8199518918991089, test/num_examples=10000, total_duration=18312.5, train/accuracy=0.9022639989852905, train/loss=0.35404881834983826, validation/accuracy=0.7415800094604492, validation/loss=1.0828709602355957, validation/num_examples=50000 +I0831 14:23:25.990882 139921981437696 logging_writer.py:48] [65400] global_step=65400, grad_norm=4.101398468017578, loss=1.0438628196716309 +I0831 14:23:53.090944 139921973044992 logging_writer.py:48] [65500] global_step=65500, grad_norm=4.57509183883667, loss=1.142104148864746 +I0831 14:24:20.209227 139921981437696 logging_writer.py:48] [65600] global_step=65600, grad_norm=4.848811626434326, loss=1.190643072128296 +I0831 14:24:47.396575 139921973044992 logging_writer.py:48] [65700] global_step=65700, grad_norm=4.29219388961792, loss=1.0560153722763062 +I0831 14:25:14.493501 139921981437696 logging_writer.py:48] [65800] global_step=65800, grad_norm=4.673631191253662, loss=1.1863290071487427 +I0831 14:25:41.623129 139921973044992 logging_writer.py:48] [65900] global_step=65900, grad_norm=4.711935043334961, loss=1.1794850826263428 +I0831 14:26:09.038257 139921981437696 logging_writer.py:48] [66000] global_step=66000, grad_norm=4.8600568771362305, loss=1.171064853668213 +I0831 14:26:36.194282 139921973044992 logging_writer.py:48] [66100] global_step=66100, grad_norm=4.595717430114746, loss=1.094131588935852 +I0831 14:27:03.332415 139921981437696 logging_writer.py:48] [66200] global_step=66200, grad_norm=4.770264625549316, loss=1.1939868927001953 +I0831 14:27:30.504551 139921973044992 logging_writer.py:48] [66300] global_step=66300, grad_norm=4.3950347900390625, loss=1.1467045545578003 +I0831 14:27:57.620370 139921981437696 logging_writer.py:48] [66400] global_step=66400, grad_norm=4.767884254455566, loss=1.1233928203582764 +I0831 14:28:24.766816 139921973044992 logging_writer.py:48] [66500] global_step=66500, grad_norm=4.739262580871582, loss=1.1042908430099487 +I0831 14:28:51.956856 139921981437696 logging_writer.py:48] [66600] global_step=66600, grad_norm=4.4891204833984375, loss=1.0323498249053955 +I0831 14:29:19.119726 139921973044992 logging_writer.py:48] [66700] global_step=66700, grad_norm=4.667444705963135, loss=1.1761155128479004 +I0831 14:29:46.262524 139921981437696 logging_writer.py:48] [66800] global_step=66800, grad_norm=4.85011100769043, loss=1.2334927320480347 +I0831 14:30:13.439866 139921973044992 logging_writer.py:48] [66900] global_step=66900, grad_norm=4.720789432525635, loss=1.1222870349884033 +I0831 14:30:40.567733 139921981437696 logging_writer.py:48] [67000] global_step=67000, grad_norm=4.72529935836792, loss=1.1959786415100098 +I0831 14:31:07.900299 139921973044992 logging_writer.py:48] [67100] global_step=67100, grad_norm=4.805488109588623, loss=1.1020101308822632 +I0831 14:31:35.084568 139921981437696 logging_writer.py:48] [67200] global_step=67200, grad_norm=4.814241886138916, loss=1.1361172199249268 +I0831 14:32:02.203559 139921973044992 logging_writer.py:48] [67300] global_step=67300, grad_norm=4.710190773010254, loss=1.081744909286499 +I0831 14:32:29.353504 139921981437696 logging_writer.py:48] [67400] global_step=67400, grad_norm=4.733648777008057, loss=1.1257764101028442 +I0831 14:32:56.536435 139921973044992 logging_writer.py:48] [67500] global_step=67500, grad_norm=4.735122203826904, loss=1.0482170581817627 +I0831 14:33:23.665860 139921981437696 logging_writer.py:48] [67600] global_step=67600, grad_norm=4.858153820037842, loss=1.1353048086166382 +I0831 14:33:50.772533 139921973044992 logging_writer.py:48] [67700] global_step=67700, grad_norm=4.845344543457031, loss=1.1502013206481934 +I0831 14:34:17.949711 139921981437696 logging_writer.py:48] [67800] global_step=67800, grad_norm=5.009126663208008, loss=1.1272145509719849 +I0831 14:34:45.096554 139921973044992 logging_writer.py:48] [67900] global_step=67900, grad_norm=5.078135013580322, loss=1.1565523147583008 +I0831 14:35:12.228301 139921981437696 logging_writer.py:48] [68000] global_step=68000, grad_norm=4.995347499847412, loss=1.1867330074310303 +I0831 14:35:39.404075 139921973044992 logging_writer.py:48] [68100] global_step=68100, grad_norm=4.884612083435059, loss=1.147743821144104 +I0831 14:36:06.735168 139921981437696 logging_writer.py:48] [68200] global_step=68200, grad_norm=5.104712963104248, loss=1.0627394914627075 +I0831 14:36:33.862208 139921973044992 logging_writer.py:48] [68300] global_step=68300, grad_norm=4.765049457550049, loss=1.1504946947097778 +I0831 14:37:01.061503 139921981437696 logging_writer.py:48] [68400] global_step=68400, grad_norm=4.813565731048584, loss=1.0259149074554443 +I0831 14:37:28.186849 139921973044992 logging_writer.py:48] [68500] global_step=68500, grad_norm=4.897980690002441, loss=1.1635754108428955 +I0831 14:37:55.302722 139921981437696 logging_writer.py:48] [68600] global_step=68600, grad_norm=5.1913557052612305, loss=1.152198076248169 +I0831 14:38:22.486143 139921973044992 logging_writer.py:48] [68700] global_step=68700, grad_norm=4.7669172286987305, loss=1.1064397096633911 +I0831 14:38:49.659597 139921981437696 logging_writer.py:48] [68800] global_step=68800, grad_norm=4.742974758148193, loss=1.0124551057815552 +I0831 14:39:16.778997 139921973044992 logging_writer.py:48] [68900] global_step=68900, grad_norm=5.109218597412109, loss=1.1678385734558105 +I0831 14:39:43.976955 139921981437696 logging_writer.py:48] [69000] global_step=69000, grad_norm=4.915887832641602, loss=1.097650170326233 +I0831 14:40:11.134737 139921973044992 logging_writer.py:48] [69100] global_step=69100, grad_norm=4.993927478790283, loss=1.1646918058395386 +I0831 14:40:38.250189 139921981437696 logging_writer.py:48] [69200] global_step=69200, grad_norm=4.781600475311279, loss=1.1242425441741943 +I0831 14:41:05.662429 139921973044992 logging_writer.py:48] [69300] global_step=69300, grad_norm=4.62204122543335, loss=1.0229254961013794 +I0831 14:41:32.767539 139921981437696 logging_writer.py:48] [69400] global_step=69400, grad_norm=4.9112043380737305, loss=1.1304280757904053 +I0831 14:41:59.912784 139921973044992 logging_writer.py:48] [69500] global_step=69500, grad_norm=4.816128253936768, loss=1.0334134101867676 +I0831 14:42:27.114762 139921981437696 logging_writer.py:48] [69600] global_step=69600, grad_norm=4.833178997039795, loss=1.157449722290039 +I0831 14:42:54.227430 139921973044992 logging_writer.py:48] [69700] global_step=69700, grad_norm=4.846080780029297, loss=1.1215360164642334 +I0831 14:43:21.361923 139921981437696 logging_writer.py:48] [69800] global_step=69800, grad_norm=4.9487481117248535, loss=1.1895010471343994 +I0831 14:43:48.535931 139921973044992 logging_writer.py:48] [69900] global_step=69900, grad_norm=4.9894118309021, loss=1.1351152658462524 +I0831 14:44:15.652756 139921981437696 logging_writer.py:48] [70000] global_step=70000, grad_norm=5.018789291381836, loss=1.1443202495574951 +I0831 14:44:42.787801 139921973044992 logging_writer.py:48] [70100] global_step=70100, grad_norm=5.0282883644104, loss=1.0329018831253052 +I0831 14:45:09.998080 139921981437696 logging_writer.py:48] [70200] global_step=70200, grad_norm=4.9962077140808105, loss=1.138913869857788 +I0831 14:45:37.333864 139921973044992 logging_writer.py:48] [70300] global_step=70300, grad_norm=4.851072311401367, loss=1.1627275943756104 +I0831 14:46:04.452016 139921981437696 logging_writer.py:48] [70400] global_step=70400, grad_norm=4.9225382804870605, loss=1.1140365600585938 +I0831 14:46:31.651588 139921973044992 logging_writer.py:48] [70500] global_step=70500, grad_norm=5.152256488800049, loss=1.1049968004226685 +I0831 14:46:58.775603 139921981437696 logging_writer.py:48] [70600] global_step=70600, grad_norm=5.0186238288879395, loss=1.0909229516983032 +I0831 14:47:25.919833 139921973044992 logging_writer.py:48] [70700] global_step=70700, grad_norm=5.124016761779785, loss=1.1382887363433838 +I0831 14:47:53.136082 139921981437696 logging_writer.py:48] [70800] global_step=70800, grad_norm=4.688169479370117, loss=1.0554054975509644 +I0831 14:48:20.281051 139921973044992 logging_writer.py:48] [70900] global_step=70900, grad_norm=5.053037643432617, loss=1.0685118436813354 +I0831 14:48:47.380601 139921981437696 logging_writer.py:48] [71000] global_step=71000, grad_norm=5.073338031768799, loss=1.1066813468933105 +I0831 14:49:14.570951 139921973044992 logging_writer.py:48] [71100] global_step=71100, grad_norm=4.900101184844971, loss=1.0835747718811035 +I0831 14:49:41.678243 139921981437696 logging_writer.py:48] [71200] global_step=71200, grad_norm=4.88961124420166, loss=1.1217281818389893 +I0831 14:50:08.829879 139921973044992 logging_writer.py:48] [71300] global_step=71300, grad_norm=4.82552433013916, loss=1.065260410308838 +I0831 14:50:36.203010 139921981437696 logging_writer.py:48] [71400] global_step=71400, grad_norm=5.046196937561035, loss=1.128748893737793 +I0831 14:51:03.339010 139921973044992 logging_writer.py:48] [71500] global_step=71500, grad_norm=5.131617546081543, loss=1.1914293766021729 +I0831 14:51:30.471650 139921981437696 logging_writer.py:48] [71600] global_step=71600, grad_norm=5.031429290771484, loss=1.1474926471710205 +I0831 14:51:57.668415 139921973044992 logging_writer.py:48] [71700] global_step=71700, grad_norm=5.11308479309082, loss=1.1210181713104248 +I0831 14:52:24.794324 139921981437696 logging_writer.py:48] [71800] global_step=71800, grad_norm=5.01565408706665, loss=1.0791280269622803 +I0831 14:52:51.925388 139921973044992 logging_writer.py:48] [71900] global_step=71900, grad_norm=5.334219455718994, loss=1.1214804649353027 +I0831 14:53:19.090880 139921981437696 logging_writer.py:48] [72000] global_step=72000, grad_norm=4.761529445648193, loss=1.051267385482788 +I0831 14:53:46.211894 139921973044992 logging_writer.py:48] [72100] global_step=72100, grad_norm=4.807604789733887, loss=1.0979233980178833 +I0831 14:54:13.343312 139921981437696 logging_writer.py:48] [72200] global_step=72200, grad_norm=5.260127067565918, loss=1.2045758962631226 +I0831 14:54:40.533004 139921973044992 logging_writer.py:48] [72300] global_step=72300, grad_norm=5.1565399169921875, loss=1.1268677711486816 +I0831 14:55:07.860650 139921981437696 logging_writer.py:48] [72400] global_step=72400, grad_norm=5.111470699310303, loss=1.1325664520263672 +I0831 14:55:34.982087 139921973044992 logging_writer.py:48] [72500] global_step=72500, grad_norm=5.174179553985596, loss=1.0892267227172852 +I0831 14:56:02.165680 139921981437696 logging_writer.py:48] [72600] global_step=72600, grad_norm=4.966782569885254, loss=1.1475439071655273 +I0831 14:56:27.239144 140117123622080 spec.py:333] Evaluating on the training split. +I0831 14:56:36.141767 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 14:56:45.669651 140117123622080 spec.py:363] Evaluating on the test split. +I0831 14:56:46.562065 140117123622080 submission_runner.py:516] Time since start: 20327.90s, Step: 72694, {'train/accuracy': Array(0.90927935, dtype=float32), 'train/loss': Array(0.32634285, dtype=float32), 'validation/accuracy': Array(0.74307996, dtype=float32), 'validation/loss': Array(1.0753398, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6147, dtype=float32), 'test/loss': Array(1.8146323, dtype=float32), 'test/num_examples': 10000, 'score': 20012.96212363243, 'total_duration': 20327.902194976807, 'accumulated_submission_time': 20012.96212363243, 'accumulated_eval_time': 313.6133460998535, 'accumulated_logging_time': 0.7590162754058838} +I0831 14:56:46.615752 139921973044992 logging_writer.py:48] [72694] accumulated_eval_time=313.613, accumulated_logging_time=0.759016, accumulated_submission_time=20013, global_step=72694, preemption_count=0, score=20013, test/accuracy=0.6147000193595886, test/loss=1.8146322965621948, test/num_examples=10000, total_duration=20327.9, train/accuracy=0.9092793464660645, train/loss=0.3263428509235382, validation/accuracy=0.7430799603462219, validation/loss=1.0753397941589355, validation/num_examples=50000 +I0831 14:56:48.630256 139921981437696 logging_writer.py:48] [72700] global_step=72700, grad_norm=4.689984321594238, loss=0.9761513471603394 +I0831 14:57:15.845953 139921973044992 logging_writer.py:48] [72800] global_step=72800, grad_norm=4.943140506744385, loss=1.0665078163146973 +I0831 14:57:43.027375 139921981437696 logging_writer.py:48] [72900] global_step=72900, grad_norm=5.51206111907959, loss=1.1972532272338867 +I0831 14:58:10.121062 139921973044992 logging_writer.py:48] [73000] global_step=73000, grad_norm=5.036368370056152, loss=1.07245934009552 +I0831 14:58:37.225706 139921981437696 logging_writer.py:48] [73100] global_step=73100, grad_norm=5.24921178817749, loss=1.1665465831756592 +I0831 14:59:04.433664 139921973044992 logging_writer.py:48] [73200] global_step=73200, grad_norm=5.332740783691406, loss=1.201326608657837 +I0831 14:59:31.532684 139921981437696 logging_writer.py:48] [73300] global_step=73300, grad_norm=5.324157238006592, loss=1.0642510652542114 +I0831 14:59:58.660310 139921973044992 logging_writer.py:48] [73400] global_step=73400, grad_norm=4.993917465209961, loss=1.1561779975891113 +I0831 15:00:26.073395 139921981437696 logging_writer.py:48] [73500] global_step=73500, grad_norm=5.163723945617676, loss=1.092266321182251 +I0831 15:00:53.209861 139921973044992 logging_writer.py:48] [73600] global_step=73600, grad_norm=4.907471656799316, loss=1.049884557723999 +I0831 15:01:20.331002 139921981437696 logging_writer.py:48] [73700] global_step=73700, grad_norm=4.820695400238037, loss=1.0931429862976074 +I0831 15:01:47.516947 139921973044992 logging_writer.py:48] [73800] global_step=73800, grad_norm=5.0171613693237305, loss=1.0809097290039062 +I0831 15:02:14.629712 139921981437696 logging_writer.py:48] [73900] global_step=73900, grad_norm=5.090182781219482, loss=1.110278606414795 +I0831 15:02:41.740900 139921973044992 logging_writer.py:48] [74000] global_step=74000, grad_norm=5.084195613861084, loss=1.1111266613006592 +I0831 15:03:08.926740 139921981437696 logging_writer.py:48] [74100] global_step=74100, grad_norm=5.148543357849121, loss=1.0917757749557495 +I0831 15:03:36.047661 139921973044992 logging_writer.py:48] [74200] global_step=74200, grad_norm=5.215298652648926, loss=1.1212313175201416 +I0831 15:04:03.173035 139921981437696 logging_writer.py:48] [74300] global_step=74300, grad_norm=4.790436267852783, loss=1.0373541116714478 +I0831 15:04:30.334450 139921973044992 logging_writer.py:48] [74400] global_step=74400, grad_norm=5.087887287139893, loss=1.052003026008606 +I0831 15:04:57.661135 139921981437696 logging_writer.py:48] [74500] global_step=74500, grad_norm=4.849293231964111, loss=1.0941190719604492 +I0831 15:05:24.786757 139921973044992 logging_writer.py:48] [74600] global_step=74600, grad_norm=5.276228904724121, loss=1.1550941467285156 +I0831 15:05:51.984738 139921981437696 logging_writer.py:48] [74700] global_step=74700, grad_norm=4.901560306549072, loss=1.084816336631775 +I0831 15:06:19.097811 139921973044992 logging_writer.py:48] [74800] global_step=74800, grad_norm=5.171930313110352, loss=1.1358134746551514 +I0831 15:06:46.203836 139921981437696 logging_writer.py:48] [74900] global_step=74900, grad_norm=4.995802402496338, loss=1.103102684020996 +I0831 15:07:13.378346 139921973044992 logging_writer.py:48] [75000] global_step=75000, grad_norm=5.466159343719482, loss=1.107773780822754 +I0831 15:07:40.521841 139921981437696 logging_writer.py:48] [75100] global_step=75100, grad_norm=5.061507701873779, loss=1.1366379261016846 +I0831 15:08:07.639659 139921973044992 logging_writer.py:48] [75200] global_step=75200, grad_norm=5.077211856842041, loss=1.091511607170105 +I0831 15:08:34.812476 139921981437696 logging_writer.py:48] [75300] global_step=75300, grad_norm=4.999211311340332, loss=0.9825881719589233 +I0831 15:09:01.940867 139921973044992 logging_writer.py:48] [75400] global_step=75400, grad_norm=5.1975932121276855, loss=1.1111698150634766 +I0831 15:09:29.065706 139921981437696 logging_writer.py:48] [75500] global_step=75500, grad_norm=5.331487655639648, loss=1.1875977516174316 +I0831 15:09:56.419100 139921973044992 logging_writer.py:48] [75600] global_step=75600, grad_norm=5.060285568237305, loss=1.1027425527572632 +I0831 15:10:23.549052 139921981437696 logging_writer.py:48] [75700] global_step=75700, grad_norm=5.08696985244751, loss=1.1923480033874512 +I0831 15:10:50.658640 139921973044992 logging_writer.py:48] [75800] global_step=75800, grad_norm=5.441874980926514, loss=1.2674405574798584 +I0831 15:11:17.821956 139921981437696 logging_writer.py:48] [75900] global_step=75900, grad_norm=4.949965000152588, loss=1.0476014614105225 +I0831 15:11:44.948333 139921973044992 logging_writer.py:48] [76000] global_step=76000, grad_norm=5.0859856605529785, loss=1.10015869140625 +I0831 15:12:12.093542 139921981437696 logging_writer.py:48] [76100] global_step=76100, grad_norm=5.04551362991333, loss=1.108871579170227 +I0831 15:12:39.270822 139921973044992 logging_writer.py:48] [76200] global_step=76200, grad_norm=5.341018199920654, loss=1.1231398582458496 +I0831 15:13:06.420224 139921981437696 logging_writer.py:48] [76300] global_step=76300, grad_norm=5.370436668395996, loss=1.0661956071853638 +I0831 15:13:33.532053 139921973044992 logging_writer.py:48] [76400] global_step=76400, grad_norm=4.97642707824707, loss=1.083654761314392 +I0831 15:14:00.694876 139921981437696 logging_writer.py:48] [76500] global_step=76500, grad_norm=5.402655601501465, loss=1.1506675481796265 +I0831 15:14:27.803941 139921973044992 logging_writer.py:48] [76600] global_step=76600, grad_norm=4.965524673461914, loss=1.0757575035095215 +I0831 15:14:55.160644 139921981437696 logging_writer.py:48] [76700] global_step=76700, grad_norm=5.203562259674072, loss=1.1152127981185913 +I0831 15:15:22.346066 139921973044992 logging_writer.py:48] [76800] global_step=76800, grad_norm=5.209282875061035, loss=1.1114661693572998 +I0831 15:15:49.472681 139921981437696 logging_writer.py:48] [76900] global_step=76900, grad_norm=5.270101070404053, loss=1.1034777164459229 +I0831 15:16:16.634410 139921973044992 logging_writer.py:48] [77000] global_step=77000, grad_norm=4.830829620361328, loss=1.0300520658493042 +I0831 15:16:43.806899 139921981437696 logging_writer.py:48] [77100] global_step=77100, grad_norm=5.38095235824585, loss=1.1462504863739014 +I0831 15:17:10.940693 139921973044992 logging_writer.py:48] [77200] global_step=77200, grad_norm=5.091732978820801, loss=1.1135560274124146 +I0831 15:17:38.063681 139921981437696 logging_writer.py:48] [77300] global_step=77300, grad_norm=5.389059543609619, loss=1.1136119365692139 +I0831 15:18:05.287899 139921973044992 logging_writer.py:48] [77400] global_step=77400, grad_norm=5.305062294006348, loss=1.1488096714019775 +I0831 15:18:32.423988 139921981437696 logging_writer.py:48] [77500] global_step=77500, grad_norm=5.026465892791748, loss=1.112950325012207 +I0831 15:18:59.524315 139921973044992 logging_writer.py:48] [77600] global_step=77600, grad_norm=4.986379146575928, loss=1.0278469324111938 +I0831 15:19:26.684331 139921981437696 logging_writer.py:48] [77700] global_step=77700, grad_norm=5.270123481750488, loss=1.0422487258911133 +I0831 15:19:54.022273 139921973044992 logging_writer.py:48] [77800] global_step=77800, grad_norm=5.093537330627441, loss=1.1071710586547852 +I0831 15:20:21.130464 139921981437696 logging_writer.py:48] [77900] global_step=77900, grad_norm=5.197536945343018, loss=1.0779982805252075 +I0831 15:20:48.297307 139921973044992 logging_writer.py:48] [78000] global_step=78000, grad_norm=4.99685525894165, loss=1.0701377391815186 +I0831 15:21:15.413417 139921981437696 logging_writer.py:48] [78100] global_step=78100, grad_norm=5.239736080169678, loss=1.0844916105270386 +I0831 15:21:42.533067 139921973044992 logging_writer.py:48] [78200] global_step=78200, grad_norm=4.996396541595459, loss=1.1201118230819702 +I0831 15:22:09.729408 139921981437696 logging_writer.py:48] [78300] global_step=78300, grad_norm=5.027043342590332, loss=1.080621361732483 +I0831 15:22:36.834934 139921973044992 logging_writer.py:48] [78400] global_step=78400, grad_norm=5.139797687530518, loss=1.1045668125152588 +I0831 15:23:03.975131 139921981437696 logging_writer.py:48] [78500] global_step=78500, grad_norm=5.142545700073242, loss=1.0673242807388306 +I0831 15:23:31.177442 139921973044992 logging_writer.py:48] [78600] global_step=78600, grad_norm=5.027669906616211, loss=1.0249481201171875 +I0831 15:23:58.327088 139921981437696 logging_writer.py:48] [78700] global_step=78700, grad_norm=4.728236198425293, loss=0.9893838167190552 +I0831 15:24:25.612567 139921973044992 logging_writer.py:48] [78800] global_step=78800, grad_norm=5.289634704589844, loss=1.0834747552871704 +I0831 15:24:52.777457 139921981437696 logging_writer.py:48] [78900] global_step=78900, grad_norm=5.052718162536621, loss=1.0009924173355103 +I0831 15:25:19.916835 139921973044992 logging_writer.py:48] [79000] global_step=79000, grad_norm=4.820370197296143, loss=1.0780344009399414 +I0831 15:25:47.022524 139921981437696 logging_writer.py:48] [79100] global_step=79100, grad_norm=4.770482540130615, loss=1.0067875385284424 +I0831 15:26:14.177916 139921973044992 logging_writer.py:48] [79200] global_step=79200, grad_norm=4.845874786376953, loss=1.0587310791015625 +I0831 15:26:41.306327 139921981437696 logging_writer.py:48] [79300] global_step=79300, grad_norm=4.985036373138428, loss=1.080036997795105 +I0831 15:27:08.442029 139921973044992 logging_writer.py:48] [79400] global_step=79400, grad_norm=5.6578145027160645, loss=1.2805588245391846 +I0831 15:27:35.620720 139921981437696 logging_writer.py:48] [79500] global_step=79500, grad_norm=5.174469947814941, loss=1.1091934442520142 +I0831 15:28:02.723290 139921973044992 logging_writer.py:48] [79600] global_step=79600, grad_norm=4.919693470001221, loss=1.0635347366333008 +I0831 15:28:29.836439 139921981437696 logging_writer.py:48] [79700] global_step=79700, grad_norm=5.0095953941345215, loss=1.1064584255218506 +I0831 15:28:57.037093 139921973044992 logging_writer.py:48] [79800] global_step=79800, grad_norm=5.111720085144043, loss=1.1639938354492188 +I0831 15:29:24.374439 139921981437696 logging_writer.py:48] [79900] global_step=79900, grad_norm=4.948285102844238, loss=1.0515899658203125 +I0831 15:29:51.509358 139921973044992 logging_writer.py:48] [80000] global_step=80000, grad_norm=5.069771766662598, loss=1.0621607303619385 +I0831 15:30:02.774015 140117123622080 spec.py:333] Evaluating on the training split. +I0831 15:30:10.658130 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 15:30:20.904147 140117123622080 spec.py:363] Evaluating on the test split. +I0831 15:30:21.794430 140117123622080 submission_runner.py:516] Time since start: 22343.13s, Step: 80043, {'train/accuracy': Array(0.91222894, dtype=float32), 'train/loss': Array(0.31514737, dtype=float32), 'validation/accuracy': Array(0.74571997, dtype=float32), 'validation/loss': Array(1.0688671, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6169, dtype=float32), 'test/loss': Array(1.8074372, dtype=float32), 'test/num_examples': 10000, 'score': 22009.03094482422, 'total_duration': 22343.13378739357, 'accumulated_submission_time': 22009.03094482422, 'accumulated_eval_time': 332.63128685951233, 'accumulated_logging_time': 0.8456320762634277} +I0831 15:30:21.838752 139921981437696 logging_writer.py:48] [80043] accumulated_eval_time=332.631, accumulated_logging_time=0.845632, accumulated_submission_time=22009, global_step=80043, preemption_count=0, score=22009, test/accuracy=0.6169000267982483, test/loss=1.8074371814727783, test/num_examples=10000, total_duration=22343.1, train/accuracy=0.9122289419174194, train/loss=0.31514737010002136, validation/accuracy=0.7457199692726135, validation/loss=1.0688670873641968, validation/num_examples=50000 +I0831 15:30:37.685667 139921973044992 logging_writer.py:48] [80100] global_step=80100, grad_norm=5.113198757171631, loss=1.169847011566162 +I0831 15:31:04.781076 139921981437696 logging_writer.py:48] [80200] global_step=80200, grad_norm=4.946497917175293, loss=1.0154576301574707 +I0831 15:31:31.871093 139921973044992 logging_writer.py:48] [80300] global_step=80300, grad_norm=5.287879467010498, loss=1.1337790489196777 +I0831 15:31:59.044991 139921981437696 logging_writer.py:48] [80400] global_step=80400, grad_norm=5.18288516998291, loss=1.0644694566726685 +I0831 15:32:26.142107 139921973044992 logging_writer.py:48] [80500] global_step=80500, grad_norm=4.887748718261719, loss=1.0537731647491455 +I0831 15:32:53.277523 139921981437696 logging_writer.py:48] [80600] global_step=80600, grad_norm=5.026390552520752, loss=1.0620787143707275 +I0831 15:33:20.473621 139921973044992 logging_writer.py:48] [80700] global_step=80700, grad_norm=4.882890701293945, loss=1.0320380926132202 +I0831 15:33:47.590636 139921981437696 logging_writer.py:48] [80800] global_step=80800, grad_norm=4.946004390716553, loss=1.0686345100402832 +I0831 15:34:14.939133 139921973044992 logging_writer.py:48] [80900] global_step=80900, grad_norm=5.132371425628662, loss=1.1387947797775269 +I0831 15:34:42.129187 139921981437696 logging_writer.py:48] [81000] global_step=81000, grad_norm=5.04193639755249, loss=1.0906144380569458 +I0831 15:35:09.236593 139921973044992 logging_writer.py:48] [81100] global_step=81100, grad_norm=5.210312366485596, loss=1.1718459129333496 +I0831 15:35:36.373287 139921981437696 logging_writer.py:48] [81200] global_step=81200, grad_norm=4.899787425994873, loss=1.0311810970306396 +I0831 15:36:03.566257 139921973044992 logging_writer.py:48] [81300] global_step=81300, grad_norm=5.077867031097412, loss=1.090095043182373 +I0831 15:36:30.693756 139921981437696 logging_writer.py:48] [81400] global_step=81400, grad_norm=5.131804466247559, loss=1.0895289182662964 +I0831 15:36:57.797051 139921973044992 logging_writer.py:48] [81500] global_step=81500, grad_norm=5.088109016418457, loss=1.1036720275878906 +I0831 15:37:24.989259 139921981437696 logging_writer.py:48] [81600] global_step=81600, grad_norm=5.099238872528076, loss=1.0408028364181519 +I0831 15:37:52.134333 139921973044992 logging_writer.py:48] [81700] global_step=81700, grad_norm=4.827699184417725, loss=1.1043035984039307 +I0831 15:38:19.243012 139921981437696 logging_writer.py:48] [81800] global_step=81800, grad_norm=4.97106409072876, loss=1.0910472869873047 +I0831 15:38:46.430058 139921973044992 logging_writer.py:48] [81900] global_step=81900, grad_norm=5.013543605804443, loss=1.0581581592559814 +I0831 15:39:13.767234 139921981437696 logging_writer.py:48] [82000] global_step=82000, grad_norm=4.964090347290039, loss=1.0352638959884644 +I0831 15:39:40.904325 139921973044992 logging_writer.py:48] [82100] global_step=82100, grad_norm=5.2746357917785645, loss=1.0858242511749268 +I0831 15:40:08.093380 139921981437696 logging_writer.py:48] [82200] global_step=82200, grad_norm=5.00617790222168, loss=1.0245529413223267 +I0831 15:40:35.237955 139921973044992 logging_writer.py:48] [82300] global_step=82300, grad_norm=5.070066452026367, loss=1.059300184249878 +I0831 15:41:02.341930 139921981437696 logging_writer.py:48] [82400] global_step=82400, grad_norm=4.902980804443359, loss=1.0498554706573486 +I0831 15:41:29.560267 139921973044992 logging_writer.py:48] [82500] global_step=82500, grad_norm=5.045276641845703, loss=1.0761722326278687 +I0831 15:41:56.724766 139921981437696 logging_writer.py:48] [82600] global_step=82600, grad_norm=4.914961814880371, loss=1.0793988704681396 +I0831 15:42:23.845378 139921973044992 logging_writer.py:48] [82700] global_step=82700, grad_norm=5.097283363342285, loss=1.1361911296844482 +I0831 15:42:51.039873 139921981437696 logging_writer.py:48] [82800] global_step=82800, grad_norm=4.956486225128174, loss=1.106457233428955 +I0831 15:43:18.170435 139921973044992 logging_writer.py:48] [82900] global_step=82900, grad_norm=5.235239505767822, loss=1.017351746559143 +I0831 15:43:45.312582 139921981437696 logging_writer.py:48] [83000] global_step=83000, grad_norm=4.994144439697266, loss=0.9846730828285217 +I0831 15:44:12.762395 139921973044992 logging_writer.py:48] [83100] global_step=83100, grad_norm=4.959412574768066, loss=1.0473713874816895 +I0831 15:44:39.885916 139921981437696 logging_writer.py:48] [83200] global_step=83200, grad_norm=5.0453901290893555, loss=1.0236718654632568 +I0831 15:45:07.014809 139921973044992 logging_writer.py:48] [83300] global_step=83300, grad_norm=5.161510944366455, loss=1.058764100074768 +I0831 15:45:34.202309 139921981437696 logging_writer.py:48] [83400] global_step=83400, grad_norm=5.181464195251465, loss=1.0942294597625732 +I0831 15:46:01.324841 139921973044992 logging_writer.py:48] [83500] global_step=83500, grad_norm=5.074193954467773, loss=1.0586225986480713 +I0831 15:46:28.453577 139921981437696 logging_writer.py:48] [83600] global_step=83600, grad_norm=5.171372413635254, loss=1.00095534324646 +I0831 15:46:55.632302 139921973044992 logging_writer.py:48] [83700] global_step=83700, grad_norm=5.123868465423584, loss=1.0444600582122803 +I0831 15:47:22.776292 139921981437696 logging_writer.py:48] [83800] global_step=83800, grad_norm=5.154863357543945, loss=1.072662353515625 +I0831 15:47:49.873981 139921973044992 logging_writer.py:48] [83900] global_step=83900, grad_norm=5.078988552093506, loss=1.1255805492401123 +I0831 15:48:17.064960 139921981437696 logging_writer.py:48] [84000] global_step=84000, grad_norm=4.7159833908081055, loss=0.9851527810096741 +I0831 15:48:44.195688 139921973044992 logging_writer.py:48] [84100] global_step=84100, grad_norm=5.438589096069336, loss=1.16544508934021 +I0831 15:49:11.513703 139921981437696 logging_writer.py:48] [84200] global_step=84200, grad_norm=5.202115535736084, loss=1.019557237625122 +I0831 15:49:38.713744 139921973044992 logging_writer.py:48] [84300] global_step=84300, grad_norm=5.058207988739014, loss=1.1060690879821777 +I0831 15:50:05.819808 139921981437696 logging_writer.py:48] [84400] global_step=84400, grad_norm=4.837040901184082, loss=1.04816734790802 +I0831 15:50:33.004980 139921973044992 logging_writer.py:48] [84500] global_step=84500, grad_norm=5.120960235595703, loss=1.1078611612319946 +I0831 15:51:00.215831 139921981437696 logging_writer.py:48] [84600] global_step=84600, grad_norm=4.86998987197876, loss=1.0453006029129028 +I0831 15:51:27.336015 139921973044992 logging_writer.py:48] [84700] global_step=84700, grad_norm=5.04424524307251, loss=1.083539605140686 +I0831 15:51:54.441744 139921981437696 logging_writer.py:48] [84800] global_step=84800, grad_norm=5.222089767456055, loss=1.130238652229309 +I0831 15:52:21.607055 139921973044992 logging_writer.py:48] [84900] global_step=84900, grad_norm=4.957637310028076, loss=1.0302064418792725 +I0831 15:52:48.714161 139921981437696 logging_writer.py:48] [85000] global_step=85000, grad_norm=4.869359016418457, loss=1.1109617948532104 +I0831 15:53:15.814024 139921973044992 logging_writer.py:48] [85100] global_step=85100, grad_norm=5.0345869064331055, loss=1.08828866481781 +I0831 15:53:43.225287 139921981437696 logging_writer.py:48] [85200] global_step=85200, grad_norm=5.167697906494141, loss=1.08962082862854 +I0831 15:54:10.336416 139921973044992 logging_writer.py:48] [85300] global_step=85300, grad_norm=5.000251293182373, loss=1.0972663164138794 +I0831 15:54:37.441298 139921981437696 logging_writer.py:48] [85400] global_step=85400, grad_norm=4.837368011474609, loss=0.9755046963691711 +I0831 15:55:04.634943 139921973044992 logging_writer.py:48] [85500] global_step=85500, grad_norm=4.785784721374512, loss=1.0627222061157227 +I0831 15:55:31.767078 139921981437696 logging_writer.py:48] [85600] global_step=85600, grad_norm=4.984625339508057, loss=1.104077935218811 +I0831 15:55:58.873961 139921973044992 logging_writer.py:48] [85700] global_step=85700, grad_norm=4.884222030639648, loss=1.0519647598266602 +I0831 15:56:26.039497 139921981437696 logging_writer.py:48] [85800] global_step=85800, grad_norm=5.123117923736572, loss=1.0741729736328125 +I0831 15:56:53.164622 139921973044992 logging_writer.py:48] [85900] global_step=85900, grad_norm=5.249913215637207, loss=1.1543035507202148 +I0831 15:57:20.296742 139921981437696 logging_writer.py:48] [86000] global_step=86000, grad_norm=4.974696636199951, loss=1.0515488386154175 +I0831 15:57:47.499129 139921973044992 logging_writer.py:48] [86100] global_step=86100, grad_norm=5.0895795822143555, loss=1.0050456523895264 +I0831 15:58:14.623135 139921981437696 logging_writer.py:48] [86200] global_step=86200, grad_norm=5.067255020141602, loss=1.0853183269500732 +I0831 15:58:41.956840 139921973044992 logging_writer.py:48] [86300] global_step=86300, grad_norm=4.729895114898682, loss=1.0786405801773071 +I0831 15:59:09.125277 139921981437696 logging_writer.py:48] [86400] global_step=86400, grad_norm=4.949068069458008, loss=1.1230840682983398 +I0831 15:59:36.226943 139921973044992 logging_writer.py:48] [86500] global_step=86500, grad_norm=5.0749006271362305, loss=1.1195447444915771 +I0831 16:00:03.335024 139921981437696 logging_writer.py:48] [86600] global_step=86600, grad_norm=5.099456787109375, loss=1.0718032121658325 +I0831 16:00:30.500654 139921973044992 logging_writer.py:48] [86700] global_step=86700, grad_norm=4.877786159515381, loss=1.0005264282226562 +I0831 16:00:57.653510 139921981437696 logging_writer.py:48] [86800] global_step=86800, grad_norm=4.819756507873535, loss=1.0957746505737305 +I0831 16:01:24.786390 139921973044992 logging_writer.py:48] [86900] global_step=86900, grad_norm=4.950500011444092, loss=1.067353367805481 +I0831 16:01:51.961876 139921981437696 logging_writer.py:48] [87000] global_step=87000, grad_norm=4.930877208709717, loss=1.0989254713058472 +I0831 16:02:19.088351 139921973044992 logging_writer.py:48] [87100] global_step=87100, grad_norm=4.9982147216796875, loss=1.0931754112243652 +I0831 16:02:46.185979 139921981437696 logging_writer.py:48] [87200] global_step=87200, grad_norm=5.060156345367432, loss=1.0923956632614136 +I0831 16:03:13.343705 139921973044992 logging_writer.py:48] [87300] global_step=87300, grad_norm=5.031198024749756, loss=1.1320167779922485 +I0831 16:03:37.832631 140117123622080 spec.py:333] Evaluating on the training split. +I0831 16:03:46.305410 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 16:03:55.862621 140117123622080 spec.py:363] Evaluating on the test split. +I0831 16:03:56.747779 140117123622080 submission_runner.py:516] Time since start: 24358.09s, Step: 87391, {'train/accuracy': Array(0.91685265, dtype=float32), 'train/loss': Array(0.29532543, dtype=float32), 'validation/accuracy': Array(0.74799997, dtype=float32), 'validation/loss': Array(1.0682663, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6222, dtype=float32), 'test/loss': Array(1.8081042, dtype=float32), 'test/num_examples': 10000, 'score': 24004.935247659683, 'total_duration': 24358.087627649307, 'accumulated_submission_time': 24004.935247659683, 'accumulated_eval_time': 351.5444669723511, 'accumulated_logging_time': 0.9268791675567627} +I0831 16:03:56.800516 139921981437696 logging_writer.py:48] [87391] accumulated_eval_time=351.544, accumulated_logging_time=0.926879, accumulated_submission_time=24004.9, global_step=87391, preemption_count=0, score=24004.9, test/accuracy=0.6222000122070312, test/loss=1.808104157447815, test/num_examples=10000, total_duration=24358.1, train/accuracy=0.9168526530265808, train/loss=0.2953254282474518, validation/accuracy=0.7479999661445618, validation/loss=1.0682662725448608, validation/num_examples=50000 +I0831 16:03:59.660812 139921973044992 logging_writer.py:48] [87400] global_step=87400, grad_norm=5.015869140625, loss=1.0852726697921753 +I0831 16:04:26.847290 139921981437696 logging_writer.py:48] [87500] global_step=87500, grad_norm=4.934142589569092, loss=1.0122350454330444 +I0831 16:04:54.011933 139921973044992 logging_writer.py:48] [87600] global_step=87600, grad_norm=4.785815238952637, loss=1.0094504356384277 +I0831 16:05:21.114662 139921981437696 logging_writer.py:48] [87700] global_step=87700, grad_norm=4.819437503814697, loss=1.0377020835876465 +I0831 16:05:48.229192 139921973044992 logging_writer.py:48] [87800] global_step=87800, grad_norm=5.248297691345215, loss=1.0672426223754883 +I0831 16:06:15.411919 139921981437696 logging_writer.py:48] [87900] global_step=87900, grad_norm=5.049840927124023, loss=1.182447910308838 +I0831 16:06:42.554839 139921973044992 logging_writer.py:48] [88000] global_step=88000, grad_norm=4.909049987792969, loss=1.1345601081848145 +I0831 16:07:09.704229 139921981437696 logging_writer.py:48] [88100] global_step=88100, grad_norm=4.871236324310303, loss=1.0541579723358154 +I0831 16:07:36.905705 139921973044992 logging_writer.py:48] [88200] global_step=88200, grad_norm=4.71498966217041, loss=0.980340838432312 +I0831 16:08:04.032468 139921981437696 logging_writer.py:48] [88300] global_step=88300, grad_norm=4.851357460021973, loss=1.0380288362503052 +I0831 16:08:31.170693 139921973044992 logging_writer.py:48] [88400] global_step=88400, grad_norm=5.00673770904541, loss=1.0429959297180176 +I0831 16:08:58.612538 139921981437696 logging_writer.py:48] [88500] global_step=88500, grad_norm=4.872030735015869, loss=1.0394172668457031 +I0831 16:09:25.753389 139921973044992 logging_writer.py:48] [88600] global_step=88600, grad_norm=4.863335132598877, loss=1.1059322357177734 +I0831 16:09:52.876737 139921981437696 logging_writer.py:48] [88700] global_step=88700, grad_norm=4.751591682434082, loss=1.0520637035369873 +I0831 16:10:20.079880 139921973044992 logging_writer.py:48] [88800] global_step=88800, grad_norm=4.943983554840088, loss=0.9909031391143799 +I0831 16:10:47.219199 139921981437696 logging_writer.py:48] [88900] global_step=88900, grad_norm=4.988236427307129, loss=1.0989186763763428 +I0831 16:11:14.363308 139921973044992 logging_writer.py:48] [89000] global_step=89000, grad_norm=5.223843574523926, loss=1.0100619792938232 +I0831 16:11:41.573462 139921981437696 logging_writer.py:48] [89100] global_step=89100, grad_norm=5.099555492401123, loss=1.146026849746704 +I0831 16:12:08.690459 139921973044992 logging_writer.py:48] [89200] global_step=89200, grad_norm=4.832293510437012, loss=0.9931567907333374 +I0831 16:12:35.787643 139921981437696 logging_writer.py:48] [89300] global_step=89300, grad_norm=4.909796237945557, loss=1.0210639238357544 +I0831 16:13:02.978947 139921973044992 logging_writer.py:48] [89400] global_step=89400, grad_norm=5.039376735687256, loss=1.0735878944396973 +I0831 16:13:30.304175 139921981437696 logging_writer.py:48] [89500] global_step=89500, grad_norm=4.947811126708984, loss=1.0446226596832275 +I0831 16:13:57.425578 139921973044992 logging_writer.py:48] [89600] global_step=89600, grad_norm=5.052218914031982, loss=1.1364977359771729 +I0831 16:14:24.608335 139921981437696 logging_writer.py:48] [89700] global_step=89700, grad_norm=4.959063529968262, loss=1.1110379695892334 +I0831 16:14:51.763933 139921973044992 logging_writer.py:48] [89800] global_step=89800, grad_norm=4.735313415527344, loss=0.9469889998435974 +I0831 16:15:18.876302 139921981437696 logging_writer.py:48] [89900] global_step=89900, grad_norm=5.27932596206665, loss=1.118708610534668 +I0831 16:15:46.064100 139921973044992 logging_writer.py:48] [90000] global_step=90000, grad_norm=5.174208641052246, loss=1.0775628089904785 +I0831 16:16:13.172710 139921981437696 logging_writer.py:48] [90100] global_step=90100, grad_norm=4.795698642730713, loss=0.9834789037704468 +I0831 16:16:40.271463 139921973044992 logging_writer.py:48] [90200] global_step=90200, grad_norm=5.2726731300354, loss=1.1191699504852295 +I0831 16:17:07.456261 139921981437696 logging_writer.py:48] [90300] global_step=90300, grad_norm=5.660727024078369, loss=1.1786872148513794 +I0831 16:17:34.584173 139921973044992 logging_writer.py:48] [90400] global_step=90400, grad_norm=5.172414779663086, loss=1.146747350692749 +I0831 16:18:01.710936 139921981437696 logging_writer.py:48] [90500] global_step=90500, grad_norm=5.10603666305542, loss=1.0421147346496582 +I0831 16:18:29.112364 139921973044992 logging_writer.py:48] [90600] global_step=90600, grad_norm=4.955665588378906, loss=1.0542106628417969 +I0831 16:18:56.230792 139921981437696 logging_writer.py:48] [90700] global_step=90700, grad_norm=4.948988437652588, loss=1.077939748764038 +I0831 16:19:23.355799 139921973044992 logging_writer.py:48] [90800] global_step=90800, grad_norm=4.924530982971191, loss=1.1202385425567627 +I0831 16:19:50.532552 139921981437696 logging_writer.py:48] [90900] global_step=90900, grad_norm=4.879987716674805, loss=1.0456531047821045 +I0831 16:20:17.649899 139921973044992 logging_writer.py:48] [91000] global_step=91000, grad_norm=4.910399913787842, loss=1.0538737773895264 +I0831 16:20:44.790298 139921981437696 logging_writer.py:48] [91100] global_step=91100, grad_norm=5.165282249450684, loss=1.0837771892547607 +I0831 16:21:12.009988 139921973044992 logging_writer.py:48] [91200] global_step=91200, grad_norm=4.8610310554504395, loss=0.9770346879959106 +I0831 16:21:39.144471 139921981437696 logging_writer.py:48] [91300] global_step=91300, grad_norm=4.913155555725098, loss=1.0742205381393433 +I0831 16:22:06.258479 139921973044992 logging_writer.py:48] [91400] global_step=91400, grad_norm=4.9798479080200195, loss=1.0733672380447388 +I0831 16:22:33.456811 139921981437696 logging_writer.py:48] [91500] global_step=91500, grad_norm=5.006542682647705, loss=1.0828425884246826 +I0831 16:23:00.795861 139921973044992 logging_writer.py:48] [91600] global_step=91600, grad_norm=5.278910160064697, loss=1.0683127641677856 +I0831 16:23:27.930214 139921981437696 logging_writer.py:48] [91700] global_step=91700, grad_norm=5.118814945220947, loss=1.0788077116012573 +I0831 16:23:55.119932 139921973044992 logging_writer.py:48] [91800] global_step=91800, grad_norm=5.117591381072998, loss=1.0877797603607178 +I0831 16:24:22.222014 139921981437696 logging_writer.py:48] [91900] global_step=91900, grad_norm=5.371670246124268, loss=1.0794960260391235 +I0831 16:24:49.341783 139921973044992 logging_writer.py:48] [92000] global_step=92000, grad_norm=4.974464416503906, loss=1.0612086057662964 +I0831 16:25:16.525414 139921981437696 logging_writer.py:48] [92100] global_step=92100, grad_norm=5.135711193084717, loss=1.0713037252426147 +I0831 16:25:43.640805 139921973044992 logging_writer.py:48] [92200] global_step=92200, grad_norm=5.039716720581055, loss=1.0348458290100098 +I0831 16:26:10.762953 139921981437696 logging_writer.py:48] [92300] global_step=92300, grad_norm=5.326661109924316, loss=1.1817471981048584 +I0831 16:26:37.966988 139921973044992 logging_writer.py:48] [92400] global_step=92400, grad_norm=5.176074981689453, loss=1.1199016571044922 +I0831 16:27:05.114243 139921981437696 logging_writer.py:48] [92500] global_step=92500, grad_norm=5.160790920257568, loss=1.1366130113601685 +I0831 16:27:32.273334 139921973044992 logging_writer.py:48] [92600] global_step=92600, grad_norm=4.847156047821045, loss=1.0833344459533691 +I0831 16:27:59.647379 139921981437696 logging_writer.py:48] [92700] global_step=92700, grad_norm=4.866337776184082, loss=1.017371654510498 +I0831 16:28:26.782457 139921973044992 logging_writer.py:48] [92800] global_step=92800, grad_norm=4.731010437011719, loss=0.9778791666030884 +I0831 16:28:53.895708 139921981437696 logging_writer.py:48] [92900] global_step=92900, grad_norm=5.437882423400879, loss=1.055895447731018 +I0831 16:29:21.076274 139921973044992 logging_writer.py:48] [93000] global_step=93000, grad_norm=5.451284885406494, loss=1.122131586074829 +I0831 16:29:48.190765 139921981437696 logging_writer.py:48] [93100] global_step=93100, grad_norm=5.274907112121582, loss=1.1346938610076904 +I0831 16:30:15.287895 139921973044992 logging_writer.py:48] [93200] global_step=93200, grad_norm=4.731269836425781, loss=0.9297258257865906 +I0831 16:30:42.487869 139921981437696 logging_writer.py:48] [93300] global_step=93300, grad_norm=4.782537937164307, loss=1.0598220825195312 +I0831 16:31:09.603862 139921973044992 logging_writer.py:48] [93400] global_step=93400, grad_norm=4.7977294921875, loss=0.9936653971672058 +I0831 16:31:36.703683 139921981437696 logging_writer.py:48] [93500] global_step=93500, grad_norm=5.594959735870361, loss=1.1642554998397827 +I0831 16:32:03.897655 139921973044992 logging_writer.py:48] [93600] global_step=93600, grad_norm=4.765223503112793, loss=1.0357654094696045 +I0831 16:32:31.217403 139921981437696 logging_writer.py:48] [93700] global_step=93700, grad_norm=5.343916893005371, loss=1.1298787593841553 +I0831 16:32:58.320787 139921973044992 logging_writer.py:48] [93800] global_step=93800, grad_norm=4.928273677825928, loss=1.0203505754470825 +I0831 16:33:25.524907 139921981437696 logging_writer.py:48] [93900] global_step=93900, grad_norm=5.0251593589782715, loss=1.0504157543182373 +I0831 16:33:52.632230 139921973044992 logging_writer.py:48] [94000] global_step=94000, grad_norm=4.854905605316162, loss=1.0002855062484741 +I0831 16:34:19.755504 139921981437696 logging_writer.py:48] [94100] global_step=94100, grad_norm=5.069383144378662, loss=1.021092176437378 +I0831 16:34:46.945108 139921973044992 logging_writer.py:48] [94200] global_step=94200, grad_norm=5.237695217132568, loss=1.131460189819336 +I0831 16:35:14.062508 139921981437696 logging_writer.py:48] [94300] global_step=94300, grad_norm=5.206452369689941, loss=1.130422592163086 +I0831 16:35:41.189255 139921973044992 logging_writer.py:48] [94400] global_step=94400, grad_norm=5.077343940734863, loss=1.0562909841537476 +I0831 16:36:08.376635 139921981437696 logging_writer.py:48] [94500] global_step=94500, grad_norm=4.61035680770874, loss=0.9722060561180115 +I0831 16:36:35.497645 139921973044992 logging_writer.py:48] [94600] global_step=94600, grad_norm=5.036670207977295, loss=1.1185842752456665 +I0831 16:37:02.648437 139921981437696 logging_writer.py:48] [94700] global_step=94700, grad_norm=4.972769260406494, loss=1.07618248462677 +I0831 16:37:12.860914 140117123622080 spec.py:333] Evaluating on the training split. +I0831 16:37:20.238226 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 16:37:30.553167 140117123622080 spec.py:363] Evaluating on the test split. +I0831 16:37:31.414953 140117123622080 submission_runner.py:516] Time since start: 26372.75s, Step: 94739, {'train/accuracy': Array(0.9225127, dtype=float32), 'train/loss': Array(0.27492654, dtype=float32), 'validation/accuracy': Array(0.74825996, dtype=float32), 'validation/loss': Array(1.0668399, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6223, dtype=float32), 'test/loss': Array(1.8116835, dtype=float32), 'test/num_examples': 10000, 'score': 26000.92212820053, 'total_duration': 26372.754496335983, 'accumulated_submission_time': 26000.92212820053, 'accumulated_eval_time': 370.09621691703796, 'accumulated_logging_time': 1.0008885860443115} +I0831 16:37:31.471959 139921973044992 logging_writer.py:48] [94739] accumulated_eval_time=370.096, accumulated_logging_time=1.00089, accumulated_submission_time=26000.9, global_step=94739, preemption_count=0, score=26000.9, test/accuracy=0.6223000288009644, test/loss=1.8116835355758667, test/num_examples=10000, total_duration=26372.8, train/accuracy=0.9225127100944519, train/loss=0.2749265432357788, validation/accuracy=0.748259961605072, validation/loss=1.0668399333953857, validation/num_examples=50000 +I0831 16:37:48.884779 139921981437696 logging_writer.py:48] [94800] global_step=94800, grad_norm=4.971933841705322, loss=1.029255986213684 +I0831 16:38:15.970484 139921973044992 logging_writer.py:48] [94900] global_step=94900, grad_norm=5.387161731719971, loss=1.2059314250946045 +I0831 16:38:43.073299 139921981437696 logging_writer.py:48] [95000] global_step=95000, grad_norm=4.598382472991943, loss=1.0087381601333618 +I0831 16:39:10.243186 139921973044992 logging_writer.py:48] [95100] global_step=95100, grad_norm=4.945769309997559, loss=1.117740273475647 +I0831 16:39:37.345891 139921981437696 logging_writer.py:48] [95200] global_step=95200, grad_norm=4.933661460876465, loss=1.0619826316833496 +I0831 16:40:04.449617 139921973044992 logging_writer.py:48] [95300] global_step=95300, grad_norm=4.812836647033691, loss=1.0082528591156006 +I0831 16:40:31.630868 139921981437696 logging_writer.py:48] [95400] global_step=95400, grad_norm=4.993622303009033, loss=1.0789048671722412 +I0831 16:40:58.755468 139921973044992 logging_writer.py:48] [95500] global_step=95500, grad_norm=4.945078372955322, loss=1.029114007949829 +I0831 16:41:25.850048 139921981437696 logging_writer.py:48] [95600] global_step=95600, grad_norm=5.506173610687256, loss=1.0796664953231812 +I0831 16:41:53.027635 139921973044992 logging_writer.py:48] [95700] global_step=95700, grad_norm=4.812524318695068, loss=1.0255804061889648 +I0831 16:42:20.138133 139921981437696 logging_writer.py:48] [95800] global_step=95800, grad_norm=5.079554557800293, loss=1.0806682109832764 +I0831 16:42:47.446392 139921973044992 logging_writer.py:48] [95900] global_step=95900, grad_norm=4.843756675720215, loss=1.0355265140533447 +I0831 16:43:14.636779 139921981437696 logging_writer.py:48] [96000] global_step=96000, grad_norm=5.344239711761475, loss=1.0372860431671143 +I0831 16:43:41.769782 139921973044992 logging_writer.py:48] [96100] global_step=96100, grad_norm=5.046606540679932, loss=1.069011926651001 +I0831 16:44:08.914793 139921981437696 logging_writer.py:48] [96200] global_step=96200, grad_norm=4.862529277801514, loss=1.0125499963760376 +I0831 16:44:36.067824 139921973044992 logging_writer.py:48] [96300] global_step=96300, grad_norm=4.772211074829102, loss=0.9638989567756653 +I0831 16:45:03.187988 139921981437696 logging_writer.py:48] [96400] global_step=96400, grad_norm=5.297403335571289, loss=1.0052752494812012 +I0831 16:45:30.305159 139921973044992 logging_writer.py:48] [96500] global_step=96500, grad_norm=4.831307411193848, loss=1.0771822929382324 +I0831 16:45:57.633917 139921981437696 logging_writer.py:48] [96600] global_step=96600, grad_norm=4.93776273727417, loss=1.0503265857696533 +I0831 16:46:24.720684 139921973044992 logging_writer.py:48] [96700] global_step=96700, grad_norm=4.926252841949463, loss=1.0314359664916992 +I0831 16:46:51.854652 139921981437696 logging_writer.py:48] [96800] global_step=96800, grad_norm=4.806466579437256, loss=1.0545587539672852 +I0831 16:47:19.287585 139921973044992 logging_writer.py:48] [96900] global_step=96900, grad_norm=5.284399032592773, loss=1.0499145984649658 +I0831 16:47:46.399607 139921981437696 logging_writer.py:48] [97000] global_step=97000, grad_norm=5.063129425048828, loss=1.0725464820861816 +I0831 16:48:13.499499 139921973044992 logging_writer.py:48] [97100] global_step=97100, grad_norm=4.952840328216553, loss=1.0261549949645996 +I0831 16:48:40.663969 139921981437696 logging_writer.py:48] [97200] global_step=97200, grad_norm=4.983527183532715, loss=1.060793161392212 +I0831 16:49:07.774064 139921973044992 logging_writer.py:48] [97300] global_step=97300, grad_norm=4.850233554840088, loss=0.9238430857658386 +I0831 16:49:34.927176 139921981437696 logging_writer.py:48] [97400] global_step=97400, grad_norm=4.886474609375, loss=0.9270267486572266 +I0831 16:50:02.119071 139921973044992 logging_writer.py:48] [97500] global_step=97500, grad_norm=4.906455039978027, loss=1.0014539957046509 +I0831 16:50:29.256345 139921981437696 logging_writer.py:48] [97600] global_step=97600, grad_norm=4.964179992675781, loss=1.0640369653701782 +I0831 16:50:56.393060 139921973044992 logging_writer.py:48] [97700] global_step=97700, grad_norm=4.991407871246338, loss=1.0406910181045532 +I0831 16:51:23.609820 139921981437696 logging_writer.py:48] [97800] global_step=97800, grad_norm=5.157801628112793, loss=1.04185152053833 +I0831 16:51:50.761999 139921973044992 logging_writer.py:48] [97900] global_step=97900, grad_norm=4.951785087585449, loss=1.0414764881134033 +I0831 16:52:18.110990 139921981437696 logging_writer.py:48] [98000] global_step=98000, grad_norm=4.865335941314697, loss=0.9932011961936951 +I0831 16:52:45.336139 139921973044992 logging_writer.py:48] [98100] global_step=98100, grad_norm=4.908229827880859, loss=1.08207106590271 +I0831 16:53:12.446389 139921981437696 logging_writer.py:48] [98200] global_step=98200, grad_norm=5.233084201812744, loss=1.1257624626159668 +I0831 16:53:39.566317 139921973044992 logging_writer.py:48] [98300] global_step=98300, grad_norm=4.57635498046875, loss=0.9375048875808716 +I0831 16:54:06.767330 139921981437696 logging_writer.py:48] [98400] global_step=98400, grad_norm=4.648318290710449, loss=0.9683881998062134 +I0831 16:54:33.923656 139921973044992 logging_writer.py:48] [98500] global_step=98500, grad_norm=5.20836877822876, loss=1.0753921270370483 +I0831 16:55:01.043568 139921981437696 logging_writer.py:48] [98600] global_step=98600, grad_norm=5.0288896560668945, loss=1.0395619869232178 +I0831 16:55:28.223949 139921973044992 logging_writer.py:48] [98700] global_step=98700, grad_norm=4.966915130615234, loss=1.0096914768218994 +I0831 16:55:55.359783 139921981437696 logging_writer.py:48] [98800] global_step=98800, grad_norm=5.030628204345703, loss=1.048041820526123 +I0831 16:56:22.520781 139921973044992 logging_writer.py:48] [98900] global_step=98900, grad_norm=4.924155235290527, loss=0.9764904975891113 +I0831 16:56:49.927790 139921981437696 logging_writer.py:48] [99000] global_step=99000, grad_norm=4.966123104095459, loss=1.0577075481414795 +I0831 16:57:17.075845 139921973044992 logging_writer.py:48] [99100] global_step=99100, grad_norm=4.928457260131836, loss=1.0316542387008667 +I0831 16:57:44.222620 139921981437696 logging_writer.py:48] [99200] global_step=99200, grad_norm=4.73695182800293, loss=1.0671933889389038 +I0831 16:58:11.422719 139921973044992 logging_writer.py:48] [99300] global_step=99300, grad_norm=4.940418720245361, loss=1.032058835029602 +I0831 16:58:38.562106 139921981437696 logging_writer.py:48] [99400] global_step=99400, grad_norm=4.9105377197265625, loss=1.0081161260604858 +I0831 16:59:05.655372 139921973044992 logging_writer.py:48] [99500] global_step=99500, grad_norm=4.896152973175049, loss=1.0679348707199097 +I0831 16:59:32.848101 139921981437696 logging_writer.py:48] [99600] global_step=99600, grad_norm=5.053324222564697, loss=1.0520259141921997 +I0831 16:59:59.966859 139921973044992 logging_writer.py:48] [99700] global_step=99700, grad_norm=4.7733683586120605, loss=0.9843765497207642 +I0831 17:00:27.103710 139921981437696 logging_writer.py:48] [99800] global_step=99800, grad_norm=5.040751934051514, loss=1.036278486251831 +I0831 17:00:54.320984 139921973044992 logging_writer.py:48] [99900] global_step=99900, grad_norm=4.678987979888916, loss=1.0400317907333374 +I0831 17:01:21.427415 139921981437696 logging_writer.py:48] [100000] global_step=100000, grad_norm=5.248448848724365, loss=1.0327235460281372 +I0831 17:01:48.776822 139921973044992 logging_writer.py:48] [100100] global_step=100100, grad_norm=4.770143508911133, loss=1.0261389017105103 +I0831 17:02:15.972619 139921981437696 logging_writer.py:48] [100200] global_step=100200, grad_norm=5.7654500007629395, loss=1.0709962844848633 +I0831 17:02:43.088274 139921973044992 logging_writer.py:48] [100300] global_step=100300, grad_norm=4.840331554412842, loss=0.9635717272758484 +I0831 17:03:10.215442 139921981437696 logging_writer.py:48] [100400] global_step=100400, grad_norm=4.840884208679199, loss=1.0655407905578613 +I0831 17:03:37.387327 139921973044992 logging_writer.py:48] [100500] global_step=100500, grad_norm=4.790011405944824, loss=1.0359106063842773 +I0831 17:04:04.505374 139921981437696 logging_writer.py:48] [100600] global_step=100600, grad_norm=5.409252643585205, loss=1.0338610410690308 +I0831 17:04:31.630091 139921973044992 logging_writer.py:48] [100700] global_step=100700, grad_norm=5.58707332611084, loss=1.0350655317306519 +I0831 17:04:58.826701 139921981437696 logging_writer.py:48] [100800] global_step=100800, grad_norm=5.060688018798828, loss=1.0704023838043213 +I0831 17:05:25.949663 139921973044992 logging_writer.py:48] [100900] global_step=100900, grad_norm=5.127038478851318, loss=1.1521246433258057 +I0831 17:05:53.079058 139921981437696 logging_writer.py:48] [101000] global_step=101000, grad_norm=4.926717281341553, loss=1.03409743309021 +I0831 17:06:20.489116 139921973044992 logging_writer.py:48] [101100] global_step=101100, grad_norm=5.027322292327881, loss=1.0406004190444946 +I0831 17:06:47.637085 139921981437696 logging_writer.py:48] [101200] global_step=101200, grad_norm=5.033076286315918, loss=1.0904371738433838 +I0831 17:07:14.760112 139921973044992 logging_writer.py:48] [101300] global_step=101300, grad_norm=4.825054168701172, loss=1.040416955947876 +I0831 17:07:41.959133 139921981437696 logging_writer.py:48] [101400] global_step=101400, grad_norm=4.843383312225342, loss=1.0063362121582031 +I0831 17:08:09.066161 139921973044992 logging_writer.py:48] [101500] global_step=101500, grad_norm=4.744531631469727, loss=0.9552044868469238 +I0831 17:08:36.187361 139921981437696 logging_writer.py:48] [101600] global_step=101600, grad_norm=4.973448276519775, loss=0.969937801361084 +I0831 17:09:03.369534 139921973044992 logging_writer.py:48] [101700] global_step=101700, grad_norm=4.982604503631592, loss=1.017988920211792 +I0831 17:09:30.476511 139921981437696 logging_writer.py:48] [101800] global_step=101800, grad_norm=4.846097946166992, loss=1.0159226655960083 +I0831 17:09:57.585523 139921973044992 logging_writer.py:48] [101900] global_step=101900, grad_norm=5.068576812744141, loss=1.0338603258132935 +I0831 17:10:24.794069 139921981437696 logging_writer.py:48] [102000] global_step=102000, grad_norm=4.601844310760498, loss=1.0067147016525269 +I0831 17:10:47.435871 140117123622080 spec.py:333] Evaluating on the training split. +I0831 17:10:54.791000 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 17:11:04.372564 140117123622080 spec.py:363] Evaluating on the test split. +I0831 17:11:05.261120 140117123622080 submission_runner.py:516] Time since start: 28386.60s, Step: 102085, {'train/accuracy': Array(0.92524314, dtype=float32), 'train/loss': Array(0.26571548, dtype=float32), 'validation/accuracy': Array(0.74825996, dtype=float32), 'validation/loss': Array(1.0670127, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62280005, dtype=float32), 'test/loss': Array(1.8185028, dtype=float32), 'test/num_examples': 10000, 'score': 27996.79708981514, 'total_duration': 28386.60117483139, 'accumulated_submission_time': 27996.79708981514, 'accumulated_eval_time': 387.91968870162964, 'accumulated_logging_time': 1.094411849975586} +I0831 17:11:05.324524 139921973044992 logging_writer.py:48] [102085] accumulated_eval_time=387.92, accumulated_logging_time=1.09441, accumulated_submission_time=27996.8, global_step=102085, preemption_count=0, score=27996.8, test/accuracy=0.6228000521659851, test/loss=1.8185027837753296, test/num_examples=10000, total_duration=28386.6, train/accuracy=0.9252431392669678, train/loss=0.26571547985076904, validation/accuracy=0.748259961605072, validation/loss=1.0670126676559448, validation/num_examples=50000 +I0831 17:11:09.882044 139921981437696 logging_writer.py:48] [102100] global_step=102100, grad_norm=4.816553592681885, loss=1.0145237445831299 +I0831 17:11:37.028964 139921973044992 logging_writer.py:48] [102200] global_step=102200, grad_norm=4.984592914581299, loss=1.1294680833816528 +I0831 17:12:04.425789 139921981437696 logging_writer.py:48] [102300] global_step=102300, grad_norm=5.106274127960205, loss=0.9978965520858765 +I0831 17:12:31.558324 139921973044992 logging_writer.py:48] [102400] global_step=102400, grad_norm=5.079766273498535, loss=1.121262550354004 +I0831 17:12:58.708250 139921981437696 logging_writer.py:48] [102500] global_step=102500, grad_norm=4.828175067901611, loss=1.0665109157562256 +I0831 17:13:25.881018 139921973044992 logging_writer.py:48] [102600] global_step=102600, grad_norm=4.669379711151123, loss=0.9746636152267456 +I0831 17:13:52.984159 139921981437696 logging_writer.py:48] [102700] global_step=102700, grad_norm=4.991742134094238, loss=1.1489107608795166 +I0831 17:14:20.116366 139921973044992 logging_writer.py:48] [102800] global_step=102800, grad_norm=4.80378532409668, loss=1.0352352857589722 +I0831 17:14:47.312899 139921981437696 logging_writer.py:48] [102900] global_step=102900, grad_norm=4.9764556884765625, loss=1.0332578420639038 +I0831 17:15:14.437871 139921973044992 logging_writer.py:48] [103000] global_step=103000, grad_norm=5.156059265136719, loss=1.029948115348816 +I0831 17:15:41.544961 139921981437696 logging_writer.py:48] [103100] global_step=103100, grad_norm=4.815242290496826, loss=1.0263473987579346 +I0831 17:16:08.741253 139921973044992 logging_writer.py:48] [103200] global_step=103200, grad_norm=4.93530797958374, loss=1.0087791681289673 +I0831 17:16:35.869815 139921981437696 logging_writer.py:48] [103300] global_step=103300, grad_norm=4.677610397338867, loss=0.945088803768158 +I0831 17:17:03.222673 139921973044992 logging_writer.py:48] [103400] global_step=103400, grad_norm=5.112030982971191, loss=1.0652318000793457 +I0831 17:17:30.409344 139921981437696 logging_writer.py:48] [103500] global_step=103500, grad_norm=4.839086532592773, loss=1.0880540609359741 +I0831 17:17:57.508434 139921973044992 logging_writer.py:48] [103600] global_step=103600, grad_norm=4.899280548095703, loss=1.0457367897033691 +I0831 17:18:24.625426 139921981437696 logging_writer.py:48] [103700] global_step=103700, grad_norm=5.151354789733887, loss=1.0618921518325806 +I0831 17:18:51.822935 139921973044992 logging_writer.py:48] [103800] global_step=103800, grad_norm=4.66310453414917, loss=0.9681028127670288 +I0831 17:19:18.946664 139921981437696 logging_writer.py:48] [103900] global_step=103900, grad_norm=4.652068614959717, loss=1.000834345817566 +I0831 17:19:46.069144 139921973044992 logging_writer.py:48] [104000] global_step=104000, grad_norm=4.9891533851623535, loss=1.1246747970581055 +I0831 17:20:13.238491 139921981437696 logging_writer.py:48] [104100] global_step=104100, grad_norm=5.003024578094482, loss=1.1289877891540527 +I0831 17:20:40.361902 139921973044992 logging_writer.py:48] [104200] global_step=104200, grad_norm=4.8928680419921875, loss=0.948710560798645 +I0831 17:21:07.489720 139921981437696 logging_writer.py:48] [104300] global_step=104300, grad_norm=4.915801525115967, loss=1.0313105583190918 +I0831 17:21:34.904863 139921973044992 logging_writer.py:48] [104400] global_step=104400, grad_norm=4.899465560913086, loss=0.9102969765663147 +I0831 17:22:02.035527 139921981437696 logging_writer.py:48] [104500] global_step=104500, grad_norm=5.334686756134033, loss=1.1130014657974243 +I0831 17:22:29.170559 139921973044992 logging_writer.py:48] [104600] global_step=104600, grad_norm=5.161337375640869, loss=1.0932420492172241 +I0831 17:22:56.350844 139921981437696 logging_writer.py:48] [104700] global_step=104700, grad_norm=4.97885799407959, loss=1.0265713930130005 +I0831 17:23:23.487620 139921973044992 logging_writer.py:48] [104800] global_step=104800, grad_norm=5.021780014038086, loss=1.0154505968093872 +I0831 17:23:50.604681 139921981437696 logging_writer.py:48] [104900] global_step=104900, grad_norm=5.090000152587891, loss=1.1012554168701172 +I0831 17:24:17.771459 139921973044992 logging_writer.py:48] [105000] global_step=105000, grad_norm=5.07289457321167, loss=1.1202216148376465 +I0831 17:24:44.887888 139921981437696 logging_writer.py:48] [105100] global_step=105100, grad_norm=4.904410362243652, loss=1.0085796117782593 +I0831 17:25:12.013000 139921973044992 logging_writer.py:48] [105200] global_step=105200, grad_norm=4.782900810241699, loss=1.0124746561050415 +I0831 17:25:39.188367 139921981437696 logging_writer.py:48] [105300] global_step=105300, grad_norm=5.047495365142822, loss=1.0666139125823975 +I0831 17:26:06.376990 139921973044992 logging_writer.py:48] [105400] global_step=105400, grad_norm=4.982466697692871, loss=1.0837852954864502 +I0831 17:26:33.667606 139921981437696 logging_writer.py:48] [105500] global_step=105500, grad_norm=5.097845554351807, loss=1.0608372688293457 +I0831 17:27:00.846345 139921973044992 logging_writer.py:48] [105600] global_step=105600, grad_norm=4.901784896850586, loss=1.061020851135254 +I0831 17:27:27.937879 139921981437696 logging_writer.py:48] [105700] global_step=105700, grad_norm=5.139249324798584, loss=0.9967049956321716 +I0831 17:27:55.043657 139921973044992 logging_writer.py:48] [105800] global_step=105800, grad_norm=4.9657883644104, loss=0.9808757305145264 +I0831 17:28:22.249844 139921981437696 logging_writer.py:48] [105900] global_step=105900, grad_norm=4.81260347366333, loss=0.9211606979370117 +I0831 17:28:49.376947 139921973044992 logging_writer.py:48] [106000] global_step=106000, grad_norm=5.029510021209717, loss=1.128256916999817 +I0831 17:29:16.529731 139921981437696 logging_writer.py:48] [106100] global_step=106100, grad_norm=4.822310924530029, loss=0.9786746501922607 +I0831 17:29:43.701266 139921973044992 logging_writer.py:48] [106200] global_step=106200, grad_norm=4.60256814956665, loss=0.9673628807067871 +I0831 17:30:10.798460 139921981437696 logging_writer.py:48] [106300] global_step=106300, grad_norm=4.757440567016602, loss=0.9747515916824341 +I0831 17:30:37.956826 139921973044992 logging_writer.py:48] [106400] global_step=106400, grad_norm=4.741048336029053, loss=0.9932666420936584 +I0831 17:31:05.351476 139921981437696 logging_writer.py:48] [106500] global_step=106500, grad_norm=4.908257961273193, loss=0.9690331816673279 +I0831 17:31:32.473848 139921973044992 logging_writer.py:48] [106600] global_step=106600, grad_norm=4.982074737548828, loss=1.0494199991226196 +I0831 17:31:59.589552 139921981437696 logging_writer.py:48] [106700] global_step=106700, grad_norm=5.053983688354492, loss=0.9701627492904663 +I0831 17:32:26.763515 139921973044992 logging_writer.py:48] [106800] global_step=106800, grad_norm=4.995861530303955, loss=1.03098726272583 +I0831 17:32:53.885484 139921981437696 logging_writer.py:48] [106900] global_step=106900, grad_norm=4.792588710784912, loss=1.0513298511505127 +I0831 17:33:21.008351 139921973044992 logging_writer.py:48] [107000] global_step=107000, grad_norm=4.823419094085693, loss=1.0618562698364258 +I0831 17:33:48.189468 139921981437696 logging_writer.py:48] [107100] global_step=107100, grad_norm=5.0164923667907715, loss=1.0420153141021729 +I0831 17:34:15.319933 139921973044992 logging_writer.py:48] [107200] global_step=107200, grad_norm=4.682460308074951, loss=1.0484470129013062 +I0831 17:34:42.424176 139921981437696 logging_writer.py:48] [107300] global_step=107300, grad_norm=4.956842422485352, loss=0.9871098399162292 +I0831 17:35:09.616789 139921973044992 logging_writer.py:48] [107400] global_step=107400, grad_norm=4.9920334815979, loss=1.0839574337005615 +I0831 17:35:36.767270 139921981437696 logging_writer.py:48] [107500] global_step=107500, grad_norm=4.862011909484863, loss=1.0089914798736572 +I0831 17:36:04.167184 139921973044992 logging_writer.py:48] [107600] global_step=107600, grad_norm=4.7313714027404785, loss=0.9476348161697388 +I0831 17:36:31.360007 139921981437696 logging_writer.py:48] [107700] global_step=107700, grad_norm=5.070718288421631, loss=1.046877384185791 +I0831 17:36:58.468824 139921973044992 logging_writer.py:48] [107800] global_step=107800, grad_norm=4.986874103546143, loss=1.0479662418365479 +I0831 17:37:25.624440 139921981437696 logging_writer.py:48] [107900] global_step=107900, grad_norm=4.962133884429932, loss=1.0109649896621704 +I0831 17:37:52.791020 139921973044992 logging_writer.py:48] [108000] global_step=108000, grad_norm=5.002125263214111, loss=1.0440912246704102 +I0831 17:38:19.928196 139921981437696 logging_writer.py:48] [108100] global_step=108100, grad_norm=4.8443498611450195, loss=1.0180463790893555 +I0831 17:38:47.093254 139921973044992 logging_writer.py:48] [108200] global_step=108200, grad_norm=4.861069202423096, loss=1.0149760246276855 +I0831 17:39:14.288475 139921981437696 logging_writer.py:48] [108300] global_step=108300, grad_norm=4.550824165344238, loss=0.9304215908050537 +I0831 17:39:41.410884 139921973044992 logging_writer.py:48] [108400] global_step=108400, grad_norm=5.09627103805542, loss=1.0834604501724243 +I0831 17:40:08.536942 139921981437696 logging_writer.py:48] [108500] global_step=108500, grad_norm=4.771419525146484, loss=0.9449177980422974 +I0831 17:40:35.989629 139921973044992 logging_writer.py:48] [108600] global_step=108600, grad_norm=4.995847225189209, loss=1.0224610567092896 +I0831 17:41:03.101032 139921981437696 logging_writer.py:48] [108700] global_step=108700, grad_norm=4.897313594818115, loss=1.0274909734725952 +I0831 17:41:30.192483 139921973044992 logging_writer.py:48] [108800] global_step=108800, grad_norm=5.478229522705078, loss=1.0703835487365723 +I0831 17:41:57.361168 139921981437696 logging_writer.py:48] [108900] global_step=108900, grad_norm=4.7718186378479, loss=0.9743497371673584 +I0831 17:42:24.492063 139921973044992 logging_writer.py:48] [109000] global_step=109000, grad_norm=4.874917030334473, loss=1.0296489000320435 +I0831 17:42:51.605765 139921981437696 logging_writer.py:48] [109100] global_step=109100, grad_norm=4.872949123382568, loss=1.044224500656128 +I0831 17:43:18.762778 139921973044992 logging_writer.py:48] [109200] global_step=109200, grad_norm=4.678082466125488, loss=0.994694709777832 +I0831 17:43:45.873442 139921981437696 logging_writer.py:48] [109300] global_step=109300, grad_norm=5.002723217010498, loss=1.0262789726257324 +I0831 17:44:12.988558 139921973044992 logging_writer.py:48] [109400] global_step=109400, grad_norm=4.6901960372924805, loss=1.0299370288848877 +I0831 17:44:21.356922 140117123622080 spec.py:333] Evaluating on the training split. +I0831 17:44:28.064976 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 17:44:38.851624 140117123622080 spec.py:363] Evaluating on the test split. +I0831 17:44:39.746653 140117123622080 submission_runner.py:516] Time since start: 30401.08s, Step: 109432, {'train/accuracy': Array(0.92867106, dtype=float32), 'train/loss': Array(0.24968423, dtype=float32), 'validation/accuracy': Array(0.74715996, dtype=float32), 'validation/loss': Array(1.068822, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62600005, dtype=float32), 'test/loss': Array(1.8233396, dtype=float32), 'test/num_examples': 10000, 'score': 29992.7430062294, 'total_duration': 30401.084956884384, 'accumulated_submission_time': 29992.7430062294, 'accumulated_eval_time': 406.3058910369873, 'accumulated_logging_time': 1.1918253898620605} +I0831 17:44:39.793625 139921981437696 logging_writer.py:48] [109432] accumulated_eval_time=406.306, accumulated_logging_time=1.19183, accumulated_submission_time=29992.7, global_step=109432, preemption_count=0, score=29992.7, test/accuracy=0.6260000467300415, test/loss=1.823339581489563, test/num_examples=10000, total_duration=30401.1, train/accuracy=0.9286710619926453, train/loss=0.24968422949314117, validation/accuracy=0.7471599578857422, validation/loss=1.0688220262527466, validation/num_examples=50000 +I0831 17:44:58.673235 139921973044992 logging_writer.py:48] [109500] global_step=109500, grad_norm=4.8015217781066895, loss=0.967065155506134 +I0831 17:45:25.770812 139921981437696 logging_writer.py:48] [109600] global_step=109600, grad_norm=4.630969524383545, loss=0.9466532468795776 +I0831 17:45:53.079046 139921973044992 logging_writer.py:48] [109700] global_step=109700, grad_norm=4.709744930267334, loss=0.9260460138320923 +I0831 17:46:20.243385 139921981437696 logging_writer.py:48] [109800] global_step=109800, grad_norm=4.770446300506592, loss=0.9334539771080017 +I0831 17:46:47.337080 139921973044992 logging_writer.py:48] [109900] global_step=109900, grad_norm=4.748386383056641, loss=0.9647608995437622 +I0831 17:47:14.449060 139921981437696 logging_writer.py:48] [110000] global_step=110000, grad_norm=4.774829387664795, loss=1.0267947912216187 +I0831 17:47:41.645035 139921973044992 logging_writer.py:48] [110100] global_step=110100, grad_norm=4.913699150085449, loss=1.0453882217407227 +I0831 17:48:08.766111 139921981437696 logging_writer.py:48] [110200] global_step=110200, grad_norm=4.940305709838867, loss=1.0344092845916748 +I0831 17:48:35.861966 139921973044992 logging_writer.py:48] [110300] global_step=110300, grad_norm=5.014720439910889, loss=0.9563100934028625 +I0831 17:49:03.060504 139921981437696 logging_writer.py:48] [110400] global_step=110400, grad_norm=4.837789058685303, loss=0.9863467812538147 +I0831 17:49:30.170610 139921973044992 logging_writer.py:48] [110500] global_step=110500, grad_norm=4.854002952575684, loss=1.048712134361267 +I0831 17:49:57.268394 139921981437696 logging_writer.py:48] [110600] global_step=110600, grad_norm=5.102974891662598, loss=1.0185803174972534 +I0831 17:50:24.438687 139921973044992 logging_writer.py:48] [110700] global_step=110700, grad_norm=4.728212833404541, loss=1.012747049331665 +I0831 17:50:51.717043 139921981437696 logging_writer.py:48] [110800] global_step=110800, grad_norm=4.921494483947754, loss=1.1082319021224976 +I0831 17:51:18.844573 139921973044992 logging_writer.py:48] [110900] global_step=110900, grad_norm=4.841588020324707, loss=1.0290318727493286 +I0831 17:51:45.998288 139921981437696 logging_writer.py:48] [111000] global_step=111000, grad_norm=4.921259880065918, loss=0.9880991578102112 +I0831 17:52:13.090671 139921973044992 logging_writer.py:48] [111100] global_step=111100, grad_norm=5.631544589996338, loss=1.126893162727356 +I0831 17:52:40.203871 139921981437696 logging_writer.py:48] [111200] global_step=111200, grad_norm=4.799305438995361, loss=1.032954454421997 +I0831 17:53:07.387798 139921973044992 logging_writer.py:48] [111300] global_step=111300, grad_norm=4.798568248748779, loss=1.0452944040298462 +I0831 17:53:34.508045 139921981437696 logging_writer.py:48] [111400] global_step=111400, grad_norm=4.585724353790283, loss=0.920928955078125 +I0831 17:54:01.622190 139921973044992 logging_writer.py:48] [111500] global_step=111500, grad_norm=5.304590225219727, loss=1.12534499168396 +I0831 17:54:28.798507 139921981437696 logging_writer.py:48] [111600] global_step=111600, grad_norm=4.875804424285889, loss=1.0154452323913574 +I0831 17:54:55.965641 139921973044992 logging_writer.py:48] [111700] global_step=111700, grad_norm=4.530728340148926, loss=0.8983117938041687 +I0831 17:55:23.312721 139921981437696 logging_writer.py:48] [111800] global_step=111800, grad_norm=4.78719425201416, loss=1.0687205791473389 +I0831 17:55:50.474258 139921973044992 logging_writer.py:48] [111900] global_step=111900, grad_norm=5.047825813293457, loss=1.0879746675491333 +I0831 17:56:17.608685 139921981437696 logging_writer.py:48] [112000] global_step=112000, grad_norm=4.706247806549072, loss=1.0193214416503906 +I0831 17:56:44.720974 139921973044992 logging_writer.py:48] [112100] global_step=112100, grad_norm=4.9735283851623535, loss=0.9823682308197021 +I0831 17:57:11.935734 139921981437696 logging_writer.py:48] [112200] global_step=112200, grad_norm=4.790675640106201, loss=0.9636830687522888 +I0831 17:57:39.060117 139921973044992 logging_writer.py:48] [112300] global_step=112300, grad_norm=4.853516578674316, loss=0.9086178541183472 +I0831 17:58:06.174798 139921981437696 logging_writer.py:48] [112400] global_step=112400, grad_norm=4.911617279052734, loss=1.0415003299713135 +I0831 17:58:33.391870 139921973044992 logging_writer.py:48] [112500] global_step=112500, grad_norm=4.859004497528076, loss=1.0765695571899414 +I0831 17:59:00.515105 139921981437696 logging_writer.py:48] [112600] global_step=112600, grad_norm=4.956579208374023, loss=0.9893777966499329 +I0831 17:59:27.635306 139921973044992 logging_writer.py:48] [112700] global_step=112700, grad_norm=5.2413811683654785, loss=1.1191129684448242 +I0831 17:59:54.830418 139921981437696 logging_writer.py:48] [112800] global_step=112800, grad_norm=5.1168975830078125, loss=0.9290059804916382 +I0831 18:00:22.167179 139921973044992 logging_writer.py:48] [112900] global_step=112900, grad_norm=4.999540328979492, loss=1.013077735900879 +I0831 18:00:49.297250 139921981437696 logging_writer.py:48] [113000] global_step=113000, grad_norm=4.833617210388184, loss=1.0511133670806885 +I0831 18:01:16.487300 139921973044992 logging_writer.py:48] [113100] global_step=113100, grad_norm=4.836831092834473, loss=1.0662904977798462 +I0831 18:01:43.607482 139921981437696 logging_writer.py:48] [113200] global_step=113200, grad_norm=4.921950340270996, loss=1.0437818765640259 +I0831 18:02:10.731037 139921973044992 logging_writer.py:48] [113300] global_step=113300, grad_norm=5.1489129066467285, loss=1.0652389526367188 +I0831 18:02:37.900922 139921981437696 logging_writer.py:48] [113400] global_step=113400, grad_norm=4.662088394165039, loss=0.9299733638763428 +I0831 18:03:05.009783 139921973044992 logging_writer.py:48] [113500] global_step=113500, grad_norm=4.667788982391357, loss=0.949915885925293 +I0831 18:03:32.130758 139921981437696 logging_writer.py:48] [113600] global_step=113600, grad_norm=5.010042190551758, loss=1.1198800802230835 +I0831 18:03:59.302001 139921973044992 logging_writer.py:48] [113700] global_step=113700, grad_norm=5.121248722076416, loss=0.9405273199081421 +I0831 18:04:26.448964 139921981437696 logging_writer.py:48] [113800] global_step=113800, grad_norm=4.559306621551514, loss=0.9412976503372192 +I0831 18:04:53.561518 139921973044992 logging_writer.py:48] [113900] global_step=113900, grad_norm=4.790529251098633, loss=0.9637941718101501 +I0831 18:05:20.936691 139921981437696 logging_writer.py:48] [114000] global_step=114000, grad_norm=4.840064525604248, loss=1.0616673231124878 +I0831 18:05:48.053134 139921973044992 logging_writer.py:48] [114100] global_step=114100, grad_norm=4.91908597946167, loss=0.9630898833274841 +I0831 18:06:15.172183 139921981437696 logging_writer.py:48] [114200] global_step=114200, grad_norm=4.756585597991943, loss=0.9676967263221741 +I0831 18:06:42.359197 139921973044992 logging_writer.py:48] [114300] global_step=114300, grad_norm=4.686849594116211, loss=0.963951587677002 +I0831 18:07:09.487534 139921981437696 logging_writer.py:48] [114400] global_step=114400, grad_norm=4.795084476470947, loss=1.0859941244125366 +I0831 18:07:36.606131 139921973044992 logging_writer.py:48] [114500] global_step=114500, grad_norm=4.795589923858643, loss=0.9288429021835327 +I0831 18:08:03.787977 139921981437696 logging_writer.py:48] [114600] global_step=114600, grad_norm=4.939456462860107, loss=1.1138916015625 +I0831 18:08:30.929355 139921973044992 logging_writer.py:48] [114700] global_step=114700, grad_norm=4.849266052246094, loss=1.0432488918304443 +I0831 18:08:58.065120 139921981437696 logging_writer.py:48] [114800] global_step=114800, grad_norm=4.512290000915527, loss=1.0457942485809326 +I0831 18:09:25.275514 139921973044992 logging_writer.py:48] [114900] global_step=114900, grad_norm=4.912945747375488, loss=1.0573993921279907 +I0831 18:09:52.423270 139921981437696 logging_writer.py:48] [115000] global_step=115000, grad_norm=4.935878753662109, loss=1.0909123420715332 +I0831 18:10:19.729762 139921973044992 logging_writer.py:48] [115100] global_step=115100, grad_norm=4.840478897094727, loss=1.0400784015655518 +I0831 18:10:46.895478 139921981437696 logging_writer.py:48] [115200] global_step=115200, grad_norm=4.797809600830078, loss=0.9615044593811035 +I0831 18:11:14.014785 139921973044992 logging_writer.py:48] [115300] global_step=115300, grad_norm=5.111965179443359, loss=1.0326381921768188 +I0831 18:11:41.126093 139921981437696 logging_writer.py:48] [115400] global_step=115400, grad_norm=4.977583885192871, loss=1.0777723789215088 +I0831 18:12:08.283304 139921973044992 logging_writer.py:48] [115500] global_step=115500, grad_norm=5.076111793518066, loss=1.0934717655181885 +I0831 18:12:35.378659 139921981437696 logging_writer.py:48] [115600] global_step=115600, grad_norm=4.905673027038574, loss=1.0136691331863403 +I0831 18:13:02.511641 139921973044992 logging_writer.py:48] [115700] global_step=115700, grad_norm=5.198518753051758, loss=1.1519660949707031 +I0831 18:13:29.680949 139921981437696 logging_writer.py:48] [115800] global_step=115800, grad_norm=4.736073017120361, loss=0.9133780002593994 +I0831 18:13:56.808542 139921973044992 logging_writer.py:48] [115900] global_step=115900, grad_norm=4.952330589294434, loss=1.0094622373580933 +I0831 18:14:23.926889 139921981437696 logging_writer.py:48] [116000] global_step=116000, grad_norm=4.890621662139893, loss=1.0090384483337402 +I0831 18:14:51.308259 139921973044992 logging_writer.py:48] [116100] global_step=116100, grad_norm=4.891754627227783, loss=1.0378341674804688 +I0831 18:15:18.443396 139921981437696 logging_writer.py:48] [116200] global_step=116200, grad_norm=4.793152809143066, loss=0.9577704668045044 +I0831 18:15:45.584517 139921973044992 logging_writer.py:48] [116300] global_step=116300, grad_norm=4.950894355773926, loss=1.021409034729004 +I0831 18:16:12.799908 139921981437696 logging_writer.py:48] [116400] global_step=116400, grad_norm=4.822246551513672, loss=1.0563764572143555 +I0831 18:16:39.939307 139921973044992 logging_writer.py:48] [116500] global_step=116500, grad_norm=4.878383636474609, loss=1.076978325843811 +I0831 18:17:07.075610 139921981437696 logging_writer.py:48] [116600] global_step=116600, grad_norm=5.049248695373535, loss=1.0727092027664185 +I0831 18:17:34.266048 139921973044992 logging_writer.py:48] [116700] global_step=116700, grad_norm=4.864144325256348, loss=0.9424585103988647 +I0831 18:17:55.769421 140117123622080 spec.py:333] Evaluating on the training split. +I0831 18:18:02.442923 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 18:18:12.374881 140117123622080 spec.py:363] Evaluating on the test split. +I0831 18:18:13.258183 140117123622080 submission_runner.py:516] Time since start: 32414.60s, Step: 116781, {'train/accuracy': Array(0.9322385, dtype=float32), 'train/loss': Array(0.23868681, dtype=float32), 'validation/accuracy': Array(0.75016, dtype=float32), 'validation/loss': Array(1.0682027, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62670004, dtype=float32), 'test/loss': Array(1.8252075, dtype=float32), 'test/num_examples': 10000, 'score': 31988.6567633152, 'total_duration': 32414.598189353943, 'accumulated_submission_time': 31988.6567633152, 'accumulated_eval_time': 423.79282784461975, 'accumulated_logging_time': 1.2480707168579102} +I0831 18:18:13.317240 139921981437696 logging_writer.py:48] [116781] accumulated_eval_time=423.793, accumulated_logging_time=1.24807, accumulated_submission_time=31988.7, global_step=116781, preemption_count=0, score=31988.7, test/accuracy=0.6267000436782837, test/loss=1.8252074718475342, test/num_examples=10000, total_duration=32414.6, train/accuracy=0.9322385191917419, train/loss=0.23868681490421295, validation/accuracy=0.7501599788665771, validation/loss=1.0682027339935303, validation/num_examples=50000 +I0831 18:18:19.041202 139921973044992 logging_writer.py:48] [116800] global_step=116800, grad_norm=4.9054131507873535, loss=1.0330222845077515 +I0831 18:18:46.159919 139921981437696 logging_writer.py:48] [116900] global_step=116900, grad_norm=5.051254749298096, loss=0.9753528237342834 +I0831 18:19:13.302579 139921973044992 logging_writer.py:48] [117000] global_step=117000, grad_norm=4.620398998260498, loss=0.9316913485527039 +I0831 18:19:40.429000 139921981437696 logging_writer.py:48] [117100] global_step=117100, grad_norm=4.73403263092041, loss=0.9747112989425659 +I0831 18:20:07.787064 139921973044992 logging_writer.py:48] [117200] global_step=117200, grad_norm=4.9731574058532715, loss=1.0987181663513184 +I0831 18:20:34.947692 139921981437696 logging_writer.py:48] [117300] global_step=117300, grad_norm=4.537646770477295, loss=0.9919869899749756 +I0831 18:21:02.041311 139921973044992 logging_writer.py:48] [117400] global_step=117400, grad_norm=4.7273664474487305, loss=1.0061153173446655 +I0831 18:21:29.132618 139921981437696 logging_writer.py:48] [117500] global_step=117500, grad_norm=5.108097076416016, loss=1.1136163473129272 +I0831 18:21:56.284788 139921973044992 logging_writer.py:48] [117600] global_step=117600, grad_norm=4.545338153839111, loss=0.8913135528564453 +I0831 18:22:23.418791 139921981437696 logging_writer.py:48] [117700] global_step=117700, grad_norm=5.002782344818115, loss=1.0397050380706787 +I0831 18:22:50.553974 139921973044992 logging_writer.py:48] [117800] global_step=117800, grad_norm=5.015805721282959, loss=1.0216480493545532 +I0831 18:23:17.726846 139921981437696 logging_writer.py:48] [117900] global_step=117900, grad_norm=4.866134166717529, loss=0.9511408805847168 +I0831 18:23:44.824032 139921973044992 logging_writer.py:48] [118000] global_step=118000, grad_norm=5.296024799346924, loss=1.0942792892456055 +I0831 18:24:11.987036 139921981437696 logging_writer.py:48] [118100] global_step=118100, grad_norm=4.74285364151001, loss=1.0383933782577515 +I0831 18:24:39.418505 139921973044992 logging_writer.py:48] [118200] global_step=118200, grad_norm=4.80192232131958, loss=1.0082181692123413 +I0831 18:25:06.535968 139921981437696 logging_writer.py:48] [118300] global_step=118300, grad_norm=5.248069763183594, loss=1.0115995407104492 +I0831 18:25:33.649614 139921973044992 logging_writer.py:48] [118400] global_step=118400, grad_norm=5.025343894958496, loss=1.0583720207214355 +I0831 18:26:00.853097 139921981437696 logging_writer.py:48] [118500] global_step=118500, grad_norm=4.903810977935791, loss=1.0537958145141602 +I0831 18:26:27.987102 139921973044992 logging_writer.py:48] [118600] global_step=118600, grad_norm=4.403096675872803, loss=0.8989332318305969 +I0831 18:26:55.108985 139921981437696 logging_writer.py:48] [118700] global_step=118700, grad_norm=4.4300384521484375, loss=0.9214639663696289 +I0831 18:27:22.310014 139921973044992 logging_writer.py:48] [118800] global_step=118800, grad_norm=5.030410289764404, loss=0.9861904978752136 +I0831 18:27:49.407207 139921981437696 logging_writer.py:48] [118900] global_step=118900, grad_norm=4.825359344482422, loss=0.9991657733917236 +I0831 18:28:16.543148 139921973044992 logging_writer.py:48] [119000] global_step=119000, grad_norm=4.684498310089111, loss=1.060196042060852 +I0831 18:28:43.735997 139921981437696 logging_writer.py:48] [119100] global_step=119100, grad_norm=4.875424385070801, loss=0.9987229704856873 +I0831 18:29:10.835009 139921973044992 logging_writer.py:48] [119200] global_step=119200, grad_norm=4.712450981140137, loss=0.9542218446731567 +I0831 18:29:38.172137 139921981437696 logging_writer.py:48] [119300] global_step=119300, grad_norm=4.826153755187988, loss=1.0757598876953125 +I0831 18:30:05.349739 139921973044992 logging_writer.py:48] [119400] global_step=119400, grad_norm=5.214629650115967, loss=1.09479820728302 +I0831 18:30:32.492654 139921981437696 logging_writer.py:48] [119500] global_step=119500, grad_norm=4.917186260223389, loss=0.9912127256393433 +I0831 18:30:59.632744 139921973044992 logging_writer.py:48] [119600] global_step=119600, grad_norm=4.845108509063721, loss=1.0090383291244507 +I0831 18:31:26.818806 139921981437696 logging_writer.py:48] [119700] global_step=119700, grad_norm=4.79254674911499, loss=0.9623435735702515 +I0831 18:31:53.944190 139921973044992 logging_writer.py:48] [119800] global_step=119800, grad_norm=4.827485084533691, loss=1.0105478763580322 +I0831 18:32:21.072324 139921981437696 logging_writer.py:48] [119900] global_step=119900, grad_norm=5.138244152069092, loss=1.061623454093933 +I0831 18:32:48.263283 139921973044992 logging_writer.py:48] [120000] global_step=120000, grad_norm=4.867122173309326, loss=0.9511703252792358 +I0831 18:33:15.407915 139921981437696 logging_writer.py:48] [120100] global_step=120100, grad_norm=4.849661350250244, loss=1.018594741821289 +I0831 18:33:42.532533 139921973044992 logging_writer.py:48] [120200] global_step=120200, grad_norm=4.750497341156006, loss=0.8650225400924683 +I0831 18:34:09.930577 139921981437696 logging_writer.py:48] [120300] global_step=120300, grad_norm=4.86531925201416, loss=1.0140812397003174 +I0831 18:34:37.038411 139921973044992 logging_writer.py:48] [120400] global_step=120400, grad_norm=4.801098823547363, loss=0.9922773838043213 +I0831 18:35:04.151596 139921981437696 logging_writer.py:48] [120500] global_step=120500, grad_norm=4.6791558265686035, loss=1.0003509521484375 +I0831 18:35:31.313892 139921973044992 logging_writer.py:48] [120600] global_step=120600, grad_norm=4.931272983551025, loss=0.9591997861862183 +I0831 18:35:58.404222 139921981437696 logging_writer.py:48] [120700] global_step=120700, grad_norm=4.88363790512085, loss=1.0076533555984497 +I0831 18:36:25.517656 139921973044992 logging_writer.py:48] [120800] global_step=120800, grad_norm=5.018738746643066, loss=1.0324792861938477 +I0831 18:36:52.701022 139921981437696 logging_writer.py:48] [120900] global_step=120900, grad_norm=4.772478103637695, loss=0.9745576977729797 +I0831 18:37:19.848177 139921973044992 logging_writer.py:48] [121000] global_step=121000, grad_norm=4.859785079956055, loss=1.0034048557281494 +I0831 18:37:46.951661 139921981437696 logging_writer.py:48] [121100] global_step=121100, grad_norm=4.847606182098389, loss=0.9201053380966187 +I0831 18:38:14.178013 139921973044992 logging_writer.py:48] [121200] global_step=121200, grad_norm=4.777750015258789, loss=1.0270113945007324 +I0831 18:38:41.274871 139921981437696 logging_writer.py:48] [121300] global_step=121300, grad_norm=4.8823041915893555, loss=0.9671269655227661 +I0831 18:39:08.602985 139921973044992 logging_writer.py:48] [121400] global_step=121400, grad_norm=5.024406909942627, loss=1.0458624362945557 +I0831 18:39:35.778955 139921981437696 logging_writer.py:48] [121500] global_step=121500, grad_norm=4.80247688293457, loss=1.0041316747665405 +I0831 18:40:02.876682 139921973044992 logging_writer.py:48] [121600] global_step=121600, grad_norm=4.980691432952881, loss=1.0961484909057617 +I0831 18:40:30.010871 139921981437696 logging_writer.py:48] [121700] global_step=121700, grad_norm=4.458730697631836, loss=0.9423577785491943 +I0831 18:40:57.216058 139921973044992 logging_writer.py:48] [121800] global_step=121800, grad_norm=4.707841396331787, loss=0.962086021900177 +I0831 18:41:24.338363 139921981437696 logging_writer.py:48] [121900] global_step=121900, grad_norm=4.558718681335449, loss=0.9423974752426147 +I0831 18:41:51.459211 139921973044992 logging_writer.py:48] [122000] global_step=122000, grad_norm=5.03710412979126, loss=1.023555040359497 +I0831 18:42:18.626781 139921981437696 logging_writer.py:48] [122100] global_step=122100, grad_norm=4.719923496246338, loss=1.0190829038619995 +I0831 18:42:45.736013 139921973044992 logging_writer.py:48] [122200] global_step=122200, grad_norm=5.1880717277526855, loss=1.0732078552246094 +I0831 18:43:12.857544 139921981437696 logging_writer.py:48] [122300] global_step=122300, grad_norm=4.876638889312744, loss=0.989694356918335 +I0831 18:43:40.235600 139921973044992 logging_writer.py:48] [122400] global_step=122400, grad_norm=4.560052394866943, loss=0.9179947972297668 +I0831 18:44:07.381761 139921981437696 logging_writer.py:48] [122500] global_step=122500, grad_norm=4.935833930969238, loss=1.046614170074463 +I0831 18:44:34.506815 139921973044992 logging_writer.py:48] [122600] global_step=122600, grad_norm=4.890375137329102, loss=0.9885494709014893 +I0831 18:45:01.690138 139921981437696 logging_writer.py:48] [122700] global_step=122700, grad_norm=5.052093029022217, loss=1.0177650451660156 +I0831 18:45:28.856976 139921973044992 logging_writer.py:48] [122800] global_step=122800, grad_norm=5.1417107582092285, loss=1.0142896175384521 +I0831 18:45:55.972749 139921981437696 logging_writer.py:48] [122900] global_step=122900, grad_norm=4.769492149353027, loss=1.0279736518859863 +I0831 18:46:23.167765 139921973044992 logging_writer.py:48] [123000] global_step=123000, grad_norm=4.720660209655762, loss=1.0053718090057373 +I0831 18:46:50.302041 139921981437696 logging_writer.py:48] [123100] global_step=123100, grad_norm=4.864187717437744, loss=1.0110008716583252 +I0831 18:47:17.426375 139921973044992 logging_writer.py:48] [123200] global_step=123200, grad_norm=4.775469779968262, loss=0.9957265257835388 +I0831 18:47:44.614063 139921981437696 logging_writer.py:48] [123300] global_step=123300, grad_norm=4.908729553222656, loss=0.9391661882400513 +I0831 18:48:11.764560 139921973044992 logging_writer.py:48] [123400] global_step=123400, grad_norm=4.527639389038086, loss=0.9336605668067932 +I0831 18:48:39.127832 139921981437696 logging_writer.py:48] [123500] global_step=123500, grad_norm=4.878313064575195, loss=1.0328840017318726 +I0831 18:49:06.284293 139921973044992 logging_writer.py:48] [123600] global_step=123600, grad_norm=4.7726359367370605, loss=0.9394985437393188 +I0831 18:49:33.409317 139921981437696 logging_writer.py:48] [123700] global_step=123700, grad_norm=4.968863010406494, loss=1.005143165588379 +I0831 18:50:00.535435 139921973044992 logging_writer.py:48] [123800] global_step=123800, grad_norm=4.9497270584106445, loss=1.0437365770339966 +I0831 18:50:27.691360 139921981437696 logging_writer.py:48] [123900] global_step=123900, grad_norm=4.778032302856445, loss=0.9754948616027832 +I0831 18:50:54.823463 139921973044992 logging_writer.py:48] [124000] global_step=124000, grad_norm=4.721354961395264, loss=0.9422352313995361 +I0831 18:51:21.915239 139921981437696 logging_writer.py:48] [124100] global_step=124100, grad_norm=4.710768222808838, loss=1.0220838785171509 +I0831 18:51:29.447714 140117123622080 spec.py:333] Evaluating on the training split. +I0831 18:51:36.015763 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 18:51:45.901681 140117123622080 spec.py:363] Evaluating on the test split. +I0831 18:51:46.778075 140117123622080 submission_runner.py:516] Time since start: 34428.12s, Step: 124129, {'train/accuracy': Array(0.934949, dtype=float32), 'train/loss': Array(0.22692154, dtype=float32), 'validation/accuracy': Array(0.75034, dtype=float32), 'validation/loss': Array(1.0701694, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62640005, dtype=float32), 'test/loss': Array(1.8303214, dtype=float32), 'test/num_examples': 10000, 'score': 33984.70328807831, 'total_duration': 34428.11667752266, 'accumulated_submission_time': 33984.70328807831, 'accumulated_eval_time': 441.1199595928192, 'accumulated_logging_time': 1.3386790752410889} +I0831 18:51:46.826862 139921973044992 logging_writer.py:48] [124129] accumulated_eval_time=441.12, accumulated_logging_time=1.33868, accumulated_submission_time=33984.7, global_step=124129, preemption_count=0, score=33984.7, test/accuracy=0.6264000535011292, test/loss=1.8303214311599731, test/num_examples=10000, total_duration=34428.1, train/accuracy=0.9349489808082581, train/loss=0.22692154347896576, validation/accuracy=0.7503399848937988, validation/loss=1.070169448852539, validation/num_examples=50000 +I0831 18:52:06.444149 139921981437696 logging_writer.py:48] [124200] global_step=124200, grad_norm=4.573788642883301, loss=0.920998215675354 +I0831 18:52:33.563393 139921973044992 logging_writer.py:48] [124300] global_step=124300, grad_norm=4.830995082855225, loss=0.9547917246818542 +I0831 18:53:00.678858 139921981437696 logging_writer.py:48] [124400] global_step=124400, grad_norm=4.95279598236084, loss=1.006866216659546 +I0831 18:53:27.825994 139921973044992 logging_writer.py:48] [124500] global_step=124500, grad_norm=5.081884860992432, loss=1.0316588878631592 +I0831 18:53:55.167323 139921981437696 logging_writer.py:48] [124600] global_step=124600, grad_norm=4.902658462524414, loss=0.9918402433395386 +I0831 18:54:22.310033 139921973044992 logging_writer.py:48] [124700] global_step=124700, grad_norm=4.9617815017700195, loss=1.0519022941589355 +I0831 18:54:49.476877 139921981437696 logging_writer.py:48] [124800] global_step=124800, grad_norm=5.21953821182251, loss=1.0418155193328857 +I0831 18:55:16.615042 139921973044992 logging_writer.py:48] [124900] global_step=124900, grad_norm=4.767873287200928, loss=1.0012626647949219 +I0831 18:55:43.716599 139921981437696 logging_writer.py:48] [125000] global_step=125000, grad_norm=4.882230758666992, loss=1.0082824230194092 +I0831 18:56:10.894781 139921973044992 logging_writer.py:48] [125100] global_step=125100, grad_norm=4.934769153594971, loss=1.0723061561584473 +I0831 18:56:38.040286 139921981437696 logging_writer.py:48] [125200] global_step=125200, grad_norm=4.677310943603516, loss=0.9543856382369995 +I0831 18:57:05.155179 139921973044992 logging_writer.py:48] [125300] global_step=125300, grad_norm=4.915539264678955, loss=1.093650221824646 +I0831 18:57:32.322885 139921981437696 logging_writer.py:48] [125400] global_step=125400, grad_norm=4.973422050476074, loss=1.0231976509094238 +I0831 18:57:59.478588 139921973044992 logging_writer.py:48] [125500] global_step=125500, grad_norm=5.413672924041748, loss=1.0127782821655273 +I0831 18:58:26.849849 139921981437696 logging_writer.py:48] [125600] global_step=125600, grad_norm=4.966064453125, loss=1.0411808490753174 +I0831 18:58:54.030484 139921973044992 logging_writer.py:48] [125700] global_step=125700, grad_norm=5.225245952606201, loss=0.973819375038147 +I0831 18:59:21.154867 139921981437696 logging_writer.py:48] [125800] global_step=125800, grad_norm=4.876797199249268, loss=1.0271416902542114 +I0831 18:59:48.248754 139921973044992 logging_writer.py:48] [125900] global_step=125900, grad_norm=5.042271614074707, loss=1.0656102895736694 +I0831 19:00:15.445909 139921981437696 logging_writer.py:48] [126000] global_step=126000, grad_norm=4.799026012420654, loss=0.991119384765625 +I0831 19:00:42.560408 139921973044992 logging_writer.py:48] [126100] global_step=126100, grad_norm=5.110009670257568, loss=1.0665645599365234 +I0831 19:01:09.677767 139921981437696 logging_writer.py:48] [126200] global_step=126200, grad_norm=4.877458095550537, loss=0.9479408264160156 +I0831 19:01:36.863784 139921973044992 logging_writer.py:48] [126300] global_step=126300, grad_norm=4.945985317230225, loss=1.070550560951233 +I0831 19:02:03.983479 139921981437696 logging_writer.py:48] [126400] global_step=126400, grad_norm=5.136666297912598, loss=1.0094184875488281 +I0831 19:02:31.098831 139921973044992 logging_writer.py:48] [126500] global_step=126500, grad_norm=5.316727161407471, loss=1.1089749336242676 +I0831 19:02:58.278690 139921981437696 logging_writer.py:48] [126600] global_step=126600, grad_norm=5.015868186950684, loss=1.050066590309143 +I0831 19:03:25.607229 139921973044992 logging_writer.py:48] [126700] global_step=126700, grad_norm=5.037516117095947, loss=1.0428359508514404 +I0831 19:03:52.727792 139921981437696 logging_writer.py:48] [126800] global_step=126800, grad_norm=4.710676193237305, loss=0.9986215233802795 +I0831 19:04:19.907614 139921973044992 logging_writer.py:48] [126900] global_step=126900, grad_norm=4.846414566040039, loss=0.9555006623268127 +I0831 19:04:47.033407 139921981437696 logging_writer.py:48] [127000] global_step=127000, grad_norm=5.098504543304443, loss=1.0855990648269653 +I0831 19:05:14.155210 139921973044992 logging_writer.py:48] [127100] global_step=127100, grad_norm=4.994109153747559, loss=0.9869741201400757 +I0831 19:05:41.328735 139921981437696 logging_writer.py:48] [127200] global_step=127200, grad_norm=4.615044116973877, loss=0.9512423276901245 +I0831 19:06:08.423931 139921973044992 logging_writer.py:48] [127300] global_step=127300, grad_norm=4.9811577796936035, loss=0.9851111769676208 +I0831 19:06:35.544130 139921981437696 logging_writer.py:48] [127400] global_step=127400, grad_norm=4.78759765625, loss=0.9928128123283386 +I0831 19:07:02.727712 139921973044992 logging_writer.py:48] [127500] global_step=127500, grad_norm=5.289881706237793, loss=1.0777745246887207 +I0831 19:07:29.838753 139921981437696 logging_writer.py:48] [127600] global_step=127600, grad_norm=4.982982158660889, loss=1.08706533908844 +I0831 19:07:56.975401 139921973044992 logging_writer.py:48] [127700] global_step=127700, grad_norm=4.959151744842529, loss=1.0438778400421143 +I0831 19:08:24.384214 139921981437696 logging_writer.py:48] [127800] global_step=127800, grad_norm=4.983370304107666, loss=1.037099838256836 +I0831 19:08:51.485399 139921973044992 logging_writer.py:48] [127900] global_step=127900, grad_norm=4.648209571838379, loss=0.9618728160858154 +I0831 19:09:18.613826 139921981437696 logging_writer.py:48] [128000] global_step=128000, grad_norm=4.9036664962768555, loss=1.0325816869735718 +I0831 19:09:45.807535 139921973044992 logging_writer.py:48] [128100] global_step=128100, grad_norm=5.317135334014893, loss=1.0063817501068115 +I0831 19:10:12.965635 139921981437696 logging_writer.py:48] [128200] global_step=128200, grad_norm=4.850796699523926, loss=0.9372588992118835 +I0831 19:10:40.083278 139921973044992 logging_writer.py:48] [128300] global_step=128300, grad_norm=4.68488073348999, loss=0.9390082955360413 +I0831 19:11:07.253661 139921981437696 logging_writer.py:48] [128400] global_step=128400, grad_norm=4.774626731872559, loss=0.9705464839935303 +I0831 19:11:34.364708 139921973044992 logging_writer.py:48] [128500] global_step=128500, grad_norm=4.4585113525390625, loss=0.9783653616905212 +I0831 19:12:01.478414 139921981437696 logging_writer.py:48] [128600] global_step=128600, grad_norm=4.74650239944458, loss=0.8961331844329834 +I0831 19:12:28.664198 139921973044992 logging_writer.py:48] [128700] global_step=128700, grad_norm=4.754620552062988, loss=0.9165308475494385 +I0831 19:12:55.764052 139921981437696 logging_writer.py:48] [128800] global_step=128800, grad_norm=5.231456279754639, loss=1.0824346542358398 +I0831 19:13:23.103730 139921973044992 logging_writer.py:48] [128900] global_step=128900, grad_norm=5.0074849128723145, loss=1.012521505355835 +I0831 19:13:50.268504 139921981437696 logging_writer.py:48] [129000] global_step=129000, grad_norm=5.1197333335876465, loss=0.9348454475402832 +I0831 19:14:17.378738 139921973044992 logging_writer.py:48] [129100] global_step=129100, grad_norm=4.990143775939941, loss=1.0237383842468262 +I0831 19:14:44.517954 139921981437696 logging_writer.py:48] [129200] global_step=129200, grad_norm=4.713932514190674, loss=0.9269334673881531 +I0831 19:15:11.711831 139921973044992 logging_writer.py:48] [129300] global_step=129300, grad_norm=4.699375629425049, loss=0.8953531384468079 +I0831 19:15:38.848545 139921981437696 logging_writer.py:48] [129400] global_step=129400, grad_norm=4.655710697174072, loss=0.9238196611404419 +I0831 19:16:05.955716 139921973044992 logging_writer.py:48] [129500] global_step=129500, grad_norm=5.032966136932373, loss=1.0032687187194824 +I0831 19:16:33.145720 139921981437696 logging_writer.py:48] [129600] global_step=129600, grad_norm=5.058819770812988, loss=1.0229711532592773 +I0831 19:17:00.257283 139921973044992 logging_writer.py:48] [129700] global_step=129700, grad_norm=5.01087760925293, loss=1.009272575378418 +I0831 19:17:27.399976 139921981437696 logging_writer.py:48] [129800] global_step=129800, grad_norm=4.770064830780029, loss=0.9408623576164246 +I0831 19:17:54.825585 139921973044992 logging_writer.py:48] [129900] global_step=129900, grad_norm=4.825511455535889, loss=0.9374531507492065 +I0831 19:18:21.925579 139921981437696 logging_writer.py:48] [130000] global_step=130000, grad_norm=4.924545764923096, loss=1.030064582824707 +I0831 19:18:49.084409 139921973044992 logging_writer.py:48] [130100] global_step=130100, grad_norm=4.893178939819336, loss=0.9906936883926392 +I0831 19:19:16.250165 139921981437696 logging_writer.py:48] [130200] global_step=130200, grad_norm=4.788145065307617, loss=1.036426305770874 +I0831 19:19:43.355356 139921973044992 logging_writer.py:48] [130300] global_step=130300, grad_norm=5.164313793182373, loss=1.0315181016921997 +I0831 19:20:10.496792 139921981437696 logging_writer.py:48] [130400] global_step=130400, grad_norm=4.8989152908325195, loss=0.9652845859527588 +I0831 19:20:37.688223 139921973044992 logging_writer.py:48] [130500] global_step=130500, grad_norm=4.685306072235107, loss=0.9543024301528931 +I0831 19:21:04.829024 139921981437696 logging_writer.py:48] [130600] global_step=130600, grad_norm=4.865204811096191, loss=0.9873088598251343 +I0831 19:21:31.959319 139921973044992 logging_writer.py:48] [130700] global_step=130700, grad_norm=4.917021751403809, loss=1.008719801902771 +I0831 19:21:59.137543 139921981437696 logging_writer.py:48] [130800] global_step=130800, grad_norm=4.769372463226318, loss=1.009446144104004 +I0831 19:22:26.260018 139921973044992 logging_writer.py:48] [130900] global_step=130900, grad_norm=4.57985782623291, loss=0.8652845621109009 +I0831 19:22:53.599110 139921981437696 logging_writer.py:48] [131000] global_step=131000, grad_norm=4.736083030700684, loss=0.9471940994262695 +I0831 19:23:20.796348 139921973044992 logging_writer.py:48] [131100] global_step=131100, grad_norm=5.132066249847412, loss=0.9884502291679382 +I0831 19:23:47.924466 139921981437696 logging_writer.py:48] [131200] global_step=131200, grad_norm=4.5111470222473145, loss=0.9261791706085205 +I0831 19:24:15.067715 139921973044992 logging_writer.py:48] [131300] global_step=131300, grad_norm=5.067234039306641, loss=1.056391954421997 +I0831 19:24:42.229998 139921981437696 logging_writer.py:48] [131400] global_step=131400, grad_norm=5.07786750793457, loss=1.0464445352554321 +I0831 19:25:02.939299 140117123622080 spec.py:333] Evaluating on the training split. +I0831 19:25:09.578731 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 19:25:20.235132 140117123622080 spec.py:363] Evaluating on the test split. +I0831 19:25:21.124323 140117123622080 submission_runner.py:516] Time since start: 36442.46s, Step: 131478, {'train/accuracy': Array(0.9367626, dtype=float32), 'train/loss': Array(0.21946916, dtype=float32), 'validation/accuracy': Array(0.75009996, dtype=float32), 'validation/loss': Array(1.0705539, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6244, dtype=float32), 'test/loss': Array(1.8374841, dtype=float32), 'test/num_examples': 10000, 'score': 35980.75158596039, 'total_duration': 36442.463853120804, 'accumulated_submission_time': 35980.75158596039, 'accumulated_eval_time': 459.30268383026123, 'accumulated_logging_time': 1.3988158702850342} +I0831 19:25:21.171768 139921973044992 logging_writer.py:48] [131478] accumulated_eval_time=459.303, accumulated_logging_time=1.39882, accumulated_submission_time=35980.8, global_step=131478, preemption_count=0, score=35980.8, test/accuracy=0.6244000196456909, test/loss=1.8374841213226318, test/num_examples=10000, total_duration=36442.5, train/accuracy=0.9367625713348389, train/loss=0.21946915984153748, validation/accuracy=0.7500999569892883, validation/loss=1.0705538988113403, validation/num_examples=50000 +I0831 19:25:27.548291 139921981437696 logging_writer.py:48] [131500] global_step=131500, grad_norm=4.693057060241699, loss=0.945777416229248 +I0831 19:25:54.625195 139921973044992 logging_writer.py:48] [131600] global_step=131600, grad_norm=4.8968119621276855, loss=1.0125634670257568 +I0831 19:26:21.775130 139921981437696 logging_writer.py:48] [131700] global_step=131700, grad_norm=4.809353351593018, loss=1.0111111402511597 +I0831 19:26:48.898675 139921973044992 logging_writer.py:48] [131800] global_step=131800, grad_norm=4.871486663818359, loss=0.9539484977722168 +I0831 19:27:16.036075 139921981437696 logging_writer.py:48] [131900] global_step=131900, grad_norm=4.682309627532959, loss=0.9816653728485107 +I0831 19:27:43.439092 139921973044992 logging_writer.py:48] [132000] global_step=132000, grad_norm=4.694303035736084, loss=1.0098254680633545 +I0831 19:28:10.551725 139921981437696 logging_writer.py:48] [132100] global_step=132100, grad_norm=4.763579368591309, loss=0.9284906983375549 +I0831 19:28:37.674286 139921973044992 logging_writer.py:48] [132200] global_step=132200, grad_norm=4.891814708709717, loss=1.0097081661224365 +I0831 19:29:04.831570 139921981437696 logging_writer.py:48] [132300] global_step=132300, grad_norm=5.05713415145874, loss=1.0623652935028076 +I0831 19:29:31.930169 139921973044992 logging_writer.py:48] [132400] global_step=132400, grad_norm=4.963854789733887, loss=0.9778732061386108 +I0831 19:29:59.033143 139921981437696 logging_writer.py:48] [132500] global_step=132500, grad_norm=5.015320777893066, loss=0.9533590078353882 +I0831 19:30:26.222620 139921973044992 logging_writer.py:48] [132600] global_step=132600, grad_norm=5.158733367919922, loss=0.9580050706863403 +I0831 19:30:53.334343 139921981437696 logging_writer.py:48] [132700] global_step=132700, grad_norm=4.869449615478516, loss=0.9974462985992432 +I0831 19:31:20.441978 139921973044992 logging_writer.py:48] [132800] global_step=132800, grad_norm=4.914292812347412, loss=0.9936086535453796 +I0831 19:31:47.604510 139921981437696 logging_writer.py:48] [132900] global_step=132900, grad_norm=5.020999431610107, loss=1.0403691530227661 +I0831 19:32:14.722224 139921973044992 logging_writer.py:48] [133000] global_step=133000, grad_norm=5.026878833770752, loss=1.0304293632507324 +I0831 19:32:42.060043 139921981437696 logging_writer.py:48] [133100] global_step=133100, grad_norm=5.031275272369385, loss=0.9774795770645142 +I0831 19:33:09.232914 139921973044992 logging_writer.py:48] [133200] global_step=133200, grad_norm=4.7352519035339355, loss=0.9272228479385376 +I0831 19:33:36.362954 139921981437696 logging_writer.py:48] [133300] global_step=133300, grad_norm=5.1211748123168945, loss=1.0546679496765137 +I0831 19:34:03.479348 139921973044992 logging_writer.py:48] [133400] global_step=133400, grad_norm=5.059954643249512, loss=1.0632152557373047 +I0831 19:34:30.636078 139921981437696 logging_writer.py:48] [133500] global_step=133500, grad_norm=4.871829032897949, loss=0.9675968885421753 +I0831 19:34:57.782880 139921973044992 logging_writer.py:48] [133600] global_step=133600, grad_norm=4.7295145988464355, loss=0.8519372344017029 +I0831 19:35:24.894394 139921981437696 logging_writer.py:48] [133700] global_step=133700, grad_norm=5.232987880706787, loss=1.0175331830978394 +I0831 19:35:52.068373 139921973044992 logging_writer.py:48] [133800] global_step=133800, grad_norm=4.597334861755371, loss=0.9407187104225159 +I0831 19:36:19.175031 139921981437696 logging_writer.py:48] [133900] global_step=133900, grad_norm=5.184926509857178, loss=1.1002558469772339 +I0831 19:36:46.282681 139921973044992 logging_writer.py:48] [134000] global_step=134000, grad_norm=4.896554946899414, loss=1.0148003101348877 +I0831 19:37:13.710310 139921981437696 logging_writer.py:48] [134100] global_step=134100, grad_norm=5.078643321990967, loss=1.0784714221954346 +I0831 19:37:40.816236 139921973044992 logging_writer.py:48] [134200] global_step=134200, grad_norm=4.943904876708984, loss=0.9443937540054321 +I0831 19:38:07.952146 139921981437696 logging_writer.py:48] [134300] global_step=134300, grad_norm=4.791806221008301, loss=0.9844144582748413 +I0831 19:38:35.134917 139921973044992 logging_writer.py:48] [134400] global_step=134400, grad_norm=5.0971245765686035, loss=1.0815157890319824 +I0831 19:39:02.252962 139921981437696 logging_writer.py:48] [134500] global_step=134500, grad_norm=4.526298522949219, loss=0.906464159488678 +I0831 19:39:29.363158 139921973044992 logging_writer.py:48] [134600] global_step=134600, grad_norm=4.744258403778076, loss=0.9701176881790161 +I0831 19:39:56.545047 139921981437696 logging_writer.py:48] [134700] global_step=134700, grad_norm=4.755051612854004, loss=0.9071029424667358 +I0831 19:40:23.693033 139921973044992 logging_writer.py:48] [134800] global_step=134800, grad_norm=5.067105293273926, loss=1.1246540546417236 +I0831 19:40:50.799924 139921981437696 logging_writer.py:48] [134900] global_step=134900, grad_norm=4.553475856781006, loss=0.9228195548057556 +I0831 19:41:17.955907 139921973044992 logging_writer.py:48] [135000] global_step=135000, grad_norm=5.556941032409668, loss=1.0210673809051514 +I0831 19:41:45.075597 139921981437696 logging_writer.py:48] [135100] global_step=135100, grad_norm=4.87706184387207, loss=0.901208758354187 +I0831 19:42:12.382789 139921973044992 logging_writer.py:48] [135200] global_step=135200, grad_norm=4.807921886444092, loss=0.9880447387695312 +I0831 19:42:39.576366 139921981437696 logging_writer.py:48] [135300] global_step=135300, grad_norm=4.762678146362305, loss=0.9933232069015503 +I0831 19:43:06.679334 139921973044992 logging_writer.py:48] [135400] global_step=135400, grad_norm=5.029171943664551, loss=1.0584275722503662 +I0831 19:43:33.777442 139921981437696 logging_writer.py:48] [135500] global_step=135500, grad_norm=4.969114303588867, loss=0.9844081401824951 +I0831 19:44:00.961402 139921973044992 logging_writer.py:48] [135600] global_step=135600, grad_norm=4.755460262298584, loss=0.915553092956543 +I0831 19:44:28.082743 139921981437696 logging_writer.py:48] [135700] global_step=135700, grad_norm=4.861955165863037, loss=0.9271779656410217 +I0831 19:44:55.208088 139921973044992 logging_writer.py:48] [135800] global_step=135800, grad_norm=4.815236568450928, loss=1.028063178062439 +I0831 19:45:22.366461 139921981437696 logging_writer.py:48] [135900] global_step=135900, grad_norm=4.966506481170654, loss=1.0628470182418823 +I0831 19:45:49.519267 139921973044992 logging_writer.py:48] [136000] global_step=136000, grad_norm=4.62230110168457, loss=0.9347306489944458 +I0831 19:46:16.650522 139921981437696 logging_writer.py:48] [136100] global_step=136100, grad_norm=5.026496887207031, loss=0.9974421262741089 +I0831 19:46:44.097599 139921973044992 logging_writer.py:48] [136200] global_step=136200, grad_norm=4.483848571777344, loss=0.9009636640548706 +I0831 19:47:11.239625 139921981437696 logging_writer.py:48] [136300] global_step=136300, grad_norm=4.840019702911377, loss=0.938897967338562 +I0831 19:47:38.340471 139921973044992 logging_writer.py:48] [136400] global_step=136400, grad_norm=5.029755592346191, loss=0.9433825016021729 +I0831 19:48:05.498586 139921981437696 logging_writer.py:48] [136500] global_step=136500, grad_norm=5.184850215911865, loss=1.0080307722091675 +I0831 19:48:32.599849 139921973044992 logging_writer.py:48] [136600] global_step=136600, grad_norm=5.105036735534668, loss=1.0315423011779785 +I0831 19:48:59.683015 139921981437696 logging_writer.py:48] [136700] global_step=136700, grad_norm=5.1069111824035645, loss=0.9624209403991699 +I0831 19:49:26.874944 139921973044992 logging_writer.py:48] [136800] global_step=136800, grad_norm=4.740869045257568, loss=0.9307385087013245 +I0831 19:49:54.004445 139921981437696 logging_writer.py:48] [136900] global_step=136900, grad_norm=4.946755409240723, loss=0.9665980935096741 +I0831 19:50:21.128611 139921973044992 logging_writer.py:48] [137000] global_step=137000, grad_norm=5.140575885772705, loss=0.965390682220459 +I0831 19:50:48.343280 139921981437696 logging_writer.py:48] [137100] global_step=137100, grad_norm=4.808712959289551, loss=0.9474648237228394 +I0831 19:51:15.446616 139921973044992 logging_writer.py:48] [137200] global_step=137200, grad_norm=4.800200462341309, loss=0.9657973647117615 +I0831 19:51:42.785100 139921981437696 logging_writer.py:48] [137300] global_step=137300, grad_norm=4.869572162628174, loss=0.941719114780426 +I0831 19:52:09.979587 139921973044992 logging_writer.py:48] [137400] global_step=137400, grad_norm=4.549490928649902, loss=0.9319491982460022 +I0831 19:52:37.085034 139921981437696 logging_writer.py:48] [137500] global_step=137500, grad_norm=4.469274520874023, loss=0.8801754117012024 +I0831 19:53:04.225728 139921973044992 logging_writer.py:48] [137600] global_step=137600, grad_norm=4.844291687011719, loss=1.0707017183303833 +I0831 19:53:31.419691 139921981437696 logging_writer.py:48] [137700] global_step=137700, grad_norm=4.740718841552734, loss=0.867843747138977 +I0831 19:53:58.554010 139921973044992 logging_writer.py:48] [137800] global_step=137800, grad_norm=5.038604259490967, loss=1.0009264945983887 +I0831 19:54:25.647140 139921981437696 logging_writer.py:48] [137900] global_step=137900, grad_norm=4.915534019470215, loss=1.0007017850875854 +I0831 19:54:52.817166 139921973044992 logging_writer.py:48] [138000] global_step=138000, grad_norm=5.0469536781311035, loss=1.0209492444992065 +I0831 19:55:19.955843 139921981437696 logging_writer.py:48] [138100] global_step=138100, grad_norm=5.265795707702637, loss=0.9732681512832642 +I0831 19:55:47.105166 139921973044992 logging_writer.py:48] [138200] global_step=138200, grad_norm=4.8287034034729, loss=0.9235097169876099 +I0831 19:56:14.473296 139921981437696 logging_writer.py:48] [138300] global_step=138300, grad_norm=5.062838077545166, loss=1.054419755935669 +I0831 19:56:41.569689 139921973044992 logging_writer.py:48] [138400] global_step=138400, grad_norm=4.922589302062988, loss=1.004551887512207 +I0831 19:57:08.704140 139921981437696 logging_writer.py:48] [138500] global_step=138500, grad_norm=4.890969276428223, loss=1.019485592842102 +I0831 19:57:35.886921 139921973044992 logging_writer.py:48] [138600] global_step=138600, grad_norm=5.1045989990234375, loss=1.0369521379470825 +I0831 19:58:02.982329 139921981437696 logging_writer.py:48] [138700] global_step=138700, grad_norm=5.1607770919799805, loss=1.0477265119552612 +I0831 19:58:30.128131 139921973044992 logging_writer.py:48] [138800] global_step=138800, grad_norm=4.7293477058410645, loss=0.9454880952835083 +I0831 19:58:37.346294 140117123622080 spec.py:333] Evaluating on the training split. +I0831 19:58:43.922130 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 19:58:54.650277 140117123622080 spec.py:363] Evaluating on the test split. +I0831 19:58:55.538075 140117123622080 submission_runner.py:516] Time since start: 38456.88s, Step: 138828, {'train/accuracy': Array(0.9412866, dtype=float32), 'train/loss': Array(0.20869601, dtype=float32), 'validation/accuracy': Array(0.75215995, dtype=float32), 'validation/loss': Array(1.0715029, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6236, dtype=float32), 'test/loss': Array(1.848797, dtype=float32), 'test/num_examples': 10000, 'score': 37976.86044549942, 'total_duration': 38456.877252578735, 'accumulated_submission_time': 37976.86044549942, 'accumulated_eval_time': 477.49180936813354, 'accumulated_logging_time': 1.4586880207061768} +I0831 19:58:55.590774 139921981437696 logging_writer.py:48] [138828] accumulated_eval_time=477.492, accumulated_logging_time=1.45869, accumulated_submission_time=37976.9, global_step=138828, preemption_count=0, score=37976.9, test/accuracy=0.6236000061035156, test/loss=1.8487969636917114, test/num_examples=10000, total_duration=38456.9, train/accuracy=0.9412866234779358, train/loss=0.20869600772857666, validation/accuracy=0.7521599531173706, validation/loss=1.071502923965454, validation/num_examples=50000 +I0831 19:59:15.508922 139921973044992 logging_writer.py:48] [138900] global_step=138900, grad_norm=4.8234782218933105, loss=0.8900262713432312 +I0831 19:59:42.633192 139921981437696 logging_writer.py:48] [139000] global_step=139000, grad_norm=5.154407978057861, loss=1.0618584156036377 +I0831 20:00:09.770620 139921973044992 logging_writer.py:48] [139100] global_step=139100, grad_norm=4.860782146453857, loss=1.000820279121399 +I0831 20:00:36.942743 139921981437696 logging_writer.py:48] [139200] global_step=139200, grad_norm=4.904153347015381, loss=0.9543589949607849 +I0831 20:01:04.064755 139921973044992 logging_writer.py:48] [139300] global_step=139300, grad_norm=4.804417610168457, loss=0.9436750411987305 +I0831 20:01:31.404233 139921981437696 logging_writer.py:48] [139400] global_step=139400, grad_norm=4.730575084686279, loss=0.9402377009391785 +I0831 20:01:58.590408 139921973044992 logging_writer.py:48] [139500] global_step=139500, grad_norm=4.640751838684082, loss=0.8534265160560608 +I0831 20:02:25.717682 139921981437696 logging_writer.py:48] [139600] global_step=139600, grad_norm=4.636613845825195, loss=0.8963083028793335 +I0831 20:02:52.839509 139921973044992 logging_writer.py:48] [139700] global_step=139700, grad_norm=4.8276753425598145, loss=0.9443851113319397 +I0831 20:03:20.010599 139921981437696 logging_writer.py:48] [139800] global_step=139800, grad_norm=4.82011079788208, loss=0.9489043354988098 +I0831 20:03:47.125815 139921973044992 logging_writer.py:48] [139900] global_step=139900, grad_norm=4.797784328460693, loss=0.9197148680686951 +I0831 20:04:14.228624 139921981437696 logging_writer.py:48] [140000] global_step=140000, grad_norm=4.893239498138428, loss=0.9636574983596802 +I0831 20:04:41.391919 139921973044992 logging_writer.py:48] [140100] global_step=140100, grad_norm=4.616306781768799, loss=0.967037558555603 +I0831 20:05:08.501805 139921981437696 logging_writer.py:48] [140200] global_step=140200, grad_norm=5.336060523986816, loss=1.0877015590667725 +I0831 20:05:35.615040 139921973044992 logging_writer.py:48] [140300] global_step=140300, grad_norm=5.1626200675964355, loss=1.0028671026229858 +I0831 20:06:02.797121 139921981437696 logging_writer.py:48] [140400] global_step=140400, grad_norm=4.831572532653809, loss=1.0165404081344604 +I0831 20:06:30.134616 139921973044992 logging_writer.py:48] [140500] global_step=140500, grad_norm=5.160471439361572, loss=0.9574214816093445 +I0831 20:06:57.250758 139921981437696 logging_writer.py:48] [140600] global_step=140600, grad_norm=5.002828121185303, loss=0.9509700536727905 +I0831 20:07:24.407297 139921973044992 logging_writer.py:48] [140700] global_step=140700, grad_norm=5.242112636566162, loss=1.0406849384307861 +I0831 20:07:51.517349 139921981437696 logging_writer.py:48] [140800] global_step=140800, grad_norm=4.744933128356934, loss=0.9733247756958008 +I0831 20:08:18.636330 139921973044992 logging_writer.py:48] [140900] global_step=140900, grad_norm=5.228756904602051, loss=0.9783738255500793 +I0831 20:08:45.811425 139921981437696 logging_writer.py:48] [141000] global_step=141000, grad_norm=5.19848108291626, loss=1.0371670722961426 +I0831 20:09:12.936215 139921973044992 logging_writer.py:48] [141100] global_step=141100, grad_norm=4.850997447967529, loss=1.011049509048462 +I0831 20:09:40.088277 139921981437696 logging_writer.py:48] [141200] global_step=141200, grad_norm=4.384255886077881, loss=0.899743914604187 +I0831 20:10:07.275839 139921973044992 logging_writer.py:48] [141300] global_step=141300, grad_norm=4.955875873565674, loss=0.9624778032302856 +I0831 20:10:34.419960 139921981437696 logging_writer.py:48] [141400] global_step=141400, grad_norm=5.027757167816162, loss=1.053900122642517 +I0831 20:11:01.761796 139921973044992 logging_writer.py:48] [141500] global_step=141500, grad_norm=4.928769588470459, loss=0.9593989849090576 +I0831 20:11:28.934743 139921981437696 logging_writer.py:48] [141600] global_step=141600, grad_norm=5.581726551055908, loss=1.0072463750839233 +I0831 20:11:56.062657 139921973044992 logging_writer.py:48] [141700] global_step=141700, grad_norm=4.846467018127441, loss=0.9202703237533569 +I0831 20:12:23.185478 139921981437696 logging_writer.py:48] [141800] global_step=141800, grad_norm=5.165030002593994, loss=0.9703384637832642 +I0831 20:12:50.361176 139921973044992 logging_writer.py:48] [141900] global_step=141900, grad_norm=4.560389041900635, loss=0.942169725894928 +I0831 20:13:17.487421 139921981437696 logging_writer.py:48] [142000] global_step=142000, grad_norm=5.1072516441345215, loss=0.9940844774246216 +I0831 20:13:44.612477 139921973044992 logging_writer.py:48] [142100] global_step=142100, grad_norm=4.906510353088379, loss=1.015462875366211 +I0831 20:14:11.799412 139921981437696 logging_writer.py:48] [142200] global_step=142200, grad_norm=4.926085472106934, loss=0.942711591720581 +I0831 20:14:38.938037 139921973044992 logging_writer.py:48] [142300] global_step=142300, grad_norm=4.589766979217529, loss=1.0212045907974243 +I0831 20:15:06.052488 139921981437696 logging_writer.py:48] [142400] global_step=142400, grad_norm=4.971861362457275, loss=0.9675861597061157 +I0831 20:15:33.226535 139921973044992 logging_writer.py:48] [142500] global_step=142500, grad_norm=4.831449508666992, loss=0.9861917495727539 +I0831 20:16:00.579202 139921981437696 logging_writer.py:48] [142600] global_step=142600, grad_norm=5.186265468597412, loss=0.9584445953369141 +I0831 20:16:27.681445 139921973044992 logging_writer.py:48] [142700] global_step=142700, grad_norm=4.917385578155518, loss=0.9690772294998169 +I0831 20:16:54.864398 139921981437696 logging_writer.py:48] [142800] global_step=142800, grad_norm=5.349570274353027, loss=1.074756383895874 +I0831 20:17:22.032712 139921973044992 logging_writer.py:48] [142900] global_step=142900, grad_norm=4.921330451965332, loss=1.0449920892715454 +I0831 20:17:49.152278 139921981437696 logging_writer.py:48] [143000] global_step=143000, grad_norm=4.9669342041015625, loss=0.9152712821960449 +I0831 20:18:16.325883 139921973044992 logging_writer.py:48] [143100] global_step=143100, grad_norm=4.546168804168701, loss=0.8652523756027222 +I0831 20:18:43.467555 139921981437696 logging_writer.py:48] [143200] global_step=143200, grad_norm=5.047435283660889, loss=1.0079898834228516 +I0831 20:19:10.566601 139921973044992 logging_writer.py:48] [143300] global_step=143300, grad_norm=4.836946964263916, loss=0.9640302658081055 +I0831 20:19:37.755253 139921981437696 logging_writer.py:48] [143400] global_step=143400, grad_norm=4.63958215713501, loss=0.9641320705413818 +I0831 20:20:04.897696 139921973044992 logging_writer.py:48] [143500] global_step=143500, grad_norm=4.659805774688721, loss=0.8832091093063354 +I0831 20:20:32.005401 139921981437696 logging_writer.py:48] [143600] global_step=143600, grad_norm=4.766984462738037, loss=1.0243116617202759 +I0831 20:20:59.407385 139921973044992 logging_writer.py:48] [143700] global_step=143700, grad_norm=4.656814098358154, loss=1.0206990242004395 +I0831 20:21:26.529431 139921981437696 logging_writer.py:48] [143800] global_step=143800, grad_norm=4.874742031097412, loss=0.9548232555389404 +I0831 20:21:53.649078 139921973044992 logging_writer.py:48] [143900] global_step=143900, grad_norm=4.959917068481445, loss=0.9766672849655151 +I0831 20:22:20.805253 139921981437696 logging_writer.py:48] [144000] global_step=144000, grad_norm=4.921894550323486, loss=0.9540019035339355 +I0831 20:22:47.911975 139921973044992 logging_writer.py:48] [144100] global_step=144100, grad_norm=5.424224376678467, loss=1.0936082601547241 +I0831 20:23:15.028042 139921981437696 logging_writer.py:48] [144200] global_step=144200, grad_norm=4.770810127258301, loss=0.937522292137146 +I0831 20:23:42.214372 139921973044992 logging_writer.py:48] [144300] global_step=144300, grad_norm=4.941946506500244, loss=0.9944703578948975 +I0831 20:24:09.308711 139921981437696 logging_writer.py:48] [144400] global_step=144400, grad_norm=4.950103282928467, loss=0.9230032563209534 +I0831 20:24:36.420889 139921973044992 logging_writer.py:48] [144500] global_step=144500, grad_norm=4.603263854980469, loss=0.9546459317207336 +I0831 20:25:03.597250 139921981437696 logging_writer.py:48] [144600] global_step=144600, grad_norm=4.835550308227539, loss=1.0063673257827759 +I0831 20:25:30.930794 139921973044992 logging_writer.py:48] [144700] global_step=144700, grad_norm=5.041654586791992, loss=0.97824627161026 +I0831 20:25:58.053539 139921981437696 logging_writer.py:48] [144800] global_step=144800, grad_norm=4.740554332733154, loss=0.9348577857017517 +I0831 20:26:25.245770 139921973044992 logging_writer.py:48] [144900] global_step=144900, grad_norm=4.888357162475586, loss=0.9911390542984009 +I0831 20:26:52.356118 139921981437696 logging_writer.py:48] [145000] global_step=145000, grad_norm=4.8480682373046875, loss=1.0248340368270874 +I0831 20:27:19.477410 139921973044992 logging_writer.py:48] [145100] global_step=145100, grad_norm=4.893320083618164, loss=1.0364820957183838 +I0831 20:27:46.649146 139921981437696 logging_writer.py:48] [145200] global_step=145200, grad_norm=4.831686496734619, loss=0.8924084901809692 +I0831 20:28:13.746164 139921973044992 logging_writer.py:48] [145300] global_step=145300, grad_norm=4.896422863006592, loss=0.9264111518859863 +I0831 20:28:40.852030 139921981437696 logging_writer.py:48] [145400] global_step=145400, grad_norm=5.085190773010254, loss=0.9875624179840088 +I0831 20:29:08.012637 139921973044992 logging_writer.py:48] [145500] global_step=145500, grad_norm=4.982946872711182, loss=1.0134501457214355 +I0831 20:29:35.115311 139921981437696 logging_writer.py:48] [145600] global_step=145600, grad_norm=4.838016510009766, loss=0.8934617638587952 +I0831 20:30:02.261304 139921973044992 logging_writer.py:48] [145700] global_step=145700, grad_norm=5.111640930175781, loss=0.9829938411712646 +I0831 20:30:29.647353 139921981437696 logging_writer.py:48] [145800] global_step=145800, grad_norm=4.872716426849365, loss=0.9983541965484619 +I0831 20:30:56.744948 139921973044992 logging_writer.py:48] [145900] global_step=145900, grad_norm=5.316831111907959, loss=1.0415749549865723 +I0831 20:31:23.872508 139921981437696 logging_writer.py:48] [146000] global_step=146000, grad_norm=4.864204406738281, loss=0.9499776363372803 +I0831 20:31:51.057854 139921973044992 logging_writer.py:48] [146100] global_step=146100, grad_norm=4.590548038482666, loss=0.8766248226165771 +I0831 20:32:11.774658 140117123622080 spec.py:333] Evaluating on the training split. +I0831 20:32:18.350536 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 20:32:29.403313 140117123622080 spec.py:363] Evaluating on the test split. +I0831 20:32:30.303316 140117123622080 submission_runner.py:516] Time since start: 40471.64s, Step: 146178, {'train/accuracy': Array(0.94202405, dtype=float32), 'train/loss': Array(0.20745388, dtype=float32), 'validation/accuracy': Array(0.75236, dtype=float32), 'validation/loss': Array(1.0740654, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62520003, dtype=float32), 'test/loss': Array(1.8540839, dtype=float32), 'test/num_examples': 10000, 'score': 39972.98211836815, 'total_duration': 40471.640963315964, 'accumulated_submission_time': 39972.98211836815, 'accumulated_eval_time': 496.0162832736969, 'accumulated_logging_time': 1.5198278427124023} +I0831 20:32:30.367457 139921981437696 logging_writer.py:48] [146178] accumulated_eval_time=496.016, accumulated_logging_time=1.51983, accumulated_submission_time=39973, global_step=146178, preemption_count=0, score=39973, test/accuracy=0.6252000331878662, test/loss=1.8540838956832886, test/num_examples=10000, total_duration=40471.6, train/accuracy=0.9420240521430969, train/loss=0.2074538767337799, validation/accuracy=0.7523599863052368, validation/loss=1.0740654468536377, validation/num_examples=50000 +I0831 20:32:36.773552 139921973044992 logging_writer.py:48] [146200] global_step=146200, grad_norm=4.696983337402344, loss=0.9203152656555176 +I0831 20:33:03.824853 139921981437696 logging_writer.py:48] [146300] global_step=146300, grad_norm=4.8906402587890625, loss=1.0387858152389526 +I0831 20:33:30.982028 139921973044992 logging_writer.py:48] [146400] global_step=146400, grad_norm=5.150010108947754, loss=1.0185277462005615 +I0831 20:33:58.085627 139921981437696 logging_writer.py:48] [146500] global_step=146500, grad_norm=4.773468494415283, loss=0.9522666335105896 +I0831 20:34:25.218222 139921973044992 logging_writer.py:48] [146600] global_step=146600, grad_norm=4.710721969604492, loss=0.985275149345398 +I0831 20:34:52.393690 139921981437696 logging_writer.py:48] [146700] global_step=146700, grad_norm=4.9756879806518555, loss=0.9200530648231506 +I0831 20:35:19.714695 139921973044992 logging_writer.py:48] [146800] global_step=146800, grad_norm=5.028190612792969, loss=1.0484095811843872 +I0831 20:35:46.853527 139921981437696 logging_writer.py:48] [146900] global_step=146900, grad_norm=5.036390781402588, loss=0.9611273407936096 +I0831 20:36:14.034417 139921973044992 logging_writer.py:48] [147000] global_step=147000, grad_norm=5.154229640960693, loss=1.073697566986084 +I0831 20:36:41.184140 139921981437696 logging_writer.py:48] [147100] global_step=147100, grad_norm=4.88695764541626, loss=0.9591866135597229 +I0831 20:37:08.312152 139921973044992 logging_writer.py:48] [147200] global_step=147200, grad_norm=5.267993927001953, loss=0.9773340225219727 +I0831 20:37:35.490525 139921981437696 logging_writer.py:48] [147300] global_step=147300, grad_norm=4.931237697601318, loss=1.0043814182281494 +I0831 20:38:02.606342 139921973044992 logging_writer.py:48] [147400] global_step=147400, grad_norm=5.1140217781066895, loss=0.9818305969238281 +I0831 20:38:29.735273 139921981437696 logging_writer.py:48] [147500] global_step=147500, grad_norm=4.740119457244873, loss=0.940901517868042 +I0831 20:38:56.893726 139921973044992 logging_writer.py:48] [147600] global_step=147600, grad_norm=4.819140434265137, loss=0.9516279697418213 +I0831 20:39:24.006998 139921981437696 logging_writer.py:48] [147700] global_step=147700, grad_norm=4.692424774169922, loss=0.9667174220085144 +I0831 20:39:51.124324 139921973044992 logging_writer.py:48] [147800] global_step=147800, grad_norm=4.842016220092773, loss=0.9538466930389404 +I0831 20:40:18.473081 139921981437696 logging_writer.py:48] [147900] global_step=147900, grad_norm=5.224215507507324, loss=1.04990553855896 +I0831 20:40:45.574705 139921973044992 logging_writer.py:48] [148000] global_step=148000, grad_norm=5.002373695373535, loss=1.0649664402008057 +I0831 20:41:12.693593 139921981437696 logging_writer.py:48] [148100] global_step=148100, grad_norm=4.8628644943237305, loss=0.9025545120239258 +I0831 20:41:39.859765 139921973044992 logging_writer.py:48] [148200] global_step=148200, grad_norm=4.849637508392334, loss=0.9863833785057068 +I0831 20:42:06.973795 139921981437696 logging_writer.py:48] [148300] global_step=148300, grad_norm=5.189327716827393, loss=1.01067316532135 +I0831 20:42:34.101449 139921973044992 logging_writer.py:48] [148400] global_step=148400, grad_norm=4.7538347244262695, loss=0.9074214100837708 +I0831 20:43:01.279737 139921981437696 logging_writer.py:48] [148500] global_step=148500, grad_norm=5.022421836853027, loss=0.8979357481002808 +I0831 20:43:28.386916 139921973044992 logging_writer.py:48] [148600] global_step=148600, grad_norm=5.051462173461914, loss=0.9741716980934143 +I0831 20:43:55.511293 139921981437696 logging_writer.py:48] [148700] global_step=148700, grad_norm=4.870255947113037, loss=0.9767467379570007 +I0831 20:44:22.708320 139921973044992 logging_writer.py:48] [148800] global_step=148800, grad_norm=4.671577453613281, loss=0.9314350485801697 +I0831 20:44:50.046501 139921981437696 logging_writer.py:48] [148900] global_step=148900, grad_norm=5.204407215118408, loss=1.0205132961273193 +I0831 20:45:17.152336 139921973044992 logging_writer.py:48] [149000] global_step=149000, grad_norm=5.015869617462158, loss=0.9394778609275818 +I0831 20:45:44.328975 139921981437696 logging_writer.py:48] [149100] global_step=149100, grad_norm=4.957120418548584, loss=0.9935173988342285 +I0831 20:46:11.438475 139921973044992 logging_writer.py:48] [149200] global_step=149200, grad_norm=4.940171718597412, loss=0.9533122777938843 +I0831 20:46:38.543114 139921981437696 logging_writer.py:48] [149300] global_step=149300, grad_norm=4.6103410720825195, loss=0.9639682769775391 +I0831 20:47:05.703466 139921973044992 logging_writer.py:48] [149400] global_step=149400, grad_norm=5.163661479949951, loss=1.069517970085144 +I0831 20:47:32.847211 139921981437696 logging_writer.py:48] [149500] global_step=149500, grad_norm=4.9699482917785645, loss=0.9375074505805969 +I0831 20:47:59.974213 139921973044992 logging_writer.py:48] [149600] global_step=149600, grad_norm=4.9002366065979, loss=1.0006431341171265 +I0831 20:48:27.144949 139921981437696 logging_writer.py:48] [149700] global_step=149700, grad_norm=4.9004998207092285, loss=0.9163181781768799 +I0831 20:48:54.240715 139921973044992 logging_writer.py:48] [149800] global_step=149800, grad_norm=4.796760082244873, loss=1.046330213546753 +I0831 20:49:21.355562 139921981437696 logging_writer.py:48] [149900] global_step=149900, grad_norm=4.8776774406433105, loss=0.9906493425369263 +I0831 20:49:48.772783 139921973044992 logging_writer.py:48] [150000] global_step=150000, grad_norm=4.853640556335449, loss=0.941401481628418 +I0831 20:50:15.891414 139921981437696 logging_writer.py:48] [150100] global_step=150100, grad_norm=4.88116979598999, loss=0.928198516368866 +I0831 20:50:43.019940 139921973044992 logging_writer.py:48] [150200] global_step=150200, grad_norm=5.326756000518799, loss=1.0813928842544556 +I0831 20:51:10.220665 139921981437696 logging_writer.py:48] [150300] global_step=150300, grad_norm=5.198439121246338, loss=1.0039770603179932 +I0831 20:51:37.350282 139921973044992 logging_writer.py:48] [150400] global_step=150400, grad_norm=5.161378383636475, loss=1.040886640548706 +I0831 20:52:04.482465 139921981437696 logging_writer.py:48] [150500] global_step=150500, grad_norm=5.211165428161621, loss=0.953351616859436 +I0831 20:52:31.676439 139921973044992 logging_writer.py:48] [150600] global_step=150600, grad_norm=4.999195098876953, loss=1.008204460144043 +I0831 20:52:58.781216 139921981437696 logging_writer.py:48] [150700] global_step=150700, grad_norm=5.1620049476623535, loss=0.968291163444519 +I0831 20:53:25.916902 139921973044992 logging_writer.py:48] [150800] global_step=150800, grad_norm=4.895872116088867, loss=0.9047017097473145 +I0831 20:53:53.103719 139921981437696 logging_writer.py:48] [150900] global_step=150900, grad_norm=4.975243091583252, loss=1.0069422721862793 +I0831 20:54:20.467739 139921973044992 logging_writer.py:48] [151000] global_step=151000, grad_norm=5.294717311859131, loss=1.0717651844024658 +I0831 20:54:47.607386 139921981437696 logging_writer.py:48] [151100] global_step=151100, grad_norm=4.959570407867432, loss=0.9760462641716003 +I0831 20:55:14.798783 139921973044992 logging_writer.py:48] [151200] global_step=151200, grad_norm=4.945578098297119, loss=0.9806424379348755 +I0831 20:55:41.936227 139921981437696 logging_writer.py:48] [151300] global_step=151300, grad_norm=5.226492404937744, loss=0.9860178232192993 +I0831 20:56:09.038343 139921973044992 logging_writer.py:48] [151400] global_step=151400, grad_norm=4.705639839172363, loss=0.8996001482009888 +I0831 20:56:36.198491 139921981437696 logging_writer.py:48] [151500] global_step=151500, grad_norm=5.014384746551514, loss=0.9587138295173645 +I0831 20:57:03.325675 139921973044992 logging_writer.py:48] [151600] global_step=151600, grad_norm=5.4270429611206055, loss=0.888384997844696 +I0831 20:57:30.439578 139921981437696 logging_writer.py:48] [151700] global_step=151700, grad_norm=4.7115325927734375, loss=0.8674713373184204 +I0831 20:57:57.610048 139921973044992 logging_writer.py:48] [151800] global_step=151800, grad_norm=4.94057559967041, loss=1.007075309753418 +I0831 20:58:24.703542 139921981437696 logging_writer.py:48] [151900] global_step=151900, grad_norm=4.5315704345703125, loss=0.8552232384681702 +I0831 20:58:51.813860 139921973044992 logging_writer.py:48] [152000] global_step=152000, grad_norm=4.99876594543457, loss=0.99898761510849 +I0831 20:59:19.213793 139921981437696 logging_writer.py:48] [152100] global_step=152100, grad_norm=4.990969657897949, loss=0.9361461400985718 +I0831 20:59:46.315607 139921973044992 logging_writer.py:48] [152200] global_step=152200, grad_norm=4.840461254119873, loss=0.962776243686676 +I0831 21:00:13.412754 139921981437696 logging_writer.py:48] [152300] global_step=152300, grad_norm=5.073439121246338, loss=0.9943138957023621 +I0831 21:00:40.589858 139921973044992 logging_writer.py:48] [152400] global_step=152400, grad_norm=5.18810510635376, loss=0.9988728165626526 +I0831 21:01:07.716291 139921981437696 logging_writer.py:48] [152500] global_step=152500, grad_norm=4.687082290649414, loss=0.9415919184684753 +I0831 21:01:34.812962 139921973044992 logging_writer.py:48] [152600] global_step=152600, grad_norm=4.698451995849609, loss=0.8820456862449646 +I0831 21:02:02.004604 139921981437696 logging_writer.py:48] [152700] global_step=152700, grad_norm=4.995190143585205, loss=0.9909572601318359 +I0831 21:02:29.143144 139921973044992 logging_writer.py:48] [152800] global_step=152800, grad_norm=5.163959503173828, loss=1.018420934677124 +I0831 21:02:56.285755 139921981437696 logging_writer.py:48] [152900] global_step=152900, grad_norm=4.865706443786621, loss=1.025252103805542 +I0831 21:03:23.495778 139921973044992 logging_writer.py:48] [153000] global_step=153000, grad_norm=5.1189374923706055, loss=1.0230038166046143 +I0831 21:03:50.610406 139921981437696 logging_writer.py:48] [153100] global_step=153100, grad_norm=4.9897332191467285, loss=0.9786919355392456 +I0831 21:04:17.942695 139921973044992 logging_writer.py:48] [153200] global_step=153200, grad_norm=4.974562644958496, loss=1.0017662048339844 +I0831 21:04:45.111339 139921981437696 logging_writer.py:48] [153300] global_step=153300, grad_norm=4.857293605804443, loss=0.9766420125961304 +I0831 21:05:12.215687 139921973044992 logging_writer.py:48] [153400] global_step=153400, grad_norm=5.221899509429932, loss=0.9248753190040588 +I0831 21:05:39.325645 139921981437696 logging_writer.py:48] [153500] global_step=153500, grad_norm=4.836357116699219, loss=0.9237293004989624 +I0831 21:05:46.305129 140117123622080 spec.py:333] Evaluating on the training split. +I0831 21:05:53.011794 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 21:06:11.170295 140117123622080 spec.py:363] Evaluating on the test split. +I0831 21:06:12.053911 140117123622080 submission_runner.py:516] Time since start: 42493.39s, Step: 153527, {'train/accuracy': Array(0.9441366, dtype=float32), 'train/loss': Array(0.19585101, dtype=float32), 'validation/accuracy': Array(0.75236, dtype=float32), 'validation/loss': Array(1.0734808, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62710005, dtype=float32), 'test/loss': Array(1.8547612, dtype=float32), 'test/num_examples': 10000, 'score': 41968.85789465904, 'total_duration': 42493.38979125023, 'accumulated_submission_time': 41968.85789465904, 'accumulated_eval_time': 521.7591128349304, 'accumulated_logging_time': 1.5930283069610596} +I0831 21:06:12.121558 139921973044992 logging_writer.py:48] [153527] accumulated_eval_time=521.759, accumulated_logging_time=1.59303, accumulated_submission_time=41968.9, global_step=153527, preemption_count=0, score=41968.9, test/accuracy=0.6271000504493713, test/loss=1.8547612428665161, test/num_examples=10000, total_duration=42493.4, train/accuracy=0.9441366195678711, train/loss=0.19585101306438446, validation/accuracy=0.7523599863052368, validation/loss=1.0734808444976807, validation/num_examples=50000 +I0831 21:06:33.298580 139921981437696 logging_writer.py:48] [153600] global_step=153600, grad_norm=5.034160614013672, loss=0.9882874488830566 +I0831 21:07:00.386076 139921973044992 logging_writer.py:48] [153700] global_step=153700, grad_norm=5.231109619140625, loss=0.971718430519104 +I0831 21:07:27.500596 139921981437696 logging_writer.py:48] [153800] global_step=153800, grad_norm=4.946824073791504, loss=1.0024484395980835 +I0831 21:07:54.676630 139921973044992 logging_writer.py:48] [153900] global_step=153900, grad_norm=4.8576273918151855, loss=0.9792189002037048 +I0831 21:08:21.796542 139921981437696 logging_writer.py:48] [154000] global_step=154000, grad_norm=4.861319065093994, loss=0.9175753593444824 +I0831 21:08:48.900004 139921973044992 logging_writer.py:48] [154100] global_step=154100, grad_norm=4.79464864730835, loss=0.9444677829742432 +I0831 21:09:16.309222 139921981437696 logging_writer.py:48] [154200] global_step=154200, grad_norm=5.088736534118652, loss=1.0336530208587646 +I0831 21:09:43.413495 139921973044992 logging_writer.py:48] [154300] global_step=154300, grad_norm=4.904961109161377, loss=0.9074363708496094 +I0831 21:10:28.759262 139921981437696 logging_writer.py:48] [154400] global_step=154400, grad_norm=5.118888854980469, loss=0.9921061992645264 +I0831 21:11:08.410100 139921973044992 logging_writer.py:48] [154500] global_step=154500, grad_norm=5.148547172546387, loss=1.0392872095108032 +I0831 21:11:35.500721 139921981437696 logging_writer.py:48] [154600] global_step=154600, grad_norm=5.398830413818359, loss=1.0125504732131958 +I0831 21:12:02.595202 139921973044992 logging_writer.py:48] [154700] global_step=154700, grad_norm=4.869438648223877, loss=0.9486622214317322 +I0831 21:12:29.806463 139921981437696 logging_writer.py:48] [154800] global_step=154800, grad_norm=5.235264778137207, loss=1.0351850986480713 +I0831 21:12:56.934554 139921973044992 logging_writer.py:48] [154900] global_step=154900, grad_norm=5.062743186950684, loss=0.9841554760932922 +I0831 21:13:24.093162 139921981437696 logging_writer.py:48] [155000] global_step=155000, grad_norm=4.78076696395874, loss=0.9402490854263306 +I0831 21:13:51.285740 139921973044992 logging_writer.py:48] [155100] global_step=155100, grad_norm=4.757939338684082, loss=0.9149352312088013 +I0831 21:14:18.402607 139921981437696 logging_writer.py:48] [155200] global_step=155200, grad_norm=4.800047874450684, loss=0.9138137102127075 +I0831 21:14:45.720505 139921973044992 logging_writer.py:48] [155300] global_step=155300, grad_norm=4.852859020233154, loss=0.9038362503051758 +I0831 21:15:12.913699 139921981437696 logging_writer.py:48] [155400] global_step=155400, grad_norm=4.932057857513428, loss=0.8957065343856812 +I0831 21:15:40.011597 139921973044992 logging_writer.py:48] [155500] global_step=155500, grad_norm=4.969590187072754, loss=0.9624759554862976 +I0831 21:16:07.130204 139921981437696 logging_writer.py:48] [155600] global_step=155600, grad_norm=5.094176769256592, loss=0.988711416721344 +I0831 21:16:34.298892 139921973044992 logging_writer.py:48] [155700] global_step=155700, grad_norm=5.019676208496094, loss=0.9568638801574707 +I0831 21:17:01.447249 139921981437696 logging_writer.py:48] [155800] global_step=155800, grad_norm=5.048218250274658, loss=1.0014100074768066 +I0831 21:17:28.579588 139921973044992 logging_writer.py:48] [155900] global_step=155900, grad_norm=4.878860950469971, loss=0.9643958806991577 +I0831 21:17:55.765517 139921981437696 logging_writer.py:48] [156000] global_step=156000, grad_norm=4.584903717041016, loss=0.8995145559310913 +I0831 21:18:22.898393 139921973044992 logging_writer.py:48] [156100] global_step=156100, grad_norm=5.039022922515869, loss=0.9491670727729797 +I0831 21:18:50.007716 139921981437696 logging_writer.py:48] [156200] global_step=156200, grad_norm=4.481903076171875, loss=0.8540570735931396 +I0831 21:19:17.393329 139921973044992 logging_writer.py:48] [156300] global_step=156300, grad_norm=5.051877975463867, loss=0.9942675232887268 +I0831 21:19:44.521752 139921981437696 logging_writer.py:48] [156400] global_step=156400, grad_norm=4.96096658706665, loss=0.9692563414573669 +I0831 21:20:11.655639 139921973044992 logging_writer.py:48] [156500] global_step=156500, grad_norm=5.165560722351074, loss=1.0188591480255127 +I0831 21:20:38.859828 139921981437696 logging_writer.py:48] [156600] global_step=156600, grad_norm=4.84659481048584, loss=0.8869377374649048 +I0831 21:21:05.958062 139921973044992 logging_writer.py:48] [156700] global_step=156700, grad_norm=5.205613136291504, loss=1.0169597864151 +I0831 21:21:33.067887 139921981437696 logging_writer.py:48] [156800] global_step=156800, grad_norm=5.096639633178711, loss=0.9939370155334473 +I0831 21:22:00.254528 139921973044992 logging_writer.py:48] [156900] global_step=156900, grad_norm=5.016160488128662, loss=0.9745461940765381 +I0831 21:22:27.334901 139921981437696 logging_writer.py:48] [157000] global_step=157000, grad_norm=4.833003044128418, loss=0.9389170408248901 +I0831 21:22:54.476541 139921973044992 logging_writer.py:48] [157100] global_step=157100, grad_norm=5.043243885040283, loss=0.9259165525436401 +I0831 21:23:21.642843 139921981437696 logging_writer.py:48] [157200] global_step=157200, grad_norm=4.843945503234863, loss=0.903425931930542 +I0831 21:23:48.754005 139921973044992 logging_writer.py:48] [157300] global_step=157300, grad_norm=5.170466899871826, loss=0.9466366767883301 +I0831 21:24:16.067406 139921981437696 logging_writer.py:48] [157400] global_step=157400, grad_norm=4.860477447509766, loss=1.0301798582077026 +I0831 21:24:43.241278 139921973044992 logging_writer.py:48] [157500] global_step=157500, grad_norm=4.928747177124023, loss=0.9555376768112183 +I0831 21:25:10.339690 139921981437696 logging_writer.py:48] [157600] global_step=157600, grad_norm=4.862509250640869, loss=0.929280161857605 +I0831 21:25:37.461072 139921973044992 logging_writer.py:48] [157700] global_step=157700, grad_norm=5.06353235244751, loss=0.9759159088134766 +I0831 21:26:04.639404 139921981437696 logging_writer.py:48] [157800] global_step=157800, grad_norm=5.426384449005127, loss=1.0437651872634888 +I0831 21:26:31.740968 139921973044992 logging_writer.py:48] [157900] global_step=157900, grad_norm=4.889151573181152, loss=0.9819507598876953 +I0831 21:26:58.845971 139921981437696 logging_writer.py:48] [158000] global_step=158000, grad_norm=5.146424770355225, loss=0.9344742298126221 +I0831 21:27:26.020149 139921973044992 logging_writer.py:48] [158100] global_step=158100, grad_norm=4.619809150695801, loss=0.8914041519165039 +I0831 21:27:53.141311 139921981437696 logging_writer.py:48] [158200] global_step=158200, grad_norm=4.658225059509277, loss=0.8950545191764832 +I0831 21:28:20.261678 139921973044992 logging_writer.py:48] [158300] global_step=158300, grad_norm=4.793298244476318, loss=0.919258713722229 +I0831 21:28:47.663483 139921981437696 logging_writer.py:48] [158400] global_step=158400, grad_norm=5.122467517852783, loss=1.041104793548584 +I0831 21:29:14.774887 139921973044992 logging_writer.py:48] [158500] global_step=158500, grad_norm=5.153772354125977, loss=0.9115997552871704 +I0831 21:29:41.879817 139921981437696 logging_writer.py:48] [158600] global_step=158600, grad_norm=5.215651035308838, loss=0.9954379200935364 +I0831 21:30:09.053952 139921973044992 logging_writer.py:48] [158700] global_step=158700, grad_norm=5.008596420288086, loss=1.0144150257110596 +I0831 21:30:36.155250 139921981437696 logging_writer.py:48] [158800] global_step=158800, grad_norm=4.744813919067383, loss=0.8769266605377197 +I0831 21:31:03.244291 139921973044992 logging_writer.py:48] [158900] global_step=158900, grad_norm=4.899735450744629, loss=0.8875149488449097 +I0831 21:31:30.408761 139921981437696 logging_writer.py:48] [159000] global_step=159000, grad_norm=4.964328289031982, loss=0.9556666016578674 +I0831 21:31:57.536421 139921973044992 logging_writer.py:48] [159100] global_step=159100, grad_norm=4.980753421783447, loss=1.04117751121521 +I0831 21:32:24.646649 139921981437696 logging_writer.py:48] [159200] global_step=159200, grad_norm=5.03514289855957, loss=1.0239049196243286 +I0831 21:32:51.993651 139921973044992 logging_writer.py:48] [159300] global_step=159300, grad_norm=5.203268051147461, loss=0.9854373335838318 +I0831 21:33:19.085396 139921981437696 logging_writer.py:48] [159400] global_step=159400, grad_norm=5.023107528686523, loss=0.890872597694397 +I0831 21:33:46.390279 139921973044992 logging_writer.py:48] [159500] global_step=159500, grad_norm=4.850172519683838, loss=0.9252333045005798 +I0831 21:34:13.598324 139921981437696 logging_writer.py:48] [159600] global_step=159600, grad_norm=4.773884296417236, loss=0.9374220967292786 +I0831 21:34:40.703186 139921973044992 logging_writer.py:48] [159700] global_step=159700, grad_norm=5.113864898681641, loss=1.0642116069793701 +I0831 21:35:07.848459 139921981437696 logging_writer.py:48] [159800] global_step=159800, grad_norm=5.125539302825928, loss=1.0054385662078857 +I0831 21:35:35.007029 139921973044992 logging_writer.py:48] [159900] global_step=159900, grad_norm=4.812593460083008, loss=0.9136924147605896 +I0831 21:36:02.156879 139921981437696 logging_writer.py:48] [160000] global_step=160000, grad_norm=4.953841209411621, loss=0.9339091777801514 +I0831 21:36:29.313459 139921973044992 logging_writer.py:48] [160100] global_step=160100, grad_norm=4.8145527839660645, loss=0.8957515358924866 +I0831 21:36:56.490889 139921981437696 logging_writer.py:48] [160200] global_step=160200, grad_norm=5.234736919403076, loss=0.9969123005867004 +I0831 21:37:23.605755 139921973044992 logging_writer.py:48] [160300] global_step=160300, grad_norm=4.963168621063232, loss=0.9264186024665833 +I0831 21:37:50.760081 139921981437696 logging_writer.py:48] [160400] global_step=160400, grad_norm=4.898416042327881, loss=0.9302302598953247 +I0831 21:38:18.184638 139921973044992 logging_writer.py:48] [160500] global_step=160500, grad_norm=4.952752590179443, loss=0.9232945442199707 +I0831 21:38:45.320846 139921981437696 logging_writer.py:48] [160600] global_step=160600, grad_norm=4.999945163726807, loss=0.9090070724487305 +I0831 21:39:12.424799 139921973044992 logging_writer.py:48] [160700] global_step=160700, grad_norm=4.993600845336914, loss=1.006201982498169 +I0831 21:39:28.068223 140117123622080 spec.py:333] Evaluating on the training split. +I0831 21:39:34.835332 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 21:39:46.721556 140117123622080 spec.py:363] Evaluating on the test split. +I0831 21:39:47.608752 140117123622080 submission_runner.py:516] Time since start: 44508.95s, Step: 160759, {'train/accuracy': Array(0.9484614, dtype=float32), 'train/loss': Array(0.18218228, dtype=float32), 'validation/accuracy': Array(0.75246, dtype=float32), 'validation/loss': Array(1.0760537, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62740004, dtype=float32), 'test/loss': Array(1.861663, dtype=float32), 'test/num_examples': 10000, 'score': 43964.741797447205, 'total_duration': 44508.94835686684, 'accumulated_submission_time': 43964.741797447205, 'accumulated_eval_time': 541.2974145412445, 'accumulated_logging_time': 1.670964002609253} +I0831 21:39:47.672915 139921981437696 logging_writer.py:48] [160759] accumulated_eval_time=541.297, accumulated_logging_time=1.67096, accumulated_submission_time=43964.7, global_step=160759, preemption_count=0, score=43964.7, test/accuracy=0.6274000406265259, test/loss=1.8616629838943481, test/num_examples=10000, total_duration=44508.9, train/accuracy=0.9484614133834839, train/loss=0.18218228220939636, validation/accuracy=0.7524600028991699, validation/loss=1.0760537385940552, validation/num_examples=50000 +I0831 21:39:59.113121 139921973044992 logging_writer.py:48] [160800] global_step=160800, grad_norm=5.261658191680908, loss=1.005176067352295 +I0831 21:40:26.179374 139921981437696 logging_writer.py:48] [160900] global_step=160900, grad_norm=4.993444919586182, loss=0.9471278190612793 +I0831 21:40:53.297456 139921973044992 logging_writer.py:48] [161000] global_step=161000, grad_norm=5.014404296875, loss=0.9628086090087891 +I0831 21:41:20.488173 139921981437696 logging_writer.py:48] [161100] global_step=161100, grad_norm=5.071488380432129, loss=0.9343177080154419 +I0831 21:41:47.600602 139921973044992 logging_writer.py:48] [161200] global_step=161200, grad_norm=5.15348482131958, loss=0.9353330135345459 +I0831 21:42:14.714593 139921981437696 logging_writer.py:48] [161300] global_step=161300, grad_norm=4.770608901977539, loss=0.8641740083694458 +I0831 21:42:41.880837 139921973044992 logging_writer.py:48] [161400] global_step=161400, grad_norm=4.931656360626221, loss=0.9511432647705078 +I0831 21:43:22.881442 139921981437696 logging_writer.py:48] [161500] global_step=161500, grad_norm=5.000479221343994, loss=0.9763637185096741 +I0831 21:44:03.604378 139921973044992 logging_writer.py:48] [161600] global_step=161600, grad_norm=5.1131591796875, loss=1.0261188745498657 +I0831 21:44:48.982674 139921981437696 logging_writer.py:48] [161700] global_step=161700, grad_norm=5.074275970458984, loss=1.033536672592163 +I0831 21:45:34.560266 139921973044992 logging_writer.py:48] [161800] global_step=161800, grad_norm=5.155675411224365, loss=0.9512190222740173 +I0831 21:46:21.980813 139921981437696 logging_writer.py:48] [161900] global_step=161900, grad_norm=4.648331165313721, loss=0.9452334642410278 +I0831 21:47:07.258297 139921973044992 logging_writer.py:48] [162000] global_step=162000, grad_norm=5.267159938812256, loss=1.0287482738494873 +I0831 21:47:50.781215 139921981437696 logging_writer.py:48] [162100] global_step=162100, grad_norm=5.091073513031006, loss=0.9112268686294556 +I0831 21:48:35.557784 139921973044992 logging_writer.py:48] [162200] global_step=162200, grad_norm=4.943297386169434, loss=0.899752140045166 +I0831 21:49:20.494024 139921981437696 logging_writer.py:48] [162300] global_step=162300, grad_norm=4.989020824432373, loss=0.9870470762252808 +I0831 21:50:04.681305 139921973044992 logging_writer.py:48] [162400] global_step=162400, grad_norm=5.097851753234863, loss=0.966347873210907 +I0831 21:50:51.288289 139921981437696 logging_writer.py:48] [162500] global_step=162500, grad_norm=4.99167013168335, loss=0.9261490702629089 +I0831 21:51:34.583935 139921973044992 logging_writer.py:48] [162600] global_step=162600, grad_norm=4.667116165161133, loss=0.8542503118515015 +I0831 21:52:11.076623 139921981437696 logging_writer.py:48] [162700] global_step=162700, grad_norm=4.596889019012451, loss=0.8661216497421265 +I0831 21:52:58.647097 139921973044992 logging_writer.py:48] [162800] global_step=162800, grad_norm=5.100932598114014, loss=0.9826565980911255 +I0831 21:53:45.503569 139921981437696 logging_writer.py:48] [162900] global_step=162900, grad_norm=5.2621636390686035, loss=0.9855632185935974 +I0831 21:54:29.613566 139921973044992 logging_writer.py:48] [163000] global_step=163000, grad_norm=5.075774669647217, loss=1.0296882390975952 +I0831 21:55:13.778763 139921981437696 logging_writer.py:48] [163100] global_step=163100, grad_norm=4.734865188598633, loss=0.9463732242584229 +I0831 21:56:01.156057 139921973044992 logging_writer.py:48] [163200] global_step=163200, grad_norm=4.744150638580322, loss=0.8646335601806641 +I0831 21:56:45.024356 139921981437696 logging_writer.py:48] [163300] global_step=163300, grad_norm=5.2253546714782715, loss=0.9405995607376099 +I0831 21:57:29.921602 139921973044992 logging_writer.py:48] [163400] global_step=163400, grad_norm=4.965240955352783, loss=0.9429086446762085 +I0831 21:58:14.151387 139921981437696 logging_writer.py:48] [163500] global_step=163500, grad_norm=4.9924635887146, loss=0.922538161277771 +I0831 21:58:57.594089 139921973044992 logging_writer.py:48] [163600] global_step=163600, grad_norm=4.746679306030273, loss=0.8413926959037781 +I0831 21:59:42.085360 139921981437696 logging_writer.py:48] [163700] global_step=163700, grad_norm=5.060418605804443, loss=0.9192659854888916 +I0831 22:00:25.724208 139921973044992 logging_writer.py:48] [163800] global_step=163800, grad_norm=5.1425347328186035, loss=1.0337450504302979 +I0831 22:01:09.758153 139921981437696 logging_writer.py:48] [163900] global_step=163900, grad_norm=5.0433735847473145, loss=0.97238689661026 +I0831 22:01:55.284538 139921973044992 logging_writer.py:48] [164000] global_step=164000, grad_norm=5.16209077835083, loss=0.943888247013092 +I0831 22:02:39.864356 139921981437696 logging_writer.py:48] [164100] global_step=164100, grad_norm=4.9954986572265625, loss=0.9839959144592285 +I0831 22:03:23.026341 139921973044992 logging_writer.py:48] [164200] global_step=164200, grad_norm=5.195619106292725, loss=0.9774796962738037 +I0831 22:04:05.257540 139921981437696 logging_writer.py:48] [164300] global_step=164300, grad_norm=5.159951686859131, loss=0.9475569725036621 +I0831 22:04:50.261797 139921973044992 logging_writer.py:48] [164400] global_step=164400, grad_norm=5.1195149421691895, loss=0.9835314750671387 +I0831 22:05:33.932302 139921981437696 logging_writer.py:48] [164500] global_step=164500, grad_norm=5.26022481918335, loss=0.9862913489341736 +I0831 22:06:16.453764 139921973044992 logging_writer.py:48] [164600] global_step=164600, grad_norm=4.999389171600342, loss=0.872920036315918 +I0831 22:06:59.892820 139921981437696 logging_writer.py:48] [164700] global_step=164700, grad_norm=5.249314308166504, loss=0.9495363235473633 +I0831 22:07:43.791248 139921973044992 logging_writer.py:48] [164800] global_step=164800, grad_norm=4.921785831451416, loss=0.9014332890510559 +I0831 22:08:27.694732 139921981437696 logging_writer.py:48] [164900] global_step=164900, grad_norm=4.780940532684326, loss=0.9377914667129517 +I0831 22:09:10.197206 139921973044992 logging_writer.py:48] [165000] global_step=165000, grad_norm=5.033972263336182, loss=0.9830811619758606 +I0831 22:09:53.767311 139921981437696 logging_writer.py:48] [165100] global_step=165100, grad_norm=5.083905220031738, loss=0.9609131813049316 +I0831 22:10:39.021340 139921973044992 logging_writer.py:48] [165200] global_step=165200, grad_norm=5.067040920257568, loss=1.0165965557098389 +I0831 22:11:22.265437 139921981437696 logging_writer.py:48] [165300] global_step=165300, grad_norm=5.333278656005859, loss=0.9564242362976074 +I0831 22:12:05.096498 139921973044992 logging_writer.py:48] [165400] global_step=165400, grad_norm=5.0396904945373535, loss=0.9295569658279419 +I0831 22:12:48.406819 139921981437696 logging_writer.py:48] [165500] global_step=165500, grad_norm=4.853050231933594, loss=0.906808614730835 +I0831 22:13:03.891163 140117123622080 spec.py:333] Evaluating on the training split. +I0831 22:13:12.881479 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 22:13:49.510512 140117123622080 spec.py:363] Evaluating on the test split. +I0831 22:13:50.553205 140117123622080 submission_runner.py:516] Time since start: 46551.75s, Step: 165536, {'train/accuracy': Array(0.9486009, dtype=float32), 'train/loss': Array(0.1856623, dtype=float32), 'validation/accuracy': Array(0.75286, dtype=float32), 'validation/loss': Array(1.0761234, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62450004, dtype=float32), 'test/loss': Array(1.8669696, dtype=float32), 'test/num_examples': 10000, 'score': 45960.87983632088, 'total_duration': 46551.746829509735, 'accumulated_submission_time': 45960.87983632088, 'accumulated_eval_time': 587.8112554550171, 'accumulated_logging_time': 1.769763708114624} +I0831 22:13:51.013066 139921973044992 logging_writer.py:48] [165536] accumulated_eval_time=587.811, accumulated_logging_time=1.76976, accumulated_submission_time=45960.9, global_step=165536, preemption_count=0, score=45960.9, test/accuracy=0.624500036239624, test/loss=1.8669695854187012, test/num_examples=10000, total_duration=46551.7, train/accuracy=0.9486008882522583, train/loss=0.18566229939460754, validation/accuracy=0.7528600096702576, validation/loss=1.0761233568191528, validation/num_examples=50000 +I0831 22:14:14.922623 139921981437696 logging_writer.py:48] [165600] global_step=165600, grad_norm=5.152842044830322, loss=1.0235539674758911 +I0831 22:14:57.028702 139921973044992 logging_writer.py:48] [165700] global_step=165700, grad_norm=4.846169471740723, loss=0.9405519366264343 +I0831 22:15:40.309983 139921981437696 logging_writer.py:48] [165800] global_step=165800, grad_norm=4.994511604309082, loss=0.9140400886535645 +I0831 22:16:23.011892 139921973044992 logging_writer.py:48] [165900] global_step=165900, grad_norm=4.916687488555908, loss=0.9318277835845947 +I0831 22:17:06.139215 139921981437696 logging_writer.py:48] [166000] global_step=166000, grad_norm=5.230715751647949, loss=0.9093665480613708 +I0831 22:17:48.931558 139921973044992 logging_writer.py:48] [166100] global_step=166100, grad_norm=4.925557613372803, loss=0.9330472350120544 +I0831 22:18:33.566501 139921981437696 logging_writer.py:48] [166200] global_step=166200, grad_norm=5.06247091293335, loss=0.9562727212905884 +I0831 22:19:17.380453 139921973044992 logging_writer.py:48] [166300] global_step=166300, grad_norm=5.189027309417725, loss=0.9611095190048218 +I0831 22:20:02.024781 139921981437696 logging_writer.py:48] [166400] global_step=166400, grad_norm=4.735781192779541, loss=0.861931324005127 +I0831 22:20:48.558485 139921973044992 logging_writer.py:48] [166500] global_step=166500, grad_norm=4.748887062072754, loss=0.9387449026107788 +I0831 22:21:33.124875 139921981437696 logging_writer.py:48] [166600] global_step=166600, grad_norm=4.938686847686768, loss=0.9204044342041016 +I0831 22:22:18.323078 139921973044992 logging_writer.py:48] [166700] global_step=166700, grad_norm=5.0654449462890625, loss=0.9294266700744629 +I0831 22:23:00.586066 139921981437696 logging_writer.py:48] [166800] global_step=166800, grad_norm=5.326050758361816, loss=0.9896588325500488 +I0831 22:23:45.022441 139921973044992 logging_writer.py:48] [166900] global_step=166900, grad_norm=5.000044345855713, loss=0.9614710807800293 +I0831 22:24:28.135305 139921981437696 logging_writer.py:48] [167000] global_step=167000, grad_norm=4.910777568817139, loss=0.9328413605690002 +I0831 22:25:12.395816 139921973044992 logging_writer.py:48] [167100] global_step=167100, grad_norm=5.0432820320129395, loss=0.9362127780914307 +I0831 22:25:55.439225 139921981437696 logging_writer.py:48] [167200] global_step=167200, grad_norm=4.923275470733643, loss=0.8459237813949585 +I0831 22:26:37.882254 139921973044992 logging_writer.py:48] [167300] global_step=167300, grad_norm=5.096238613128662, loss=1.0115020275115967 +I0831 22:27:21.258184 139921981437696 logging_writer.py:48] [167400] global_step=167400, grad_norm=5.086239814758301, loss=0.9223483800888062 +I0831 22:28:04.920325 139921973044992 logging_writer.py:48] [167500] global_step=167500, grad_norm=5.230425834655762, loss=0.9665791392326355 +I0831 22:28:48.704797 139921981437696 logging_writer.py:48] [167600] global_step=167600, grad_norm=5.191025733947754, loss=0.9763803482055664 +I0831 22:29:33.970164 139921973044992 logging_writer.py:48] [167700] global_step=167700, grad_norm=5.307724952697754, loss=1.0581448078155518 +I0831 22:30:17.525082 139921981437696 logging_writer.py:48] [167800] global_step=167800, grad_norm=5.082458019256592, loss=0.9685335755348206 +I0831 22:31:01.313699 139921973044992 logging_writer.py:48] [167900] global_step=167900, grad_norm=4.569759845733643, loss=0.8071920871734619 +I0831 22:31:46.298068 139921981437696 logging_writer.py:48] [168000] global_step=168000, grad_norm=5.328387260437012, loss=1.050620198249817 +I0831 22:32:30.557644 139921973044992 logging_writer.py:48] [168100] global_step=168100, grad_norm=4.8404388427734375, loss=0.8947907090187073 +I0831 22:33:15.366693 139921981437696 logging_writer.py:48] [168200] global_step=168200, grad_norm=5.052520751953125, loss=0.9142324328422546 +I0831 22:33:59.090009 139921973044992 logging_writer.py:48] [168300] global_step=168300, grad_norm=5.283833980560303, loss=0.9668012857437134 +I0831 22:34:44.005503 139921981437696 logging_writer.py:48] [168400] global_step=168400, grad_norm=4.930121898651123, loss=0.9361954927444458 +I0831 22:35:26.806147 139921973044992 logging_writer.py:48] [168500] global_step=168500, grad_norm=4.663738250732422, loss=0.8189729452133179 +I0831 22:36:10.097494 139921981437696 logging_writer.py:48] [168600] global_step=168600, grad_norm=4.893496990203857, loss=0.9989497065544128 +I0831 22:36:53.327615 139921973044992 logging_writer.py:48] [168700] global_step=168700, grad_norm=4.958418846130371, loss=0.8832401633262634 +I0831 22:37:37.516379 139921981437696 logging_writer.py:48] [168800] global_step=168800, grad_norm=4.915536403656006, loss=0.9655736088752747 +I0831 22:38:27.569994 139921973044992 logging_writer.py:48] [168900] global_step=168900, grad_norm=5.123279571533203, loss=0.9623851776123047 +I0831 22:39:17.687958 139921981437696 logging_writer.py:48] [169000] global_step=169000, grad_norm=5.337967872619629, loss=0.9678080081939697 +I0831 22:40:05.584502 139921973044992 logging_writer.py:48] [169100] global_step=169100, grad_norm=4.7867913246154785, loss=0.9171231985092163 +I0831 22:40:57.784759 139921981437696 logging_writer.py:48] [169200] global_step=169200, grad_norm=5.0116705894470215, loss=0.9390310645103455 +I0831 22:41:49.225080 139921973044992 logging_writer.py:48] [169300] global_step=169300, grad_norm=4.759862899780273, loss=0.9013236165046692 +I0831 22:42:44.342139 139921981437696 logging_writer.py:48] [169400] global_step=169400, grad_norm=5.504498481750488, loss=1.0429799556732178 +I0831 22:43:36.311613 139921973044992 logging_writer.py:48] [169500] global_step=169500, grad_norm=5.349216938018799, loss=1.0471644401550293 +I0831 22:44:27.803708 139921981437696 logging_writer.py:48] [169600] global_step=169600, grad_norm=5.331131935119629, loss=0.958016037940979 +I0831 22:45:18.610998 139921973044992 logging_writer.py:48] [169700] global_step=169700, grad_norm=4.926514625549316, loss=0.9387076497077942 +I0831 22:46:06.779950 139921981437696 logging_writer.py:48] [169800] global_step=169800, grad_norm=4.949702739715576, loss=0.8957004547119141 +I0831 22:46:53.622480 139921973044992 logging_writer.py:48] [169900] global_step=169900, grad_norm=5.235446453094482, loss=0.9655469059944153 +I0831 22:47:06.506012 140117123622080 spec.py:333] Evaluating on the training split. +I0831 22:47:15.806367 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 22:47:26.548753 140117123622080 spec.py:363] Evaluating on the test split. +I0831 22:47:27.660902 140117123622080 submission_runner.py:516] Time since start: 48568.78s, Step: 169928, {'train/accuracy': Array(0.9453723, dtype=float32), 'train/loss': Array(0.18652251, dtype=float32), 'validation/accuracy': Array(0.75308, dtype=float32), 'validation/loss': Array(1.0758554, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62670004, dtype=float32), 'test/loss': Array(1.8681524, dtype=float32), 'test/num_examples': 10000, 'score': 47956.30036401749, 'total_duration': 48568.78353047371, 'accumulated_submission_time': 47956.30036401749, 'accumulated_eval_time': 608.7469418048859, 'accumulated_logging_time': 2.263536214828491} +I0831 22:47:28.070046 139921981437696 logging_writer.py:48] [169928] accumulated_eval_time=608.747, accumulated_logging_time=2.26354, accumulated_submission_time=47956.3, global_step=169928, preemption_count=0, score=47956.3, test/accuracy=0.6267000436782837, test/loss=1.868152379989624, test/num_examples=10000, total_duration=48568.8, train/accuracy=0.9453722834587097, train/loss=0.18652251362800598, validation/accuracy=0.7530800104141235, validation/loss=1.0758553743362427, validation/num_examples=50000 +I0831 22:47:57.251699 139921973044992 logging_writer.py:48] [170000] global_step=170000, grad_norm=4.998445510864258, loss=0.9480268955230713 +I0831 22:48:45.591417 139921981437696 logging_writer.py:48] [170100] global_step=170100, grad_norm=5.008410930633545, loss=0.884440541267395 +I0831 22:49:35.194232 139921973044992 logging_writer.py:48] [170200] global_step=170200, grad_norm=5.103313446044922, loss=0.9354877471923828 +I0831 22:50:25.384115 139921981437696 logging_writer.py:48] [170300] global_step=170300, grad_norm=5.174259662628174, loss=0.9878000020980835 +I0831 22:51:11.686558 139921973044992 logging_writer.py:48] [170400] global_step=170400, grad_norm=5.005756855010986, loss=0.9410803318023682 +I0831 22:51:57.122400 139921981437696 logging_writer.py:48] [170500] global_step=170500, grad_norm=5.11229133605957, loss=1.0123660564422607 +I0831 22:52:43.051028 139921973044992 logging_writer.py:48] [170600] global_step=170600, grad_norm=4.969985008239746, loss=1.043041706085205 +I0831 22:53:28.107566 139921981437696 logging_writer.py:48] [170700] global_step=170700, grad_norm=5.088776588439941, loss=0.9430950284004211 +I0831 22:54:11.931905 139921973044992 logging_writer.py:48] [170800] global_step=170800, grad_norm=4.821752071380615, loss=0.93293297290802 +I0831 22:54:55.692300 139921981437696 logging_writer.py:48] [170900] global_step=170900, grad_norm=5.208302974700928, loss=0.9638180732727051 +I0831 22:55:39.409644 139921973044992 logging_writer.py:48] [171000] global_step=171000, grad_norm=4.924722671508789, loss=0.8736580610275269 +I0831 22:56:22.629141 139921981437696 logging_writer.py:48] [171100] global_step=171100, grad_norm=5.169553279876709, loss=0.9676843881607056 +I0831 22:57:06.129951 139921973044992 logging_writer.py:48] [171200] global_step=171200, grad_norm=5.00455379486084, loss=0.9939327239990234 +I0831 22:57:48.569530 139921981437696 logging_writer.py:48] [171300] global_step=171300, grad_norm=4.776683807373047, loss=0.9372166991233826 +I0831 22:58:32.319631 139921973044992 logging_writer.py:48] [171400] global_step=171400, grad_norm=5.140227317810059, loss=0.988563597202301 +I0831 22:59:16.057984 139921981437696 logging_writer.py:48] [171500] global_step=171500, grad_norm=5.284642696380615, loss=0.9524602890014648 +I0831 22:59:59.858161 139921973044992 logging_writer.py:48] [171600] global_step=171600, grad_norm=5.172173500061035, loss=0.979625940322876 +I0831 23:00:43.826171 139921981437696 logging_writer.py:48] [171700] global_step=171700, grad_norm=5.593188762664795, loss=0.9959746599197388 +I0831 23:01:27.521092 139921973044992 logging_writer.py:48] [171800] global_step=171800, grad_norm=5.024170398712158, loss=0.9131293892860413 +I0831 23:02:12.314534 139921981437696 logging_writer.py:48] [171900] global_step=171900, grad_norm=5.120051860809326, loss=0.9485231637954712 +I0831 23:02:55.519143 139921973044992 logging_writer.py:48] [172000] global_step=172000, grad_norm=5.104409217834473, loss=0.9050700664520264 +I0831 23:03:39.857613 139921981437696 logging_writer.py:48] [172100] global_step=172100, grad_norm=5.3043904304504395, loss=0.9672328233718872 +I0831 23:04:23.892124 139921973044992 logging_writer.py:48] [172200] global_step=172200, grad_norm=5.172736167907715, loss=0.9962968826293945 +I0831 23:05:07.265544 139921981437696 logging_writer.py:48] [172300] global_step=172300, grad_norm=4.750157833099365, loss=0.8858863115310669 +I0831 23:05:50.750036 139921973044992 logging_writer.py:48] [172400] global_step=172400, grad_norm=5.1699113845825195, loss=0.9299823045730591 +I0831 23:06:34.471627 139921981437696 logging_writer.py:48] [172500] global_step=172500, grad_norm=4.560922145843506, loss=0.8254849910736084 +I0831 23:07:20.282427 139921973044992 logging_writer.py:48] [172600] global_step=172600, grad_norm=4.878159523010254, loss=0.9229831695556641 +I0831 23:08:07.807038 139921981437696 logging_writer.py:48] [172700] global_step=172700, grad_norm=5.207629680633545, loss=0.971900224685669 +I0831 23:08:51.812801 139921973044992 logging_writer.py:48] [172800] global_step=172800, grad_norm=4.814146995544434, loss=0.9414615631103516 +I0831 23:09:36.079755 139921981437696 logging_writer.py:48] [172900] global_step=172900, grad_norm=5.244767189025879, loss=0.9907903075218201 +I0831 23:10:18.279233 139921973044992 logging_writer.py:48] [173000] global_step=173000, grad_norm=5.095126152038574, loss=0.9178517460823059 +I0831 23:11:01.442292 139921981437696 logging_writer.py:48] [173100] global_step=173100, grad_norm=4.928292751312256, loss=0.8655149936676025 +I0831 23:11:44.709715 139921973044992 logging_writer.py:48] [173200] global_step=173200, grad_norm=5.063683986663818, loss=0.968812882900238 +I0831 23:12:27.780645 139921981437696 logging_writer.py:48] [173300] global_step=173300, grad_norm=4.902136325836182, loss=0.9553601145744324 +I0831 23:13:12.330918 139921973044992 logging_writer.py:48] [173400] global_step=173400, grad_norm=4.786780834197998, loss=0.8494292497634888 +I0831 23:13:56.192322 139921981437696 logging_writer.py:48] [173500] global_step=173500, grad_norm=5.114274024963379, loss=0.9875612258911133 +I0831 23:14:40.002919 139921973044992 logging_writer.py:48] [173600] global_step=173600, grad_norm=5.057380199432373, loss=0.9982177019119263 +I0831 23:15:23.230577 139921981437696 logging_writer.py:48] [173700] global_step=173700, grad_norm=5.449744701385498, loss=0.9683394432067871 +I0831 23:16:07.220607 139921973044992 logging_writer.py:48] [173800] global_step=173800, grad_norm=5.162412643432617, loss=0.9892996549606323 +I0831 23:16:52.946490 139921981437696 logging_writer.py:48] [173900] global_step=173900, grad_norm=4.876076698303223, loss=0.9287992715835571 +I0831 23:17:38.402201 139921973044992 logging_writer.py:48] [174000] global_step=174000, grad_norm=4.851548194885254, loss=1.009970784187317 +I0831 23:18:24.593317 139921981437696 logging_writer.py:48] [174100] global_step=174100, grad_norm=5.410205841064453, loss=1.0253218412399292 +I0831 23:19:07.653833 139921973044992 logging_writer.py:48] [174200] global_step=174200, grad_norm=5.18351411819458, loss=0.9238368272781372 +I0831 23:19:52.146668 139921981437696 logging_writer.py:48] [174300] global_step=174300, grad_norm=5.080911636352539, loss=0.995248019695282 +I0831 23:20:36.264925 139921973044992 logging_writer.py:48] [174400] global_step=174400, grad_norm=5.013103485107422, loss=0.9252077341079712 +I0831 23:20:43.801705 140117123622080 spec.py:333] Evaluating on the training split. +I0831 23:20:52.512462 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 23:21:08.274061 140117123622080 spec.py:363] Evaluating on the test split. +I0831 23:21:09.382743 140117123622080 submission_runner.py:516] Time since start: 50590.51s, Step: 174418, {'train/accuracy': Array(0.9489995, dtype=float32), 'train/loss': Array(0.18063688, dtype=float32), 'validation/accuracy': Array(0.75373995, dtype=float32), 'validation/loss': Array(1.076455, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6291, dtype=float32), 'test/loss': Array(1.8668636, dtype=float32), 'test/num_examples': 10000, 'score': 49951.91783928871, 'total_duration': 50590.50589442253, 'accumulated_submission_time': 49951.91783928871, 'accumulated_eval_time': 634.1093051433563, 'accumulated_logging_time': 2.748711347579956} +I0831 23:21:09.782138 139921981437696 logging_writer.py:48] [174418] accumulated_eval_time=634.109, accumulated_logging_time=2.74871, accumulated_submission_time=49951.9, global_step=174418, preemption_count=0, score=49951.9, test/accuracy=0.6291000247001648, test/loss=1.8668636083602905, test/num_examples=10000, total_duration=50590.5, train/accuracy=0.9489995241165161, train/loss=0.18063688278198242, validation/accuracy=0.7537399530410767, validation/loss=1.076454997062683, validation/num_examples=50000 +I0831 23:21:40.028460 139921973044992 logging_writer.py:48] [174500] global_step=174500, grad_norm=5.036942005157471, loss=1.0144895315170288 +I0831 23:22:24.631801 139921981437696 logging_writer.py:48] [174600] global_step=174600, grad_norm=4.71986722946167, loss=0.8249006271362305 +I0831 23:23:08.740225 139921973044992 logging_writer.py:48] [174700] global_step=174700, grad_norm=5.120420932769775, loss=0.8937488794326782 +I0831 23:23:53.352868 139921981437696 logging_writer.py:48] [174800] global_step=174800, grad_norm=4.892276287078857, loss=0.9498565196990967 +I0831 23:24:40.531604 139921973044992 logging_writer.py:48] [174900] global_step=174900, grad_norm=5.234351634979248, loss=0.901939868927002 +I0831 23:25:28.227542 139921981437696 logging_writer.py:48] [175000] global_step=175000, grad_norm=5.099279403686523, loss=0.9688783884048462 +I0831 23:26:17.906653 139921973044992 logging_writer.py:48] [175100] global_step=175100, grad_norm=5.000911235809326, loss=0.9047040939331055 +I0831 23:27:11.928079 139921981437696 logging_writer.py:48] [175200] global_step=175200, grad_norm=5.129055500030518, loss=0.9797968864440918 +I0831 23:28:01.333704 139921973044992 logging_writer.py:48] [175300] global_step=175300, grad_norm=5.034256935119629, loss=0.9143140912055969 +I0831 23:28:47.607225 139921981437696 logging_writer.py:48] [175400] global_step=175400, grad_norm=4.902943134307861, loss=0.9525947570800781 +I0831 23:29:32.923178 139921973044992 logging_writer.py:48] [175500] global_step=175500, grad_norm=4.944500923156738, loss=0.866134524345398 +I0831 23:30:19.003915 139921981437696 logging_writer.py:48] [175600] global_step=175600, grad_norm=5.275557041168213, loss=0.9017285108566284 +I0831 23:31:04.211094 139921973044992 logging_writer.py:48] [175700] global_step=175700, grad_norm=5.1265549659729, loss=0.9883629083633423 +I0831 23:31:52.504684 139921981437696 logging_writer.py:48] [175800] global_step=175800, grad_norm=5.000940322875977, loss=0.9335716962814331 +I0831 23:32:37.429608 139921973044992 logging_writer.py:48] [175900] global_step=175900, grad_norm=5.322895050048828, loss=1.0109267234802246 +I0831 23:33:23.257832 139921981437696 logging_writer.py:48] [176000] global_step=176000, grad_norm=5.2084197998046875, loss=0.923934817314148 +I0831 23:34:08.694683 139921973044992 logging_writer.py:48] [176100] global_step=176100, grad_norm=5.172661304473877, loss=0.9868496656417847 +I0831 23:34:52.916353 139921981437696 logging_writer.py:48] [176200] global_step=176200, grad_norm=5.065589427947998, loss=0.9173805713653564 +I0831 23:35:37.555284 139921973044992 logging_writer.py:48] [176300] global_step=176300, grad_norm=5.212427139282227, loss=0.9420467615127563 +I0831 23:36:24.233928 139921981437696 logging_writer.py:48] [176400] global_step=176400, grad_norm=4.928747177124023, loss=0.9401435852050781 +I0831 23:37:09.769961 139921973044992 logging_writer.py:48] [176500] global_step=176500, grad_norm=5.2131829261779785, loss=0.9882749319076538 +I0831 23:37:55.087405 139921981437696 logging_writer.py:48] [176600] global_step=176600, grad_norm=5.1907548904418945, loss=0.9286671280860901 +I0831 23:38:39.678059 139921973044992 logging_writer.py:48] [176700] global_step=176700, grad_norm=5.211300373077393, loss=0.9826594591140747 +I0831 23:39:22.354059 139921981437696 logging_writer.py:48] [176800] global_step=176800, grad_norm=4.938755035400391, loss=1.0394479036331177 +I0831 23:40:06.694070 139921973044992 logging_writer.py:48] [176900] global_step=176900, grad_norm=5.09529972076416, loss=0.9819705486297607 +I0831 23:40:49.963165 139921981437696 logging_writer.py:48] [177000] global_step=177000, grad_norm=5.128530502319336, loss=0.9570082426071167 +I0831 23:41:34.772982 139921973044992 logging_writer.py:48] [177100] global_step=177100, grad_norm=5.148158073425293, loss=0.9816645979881287 +I0831 23:42:18.207826 139921981437696 logging_writer.py:48] [177200] global_step=177200, grad_norm=5.059715270996094, loss=0.9705163240432739 +I0831 23:43:02.567141 139921973044992 logging_writer.py:48] [177300] global_step=177300, grad_norm=5.0406494140625, loss=0.9050667881965637 +I0831 23:43:46.148232 139921981437696 logging_writer.py:48] [177400] global_step=177400, grad_norm=5.232635974884033, loss=1.032281517982483 +I0831 23:44:29.680655 139921973044992 logging_writer.py:48] [177500] global_step=177500, grad_norm=5.034666538238525, loss=0.9409810900688171 +I0831 23:45:13.530337 139921981437696 logging_writer.py:48] [177600] global_step=177600, grad_norm=5.083624839782715, loss=1.0042052268981934 +I0831 23:45:59.862985 139921973044992 logging_writer.py:48] [177700] global_step=177700, grad_norm=5.054671764373779, loss=0.9332374334335327 +I0831 23:46:45.202867 139921981437696 logging_writer.py:48] [177800] global_step=177800, grad_norm=4.885261058807373, loss=0.8499916195869446 +I0831 23:47:31.006020 139921973044992 logging_writer.py:48] [177900] global_step=177900, grad_norm=4.946290493011475, loss=0.9739598035812378 +I0831 23:48:16.235295 139921981437696 logging_writer.py:48] [178000] global_step=178000, grad_norm=4.73599100112915, loss=0.86484694480896 +I0831 23:49:00.645013 139921973044992 logging_writer.py:48] [178100] global_step=178100, grad_norm=5.039834976196289, loss=0.9000093936920166 +I0831 23:49:44.841328 139921981437696 logging_writer.py:48] [178200] global_step=178200, grad_norm=5.075593948364258, loss=0.9157423973083496 +I0831 23:50:28.589150 139921973044992 logging_writer.py:48] [178300] global_step=178300, grad_norm=4.948075771331787, loss=0.9464122653007507 +I0831 23:51:12.541247 139921981437696 logging_writer.py:48] [178400] global_step=178400, grad_norm=4.939866542816162, loss=0.8853955864906311 +I0831 23:51:55.264941 139921973044992 logging_writer.py:48] [178500] global_step=178500, grad_norm=5.152597427368164, loss=1.0153712034225464 +I0831 23:52:38.799505 139921981437696 logging_writer.py:48] [178600] global_step=178600, grad_norm=5.232766628265381, loss=0.9601128697395325 +I0831 23:53:22.195260 139921973044992 logging_writer.py:48] [178700] global_step=178700, grad_norm=4.925014972686768, loss=0.9187140464782715 +I0831 23:54:05.660070 139921981437696 logging_writer.py:48] [178800] global_step=178800, grad_norm=5.127266883850098, loss=0.9560840129852295 +I0831 23:54:25.257220 140117123622080 spec.py:333] Evaluating on the training split. +I0831 23:54:33.961980 140117123622080 spec.py:346] Evaluating on the validation split. +I0831 23:55:03.584019 140117123622080 spec.py:363] Evaluating on the test split. +I0831 23:55:04.690753 140117123622080 submission_runner.py:516] Time since start: 52625.81s, Step: 178846, {'train/accuracy': Array(0.94782364, dtype=float32), 'train/loss': Array(0.1783816, dtype=float32), 'validation/accuracy': Array(0.75402, dtype=float32), 'validation/loss': Array(1.0778383, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6283, dtype=float32), 'test/loss': Array(1.8672477, dtype=float32), 'test/num_examples': 10000, 'score': 51947.29002213478, 'total_duration': 52625.81468272209, 'accumulated_submission_time': 51947.29002213478, 'accumulated_eval_time': 673.3249371051788, 'accumulated_logging_time': 3.187523365020752} +I0831 23:55:05.095427 139921973044992 logging_writer.py:48] [178846] accumulated_eval_time=673.325, accumulated_logging_time=3.18752, accumulated_submission_time=51947.3, global_step=178846, preemption_count=0, score=51947.3, test/accuracy=0.6283000111579895, test/loss=1.8672477006912231, test/num_examples=10000, total_duration=52625.8, train/accuracy=0.9478236436843872, train/loss=0.17838160693645477, validation/accuracy=0.7540199756622314, validation/loss=1.0778383016586304, validation/num_examples=50000 +I0831 23:55:25.687951 139921981437696 logging_writer.py:48] [178900] global_step=178900, grad_norm=4.7738566398620605, loss=0.9123994708061218 +I0831 23:56:12.051110 139921973044992 logging_writer.py:48] [179000] global_step=179000, grad_norm=4.919876575469971, loss=0.9970911741256714 +I0831 23:56:56.311345 139921981437696 logging_writer.py:48] [179100] global_step=179100, grad_norm=5.053291320800781, loss=0.9524286985397339 +I0831 23:57:38.745913 139921973044992 logging_writer.py:48] [179200] global_step=179200, grad_norm=5.342507362365723, loss=0.9632333517074585 +I0831 23:58:21.894665 139921981437696 logging_writer.py:48] [179300] global_step=179300, grad_norm=5.184536457061768, loss=0.9496970176696777 +I0831 23:59:06.080851 139921973044992 logging_writer.py:48] [179400] global_step=179400, grad_norm=5.214294910430908, loss=0.9824342727661133 +I0831 23:59:49.258600 139921981437696 logging_writer.py:48] [179500] global_step=179500, grad_norm=5.116392612457275, loss=0.9222437143325806 +I0901 00:00:31.647682 139921973044992 logging_writer.py:48] [179600] global_step=179600, grad_norm=5.161383152008057, loss=0.9583783149719238 +I0901 00:01:16.033063 139921981437696 logging_writer.py:48] [179700] global_step=179700, grad_norm=5.029388427734375, loss=0.8476839065551758 +I0901 00:01:58.868396 139921973044992 logging_writer.py:48] [179800] global_step=179800, grad_norm=5.040289402008057, loss=0.9052124619483948 +I0901 00:02:41.261280 139921981437696 logging_writer.py:48] [179900] global_step=179900, grad_norm=5.06241512298584, loss=0.8546245694160461 +I0901 00:03:24.809909 139921973044992 logging_writer.py:48] [180000] global_step=180000, grad_norm=5.231296539306641, loss=0.8575757741928101 +I0901 00:04:09.115883 139921981437696 logging_writer.py:48] [180100] global_step=180100, grad_norm=4.84977912902832, loss=0.811313807964325 +I0901 00:04:54.875895 139921973044992 logging_writer.py:48] [180200] global_step=180200, grad_norm=5.097011089324951, loss=0.9459633827209473 +I0901 00:05:37.857642 139921981437696 logging_writer.py:48] [180300] global_step=180300, grad_norm=4.91051721572876, loss=0.9447720050811768 +I0901 00:06:23.308517 139921973044992 logging_writer.py:48] [180400] global_step=180400, grad_norm=5.018645286560059, loss=0.9116261005401611 +I0901 00:07:06.482326 139921981437696 logging_writer.py:48] [180500] global_step=180500, grad_norm=5.501119613647461, loss=0.9736320376396179 +I0901 00:07:50.951189 139921981437696 logging_writer.py:48] [180600] global_step=180600, grad_norm=5.021002292633057, loss=0.9508739113807678 +I0901 00:08:35.152726 139921973044992 logging_writer.py:48] [180700] global_step=180700, grad_norm=4.840488910675049, loss=0.860907256603241 +I0901 00:09:19.262144 139921981437696 logging_writer.py:48] [180800] global_step=180800, grad_norm=5.266315937042236, loss=0.9292709827423096 +I0901 00:10:03.621520 139921973044992 logging_writer.py:48] [180900] global_step=180900, grad_norm=4.8468017578125, loss=0.8942832946777344 +I0901 00:10:48.640219 139921981437696 logging_writer.py:48] [181000] global_step=181000, grad_norm=4.9875078201293945, loss=0.9300423264503479 +I0901 00:11:31.095170 139921973044992 logging_writer.py:48] [181100] global_step=181100, grad_norm=5.115767955780029, loss=0.9372349977493286 +I0901 00:12:13.964455 139921981437696 logging_writer.py:48] [181200] global_step=181200, grad_norm=5.002644062042236, loss=0.8996254205703735 +I0901 00:12:57.344704 139921973044992 logging_writer.py:48] [181300] global_step=181300, grad_norm=4.869000434875488, loss=0.8720095753669739 +I0901 00:13:41.602272 139921981437696 logging_writer.py:48] [181400] global_step=181400, grad_norm=5.159334182739258, loss=0.9714869260787964 +I0901 00:14:26.678514 139921973044992 logging_writer.py:48] [181500] global_step=181500, grad_norm=4.869708061218262, loss=0.9027442932128906 +I0901 00:15:11.029421 139921981437696 logging_writer.py:48] [181600] global_step=181600, grad_norm=5.202916145324707, loss=0.9343599081039429 +I0901 00:15:53.918643 139921973044992 logging_writer.py:48] [181700] global_step=181700, grad_norm=4.824665546417236, loss=0.9024068713188171 +I0901 00:16:36.143254 139921981437696 logging_writer.py:48] [181800] global_step=181800, grad_norm=5.38730525970459, loss=0.9736725091934204 +I0901 00:17:19.628851 139921973044992 logging_writer.py:48] [181900] global_step=181900, grad_norm=4.835203170776367, loss=0.9001156091690063 +I0901 00:18:02.228013 139921981437696 logging_writer.py:48] [182000] global_step=182000, grad_norm=4.637960910797119, loss=0.8328962922096252 +I0901 00:18:46.028717 139921973044992 logging_writer.py:48] [182100] global_step=182100, grad_norm=5.118022441864014, loss=0.9740627408027649 +I0901 00:19:27.679543 139921981437696 logging_writer.py:48] [182200] global_step=182200, grad_norm=5.2687506675720215, loss=0.9869484901428223 +I0901 00:20:09.523435 139921973044992 logging_writer.py:48] [182300] global_step=182300, grad_norm=4.761861324310303, loss=0.8835271596908569 +I0901 00:20:52.386115 139921981437696 logging_writer.py:48] [182400] global_step=182400, grad_norm=5.1224565505981445, loss=0.8871628046035767 +I0901 00:21:37.216631 139921973044992 logging_writer.py:48] [182500] global_step=182500, grad_norm=4.992258071899414, loss=0.8937028646469116 +I0901 00:22:20.678274 139921981437696 logging_writer.py:48] [182600] global_step=182600, grad_norm=4.682897090911865, loss=0.8550283908843994 +I0901 00:23:05.695472 139921973044992 logging_writer.py:48] [182700] global_step=182700, grad_norm=5.14408016204834, loss=0.9484037160873413 +I0901 00:23:48.363830 139921981437696 logging_writer.py:48] [182800] global_step=182800, grad_norm=5.119777679443359, loss=0.953913688659668 +I0901 00:24:32.075027 139921973044992 logging_writer.py:48] [182900] global_step=182900, grad_norm=4.929784297943115, loss=0.8799233436584473 +I0901 00:25:15.160916 139921981437696 logging_writer.py:48] [183000] global_step=183000, grad_norm=4.88779878616333, loss=0.890139102935791 +I0901 00:25:57.887360 139921973044992 logging_writer.py:48] [183100] global_step=183100, grad_norm=4.943493843078613, loss=0.8662341237068176 +I0901 00:26:42.583266 139921981437696 logging_writer.py:48] [183200] global_step=183200, grad_norm=5.190832138061523, loss=1.0011495351791382 +I0901 00:27:26.249900 139921973044992 logging_writer.py:48] [183300] global_step=183300, grad_norm=4.979455471038818, loss=0.9568437337875366 +I0901 00:28:09.631524 139921981437696 logging_writer.py:48] [183400] global_step=183400, grad_norm=5.206514835357666, loss=0.9533624053001404 +I0901 00:28:20.481903 140117123622080 spec.py:333] Evaluating on the training split. +I0901 00:28:29.376064 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 00:29:01.784539 140117123622080 spec.py:363] Evaluating on the test split. +I0901 00:29:02.882599 140117123622080 submission_runner.py:516] Time since start: 54664.02s, Step: 183424, {'train/accuracy': Array(0.9513512, dtype=float32), 'train/loss': Array(0.17198183, dtype=float32), 'validation/accuracy': Array(0.75394, dtype=float32), 'validation/loss': Array(1.0774896, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.629, dtype=float32), 'test/loss': Array(1.8643882, dtype=float32), 'test/num_examples': 10000, 'score': 53942.5781993866, 'total_duration': 54664.01664042473, 'accumulated_submission_time': 53942.5781993866, 'accumulated_eval_time': 715.5178511142731, 'accumulated_logging_time': 3.6479856967926025} +I0901 00:29:03.294786 139921973044992 logging_writer.py:48] [183424] accumulated_eval_time=715.518, accumulated_logging_time=3.64799, accumulated_submission_time=53942.6, global_step=183424, preemption_count=0, score=53942.6, test/accuracy=0.6290000081062317, test/loss=1.8643882274627686, test/num_examples=10000, total_duration=54664, train/accuracy=0.9513512253761292, train/loss=0.1719818264245987, validation/accuracy=0.7539399862289429, validation/loss=1.0774896144866943, validation/num_examples=50000 +I0901 00:29:32.287621 139921981437696 logging_writer.py:48] [183500] global_step=183500, grad_norm=5.1103668212890625, loss=0.8879998922348022 +I0901 00:30:17.246209 139921973044992 logging_writer.py:48] [183600] global_step=183600, grad_norm=5.357842922210693, loss=0.9694778919219971 +I0901 00:31:03.037174 139921981437696 logging_writer.py:48] [183700] global_step=183700, grad_norm=5.092708587646484, loss=0.9667546153068542 +I0901 00:31:47.740716 139921973044992 logging_writer.py:48] [183800] global_step=183800, grad_norm=4.915397644042969, loss=0.916071891784668 +I0901 00:32:33.149178 139921981437696 logging_writer.py:48] [183900] global_step=183900, grad_norm=4.934107780456543, loss=0.8932393789291382 +I0901 00:33:19.549926 139921973044992 logging_writer.py:48] [184000] global_step=184000, grad_norm=4.749549388885498, loss=0.8645725250244141 +I0901 00:34:06.753557 139921981437696 logging_writer.py:48] [184100] global_step=184100, grad_norm=4.74613094329834, loss=0.899001955986023 +I0901 00:34:52.138211 139921973044992 logging_writer.py:48] [184200] global_step=184200, grad_norm=4.96061897277832, loss=1.021587610244751 +I0901 00:35:36.613556 139921981437696 logging_writer.py:48] [184300] global_step=184300, grad_norm=5.042257308959961, loss=0.9675810933113098 +I0901 00:36:20.740285 139921973044992 logging_writer.py:48] [184400] global_step=184400, grad_norm=4.943033695220947, loss=0.905627965927124 +I0901 00:37:04.519054 139921981437696 logging_writer.py:48] [184500] global_step=184500, grad_norm=4.846328258514404, loss=0.881272554397583 +I0901 00:37:48.375007 139921973044992 logging_writer.py:48] [184600] global_step=184600, grad_norm=5.079095363616943, loss=0.9245611429214478 +I0901 00:38:34.693285 139921981437696 logging_writer.py:48] [184700] global_step=184700, grad_norm=5.189856052398682, loss=0.9617631435394287 +I0901 00:39:21.428112 139921973044992 logging_writer.py:48] [184800] global_step=184800, grad_norm=5.138073921203613, loss=0.9528495073318481 +I0901 00:40:08.252868 139921981437696 logging_writer.py:48] [184900] global_step=184900, grad_norm=4.697085380554199, loss=0.9094182848930359 +I0901 00:40:55.346812 139921973044992 logging_writer.py:48] [185000] global_step=185000, grad_norm=5.10360050201416, loss=0.9181907773017883 +I0901 00:41:43.390104 139921981437696 logging_writer.py:48] [185100] global_step=185100, grad_norm=5.074392318725586, loss=0.8846005201339722 +I0901 00:42:32.771038 139921973044992 logging_writer.py:48] [185200] global_step=185200, grad_norm=5.203348159790039, loss=0.9448322057723999 +I0901 00:43:21.054605 139921981437696 logging_writer.py:48] [185300] global_step=185300, grad_norm=5.02539587020874, loss=0.9068973064422607 +I0901 00:44:08.774408 139921973044992 logging_writer.py:48] [185400] global_step=185400, grad_norm=5.111128330230713, loss=0.8813860416412354 +I0901 00:44:57.122422 139921981437696 logging_writer.py:48] [185500] global_step=185500, grad_norm=4.922668933868408, loss=0.9293126463890076 +I0901 00:45:45.310685 139921973044992 logging_writer.py:48] [185600] global_step=185600, grad_norm=5.185923099517822, loss=0.9751316905021667 +I0901 00:46:34.778699 139921981437696 logging_writer.py:48] [185700] global_step=185700, grad_norm=5.09031343460083, loss=0.9528994560241699 +I0901 00:47:22.460766 139921973044992 logging_writer.py:48] [185800] global_step=185800, grad_norm=4.994513034820557, loss=0.8622792959213257 +I0901 00:48:10.875787 139921981437696 logging_writer.py:48] [185900] global_step=185900, grad_norm=5.013795852661133, loss=0.974343478679657 +I0901 00:49:00.352304 139921973044992 logging_writer.py:48] [186000] global_step=186000, grad_norm=4.79968786239624, loss=0.8142518401145935 +I0901 00:49:48.955619 139921981437696 logging_writer.py:48] [186100] global_step=186100, grad_norm=4.993161201477051, loss=0.9759147763252258 +I0901 00:50:38.232373 139921973044992 logging_writer.py:48] [186200] global_step=186200, grad_norm=4.572606086730957, loss=0.8213826417922974 +I0901 00:51:28.050569 139921981437696 logging_writer.py:48] [186300] global_step=186300, grad_norm=5.006608009338379, loss=0.9728635549545288 +I0901 00:52:19.087521 139921973044992 logging_writer.py:48] [186400] global_step=186400, grad_norm=5.536976337432861, loss=0.9993458986282349 +I0901 00:53:12.517505 139921981437696 logging_writer.py:48] [186500] global_step=186500, grad_norm=5.303413391113281, loss=0.9438822865486145 +I0901 00:54:03.559844 139921973044992 logging_writer.py:48] [186600] global_step=186600, grad_norm=5.533377170562744, loss=0.9145150780677795 +I0901 00:54:55.914268 139921981437696 logging_writer.py:48] [186700] global_step=186700, grad_norm=5.044100761413574, loss=0.8958327174186707 +I0901 00:55:47.968154 139921973044992 logging_writer.py:48] [186800] global_step=186800, grad_norm=5.123433589935303, loss=0.9352436065673828 +I0901 00:56:42.089492 139921981437696 logging_writer.py:48] [186900] global_step=186900, grad_norm=5.096984386444092, loss=0.8700875043869019 +I0901 00:57:35.730834 139921973044992 logging_writer.py:48] [187000] global_step=187000, grad_norm=5.177615642547607, loss=0.9319977760314941 +I0901 00:58:29.354430 139921981437696 logging_writer.py:48] [187100] global_step=187100, grad_norm=5.174818992614746, loss=0.935737133026123 +I0901 00:59:21.303946 139921973044992 logging_writer.py:48] [187200] global_step=187200, grad_norm=4.944273948669434, loss=0.8643767833709717 +I0901 01:00:15.169902 139921981437696 logging_writer.py:48] [187300] global_step=187300, grad_norm=4.985434532165527, loss=0.8361901640892029 +I0901 01:01:10.068271 139921973044992 logging_writer.py:48] [187400] global_step=187400, grad_norm=5.204710483551025, loss=0.8938449025154114 +I0901 01:02:04.327478 139921981437696 logging_writer.py:48] [187500] global_step=187500, grad_norm=4.793647289276123, loss=0.9224609136581421 +I0901 01:02:18.755343 140117123622080 spec.py:333] Evaluating on the training split. +I0901 01:02:27.351781 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 01:02:57.510299 140117123622080 spec.py:363] Evaluating on the test split. +I0901 01:02:58.667820 140117123622080 submission_runner.py:516] Time since start: 56699.74s, Step: 187529, {'train/accuracy': Array(0.950275, dtype=float32), 'train/loss': Array(0.17458051, dtype=float32), 'validation/accuracy': Array(0.75412, dtype=float32), 'validation/loss': Array(1.0794177, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6301, dtype=float32), 'test/loss': Array(1.8626338, dtype=float32), 'test/num_examples': 10000, 'score': 55937.934849500656, 'total_duration': 56699.74064064026, 'accumulated_submission_time': 55937.934849500656, 'accumulated_eval_time': 755.1613154411316, 'accumulated_logging_time': 4.11983323097229} +I0901 01:02:59.162255 139921973044992 logging_writer.py:48] [187529] accumulated_eval_time=755.161, accumulated_logging_time=4.11983, accumulated_submission_time=55937.9, global_step=187529, preemption_count=0, score=55937.9, test/accuracy=0.6301000118255615, test/loss=1.8626338243484497, test/num_examples=10000, total_duration=56699.7, train/accuracy=0.9502750039100647, train/loss=0.17458051443099976, validation/accuracy=0.7541199922561646, validation/loss=1.0794177055358887, validation/num_examples=50000 +I0901 01:03:31.048881 139921981437696 logging_writer.py:48] [187600] global_step=187600, grad_norm=5.0552873611450195, loss=0.8531773090362549 +I0901 01:04:26.473747 139921973044992 logging_writer.py:48] [187700] global_step=187700, grad_norm=5.0441789627075195, loss=0.9491626620292664 +I0901 01:05:22.734743 139921981437696 logging_writer.py:48] [187800] global_step=187800, grad_norm=4.881083965301514, loss=0.8825333118438721 +I0901 01:06:17.784700 139921973044992 logging_writer.py:48] [187900] global_step=187900, grad_norm=5.276174068450928, loss=1.0084288120269775 +I0901 01:07:10.653075 139921981437696 logging_writer.py:48] [188000] global_step=188000, grad_norm=5.390972137451172, loss=0.9985030889511108 +I0901 01:08:04.931979 139921973044992 logging_writer.py:48] [188100] global_step=188100, grad_norm=5.200099945068359, loss=1.002005934715271 +I0901 01:09:00.006263 139921981437696 logging_writer.py:48] [188200] global_step=188200, grad_norm=5.166253566741943, loss=0.9224262237548828 +I0901 01:09:54.547541 139921973044992 logging_writer.py:48] [188300] global_step=188300, grad_norm=5.234634876251221, loss=0.9185755252838135 +I0901 01:10:47.165790 139921981437696 logging_writer.py:48] [188400] global_step=188400, grad_norm=5.326427936553955, loss=0.9326956272125244 +I0901 01:11:42.773521 139921973044992 logging_writer.py:48] [188500] global_step=188500, grad_norm=5.128489971160889, loss=0.8571536540985107 +I0901 01:12:34.626705 139921981437696 logging_writer.py:48] [188600] global_step=188600, grad_norm=5.2474493980407715, loss=0.98667973279953 +I0901 01:13:26.110328 139921973044992 logging_writer.py:48] [188700] global_step=188700, grad_norm=4.736712455749512, loss=0.8138905763626099 +I0901 01:14:18.096430 139921981437696 logging_writer.py:48] [188800] global_step=188800, grad_norm=5.008528232574463, loss=0.8929576873779297 +I0901 01:15:09.769220 139921973044992 logging_writer.py:48] [188900] global_step=188900, grad_norm=5.39980411529541, loss=0.9717090725898743 +I0901 01:16:05.374630 139921981437696 logging_writer.py:48] [189000] global_step=189000, grad_norm=5.026710033416748, loss=0.8971951007843018 +I0901 01:17:00.960404 139921973044992 logging_writer.py:48] [189100] global_step=189100, grad_norm=4.8860673904418945, loss=0.8346412181854248 +I0901 01:17:53.413143 139921981437696 logging_writer.py:48] [189200] global_step=189200, grad_norm=5.288318634033203, loss=0.9154367446899414 +I0901 01:18:43.858082 139921973044992 logging_writer.py:48] [189300] global_step=189300, grad_norm=5.06822395324707, loss=0.9055982232093811 +I0901 01:19:36.704804 139921981437696 logging_writer.py:48] [189400] global_step=189400, grad_norm=5.037821292877197, loss=0.880085825920105 +I0901 01:20:27.571392 139921973044992 logging_writer.py:48] [189500] global_step=189500, grad_norm=5.314779281616211, loss=0.9614152312278748 +I0901 01:21:17.397441 139921981437696 logging_writer.py:48] [189600] global_step=189600, grad_norm=4.8515777587890625, loss=0.9537127017974854 +I0901 01:22:08.126554 139921973044992 logging_writer.py:48] [189700] global_step=189700, grad_norm=5.221658229827881, loss=0.9315382838249207 +I0901 01:22:58.431252 139921981437696 logging_writer.py:48] [189800] global_step=189800, grad_norm=5.284745693206787, loss=1.0015546083450317 +I0901 01:24:30.239079 139921973044992 logging_writer.py:48] [189900] global_step=189900, grad_norm=5.157924175262451, loss=0.9224256277084351 +I0901 01:25:23.082668 139921981437696 logging_writer.py:48] [190000] global_step=190000, grad_norm=4.778889179229736, loss=0.888092577457428 +I0901 01:26:14.564085 139921973044992 logging_writer.py:48] [190100] global_step=190100, grad_norm=4.892291069030762, loss=0.9181605577468872 +I0901 01:27:08.934656 139921981437696 logging_writer.py:48] [190200] global_step=190200, grad_norm=4.901443958282471, loss=0.8633422255516052 +I0901 01:28:00.946086 139921973044992 logging_writer.py:48] [190300] global_step=190300, grad_norm=4.89575719833374, loss=0.9246034026145935 +I0901 01:28:52.790988 139921981437696 logging_writer.py:48] [190400] global_step=190400, grad_norm=4.997104644775391, loss=0.9447810649871826 +I0901 01:29:46.139717 139921973044992 logging_writer.py:48] [190500] global_step=190500, grad_norm=5.318854331970215, loss=0.9533683061599731 +I0901 01:30:37.706799 139921981437696 logging_writer.py:48] [190600] global_step=190600, grad_norm=4.854382038116455, loss=0.8871814608573914 +I0901 01:31:31.441587 139921973044992 logging_writer.py:48] [190700] global_step=190700, grad_norm=5.193007469177246, loss=0.9621533751487732 +I0901 01:32:21.884757 139921981437696 logging_writer.py:48] [190800] global_step=190800, grad_norm=5.226263999938965, loss=0.9242850542068481 +I0901 01:33:12.359827 139921973044992 logging_writer.py:48] [190900] global_step=190900, grad_norm=5.1542816162109375, loss=0.9804941415786743 +I0901 01:34:04.688625 139921981437696 logging_writer.py:48] [191000] global_step=191000, grad_norm=5.190145492553711, loss=0.956680417060852 +I0901 01:34:57.583064 139921973044992 logging_writer.py:48] [191100] global_step=191100, grad_norm=5.337347507476807, loss=0.9686191082000732 +I0901 01:35:49.485347 139921981437696 logging_writer.py:48] [191200] global_step=191200, grad_norm=5.126771926879883, loss=0.8767175078392029 +I0901 01:36:14.784466 140117123622080 spec.py:333] Evaluating on the training split. +I0901 01:36:23.840844 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 01:37:00.985019 140117123622080 spec.py:363] Evaluating on the test split. +I0901 01:37:02.092565 140117123622080 submission_runner.py:516] Time since start: 58743.22s, Step: 191248, {'train/accuracy': Array(0.9526666, dtype=float32), 'train/loss': Array(0.16960119, dtype=float32), 'validation/accuracy': Array(0.75324, dtype=float32), 'validation/loss': Array(1.0773498, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62750006, dtype=float32), 'test/loss': Array(1.8607931, dtype=float32), 'test/num_examples': 10000, 'score': 57933.47280240059, 'total_duration': 58743.22166013718, 'accumulated_submission_time': 57933.47280240059, 'accumulated_eval_time': 802.2567172050476, 'accumulated_logging_time': 4.655627489089966} +I0901 01:37:02.549392 139921973044992 logging_writer.py:48] [191248] accumulated_eval_time=802.257, accumulated_logging_time=4.65563, accumulated_submission_time=57933.5, global_step=191248, preemption_count=0, score=57933.5, test/accuracy=0.627500057220459, test/loss=1.860793113708496, test/num_examples=10000, total_duration=58743.2, train/accuracy=0.9526665806770325, train/loss=0.1696011871099472, validation/accuracy=0.7532399892807007, validation/loss=1.0773497819900513, validation/num_examples=50000 +I0901 01:37:24.228398 139921981437696 logging_writer.py:48] [191300] global_step=191300, grad_norm=4.965222358703613, loss=0.9512377977371216 +I0901 01:38:17.418574 139921973044992 logging_writer.py:48] [191400] global_step=191400, grad_norm=5.214223384857178, loss=0.9174201488494873 +I0901 01:39:12.246126 139921981437696 logging_writer.py:48] [191500] global_step=191500, grad_norm=5.250522613525391, loss=0.9576627612113953 +I0901 01:40:05.088099 139921973044992 logging_writer.py:48] [191600] global_step=191600, grad_norm=5.283293724060059, loss=0.9859845042228699 +I0901 01:40:58.027176 139921981437696 logging_writer.py:48] [191700] global_step=191700, grad_norm=5.149753570556641, loss=0.9414466619491577 +I0901 01:41:51.427153 139921973044992 logging_writer.py:48] [191800] global_step=191800, grad_norm=5.255005359649658, loss=0.9645184874534607 +I0901 01:42:45.233397 139921981437696 logging_writer.py:48] [191900] global_step=191900, grad_norm=5.094380855560303, loss=0.9526503682136536 +I0901 01:43:37.095499 139921973044992 logging_writer.py:48] [192000] global_step=192000, grad_norm=4.880415439605713, loss=0.9083675146102905 +I0901 01:44:28.529417 139921981437696 logging_writer.py:48] [192100] global_step=192100, grad_norm=4.588733196258545, loss=0.8435249328613281 +I0901 01:45:21.572044 139921973044992 logging_writer.py:48] [192200] global_step=192200, grad_norm=5.401425361633301, loss=1.003023386001587 +I0901 01:46:12.135546 139921981437696 logging_writer.py:48] [192300] global_step=192300, grad_norm=5.231875419616699, loss=0.9462592005729675 +I0901 01:47:03.087225 139921973044992 logging_writer.py:48] [192400] global_step=192400, grad_norm=4.936215877532959, loss=0.9570009708404541 +I0901 01:47:56.094242 139921981437696 logging_writer.py:48] [192500] global_step=192500, grad_norm=5.036055564880371, loss=0.9078521132469177 +I0901 01:48:49.636021 139921973044992 logging_writer.py:48] [192600] global_step=192600, grad_norm=5.109737873077393, loss=0.8998677730560303 +I0901 01:49:43.236303 139921981437696 logging_writer.py:48] [192700] global_step=192700, grad_norm=5.198482513427734, loss=0.953162670135498 +I0901 01:50:34.654071 139921973044992 logging_writer.py:48] [192800] global_step=192800, grad_norm=5.020749092102051, loss=0.925717830657959 +I0901 01:51:29.916932 139921981437696 logging_writer.py:48] [192900] global_step=192900, grad_norm=5.078593730926514, loss=0.8988884687423706 +I0901 01:52:22.856111 139921973044992 logging_writer.py:48] [193000] global_step=193000, grad_norm=5.165898323059082, loss=0.9427819848060608 +I0901 01:53:16.055649 139921981437696 logging_writer.py:48] [193100] global_step=193100, grad_norm=5.101173400878906, loss=0.8990031480789185 +I0901 01:54:08.646626 139921973044992 logging_writer.py:48] [193200] global_step=193200, grad_norm=5.036683082580566, loss=0.9112943410873413 +I0901 01:55:00.528355 139921981437696 logging_writer.py:48] [193300] global_step=193300, grad_norm=4.939996719360352, loss=0.9027379751205444 +I0901 01:55:51.927682 139921973044992 logging_writer.py:48] [193400] global_step=193400, grad_norm=5.021849155426025, loss=0.9198101758956909 +I0901 01:56:44.068458 139921981437696 logging_writer.py:48] [193500] global_step=193500, grad_norm=4.999680042266846, loss=0.8867872953414917 +I0901 01:57:35.746225 139921973044992 logging_writer.py:48] [193600] global_step=193600, grad_norm=5.057677745819092, loss=0.9064480066299438 +I0901 01:58:28.347533 139921981437696 logging_writer.py:48] [193700] global_step=193700, grad_norm=5.371496200561523, loss=0.8994989395141602 +I0901 01:59:21.092986 139921973044992 logging_writer.py:48] [193800] global_step=193800, grad_norm=4.836759567260742, loss=0.8914552927017212 +I0901 02:00:13.717348 139921981437696 logging_writer.py:48] [193900] global_step=193900, grad_norm=5.162883281707764, loss=0.896181583404541 +I0901 02:01:07.772658 139921973044992 logging_writer.py:48] [194000] global_step=194000, grad_norm=5.589412212371826, loss=1.0284178256988525 +I0901 02:02:01.709980 139921981437696 logging_writer.py:48] [194100] global_step=194100, grad_norm=5.44589900970459, loss=1.010350227355957 +I0901 02:02:54.295519 139921973044992 logging_writer.py:48] [194200] global_step=194200, grad_norm=4.789029121398926, loss=0.9262799024581909 +I0901 02:03:45.731281 139921981437696 logging_writer.py:48] [194300] global_step=194300, grad_norm=4.981706619262695, loss=0.9026138186454773 +I0901 02:04:36.785225 139921973044992 logging_writer.py:48] [194400] global_step=194400, grad_norm=5.011269569396973, loss=0.896692156791687 +I0901 02:05:27.995505 139921981437696 logging_writer.py:48] [194500] global_step=194500, grad_norm=5.18145227432251, loss=0.9410532712936401 +I0901 02:06:19.586856 139921973044992 logging_writer.py:48] [194600] global_step=194600, grad_norm=5.097621917724609, loss=0.9324361681938171 +I0901 02:07:12.263051 139921981437696 logging_writer.py:48] [194700] global_step=194700, grad_norm=5.332058906555176, loss=0.9835844039916992 +I0901 02:08:03.417804 139921973044992 logging_writer.py:48] [194800] global_step=194800, grad_norm=5.214462757110596, loss=0.9546920657157898 +I0901 02:08:55.138225 139921981437696 logging_writer.py:48] [194900] global_step=194900, grad_norm=4.8901238441467285, loss=0.8676942586898804 +I0901 02:09:47.070426 139921973044992 logging_writer.py:48] [195000] global_step=195000, grad_norm=5.327489376068115, loss=1.0178062915802002 +I0901 02:10:18.229437 140117123622080 spec.py:333] Evaluating on the training split. +I0901 02:10:28.128048 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 02:10:53.903997 140117123622080 spec.py:363] Evaluating on the test split. +I0901 02:10:55.008352 140117123622080 submission_runner.py:516] Time since start: 60776.14s, Step: 195063, {'train/accuracy': Array(0.952547, dtype=float32), 'train/loss': Array(0.17043896, dtype=float32), 'validation/accuracy': Array(0.75376, dtype=float32), 'validation/loss': Array(1.0782413, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62990004, dtype=float32), 'test/loss': Array(1.8617669, dtype=float32), 'test/num_examples': 10000, 'score': 59929.071627378464, 'total_duration': 60776.140051841736, 'accumulated_submission_time': 59929.071627378464, 'accumulated_eval_time': 838.8255090713501, 'accumulated_logging_time': 5.14893364906311} +I0901 02:10:55.513598 139921981437696 logging_writer.py:48] [195063] accumulated_eval_time=838.826, accumulated_logging_time=5.14893, accumulated_submission_time=59929.1, global_step=195063, preemption_count=0, score=59929.1, test/accuracy=0.6299000382423401, test/loss=1.8617669343948364, test/num_examples=10000, total_duration=60776.1, train/accuracy=0.952547013759613, train/loss=0.1704389601945877, validation/accuracy=0.7537599802017212, validation/loss=1.0782413482666016, validation/num_examples=50000 +I0901 02:11:09.974606 139921973044992 logging_writer.py:48] [195100] global_step=195100, grad_norm=4.962231636047363, loss=0.9323058128356934 +I0901 02:12:01.298951 139921981437696 logging_writer.py:48] [195200] global_step=195200, grad_norm=5.355375289916992, loss=1.0366965532302856 +I0901 02:12:53.375477 139921973044992 logging_writer.py:48] [195300] global_step=195300, grad_norm=5.147257328033447, loss=0.9325547814369202 +I0901 02:13:45.191291 139921981437696 logging_writer.py:48] [195400] global_step=195400, grad_norm=5.116206645965576, loss=0.9177793264389038 +I0901 02:14:35.097245 139921973044992 logging_writer.py:48] [195500] global_step=195500, grad_norm=4.9716315269470215, loss=0.901709794998169 +I0901 02:15:23.393985 139921981437696 logging_writer.py:48] [195600] global_step=195600, grad_norm=5.568150997161865, loss=0.9556702375411987 +I0901 02:16:13.965560 139921973044992 logging_writer.py:48] [195700] global_step=195700, grad_norm=4.887110710144043, loss=0.8821663856506348 +I0901 02:17:03.760162 139921981437696 logging_writer.py:48] [195800] global_step=195800, grad_norm=5.1680121421813965, loss=0.9795944094657898 +I0901 02:17:53.850679 139921973044992 logging_writer.py:48] [195900] global_step=195900, grad_norm=4.82769250869751, loss=0.9166573882102966 +I0901 02:18:43.318243 139921981437696 logging_writer.py:48] [196000] global_step=196000, grad_norm=4.985692501068115, loss=0.9455525279045105 +I0901 02:19:31.356673 139921973044992 logging_writer.py:48] [196100] global_step=196100, grad_norm=5.427358150482178, loss=0.939242959022522 +I0901 02:20:21.255217 139921981437696 logging_writer.py:48] [196200] global_step=196200, grad_norm=5.015539646148682, loss=0.903001606464386 +I0901 02:21:08.737277 139921973044992 logging_writer.py:48] [196300] global_step=196300, grad_norm=5.088765621185303, loss=0.8802263140678406 +I0901 02:21:57.734168 139921981437696 logging_writer.py:48] [196400] global_step=196400, grad_norm=5.157571792602539, loss=1.0119541883468628 +I0901 02:22:47.937780 139921973044992 logging_writer.py:48] [196500] global_step=196500, grad_norm=5.035524368286133, loss=0.8622300624847412 +I0901 02:23:36.406256 139921981437696 logging_writer.py:48] [196600] global_step=196600, grad_norm=5.196810722351074, loss=0.9526710510253906 +I0901 02:24:24.311015 139921973044992 logging_writer.py:48] [196700] global_step=196700, grad_norm=5.130920886993408, loss=0.8986252546310425 +I0901 02:25:13.479273 139921981437696 logging_writer.py:48] [196800] global_step=196800, grad_norm=5.287567138671875, loss=0.9231606721878052 +I0901 02:26:02.670312 139921973044992 logging_writer.py:48] [196900] global_step=196900, grad_norm=5.137923240661621, loss=0.8977866768836975 +I0901 02:26:49.747238 139921981437696 logging_writer.py:48] [197000] global_step=197000, grad_norm=4.969734191894531, loss=0.9156674146652222 +I0901 02:27:37.795878 139921973044992 logging_writer.py:48] [197100] global_step=197100, grad_norm=4.644207954406738, loss=0.8736706972122192 +I0901 02:28:29.002897 139921981437696 logging_writer.py:48] [197200] global_step=197200, grad_norm=5.357461452484131, loss=0.9754260778427124 +I0901 02:29:21.719686 139921973044992 logging_writer.py:48] [197300] global_step=197300, grad_norm=4.856748104095459, loss=0.8386725783348083 +I0901 02:30:11.476858 139921981437696 logging_writer.py:48] [197400] global_step=197400, grad_norm=4.912690162658691, loss=0.8917994499206543 +I0901 02:30:59.713081 139921973044992 logging_writer.py:48] [197500] global_step=197500, grad_norm=5.249939918518066, loss=0.9795364141464233 +I0901 02:31:48.810260 139921981437696 logging_writer.py:48] [197600] global_step=197600, grad_norm=5.090019226074219, loss=0.875196635723114 +I0901 02:32:39.807876 139921973044992 logging_writer.py:48] [197700] global_step=197700, grad_norm=5.377240180969238, loss=0.9122577905654907 +I0901 02:33:30.028276 139921981437696 logging_writer.py:48] [197800] global_step=197800, grad_norm=5.234877586364746, loss=0.9337693452835083 +I0901 02:34:18.058208 139921973044992 logging_writer.py:48] [197900] global_step=197900, grad_norm=5.2935590744018555, loss=1.0175368785858154 +I0901 02:35:03.789395 139921981437696 logging_writer.py:48] [198000] global_step=198000, grad_norm=5.093357086181641, loss=0.936771035194397 +I0901 02:35:50.834189 139921973044992 logging_writer.py:48] [198100] global_step=198100, grad_norm=5.023231506347656, loss=0.9631994962692261 +I0901 02:36:38.080610 139921981437696 logging_writer.py:48] [198200] global_step=198200, grad_norm=4.958556652069092, loss=0.9069747924804688 +I0901 02:37:22.574887 139921973044992 logging_writer.py:48] [198300] global_step=198300, grad_norm=5.046252250671387, loss=0.9585158228874207 +I0901 02:38:08.296819 139921981437696 logging_writer.py:48] [198400] global_step=198400, grad_norm=5.128335475921631, loss=0.9298807382583618 +I0901 02:38:59.299488 139921973044992 logging_writer.py:48] [198500] global_step=198500, grad_norm=5.211721420288086, loss=0.9638482928276062 +I0901 02:39:48.790218 139921981437696 logging_writer.py:48] [198600] global_step=198600, grad_norm=4.835295677185059, loss=0.905436635017395 +I0901 02:40:38.674258 139921973044992 logging_writer.py:48] [198700] global_step=198700, grad_norm=5.416763782501221, loss=0.9534271955490112 +I0901 02:41:31.032259 139921981437696 logging_writer.py:48] [198800] global_step=198800, grad_norm=4.994711875915527, loss=0.8898618817329407 +I0901 02:42:24.211749 139921973044992 logging_writer.py:48] [198900] global_step=198900, grad_norm=5.025669574737549, loss=0.8643783330917358 +I0901 02:43:19.190202 139921981437696 logging_writer.py:48] [199000] global_step=199000, grad_norm=4.953948497772217, loss=0.9254442453384399 +I0901 02:44:10.862057 140117123622080 spec.py:333] Evaluating on the training split. +I0901 02:44:20.644100 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 02:44:56.145433 140117123622080 spec.py:363] Evaluating on the test split. +I0901 02:44:57.279415 140117123622080 submission_runner.py:516] Time since start: 62818.38s, Step: 199096, {'train/accuracy': Array(0.95272636, dtype=float32), 'train/loss': Array(0.16754384, dtype=float32), 'validation/accuracy': Array(0.7528, dtype=float32), 'validation/loss': Array(1.078183, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62920004, dtype=float32), 'test/loss': Array(1.8614823, dtype=float32), 'test/num_examples': 10000, 'score': 61924.34237098694, 'total_duration': 62818.38122630119, 'accumulated_submission_time': 61924.34237098694, 'accumulated_eval_time': 885.0028548240662, 'accumulated_logging_time': 5.6843955516815186} +I0901 02:44:57.779118 139921973044992 logging_writer.py:48] [199096] accumulated_eval_time=885.003, accumulated_logging_time=5.6844, accumulated_submission_time=61924.3, global_step=199096, preemption_count=0, score=61924.3, test/accuracy=0.6292000412940979, test/loss=1.8614822626113892, test/num_examples=10000, total_duration=62818.4, train/accuracy=0.9527263641357422, train/loss=0.16754384338855743, validation/accuracy=0.7527999877929688, validation/loss=1.0781830549240112, validation/num_examples=50000 +I0901 02:44:59.194190 139921981437696 logging_writer.py:48] [199100] global_step=199100, grad_norm=4.854145050048828, loss=0.859321653842926 +I0901 02:45:46.969334 139921973044992 logging_writer.py:48] [199200] global_step=199200, grad_norm=5.089237213134766, loss=0.9018992185592651 +I0901 02:46:39.638283 139921981437696 logging_writer.py:48] [199300] global_step=199300, grad_norm=5.23795223236084, loss=0.9640636444091797 +I0901 02:47:33.028141 139921973044992 logging_writer.py:48] [199400] global_step=199400, grad_norm=4.986812591552734, loss=0.8238881230354309 +I0901 02:48:27.184502 139921981437696 logging_writer.py:48] [199500] global_step=199500, grad_norm=5.274386405944824, loss=0.9509237408638 +I0901 02:49:19.819572 139921973044992 logging_writer.py:48] [199600] global_step=199600, grad_norm=5.455211639404297, loss=0.9987109899520874 +I0901 02:50:12.595806 139921981437696 logging_writer.py:48] [199700] global_step=199700, grad_norm=5.5727458000183105, loss=0.9921669960021973 +I0901 02:51:09.677159 139921973044992 logging_writer.py:48] [199800] global_step=199800, grad_norm=5.080610752105713, loss=0.8806011080741882 +I0901 02:52:05.609390 139921981437696 logging_writer.py:48] [199900] global_step=199900, grad_norm=5.198128700256348, loss=0.9039404988288879 +I0901 02:53:03.700288 139921973044992 logging_writer.py:48] [200000] global_step=200000, grad_norm=4.808097839355469, loss=0.8420426845550537 +I0901 02:54:01.080299 139921981437696 logging_writer.py:48] [200100] global_step=200100, grad_norm=5.227399826049805, loss=0.8521677851676941 +I0901 02:55:00.848187 139921973044992 logging_writer.py:48] [200200] global_step=200200, grad_norm=4.977720737457275, loss=0.9349397420883179 +I0901 02:55:56.363418 139921981437696 logging_writer.py:48] [200300] global_step=200300, grad_norm=5.19509744644165, loss=0.9569917917251587 +I0901 02:56:52.878339 139921973044992 logging_writer.py:48] [200400] global_step=200400, grad_norm=4.92441987991333, loss=0.908467173576355 +I0901 02:57:46.113277 139921981437696 logging_writer.py:48] [200500] global_step=200500, grad_norm=5.271442413330078, loss=0.9538626670837402 +I0901 02:58:37.599506 139921973044992 logging_writer.py:48] [200600] global_step=200600, grad_norm=5.081338405609131, loss=0.9512836933135986 +I0901 02:59:30.615411 139921981437696 logging_writer.py:48] [200700] global_step=200700, grad_norm=4.992593765258789, loss=0.8736499547958374 +I0901 03:00:22.289846 139921973044992 logging_writer.py:48] [200800] global_step=200800, grad_norm=4.961019515991211, loss=0.9846526384353638 +I0901 03:01:14.717590 139921981437696 logging_writer.py:48] [200900] global_step=200900, grad_norm=4.981587886810303, loss=0.8737311363220215 +I0901 03:02:06.278768 139921973044992 logging_writer.py:48] [201000] global_step=201000, grad_norm=4.794268608093262, loss=0.8757073879241943 +I0901 03:02:57.674258 139921981437696 logging_writer.py:48] [201100] global_step=201100, grad_norm=4.9438629150390625, loss=0.9370888471603394 +I0901 03:03:49.692721 139921973044992 logging_writer.py:48] [201200] global_step=201200, grad_norm=5.2072014808654785, loss=0.9215210676193237 +I0901 03:04:41.853460 139921981437696 logging_writer.py:48] [201300] global_step=201300, grad_norm=5.327813148498535, loss=1.0010875463485718 +I0901 03:05:35.143156 139921973044992 logging_writer.py:48] [201400] global_step=201400, grad_norm=4.959412097930908, loss=0.8991104364395142 +I0901 03:06:28.189674 139921981437696 logging_writer.py:48] [201500] global_step=201500, grad_norm=5.096189022064209, loss=0.9464887380599976 +I0901 03:07:18.340505 139921973044992 logging_writer.py:48] [201600] global_step=201600, grad_norm=5.152906894683838, loss=0.9395279288291931 +I0901 03:08:11.856630 139921981437696 logging_writer.py:48] [201700] global_step=201700, grad_norm=4.800223350524902, loss=0.8782333135604858 +I0901 03:09:05.545010 139921973044992 logging_writer.py:48] [201800] global_step=201800, grad_norm=5.436853408813477, loss=0.9561561346054077 +I0901 03:09:58.731142 139921981437696 logging_writer.py:48] [201900] global_step=201900, grad_norm=5.081823348999023, loss=0.8984829187393188 +I0901 03:10:50.760534 139921973044992 logging_writer.py:48] [202000] global_step=202000, grad_norm=5.0473952293396, loss=0.8901992440223694 +I0901 03:11:42.116919 139921981437696 logging_writer.py:48] [202100] global_step=202100, grad_norm=5.043544292449951, loss=0.9282650947570801 +I0901 03:12:32.517793 139921973044992 logging_writer.py:48] [202200] global_step=202200, grad_norm=5.1276068687438965, loss=0.9304385185241699 +I0901 03:13:22.944142 139921981437696 logging_writer.py:48] [202300] global_step=202300, grad_norm=4.6825175285339355, loss=0.8543671369552612 +I0901 03:14:15.851419 139921973044992 logging_writer.py:48] [202400] global_step=202400, grad_norm=5.577588081359863, loss=0.9244983792304993 +I0901 03:15:07.560393 139921981437696 logging_writer.py:48] [202500] global_step=202500, grad_norm=5.269078254699707, loss=0.8966765999794006 +I0901 03:15:59.912837 139921973044992 logging_writer.py:48] [202600] global_step=202600, grad_norm=5.241589546203613, loss=0.9816130995750427 +I0901 03:16:55.945526 139921981437696 logging_writer.py:48] [202700] global_step=202700, grad_norm=5.06398868560791, loss=0.8747780919075012 +I0901 03:17:48.090250 139921973044992 logging_writer.py:48] [202800] global_step=202800, grad_norm=4.878236770629883, loss=0.82566237449646 +I0901 03:18:13.433534 140117123622080 spec.py:333] Evaluating on the training split. +I0901 03:18:22.168696 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 03:18:50.475713 140117123622080 spec.py:363] Evaluating on the test split. +I0901 03:18:51.631839 140117123622080 submission_runner.py:516] Time since start: 64852.71s, Step: 202849, {'train/accuracy': Array(0.9549585, dtype=float32), 'train/loss': Array(0.15947141, dtype=float32), 'validation/accuracy': Array(0.75386, dtype=float32), 'validation/loss': Array(1.0782359, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.62990004, dtype=float32), 'test/loss': Array(1.8620335, dtype=float32), 'test/num_examples': 10000, 'score': 63919.883076667786, 'total_duration': 64852.70789504051, 'accumulated_submission_time': 63919.883076667786, 'accumulated_eval_time': 922.9353837966919, 'accumulated_logging_time': 6.243690252304077} +I0901 03:18:52.048099 139921981437696 logging_writer.py:48] [202849] accumulated_eval_time=922.935, accumulated_logging_time=6.24369, accumulated_submission_time=63919.9, global_step=202849, preemption_count=0, score=63919.9, test/accuracy=0.6299000382423401, test/loss=1.862033486366272, test/num_examples=10000, total_duration=64852.7, train/accuracy=0.9549584984779358, train/loss=0.15947140753269196, validation/accuracy=0.7538599967956543, validation/loss=1.0782358646392822, validation/num_examples=50000 +I0901 03:19:13.379885 139921973044992 logging_writer.py:48] [202900] global_step=202900, grad_norm=5.142001152038574, loss=1.0034464597702026 +I0901 03:20:04.308067 139921981437696 logging_writer.py:48] [203000] global_step=203000, grad_norm=5.087765216827393, loss=0.9284964203834534 +I0901 03:20:55.790158 139921973044992 logging_writer.py:48] [203100] global_step=203100, grad_norm=5.133492469787598, loss=0.9225966930389404 +I0901 03:21:48.814392 139921981437696 logging_writer.py:48] [203200] global_step=203200, grad_norm=4.966383457183838, loss=0.9408687353134155 +I0901 03:22:41.410773 139921973044992 logging_writer.py:48] [203300] global_step=203300, grad_norm=5.141142845153809, loss=0.8789050579071045 +I0901 03:23:32.975736 139921981437696 logging_writer.py:48] [203400] global_step=203400, grad_norm=5.227304458618164, loss=0.9915010929107666 +I0901 03:24:24.918745 139921973044992 logging_writer.py:48] [203500] global_step=203500, grad_norm=4.931765079498291, loss=0.8931645154953003 +I0901 03:25:16.512228 139921981437696 logging_writer.py:48] [203600] global_step=203600, grad_norm=4.994811058044434, loss=0.9263933300971985 +I0901 03:26:06.918324 139921973044992 logging_writer.py:48] [203700] global_step=203700, grad_norm=5.153566360473633, loss=0.8871217370033264 +I0901 03:26:59.221386 139921981437696 logging_writer.py:48] [203800] global_step=203800, grad_norm=5.4974565505981445, loss=1.0241471529006958 +I0901 03:27:50.347560 139921973044992 logging_writer.py:48] [203900] global_step=203900, grad_norm=5.024063587188721, loss=0.943141520023346 +I0901 03:28:44.751027 139921981437696 logging_writer.py:48] [204000] global_step=204000, grad_norm=5.485108852386475, loss=1.0129190683364868 +I0901 03:29:35.367842 139921973044992 logging_writer.py:48] [204100] global_step=204100, grad_norm=4.938610076904297, loss=0.9177238941192627 +I0901 03:30:26.949898 139921981437696 logging_writer.py:48] [204200] global_step=204200, grad_norm=5.197873115539551, loss=0.855902910232544 +I0901 03:31:19.963916 139921973044992 logging_writer.py:48] [204300] global_step=204300, grad_norm=4.938788890838623, loss=0.8780633211135864 +I0901 03:32:12.997174 139921981437696 logging_writer.py:48] [204400] global_step=204400, grad_norm=5.16373872756958, loss=0.8628950119018555 +I0901 03:33:03.203662 139921973044992 logging_writer.py:48] [204500] global_step=204500, grad_norm=5.19012451171875, loss=0.9914536476135254 +I0901 03:33:53.742640 139921981437696 logging_writer.py:48] [204600] global_step=204600, grad_norm=4.852474212646484, loss=0.8646405339241028 +I0901 03:34:46.189434 139921973044992 logging_writer.py:48] [204700] global_step=204700, grad_norm=5.131824970245361, loss=0.910761833190918 +I0901 03:35:35.398241 139921981437696 logging_writer.py:48] [204800] global_step=204800, grad_norm=5.081014633178711, loss=0.9338768720626831 +I0901 03:36:27.425194 139921973044992 logging_writer.py:48] [204900] global_step=204900, grad_norm=5.310151100158691, loss=0.9057902693748474 +I0901 03:37:19.602013 139921981437696 logging_writer.py:48] [205000] global_step=205000, grad_norm=4.9742021560668945, loss=0.855685293674469 +I0901 03:38:10.250192 139921973044992 logging_writer.py:48] [205100] global_step=205100, grad_norm=5.082778453826904, loss=0.8521609306335449 +I0901 03:39:02.668349 139921981437696 logging_writer.py:48] [205200] global_step=205200, grad_norm=5.022729873657227, loss=0.849153995513916 +I0901 03:39:52.714936 139921973044992 logging_writer.py:48] [205300] global_step=205300, grad_norm=5.30662202835083, loss=0.9881296157836914 +I0901 03:40:44.900432 139921981437696 logging_writer.py:48] [205400] global_step=205400, grad_norm=5.740540027618408, loss=1.0138365030288696 +I0901 03:41:38.015750 139921973044992 logging_writer.py:48] [205500] global_step=205500, grad_norm=5.424652099609375, loss=1.0562061071395874 +I0901 03:42:29.776018 139921981437696 logging_writer.py:48] [205600] global_step=205600, grad_norm=5.0391154289245605, loss=0.8735479116439819 +I0901 03:43:20.968653 139921973044992 logging_writer.py:48] [205700] global_step=205700, grad_norm=4.934131622314453, loss=0.8910331726074219 +I0901 03:44:11.798535 139921981437696 logging_writer.py:48] [205800] global_step=205800, grad_norm=4.737799644470215, loss=0.8327043056488037 +I0901 03:45:00.570166 139921973044992 logging_writer.py:48] [205900] global_step=205900, grad_norm=5.240020275115967, loss=0.9855616688728333 +I0901 03:45:49.874439 139921981437696 logging_writer.py:48] [206000] global_step=206000, grad_norm=5.215758323669434, loss=0.9970656633377075 +I0901 03:46:39.591671 139921973044992 logging_writer.py:48] [206100] global_step=206100, grad_norm=5.039385795593262, loss=0.7976914048194885 +I0901 03:47:29.685153 139921981437696 logging_writer.py:48] [206200] global_step=206200, grad_norm=4.991382598876953, loss=0.9156073331832886 +I0901 03:48:19.039177 139921973044992 logging_writer.py:48] [206300] global_step=206300, grad_norm=4.636569023132324, loss=0.8123666048049927 +I0901 03:49:08.481129 139921981437696 logging_writer.py:48] [206400] global_step=206400, grad_norm=5.1354594230651855, loss=0.967815637588501 +I0901 03:49:59.253669 139921973044992 logging_writer.py:48] [206500] global_step=206500, grad_norm=5.37056827545166, loss=0.9465696215629578 +I0901 03:50:49.545297 139921981437696 logging_writer.py:48] [206600] global_step=206600, grad_norm=5.5642313957214355, loss=1.0446619987487793 +I0901 03:51:41.427504 139921973044992 logging_writer.py:48] [206700] global_step=206700, grad_norm=4.826991558074951, loss=0.8509694933891296 +I0901 03:52:07.605539 140117123622080 spec.py:333] Evaluating on the training split. +I0901 03:52:18.121912 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 03:52:39.711435 140117123622080 spec.py:363] Evaluating on the test split. +I0901 03:52:40.791242 140117123622080 submission_runner.py:516] Time since start: 66881.94s, Step: 206754, {'train/accuracy': Array(0.9528858, dtype=float32), 'train/loss': Array(0.16299346, dtype=float32), 'validation/accuracy': Array(0.75408, dtype=float32), 'validation/loss': Array(1.0789212, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.629, dtype=float32), 'test/loss': Array(1.8630679, dtype=float32), 'test/num_examples': 10000, 'score': 65915.35211348534, 'total_duration': 66881.94443774223, 'accumulated_submission_time': 65915.35211348534, 'accumulated_eval_time': 955.9324560165405, 'accumulated_logging_time': 6.702232837677002} +I0901 03:52:41.289181 139921981437696 logging_writer.py:48] [206754] accumulated_eval_time=955.932, accumulated_logging_time=6.70223, accumulated_submission_time=65915.4, global_step=206754, preemption_count=0, score=65915.4, test/accuracy=0.6290000081062317, test/loss=1.863067865371704, test/num_examples=10000, total_duration=66881.9, train/accuracy=0.9528858065605164, train/loss=0.16299346089363098, validation/accuracy=0.7540799975395203, validation/loss=1.0789211988449097, validation/num_examples=50000 +I0901 03:52:58.312817 139921973044992 logging_writer.py:48] [206800] global_step=206800, grad_norm=5.094330787658691, loss=0.8758286237716675 +I0901 03:53:48.493738 139921981437696 logging_writer.py:48] [206900] global_step=206900, grad_norm=5.102592468261719, loss=0.9203047156333923 +I0901 03:54:39.537535 139921973044992 logging_writer.py:48] [207000] global_step=207000, grad_norm=5.244966506958008, loss=0.9406728744506836 +I0901 03:55:28.573068 139921981437696 logging_writer.py:48] [207100] global_step=207100, grad_norm=5.275124549865723, loss=0.8883792757987976 +I0901 03:56:19.056848 139921973044992 logging_writer.py:48] [207200] global_step=207200, grad_norm=5.062265872955322, loss=0.8934289813041687 +I0901 03:57:08.751048 139921981437696 logging_writer.py:48] [207300] global_step=207300, grad_norm=4.986675262451172, loss=0.8484734892845154 +I0901 03:57:57.188901 139921973044992 logging_writer.py:48] [207400] global_step=207400, grad_norm=5.077184677124023, loss=0.9095171689987183 +I0901 03:58:47.225306 139921981437696 logging_writer.py:48] [207500] global_step=207500, grad_norm=5.179933547973633, loss=0.9289082288742065 +I0901 03:59:37.140421 139921973044992 logging_writer.py:48] [207600] global_step=207600, grad_norm=5.29863166809082, loss=0.9422008395195007 +I0901 04:00:26.740118 139921981437696 logging_writer.py:48] [207700] global_step=207700, grad_norm=5.251464366912842, loss=0.9219319820404053 +I0901 04:01:16.944646 139921973044992 logging_writer.py:48] [207800] global_step=207800, grad_norm=4.6855926513671875, loss=0.8318419456481934 +I0901 04:02:06.936469 139921981437696 logging_writer.py:48] [207900] global_step=207900, grad_norm=4.9897003173828125, loss=0.9507322311401367 +I0901 04:02:55.379743 139921973044992 logging_writer.py:48] [208000] global_step=208000, grad_norm=5.17241096496582, loss=0.8975096344947815 +I0901 04:03:42.229098 139921981437696 logging_writer.py:48] [208100] global_step=208100, grad_norm=5.391748428344727, loss=0.995088517665863 +I0901 04:04:32.189775 139921973044992 logging_writer.py:48] [208200] global_step=208200, grad_norm=5.198868274688721, loss=0.879770040512085 +I0901 04:05:21.101728 139921981437696 logging_writer.py:48] [208300] global_step=208300, grad_norm=5.233975410461426, loss=0.9455336928367615 +I0901 04:06:10.929483 139921973044992 logging_writer.py:48] [208400] global_step=208400, grad_norm=5.496593952178955, loss=0.9707695841789246 +I0901 04:06:59.734660 139921981437696 logging_writer.py:48] [208500] global_step=208500, grad_norm=4.978848457336426, loss=0.8897007703781128 +I0901 04:07:50.025379 139921973044992 logging_writer.py:48] [208600] global_step=208600, grad_norm=4.978503227233887, loss=0.8678510189056396 +I0901 04:08:40.087639 139921981437696 logging_writer.py:48] [208700] global_step=208700, grad_norm=5.045394420623779, loss=0.887261688709259 +I0901 04:09:28.547085 139921973044992 logging_writer.py:48] [208800] global_step=208800, grad_norm=5.402185440063477, loss=0.9732722640037537 +I0901 04:10:16.197018 139921981437696 logging_writer.py:48] [208900] global_step=208900, grad_norm=5.222142219543457, loss=0.9825078845024109 +I0901 04:11:06.903146 139921973044992 logging_writer.py:48] [209000] global_step=209000, grad_norm=5.459487438201904, loss=0.9618760347366333 +I0901 04:11:57.181123 139921981437696 logging_writer.py:48] [209100] global_step=209100, grad_norm=5.219305992126465, loss=0.9284542798995972 +I0901 04:12:49.172547 139921973044992 logging_writer.py:48] [209200] global_step=209200, grad_norm=5.160443305969238, loss=0.9863014221191406 +I0901 04:13:38.323202 139921981437696 logging_writer.py:48] [209300] global_step=209300, grad_norm=5.289532661437988, loss=1.007180094718933 +I0901 04:14:28.226255 139921973044992 logging_writer.py:48] [209400] global_step=209400, grad_norm=5.095505237579346, loss=0.9366816282272339 +I0901 04:15:19.582709 139921981437696 logging_writer.py:48] [209500] global_step=209500, grad_norm=5.150091648101807, loss=0.9685860276222229 +I0901 04:16:08.088450 139921973044992 logging_writer.py:48] [209600] global_step=209600, grad_norm=5.289364337921143, loss=0.9760082960128784 +I0901 04:16:57.685092 139921981437696 logging_writer.py:48] [209700] global_step=209700, grad_norm=5.325281143188477, loss=0.9625385403633118 +I0901 04:17:47.713113 139921973044992 logging_writer.py:48] [209800] global_step=209800, grad_norm=5.133907794952393, loss=0.9025450944900513 +I0901 04:18:36.650622 139921981437696 logging_writer.py:48] [209900] global_step=209900, grad_norm=5.200686931610107, loss=0.862160325050354 +I0901 04:19:25.528101 139921973044992 logging_writer.py:48] [210000] global_step=210000, grad_norm=5.0692644119262695, loss=0.9429908990859985 +I0901 04:20:11.494866 139921981437696 logging_writer.py:48] [210100] global_step=210100, grad_norm=5.548826694488525, loss=0.9137258529663086 +I0901 04:21:00.616386 139921973044992 logging_writer.py:48] [210200] global_step=210200, grad_norm=5.013029098510742, loss=0.8738904595375061 +I0901 04:21:47.690350 139921981437696 logging_writer.py:48] [210300] global_step=210300, grad_norm=4.992095947265625, loss=0.9276241660118103 +I0901 04:22:35.217660 139921973044992 logging_writer.py:48] [210400] global_step=210400, grad_norm=5.023890018463135, loss=0.8646029829978943 +I0901 04:23:23.607188 139921981437696 logging_writer.py:48] [210500] global_step=210500, grad_norm=4.9644293785095215, loss=0.8720778226852417 +I0901 04:24:09.493642 139921973044992 logging_writer.py:48] [210600] global_step=210600, grad_norm=5.156888008117676, loss=0.9716874361038208 +I0901 04:24:56.879058 139921981437696 logging_writer.py:48] [210700] global_step=210700, grad_norm=5.001516342163086, loss=0.9196484684944153 +I0901 04:25:43.094339 139921973044992 logging_writer.py:48] [210800] global_step=210800, grad_norm=5.292616844177246, loss=0.9546539783477783 +I0901 04:25:56.720865 140117123622080 spec.py:333] Evaluating on the training split. +I0901 04:26:05.505845 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 04:26:35.908888 140117123622080 spec.py:363] Evaluating on the test split. +I0901 04:26:36.978410 140117123622080 submission_runner.py:516] Time since start: 68918.14s, Step: 210830, {'train/accuracy': Array(0.9546795, dtype=float32), 'train/loss': Array(0.16126078, dtype=float32), 'validation/accuracy': Array(0.75395995, dtype=float32), 'validation/loss': Array(1.0778822, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6302, dtype=float32), 'test/loss': Array(1.8621116, dtype=float32), 'test/num_examples': 10000, 'score': 67910.6963338852, 'total_duration': 68918.14360046387, 'accumulated_submission_time': 67910.6963338852, 'accumulated_eval_time': 996.0133640766144, 'accumulated_logging_time': 7.245025157928467} +I0901 04:26:37.505176 139921981437696 logging_writer.py:48] [210830] accumulated_eval_time=996.013, accumulated_logging_time=7.24503, accumulated_submission_time=67910.7, global_step=210830, preemption_count=0, score=67910.7, test/accuracy=0.6302000284194946, test/loss=1.8621115684509277, test/num_examples=10000, total_duration=68918.1, train/accuracy=0.9546794891357422, train/loss=0.16126078367233276, validation/accuracy=0.7539599537849426, validation/loss=1.077882170677185, validation/num_examples=50000 +I0901 04:27:04.748293 139921973044992 logging_writer.py:48] [210900] global_step=210900, grad_norm=5.153116226196289, loss=0.8613507151603699 +I0901 04:27:50.575899 139921981437696 logging_writer.py:48] [211000] global_step=211000, grad_norm=5.315664768218994, loss=0.9884841442108154 +I0901 04:28:38.127283 139921973044992 logging_writer.py:48] [211100] global_step=211100, grad_norm=5.348547458648682, loss=0.9610521793365479 +I0901 04:29:24.156081 139921981437696 logging_writer.py:48] [211200] global_step=211200, grad_norm=5.133537769317627, loss=0.928541898727417 +I0901 04:30:09.637515 139921973044992 logging_writer.py:48] [211300] global_step=211300, grad_norm=4.98984956741333, loss=0.8715340495109558 +I0901 04:30:57.871524 139921981437696 logging_writer.py:48] [211400] global_step=211400, grad_norm=5.457026958465576, loss=0.954578161239624 +I0901 04:31:47.550090 139921973044992 logging_writer.py:48] [211500] global_step=211500, grad_norm=5.10288143157959, loss=0.9043671488761902 +I0901 04:32:34.127412 139921981437696 logging_writer.py:48] [211600] global_step=211600, grad_norm=4.988109111785889, loss=0.8917821049690247 +I0901 04:33:22.135970 139921973044992 logging_writer.py:48] [211700] global_step=211700, grad_norm=5.135811805725098, loss=0.9228299856185913 +I0901 04:34:11.039785 139921981437696 logging_writer.py:48] [211800] global_step=211800, grad_norm=5.723660945892334, loss=0.9441190958023071 +I0901 04:35:00.645313 139921973044992 logging_writer.py:48] [211900] global_step=211900, grad_norm=5.128077983856201, loss=0.8854506015777588 +I0901 04:35:47.506135 139921981437696 logging_writer.py:48] [212000] global_step=212000, grad_norm=4.993853569030762, loss=0.9233067035675049 +I0901 04:36:33.932691 139921973044992 logging_writer.py:48] [212100] global_step=212100, grad_norm=5.114875793457031, loss=0.8868685960769653 +I0901 04:37:20.803683 139921981437696 logging_writer.py:48] [212200] global_step=212200, grad_norm=4.984254360198975, loss=0.9018019437789917 +I0901 04:38:06.031008 139921973044992 logging_writer.py:48] [212300] global_step=212300, grad_norm=5.001010894775391, loss=0.874775767326355 +I0901 04:38:52.582712 139921981437696 logging_writer.py:48] [212400] global_step=212400, grad_norm=4.961604118347168, loss=0.8020590543746948 +I0901 04:39:38.552688 139921973044992 logging_writer.py:48] [212500] global_step=212500, grad_norm=5.018125057220459, loss=0.9557503461837769 +I0901 04:40:23.841095 139921981437696 logging_writer.py:48] [212600] global_step=212600, grad_norm=4.893603324890137, loss=0.8804634213447571 +I0901 04:41:12.472225 139921973044992 logging_writer.py:48] [212700] global_step=212700, grad_norm=5.377978801727295, loss=0.9779345989227295 +I0901 04:41:57.533085 139921981437696 logging_writer.py:48] [212800] global_step=212800, grad_norm=5.202916145324707, loss=0.9589000940322876 +I0901 04:42:43.749400 139921973044992 logging_writer.py:48] [212900] global_step=212900, grad_norm=5.011430263519287, loss=0.9092808365821838 +I0901 04:43:30.002455 139921981437696 logging_writer.py:48] [213000] global_step=213000, grad_norm=5.072605133056641, loss=0.9081613421440125 +I0901 04:44:14.980884 139921973044992 logging_writer.py:48] [213100] global_step=213100, grad_norm=4.945305824279785, loss=0.8509699106216431 +I0901 04:45:00.253943 139921981437696 logging_writer.py:48] [213200] global_step=213200, grad_norm=5.066500186920166, loss=0.8990758657455444 +I0901 04:45:46.605411 139921973044992 logging_writer.py:48] [213300] global_step=213300, grad_norm=5.114422798156738, loss=0.8816932439804077 +I0901 04:46:29.565409 139921981437696 logging_writer.py:48] [213400] global_step=213400, grad_norm=5.010619640350342, loss=0.871833860874176 +I0901 04:47:14.459747 139921973044992 logging_writer.py:48] [213500] global_step=213500, grad_norm=5.043308258056641, loss=0.8404016494750977 +I0901 04:47:59.075810 139921981437696 logging_writer.py:48] [213600] global_step=213600, grad_norm=5.099091529846191, loss=0.877414882183075 +I0901 04:48:42.906985 139921973044992 logging_writer.py:48] [213700] global_step=213700, grad_norm=4.742823600769043, loss=0.8831362724304199 +I0901 04:49:25.518404 139921981437696 logging_writer.py:48] [213800] global_step=213800, grad_norm=5.0153117179870605, loss=0.9237356185913086 +I0901 04:50:09.611935 139921973044992 logging_writer.py:48] [213900] global_step=213900, grad_norm=5.275447368621826, loss=0.93492591381073 +I0901 04:50:53.761101 139921981437696 logging_writer.py:48] [214000] global_step=214000, grad_norm=5.365634441375732, loss=1.0174566507339478 +I0901 04:51:37.222412 139921973044992 logging_writer.py:48] [214100] global_step=214100, grad_norm=4.815036296844482, loss=0.8611651062965393 +I0901 04:52:21.577841 139921981437696 logging_writer.py:48] [214200] global_step=214200, grad_norm=5.113801956176758, loss=0.8977988362312317 +I0901 04:53:04.008725 139921973044992 logging_writer.py:48] [214300] global_step=214300, grad_norm=5.307579517364502, loss=0.9361207485198975 +I0901 04:53:46.808907 139921981437696 logging_writer.py:48] [214400] global_step=214400, grad_norm=5.079486846923828, loss=0.8900383710861206 +I0901 04:54:30.006645 139921973044992 logging_writer.py:48] [214500] global_step=214500, grad_norm=5.225406646728516, loss=0.9650312662124634 +I0901 04:55:11.449355 139921981437696 logging_writer.py:48] [214600] global_step=214600, grad_norm=5.327311992645264, loss=1.002485990524292 +I0901 04:55:52.758100 139921973044992 logging_writer.py:48] [214700] global_step=214700, grad_norm=5.308073997497559, loss=0.9312092065811157 +I0901 04:56:34.673197 139921981437696 logging_writer.py:48] [214800] global_step=214800, grad_norm=4.910130500793457, loss=0.894026517868042 +I0901 04:57:14.268891 139921973044992 logging_writer.py:48] [214900] global_step=214900, grad_norm=5.282629013061523, loss=0.8717041611671448 +I0901 04:57:55.310437 139921981437696 logging_writer.py:48] [215000] global_step=215000, grad_norm=5.073805809020996, loss=0.8984159231185913 +I0901 04:58:36.471621 139921973044992 logging_writer.py:48] [215100] global_step=215100, grad_norm=5.038486480712891, loss=0.8802030086517334 +I0901 04:59:20.095498 139921981437696 logging_writer.py:48] [215200] global_step=215200, grad_norm=5.722534656524658, loss=1.04356050491333 +I0901 04:59:53.114450 140117123622080 spec.py:333] Evaluating on the training split. +I0901 05:00:02.184293 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 05:00:33.022626 140117123622080 spec.py:363] Evaluating on the test split. +I0901 05:00:34.161932 140117123622080 submission_runner.py:516] Time since start: 70955.26s, Step: 215277, {'train/accuracy': Array(0.9546795, dtype=float32), 'train/loss': Array(0.15797712, dtype=float32), 'validation/accuracy': Array(0.75408, dtype=float32), 'validation/loss': Array(1.0774853, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.63310003, dtype=float32), 'test/loss': Array(1.8600335, dtype=float32), 'test/num_examples': 10000, 'score': 69906.23444628716, 'total_duration': 70955.25825119019, 'accumulated_submission_time': 69906.23444628716, 'accumulated_eval_time': 1036.815337896347, 'accumulated_logging_time': 7.803174257278442} +I0901 05:00:34.584003 139921973044992 logging_writer.py:48] [215277] accumulated_eval_time=1036.82, accumulated_logging_time=7.80317, accumulated_submission_time=69906.2, global_step=215277, preemption_count=0, score=69906.2, test/accuracy=0.6331000328063965, test/loss=1.8600335121154785, test/num_examples=10000, total_duration=70955.3, train/accuracy=0.9546794891357422, train/loss=0.1579771190881729, validation/accuracy=0.7540799975395203, validation/loss=1.0774853229522705, validation/num_examples=50000 +I0901 05:00:41.165721 139921981437696 logging_writer.py:48] [215300] global_step=215300, grad_norm=5.209726810455322, loss=0.9245810508728027 +I0901 05:01:23.349060 139921973044992 logging_writer.py:48] [215400] global_step=215400, grad_norm=5.404288291931152, loss=0.9539520144462585 +I0901 05:02:06.593559 139921981437696 logging_writer.py:48] [215500] global_step=215500, grad_norm=5.296565055847168, loss=0.8851534128189087 +I0901 05:02:47.897917 139921973044992 logging_writer.py:48] [215600] global_step=215600, grad_norm=5.240959167480469, loss=0.9085546135902405 +I0901 05:03:30.917989 139921981437696 logging_writer.py:48] [215700] global_step=215700, grad_norm=5.009771823883057, loss=0.8803013563156128 +I0901 05:04:13.660840 139921973044992 logging_writer.py:48] [215800] global_step=215800, grad_norm=4.938933849334717, loss=0.870991051197052 +I0901 05:04:56.411501 139921981437696 logging_writer.py:48] [215900] global_step=215900, grad_norm=4.885268688201904, loss=0.89155113697052 +I0901 05:05:37.250045 139921973044992 logging_writer.py:48] [216000] global_step=216000, grad_norm=4.846563816070557, loss=0.7862259149551392 +I0901 05:06:19.300134 139921981437696 logging_writer.py:48] [216100] global_step=216100, grad_norm=5.349638938903809, loss=0.9574717283248901 +I0901 05:07:01.140778 139921973044992 logging_writer.py:48] [216200] global_step=216200, grad_norm=5.270731449127197, loss=0.95424485206604 +I0901 05:07:43.844132 139921981437696 logging_writer.py:48] [216300] global_step=216300, grad_norm=5.266517639160156, loss=0.9599326848983765 +I0901 05:08:27.670740 139921973044992 logging_writer.py:48] [216400] global_step=216400, grad_norm=5.193632125854492, loss=0.8684572577476501 +I0901 05:09:11.918064 139921981437696 logging_writer.py:48] [216500] global_step=216500, grad_norm=4.908528804779053, loss=0.8023777008056641 +I0901 05:09:55.098418 139921973044992 logging_writer.py:48] [216600] global_step=216600, grad_norm=5.337294101715088, loss=0.9144799709320068 +I0901 05:10:38.028306 139921981437696 logging_writer.py:48] [216700] global_step=216700, grad_norm=5.105943202972412, loss=0.8863859176635742 +I0901 05:11:21.334683 139921973044992 logging_writer.py:48] [216800] global_step=216800, grad_norm=5.027052402496338, loss=0.920903205871582 +I0901 05:12:03.547258 139921981437696 logging_writer.py:48] [216900] global_step=216900, grad_norm=5.284718036651611, loss=0.9968641996383667 +I0901 05:12:46.490614 139921973044992 logging_writer.py:48] [217000] global_step=217000, grad_norm=5.220391750335693, loss=0.9313672184944153 +I0901 05:13:28.652569 139921981437696 logging_writer.py:48] [217100] global_step=217100, grad_norm=4.980034351348877, loss=0.8858731985092163 +I0901 05:14:12.150613 139921973044992 logging_writer.py:48] [217200] global_step=217200, grad_norm=5.039654731750488, loss=0.9210734963417053 +I0901 05:14:55.116343 139921981437696 logging_writer.py:48] [217300] global_step=217300, grad_norm=5.334781646728516, loss=0.9748035669326782 +I0901 05:15:36.388195 139921973044992 logging_writer.py:48] [217400] global_step=217400, grad_norm=5.16457462310791, loss=0.9219173192977905 +I0901 05:16:19.063783 139921981437696 logging_writer.py:48] [217500] global_step=217500, grad_norm=4.841170787811279, loss=0.8522967100143433 +I0901 05:17:02.017367 139921973044992 logging_writer.py:48] [217600] global_step=217600, grad_norm=5.153195381164551, loss=0.9608635306358337 +I0901 05:17:46.989598 139921981437696 logging_writer.py:48] [217700] global_step=217700, grad_norm=5.029423713684082, loss=0.8846449851989746 +I0901 05:18:29.592622 139921973044992 logging_writer.py:48] [217800] global_step=217800, grad_norm=4.802762508392334, loss=0.7918198108673096 +I0901 05:19:10.966532 139921981437696 logging_writer.py:48] [217900] global_step=217900, grad_norm=5.22661828994751, loss=0.8689782619476318 +I0901 05:19:52.261264 139921973044992 logging_writer.py:48] [218000] global_step=218000, grad_norm=5.332197666168213, loss=0.9813300371170044 +I0901 05:20:34.368326 139921981437696 logging_writer.py:48] [218100] global_step=218100, grad_norm=5.283938884735107, loss=0.9606820344924927 +I0901 05:21:17.100504 139921973044992 logging_writer.py:48] [218200] global_step=218200, grad_norm=5.034923076629639, loss=0.8902592062950134 +I0901 05:21:58.637321 139921981437696 logging_writer.py:48] [218300] global_step=218300, grad_norm=4.824457168579102, loss=0.8664211630821228 +I0901 05:22:40.408616 139921973044992 logging_writer.py:48] [218400] global_step=218400, grad_norm=4.872227668762207, loss=0.915947437286377 +I0901 05:23:17.568780 139921981437696 logging_writer.py:48] [218500] global_step=218500, grad_norm=5.425260543823242, loss=0.9741702079772949 +I0901 05:23:50.053129 139921973044992 logging_writer.py:48] [218600] global_step=218600, grad_norm=5.214247703552246, loss=0.9263602495193481 +I0901 05:24:22.807626 139921981437696 logging_writer.py:48] [218700] global_step=218700, grad_norm=5.211658000946045, loss=0.9429059624671936 +I0901 05:24:55.488836 139921973044992 logging_writer.py:48] [218800] global_step=218800, grad_norm=5.136645793914795, loss=0.9234890341758728 +I0901 05:25:28.068797 139921981437696 logging_writer.py:48] [218900] global_step=218900, grad_norm=5.319687366485596, loss=0.9385421872138977 +I0901 05:26:01.470860 139921973044992 logging_writer.py:48] [219000] global_step=219000, grad_norm=5.11089563369751, loss=0.8951600193977356 +I0901 05:26:34.367853 139921981437696 logging_writer.py:48] [219100] global_step=219100, grad_norm=5.02082633972168, loss=0.9241605401039124 +I0901 05:27:07.692725 139921973044992 logging_writer.py:48] [219200] global_step=219200, grad_norm=5.103062152862549, loss=0.9026141166687012 +I0901 05:27:40.676816 139921981437696 logging_writer.py:48] [219300] global_step=219300, grad_norm=5.183253288269043, loss=0.888795018196106 +I0901 05:28:13.237639 139921973044992 logging_writer.py:48] [219400] global_step=219400, grad_norm=5.213377475738525, loss=0.8955112099647522 +I0901 05:28:45.870908 139921981437696 logging_writer.py:48] [219500] global_step=219500, grad_norm=4.849860191345215, loss=0.8460988998413086 +I0901 05:29:18.845524 139921973044992 logging_writer.py:48] [219600] global_step=219600, grad_norm=5.014862060546875, loss=0.9206774234771729 +I0901 05:29:51.699751 139921981437696 logging_writer.py:48] [219700] global_step=219700, grad_norm=4.95686674118042, loss=0.8829386830329895 +I0901 05:30:24.895580 139921973044992 logging_writer.py:48] [219800] global_step=219800, grad_norm=5.321438312530518, loss=0.9632787108421326 +I0901 05:30:57.589187 139921981437696 logging_writer.py:48] [219900] global_step=219900, grad_norm=5.416116714477539, loss=0.9298134446144104 +I0901 05:31:30.753657 139921973044992 logging_writer.py:48] [220000] global_step=220000, grad_norm=4.997304916381836, loss=0.8993109464645386 +I0901 05:32:03.754537 139921981437696 logging_writer.py:48] [220100] global_step=220100, grad_norm=5.542360305786133, loss=0.9184138774871826 +I0901 05:32:39.507214 139921973044992 logging_writer.py:48] [220200] global_step=220200, grad_norm=5.029839992523193, loss=0.9027119874954224 +I0901 05:33:15.323236 139921981437696 logging_writer.py:48] [220300] global_step=220300, grad_norm=5.060340404510498, loss=0.911401629447937 +I0901 05:33:49.959213 140117123622080 spec.py:333] Evaluating on the training split. +I0901 05:33:59.012092 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 05:34:27.607859 140117123622080 spec.py:363] Evaluating on the test split. +I0901 05:34:28.681519 140117123622080 submission_runner.py:516] Time since start: 72989.84s, Step: 220397, {'train/accuracy': Array(0.95649314, dtype=float32), 'train/loss': Array(0.15317032, dtype=float32), 'validation/accuracy': Array(0.75404, dtype=float32), 'validation/loss': Array(1.080798, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.6341, dtype=float32), 'test/loss': Array(1.863171, dtype=float32), 'test/num_examples': 10000, 'score': 71901.50822019577, 'total_duration': 72989.8373169899, 'accumulated_submission_time': 71901.50822019577, 'accumulated_eval_time': 1075.351613998413, 'accumulated_logging_time': 8.280278444290161} +I0901 05:34:29.137010 139921973044992 logging_writer.py:48] [220397] accumulated_eval_time=1075.35, accumulated_logging_time=8.28028, accumulated_submission_time=71901.5, global_step=220397, preemption_count=0, score=71901.5, test/accuracy=0.6341000199317932, test/loss=1.8631709814071655, test/num_examples=10000, total_duration=72989.8, train/accuracy=0.9564931392669678, train/loss=0.15317031741142273, validation/accuracy=0.754040002822876, validation/loss=1.0807980298995972, validation/num_examples=50000 +I0901 05:34:30.320527 139921981437696 logging_writer.py:48] [220400] global_step=220400, grad_norm=5.227998733520508, loss=1.001722812652588 +I0901 05:35:00.806415 139921973044992 logging_writer.py:48] [220500] global_step=220500, grad_norm=5.122833728790283, loss=0.8808613419532776 +I0901 05:35:33.037385 139921981437696 logging_writer.py:48] [220600] global_step=220600, grad_norm=5.226979732513428, loss=0.9029439091682434 +I0901 05:36:12.726135 139921973044992 logging_writer.py:48] [220700] global_step=220700, grad_norm=5.215989589691162, loss=0.9803972244262695 +I0901 05:36:55.875290 139921981437696 logging_writer.py:48] [220800] global_step=220800, grad_norm=4.980408191680908, loss=0.9081791043281555 +I0901 05:37:30.385524 139921973044992 logging_writer.py:48] [220900] global_step=220900, grad_norm=5.0105438232421875, loss=0.9124945402145386 +I0901 05:38:04.561352 139921981437696 logging_writer.py:48] [221000] global_step=221000, grad_norm=4.8925580978393555, loss=0.8812861442565918 +I0901 05:38:38.552753 139921973044992 logging_writer.py:48] [221100] global_step=221100, grad_norm=5.167718410491943, loss=0.9017781019210815 +I0901 05:39:11.408661 139921981437696 logging_writer.py:48] [221200] global_step=221200, grad_norm=5.5954718589782715, loss=1.028396487236023 +I0901 05:39:44.674027 139921973044992 logging_writer.py:48] [221300] global_step=221300, grad_norm=5.13559627532959, loss=0.9290332198143005 +I0901 05:40:18.239992 139921981437696 logging_writer.py:48] [221400] global_step=221400, grad_norm=4.8653082847595215, loss=0.8863214254379272 +I0901 05:40:52.343083 139921973044992 logging_writer.py:48] [221500] global_step=221500, grad_norm=4.96597957611084, loss=0.8650926947593689 +I0901 05:41:24.922185 139921981437696 logging_writer.py:48] [221600] global_step=221600, grad_norm=5.050286769866943, loss=0.8762205839157104 +I0901 05:41:58.115599 139921973044992 logging_writer.py:48] [221700] global_step=221700, grad_norm=5.270134925842285, loss=0.872877299785614 +I0901 05:42:30.891085 139921981437696 logging_writer.py:48] [221800] global_step=221800, grad_norm=4.951303958892822, loss=0.8943935632705688 +I0901 05:43:03.957434 139921973044992 logging_writer.py:48] [221900] global_step=221900, grad_norm=5.104475021362305, loss=0.8794708251953125 +I0901 05:43:37.074986 139921981437696 logging_writer.py:48] [222000] global_step=222000, grad_norm=5.27532434463501, loss=0.9702441692352295 +I0901 05:44:10.108567 139921973044992 logging_writer.py:48] [222100] global_step=222100, grad_norm=5.118463516235352, loss=0.892654538154602 +I0901 05:44:43.465187 139921981437696 logging_writer.py:48] [222200] global_step=222200, grad_norm=4.997800827026367, loss=0.9249782562255859 +I0901 05:45:16.542014 139921973044992 logging_writer.py:48] [222300] global_step=222300, grad_norm=4.979686260223389, loss=0.8402267694473267 +I0901 05:45:49.462439 139921981437696 logging_writer.py:48] [222400] global_step=222400, grad_norm=5.282010078430176, loss=0.9425011873245239 +I0901 05:46:22.748912 139921973044992 logging_writer.py:48] [222500] global_step=222500, grad_norm=4.8539509773254395, loss=0.896287202835083 +I0901 05:46:55.259054 139921981437696 logging_writer.py:48] [222600] global_step=222600, grad_norm=5.307825565338135, loss=0.9276976585388184 +I0901 05:47:28.995963 139921973044992 logging_writer.py:48] [222700] global_step=222700, grad_norm=5.4053521156311035, loss=0.9499086737632751 +I0901 05:48:01.526663 139921981437696 logging_writer.py:48] [222800] global_step=222800, grad_norm=4.743131160736084, loss=0.87527996301651 +I0901 05:48:34.484347 139921973044992 logging_writer.py:48] [222900] global_step=222900, grad_norm=5.093332767486572, loss=0.9245432615280151 +I0901 05:49:07.033991 139921981437696 logging_writer.py:48] [223000] global_step=223000, grad_norm=5.076138496398926, loss=0.8587645888328552 +I0901 05:49:39.648826 139921973044992 logging_writer.py:48] [223100] global_step=223100, grad_norm=5.060522079467773, loss=0.9442175626754761 +I0901 05:50:12.195954 139921981437696 logging_writer.py:48] [223200] global_step=223200, grad_norm=5.331088066101074, loss=0.9304803609848022 +I0901 05:50:44.922413 139921973044992 logging_writer.py:48] [223300] global_step=223300, grad_norm=4.628348350524902, loss=0.8376320600509644 +I0901 05:51:17.169647 139921981437696 logging_writer.py:48] [223400] global_step=223400, grad_norm=5.115884780883789, loss=0.9621231555938721 +I0901 05:51:49.531630 139921973044992 logging_writer.py:48] [223500] global_step=223500, grad_norm=4.840067386627197, loss=0.8676809072494507 +I0901 05:52:22.092365 139921981437696 logging_writer.py:48] [223600] global_step=223600, grad_norm=5.200618743896484, loss=0.922809362411499 +I0901 05:52:54.440253 139921973044992 logging_writer.py:48] [223700] global_step=223700, grad_norm=5.385339260101318, loss=0.9958111047744751 +I0901 05:53:27.428703 139921981437696 logging_writer.py:48] [223800] global_step=223800, grad_norm=5.034536361694336, loss=0.9147318601608276 +I0901 05:54:00.167588 139921973044992 logging_writer.py:48] [223900] global_step=223900, grad_norm=5.315379619598389, loss=0.9217160940170288 +I0901 05:54:33.676805 139921981437696 logging_writer.py:48] [224000] global_step=224000, grad_norm=4.949201583862305, loss=0.8892548084259033 +I0901 05:55:05.922954 139921973044992 logging_writer.py:48] [224100] global_step=224100, grad_norm=5.261760234832764, loss=0.8990651369094849 +I0901 05:55:38.637033 139921981437696 logging_writer.py:48] [224200] global_step=224200, grad_norm=5.170938014984131, loss=0.9093736410140991 +I0901 05:56:10.971346 139921973044992 logging_writer.py:48] [224300] global_step=224300, grad_norm=5.045515537261963, loss=0.9224607944488525 +I0901 05:56:43.602962 139921981437696 logging_writer.py:48] [224400] global_step=224400, grad_norm=4.950613498687744, loss=0.8751481771469116 +I0901 05:57:16.658598 139921973044992 logging_writer.py:48] [224500] global_step=224500, grad_norm=5.239254474639893, loss=1.0115783214569092 +I0901 05:57:48.816482 139921981437696 logging_writer.py:48] [224600] global_step=224600, grad_norm=4.98504114151001, loss=0.8933776617050171 +I0901 05:58:21.240139 139921973044992 logging_writer.py:48] [224700] global_step=224700, grad_norm=5.155628204345703, loss=0.9398466348648071 +I0901 05:58:54.180905 139921981437696 logging_writer.py:48] [224800] global_step=224800, grad_norm=5.004520893096924, loss=0.9245262145996094 +I0901 05:59:27.053357 139921973044992 logging_writer.py:48] [224900] global_step=224900, grad_norm=4.967422962188721, loss=0.8815662264823914 +I0901 05:59:59.755025 139921981437696 logging_writer.py:48] [225000] global_step=225000, grad_norm=5.359360218048096, loss=0.8634929060935974 +I0901 06:00:32.136827 139921973044992 logging_writer.py:48] [225100] global_step=225100, grad_norm=5.344351768493652, loss=0.9146206974983215 +I0901 06:01:05.046123 139921981437696 logging_writer.py:48] [225200] global_step=225200, grad_norm=4.944432258605957, loss=0.8850175142288208 +I0901 06:01:37.888939 139921973044992 logging_writer.py:48] [225300] global_step=225300, grad_norm=4.997131824493408, loss=0.8535712957382202 +I0901 06:02:10.788561 139921981437696 logging_writer.py:48] [225400] global_step=225400, grad_norm=5.1853742599487305, loss=0.9141325950622559 +I0901 06:02:43.543237 139921973044992 logging_writer.py:48] [225500] global_step=225500, grad_norm=5.451545238494873, loss=0.9966766238212585 +I0901 06:03:16.426238 139921981437696 logging_writer.py:48] [225600] global_step=225600, grad_norm=5.127327919006348, loss=0.9304978847503662 +I0901 06:03:49.164386 139921973044992 logging_writer.py:48] [225700] global_step=225700, grad_norm=5.021611213684082, loss=0.7990228533744812 +I0901 06:04:21.917794 139921981437696 logging_writer.py:48] [225800] global_step=225800, grad_norm=5.151325702667236, loss=0.9331804513931274 +I0901 06:04:54.347917 139921973044992 logging_writer.py:48] [225900] global_step=225900, grad_norm=4.946349620819092, loss=0.8601771593093872 +I0901 06:05:26.891231 139921981437696 logging_writer.py:48] [226000] global_step=226000, grad_norm=5.186084747314453, loss=0.9174197912216187 +I0901 06:05:59.311913 139921973044992 logging_writer.py:48] [226100] global_step=226100, grad_norm=5.30026912689209, loss=0.9434544444084167 +I0901 06:06:32.575624 139921981437696 logging_writer.py:48] [226200] global_step=226200, grad_norm=5.316706657409668, loss=0.8915654420852661 +I0901 06:07:07.789540 139921973044992 logging_writer.py:48] [226300] global_step=226300, grad_norm=5.076333999633789, loss=0.8932842016220093 +I0901 06:07:43.415657 139921981437696 logging_writer.py:48] [226400] global_step=226400, grad_norm=5.329573154449463, loss=0.9848429560661316 +I0901 06:07:44.845485 140117123622080 spec.py:333] Evaluating on the training split. +I0901 06:07:54.405077 140117123622080 spec.py:346] Evaluating on the validation split. +I0901 06:08:20.136599 140117123622080 spec.py:363] Evaluating on the test split. +I0901 06:08:21.334306 140117123622080 submission_runner.py:516] Time since start: 75022.37s, Step: 226406, {'train/accuracy': Array(0.95523757, dtype=float32), 'train/loss': Array(0.15633641, dtype=float32), 'validation/accuracy': Array(0.75372, dtype=float32), 'validation/loss': Array(1.0817096, dtype=float32), 'validation/num_examples': 50000, 'test/accuracy': Array(0.63430005, dtype=float32), 'test/loss': Array(1.8659685, dtype=float32), 'test/num_examples': 10000, 'score': 73897.11191749573, 'total_duration': 75022.37075853348, 'accumulated_submission_time': 73897.11191749573, 'accumulated_eval_time': 1111.5350549221039, 'accumulated_logging_time': 8.781720638275146} +I0901 06:08:21.766742 139921973044992 logging_writer.py:48] [226406] accumulated_eval_time=1111.54, accumulated_logging_time=8.78172, accumulated_submission_time=73897.1, global_step=226406, preemption_count=0, score=73897.1, test/accuracy=0.6343000531196594, test/loss=1.8659684658050537, test/num_examples=10000, total_duration=75022.4, train/accuracy=0.9552375674247742, train/loss=0.15633641183376312, validation/accuracy=0.7537199854850769, validation/loss=1.081709623336792, validation/num_examples=50000 +I0901 06:08:49.630580 139921981437696 logging_writer.py:48] [226500] global_step=226500, grad_norm=4.998298168182373, loss=0.8191386461257935 +I0901 06:09:21.311222 139921973044992 logging_writer.py:48] [226600] global_step=226600, grad_norm=5.275415420532227, loss=0.9888912439346313 +I0901 06:09:53.241525 139921981437696 logging_writer.py:48] [226700] global_step=226700, grad_norm=5.114013671875, loss=0.8066397905349731 +I0901 06:10:25.546577 139921973044992 logging_writer.py:48] [226800] global_step=226800, grad_norm=5.544271469116211, loss=0.8944303393363953 +I0901 06:10:59.377294 139921981437696 logging_writer.py:48] [226900] global_step=226900, grad_norm=5.243582725524902, loss=0.8191415071487427 +I0901 06:11:34.209624 139921973044992 logging_writer.py:48] [227000] global_step=227000, grad_norm=5.234365463256836, loss=0.8590567111968994 +I0901 06:12:07.990851 139921981437696 logging_writer.py:48] [227100] global_step=227100, grad_norm=5.308478355407715, loss=0.929723858833313 +I0901 06:12:42.226760 139921973044992 logging_writer.py:48] [227200] global_step=227200, grad_norm=5.261415481567383, loss=0.908229649066925 +I0901 06:13:16.528399 139921981437696 logging_writer.py:48] [227300] global_step=227300, grad_norm=4.932633399963379, loss=0.9209676384925842 +I0901 06:13:50.283041 139921973044992 logging_writer.py:48] [227400] global_step=227400, grad_norm=5.036497116088867, loss=0.872220516204834 +I0901 06:14:24.411114 139921981437696 logging_writer.py:48] [227500] global_step=227500, grad_norm=5.268780708312988, loss=0.8036066293716431 +I0901 06:14:58.439174 139921973044992 logging_writer.py:48] [227600] global_step=227600, grad_norm=5.1295928955078125, loss=0.9697645902633667 +I0901 06:15:33.926703 139921981437696 logging_writer.py:48] [227700] global_step=227700, grad_norm=4.784746170043945, loss=0.8296281695365906 +I0901 06:16:09.558137 139921973044992 logging_writer.py:48] [227800] global_step=227800, grad_norm=5.042296886444092, loss=0.8889498114585876 +I0901 06:16:44.029170 139921981437696 logging_writer.py:48] [227900] global_step=227900, grad_norm=5.198847770690918, loss=0.878498911857605 +I0901 06:17:18.258201 139921973044992 logging_writer.py:48] [228000] global_step=228000, grad_norm=5.0690741539001465, loss=0.8548619151115417 +I0901 06:17:52.013542 139921981437696 logging_writer.py:48] [228100] global_step=228100, grad_norm=5.341606616973877, loss=0.9951151609420776 +I0901 06:18:26.943684 139921973044992 logging_writer.py:48] [228200] global_step=228200, grad_norm=5.152012825012207, loss=0.8384189009666443 +I0901 06:19:00.484176 139921981437696 logging_writer.py:48] [228300] global_step=228300, grad_norm=5.269826889038086, loss=1.0044865608215332 +I0901 06:19:34.559245 139921973044992 logging_writer.py:48] [228400] global_step=228400, grad_norm=5.242280960083008, loss=0.9437057971954346 +I0901 06:20:07.930946 139921981437696 logging_writer.py:48] [228500] global_step=228500, grad_norm=5.353850841522217, loss=0.9914513230323792 +I0901 06:20:41.809294 139921973044992 logging_writer.py:48] [228600] global_step=228600, grad_norm=5.195721626281738, loss=0.9520765542984009 +I0901 06:21:15.749988 139921981437696 logging_writer.py:48] [228700] global_step=228700, grad_norm=5.550253868103027, loss=0.9426811337471008 +I0901 06:21:50.011523 139921973044992 logging_writer.py:48] [228800] global_step=228800, grad_norm=5.287452697753906, loss=0.945716142654419 +I0901 06:22:24.648372 139921981437696 logging_writer.py:48] [228900] global_step=228900, grad_norm=5.613103866577148, loss=1.0503196716308594 +I0901 06:22:59.728841 139921973044992 logging_writer.py:48] [229000] global_step=229000, grad_norm=5.046504497528076, loss=0.9111925959587097 +I0901 06:23:33.351124 139921981437696 logging_writer.py:48] [229100] global_step=229100, grad_norm=5.127466678619385, loss=0.9772840738296509 +I0901 06:24:07.108847 139921973044992 logging_writer.py:48] [229200] global_step=229200, grad_norm=5.161699295043945, loss=0.9445910453796387 +I0901 06:24:41.010252 139921981437696 logging_writer.py:48] [229300] global_step=229300, grad_norm=5.541453838348389, loss=0.9734501838684082 +I0901 06:25:14.514416 139921973044992 logging_writer.py:48] [229400] global_step=229400, grad_norm=4.963949680328369, loss=0.85980623960495 +I0901 06:25:48.672599 139921981437696 logging_writer.py:48] [229500] global_step=229500, grad_norm=4.632791519165039, loss=0.7670915722846985 +I0901 06:26:22.443649 139921973044992 logging_writer.py:48] [229600] global_step=229600, grad_norm=5.357014179229736, loss=0.9371649026870728 +I0901 06:26:55.964395 139921981437696 logging_writer.py:48] [229700] global_step=229700, grad_norm=5.409812927246094, loss=0.9799932241439819 +I0901 06:27:30.365059 139921973044992 logging_writer.py:48] [229800] global_step=229800, grad_norm=4.934323787689209, loss=0.9098383784294128 +I0901 06:28:04.199832 139921981437696 logging_writer.py:48] [229900] global_step=229900, grad_norm=5.275177955627441, loss=0.8990069031715393 +I0901 06:28:38.983226 139921973044992 logging_writer.py:48] [230000] global_step=230000, grad_norm=5.252416610717773, loss=0.9163826704025269 +I0901 06:29:14.117155 139921981437696 logging_writer.py:48] [230100] global_step=230100, grad_norm=5.542680740356445, loss=0.9463514089584351 +I0901 06:29:49.230474 139921973044992 logging_writer.py:48] [230200] global_step=230200, grad_norm=4.813201427459717, loss=0.8499897718429565 +I0901 06:30:24.843950 139921981437696 logging_writer.py:48] [230300] global_step=230300, grad_norm=5.253739356994629, loss=0.9262945055961609 +I0901 06:31:00.542137 139921973044992 logging_writer.py:48] [230400] global_step=230400, grad_norm=5.518209457397461, loss=0.9405405521392822 +I0901 06:31:34.637993 139921981437696 logging_writer.py:48] [230500] global_step=230500, grad_norm=5.37162446975708, loss=0.9204640984535217 +I0901 06:32:08.320124 139921973044992 logging_writer.py:48] [230600] global_step=230600, grad_norm=5.3649091720581055, loss=0.9387766122817993 +I0901 06:32:42.859477 139921981437696 logging_writer.py:48] [230700] global_step=230700, grad_norm=4.926331996917725, loss=0.9076877236366272 +I0901 06:33:16.335926 139921973044992 logging_writer.py:48] [230800] global_step=230800, grad_norm=5.255847454071045, loss=0.936400294303894 +I0901 06:33:50.130678 139921981437696 logging_writer.py:48] [230900] global_step=230900, grad_norm=4.9094390869140625, loss=0.8191436529159546 +I0901 06:34:24.389266 139921973044992 logging_writer.py:48] [231000] global_step=231000, grad_norm=5.118624687194824, loss=0.8749614953994751 +I0901 06:34:58.338091 139921981437696 logging_writer.py:48] [231100] global_step=231100, grad_norm=5.494625568389893, loss=1.076176643371582 +I0901 06:35:32.252051 139921973044992 logging_writer.py:48] [231200] global_step=231200, grad_norm=4.9834160804748535, loss=0.8773630261421204 +I0901 06:36:07.433776 139921981437696 logging_writer.py:48] [231300] global_step=231300, grad_norm=4.956263065338135, loss=0.8738206624984741 +I0901 06:36:42.627259 139921973044992 logging_writer.py:48] [231400] global_step=231400, grad_norm=5.057019233703613, loss=0.9189537763595581 +I0901 06:37:18.576854 139921981437696 logging_writer.py:48] [231500] global_step=231500, grad_norm=5.453633785247803, loss=0.9052627086639404 +I0901 06:37:53.031754 139921973044992 logging_writer.py:48] [231600] global_step=231600, grad_norm=5.307099342346191, loss=0.9718407988548279 +I0901 06:38:26.941447 139921981437696 logging_writer.py:48] [231700] global_step=231700, grad_norm=5.044972896575928, loss=0.8993996381759644 +I0901 06:39:00.883882 139921973044992 logging_writer.py:48] [231800] global_step=231800, grad_norm=5.386070251464844, loss=0.855185866355896 +I0901 06:39:35.093232 139921981437696 logging_writer.py:48] [231900] global_step=231900, grad_norm=5.1513142585754395, loss=0.8947547674179077 +I0901 06:40:12.333906 139921973044992 logging_writer.py:48] [232000] global_step=232000, grad_norm=5.191963195800781, loss=0.875298798084259 +I0901 06:40:53.843374 139921981437696 logging_writer.py:48] [232100] global_step=232100, grad_norm=5.2257843017578125, loss=0.8918358087539673 +I0901 06:41:35.361425 139921973044992 logging_writer.py:48] [232200] global_step=232200, grad_norm=4.804821968078613, loss=0.8751260042190552 +I0901 06:41:37.252260 139921981437696 logging_writer.py:48] [232206] global_step=232206, preemption_count=0, score=75892.4 +I0901 06:41:38.306414 140117123622080 submission_runner.py:857] Final imagenet_resnet score: 75892.38308477402 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/eval_measurements.csv b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/eval_measurements.csv index dd38dd520..94eaaab51 100644 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/eval_measurements.csv +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/eval_measurements.csv @@ -1,39 +1,39 @@ accumulated_eval_time,accumulated_logging_time,accumulated_submission_time,global_step,preemption_count,score,test/accuracy,test/loss,test/num_examples,total_duration,train/accuracy,train/loss,validation/accuracy,validation/loss,validation/num_examples -58.0493426322937,0.0,55.403265714645386,1,0,55.403265714645386,0.00080000004,6.912909,10000,113.45354461669922,0.0007573342,6.9125247,0.00068,6.91273,50000 -80.8735921382904,0.0282888412475585,2051.405519247055,5091,0,2051.405519247055,0.025600001,6.2957478,10000,2132.3507702350616,0.03683036,6.0105863,0.03464,6.073887,50000 -104.81943488121031,0.0968890190124511,4047.7705268859854,10483,0,4047.7705268859854,0.0092,7.565653,10000,4152.775046825409,0.011838329,7.4746013,0.01248,7.483133,50000 -128.34734845161438,0.155996561050415,6044.017327547073,15877,0,6044.017327547073,0.0046,8.634853,10000,6172.654530525208,0.004803093,8.611359,0.00446,8.60309,50000 -153.84260439872742,0.2327916622161865,8040.030968904495,21268,0,8040.030968904495,0.0025000002,9.078246,10000,8194.285801887512,0.0022919322,9.054386,0.00214,9.063879,50000 -177.0388777256012,0.2861685752868652,10036.045359134674,26661,0,10036.045359134674,0.0015,9.203677,10000,10213.59507727623,0.0014748087,9.214938,0.00156,9.203479,50000 -200.61437606811523,0.3487226963043213,12032.575147390366,32053,0,12032.575147390366,0.0014000001,8.848667,10000,12233.807253837584,0.001434949,8.892026,0.00172,8.847008,50000 -223.4686014652252,0.3977727890014648,14028.56815481186,37430,0,14028.56815481186,0.0015,8.609212,10000,14252.74825811386,0.0015345982,8.616239,0.0015199999,8.605657,50000 -245.62259697914124,0.481757640838623,16024.628784179688,42810,0,16024.628784179688,0.0013000001,8.466099,10000,16271.091526985168,0.0016143175,8.523104,0.0016399999,8.457493,50000 -267.5842914581299,0.5713906288146973,18021.00621008873,48190,0,18021.00621008873,0.0015,8.331606,10000,18289.564914941788,0.0017737563,8.355035,0.0013799999,8.319946,50000 -289.2678849697113,0.6422865390777588,20017.06850886345,53567,0,20017.06850886345,0.0011,8.248967,10000,20307.42605495453,0.0015744579,8.254096,0.0013799999,8.238354,50000 -310.8434724807739,0.6898555755615234,22013.158291101456,58943,0,22013.158291101456,0.0014000001,8.245119,10000,22325.183495998383,0.0011360012,8.292785,0.0013799999,8.236893,50000 -332.6784863471985,0.7396748065948486,24009.70991158485,64321,0,24009.70991158485,0.0012,8.277351,10000,24343.664563417435,0.0013751594,8.30454,0.00134,8.2691765,50000 -354.2806496620178,0.8288896083831787,26005.58596253395,69699,0,26005.58596253395,0.0013000001,8.307022,10000,26361.27652549744,0.0016541772,8.343106,0.0013799999,8.300176,50000 -375.2546329498291,0.8931126594543457,28001.82784414292,75079,0,28001.82784414292,0.0015,8.351302,10000,28378.600981235504,0.0012755102,8.378688,0.0015,8.342849,50000 -396.9780945777893,0.9690923690795898,29998.31781888008,80458,0,29998.31781888008,0.0017000001,8.418403,10000,30396.9351439476,0.0012555803,8.448164,0.00148,8.408929,50000 -417.75972962379456,1.046638011932373,31994.878091335297,85840,0,31994.878091335297,0.0019,8.485265,10000,32414.399368047714,0.0013950892,8.513527,0.0014599999,8.475016,50000 -439.3616163730621,1.124169111251831,33991.13047719002,91222,0,33991.13047719002,0.0017000001,8.573786,10000,34432.37569451332,0.0014548788,8.590439,0.00144,8.561899,50000 -461.238737821579,1.1705234050750732,35987.429832696915,96604,0,35987.429832696915,0.0016000001,8.656847,10000,36450.645072460175,0.0013751594,8.670943,0.00134,8.643026,50000 -483.01501297950745,1.223264217376709,37983.7277469635,101983,0,37983.7277469635,0.0016000001,8.761139,10000,38468.81688165665,0.0013153699,8.769457,0.0013799999,8.745812,50000 -503.7032513618469,1.2757625579833984,39979.80666279793,107362,0,39979.80666279793,0.0014000001,8.886436,10000,40485.68164682388,0.0008569834,8.899464,0.00134,8.870979,50000 -524.8504509925842,1.324159860610962,41976.64593648911,112744,0,41976.64593648911,0.0013000001,9.029619,10000,42503.76131486893,0.0012157207,9.044412,0.0013199999,9.01385,50000 -547.0459926128387,1.3902931213378906,43972.96816420555,118126,0,43972.96816420555,0.0012,9.145188,10000,44522.39081430435,0.0013153699,9.158456,0.0013,9.12994,50000 -570.2733144760132,1.4398610591888428,45969.52896118164,123505,0,45969.52896118164,0.0012,9.25315,10000,46542.27374839783,0.0012755102,9.268085,0.00128,9.239881,50000 -608.6497070789337,1.4954533576965332,47966.240132808685,128650,0,47966.240132808685,0.0011,9.347084,10000,48577.46176099777,0.001434949,9.360594,0.0013,9.334594,50000 -646.7725491523743,1.700319766998291,49962.06882119179,133726,0,49962.06882119179,0.0012,9.396325,10000,50611.829721450806,0.0012157207,9.420927,0.00136,9.384761,50000 -676.9540636539459,2.354547500610352,51957.18427658081,138508,0,51957.18427658081,0.0013000001,9.4097805,10000,52638.24155449867,0.0012755102,9.419978,0.00136,9.398717,50000 -717.1750359535217,2.830092191696167,53952.60812115669,143180,0,53952.60812115669,0.0012,9.415645,10000,54674.6584277153,0.0013153699,9.403687,0.0014,9.406859,50000 -776.5972204208374,3.354834079742432,55949.09929037094,147523,0,55949.09929037094,0.0014000001,9.402309,10000,56731.34460759163,0.0015943877,9.445664,0.0013799999,9.394714,50000 -817.6529867649078,3.770930528640747,57944.53552174568,152348,0,57944.53552174568,0.0014000001,9.372682,10000,58768.525552749634,0.0014548788,9.38589,0.00148,9.364571,50000 -859.5641634464264,4.192249059677124,59939.89631223679,156638,0,59939.89631223679,0.0015,9.373345,10000,60806.5135269165,0.0015146683,9.388604,0.00148,9.365797,50000 -901.0739798545836,4.68764328956604,61935.455322265625,160837,0,61935.455322265625,0.0015,9.357764,10000,62844.32478642464,0.0013950892,9.382287,0.00154,9.348522,50000 -942.0960912704468,5.1383795738220215,63930.97182655335,164985,0,63930.97182655335,0.0015,9.366373,10000,64881.60176515579,0.0015943877,9.390181,0.00154,9.356932,50000 -984.6689186096193,5.627253293991089,65926.25481128693,169039,0,65926.25481128693,0.0015,9.372903,10000,66920.25790929794,0.0015345982,9.38863,0.0016,9.360721,50000 -1025.479611635208,6.089812755584717,67922.46532917023,173140,0,67922.46532917023,0.0015,9.398773,10000,68957.95928025246,0.0012356505,9.44169,0.00156,9.386361,50000 -1066.5990326404572,6.489320278167725,69918.275775671,177337,0,69918.275775671,0.0015,9.381394,10000,70995.66643381119,0.0014947385,9.390521,0.0016999999,9.368401,50000 -1108.525652885437,6.931232213973999,71913.75967979431,181398,0,71913.75967979431,0.0015,9.403217,10000,73033.81380438805,0.0013352997,9.425164,0.0016999999,9.390179,50000 -1146.3552420139313,7.326283931732178,73909.05753827095,185561,0,73909.05753827095,0.0017000001,9.43556,10000,75067.66927337646,0.0014748087,9.436946,0.00168,9.422385,50000 +64.27141284942627,0.0,52.91309928894043,1,0,52.91309928894043,0.00080000004,6.9129095,10000,117.18459677696228,0.0007573342,6.9125247,0.00068,6.91273,50000 +125.14888834953308,0.0383615493774414,2048.840027332306,6574,0,2048.840027332306,0.44320002,2.5707448,10000,2174.077612400055,0.63345027,1.5319011,0.57534,1.8242104,50000 +147.4297194480896,0.0868222713470459,4044.9069986343375,13920,0,4044.9069986343375,0.5343,2.1329305,10000,4192.527692556381,0.7597656,0.9458039,0.66716,1.3747368,50000 +169.818678855896,0.1349682807922363,6040.991535186768,21266,0,6040.991535186768,0.5623,1.972755,10000,6211.1035261154175,0.8064812,0.74569297,0.6971,1.2396864,50000 +191.76151204109192,0.1867382526397705,8036.958885192871,28612,0,8036.958885192871,0.58250004,1.8919845,10000,8229.120635509491,0.83091515,0.63762105,0.71488,1.1683154,50000 +213.9108157157898,0.3050100803375244,10032.998893737791,35959,0,10032.998893737791,0.59010005,1.8540716,10000,10247.509325027466,0.85227996,0.538871,0.72279996,1.1272918,50000 +234.2279772758484,0.3990471363067627,12029.003738880156,43305,0,12029.003738880156,0.6014,1.8353652,10000,12263.979522943497,0.87549824,0.461255,0.72856,1.110315,50000 +254.1667010784149,0.4768648147583008,14025.104043245316,50652,0,14025.104043245316,0.6049,1.8242571,10000,14280.152136325836,0.8825733,0.42156616,0.73304,1.099628,50000 +274.9797277450561,0.5500438213348389,16021.15643143654,58000,0,16021.15643143654,0.60840005,1.827331,10000,16297.145352840424,0.89333546,0.38592175,0.73793995,1.0922809,50000 +294.29212617874146,0.6333343982696533,18017.071206331253,65347,0,18017.071206331253,0.60940003,1.8199519,10000,18312.510173797607,0.902264,0.35404882,0.74158,1.082871,50000 +313.6133460998535,0.7590162754058838,20012.96212363243,72694,0,20012.96212363243,0.6147,1.8146323,10000,20327.902194976807,0.90927935,0.32634285,0.74307996,1.0753398,50000 +332.6312868595124,0.8456320762634277,22009.03094482422,80043,0,22009.03094482422,0.6169,1.8074372,10000,22343.13378739357,0.91222894,0.31514737,0.74571997,1.0688671,50000 +351.5444669723511,0.9268791675567628,24004.935247659683,87391,0,24004.935247659683,0.6222,1.8081042,10000,24358.087627649307,0.91685265,0.29532543,0.74799997,1.0682663,50000 +370.096216917038,1.0008885860443115,26000.92212820053,94739,0,26000.92212820053,0.6223,1.8116835,10000,26372.754496335983,0.9225127,0.27492654,0.74825996,1.0668399,50000 +387.9196887016296,1.094411849975586,27996.79708981514,102085,0,27996.79708981514,0.62280005,1.8185028,10000,28386.60117483139,0.92524314,0.26571548,0.74825996,1.0670127,50000 +406.3058910369873,1.1918253898620603,29992.7430062294,109432,0,29992.7430062294,0.62600005,1.8233396,10000,30401.084956884384,0.92867106,0.24968423,0.74715996,1.068822,50000 +423.79282784461975,1.2480707168579102,31988.6567633152,116781,0,31988.6567633152,0.62670004,1.8252075,10000,32414.598189353943,0.9322385,0.23868681,0.75016,1.0682027,50000 +441.1199595928192,1.3386790752410889,33984.70328807831,124129,0,33984.70328807831,0.62640005,1.8303214,10000,34428.11667752266,0.934949,0.22692154,0.75034,1.0701694,50000 +459.3026838302612,1.3988158702850342,35980.75158596039,131478,0,35980.75158596039,0.6244,1.8374841,10000,36442.4638531208,0.9367626,0.21946916,0.75009996,1.0705539,50000 +477.4918093681336,1.4586880207061768,37976.86044549942,138828,0,37976.86044549942,0.6236,1.848797,10000,38456.877252578735,0.9412866,0.20869601,0.75215995,1.0715029,50000 +496.0162832736969,1.5198278427124023,39972.98211836815,146178,0,39972.98211836815,0.62520003,1.8540839,10000,40471.64096331597,0.94202405,0.20745388,0.75236,1.0740654,50000 +521.7591128349304,1.5930283069610596,41968.85789465904,153527,0,41968.85789465904,0.62710005,1.8547612,10000,42493.38979125023,0.9441366,0.19585101,0.75236,1.0734808,50000 +541.2974145412445,1.670964002609253,43964.741797447205,160759,0,43964.741797447205,0.62740004,1.861663,10000,44508.94835686684,0.9484614,0.18218228,0.75246,1.0760537,50000 +587.8112554550171,1.769763708114624,45960.87983632088,165536,0,45960.87983632088,0.62450004,1.8669696,10000,46551.746829509735,0.9486009,0.1856623,0.75286,1.0761234,50000 +608.7469418048859,2.263536214828491,47956.30036401749,169928,0,47956.30036401749,0.62670004,1.8681524,10000,48568.78353047371,0.9453723,0.18652251,0.75308,1.0758554,50000 +634.1093051433563,2.748711347579956,49951.91783928871,174418,0,49951.91783928871,0.6291,1.8668636,10000,50590.50589442253,0.9489995,0.18063688,0.75373995,1.076455,50000 +673.3249371051788,3.187523365020752,51947.29002213478,178846,0,51947.29002213478,0.6283,1.8672477,10000,52625.81468272209,0.94782364,0.1783816,0.75402,1.0778383,50000 +715.5178511142731,3.6479856967926025,53942.5781993866,183424,0,53942.5781993866,0.629,1.8643882,10000,54664.01664042473,0.9513512,0.17198183,0.75394,1.0774896,50000 +755.1613154411316,4.11983323097229,55937.934849500656,187529,0,55937.934849500656,0.6301,1.8626338,10000,56699.74064064026,0.950275,0.17458051,0.75412,1.0794177,50000 +802.2567172050476,4.655627489089966,57933.47280240059,191248,0,57933.47280240059,0.62750006,1.8607931,10000,58743.22166013718,0.9526666,0.16960119,0.75324,1.0773498,50000 +838.8255090713501,5.14893364906311,59929.071627378464,195063,0,59929.071627378464,0.62990004,1.8617669,10000,60776.140051841736,0.952547,0.17043896,0.75376,1.0782413,50000 +885.0028548240662,5.684395551681519,61924.34237098694,199096,0,61924.34237098694,0.62920004,1.8614823,10000,62818.38122630119,0.95272636,0.16754384,0.7528,1.078183,50000 +922.935383796692,6.243690252304077,63919.883076667786,202849,0,63919.883076667786,0.62990004,1.8620335,10000,64852.70789504051,0.9549585,0.15947141,0.75386,1.0782359,50000 +955.9324560165404,6.702232837677002,65915.35211348534,206754,0,65915.35211348534,0.629,1.8630679,10000,66881.94443774223,0.9528858,0.16299346,0.75408,1.0789212,50000 +996.0133640766144,7.245025157928467,67910.6963338852,210830,0,67910.6963338852,0.6302,1.8621116,10000,68918.14360046387,0.9546795,0.16126078,0.75395995,1.0778822,50000 +1036.815337896347,7.803174257278442,69906.23444628716,215277,0,69906.23444628716,0.63310003,1.8600335,10000,70955.25825119019,0.9546795,0.15797712,0.75408,1.0774853,50000 +1075.351613998413,8.280278444290161,71901.50822019577,220397,0,71901.50822019577,0.6341,1.863171,10000,72989.8373169899,0.95649314,0.15317032,0.75404,1.080798,50000 +1111.5350549221039,8.781720638275146,73897.11191749573,226406,0,73897.11191749573,0.63430005,1.8659685,10000,75022.37075853348,0.95523757,0.15633641,0.75372,1.0817096,50000 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/events.out.tfevents.1778625035.77174c5fce7f.13.0.v2 b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/events.out.tfevents.1778625035.77174c5fce7f.13.0.v2 deleted file mode 100644 index 5c6ba2a895f4a20a8505255251c01e4e1a855241..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 430147 zcmbTWmaT}%w>!Wm8mGwK<1DXMNx!OsDxA!LQ+IY2vO!# z$XEy=e6Rg^t#z$utuwsm-*fKk?B8>*z4!C%eeYdAx&Qy??oG4jN^)}PYhQZZ8_=@8 zy{nz$0t55KvlqDcF&f)mS<`*?0=I>(bDS0~F_^i~b+&|g)hZN?H;NB7yT zrE^}J=>1NsjL$V!QhU51@K;)=cFJv)<((F~xeZiStkGCbQ@)PML$l#^m6eng+-BN2 z*}1xXyREmY>+jm|b?aUz?SJ{ZR!6zLveH~vyIC_9EOd4LZ$RbOn0^rONm+5W@$Z0o z2Q(%sWsLLv|NptJvW{}yxlRl1?VM(~xzBd_Z^WUHe(J7WLRm6McKH?fKM zwf!nJI(1g=sNBH)*JskjZstr6SG$=@|C@7X!;n6Z({9U?JKcWgOgC;eRjFG2PfiDA zZDr;E&!In$xyH=rp;S9J-Sl^c?XY*Rm9lmIWEd(NC^uPb=j1rc&fRgL>;@?eG{Ozq zJxXm|@84mY6nef>>d@s+n4WSc<%U!ky;W-6nXT$@*gSN=?;yE@=MO8b?e!<9o3fE| zGb(6-hx3fto_5YIPP6~CUCwgQQ240awjRmt`8)7=Sqm;AX#qVLTBBi4E&Bf6l9pO=QnS&qBDg5C z)3M4F2AFkr>R>it+f6VrKzl;T-)u}4)ovU^FD0aKq&Vj& z>)LACz~6C8KP)(;l(_g$Tz6$-<>qvp)55uP9T&`n#9y`W=_MIOjO_k9@y^9*&y?n` z`jaR=5_t88kj2rT*7x}x(m3R0vC`@Fe?s(?J1eWWxXzyK?DFg5pD}Zx#{&2NHbG`{ zLiai58~jeXcxBR8CAAHc4ok>>9F+c|)%u!$5amHXIPUd69CyqlQi)I!;@c z-RtrDgB&x-zg#KOmt=?`7K5O%-@B+@0!YoU0cjE>uBI$x+0kr-=(TEj3m{g}q?`{4 z(QSwsu~FWY04Y8~GQ^PW2>IlH#1tUIb$4bSM;? z=zA*w(leH1h#_AQa$%uLe}J5c^2LvD6yf*J4g zW=sRfb`6pthNumOeWW04z*T_wEt~OHf~3}#g^am$7B`|-?A_Y{S?x{A`H=R_yJJSg z!>eWhX%s**#E>os8Tv6T6d?I3@n0oKWF1*Z@e>B3mR*nz5Y1MkoDbPIau-6nHY?Ky zh>aG>5JM~wVmvS<4Io)}d=*~(`eN#>APX@c$Uru_#smXoLlG(GL)NNaMu_$CdnEw* zTt+g)5Icm-QdL<8kOht}nn;i`4n);64L8Cna&aSoD7ukyKBU2dD|j9WJgyJ{kb%od zh8VI8Atxs9N(abR&2j1yBv(88RN_fFA98e;8fJVyFzf|DLQaqjAw=#81Ih5Z z29TREl@1c*H3zbNAp?1yK0ys2&k{*FAF{Lj3_=2@UR4H2`%@%C2=RJ@5L5Y&w*eCS z)WKDP05V@}iQE6=b0U(-1 zAC^myWEEM+>l#OKBhHU=Qvk@M7o?mInS0#?Gj^IU8VQj2cO*jyF{@<`U*mnpxbFhU zgEQ{yB}ix^SxCxiW@&lVc{mjyXD^a+KBWH1P=pM$dKCtc&DThV5aOkTkkh(GsQ~%d zC1GaCeup<75fAL9O*`7lm4o^ui)0Xs;B5MtIEA%3^N{s2gDc<^jZOkA1QP z@oy>%S*XQyWaNX%eE@PimXz}$cJ(w6vR=V!FhC9@k_;gvT^Au^huED0$omcnXC+7l z2NLz-Fm6P~N$)8Dx!;wP^C7lR6)A6JAL2CL6*D$!7T*TQh#-<7gqU?li21QD-vM&xl1rWhNoghv zIeCfch?m`kQh@Z%BjtQZ!?RWhi73{Z1(3o5k|Bh6^+ky3-A;HOY4qdhT?rD_To$7K zp6N)VT^Ysz8K;o+r<@NNr!x*A+WlXC14y@qBtr;EAApd-6U+3VBazOHN+gI^3t7lH zyI90Yqap$q+*1ED#b_p0x)c3&y&obW~xq>=+!Fe(}$ zF_|HFqvz~GQqG6iZ~2K3!>v~v0%X?`k|Bhc%|*zgk?qF-q`B6ImJ*~;UA80NDjA56 z#jdl^k@ANOhJgM(cym3krsVPh7gkOh7heTGcN;V`^$~`5+t&%EF>W*1~wGjoDUiDr4D9XDt)N{kjV!~h7e-54k5*vms$X%j>D{B5+t{sY)4vmWA>5La&9<5 z^=c6*=R?j;pNNn{2Zp=_$or)vLkO|;K}gHOq)LFe&50T-K@v4(A)bes^T;fd4UW(e zM;B7ghpc)#03oF}<>~`umM6&&LcIJCvet6$UVs#Ddt)O(g0*BJCt{eTrC%A)9Uu?F zNI4&JwmWnC?yUFoMF8m;O)`X#NPmPh)ajT3kl5bMW=oJN4&+&lefViMo$<35Kr#oC zay}%YD|0h+Q>JDiK&A{Q8A3?<7K9wTKR62@?Vi~zkRT7*%XXx9Po^WM&U&^1$gXly z&WAL1V{Y`AW)JQFko%uWh7eM;6(OzeS%m|{DO?TTpuN8d zjKPIBX!a*aFqC_0A~U07)S&OMe!pkXq;ctMY9!?ErgGzo>*FRF4)Y3uA;S@EN!nkH zd&uE> ztUP?z7S?0*FjD^)!@1$+wK3@DhU`5$M}5a;KBLe>R_+_8eEl^Z3m7a9kKRheQgG z3(X_&<6C#ALkLW3D>;*te>b7qA+i5iIOtzGKWe9Rh=`NNLl#V9hO0XM&_+nDa*`p0L`EZ|_(t$c=t!x4lj9O3w393(a6%+Pt}Z*Y1|Us) zl5##|=!B1$vD!Yn6Lh3`f07}Dq{kv8F7M1}fcTFxPm>_k9EiaXW*_nLInov&eI}7| zKBUGEWrWAq{9B*K7g2h zAmx0>wgb%72Ak4*0|8Rw8_5tt%#I^uNO)TdfLy;jxI}{Z8^}UFl`x0jVYwR+5>-sf z`H=Y&+G9udeIMxxko7M~h7e+#f{+luEIf}iP7Qi3K`J+-suh70Fp6|WC$U0*$Byf+vGKLs&kz@!V8dngq?}6nYfarFv+g^fX8_RaY$e&qS?eEpS4oge3 z8!6{Q>P}%s9Xc;nz>c&tAsIr5*)@cW(piqrNF==JtuI01y2(Pk6`Au$^wlws05b10 zDd$73?{9z|DVb=BgCiGzk_;il_69;$4ECG^9hscIw1)%<>@Ev=`I&)KzI#{?AkLRb zIUk}h_yq1F?d#t&hNWe5gJcLHUikKQ2-1!PXk{n2e5K{CIA*0B(j$*#IM#i{4^bY7UOpT5;ImGO~w;qzhxvt2$6e&kikBg3!o#d%}&meAl|)YA#vZf;zsnW_Ui%ri0&{_ z&WF@iWbPvjYNmM*I-+JpGK3I~QiSYo+^hyb)Du5DOOP@SWK3J;%x>21>nwo8oFe6X z$j&oNS5uM?uLVeO2FVaY%$_5}w7YRIKu#BoSRz4k`^a|0_Ej);BqX+9B0#E(NI4&( zUy~VR*H(^`1IWuVk|BiHzC=i0vqw09_)&kUmjp@dD+>uwXO@=!<~>^hGO{Nr=R

    #10Gh6Z+AgS9(IUiD; zVT>87Wq#iQlC_&;2qBRb2-#j`-xrpasa}I^5~PX)IUmU!e#_(4mH@=DJ1OTw3?8pW zi2KErIB%h7N-~6y^bZJ0ZBzXUAjh5hgh`MGrm`K`@6YV0&iiwo0;G5`Dd$7HE!+_j z{YrZ&bi{8J$q+({J|QGLCwdh?)(rHDmLTcb~Y>-43O(Sx$zPts=q8G@+#DOzwO z4m#2`lw=4Y8s8Bj|Dz1g?8AosJSjnZ&14}hhci2S$-5PJ9*MLf<$TC%=bi{re-n2c zAh#!z3^7Cw-)8e_)v+tkk)u5ZWk`_M97wlD%;9&=p5#f;k<|XAoDVtkZ4E+l>LeNg z#CIsk5JD1bAS9^moE-p(D%qSPLGn0nsM=f1{6gGjXY(B({oasrKBQEK`6(d(Eq5b8 ztUi(qA;hc}LZW-y)B=c_RmLp|k~~m$X?2qe!B4aA7sCpGyqrSH`H&(7=9$(#yZ7T! ztI(cg2qAg$2ua%Hq6CoJ%jF+RkkCQ0ki62(2pM-T3`auetS9As$i}JwV2$jC&?QsyUFuIa?5t+sAkvbmU+LDd$5zmRe$lLEZXO0Magp zWC$S@3J7U-pjT~xI6T`>E&H}`ID9I2) ziW(!ty_x+5fW%o3H_^2Ig)d8JOu`H(B=a}lDY|GoksHXTTY5F)3B zkR>g?20=%Lg`JyNzK5&f9+$e=Fk-a;l_gtXRjTnv!$7fFT?lGqX< z*WRzho1xS7{iaKhz|pdh(OS$`l)=>p@VeH99;93V`Oe%QaqMV{mrP%pk_;h)ecx$= zo6Dy`M}{XA%#|RO97y5(P55c{oV2nFbj0lpDd$7FsxiN5Vl<|##*SPd8A3>&I(DS- znu%WlVq2sAA_-DBMz$l})-e|!F6Vy0ojqTfl=C5H-W|seBlzQGfTFOG=9x*4xW?8OypS^k$Dd$7>PG%0j_2%gs0;G8`$q+&+G!Sy` zafS_aBsw*8odk)rl7&RoU?4Nvs^YVDi!PCJK4j}TYwUqAvNGnNZkUhV}dt6CTi>?oXLH586B;#KWvY%>W{~89_hi)cW zvV-j7t5@LVNgX@?NLY^@LrDE!46++vT7yANU*5}sLG~uQNzgwXWS7}I>szb7!XUfW zev*146&CAbYy=C479t`-mzYsr8SNNa1lIR|}5|BaS3o zfkF2ClO*Ne4YGf`wtY4nD;3){*(M!gyvNEOD|1ws?-Q%<*t@|BAKi(R^N*G5FEDp@ zF6q|Amk`u8A{oNZM5R4`CTo{z`2obT``|DMQpSN8r7)-Am5Xxl_R`yaq?`{a9ee?I z)!t3xZo)yo&QOveghXm1MAiFz4nU^g*%U27a>vPbWaB|U+=y2x)*YZDwTnnOA95>` znG(By_=W>Oil37VA;ho)Lb{AgYXy+OhWFwnNaA=|$fi2XkB9Uj#hwAtYTFA^oSf8U~OGRizXO5wN*zP={m)A(ovGQhL}B-vQImu2+Txsp3G=Ml;{JPbANo2@sR{q?`|VJFNpk z8mt_HuLN%7MlytuqRt3ulD_#qbR@3UjvNW{V4`eCd~%pEg{hti_)6fIdZe5WDPQ^s zA%osk--3<=tC9>M#6=$=SAV*U2Z+Ox>{}8feUdCh&WE{nlhbd~Y3NA%HKd#m@$A85 zth%jD1oK9nF)*0K--Q}#P|3;V7BarP(We2DQV=3LY~ z{z5Kvq(?Bx5JCbC5%M_G5szAS-E5vo5Z}qNknxRuaapYBu>J`^u9}i^J|r|`J!a^( zonitIFLRP1glHHcw zza$w#h*@`p+%-Ljqde1!)oPaix+{}xBMUiGgIQV%SI^_G38z=2oDbRFliAtpjd_l* zsax@ZWC$U7JrI(amv96+5?eg5t^^64Dhruz&wNEGn7hIZmR7&lq?`|#yo$MYQ_m@; z26RNLl4J-Ww!IJ%m=}ce7LjVcsuHA{18KjBIkVen>!<*vodzlALq4BBf=g>`%h7!R z(pr~f2q6`{5i-f$Zz*)dVpG1l1Sy&(+Yz}t%%$_%%hq28NJ20v=R@ANNkYiVp%d}| zGBb>12q9j55u#HSj<=UeeCu?OAeqx;As62>Ur~xf<dxIHGG(kjLup56Cev7ka9l6$o4dL}4U%*?RyRI}{WEq%eh)^C5E{x5uYeCa%@RiTwNN zBtr-(8jO&FWjpYtXzIZmrc02pnX-_^)0w3;yn`V=>fAe=l=C6R9?WB$Up_N4Np{gB zLkMvhf{;aumq$QHR$V$XSAuwP{`6SqSY~MrR*l2AyFI*4%K4C?&6#u2m*X3)U}>Ez zAQ?i4oCQKW=M`-LNV{ik7fFy(4#Z+9b7s%A*o{xoyOfi1K4j#Tc-%+M9ZZY>h}vh8 zA%p}DLx`rm$##HDtu^QqG6i&SB0)ElfL~14zf0 zBtr;E9ElLmE!*&^l>zpp{t_f`t}G<mO zki0Pnd3Esf4cJF|c#MpYAcYRH9l6?=+1V@h7WV;&=_*prhcvvy%v;pgnEnEm)|d4p zLkO|8LdY&ZgLVL!IV2%gf}}diLM9|JR|$>3TlEG=Y?ctKk_5;yqs$;xv`Hw=tx$A71;7uSd%HkTXfU@GHu5XQiP4IjBl9 zgbNErvBq{r;+qcYFn6T<5~NjV=<@}7D8#zEbw5FmM8Btr;^ zoP-dQ!g`&dBkMlzy(&R+on$-Gt(5su)M)*wX|S})<@Wt4=R@4CF|S_wG)q1oAO{pl zh7e*n86h6}&JzLBq~Y~D5+u=C7Sg8{b3zO}l!vpeZJLvEK4f>`6kJ*pofPZ=qOL(Q zgpl+p2nj#%2_G$pFzHq#L4p^^Lc&HfJNuH>aSNd%C(TJYA97NY`OUs)@49CCGz?vK`54$}FvXxpMqlE3w^3IUjOPBNZXTn)(?)M~qBK zh7jU110hHC4&aTRfe9bKNsx3GS;$v2<_lkB$rWW-T02unIUjQJ^g)Ci8$Uf5AQRF_ zh7clWhmbL!wbP&@4rW@lD}G)3janoNS>?&>BjqOJp8-U9I4S2t3Ue1BWO8=H9{@RJ zMKXktKzoEVc{QURKsG!Y-9Uo)y2?UaZ!=#nIs~Q~0wn!8Dd$64)?h|r9NRfv0?5e< zk|Bg>%tA<1FIxqG7^xp`CP7|vAT>&v6XL8D^GyI^qeII1kh*=Dp@V{y8u+NQvH{5u zLK0^qWVG-0z5uEFwn#&Q=9y*ZOQ0p#wheI^p5ngf}V&n&IN_>dm}Iq{K{^C9i;FuxEJ+G$<@NZ5ChA%s-S zM@V$1rE&m?Ym;X#L5dd3cBJkF<^?3jN)O^AYVCDNIUnM3@Emqz;l%g&(BKk7k|Bh6 zIU~f;WC*@CdqcXx7zvWOL>BU;HuDvw%vV_yoXryiDU>N zhOP+7`r?Pzwc>u3&XXV&9LVw;45WRUCjQh-P<;BQoDXT~>xB@d4#&cwBcoJEh7gkO zhLDyIYkELON?cUjB}o1<*^Vr&WM1PM^d=fF&}A+o<$TE4rkMyicxEa-qUOJrWC$Ua z9tg2fYTpbXb8q)uDM3<}%R&A|Aty-Q3 z46?W0(dp0D{M{gXSM9C1Nhe!>@_|A2$NNdzUk$PkjA6Db$X?GI2HBq>#^VQ7X=#e`**t71m?+O;Z0CgX|_x(=jOcmdj!oWIunG1pU)NcA3o! z+%h)@2HCZrk<`B*WPfcfHwg9`=e)UfVUYd(8Kp6HYROpNR|enQVWV@f0BEIzJAPAYLnFA!}oqk>}*ft2hBO zyE`f8L%hSs;ja3iXvzeD3^FAdLWtZlY;J05tAPM%o~IHaK}tCg#a7IyTCKuj96+?c zOUn5Wh2t;r>=iU5u`@vWKOq@HNZ@jWoU+`B&q&-2c8!%F*{ftbvT7dlyYyI6#(r2@ zt@e;|KBT^m1416uY#0xao-rgt2+>%Hkp4z(GNB`tqxT<`AaSc@A@#N}Kj+-XoHz%N zLDr<452-Pbd2j2(K~47qq?Rqo5JD1HA*5DyKxa7qs|VgVEkOdkYRF3TE;B8$So%o` zzDBg(N!9Zy*@F*ZPdeKw$oKsHcP^fn09f{08)r(EN}lDg zE_01jU$cE%uiL!tJ9Ti6-xK828!VxYREW@0##lBM*GdD6O?$hmlBIABr!RRQ zn4hQqU9x+EB}t#E=UYY%>Wr4PCt807%ckyBh|rR_2K%B{qt<1x983+mBUw_{axC`R z%-+(iSL7A+HlQcb|}tzG$J$r8DaW9fLG z`E93rw%idcJ3CYLe9OXf_Gno;TcZLjCB{^U&|c~r*i zFYmUk8xEFr>#2IarPU4QHjS}q^>eW74xmDW7DaEg3^j081xv|~A@3zi8PDR|o%x0L zrI+_gu-GULCG~vEfGx~7hAS4%L%?FHLWKw|d5opQg)mP;{@*tO)-$9xMuPlp9Ev;0+v$x-+xC z)c?`N6D)N;QT2SwP_;W~k(-cY0hY7hsSu$>cLQ3cyogZ(%WZAjW|F0fXSsEe*D!YE5n3u3%TbTsreF#E6s93r9{6(l@?t7uNwB`y z8!QoW;iR5#>DTBd_Qh_7#%-`{Rir|MmK0y?ON&!S7lY+a@0^a3CEbr>iA-Y7FLAl` z{J;`3h^psXaxZs4i%~`08DQ}qL4^n{UVdmfoBbmjEW66}jU`LeMvle%8nf?DZD8^O zESIXNdcI};WM*!tVs_{ku&k9cAR8gHSZqW~(KF)$uzagyVH6~=-zCw z_@1Wf`IdTzm?zpQ!jBh$rQ|#nBD54TmgUcz&j5?xv`3>QOP)WcFM|x2@0Ru6r<@1N z={ZzA-%=6X1ivSbQJRV8%A5sMh|m)0k9}FbW%*sOG~cdeEm@KSIF_UC%#|0{_&4%k z@r|JB`Icovm_I(zHbu=C`cip-3K3cy0?^XApcd}?Bd%K7NtV#f9E+z8v+p;HxSR(T z|2tGY-%_}k+4s*kc-8|f2Om)(LW|*Mv~20+yA=Ae+WW*j$x_X;WOrbuVngLr>Vf5W z09DVoD0gCRu^rmq^e9+vhfpCxOBG|;y(O$0SaeUlbdxMaTR43gUd$ZFR*Z@X1xup~ zR6XBvQ7;CMV_SRgr~?+GD^!TklD-A|V$gS|A6TYaXs(njnSmV3fW0elGv18({1GhA zt*CmwWzTG87AVVf6aIctoI-^NExv(hdGu!H1F*PnclVJju|XV*^(*E$)?~bTFj&kY zsd~Ppm3cmXAmv&r8=)_9u~dlAVi|;%-J7lifyML2i7k@Fe=Emwx*2nRQTh}U1QxaX zR6XC)Z2`0I+dO-y3zkzQREW^hd@EW)J01EC7LRssLnTWE&*D?f^yTy(-PT}v+L@~7 zTQqW-IFSZ+*^GllBz8zo*>q^!0 zEvn6!nTPmQr7giSyEhdgw8U=1zPPLmI1PQtRrEM4SyF;ImR(uQ^|7goMzsga_r_E` z-}2-+^T^wnp6d^TWm79EL}+mdMvL9z7#FaZ?}<&8EMXxW%bYGuUxpQru?0)y0ji#F z89L4e_x%V>w{2k2PozSGmR=!f+3&mh99llTJ1bedwsR~GzcF(QFV<~+50>3>?@2x1 zqOZq{eHp~a)dI^_MJhySk=u@zZhbr9wTo$AjItz4DbJETh}rk2)IEu}*nMgllX|}8 zaMKFx%l6VkzF?VHj|vf5vKh;6j|WlEm+irG^CV054o+W|7Bb(*+zKbIfWDaTq3ZdT z`iXPU68dBOCa_G4p+ba~z#Z6^dB=v}@BO%+iFYMSTqwtqyM%dq;^oF>*T9mhc$C!h zExxOm=6)-%+yEAnMpTHEWbyZ(3g!RABrVQ;7*QZs~U40E4V*vAXvJ;q3Zb- z`BTiy|HQ1dqrmdzBNZaFXzWDGwlu9fU`dW^`&zP8@+=;0m}?h9M+D;UeT5TLJ>Qb2 zRDx@)-IPyHz%o6J3K3e$7|R@m*9KszwcO>SWGURm>C48u%v4>%^>BOW%P4QEo^L65 z9Dx?~he~I_V(d?a2rY@burH%Zi;jb(&c~xaBunaUj%9FN#_~1Y2JegQ_(9e4Eml*q z(c;|eQ7f=)t&>DHLTK^ajh4n2?N@_kvC(6N_rI?b&xjz|yp>K_kiH9nP`bS;2fyzB_nmU9e>9QuTaG zY9-U$yvB`cfaQ`Q6(Y1KhNI>ApqMvcN&0BrQnHlsEajSvrSf9?<6trRN!9Z$5vkGG zmp+%&uY=`VoylY)gqA$UGHb$Ry!Sfr`d%%`lDmh~m#|rkWtpMg8L)&GQuTaG$*5|y zcTvUD<4&$m>cxrKcx zsQ<_hEEDah5TQjk5-pz;%@%>h=!k89$x_9$44uK8_s!N!cZI&RKS9;=ErY|<@OS0t zVVg6-^6@McBD7R6mN9{IwZQUK`H+QVd9at$mo2TCzUWs^z5$ltrc^!OGCZXS`_kg_ zat*MoRi{FPmXy8Nmlk#_hk@nf>i1TXC4C>qax;YaZh7!WK`~fn`%(3LOP|Tiwegy+ z-Auq@5JZItEnfT3a{1gryuaV}b9)=f61AUWxs=Gf1~7h~O95E=%4w5&zUAmm4_sqv zBirJ=r0$ATh|prOA1$#{UwnnW?6|ySmSpjb;#jt=VSXdqm}qN3elzEYA z>^bwF(3jGCREW@`5`~ujO==W^rdN|2C z|8nqpuxO=IAwo+bV;Q8Y6#e%YOe@^`SW%6gVsrw;FubJHC*4Xn;K4epK10j2xjv(KBJNZ z)AgB;RFHMrd)eLozsKUNW-qo=%6vrYhNU(MIY)NX`Z zw*P)zD;R@5{)3A8p9bt>Kc>HfLGsKxTgitnp0KZ5poPrueUb*jidxZt3X)co%!vJJ zGX}V;<*~D-zdw50VeeimbylN7{^g8)=LLRGy8aFeU$^dsQdMgzOm@h=zu7mOjn_9R z3V?;Vv^^E@7gP4yN^-cw8|_%HIP~|XoWAgFx6*@7RMI~kvzOVx$L|B92mFqeJ9z%E zlBO{g`}cG9#|v8E(aG`5;IP5J0|!ku*rin5n+p6#gZ34+3vkfhRJ(saI7*BgNM#C- z61p+?5q^u8_l2h~CJv)w{@tkk)Z?3`z|nKaiaS2i5hwXSN1S%bZI$Jn7P`3&R939f zSWZ*^wj49*wc$-#60CLm^;A9o==mXoxwIGJuxbE&E+Ye|5aH+IzGXzPwHNc^}r;7ZoZ*XenYWgI$V!z><9BQl4bV zkK^>^O*Ug08NaR{^rhQvs-ADT+k|=N?EE?jE?}{GNQDS3v2oa!_Jfmaf#sZ+%00=F za)e`+D>O561u*5o}C30+<99VwlP$5E#Tmo7?EckK+EF0SteUvPv zJWCx_W{qVTtIC6=rY=>_x9ph}hYNLiL!%tK-o!I(W_`uy|EaAwo+TV=2A!brkF`3)j|YFIftY zar&ZEhuQbD747k9sBN35dcGz7tu`LV+GOg7g2iVW6(Y1G9>cyw9bNqrEGP5r^(9Mc zGRLCo&7Aj7r?r_5eL4P+s^?qOin6dTtEP2O2FsvlREW^xnT(dA8vT}orQMX^9+D;U zIL9Ja!u%F%zwJmNScc7{>iL$YL))Nb%B{Zm6hVXeREW@Gb{sAFp{LEjVtV|1f63x~ zf@3NB$?Pw?4C_{b#XgOy=UeI&hodFH%B>^xW#}a;L}*bwftJC(+bhA+)3U)Z$x_C% zteeC9C75Srh0VY+W-?XJw{)4pJlei(rO92eG_#{ZgqA$UQnl&Z5U?~^Gj6P8$vw&G zi`+$~FQ-adtp!WpO;kPK^21Ud_x(9v8{t*#cH5{Bp(XSr_ND*bz0<&Q;m$f6$D zv6$#Ezs1Tll5xtjQxR3qw~VX7+@4$#H4T4@tt_KLgcjQrv^<&?ftG9MZ_JV`!KXNu z>E+9Csn5zERR{WFluOm~EnDx0VaOp*Bb+Geo==4cExM=BQf8aE2Kv(WW0M7vrHW@s zT*7?k-!-FgH?TPWpz8US%?4x9lI!@$750~bbyUel2rU(irABL?=3sd^W7rbO^58V5 zFRS-4=a=U5AI$)Zl_OOzu=wmni}mBW_;iQ*A}T~^NjZ&u`M5-{4_Kz04fc{O>1Q|= z@B56U)A^|}V9B^d)$=Wvo_s>f>ml*|p)a4VQ6WN$*BP|bxstU7EK|2V@RKZ2sT@m5 z59WLF;eFW_V2KZ->iL%XJ(%xfZWB%Iz;ZB(3K3c?Qqj`)!{ZHLF&?9|O|tl&|08zU8yq2s|dwJwN*ZSX$drAwrAFS+r=Hj_C!KfZG$pB+F}_rTIPP zyXDe3iPynW@|ddUTfSy|z%`aQTNkI*J)Tn`LQ5fIvDukX0+yU_A<>d0FOAcejeD6p zzWu&>p9jlZ`R$~hZ|UGx7yqE}h3|7tL0=lyr$U65$TaNB@WpF~gQaj-cD!UsKF6`d z`7ObZbopAFjbO1DN7eH!T5rZ<$mE#iIOXYLLxl(}4(HHfG$#nJT^!PDmLgd~(>azy zW0-Sg==d?Oz+&8ks^?pl7UZDCwP4gN=!v&3{{ zEN?Tnq<|$`VLGYjTN=ACQ_&latZoOEnkrO?&{D-%GBiy)!WvuMWJ8W*DazpVrL!LM zJvsW#?l7>#wWjL%mUbJNCn8P7(d70m$&z`VW0}35 zIgVwPYvQ%>okOU4zNO-7Cay7~{_V=aGI=x=BDDCPM@!zThaaIYU7E{3lq|6qIF^8s z%)b9<$;mY6%g=UHJ>Oz+=pb6!9$kb_L%P45TT`ru}s>W77T0btjoPB$&!DO)0YBA z<_XoNc{lLMuI|gIdcLL5x*_(ZbK^b!VEM9!3K3djFJfPA1S(tu%W7}snw7tr6#%L-;Xz7)O7KJxG@jL&EP7$h-#Vea* zX*PoSK9;z$sTEjeccbe0mLc7ldwgc!pU(!1fe95Nw8&+nWlgt9q0kq{);H88ODWHi z@`;_kJ=wY)SUk0fzZ^;sPg=10d z6a#OeCieNd{kqQx7G_If}wdTBqV0n~VG*Gfs@+|gA z%(-%1rx6dpGWHHt&$rZZdWav0t-BB2d!77<3K3e$7>mBQM>Q3otKJ*W_&#q%m!K06M;^UKo72c}Dw$ZH%+8vAxF zJ--@?V3|3Ss^?pR$1zXZ&$qq38Z0B{Qz1f&*)_Bb*z@fh^rhB`;<=K=`#Q%`{e(I1 zA6Q+0Z`r6wqw4vV=kIo6U(Vl+asbPnOH_!^qIexGZC*#pfhGEr$|A{9#mg$Il<-g5{V(BeD@fOCDnh9H>_l2Ar1|29@%6llLz@b8(}7`kZxv$@{%? zskpzIymuSGY*~{3=2|d$AGMH*`kyB6TYl3IfysLZPbyM8c^|Vo6PZS)?X+M;{qUlK zq!lGIc|T|%18h6fG!7>3SNl>S|8nxaAn)dUn7mgHq{3t;?+x-=<4(42VZmKkm}_=W z0e>-hZ@0ZICLMn>@+VB*hwPz}{^{hs%myYsGk*$`_w^4@v41~#U#Zj;C-1eY_IbnP zy<$8S_>U&<6TP?M4=m2pCgXUwb7#@B-n>|_yqHPV^DP^0F=v8Z&O0B2@f}q_)$=Wl&s@a56c~6Zfu-^h6(Y3g z-b71H``7nijm16H4U;TYJWJ?Z#`4*J3qIYktDLImTP`+bel=<2q_qZ1=PD{hXsKW< zTAx~agGJffAzHFLxW(zqr3mJiMzg~C4PcG=1W@&SOHqs3_^ZjmZ)X#*3<;q^gqD@4me5zsUDBlstne=B(m<-7Z@F}(5ZBm%g`rcyGBA`15n8QAe9OF}%;n<^dQskBS=@~Z5n3#6 zqs4KDCSH{*|7n;ZS$ywsERKDcS6MVEn7$Y+eo9?QJ>O!s?G^T=kB(*)SPnL#LWCBT zJ7_77TCN9u`ElGXN3y);S$g$lzMY#L>pC7Rqs~$Fe9Mjr`_MA9bc7>VY_q5kp{0Yea4wmZKr+9_Y{RR~xv_uwQ zUv3&k;=4Q@78gB`EXjo&i{&BaI5xqx1-?J#f)`cKx3n!~rq=vRu8o1d6!=mhLW@Ho zT22Ok+6ikcNV&~3$r5^(V_D$Moc9AC)OCV2_Fa{#=Ua5nGVi~9v*pzbu=G%;LWCB> zyJ%6#DNzH<#wG6MlBJqwc@oQ9EsZ~KHUTU#Yp8m@WnvTNg$H-DFW}RI8~mscp{0tk zl-3?q1eWr{sa2At=pLspR>{o1U#oK}Uiocuf~x0REEVMNT-j&Q&<@a-d1+LL(2{-+ z`_ko<^CYmC6uzwS@z-6S%=;WmWnJcbvZ6Tl>;sM^HxLsW>+ zV)+0qCqsIj0n6onvsERF|3i*th&^LDGtk@-`VwzJ)$=XW9%rMaPyA6YumoFDAwo;@ zhiF+6Xa627%ZJ6POO^_rrOtWgs@!d*wGF`%HUK2&6rAs2rWg7 zWry|L6tMUmuhNk$`Hwh#xh2OO$2x^Nw*s6ON^UFY|qD!^DR8`{ivJRnNDWSu@kOUo89eg1!X2qe6rhxhH68 zGBi{L`tmKeWQ1fXY!&L9%2QbNX`TnLBQViQ2x`U@31y)$=W1{UR~sbA<&?5;yEXg$ONy#n_ia zMmzGr@@atAbjcD|!m;$UXO79O42RVQi}whso^LsF!w~oVU7>yPyJh|HREW@GU4oW^ zoP2Yzd@emSSF!|_ax5o8nLEC*1x?$4W%ygFo^R>Kj#l^cTZVsE!Rj*=BD83fqQ#`` z(JNpv9sJ5gvQ+XcS64C?HM^EQz_Mf%RnND$4q*<4#bbik!2aShfeH~?${0)axVm^v zsobN-GRac-l+zc_V8(L0L$W$pMz5yo`4%se8Mwdn{kZl%SS&VBAwo;yQ|wFON!>8$ zOSjoG*GZPtXB>;#Wab@s*^Rs5q}SO6R6XC)ehKrQ_`|j~)4;NOF%=@Tcs@f*nPJ{0 zu;ibP_m?b@&p8$s_OE4M_*8{Ybq>y?>iHH0TV`5)(Bd~Yz%nqG3K3e&o};B$d7mm+ zm2rHp4;*Ooc&o9pU1fMs@fs-ADr*O`o#q8%4(z~W&_g$ONqjOEelFB_pRSsxr@ zB}?uLPG4L)Gn2&6wB0JeqWPVw=UdLtn~xTyU3vy!8B=>6*$AN}^ab{1|HE{=Zux4# z>7$Y*@g>Ky@*Z=pe0SsKSg@@3r0V&W`Yk4+W$fjdSHLpBiwY51Y+s_K(i1Y_!Y;pWjxcD1-o+hfn{UuH>93#sr>N(EmbuK;}hM(>r){@i|#A5c*WF=2g}~i zT{0z070>eIJaZjJKBuq)ST4#Ll6tiL!x=T4#Jbb?|Zu#9;@g$OMvudy#iNxk!-FF$J~ z-H|NmZ#b4?kC@-c>gUfd1dC^Vs-ADz(4YB@++^4dYp?`1r9y-juQzBp>$7D(SlSQ% z@K~}$z2#WubYs?-rJ_5&mvr?gs-ADLwJ5~CbjkWU9W4DPP$5E##ap!ae~HF-MLYEy z@EaR6XC)I)?eP)@$Dn!M|;5H--ukT2#u>vcy%z82Yk1 zz~jASdCjvd-^?77*GB%tH@P_Mpz8US4t~MdmtjlikA%K>MNlC^OCe)vI$%mVSSH7$ ze3LAB6`Z~-y2V`Q*W9Dv3VnH&K-KdtCw#7;#kQd9Ot5@CNreb4krmh%t!|kEz*707 zPVG;>t|uqI<5*tSWA2GJi&O0jmRhfUB*v&Oz8=q>?EaUxaEx14=H8!ab|cj`i43R9>Mp{4mJw8*7Z7lK7T z=%~45so+`ilNrnWxaauylb`OS>iL%PKArKH+;rHvo?t26ON9t6MT|u^BRUliL$~T05{Wegz}1gT+Cc3K3djKVx5v*O-`qNgSXM{Q zvy&`gUpN-aP0Vpj$7M18thi%R*s$@_>!PjI6;-S7AeChyyKrsDo;^4>b2J|3X1#r#Zw$$Kp$ zD(Zilyw}c{yA~$zAN8gp#gq5ly-p&t&N!$0a5fn|fC`dUl+5IP`(Orm`C4Rsn7nT_ zj0*XellQHTHGKn<_XT69FxkobQ8i6)VP-t&Tn`qe<3uXpFDCE%cIb~ual7vIg~|K7 zQ>mnXI(aX%fyp-YX2Ilr(JU(V?lHD9Ei!Uzrmn?DB9E;LO z=9l?^V|BlQCE*%X&$qbsW2U1%Kd8Vd&rx@%5TV7o8ZF;5`&xr#z}#~ok|ppv$FgV^ zbH${)U)mF}jCY~x`IaX6%*5)QCe!dazjmHfh|r?(9W4u19NhsH4efV(BugdF^1_PQ zU)t2Zw+k!>4XAp)rTYQqoy4hcPT_OpVLhl2p{0zm3{9&^kN*x!pkEmLpmpOQ?FjCG++?JSLx&*XjtzWcSxph|r=THx!-#Rk0i02`mMV-d>O_ z-fK7($EnOOjhuTXOTkjFjH>5b0vgoD_mhRmYYYd=p$aNQXz^w&P7^wLf#pa>ud9-! zjAwZn$gHtW_v$?dOI!|B&$m>BGY63A&eL#z*>jT$5n9R^OJT<`x?s_*>u^W1)-Jv1CbH z$FW3>T7VV}HMQZ;m$%ocdcNhtGUjC`3WekM!5V8&NQDS3iHya!ul6gjSY}n0NtR%5 zjwMi^SssDTd-0ts&2Lcke2a>K6Mi6dmntiOrL2$&5n5DhVqd)0PuT{RPx+?rB}*00 zl554>jqBaE{y^xX8@#B-&5m!SDzIsJhO5n3KFmfjAk_=M_M`^WO1f8F&- z_u*I;EoQEm=nOVl4i>BVR6XDF+hr$CLp)dE-n>COu zQ5!fGpWaMgf^%x&cm8jesd~O;O_vwAztm1XhF2KfZ%`pZOB7={=VK5KmKQS|nn@O4 zUyfyn6?1;+{xCre`clJ@s^?qoMGi&F);39b(3j^fREW@`D35(H9(Nu8ZciV@OB#~p zHP73CC`u3 zm%8k`aE|+I8w{4L`cysNvc~WoS`=?y{REcCrc{W~lE+xeXAVCCYpllE-o}z8c_YV? z?8pph-PQ=kzYOktiK^#YjvQbfMaVku)gCMzuTddFOSjs%#^N`X$AYDE#$FT261s_F zN&Uzi$9nD@Hvsz5B8RHyTfEYkr)>{@IfhRV4ZKN(2rZ$ErM*fKeuue~ao1e3RP!t` zCz$ho(VE+*!BU<>)$=W>Da=*ufCbkcLSLTTq(X$2YR2N~s22@=$(o}-MzR$7bNaHc zl=L5hSd3k$dcNh&Hgnuxl4@z-yVV^$sSu&XwhpebTOW$??m}9IpS5Jk4B%Ki zEtv1gn?IkxcdhlgLe=vvRmGdo(sSdCMCgnAEhH*+kb z7BbgiHX9GaCsfz0pz8USdCi%9|5~jaoLrgbO@#<8x(e8rH;$#dz>?IZdY)wQ-@>t+ ze#BhsQ(rxD5Uerv-c&u`GWq=->`VKNdu~Esh7X`Zgcg6svcd2uK0#F1-oah6RPZeG z@|bg_Q}toI4&yqXs^?oytYMx`Pw8AAuTA%Nqe6t13dST-<;bZK}1TMX>?nq~T3p%z7LUzTh|tnn3H!36X8c=NV~H1+?vyNE+c=i2 zdTzKGA@APlfJNgvRnNElILQ2wnb137nP6#FK!pe`UX11Gn1*w}()DQkKFLzbvp8rl z=lwGt-3!6uf10Z2TS|hM+q^9ob}*+eQ+_&Q zUzUwCRRznWWU8KT$(wW-Lk7&h`vv+kAe9OcS}c@tjcp#%t3Oyw_D?=8S>i%CmY<%? zwTp_@);P7c;|NvHx18D)iUUr&=SX3o;K`?PunVHZAp_Xf!jHj*~^*l zFqfuJ2!OtH%%kf0mcyfNp(Q3NvH@5o7E&QXi%MPW%l@vJ_|C|Xh%Z@^rIKfv!~R`` zW}6l8y5;zTR6XC4*p7LdL*)}!Kd@vTr9y-jcgFJi?m#(MV~xGL-HFt3zMg9EMzVPC;#k~`nQ6kR zvTwD(^6COr&$ry`n2eS-5f2+eU-n<2LWCCi`q-D0&*x*naye+^C&^OAvlwn+)>wSc zr({AhL&CSn)r9&LibZ4LW=`q*--IdEm$seKJZhrad)boZ*fZPgx8b%On2D@eYtH)g$OOVjAey~;tQ};TfS4Q`gJ`yF^pqT*~Oe+ zb_^)b1Iw5xR6XA^t%~{KU^;r{F|e51Qz1f&VFO%ama}gcfn`vDR%6K$9L}-06fu3- z?qOCJEZw(L^?XaL!y{Z{Z>>|z!LoM`6(Y0*GnU)UoPEGjIy$hGWU1m=qT`rr;~xe; zzXFy%)>J*;vN);=EhQ6AS%D>g1{LyuoZX3APVM(L@DMT&ndf;PLXvqVbD7FaGDI?z zgb*qr6+(!jRFWi7DG4E|gd~IzLI@!V@AF*0<2{~rF4y%8-#=iT$Njn2wTFA}wfC*G z6fjH5kmC_xF>QCTnXqIYl=P+U6}BJq8gSSdEGN5?dby=y)DENTV8$Igu*~mG>gATN9+%Nlv7a~2D(o6eLX?&` zX1RAHq8Rk0)tWiogvHNCVo4p#`f}i&DZb6&t`n)3TUNHbik1n-r<{boc)OAirKMSA z?8~+dIh(J z!Nre3UyP?NnJO&dz7osL&g|NS_x5tZ(3k0RNxj^%?-@JiZ{TQ;Z`he*NkWvCaAvu# z`5E_PAE$=Q5f)EBiDgz9cC76ANxu(R2G1k)a!bPrLD-j@ZUgZiUxg(kL}{s275g%G zWesx}a5iz(u@sgwe^~xz@_t9NYdHO_KXK+Rn7m)&Ns<(k_w_T&;Hla6SHX@jd4Jc3 zr2Nz5eN2z5_@TNRchQ2$`!|6k?(Zh=FNd;EwrN7s$}o9fKa@oMPm}k9omY;9$$S4u z5~-fN|LvXka~vC|z=|@BBSB(C{Wf_&;XDKWvM#p`Chs>~B_aQD@;>8rJ3E-X-<(3i zexJM_(s>FF1HSHWuMG=R=MD+@o5}m1Vbd`wwoCs@FnNFI0ZIC|llQ-UK+U$hn#1IM z!7~#3&y)9GR?NW3dxKkR+QQ`hyVoS}UrpX;1zyL=d)>ZUPQ#IC-ba$D+)L=Oy~MF4 zOQyr*{rVN1{@P3Yx5@jf6>l$tBSFBV>dVC*r%1N<%+O_f&rQZA_>YI1O-a3c@7bpB zDcpN*)|eO#-%FQ;Bt-eW*sYQn}!P3NH@LFNXIwCm|j5c6r8)H7VaD(kO+=kT4 zEr;B8qovSd93Is|){+pVC5u_6H7R!rEPG5=ZW5Mwe~D$DfjNGRa}&mef#s4Zsh3;o zx@BU>{kJ=BLtkzzBq2)6AZ+zodWco0|?ELxQyEcU??%kFk;jTLNafbS-Jy@J%sEf0-$ zVqf$tt;2JOv~?szY0;>GeJOq2V;fj%ng!n!mTZ~j(fZl=G0L`_`y95Ho{LDm+|ng@ z7=~o0cE#tF`z$9RN{bn@Bo(wSIIo(dUT%3-j7{Nu z$r;uQEXfT?h|v+d>StLQt>msT@Lz1&iE#x(3p z#^+`DKO~m(Nr=+o&MYODXf*&!OwO4v!jdnu)IQD5lk+PMd5V^gq+V{Z54nk!1}|er zKwomckr1UNpINp%Xi*g`&BN+weEaiGNk*uoFXu+FzFbPQ!RaWMXi_h?+!(wUE%R^B zZ~{x0OC&^TF{y=X>{*p$e1HGMyb)!DCFZ2W;(M80EnO25j8EMR`bO&I7V~5_g>&N2 zL43cmNwJZCeF&u`hFK~yk$SmhdxLqn z#`1Lx@yTI}91^0mI55jMPooFWm+X!WnhQ&g%;Nlm-Cd}jH|Zo;g7rzg+;VzxH1_5F zC(TFD7vFv)L}|%kmiPDf8GvPL&G8+CB`I9e7h4DRl^$(BuQyn}HX`+M%hM62*q7Rk z=G6tu-c}?;X&G4u*H~Q5ZMeNO)Q;^YETN|*mL7?0Kjt1jrz2SQ*CX|E%X@t`N!+^2 zC=IZTYeqtpmQZF%UsYud^rgb%_x*&$ zcNX{Nkr1V&T3zhRI`4ITz~V8c;c#IokXas0XWNToa7X+{-LEF3UTzsOg6+prS9CRp zzI>cXLX;L;W=UE6$rLO$b)CiuOJ;-gU7wWLoHx*yKFvtI+|q3*yW{)QJn|k` zTD2n~N=qiQ%(i{m7%YZX5mSXFE>dE7x`FMJA1%@(TKt%0s!!rXuuSPUda`Or26Wm_%bT-=xORmhK5yPfEcJIIU6D)c?NWI*Wy67GrE343?Ly& ziv_cs@UzK>zO)*AcD=BqMoao~Y7MhgHjTo!p9~sJ>gASEeZ#OXC48gNGGZbLQCd=& zC9~n2!(gd=s>N1e36GIj%+B*`N^OU&gT7pfA@y=g&xXg*vZM1CE9gs^%Opf;(QSZh zEbaDXTd*Xpn(QGgp64W%Wz#M2V}xq&hyqJl2U0J$G+tW{*I2(5H=M!Zx0Qq_EuPGx z+sI-mSUerK9}t!znWbO_+xAy)JlY5><=sfV++um?EPf%=Ck?p_mV>)Vh|*HTER`ne z6$eZ8s|EhTk`*iIOU@=X8M$ghcl`abB$d?5EhWaX>-;0yI6t1|1;va?8%$Y(m1N`jI2hmqGp{L}`gR*Tw_{M+4wZnf&vnvw6to3eTjW@ zqZ?RuHn@IKSX?ehEHmD-bN;i_Oz=I6rCX7Dxh2Wc11(28of;2)iS0;2lol6eNjtq5 zuUj53ERiHEc{0n{1MGCYa^DHM(3en6&A;m9miFD)O`KcigMGmgQJ#b-EqTncw$hbL z&=>E79(RN#?V_YF8HQ|)S-rkj1}yDwk$SnM(4YM$AB|BVF+=PERM2dbwqL2pboBke{IieYrE8geWcbnqXgQ@2@@< zEVipleGnG=%M#12R_s{0>2sS=u*Qx!kb1f0{#f>8*M|I#d%?2Gg@hJcqgTVZs}*40R_VN=r7gR4RSP8Z7#w z4woqW^Llc^6-i%$0@?i+k8U~mKfRF?Nxj^Xna8%5ht{6Iz*1=j2~k=Mn&KL>>GEqe zSO&MwC@(C*S0$D_Q+5tht)LK}PH)?p)XOb#=Io7!bu-Q!085n~Bt&TmW|jh-Vffa7 z9?b^V5El0YiRF6|+b8!3e1iMrx$Q~4+;Sy70*{rCEu3*b)=`gyC@qznVPB3_&~k>p z=$)`>C@lFhONAV^PhNTK<`URm#zvBQxn)nr9<*FYK05~b(kqUHC@t2^(yGw}Pq5t3 zzSv4wGOkJbVj9HuW3z_EEe1<^eNr#C#9D4gOa9UhI$){Pf`lk78O&m{d#f2(4t3P- zEG#jJ63d}oY>nM(U3L>#!u3hL+_F6B09v+nH&_amxPBx=Y3bD**VurDVK}R>cEa$U z!s2sXV)-_oy?XzA@*1=>nnCL2mMN#$4ct;kf8tYWE$5LCrNxI?^d^jmgubk9Z$C&_ z9FioKCnoIeCl#$v?S;NX>5+Q5rTZ;U?2Gf~RyeB=*N22CEt)N`FSizl;=Lpa*XK?&}w}1WTuzq+V``(7%qB-3_MTeyr?05~8#uF^kKvmX=^SyMF#mVF^u< zSlZ5Gzs0Jh>a+&S$*rVbZn@>Q0WCWR8kB^-jN3^cMoNW%1_xrDdx)#QC%;1`^{@10@%1oje>hmcI=UZMxvOC)M6y9Sf@gGP~< z|2BF5E^;%TIo0*qwpi?OJZ?z#p3D5$okh2mrSaG7>7%4xzW1E+el+eqJ3Y);4&O_k zlO#m>y?C(iCEmLfzFWOq>sVW1DUexoa@c_8yy7;-a3uKFiPXz2Cym&9NyAo$HwVl8 z?j%HMDPR^?CwKfex(f+;>xCsVRnnKS>)6E(ujV;8t*(8G)XOctj_kdp>n~2Z4wh2) zNr=)iwG|%K`WknRg(Gz7n-jMROWaL~#m}1cU^xOJd2X&-(IUo_RJ{Vw#Y8x#isU_qe^3II}()`V!HG zgeWb|T4P^cH9tBFEH+y%9}pJj+Y(FWDt6|ybJwjgVDa8b>gASNgV~u=#QSoaz!JTm zgeWb}%<>>L0I!x7Z`UzESaM|+J4beBanaMyJ;8FiE2)=TX4H(rzQk!Ri3E#pZxW)k z!7fu?e}DE4SQ>?q5T#{MJ6vPo zjmO}>1Rj4}Ek{@a?@27Hm$0jH(SiqSX|iQX9XGgeWcL+hbo` zOgprMgAT}yQiR~v9GZY^u_B52~k>>FiX`5 zWfy=Y#OFW%@1Jd5Jc+bmMW9j`(p;@*THv3wGJa8N=q8E z9KTw;30RJQtzSV{A~PkH(%aZn^aAVME5Y(!qrqSGa?8xq>DZT<+Fp3IG@=X%QCjpn z;2OK+v<4@M2W?qbLs+~XN-Sn$*c$T)>#z$f@jFSq+|sQyn~MHX*mwu5v84SZL}~G2 zma~Sz+o3Nf21PU!7W+pM%k{a;GQPxy7_eL)Me602)7KKPFRfA+WP#<{1QMdOXmrHB zq}s*eoyGolUbPaIY?-C%Saz4RS^5rq8mic1QZKg*&}Hu_sW2d<3Rs@LAR$VN8M6e{ zuCNLEvM|b|v#=yQmh`2?Om=P=*u@_2URz!y^>WLDJ?w6r=4!*YV40doLX?&SW|?^M z`g5?1d9%5vumopGEJwewspx(ytn(kcd>7~ zg5_#t)zQL|FSFFG$$leSEM2<=EE6u0dbwrG{jF$u6g}`O^u;WZgeWcf%%Xqogb!GB z=1ejcmW-#8zD#J&&d2sv+Ikl(WfqfqxuyFo_J!J&~wl$ID~X*7HazRh9n zm3(ty@yV80LR&J+IkU$Tp)c{Nq+V|M)ioK{*or$&cn9lD1_@DG8g<6LOd41jpHMw| zu7#DbI6RkFBIdH|FpdXeSA!+}3aOV{^0On+azC*@UW4s)gM=t84$Sg?x9daLUW__T zwG);cnZ;lLJH9m8)hZ4wkv&Mg+)^-+Jvlsa!I%NCy%Y{0AxcXQv)po-GY>4@S5G<$ zOVSHTU&8yda~Koj^bKI~PbKwoOU>=jZG*%auHaD^!wo| zETJzYmYe46__8TD1>ffIZ5XMSTV8ZwAq5}D^n|_~HzFZQODMCPX^~O^EZ?Ve^%NG5 z9EqjUA$C4i`eJ9C@(i0p>gATi;_Nqa<)!&oz+$+FgeWc5bg?g0Q*vUGAz>+y zSz0@?skKp_-Zf#3_32IO<(3-*O5u&+OIZgp!P0Cn2~k>XnPp%>Y5XnLZr%AHVaa?Y z=}VJY?DxJ~PK%LXiFGFRa*MlDKJLdZ=9Df4mi^mFh|-eDEN?&l#3$~OC*_|Kmbljv zOH4g>-LkFw(~e+iH=ESUEu#*1K+DjHtww{zW+4etTKaXxHTJTx)d|>MToZdyEbM%fVZuUT*OmGzvqs?DZVMa^yYd{lGMvBb}4Vr61`LxuaBigkr1UtOAq_9VaHf}dhp%ri#LQNS7xbd z%O<^gMw+(*%fft8FSjfRU_XM)dN`MZzU(X_Axet{vz%}7FbVo{WOTKA!jk${(wE)* z|6pypT)=xtV@ycB+|tB)JNBhcqb_)!{Bs5gQCd=&#Z{{~{=enA>eHSIOZYp9>S49#Ik$Smh>bA{jX*TbiF<4e-k`Se(h*>(ms#y#yr3dQ#6qc+!Nng@`vA!g~ zx)=d{$s0-P<(90uY$Q>~d{1$(_>3nZO3O@rTw{l(T*7O8AMH#_7yWrXIsSvhlBvhe z`D<+T=nR%+PNZILaf;f6eR&%D34d35yOI#4C7xN_3LAcfzVtYLrn0aEew0{R9b*&T zwW=M#bAH=1q+V{BYP=sUsh8^D@ulf`5~8%U>W+PxI>RdgEPj?6b%e#`lfWLC3w80_vc`NLe2bHH2nQ^*%Q(?)ISz->c?WIo2 z0z4l(8%yftmN4HW>`R%7y8>Z*seYA&C@p!+@~I#o9W2{FY-=YhY59`A6kKDL{+9+{ zhV3P%kkrd94HJ8#CCFj>C$yAs{p&+0E#rIO8nb)!^h8Y#W;LA z-TM`(ms=jjx?^7=dgXM7z8wBQLX?(zJ+Uul_G9NjUlLkOHV_v3uM&%i@l5;}*(e@1Is6U5~8%&GmA+>hl*fXG%>(PSh8i7c7|*}wziwcBiLS^ z%pmo0i?;*Y_Lpm9;qU#Nc_c(>$!3a*A6;KX$N`P?cn zt-jWsBq^rVyN8s--SgwUX@g-}{g@|7`KM|1`cJ~~L#-UW@;*+h`;fT5n^w1Pz&=@J zyDBeWT76>xiTaK#FX#ESZDT7AH11~iT` ztpn5QHDXD~f1Fm&TVx*t)9NiQld#{X)k`G0;Eego(%QzbFypS1fWMhmZ~V*^lh*d~ z9tG3trEig>e><)I+Xt*OxlRF0tKYsyV*hzsJu%S{r`3x`-B||H>YK7i;J=zy4;)gp z@9=VE%ako`HG7uDtYua^{hjdnoE|wOQ@NMu(+l_Bi%X9A2-E7;?@7#mn^vE^TSpi6 zo*|~bGsPY!;g3B|$BWm+w8d_hirb zRi+iEOVt9*P{W)V^0ng5NjwkhUOZs33 zJj{1_j+AKD~DX~|%g?t0t(!BVZ1j+?N=6iF=0>awfYX9|oHz!Fr6)XObdheFYk z7y8W^9huF4u9} zAOC}dC@o3MGNVFz0azv(2A&g^Q2P?UANLbqvSa12m>Ydzjk!)D^>WMG#CYt>AA@rt zu)TDhMM9Jooqo8+{QG9s2g|QL53UG{$9jpS(>8YF*|gvcPTpG`BK2}h**Wa;aluf1 zeCnp#Q4*rGcrZ(oQ+7D59%@`QRagpSmcUGQ^`~zM4=?D;$C;#FZmHFPjlX1>k39%| z$(~O_l$HW!d3Ae2d+1BpFtdBYlDR?B7aJdTtb9N73%-;1!dOx-x0EVi+sj(LT%5k$ zJ(+|kEmQmB8VmmsFcd5uANxNQmbi@)i-QNV)U5sk|GUFAht$h0OPa8k1HEr;jqky- z&m$p9OB}PzYBTmJSgu!o`bJp%93+;!)$A*cSUS81tg)VTNWI)Lr|MhW_U8@g-U`-O z>82z^X=ye9`|_)MiEOY8?J@GRusAzPEN42fpB`F1_i!>YVIZlOTZ~&IqUA)8A8s$s zBS?tS;>;{f<63!wrT>1rpTd$Wvn+bfu9luR*^cic_wXb2a!c6tU1)jbJghYIWqdFR zQCf1DW#FLIc+MZ@c&hY|KkxdaI!XGnem2{kHgA7u1gx>fZlqprFnyWOd0!9TNnEiLsh3+GXH7uMyi$+KLtmzL zCm~8ptwGqA`=6_2L0_82Y-lPhMKX)iX|^95y+i}A{O-9->gAT^PFK-l6+PetSRN*m z5T(VAS;{>vj!$()jlJJaSh6-t`ch&j`z_{ta)|+~vG!9*z1%W0Wg}V|&Az=KEIMW+ zL}|%lmhjdc@UG90H_h~fC4P&xy#We!U|FR{LX?*B zL$EKFO&j7zVw+p2R}yGwV9K_a=sF&ms_@^w#5+rtTFYWFL!H_5TzxJ zS?c%gr3ID-Pu7?TOQfsBvidl?)|aw-Yk9Ercunf%mJf$Up{4VeY@90S{gH$yE&4-o zjk&a1e;h2@35w0JSg z>SZTzTK!eoA{$|`cb8ZK-PrHSB4-EOkM*lY>gAT@5AUMoutO+L6cyDaAxeve0rsVu zt}i|X-TUAudtu3zSt1v*dlv7$Rqh9CtV$rMms{>Turcu#7b@aYYM)M!5T(V8S#-RP z;)HjpgZnlMOTu%#vAwO-N{*DQ^Xq*GZ&aZYi;w^~JnWbG)AXHjRWRE&0rn6k&$b9>*8# z3lWwK4@qD2PqOpm%~mB_f#tn5sh3;og@@uAEBW5PDfA`Uj)W*JCc|-!>6;wHYp|Bz z-bV;a%r1$g9)FWd%JvcXA9udR^Zu%rTV6Zp;)sV|(K?)3+g+A~C@nF}Vq`lKr`Bc` z?|ng7e0EDLwL{squU)cc59o_^IZ`jT)C)L(eHlC54eujY)FvTHOQR9kmvJ2?8w9|CLaNedF9v^X%!$CNQnupg_n z^U`f$$&p!X`>{*G4?J$AgXM!4sh3;ydTm9E+1#xiz~b&lLX?&qX3?{qfNQMP`>&b8 zlH@7r%R6qlS$sX-ORD&d)XOcpr`Zl(^ z!VJ~)|mZAQZKiR_uY?{jxDyl2a8Q12~k=yndSMEpAEs{^~1cJ zu*7*uEUC%teC&BYWm>){nk0B zs|kzW0g2_k_5%DE!AG9Wg}#jaK+3vy; z?jx}Thkvf@1bulQLh9v~s0(Zdaxdl4bm+^a2oj>S=o;c0tEgeg{ufTS+(2RR zJS4H?Zf3WDOFb~@50>w}Nxj^%q(Tck=Pz6wV+@wm!6Za!@nn{)HHYv&yEL|$j1-n4 znI*k6JMP!5+^8~G#_cBca?90bZ00H`z0NYQv^q#al$Ii9IW@!00S27c_q}f{EawlG z_`CPSmz|!1L+*VFx*mnedzWD(NilhUwX!C*XrsyZ@-TV7(TJq{=gIrI9zohLc|ULx ziTk_B`;n<7@RM!3dL$es@4L<cgOF(TmM59VMRS# zOoGIU`fc+5@^=Qj+pMt(Ox{;sK|=oH!eY{aq zbC|pzvX=z@tI7Lpi)J`^zkcv{Z`ey%`H)QIUZSrN?!A9Tbq|Nh`?Uci=D$tekDfBh z3XagtdybwW_BdI8>~VgZykB{Q{nA*Zm*xa({qP`CFW-BvGVsMCbev&-oGO?;l7uL~ zmznH)dD6V5FIa|sK5Ql|@xBsE+e>URvi<9$!C*=MN$Ta6>+`px#m3MI&zz2x^8M>W zC@t~K(&}@^nqb*h=H((`3G|a#JffH2$JkQVb1_(QtCD)TCGqNJ3>j$h2(SE}tV2SS zmR4hNz3%UyRs$?lZs|~F3ijo;?ZmTS8Q+tHC@p!+ zvP#zuxBcUV7dHz_n!ltkVYQdz#~5zopbdTbUgFYU^>Rz#jx886VB?drV0l%JgeWcJ z$Ke_adXx7K)>z@-Z##q~GC*RvvwRkQj73YQ7=oqc7*a2{Sg&Kx__`M4eFn=2V-ljY zL^8{MtrT2iRl1G#5*DvOiN&iJyD6B|Fm@PN+V>;%a*O7P9r%SDozxtky?#EFgeWcb z#$#VDT=#4Xeeo{lc|=(3k4h|&AJ`9`^=ldzM@u*u3&lEFeyY>vSpS{P1Pw(oOfTeV6QZKjYy)neT%#U`#U$c2SBt&V+W|q}K zJ@GDS?3l)p!jcdq=}QkyQ~VfOH4foBiN{PL^>Rzca%|G8VWYH0u)S=bMM9Jog9*6C zD%aoM0{Rj?-SUF41Rs-FK5b=d>}QC3Rp`sy`J`TMnf|~4zmRe#Q*MLBW(f&VT7sFy ze%+NiV9{*gn~ee$7>Bt&VcJQ4dc zz1QPXU@0E??Y6Mw%Pc|9*s-!{&CeUbGV21Vms^4@&R}2qe!Kh;`cnNG2~k?CnWf*7 zYo=gvdC>8puw)#U^rfBQQv4V}Z6417%dKouFSmU8c@{(72Av%W7ROu?qO@c%OWuK2 zqrq}N*!qRA#DqvJ;|o|{CSL3C5G)C2NWI+B(>ekz9~*bu3YLE7Nr=+Y%NW;KfkE$j zU~&1F{9agmPDm`tv)PTbVykBdg2ktZ)XOdBO3X%!uIrW$V7a52^4Et@T6~!0X{-r8 zMNl}S);D2s2$fiz9ysn1T+nA%ORuyms|l7oUs5l(qmiA_kjx|eATmS3+(z1-3*cQAe-N3zea21~<_Bt&V^nZ(xEk2?|2 zmy=x^HxL$&Qxc1#2eUjp z_?s!dY1wOlrSS?%lCh42C@q7g;~MKYZU9b2f9r5#x3C1BlURCw zV%L;{FFP88C0L)-%Pmu1vEO0~KD3?ymRbWyh|&_sEH_pks|(vprHa42g~cUSV!1G# zom*b_s*U^P2{lQ*+;Xse8usOBAD50`32#V3l$P={urGHzovr|u%k}#m6_z}iW$aY8 z?fWm9fdA85t0k$ITc)}0Ma$3#t5;w#?MOnDmL<%R@Vx&Mu#Ae?9VRSk=OumVHJ(jH zuU%&04{L1MWl}G2*a-j&RZ7M933iN#-=o%5Uj zdWhHh&Q~J!a?6Hcr_s{CVwDwOSyz*UC@uOkagF808{*r7Z0gj#EG%9ZB$iU8*tzA_ z&DOfmmo?2uz1*U$$FB3o>f1Sj#ibnyQChs1Wx>D$TVaju+OQ}^SnMxKEOwpQ`B=Z5 zr|{nE%dMndZqZ5lf_?G*;F_NDEEGx=cg_!fOvSh8i7{>|7LGkWQ9 z5BgHQM9g3Ha?8Zt>_q748Xx>!nOc^FC@p5pGI4n`{12l2IiIqGCE=2!FGC{PDXz^i zH*2t5{XpvFmeNnz{r#TXhv0v8Zu&+-l$Hc$X&7&V{|!~?{m|FK5*#nF1omUsVTz}J z^974h0jZZ;`o`?UH8$J0XdtYyPz{5>K7`WJ&J@>}^|fCeV2#a*-|&QMv z@xpQUu`lmPz1$M{vL~*w;U|7x1k1)RBt&U3nT>0#(Y|JQ+~4(nUL|3Pxhk>f)L_S# zdG`|%!ID!z>gARjX6(A9UNZx0u(T*P|E~|Bw8St=o{_;V=*xpH2Wtt7PlCkKM3-H1 zTzLMpEm+KYlX|(OM>QYpOXmV7ZLmZPCLu~oBQxxa_Q}I|4L0)H`^LiJa7|+ARCzUi zjBh)22Z811EmAMH7&sSUh~2ZWG_ZKyCm~9U1GB6c7L z>Ud4*a3rahTWsF?p(VL(aZ9iqjUypSOAfPG1Qp=6zxTjGU13Q|l=S6^E<1;*(em>i z=u4RZQZKhG_uGM%YuhVof@RzZ5~8$>oP%rZxX&wp*j^?)jO{Hfq1PprSKp`M$7pib zX(3qpz998-OZ9bZcq5?IvLLWjdPhQ(mQZF1YNG!dEUwqf3>6lSB#EW64m*dbc%yP{ zu)Iwp^>RxyJ9Zsra=-|@Zt0UrLX?(jbFnWW-=fRIfb-<&;fBI;{twIl4LC2GWp3GV z*6i8Km(7~J^8de!I_tRyPJiEM;a?jj?{~(NB*o-?-xF-|KDyGWJ}`NI{x(VZ&y)94 zikG&8$@@B)B<}Af?+p|AV0!aYu`qf6>=}vrpC<1OJu4i7$@{}^Nu+x6zN>u%e)kP$ z9Kxwrqt7Hrtf=26?~CU$p!WDat6}o~O%VzCkCXSiy-#j|$$MkX5r3_#-zV>l3XkEu z{8Gai%V1$vDn|nTX7Yaef)GraXJa%JCht9}kfeV*dH>r7Y_jCg7?`}*X-{JRdGh}2 z!u>dT@0Xfj2b1^X^+@2qn!KO6PX{OO=lzJoe_l%HOEQ&viN0(v(Wh5Zdzid0Hq)(Q?^)yGIy{_oURtFZ90~HA zNr>`$v1Q+jamgA{a3rug+hc~XWF||F1Qt8kw8y!C@%WEaUYAI{+_F4B5iQ;C8*T>6 zn(HJ)X~|@kg|QRyAKQcVmd+QJxD<)yayRy?snZgx5a^3>KT8hsS?ZASGM?-hDJ3%2bLNANWI(=>&X5OWlnu7d?sGk zfP^S5TJy0l-DmdS2z@cs9q%A4xiX8jKfCKQP4gDsDGBXO>gAS}tJzO4^JneSp)WHA zlMtoFf?4`@zgGn;D~r2q6PDDQlD>>u&n8xG3?$^NLR_EDnVF|w_vDDbS1V6^Gm*4R}B8E33^>Ryk20Npz(9dQN^yPXx z5~8%|n&TR?G!9z=77xGLKEmR8TVgRB$Br*!S2xEeB9?}bdbvfT9lOHl^DZd_EWOW= z5T(VFS;pL-wFWHrcFsO3EJZSl-79v5vDECc-(ih4JV5H@mL6|nu(^JV-SffXd4z;0 zEk(?-s35B(SnStCg$YYmnxrqUZQ1cf=hsMlANgG;QZKidJYzfd64xqPf#r=W2~k>R zF2FVReo>n#U|CdOGe%hA??^1cmaH#(EQi_2Y!CDD*sjOI7^34LkSg4D|`)6$)=FVz};^n*3lvI7ZGT3ndL z$YCD-7eVlnCwGME`mUs2Zb{s>5-kf4THwEFx8F@dl$P-pxW=k3 z-VqCy?|m#^3rplZiKX)ewjV3#S$!&4bPY(o+~R)D7cB$5Od3F6${3Omr6rPC_IsYe z^Rcewy*~+y*L{g4Y6Cm&2U`5H1B*{}QZKhG?-q`hvU``AfyK8z2~k?=EyBKB_)vcz z^riUPr$xeI|3G5dznD#}Rn}{P=axpBNWI)rT#vmVWKo5Om%;Mhm4qlQ_ROMPeJ4KM zF-Na+sbYWL^~si5wui7OoK?kp;Cp=n`jUFNrI8bR>ZXs|n1x_DGn9lVE!oU6yp)F) z^rcA8v68SPWJ>ySGMn|~olD#kSYw^~k$Sl$a}V2(RnKVZ4i2gFUuM}E!=|Igj_~RNeYu}a>gASF=4_HUc9oGUSa#=<5T(VMS#;{npAVL4N8ELV zCF8NAFJt-p$*K*{tOFLO5K=FT~Bu&!uX&|EtJEFR}ch|>IRvuaX!Imb~*ML}~G17TYT;aGJ2VshOd$I6ReD zE`+n=%cKgqdSKbOkkrd9jxP_P<*ku5KHcGKMM9Jo&866vPhl>2_d3KZYLc+z$Shv2 z%yQ5o3@3@NSdx0V)1_=q)4la*N4Z_Vi%3sR>>iAFzsqC@mh$a&f`nk6;PZh*>Kv1v1O8>Fn z^hv$kV%eVkaLDP{0w*J_`;ic(#hF=tO$mwwOP#~5j|)q#%u?wV+m9`|wJ;z0QvNup zms@ISZpCf?RE@aS(3jz7NQlys%PfPevnGQjblc1`!jk%0(iig`Y^l#KG{<|dU-poC zxuxVPf3$?uZ&e&D-@HkP(qgzA*I2~SJ=$QY*8k*rVF`aDvDjF!`(ih}_Tv+(9>Yn! z+!7gOh!)$ORq=lmSB)hhN=rDiG--b<5G-z+i(L~I&s>S++X*&-S10WdJ~`ZcC#jcP zI(*xKmV*^OXG32$?M_t)wqIW?AA_8Wp+m9axU_A@y=g z3T((1s{D`0s$goG$9SE~ccu^=Hz%MxZ8b@nh`Q~J33MHyj9`ylDdrEcuEn5#t)-g{l=L+a(0p1F3| z7rlTX{lPLmkc22LY0TnyL(31=*z@-tw1p+|qr|ezp8b9q*>1NV^yNVTsh3-p_h%<{ zw{L0Te?n>$>-g7)P+Ig?;u@=K+X9~;a=NmZ<(7R1 zu3}%lw@ku&Nr&!|5T(V7Sq%H08v}j0pdZm(SnTs9mKP3eKQ=7&`F!Zhy#b_NZkgLX z3N3CG^6{F|jo~CjY0+4Ped$}v!3``QK9%VxEZH)PsTVs>?o?XC11ueXkb1c#uPeI- zt2tcrHEb{AN|ycWLntj~%wkc$GTuk_syS6(SQ0)<`ckbb-)t=$4uU115viA3BClP+ zzFhJz{|zjoTagf@C4pJIMpg2JzN8=9(O+1Czep@8R%`HMRC#yB2rNTRlX|&j=TaD-^)x^kA}(n*S$!R zV)A~@V>WrOlVcYIllP+slazm&yw5v)0zcGVt)gh0ydOp4{%-Q#WilqyeSoxK0;16mZHnFo{icb!P=KTqD@o9K^|_jcY-QepCb(KZtJuO{!uPw9=5_rouY zC=Gjw(z{5eaxc-J?Ijk?%en!R_m#Xz%zvA_9~S){?=Czy+ifWJIPPC1d(WJ^?4;*s zlRCKftYb;)<$KQ?niFsr+^yy967aozUrs`l--|o@UeeAk$M^cIdiHXXu;j}u`nA|q zhO&EhP(kO#w^Q^CU!RF|os=TKKPYe8=jT2JzH*AtXd;abT9m$rte*t4}<;ISWgU%+frX-HAU| z;i@}WZXYJ~a!ZRh`_WRnPW7c=i3%biN=pv2EUfzoe*u2`yluO%Bo#^eV%L@}^=(IL zb_L6)U!-1c`FxIDmD7s2jQ}Uq0Hh|FDnx)D@$to35&;1iKW$Ow#E$Y zX5c&UoGX%gxn;ExoAPY_ZXWK(s@5bSN=vnM*q5y(@8Y{Xk`v613rm5_vN@AoG3j}+ zKmG!|cazl1E%SP@ZNEp|Q@@}u`u9nQ(qhXjh4bfifxc)x_c|jinZG1`aphCd9Y3G! z3v2AjJ5n#Vw7$W9%^uAt>I;@$UrC74lF2M7Z@Y#PZVV{Ls?T^lama~-Ld zTk7bsbIa&7+gz}?Zz3T|OAxax$!2#+TV8z7UsxiyN-VcdvFk7siiMs8%bM$?UT&GP zknNM(jgQO#%dXobL}|%n7Uv7bQ(=1vf6-)wuz0yhEc#2?q}LpyM!Ue$HILNGE%io4 z;`Xwl+ys1QgAS;d)fQ^ z%Ksb}2$t$SNQlys&Mb8-YHS9}S@Y#9g~i=nVma%=j{D{J)WFHeJq0`5mLAMfqSNyqU}?9$x{I)6Y?t)q$};v8!Te1>@%~G{Eu>y`Zn??pps@Jtlvp~qVCOJR2242wmOyt>FSm@S$bOu5IUI>k z4%>N>5TzxRS$0+|qYG>7+?DqM!s6f|vCK1P*T*i%uj7 zQCd8irBwUT}SULUkn?_iz-mi>t&L}@8v zme$)F;@m~qLoSbnCDT*Vmn}Bz{{G&RdNZLfeS=87+;Z4`BU%_i5yeC3mcFm5-5tg{U5{u0XbNm=BUOmU_F!>gwUT&$7;DI6GgFk%-%fRI%L}{7H zEYEvnEd$Gs4aGkSi{CzprQSDo9cE{#6L>v&qzkE+TiRJ{LQB+EQ+yJ#lLrY=T6QqY zmWSbuz|!f|nD4^kykBBjIh1WLCmYwr_rzC>A@y>L=0kS!aC2cAzQ5lfo`fhZ@yya~ zTqypB>b;vznwo!JPtKKD>~^!Cw`WdvX#(3z!v~~ZZn?5%JN9M%{AT#3uqMw)h|=oA$)sLxsg^n(Ej`;`4TJ4v?;R4Nw6xlU z+rIbM;@!crD7Z~cVF^DVv4m-|^W=HimLs7r5vNJL+>$yX3@ugLuGtTk)#pfv(lVJ@ z(sjFy2TRy+vqr+=c~D~cIFkJqdlWdOELb$glX|&DLyMhT8d~1y2A0SvBt&U(VU|W~ zy==j9AuzbLuoTHG`3Kp3v9&e(RRN227^# z6l~tjsK2mW^3nXeY4yU?44jzuux)+|rqw_8BT0&B^=WI^3uz6;PAUP@>U#}H%0Er3 z*MGuB=$mU?Glpq(cPA3}chl;Q$!wTAJUinkOsl86lBoY_T0QTR?^u{tkJ(Kk)zj+n zWc4%!3x_SkU|QX} z8HxSpY4zNuCvjT+wogq%m{!+rO9KDZw0gbH<8fO3ulC(h&->rVNb)ykt?-_VVVi{1# z&NkYNpNU6;oX(_PZkcFhgqF#Z+v8nQt)3)AX&KKf-mOPrUoKy_94{;`hb5L5y_m)1 zkP}`l?SF^V%PqE!?Ckv6#=1wKFHIkk5T(V5S?ZsOsts$*%*J<`u;j@s)$Xt}+8rA- zYr_6CY#phWTV79PXHHuq9^f;J%Qle^r6rPCimhKc2P}a{-^~@4G+#+ywvK0Csmu6^ z_^!2&VWeJesegn0G9SIOeF5|(>?{dUT3$0ty{k#nz>@CXcZsk>`bjKt@0i7E;o{QJ zm(BY~z1-q&&nz{wL-3czu)`!oX{omv_lXNao3;W=twEM6g~jWL#Ii7s-SKVeyBhx` z&~p^2ms{dDuv_33`>OARzL-oPAxg_=W@%H)x(F;4+n?SbEcX5q%jEBDds)4>0RQdu zfgY)sTaFB7mt0rhu6_Y5_xg|!rNy3EZu$2%1&dxtX%}J1mRXuSVpp=BRnWqTRfh;t zFSmrfbisY{!_z0C!BR4ggeWax%#yRd?{=_^=&rX*SP}vxeYvE=`r=ceiz#d`-3O3* zxn=jA4QRR6^&UtBSoR+;eNtHRWtJ^<*wxZ5D`toTCyMyK3E(zaNaAXvzms=j3 zW0vP5zFC50%LEdlv;;HDrrMJ{z@nFM3y5e1*%S-Ph35!p##4;(6ZTt1-gy1{x9aocjx#ia?b_r>* z&E3<`7oCkHL}{tK6}SCf&uqgAU83)xcy zXF7IR2$nMbBt&VkWR@{O%j!a3w(YK6Q&{4{B$lkZ?3$9LUSu(_gn5&Cx#f`^yBoK2 z@nZbLx&3z; zL0>v{At6f3EoQmszWFLxcI0etEiBIA63gk;Y>oZgH2paArSJo(ms|Q?3&QQ?%T%wH zU@7=ULX;NGZMf}!nRN`G==Rcm)>&9`WtKogc5dmrWFAfxeCgATIYuJ7)$FV=Y zFSdIb5~8%sXO{Nizi>Y`D9dE9u!NtHSb7X)=gDqvU-2omdap^n+%m>segEO*%9bfx z+G_SJi&@L80ymezbN-|cBt&W1$1FV@n)HIccsK-)78cJ4i6yWuJ65)T-XEX3xv-7Y z%PlAK*tzA1%hN8w8oRlRgeWaZ%u-=>-ea&_=>FJPSc+tp*>P<0zVgR)yWLS zGweTyZ&onOFLn4iI!N>-m>2>V-IYs1QxAlq+V{h zIfz{!^IX2k8!RtglMtol3bW+j&%pod%N#P&Nm%k^mazuxxZkGr1H7hG#*oy@ExzAQ z;2LYPECJtrn`}%%l$HW!Ihx*iJ@jRH&RRENNsE#6r4fI!t3kC5_{3eqkEC91IdYga zw_oyw+0d6Vg(O62Y2$|5{`x-6N`a+}MY5-`M4ppain+6=Zl-9C{sL=k%yLpMw+zT- z<1gb6zQYN;N^3}n(lV7<%)X4pDV%CK%?=BTSFFVHt1~-?@woXC?;ScSwlR62~k*>pjG`o|PNu6)r5r;Tzz{^SQ6qSeQ^z8llM{U17<*9PNkB1xh2KjA1$F? zlY*fydoxIg($dTwxBdMGTbu*ST=SM!g(dic#45!jR94y8&Y*K~A{i4Kj@Qp2gr5T^*j|0o7`J`TMx$o_W=lt%5 zGtHqdd_TP=L$m+mK`fgEUtDGES9xMz1&jmAe-$eW;Phl$HE(t z5T&KTcHH(uK9{%;eQ|7h__MJ1T$Wgx9b@~k36-bdf2huFPwM5CfTQef4#wfVn?heU zb|oQ7iy^ZZ&7R^0mJcy$KZV8Nio|j(itWd;bf)3cP(@crz1-4s4ZA14?M`2OYry6k zBt&W1z$}jkN8$D4d#h?_mHP8~a*oWBx0<~aZ{MM5_*R+;4y0ag`QGC`?hNbBorUMg zTU|(q(h|-r4iV9*(3cg_=2e6x>8hkJ-i6lqF`h+C`V8C4w9BMkZkZl;21As{67@e#-rIWZtqPO( zmd+$nJ$e5dGj!czoU*9yPJ+aW`fc+5{Xqs?xAjIQOy2wNAtC>9^8P_F(Y)Ylf5er+xS0d${s1lT6iK!rqtdB}S)Y zRf5TT*CZ12-zM)nn7C@d-qU5{kDg+WlljLU=l^kbAAUW5?;rT>P*(QJ$|l)cMs{TH zEqjx_LQ+(cN>LO=k|L>)$}Cw$i_%gFA%qZqpZEFQem~daa?X2v|A6bdz3#8;obz1Q z^F06Pyw7XQuA0s?eLn+M&viB=UcP$X`pT{VuV*=9N6~Ial0@~t6!T;MOT%vYWx?|6 zLA!y%5^+~zdC`kKjnk`k<0DvFKOymQ%cuZHJllA)aY;ouqt(eENtBj!W*M?#3*IgL z>gGIBSUm4ZEF+JwC%(?79Or_?^%RMhTRbBz(6TZ2V=`Fc{Yes~Mf)J0ZRqq2z|P37 z9q&#Q7R&n*%dQZ1rSW(`W@WI{?@i+6mM&j4@KEOeu@TO_5(7yRrNxk0x~_S(8^+kt zM>S>(OPERsZNS;s8t5plJ^GN$#? z#ljNvKvI`RUTp5CbwA!5EC!uPyxcO`n{|79zwH(emI1v<5~anPS?1RpcnmB)ZpYRL zOW;F^<>wvt1S>i-!5S=ur%1fqQdBn)>r$-eO&hR;`jaF|%Ts3Q8Yo?VDmz1F8XSUS`r@p8+@)X_L!g2yJ|e5t2Pk|-@&4mkJq3Y;Cmq8VghAuL5Q zOTl(_C+ptM&iGx0>)|9`ZrQttJ-u$a`o|Ec%e6?7L}}^6EWKWYO$19(_tpD_B|TPB z7u(is--Rl!qwuO}kRgeeTaI}(#JZS#xA+E@$wnlJ(z23S?p_Ns1&iy=U}s^u`AA}! zG>Sc4NM1b99>&<#`y^g&IsWJ<&KJYXYpGz_5KodQEytNfZ;&$tFZKbBak z?_^hiX*Ug;f@ND*5-+zn$FgJPC%-+|Nxag4BvD%8n5Fr=MU$W|qq+}1BP`Z&5=&)& z)>Tk-`H_uanK^*O%PpTYZLlt5!|Gy3k=t;RL}@AGh;zSu?TxsH_l7x+fx?n6vm}*Z z=VJv{3G={`xP-*ZEvYuFE*`pl-oO~sG9^irmTt`Q{@ByTP?uIeUR@QIxOhojcHLkd zMQ6qwI0zP#GOvHf%Pj%B*dBN{Bi7yri%Uh4L}^*fEC=7UIt7*}hgy-s5|SXXJhx+a zvaELB#=CMej*@t}rTelvxQ;~~NxBY}2Pa7qrNxz5mIRN$JB-QgjbenwHBn-jk5-+#(JiRUWOu9|t<(58a>=8x#2}9Sy7+d;)BvD#Sn5F7)y|ZB1IAzx-Vex$`vHW<&jxVF` z9>rs2B@+@aw-iig1J&BPpb=P>FC|Ho7AIyIU-MK2uoT`3{3$GU$r4NL8SKg8sA(Sf zhU%+_BwlW*Q=)VBVH2*gm8h<_YDUeyZN3*)D-<5=W z5@&>vc)2BA&mPyY?+q*BK8vR#ND`&x8?*eXISxBki;o^%MOczjBz4i-%#Qn$BI^%; z`BHWYiI-d48_YsW>z5eXnGWybm?qK=fH|~DN z%Pkf43(#^qa^-xeOT32d?<7h~GiRLpE4>`>djGo7=OMy!E<HHjoqTJ|zaL1sTQ7-JKH=1diqLYZY!Ae;NG3f8eV?ZZjD+%k1F zyRK~YaDEP00wYNhr6rVE>KS)j2o?*w>n6gI@?27vSut$SNE4eDzA(l#50iMg#rgOh zoG)Ewf4l>hv&TsirR6QNXuEzH1D3>gUzZ6>M3%&2GLDV0y{7Zh!1B5qiI-b6Z?Oj- zUel5vfu&?Wl0<1~=z??q$&$hS!E*FXyY<52nJuws?PAX?-h2+iw+A17A@OpH|6F#T zzuDNxwqSYsizHE6#xRSor5E0VJ#pZ$nXp*qNGyc`Yzfjb?}hjJ!Vi;pxn+L!Lpa7d zt-CW5EOy695~anQS=t^mJ`Z(i6Mx@ISn_0+5wqF7zVPsu&tQxhrIL8LrPZMWXgT$1 z(O$5;&LK&ZmLO*FyEUjcSWeC16WQkCGm1g*NkUqnYh=; z1}wIwB#F|J%Pb>Sp2cT;-||gdg(WaoV)=Z2HvWv+nm_h|rRz5mFSpEV)drK?+p@|C zEX7L9`~4%7mO6)V?$`DkkKa<;YwF@9EY2?_7SA{AnME<%Z=qm`+e+f)mJ2Vv(4t-S z&{D8C>?TQ+79(aUZt?gejIj=@WBr7sNM_k-#hzJI2x%}BEVZ_fc)4X;j}vIg(=1sB zEI)RUBudLRX0ac(c`jH?CiVysmh@MWy7aik>Jql^g#lP5&LZ)0OUqY>(bA{IJG>Sf zxR4}KTFx=c>1I`Lf+gwG%4@=M^R>j%a45SLYkQ*#c13&sBJpyIUPM{E*O%OWm>tl$L5oaPHS?X@KADiRks?v9MUbkyrx0vU7g-9d%m37;_m!;^mg0z_B>S+E=U* z2$uCDNfM=H5VPn#XpeRIsyFPJu;j}uFVrVA1PF;^mf;&qt#rv-lu=usky$ zNtBii%<`y80G`9_p0O=QSmNGF>f+|Y&Mn&pylDxR+yx|FZm}=J-h@2VdBGvD6s;gh zlolUmSuy#|NSH6{GP2(ZOUOHk<);ff?zdiHR|J;tuSmSyvfG=jW1T`bb^we1N0LNo zNnw_`O)T+k+a{s)zY2?MzQm$Ghg}jL&x&XVmKV23yxelj&IZTWiQTX8y9zn?NfM={ zqASk**`+N)pe_^M8I>se*F9LX_Y%wcL#(&Q>hiq(V3}M<;^mfyw=!@Y3%Gm^`z~4) z8}a)`C@p=NrQDPw_|2{##$lSmk|nbQ)@0}WjzuT&jXSN$BwlW*x|-eRU#q4662{ou zStN(P?yc^NW9$A`3SoY^Ki(;eo&W=T}cw9 zrK}syeFNuv-JrqwwN0-M!V>*I%m3Tpyn6Pck;g6bAT5l?L-UpC$f7f~cbuW7- zpnB}p2he%n=`u<6f9kxiT7CCO=)6CDoupKE-tT>sip--sHVlLj)g_uF5+mx5&ilyo z3|Q~=i+a#`e>9dP`H!9VZOgu#3!V32NhH~yo%biEy~2KZ*Y}=tU|>e4kpzF!d7rlN z4QARga*h>r-Z#o2nf`6({U1LtyizYM=)Bi>OVa*x=Y5B*x3TkH?{?NW=)C_?Kob6| z&ij=$Hely{@lM$euu4oWA~}_-L{qj(q!ei4p7-BN_Wr#}{I|~gMu!4$PvW6Qo4SY< zr$DxPdgrmD+MRD{c!b_Np2W*n&ni~?@Tj)!&W+;mzl@zuk|_U|9_)WP^1>n=&YU`3 z_BIffq)(C~f#GwuuaEt=Asu1f&bvY4$H50P}+{@FVE{T_0?rN|z+H>bW-U5qO6Ou$}Im#@hO4Z#0mYm7G z#t4i17m20eE!Ii=!761N)aAxp5-+zHzpIO9v^9Oh+ks{K5|Tt|iD8zw=LtW+(yrs0 zslsAjD6vG>+=xTHacU6m>GiG`iI-b8uRM)OJ}-}Y2bRwRNfM={#8EukSYE2Y5-hL! zCz=RLuFRru!z}UlULS?Jd>u~W<(8$Ne9-cC*v;8sDHuzVC@r0sWw+mq*I;?*skvNO z?tPWi#rrP1YHIne#agf|^ds>KOB@@hV^>$fuQqQ9B1x2%`OIQ{(b^L%t$OxZFD&Q2 zNh}qT*x7lX>PxW8Q*$7Rms_eHd53jr65IznyvL6qNt706W;wgQA?{PMVyKsyusD2| zShj3rcNjzSnqrq{v9TmxZuvD~JX(60jfsFUR%`V9++&|`ieA?(mzS`3{*+jb z&t=E`3njnK1WVP+BwlXW7S7JcvTP=ng1Y>;Mv^Ek_RKQ>eKzhJx632RPgpE}Nh}?v zvoY4Nu^E0JX4@GOFSkUTWZi_fu2$&;mX#Ms5~bxPvy`>?G z*0JUdjBmggtNeh(%Pm#ku#Q#B+Uu}GqD2BpqO`Pd$GP9G(SiG5N!afjEi8dMB$o1% z*ckI@Gz|BTS>QtAvmMU4*0iI-be?ux`QR%dAYZZO7DK9M9!i#4;by>QYH3RMX z-@Z0r$%rBGa?ABk?A6dx&bx!4E_ahi5~U@dS^A8)TNcLHrW-l$gyp8W#L_U1tz&!R z&3{633YFa2tH1dt^LGqO>&jz_~xl>`oi7d~Z_ktFUN2h~NusojW0uM-Ot*pMm00Y4us6Fl9*f5g zi2)fTUT!&@aT4dt^@7yPP?ul1B#F{u#w@#?`+bGFG-+SFhp=SHEMKBnhs0v1Tfbnw zOzJ`6<(BX9>;(0R@#6Misnwq(QCb3+C2R4D8(`T#Vd4N`iP|Tr%Oov!TDX4oZX9C^ zW|4TgC8Z9#$=0!|%L%aPFCW{0nE;7zvB7wZt;v06Xr#ycCc3E*ut; zc)6up@3}a}N(7J81IwtDB#F{e;{?wA8aFIfLS1SYUmq_lb~X~r`c!sq*Yhx7$yQVT>7+@LeS=q5CD4d1u+Na{rqBc%Q$Y7KxWz zW`1Me>?(X%eIQu!YLO&LODeM*Ft333#}2Rmu}N6m?If0fT-JpX7L=n6W6XL7iI-dA zEY9EWlx;@Q%lUn-5h2*>6r_Q`2mT=?A;dp84E8p90!)m$4I=~ z;^38oN%Y?*<4$3Ryh#$Jr9ZP=X)|sTjImbH3v7fXS7sSx$BvaJ%W4>d<;`OfFSpd~ z!(Mfpyk^p87-J48B#F{u$}FE(ptDa*Nw6 zTdYg;tq&K#a$*WeqO_c3mePyZ{>qC?i@OWUxq}jmkq^5T%joD@8Z6n{NW9$Qyr&o* zU$(jY!Z#uJSdb)2OA@pAZ}8X(b@}zW>uF(eaFAHOjbpbRY_*r*ThQsBNW9z<|AlQh zJ=Vr)5!9vFPm)AwDer}IKd{;DD6kyNJQg4OZWo+OFV;?69wQ*YsM-)!iL2g2fcNMdp6$;Mb@_I=zXu-RA=FSq#FuxES)DV4^6 zMQbWaqO?3>md;sq@E+{*mF|hcV(Bcg=uTqCmk$+VyFy(;H<5U`rPq^vSeLf@O5(jf z^PMD#(o*Uq&izqlw%IVoZr6C4E-ZO6%ZNv;F57i{dxB+SLlQ5y?6xe9YjUFY?b}e7 zVJ%1!rKJnAl${lZeM(w=`n(dB7#B%hoUSlSxykcqgQe0C5-+#xa|*+{92mW?JXn4h zlO#&ZLT1@C-C-EirS*vwAB83Gu*BjV#BMoG_AQIwQcLwE@p8+ZWmahEy1BG2SSkgP zBudL+X1Or%O)OZp-gPe$7Uv@p%a;hYj=7pl!1Lr8@g!buDc^fPTK1+tnE;j=sU(Tg za*tUK1UpRyOVzEV%as4?esYn_^7TABhxs_f^D>OF(-tINZW(!)eYa=U%kS8&ZedH3 zC@nvk<$S-5*TB+bVfRYHlI|+0%aTK^E~5wB!nf2CR*`tQ<<9cqSeKN=$@8Etu^UMe zrA7Y~u49+y*_8*&sxvEU3Cm44iKU?#>xZdZBmD`~<=I6NFSqELvE5I)FFaleEI+Q2 zBudL1X1Q{@{Ccox^}63kSiFx)ELGgtrAME)t44subRdbBTbdTSVO`3ZjyD0z*%2g( z(&E4@(QQmiLWA>|HiugYOU(Z)f75xtEIA0fzm3*pc|qrWLRXSW(RshHD|;zYcj(s? z=)AAomt^^;&if71!|_Kg=+*Z=blzvrCh7jJ^WMjWy%P|*{!j>X-fv$-QvIJg@4qBu z6^G9IORGpqb?5!ka$k|@G5iO9#d_;Tl1PlGKRWNdZZKf2TI*Ir=lv8jlH@;j-n(1( z@`TR&=6gu8KRfTQfBK63@`f*K2E)L7U_%o8P3Qf+UOzBXs$X^rbl!hINHYD~&ig-p z;7sG_E6{nLe4M2H=gxbBj2P^^-(KzLE9ktxdWIzYSDp903U_1Y{qf~@dci6YdV%Cr zt`gpCmGGax>ox3oAAXsn`EQ-~D*{f1!s^*<`A&VY;#eP(te!VUv8UH{##rE`bq^5%(y{%ppmXmO^G(KdaFouq2#rJV03D+$D7}n!)a3U+}1C36^0O zNW9zo6vnz4p?-a zk$Aafc2o9r;o8nS55S_AO_C@rw#;(i!DIZo)|(1XX9$bg35n&QCVT4h%y`HWur#ho z;^me`*{$(v)_C%iEU={NkR(b=1hWj>HwLd}FLkN6P*}2LmR?QRlla^33KPIGs2_=! zTh7?`!1)NHjpGri|%Qh`#DWE<9@NlJwi7Li?5f&V)%{i>9wYNr#P@Q_95|d zi@v`k`qE7C-}Z`i(FsM1X#e7Hqt6*7pfh18{ESV*H?g;$iVf9|kZG@#jW(kOAYjU~IHIu>8dk~42Tk6$j z`(WAZb&UthzL6w}(sG$ubYApMfV!+(y4Fcpl1@qLqHoQfk`ABT7~i_tx|YPtEqBMT zcdr{1wtWqj-djl$r6rG9)<-O?4we~Vp~r+J)LUYyvWXq{@0cFMZo)5-BwlV=Q{5QH z*p4gZ@#*!Gha`#8qT_>e|I6G0+*`f+>U?iuaX&4wJlM`GJ7;XieM;_aCh>C1ytY@+ z;-B~tzYw~87fGVDjA9m-i<#JYKgO+BfUubRNG#WTu~Xa*+TKrLz9gO`@p8+nl5Np4 ze_$8fFSe9FNuspuWR}HK8cl;SR$y&@MOboWmPjXdtn>-_fNw$$b0YC_OOFym(bD7G zJ1>|oZH|&8N=qQK1h)uE0n7IAkekAC?~J4_Z)dZIf?AGgQ=u+ecSyY4lCo|RT9!=T zo(q-_F(ir7lEWrFfZ%a;-2@>HAuYN;-AUh?szyU z0XtW0bx0DWWf-%xtP_R%U=7nXPZyR#nWd-}d$TKPaH&Z!#^&gec)6ve|0Nt_BbH3S zeSN&UkR(dW7G`-rvF<~dFUi;9UkXc#pQJANhuM>oROer~r`OuMBwlXWIGH^uX@24H z6c}SS9+MSK9cVi|OuJt=v-y8St@X!({wkm zog`6Oo-)hN>K|r+Mf0_PD`9aClvtKEV%INTm!Iv0x|I1u;^mg1$?OERPx)d$V7@f| zL6RsfT4!bF^az>>e0#LF#tA+cDOl+GIXJ*8R) zNfM={53}rSmfjHRVmG39Z(&KlD5=X=3wHeyr`>HHSXxaa@p8+S(`*4T+Oc;8)aBU> zl0<1)$t(~UPEdEQGrgCJT#oh{V!z1UnydaXfJ?GCp@5-+zT6@}t_@i2Hg5iAeVND`&x0khOT zP%RKFQ3HnT5f-yhiABeXt;z2X42^~{wt57Kms_qG_@c$FOBOzRoj0B&QCf=md_7mqKMD^HjNmQ|)CiPG|oSu*p=ty_{o^|;;k~|P#ryw`ms{pHVE2_dC@pr(ViEWj z?~RXneEX@enBS0C*4i*j)bQzRz;Zi<#LF%IGucVqx{B%8LvkdGBvD$fGs}X@x|N|W zA7<6h5|&(r%Pp>#Z)1}0K3cdYUm8V{C@qtiCGU(8_9_e< z^6{6jINX+4ezs%hV^h+y@%jGb(Ij4Oc{%z3S}fL|IuGmEq{$?S(z2IXszeOw17plX zyH`1_zwRd&$}Gnd*|GB3uJrpb#wye!@p6lq220XC-4*-#a+;7NN=qoSWcEHc0W4*0 zS8EGPN~EMN7XsO_vRJM0o4^u(h{Ve+mEs3tT@D0I+6r}9af~EUTHZ3t;i0Jmz*76{ z&AP%85hbyVSqh^y#iq89$tIOf3*$i*%8qj&4xR_-5r_TF1 zVWIe=?&~z#1fBQA){u06*LlD92YV-=eX_nabl&&bL{j~qI`3O_nw1Eh_j7iVlQRcJxGGT>Aat1Rt7U|%xu{TI`0ckkxc)#^Zt(?IH769IOx0&JV(<0 zbLag~zhvyZPsxZK1)cZV!6f0o>b$?5wI4h0Gd7>cFOzQ$Cpnd?L~FK6Oo({72s-b5 zZ<93tt@D0yYP-6ydInCgZ6#J5&pVRU)9WC+i|yk#2j9VJF^j~@S5GsoCb)WTsyT22 z90@irBuSM2%NX{*^d7t94p`=HdE8N0ETbisl}@aacw5EE_|>!479?J7DYe2DEq*#* z@Eu=YTarX+F=v(yBTu#iOL#@y-olb6vvj)3>T;&DV>LLcJ+4XO<(BzRMx&+s@Fs3h zm--D!5~U@GSvq{bgJLvKNGK-s*fS@a*N3x_J`bbYl$5a zT2n|8r6reH)?GQX80zx+hOe=(1m2Taf^6B9d0DR+!=Ntr?~{1BC7}iDSdAJ{2|FW` z;z<&vrOpLhihXDJt^v#JAEhP>i}QVnrCT%h^tx`D_gY|yFH`z=yxej(H~`03-PT3R zV2rs{BuSJOBW7t|yZAA%=q&9zM_7tvmdLv7%DiJ_ZYQvmT}R^OmR`?O(GnKw?gAE# zZ6t}(vW;0TI8Md0^Q2wdmk3Mx14&(6*D%Y2N%@t)a?Xat%Pq?ek3q|uRqlAKigzSQ zl$LYMQtLxCJE+UVm{?O`x%p6H>7dP?EY8p0gKu|SenR5qmPPNZ(Bj=@ChqOwo9V4-%#IdVc{i=DmZ(Q;X5-+!;_OnMzgC;H9VT@(oB}tT) zYJoWSlZQ;k_sGMC4BsOx*0B=Hw6bi^m1Tn~*uxmRI-11GEoVNnyFX4Q`RQPpHH9Qm zS_Uyo35zl_VT|3r=3*x-`7+D(ovbdMdk(Y#%ZX?bFSi(4uq62@GoOLQ^btv-v}|CO zIs1=H2FuI`IWEEy_efHghBw%n93Oo8I9MJgl6bjgYaP~6yCQ0t~l<6qp_L9%PqSWvb_q*&Rfv{EM7ZF5~Zatv*;JM zz`H;7$LE9zOP0*yH<3MAte<#nDAdJj28owja^g$&AEsHpTzQQ(b7wD|y?V{8*CvKw zX+EDMQCe0rOFNykBB)E<*4?9oB`QHum+qRZ`*!S7%OJ2+=|tk?mX>GOlail(PvF~w znLSAorNxt3y2huh1k0F23O~G}n#g!rWeVD-$ zNfM={2eT|p%Eqsqj`i#NMp%-bNa|AK9y`8lsc%>jERMk>UT)dZZy#D*cU{0eiiU)f zBudLtX6ZP!-d?Cn_QhSFg(dW<#L}`Y+siX`bE{Ec=^IYs<(7Bt*_-w!-(NI`G3Fgf zk|-@lnI&&`@_4Ysl~4F3Ebhq?%lu@vCZ{}}avLm$)k(bEvaIh=tc!>7`?+9QTc0FR zT4I=GwDlUiep%VCOt}hwJ@qk9kysqBvN6_rvJ1WiePbAjms=XtU}MbDG6BD+t234) zQCdm_H-?ox?x#e1k9cVEd-mf)S&h92j zl$QC-veAFO9$0?PdY~&T=TapW-CFGUGH7w^N~p`eB_v*Mxfje9_93AcFM(zF8j?h5 zab}i%UdEwd3G7o{Pgoq%B$kKe*y+c%t$UV(C2k0bms?gJx{P%h>r;CjSPVy#BuYy( zv#fmDsux%u9$VB^SPEs9F;T28SN3$?0hZdQNxa-Lz^ptT_xtz1&=@R-&yyrdOA)jD z+<5@+^Vdms>MJZM>5{r!ZOP`#vbs&MLt;@DiI-ajS>8JRrO$SRUDsc)4ZV6gKH*F4e$u{;O%7e*Xxi#hzI@ zeq7WUECa^RnPm(AtH<@Mm z-37QeVVvu^dBTz>voz7#gg?WqY{(U`Tt7+T<(7QQbWGy3H@P=hj`)!zO3P z!g4cLV!70i-G>?fba7p-SYYoqb)j*GQO5~Zc_Wt{u1I%Hv8zDJ)u zDJsWp=NusojW0to~ZSdQJsnMDK z!eaeOVzK7?#1BhcXbYD2aU@=DFaR{ipf0iYB#F`z!Ymmsx;wy=Ke?P#fQN%-<`zEEw*O^&~mS7r)^+4=uMI+EhCv_a)35|6VkQguvB3&dn>Wz zoMz8nHSKG&zKdQYUT*1jk-eds{bAHou(TgYk|-@^%yPbTHogg&6u;$#uw=oTDE`ZZuF^@1c(S^}8m^ZbkHFvd2%jLa97sCSaOSUIqJ zef9K5n!_0T^nt|7EsnG2pk>?b?bE@M_>Ck{TC$i$Ym!$EShD-o`X(&C`4Wpse|G)S zYrvVyU@0@4#LF$MjmID#dx>RT zC3ao8yM={6ShCznyxh_#l+FD%wKw83i}T(jiPAEJS;`I^t__w;OC7a@r9fuMY{ZV0 z-HP@Vf~EgD5-+zrZ_OSQ-Y?J}1a+yjjU-W8HZeTwMzqoR7Wmt}84l|Fir}=l%IU zf!O4}yW80?=)6Dhkz`VI-nR>7xBWJ}b+Uub`;I?ImVfHJ&p+;uKPp^j#d+wwuU6*t zZ{FW^-X{!YziiB%Sv8>Z{$&M{>i^Vv|L*0+kI;F)y#`6C?!2!R9E;4L%8w7gh&rf4 z5{VJ@N9X%s5Pd4HiPN%9{%@1GB|sSTa?k!?t_KRfT2AJoFdFMXb|E)2|l z9Z7<}>Ad%vQxP)_d2K!sI`8ZEAesJc=lvf)aC0X;J?Ok|IDn-6=g#~6M{}_A{({aS zOX$3h97YoUtIqqN$Bx)}pT4>??)$KO49TfnCH&Yb(X-#68qj$^U@}Sb-#YIXy&Mn> zM}qk~AL@z~C-kFa^=#dbedNpM@Oy_TM^xEDw)&x^6Mz#yG5Ghp6c^&#G2 z%ZU&6c-6GpYyC7~aVV5niVa}zD?1su;22xxN8;rc|H)Nwj8$HF3@y8ZND`%`d>GFC z26j4QV2t&i88%N?3T2jOP1x784XhvEgu0wvNaE#|b$bkO?$>!^jqmtgT1k>9ExnjU zueI(Qu;iVnv_e=?zDnw1UxTg5jf*$j43&Ay)t`2DBFvYmK_rRNvYc6_ zuH4=OEb~MAZ4j1-ZxYLg#jN|bnA1}$uz24h@p4O5H}(bU+y}Y_U|D&eBvD%2nPqg| zl{BzKx|#0~7SHbz%a~#8_@ZH$hI_8GIz{5;mglzJaf~%uUbF}-Tm4BArR5Q`H2YSv zF<72Fc(zwqEQ=(TJC)hh?Cv8SrbAuylSsVWa<>}$rb^?`JJ{jfF`Xn)T1tiE+`m+> zzymC~$rT)gB~NDQ=gQWx^2@r81OBy~C0k#%}KFICHM0b|V2 zGsRz6oPS9yng;A?+$Wv$I$&vZmc+{~(J5zfjLp<&Vg(kBizJECa*tWcJ5|Idz5}wG zUlNugnI-Ep+rPi;OQ#25dHI&a%PnyMH_-B~WHWs0#^W_GoyI6)?s;$DY3{EH`&ZEYSt*__E{k;a^~hnnvQ~mYk*R4~buthx^C4 znUEw(%N%C;6+ZyG)((#U9w#i`J0%wFlFahH-(5ejWX~Y+a?8Hsi?J>hYFgibx|Cl) zk|-?>%%b;d#UQB5NJHaPVX@vNvBc=IbC}o;YsP?O*9a0Xx3p|!f%9c@)Fb?^!nE-u ziPCbLSq_c!!u^#yH{1C_Sn_3-tW|6svrRMm0v5Y=BwlXu(PrzI#>l*G#|-!EUlB$3be?}7Q^ok5Z) zEz_8#&L%w{u&j)kTC&n#Pkmf1B$gs)c6>PyG6*}0hCU_na?8HM<&iJ!iAl~9D33p_ak@pBJpy|nb!x=GP>;S_h7L-OOhxp z5zNxH&o|u9Z{YA>)rBQXW;ywUJ$%v{klV_gMJEe7HB{>;ZDUT!%s@HJXC)h^=?mhs6XiPEBb z9q0bh?JM!gVvm4rEri9_N@8i@z&aya6|BX#ZP#2O@p4OvShg6pdhLr{o_lVPBudK! zW@&qH?g6OFqj9nAg~e{K#In$y)n&}6XGJi^zE3Cda*JW>WSlQ+UDj$qUA*RzBua}V zvoy<{iu;jA4bbQ*ECn*l%=YY5%DBslT3|VAMB?R^N~Y|d($m0UE5Pz%97&?ITxOO> z9cN-)MzkAgC@e|)Bz1Y-j=iP!Xh+0ruy}VP@p6kxb=I}!5;4jZELMF<5~U@NS*mnz zu@=Udo{iTCVF|UCSaPSabu4IjbL=RZ=tSb>mMt!s6>r@#-XAZZV2KiSwmwtLG_TsZyULQCdbZONXqw z*zK{gar0TiVs0z3>>S7HqVe!Pe(xfnHHnv7<~oF+#qQ0Q&tM7bNRlWmJDJ77y)X~z z;;g@Qk+9^-EQaOTeV9*&Chdp1RPRCJ<(BzzL(uYTXwGV|tn5#cC@q1^QtQ!u{60*o z|M}Iza&Nz+E|#BJSG4zv*nF_KTq5yuOJg^-e}8c8nkT?wbe$woT5_0WW>v>EP?t+q zKQ;@?IXj7EQ8l&%8R{&m0+##sBwlX0n0*}UV&$+KulJuFCP|c*S~qd-kL;U`*Odp} zkKHXS4hJNbFuw0~`jJMVVA=PL#LF#jVw<5w%l_OdsLP=e<9`1LrDYhilyKA<0d?t; zYhf!ag))n^5$h;&YGQ~T-bwFCyxcM?{3%+(3ud$Z`@fMSO3M~z*_k>s7A(7}#~c!t z6njZsjK{LMKg4WR4AiAq0}?N{oQ!5~b{*Poq62kl*qkI$TKt%0_RvAiz~YkB%tKfr z4oWO(Z`fh3grDguusqKo@p8)t{gGIgwrg6XfhG1ONuso*GmB$L)-|iszQCO1Am^Ba{|zW-+x33jxcBDq9~4OW+}iWw{Bv z*HPAn|fb#bkC5R&$}lb+9ra3<&~@(~I%1g(clZQkUGv?Cn9H zOa`UjnlDwH6U*7cH z90`_QsU%)*nVZXQfSiwh=>%geHJc<+T8=ZzM%{(qpf2W>pZ*XQ?;{e6Wf5D)j-=JV z`CCw> zFm)@D1b@@59&1a( zW8wnc>RXzVg#W5rJ@23^cB^MpJl_{qycO+8PSq-5){d&zT}Yb$)~)^} zd)q=d5;RG!T|*4BeA(*RxF$Q>I8=MBGmQ0vStMS*diqxzhDYcWgKClRzdXw$NmTz! zv2N^tnOSeeE3k}oF=`+zac+_$L0%@i(rB~x^e3=rpCIvaOG0f2w6yb@Yzvk^ACg3A zSdiwV^IL)+CA2QY;FW z;zu)@;vRU@S57w+mMocNOnLS$X{}9z@Tt$%jU--fnSXy4S{8UbxeS(|9VCg;(t%mN zN1LpIx&LY2`4PersD_(?rcr4T)a(^C@oRU(%`hg zJ+OQmyl;`P6v!;QBbY@$V8mju^c+p%<(7lXy|FGAo~_2c*1k+8NtBjv%rYmY0^Vgv zEp=_Ruq1g(>e4KnU5Ga9V%ZifCo7V8x#hMMn=i}iSHiA>(zQqurKQasT*sbdZ9D~a znfkJDv#^AENi3#|+0%tyJ0E+2WyuE;FSneS$~uXkHOjzqn2z5`5~XDZvp8K1m;;s- zPv-0v7Wb19OR@K?F6H-a!47YW=OkWkF*1CIW9-VVc5h(5ynRiQC@pr(@+ma)7FZ$< zIoS$}`6-EI&Rn)8?|V4>7StuhgT%`%iw5mNi@T$R30N#nlO#&Zb!NGHYG-AzEE}J8 zNLX@ZmQL-M<>SP;hG024gT%`%qjcKiMep2Fb?{4+{T7fUN=pH=yd3$!4J@DLxAzd1 zd)|_|ET6;9E&I0^hg}8pR+4zRWkof%PyE%H-%CMVYHcJ*l$K`EIQK`T7o7yl#eH*p zgyr06iRF$j>%Q%HsNs32%Na)!FSnfNKL*FxmZTR6U@>{` zi-V8EGUYtG2{o&4?WbT#^dj+cOWVB>XldQFJX(Cuk|avYUS_H77lzlBKQ|YJ3QM8P zlKGRprS@ZN)6!u1x|_tyEwLWXXsK7?Z2{EfxeZC8w1hH?L1EZVsEd)e{%v7NIU}h{ z*+uM$@1tAQTY=@mc@i(TG%|aQmdqbUZeR(zOp+)qZ<(dxqm4(wGWXr~hr$x!E3vq> zV8_Y^WrMWAGUqFams_%~9ze?llTh5FsCIF)-#)3f;+>*r0Et{%LMvKk3TgAXq%#I{cT5_4CsQ2C`FvgAsCj1bVz;hDI_XJj#{kujD zhPu=ZBk^)eKevu(`4DaU2rL~UNfM={&OMy_rB`oB1WRqd+8R~%MeYE0&*g2a-vlGUYufptl! zOZa0FFSqwVb%mkq;Xn}a3j zDT$X`J{Ii9xqrNJs5{ijdv{bu~bAM|2W{zO_@o8&kVX?j_vCK(h_s23DoW*;v zp1(-E-16CrC8^NqK7J3jTG?&Ce}vL9h*{FCL!!V^+UjB-VabWM1(x=j4TcCyT#%$Lqx9LEDpdzP!cMQR z)kwVDQo8F0w7BnEhHpb9)FVlh79VCAxp~S*u=u!*8Z9g#!4k{tP&Qvmnk;$;^QFcq z5-+!mxxqS#-|Xmw9eCgUND`$bg;}%*US0y0p0!=32#afo#8T3i-M3sewLQK$Y~Pr~ z%Pl^s->@znc3i|=0~WL`*4nWH(x72}P8CiiOQCe0ri^--m{6=@(sw>tBOVnjaU3NZW*DsdCoM(ciSv?Xj zx8&Dhb@BFVf@7?uE=i)acrwd|J*P)OUEWj)+b%4=S0onyk!!q%NNZvTL!_B#F{;lv(;cxq{#6z4@SypRl-xODyB&u+zdvUB=_{7xN?%FSkri$wkZH zLECW8`*rCgiP93oEG8Rl`a)ewx~vEm7V`*+rRp&DroDM^A%5Go({U0nw>%lhZWs>S z-5AH1kq=3tw3LX!xgTXv58vp{3O^koEV(jEoRRX z@DUhe!yb_&N=qkZG29s+0CgFA|KlBDxpz%cm!(hGKJmI<7t4UfI)TK?EoZ7Fpe6NT zi*8_9ol24@E%TYB;msT2VDXzj__45@yDqUbT*%%cxI5)ZJ+Mq#N#f;}=O=B@vgfm2 zEwJp^K$0je&dgHV$ZZc;CQaCqA}kI!Bo<>IwkBU5@CwJ+y_qClZYll69xay!H^AN= z%LOEf(h|)q4K9E04Hlgnk=ephD6>?&$gahFG=}>?UF`3Xc)3MaQyY&j=e{<Zq_~oJxh3_2CDx^<@xfy-#>#n- zBua~3EUsho7ms-WmXX1O3xy@(mc(M0&7N1*T9WJy7VT>!UT%3=^B7tx)Ozs>ECzQ- z5~XDpv*cN%+kmC`*Y(A<|GEe3d0S!`=)%^qB{S|#18-B;jEo)}8lZU*o$MO5g zugZ}mN{cyozcVR1?>m`P z#HnLpI`1uX-s_epS^lZ>{*Qg)KjziA2%YzrwMn|a>%6yoR{?+7aE-@;L z>ECwV|M3Ga`c0h=o%f?gk+lEZd0*=3ckI01GQqGCblz{AL=yh1&U<~M6WDqGX_7DQ zVNq)q$*Ei=&a+kGNSnI>(0Si{0ZH@UI`1d@Y2o{1&-WQr7AuZrq-6E1vzy&zD5`m( z7p$K4jwD{bdVYCw7>@*B!di#Gk-*H2BvJk^H`)KPX?0wfWasJOJ@Nfy zwIBD4%q)57cf8y(@4zUm%dSsL@Qa83nk0$R(&76*BudK^W@*r>1$ITd95d)FEY8sqi)(v! z<0j478J}QPIZfi_mb0zdS)g-<72Yj)lP?w4$R`n5KF3yOO%tEn{P> zm%ZO=7-I*!k$AbqNk^8xA*KCoFLfC3Oib&W@F*`gt?} zOLI>WFSk5jI z#GWJZa?7UXJJ7NvX=X6g#XN*0QCjN7;oNuY{q;3iY8aoi6c*Q5iKP`^$EK`_!Zq1o z5{Z{vJ`M3fi@^as?C`dlMUp5jBblX9Zk-gcJWqRaKv>KkNi4a%Yb|44qxn#mU%4b+ zZb|1K%{V)Fdp^`9<2^~Dw3so=qK+xW!IJZ$+Yw>Ol37CC*`2KEYx=hYON;j;UT%rf zV&7BRR;<=Yu>2?_NtBiVW|^Ac+zTubn@l~0CF-%HE+358-O`qyrn-VK@$BwlU_&3=xS zxErB8!7?X>BvD#w#N*t*xFZkuj*9K69V{$%@e)fxJbU7MXV1%>VDTGB;^mfs9Twm@ zf2^r%Ww6W`L6RsfLzu{zY@hgz(Ij4O*tZm< z*d8q5k4O@wWfQYx%rB37;S681{*JIDB}nR$vz>KF7!=dPT_tXAA@Op{_(=Ahs}U1! z>;OxOIZ2|l_%h4o{i|@k^f;UNNLWG>C6-?4>>Q?X%~26B#$tj!Rm5s4R)(v zUq+HBE&ZA0^u3>7z|!ZTHa42@Jcf36`L0 zB#F|J#4K6uGI9U@rmqi_7Z!&Ui6v+RdrPh2<%T1{l3qaK<(8_2>?B^;TEFSk~ZX`*RmM+Xv z|KlWlL)GQ+g5kmvlOd_g@GN$Xd^-FIcHx{)A@OpH_D}X)Y~k>$W1ucavPcr8Wg)XX zSajST>T-J5v2nr@m?^Q;tIfXK`D@YKE{lKbKgxZeYik8P8T4faTPE5-+!O9qEdej)mIzc1LhLNuspeW0ueH z^A?{m$COr5~bxfvlO*!VGWk) z&s-gaC0}MK(UZ*=$8p;=z;Y{x#LF!cd$8h;%V^vgELLwx5~Zb(Sv<WlN&27g)MKCh>Ah+wW{mzOlAcI#|vplO#$@tEV{kElxyPL0t^>t9uJe z$V-Xk^IO(UxU8>^F<55xAn|fb;F&Wx#=>Hk;Prm{{v?UgGL2atR?x+75ZxWN<-D-C zzLHpiBiT7jNGwgO zvTNimyQ=jCi}w!_FSqRO#n!QX&JBIQGPv~S-#z0 zrZ_x0H?B|}ES1lZc)2CvT7OJ3d$>gaSS&+G5~anGS&H?uT@RLVYZ|;1mI9fjnq%Qm0us6F} zx|hJ6j>FrKc)3NF@2XiUEgg40s?(VyQCjkt<;JcIJWsaSw&%OBgyu^uwyoKH%U5mN ztbn={w<7U!i~Wr@SeF@VZm)*AR6jtHC@nfEIQO+(YQKaA=Se-@m#X&H{p8&LS^lQ; zK6cVE?EW@Ats4TJ_mP80CPnA{%Zruq(Pd_z{nw%MK6)I<@=u-jPfDJ}A60m5_!;QD z*O*Sy{axq%m2T{pg?w$B2A%hHO-QQ$Q|Gf&R2+JEl6U*4sp;V{ke z<;rWUnLB&w?A2=qo($dso%fOMB;mj6yw~1#8awY(*4M{71I@fiPUR}miLDZ?HTO-2 z&iiWqB+Y;8ymyEV#Qn+Mv{_zWtT^uPC9CHa4|XJYkm!h=#5=o?c=_rXUU?X7NW~twi-Tk?|sy)8n@A8Di%PkoX;;=3y z`rX4$uY(yRiP93tEH{ICJ%+k?RO{JHSnhq4)Me8UW;qd~ZH{%RPvYg4D--6TmMw=@wG)^%ja13GxW-wo(OY4wHt|-Tl^mw;C#tlRBb0% zZuTWflomr~aXS--cjWsG9nHAX?A!KFCOjcwQ zg;GRCQiLSC2+0aX5+xxCMJg&n5rx$E{W`zP<>!98JLfe%{eQn-p4aZ@JjQw4ALqfY zE&UFJ>;ub)i6mZbIkBe?)@9#_yK}(OW(G;3v^-~)mSJ07LtUOsYU(U3o&^%ivJkeC zbu^*d39#7CCh>BM?w-atU-nrRgn`9nAxWaNlz)YDfBU`6e6YmaT((%xgW zQ~_5@JtmNNxn=)mLmXqz*HtYI7GragL}{^Nmaf?kqrnpX;oUJ|2``jbf(Ek=@1EDv za8>T>Hxe(mRPMmK*7VCf!fy3zg(Qj6;>#=<-d%OU@}tG*v%+$`NMecW&M!Z<>Ysu! z)~zFnms_Iq*hO8D<~IkZODhACL}`g*maktw;A*L%r^6LtakeefO1EsovZWR*wzlq9 zRuXL_IjIl5GND`&Rn^_)4d9H&o7F5js zm9Thkl31o{F-xz{`cuJDuPlj|TcWp|#4#2={}gttW>g_bl$ID~xsu%l&y~wOax#R) zcC*AXw(UmzmmX~2827vnK1kx_ma-nj(6S@wDL(a?cZ?)aT8h2Kxqs)0aTe4?V`H<= z!jdMl7_3-_mWJWYW zef%9HmM+KHw+Acj9)Vvu_bN}~|D9`cAY4&)TvIAC@mh$GBf_kaj@)5{a!&> zT(?LpJHpxh%dWO-w?JK%eIW62i@|=@@*dKET`#a)|4Nc5Es@NU)9Ri!SZty?))bap znZ>d(YiNEn@AN^iRC_|=<(8DvAvnhNt=NRSpS(yUNtBi!%rdo+2X?NkGvA{lED4U1 zx}2!W>T+)9{1~u2wIuO!%ky?G&|>J?Cmrh2-4Aeg#VfQxY$?bgs$n_ZPIAi_a{&Oe9H^7FT9z9#w4#%$Ks2X7v=7&D$iF zk^R^)`FWH1!C>i8i^R(<8gp-AUBca441l_r>5wE!%S~p9+2GX*EaQxK8VO5=%yOka zv$S%}#NF4{j3M!I%dbkO(ef>{LlIc|O(sc{mRx4h7!rwZ5H)WfZXztv&XT%}vtsKo z57uYm_hFiKC-HL2<76GQgvIv9`xl$OB#F|}EEVT|>zW3*`-%Ch+GfHMxLsn=KgaI( zKLnQV2xH79gT%`%IX}wbHB8B2zaBwdO5~6vN{bn@%w4kSCXBJZ1E!h_i~A0VrH3XP zW20l-lfY8Pgv84&hB=;Cmj%{Mt-;b`97&?IY-N_!C#vIh^5_Zfmco)Rvjj)5YnXM# zZux+v+)NTLx0r3dix!g(-`;@bg(XR%w1hCr3(snSP?z2DKdpr&(M3|1^*QYRWzdZo z^TCq!ip0w;b+(_wF?OKOP-C#{NGC~@mXFM06tf+7*y*xU&qi26U1gSYtU0?@r|1V@ znYNF_%PniYY_Tr4JFKw-%ZbAziPEBzhI9W&tro4JE>H8PItYuOo5T`4fvsBxZTgS_ zmbN9P{f?Jg+&T_L%ib8PHDEC=OOhxprpz*6*3~v(v6&L#A}o$OC6=a@*$z7kKMun6 z|TmJS}KI#8FD*E;zMOYAO5U1}w=H6`~4MFwC= zjUe%Ii&2lUIL68@$ia1(cQGW1(vra}W@C@0f<-se@073vxl1gIXR>Sl*7tfY1B=lY z5-+!;?_mqMgIiX^U1y|EEND`%G7_;QOZ`KLw@^Va#>%vkXv$Sr{u3_e_s$U)IVltJ) z%PsrYv3umR>6*9UIA%SEBvD${GmGDR%mI zyxdZ=!FIH)`gmv+SO#j6BuYyPvuO98VG5Skhh9AsmgAlhi{oT=OfG-Sy%|^zo+R;d z%Y$80(NesM27U)V{2WQ5v{ZkKbHAlcZ(O&0Q?XB~usC~3EH##}X`B4CZU?Yzt3~4F zmh@loXxUa(19zlqt3#3~Ed!aQ_1-G@c6w~wvMgcAmRTCjWp(lBX}SU|F-as|ZrRq@ z6fGwz?CuF;%-}6aqO`1H7X1T5!eER|JAdM4m*p*gs(0N~bImz-*o%gqROZtx*ix)%Z{TUmQ?(aJ9dsb(^tWo1#r=jzH zf&)qQf9kyN`KaYc=)4c{ASucxycP1?jCdhk<$MI!W+1o%gfe>0u_5PS@r_ z=e^%;lIh=e-v99fi1ia*f9t&WKh<$D96j6E+%2X3*HfQM`y@xt z$p&oor}6YdU0|$Nbs+Kbqo?~Kb3A&sd2X8l|Cfb3ND}4$C6WDKx)!yq4VKmKiz*0< zx3|O+(Vabf3fWp}16VeHCGm30o3`v~WA~J|Bf!$*7fGVDR7}TdvZ3FM6<|3xt5Z#3 zvE47Re9>Z0NgrM<^8zem7LjMKYRrKJzE=sj!g1C|o4_UH&p zn#?lzGP^oAY1tN!r;9p~c)6ufYkwSLX`v>Mp)UP9 zM`GDqgX+ycu!SCT|&S->n7jlzb5CBo6#OjtsEC6{$*oVZ+EeE^#pk=JXHCL$1jlm>|(h|!oEmmmXgSy16`(i09oBbq~xc=;1 z=^aqvCXBHGTS&azV(R`J$Jh?9+SncC;6jopEhRE>?iXyC-yAFpdQD#@EEzJ3(MPuO zySvir8(>KoMB?R^`AtTk<(;3y6{yRJ5hRJyqR%XPCN&#_<#RI^8)1n)BB@Km2h1`v zA{o4 z0G5RHBwlX0a9gtUj^KwXw(*xL}|IlEP<^J4uZvK$V@L`$(LE8&$EYu3wQR# zH@mu;l6bjg{h~5B#w?a;9D}-ypGcA@Ed|Un!ft~nSSB^{@D-NC|L;akvyie;PC93->D@l4%g5gBE(nX?Nr~mf7IqD@u*%d) zU>P!m#LF#TpB=|zvbF79oG*Pwkt9mXPG+fCe^m*vi(CTpWuhHPqO^oFOU{i1{4QMCYv=9=OP0)HvzHx{H@lmH3?PZ}Z_ji&cO3OrMdG7V$6!G}Hd$!W6@J63DOka)Re zsTS*uY;NRt3M{vtkR(crGqbF(cp(xj(>^`U5|#p)CAA;B-|y`okNaQ^&nNM6%R!T_ zc&?mV=M&ET9>wB*{|Ke!IBUT$$6 z=8SW{_#&TaP?zV&NfM`KhF8prge5#sVmY*fb>Xa^a3US%OZH6?FSnT4 z^+St!`ycprdeA+RL}_W1g>(OdrNwHnI3M!Wtohe^^6|40OPexmjJ;{3hi`Y>JxSu_ z7Pp*GwB*%$h;QATKSz=%En}Faj>d&1U@5r$q>8XOpOaW#w_qI-&zFD0w{FHiAn|fb zkCq3}qFKka1&pzZ2_%Wq;=n9=MR6fu@%`AMuCQdwEH!qo#h=k&%YwZy#@zEryxcNt z_W_*yMPqv5+Z}6)ND`&x60^K@o{i`IjE>8i2us{~NnIwkWb0$2D;9kKi+>LiFSi8m zV(0y-o3p~9E(iLPBudLWW_dLCswULM@X=8{VYw6}vDB)^u3_rTPQSt79aI(sGVjR>XCR0n5&qSW{t%yeO&53S-tu+^y*~+%;hBQxY$?6qP@SmT#Yf zBA_l=uSgQ5C5>4^>Njc!maMJXlZD0qlEm^vlP%FtsS}QKKh&MX%Pkre*lQK8X4`GS zvTi>~qO@p##JRu6F1-R+Vs;FiEiA5=B^K@F?AF6)&7hd$@Ozu;j`tpU*H$!#TH{!1Db$iI-c->2E|!&EfTPz;Zl= zBvD${Fw2d+z-mwzo7L}D3QNKjNnLziv*%*%E4`fzmJ*vtyxdaqBU_l#s@}RASZ;15 zNt70UX7R6Mj$NKsE1TL0OUPA;WnvR{ZJ8LfW(Zgo+md*>W$%fmcn!0+&eSnrDeXj( zC@n9UC1+s41E`B##?Gz6;u9jXc(CKxUZ<%)!Ez&q#LF#Rd|3DGl-SP4!P4jlNusn= z{)BVC_ONmOV6i-WW0$aOz9zBsN@erKa?sLpV2Sr9@p8);4|aj-+^QUQ-zEi-BuYy^ zW?7{OOj?hJpl7QL&p}cLTTyAEafJD zC<_+7jL}iTk{BkbOP96mn*YVzA9`S^`ZP&ifnZ46FwhG?=x*llK2@MehEA8HG0-Ph)0Q=B&X^qVcU%zB^DmMehoVBV8t@nt)BE?cab>a#NL*i2_hrella`C{i9&4uYE@1@+wNba|JeS?jN-K0y%@OQW^yXDrQ2!FS`@d?E32%eUz}@l0^x zL`q+=hnTg@hgEZSCV+SW!6o0nm^eg z3cCsSZy-sOmS|=%tI!>vUY~9`zpk(p$Sn1~v&BL04fBq|7|ZHN;^h{bre0W=F=IEs zgSu=rAW4*#LS`AR*$G$VLo~`a6_(^kNnJi(Wn=7Z**VyS<5-!*%Pqx^A4W^WE_dwo zDqWi-QCix6#zS$F(LJyWXSVaTw!#vAS7LFm$li$9=d{un>hfeaiI-bCHDtG5C#o-; z2X(o+pCnOQW-!a*;k}B(e7QTWQa52aeotaqGl`9{?We}zTLe1ONxa-LWm!YKC>b%e zC0evCND`&Romu>sJ#`1mP>o4_g~d5aV$p8J_V!3RyfY2zva%A1ms|Q=VI2~mb)1hv zT`X&oBuYyJv$#7I;);Cc^ie~FC0l0EEX5jVhDRO5&K0fsBwlXWQ-iITj4;2v3oHw* zNfM7?p>*xGXXpFD!ASWa#r zNtBjWUvTapo4OI7V9iV0GF@0M-IrLF)?{~0-M57&z!VlU?WVI99FNx9(uM z{*@$AS|&5gh@q>-KwW+shR+ig?-+??c`{r1?euW{C9sqjMB?R^%uj5-1aulS5G?mh zND`%G2eVww?}96gsUzxH3ybXoiRFbeJ0|O<8+8ZECo>W+w~Tn*9qZC+)ogs~lR2Fv zQCdQorQNh}+?(+BcFT3bk|wk4=*?Dsy_POI1eRrcNW9$g+Bz65je<6`g1WRnNRlWm zpPA+Kvtzelj1??AYA-C2v68y9_{3J_0t#B--cjKzNxa-r?|~~?!XC6Y1dE0(Nusng z$;G+fv0Wv65+4?jyF*y~A4)7YMzS#$GG*E#s7pdHiI-d2JYY|+-!xCfy41Nrk|-_X znB{T6B>ZMq&W1jo!s7Z!Vrh7vo%fHFJ?sva9!?})ZW)*shIRQ8GSCR>QrnFrQCb|C zrT$9&L>OZ`-MaV+ORmfk&sTmQTkL-UmOJ}MyxcNuB3q`+>9n;LSQ`70BudK_W@&eF zg(Fye_w79;ED4V#b=kIs{fu{8#^HB+S|1_ta?63~H?S^QSq*1_WlR7`qO@c&i=%d< z3t(C8_UVGKgv3cKq2t)caBFDp-wAbTv5&;dEyhF3;F~IgOJ*MdOS{7)iPF;GE6)AA z1C6np@Z^_n{|KdJBeUez z*RKbb+e^ITge5~}>5|R5Je$03g?mO?tt0Vr%X8<(Xj$C%1Fl_UZ6-;SmLO))?OSg; zSd2X$BneCOQ%PMCA~)dAC^XV)19jeknV^b zMK^}b$`%&)XA+CXCiXP0#?NI#!IE!E;^mfedhB5MvX35iM;)6;k|-_4%renFbRg8_ zef19(2*Dikf z`IoBw*Sclmb4gvif>_7u`wOnap)QqAlX$ts>ug`NoK5L~9jk4ENfM>yG_#zwC_4`7 zvd{W!1z`z&A+Zej&h9IZl-P}L5Jj&e@p4Q0u<2;otr4UHbqU@~k|-_7%#!o?S~OVl zx3{S&EPja+%bY#zt31(spj z4_gUKmdw(iEL(E4m~aQb)BCd_iI-binkHafw)=jG1xsObl0<1)!7O7mH|m1reoTI6 zVTpYysY|>*yG~w~-w*eRzmh=W<(A^>&!fddBYrYi8oVY+l$PVn628jhAy{T_f7?e` zf?i21-sRana`!4#@LcKSLE`0>mRs3}CiH&jHUi6$10;#k@|;<6Kk1eNONF0pjfKTC zSz=jd&5mOyUv-Rt`7)wZyWjD0ORm>mtV>yqwfJ3y*X2kOrKNm6&V4PrQQyIG`|ixq z!cri!JmtGjfj#SBwlXWqkk3aGA1pmI9LkzlO#%u6|wup}Nhwn|u>( zxsM&Io{l75ZprD)F1AfmvmBu=V_ZoRrKKaYM3$bq0LIv{82=l>5}7Hf%j~VJPw7Pb zp|?;MuLKe=w@g^e&Xpy9E~*HY^{+`1rDYbgj7}fi6fCoJir*C$|92A08C|xfw6)Vl z+`ZAaC5e|?vNGB8m#RN!?gh)C4kU@v;=wFWCfa$U#d^{sVR3ygu?+HJ=a=xMr*oh# zH*83}+;XnsI=ru3vCpM0SRUAuBuYypvpfo}^#v>eNxmcKK1Zt(QMo(%O=E zx#eiLYj_+xf1?(j_s@1ANtBj0Kk+#BGtx$kmk^Qp3 z4Qqm-EBZnNN%eo~if;dM-&N>}o)tqv$qW0=I=oMjx znCZ>*M}46y`bj#;^l!VO|M-DnAv5EkD?0KMN&C-T(YssL!mjAuReWzjSM;WDB;mj6 ziY_((26ja|t7~D``QN&tgX$DcfTQR1p3eE= zh!gU^BhDXRInP+fRvEPGOu{d>ni`RK`O&jrB0CcpHq*!bFS`vVNtFMWsqFu9cGq>> zbLH}m=f&&%^>o4Kqr~EG#yVDCdxl5BynVBW#LF%3yD&?1`oh*wSVrGuSF{@v_F)I!ELRdQx76s#-o5Tz#V7?VCY~gT(sGko9$B0E z!x*#OKfAiHWXLRa#rOCVPW&KAl$Nc`a>vIE_bFMj zvq?{3$(LEa6|uQrJ0W`<)WvrniI-c-E=|C?bZGE>6M@e{&Gyxd~grvNQSJ6k*ki{?F&L}~fREE5fq@Lkfx+&mLu z3C)#RzOy^?mxJ=~{L-wD#LF!Q_wPcB>t{dwvfjJW`M-aJ(xUSV=e|?MfRRv_oR2zrLbhl zEG0g(YnYPJJFxpUDul$#E#D5sp{3rIp^Ly$`!-3Uv;;HD@u~INLtQpFYAhF)*gQ#H z`n+eW3_o2C;`_=k-AKILa>vdQExsPJu7YK+7fGVDWH8I3bj=T7>EKm-gRlhUODu1D zvlV&Q1-0+P7>n{H@p8+%#V%-h^>uRrSazKxNtBj4g*f+TKTduD7Tvc`9EHX6yTszY zmo0R<7*DJUmT8|zyxg*+FFQ$9Irq3DSgi9%5~XDrv!rBKWP5v558Ndz1u~0<9{U-| zLqm3hCA~h0ms{o@SdPmi$#>3RXJmL&l0<1)&nyf#rbmD+2lD`s^*a-N=pi}Oza<{2X$HMXdNgl#|tEu>iwBTqeiJX zuyhY5@p4OmANv+Tbp0Sejq&c2!uMe@QH!i`X&Qa>=Fn zV7YvU#LF!`T0h6SjBY>ZCs+k?Q7IFckv%PMAB)ib3rSeotF^FUbQ3MF;P*J0N%H!K1#gXO>; z5-+!uZ9E5$V@b^)lV!hi~k7{ zFSm>;VE1BeeTynUU3Q!$NtBjEW;tus4Lk6(wHK!di?^-jpX*~?65091ZmQK?uoyHX z@p6mX#WOg@p7lDj6)Yv2lO#$@nPP+C6Cn4Wj>Yp!n>m-?3yZCt#4>9ZyGK5~B?X`Q z%(f!&a*J2D9JJ(E>ET+Nsq%p)Qjf zka)S};39S}HnxIiIMn4-Gm=DU>CP+{{Hx+|tkmgMg~AfKQBoK8eryY?MmeMUg2m|~ ziI-bKzouhd?2GT32$mx^ND`&Rlv#pz&7J~v>D9QHR^7kWVEs2qEHet(Mhug0YFz+J zO+6AXx7=Pf5G|)}?w$^oPMt{-rDZ;|^t#Z<3@kZUovR6p>t>0i*D`jlEdI9iH?Xw5 zP2%O2iJRCzBrISw9+TfilO#&ZW@dTUFa9c6mbjj+FD$t-%a!@;T)A?C8}8$4vX{im zEwwK(%YgUuDnVVY941MWmi^2!3`9Ai{;zt*A?2MeynZ(O2 z`kEJUjCBlp8v+(hLy|;k31*fPrDi<=%XG8-?S&=8L1HP}n%(bT{@Qf`jIl}PBwlWL zQ{N3O)wIrB21}oLB#F}UfLV?Y8I1E~tAB2HVe#1_vD8gwmVhktK47UCOycF1y4I}4 zYQ&MVnPB;KjU-W8GMHtt$<6|(OVRRM1B7L>qr`IYAe%1_`#;5X{{0(Byxh{PVJKdY zo$XP5Gg#cWkR(b=@#1*i&&<2k9V{bO7n%r5hRhQAmaV}y9C-%sUvf5*c)7*mD!U%5 zb@!G9Snh2jNtBj4%wm3F*nF^@f4|5~SfZUIb-C1u)n)76!)0NNHU3KC<(8qtSv#`( zlg~k5i7y~Yl$K7+vMeteJ6CL?-kA$a;8uyH?rHV{fTzj$nPBObK;q?=)xi~Tj7{F^ zi}x=zQb-b|Wf-#zTW%Hvbs1&cXQ8mTZOB#F{8n^{WeKd=PL=)NJVge6~Qd6dHLU(O6k!?nK6#w1>D3Dj`KWAe96{xr4_GF@CGm2LdviAFN)Aay%d(FoiPAEV zSuS4~kOXznEip7!SPEnoQxmqPG_%_ed=s)w1rjf}Xg_A_V^OzC;v3zAwMi1CWg4^O zn53=(OG?`5=faY_OH!8s)!F&wMmOteP?y4wBwlXmw3{vFOuY5E9gMLD-$)XrWfil` zG!4uK%g}e_(u5`4U1HfioSo((uEqL+<=ht%FSm^S`U~fa)6~+~;l1rANusp4F^j#; z;$P5wdgO4&Y+-r%Kg-|j75&icAa>^7SRQ;6x}wVmkW7lM=msI$c$uj+V5Kv3MgIsQ zS^lXjx=g=w_@jngo^lVmqTNGCy1(m+{@I-UvV})ye1fj%HQ^-H|EVjw)Qn!i&=ox> znxs^BMVk)Uk4(SZ;`d-goqS9Zi4pZjSM-GO3^*oG>pFBrXFVrL{$p3PW65<*p)1-g zUjO&F`m-y#YODVEFs*8z2gP7u-mKQ;cY?p^ijF&Pgqcp{FYth_=xcRIrhnTN{l^bn zGwfLh=!!nrh@}1JuIMFab+IeDzsryz&=q~7B}w?Nx}vN5J;Sc((pp;1aKx+Af#g&j zC2aSwqr|(2Y9*j6`gvE9=D&4CKlI#nAC8`}LH+W?5$E`B$Vt+BvD#!F^j{4C!@jANBdnlVab+R!uS*48kvLe z%X%FelX$shd^y(PZK$~guQp!HCrOl+L}n@M)^j^phAmvBEi7?+By~A=m386l|6T>( zzYJVK;^meY+W}Y?uZf;7pe}FLlO#&ZS7!O%JNXw_(u|XIgyoW_#Bz8by9zA4-M|Ot z%k42FUT*o~7mSwN>?(R-@t91KC@mFB;+@9nn(5f>(RWDs*23cLC9zm`Wmirejt#?i z7E?`0yxig+BB55CLY4Q)?7wA84639%u;tMdm3lxv2GrWv6fRv zyxijJavkfkd!ja4YR)D}l$Ht1vVX^X+}opC`PhNN61h)O7pIkMg>mhm)bU`sSA)dM zEh(Ey;$>{`v@~2X*;Ah+QCgNVi}oD@>|Dw9doxm4{JkZXDtYXo;O9}JZNW16G>Mm6 zawD3d#b8mNO)&RkgGmylWgD}2#CqUf(I?9lOcWN^{SwQnu56Xzy+xI^U|C;>#LF$S zjM%;lbL_8Ths5eeB#F}E$1D|>7wW($ zUAPzLOLSfo?twQ@izHE6t}%NF!C7+^QGP+l0<3w#4Jtj)x{^iF}Ew&3yY7B#9~{8JzaR|*V+awB_EJ@xuv%| zOVTR)=vuI3#*-vUi)JZ2@2|=n_YUe(<=DL)!m|00#G+%te#YhLAw9uj5J=+X7FYYT zIL1tt8jl2v;U$tpY0+gC^J(h~!Sb-gB2Qt-kXcF`X7lB0i`uxt*ta~1ms=d#uzj$6 zPB!-iOWEoqiPB=gESjUgn}B7=+@M3k674IgOQpW-{r-NnFVuy)gjtbzx#h-Owpv=O zeM5Xhb@VEdL}?k#EJxk?odwI2AeU3Z5_ni*2`I_H<$Q%euIqU+JGDUT$e$$XecyZkdfweDg}& z{{17AmW#|X=F9Lru$=td;IXiT9+g-Qj%I5Yi*{{l3fT!s2*bVzFOri9e%e8BZKzOZ`Z^+>)B)gGsDfq~IIf zU;IfDrKLn^Jn!rMaK!zUt5}uH5tb~OWx6GM;_ETY?m{r;lKHe1pWxEuC+(o72eDWqZNm zcZVcVTJ)J^`Q4Ye*IGmS%BAc7wQd=7Qeqiw!0K{r%Ee@`=(>=2xkdZd7_3Xn&PEr( zQshCBC@sU8W#NS7*d1lu$FGvGc%G72x~^sa(oy5bE{_UHyxj6ZlXZI-80^{xb-8%H zc-{a1M<^{Dm}P6~L3|QFY`$SLktW$+QWwKYY`*wA#vcRA^yMU8Zqd3Qgk$XSOP6|J zX=6i@C@r4MvftAS_a}S&bw_()2@jB1YR9mr*Ns{?^##kLLJ}{xj0s}bFd>@ja35a_ z4VT|PLTNe6EO(06y$xgR*F0MTVL5(UVu|Q67yqRf{F`Zl<>gWmFSm5sI1-a|_dbDd z+xo5{NtBj*%+kR>7QYOB&UnN?VR1equ>{v;=lxYC*H?n2#~2bXw7-NQQ93~4(T%e>bOOLa;e>M6Jz6E``Gl`d5>;`*dT|O?*84q>2 zZAg+REj2Xo8s_xWaiL&IJT!2Quv|JTvAADnb-6!%7~bzUIz{5;mV_-+(Xwm)K)e^T zI!}@)Ep3@4SEFt@u-NsjXeBJ(=OmT^A6Z=vb`ReS^F@CsiI-aft{p~8ZBHG%hUq<; zBvD!hG0TmqMRmcFx3b+@VX-|gu{2)C&Xo(=l|2fU8FnOIZV8KLN$ihY$L}g^a3V>R zmg&sm<#D(zSQfqd;UFw&GK;S^d&alEqX0klfbek%khow*b5|y(z26T zW-gAn0(B|zdE8!M@xLIkwAE+#`_tx4#BW68{~+;l%hOuBaK4=D|MeWyrC7=Fzkh_% za*A2>ug}Bh$ks2e92OSWV2LFrnmyxN^<^jS;e9EM#LF!&n)gPFk!dC#lYO&D5~U@a zSt{1ZeFk+|JwE%au;j`tH-@qGvC=(v7r}f9aU=0^%g^>b(c5!($lV$b-=Q&B8it<9zA)3mKlrs;dy_sHc6tid}Ed# z<%^90OU5CCNMQ-NB(W47U`tz%MhwMobq1ay@p8-9*&Xp-%s2l?64Yf`5J{r6R4Rk# z{T~BsP6JD+6V{J~#pklbQdX1AeV6B>a80TEc@i(Tw3zS^>oR!IfLdVLeU&6pT3Rs6 z!4B2%&92K!wNiv-^A(AuSu%TU+TeaZ?%-A`fW*r!KkeAAs|IejOMvBlGD)Je^ktTf zW4C{R#=z_a7vBks^ZzV=)3I8*-7)N%topKL7wA}BvVvq%bgcHc#M&lLxmTSB9joKk zlPv$aV|BI99S`VO?dCwz{awduQ*-vqw%3oT4jrp)x06)=r;gR5?$xd%bV>XpMJ$$#uveLA4)SLj&XoqOr` zxcakWwcS@^>@%&^Y2h0fn7Lc8{7&#U9jm<#4985b`wtxk9jm%-B-6j`SpDM%KKvM4 z4mwtwc#*XK+_758uqAe^MtuL=A39dA93%<)DElbKPH)!Sd3D#LF!?{_M_t z$Kn3?B;I}xNusojVHVTr3-IlZ-A3*f!V+{#V(Dql7CMg)TGt9JD@Krbxn*SFeyod2 zhX?PWF7?NgBudLdX6ZFz0q%JpFk;10Vet%?SgK|+%kCOgM}y_<1QIW|*zS6Z7K5!b zn}H?LoFq|N9GE3^>A>p%Oz%M95f*iEbk^d?huyn2#H1ODtofHW5;Yau-uwS;^mfrYHW|yaRny#p)M!p zlO#&ZLuM)dF)RlxArFpv3d`|35{qSN*0H*GQOC+)aZVuda?6Z>(>PyxylQ}7)m)KG zk|-_jm}Tw$>IcBmWZPo{l75Zt*NO9xYW)pRNjZS?EfV zC@m$m@Vsxh&f+~-uD0rMT3E7W7X5PUzVcZe7rd{mpikoE7L7jh&=Pyi=qXs1_asS_ zmU_(ceRfUUTiwR2+GSyhyDO>7s*mivzw&4rektsOIf<8Bx^-o54_bDabPOzYEJ+fj zr8Beqa-WW4Y-7*;w}j==J&9#T5_?l+ytP>>%oo#vBwlVwi+_N1$&4NE36|0(B#F{u z!Yo_dhR4Dfd)leVePQvAl2{s)WXI(F#?QNf#ln!p%PmtjIHRTh)Fix4o^M2wC@piD z#bHD28n6s}G%8+LY@;QX)6VQ=xhIWv7J}u-3lcB4R7%J|OZ(KZb)hb)sU(TgV#_S0 z*ENm@i}Ctx$-tV?wr5-+zz&t?1h<>cSP^Gl@`B#F{;j#;MPerXGJ znOv0gRapFEB$g`Q*gDMO{xvqj7z@{W`8!^2SyYSFW!vjoHc*%SRY(%0C5l-x)i z!u4tjOG2!qE|y!^HB7u-MSL2!I)=o{Eoo`ScpM8^+OR)Z+Q*Y5N{e<`yoNC-*%_b2 z*Eb8$6_${P63e;_wsw)y&HX#n<%cvE=)33lIpiX=&tmR-y;R#z80iXPq_IZjvt zpGYhRi$AmMN!wf-=1X>{jk&P6 z$4e~Z4VYzJ%cpg~Vh}~*<(Bqt{#chEt3z*t#rH8uqO{y*mh;^w^#aSx3jG%fOTNr< z?*qH$PkI-C`(L_ilX$shNL%(sgs*;H1Xz|dAW4*#m&|gq{ww_c*h`my6~dDER8p6! z9oY3)jl5F01Hx`E5-+!G{K8%l*?M;20I(z;B1x2%JZ70!>4+cH#ct|EJ7Eb;kXRa3 zWb4UI?RC6hj5%76c)2C{=PMjzTc%fb2TPA7B#F{exg4JNV|VBIg5}p717~6JdnU0I zo6G9by79o?VA;2X#LF$u0zac=?5}<2!4kTfBvD#gGRxDt8}q^9@y>UTusA-KSmqA5 z!Jjc=@acwNshdmU<(5(wd6;BT!qBB)8CgJ*C@uY%CCe=tk7M^EqYev8mdw&5bUywJ zqa8Cs!IIP>`ggqCVwTU=$KF+qb^*(P&LoM_GKpEf%_(jKb!pn8Z=kTmzL3aS@=^bi{{fuiY6Fb7de6{Gw?*xC- z8F^yoSj=Q|`kXFwMw%@rnf`5O30IhNS)H&Pe^Cov<@<+lGi^&>88q zi6s12osm6sI^$7d@!$_-;fPmd8_B6WN<3yqiQ5_58bD`c|D7bwf9s4K*+maK@aE*F zW{M+D^8b!Fx@8-dEwx~=wROL;n#HRYYgDRiS9aGlYO^Ok^(iQx_B&pF^sM2pfoHXA z=`Sn8|K+0=NuvC}B(VRBR^@rFa3;{xxR5I>;jbhXjY(`(?!@3)ec-I-u1(_QmaXU6 zC5wJoX9KWAHX%urmKR+D(S zWlGx%IQO*LZvFw5-F762($bh&>^!gFSI!g4HLN8p*)mIsjjVZYa37QMU|Cy(#LF#5 zbX{=1c*k{H2bN+DND`%`C$kjz4y*-rxikAyV_}I)kex7OPV(U7#*Tai@9+i+7sD61a}7GQ=F4c>yfLUXggY<*eU0 zw0u5P9QX1xOD9Q`mMzS(^5kReD4LhjVX&~+zL8j#ZeWL?GEIE{_j$tq?=SxLl($-C zK9G31#rE4Vv`n$CW({>2_mw13T6~zL&(O1ISssu(Mp)8hmiFV>+%IwM;C`@F3?cDy zi*Y#z>{#_(xC!4U3yC60l$OiPGP+jgAsA!!&BLY(OXOQgU7Ri0zHyVk-Ha9JDZiy=}DyxOXHA7;#bDCX#9C|!q4OmKdAn|g` z_w!eAzI@oDi``Lv-ANLqrBnqx?+2$PR)f0SJ*96iEV(jEpD=bD%kQ=g_g0^HiNwn- z-#)Wt%1w!0-C^#}yG4>HE%lk@OkPv$TuHXy;wCH!nUcC3+`#HG{bB<=SH=dDc)7)@ zHv5uGo0&UsyukNi3!VSr?AmhZKDnW4&vWc)6w9 zu239f-5zc^0d*CNHbViHfKpJ2G7{4+{e%6FV89`)`kg7hRhOPg{{bsUD7cPEPKz9c)4ZMhgdu& zpYyGUbHC8rkR(dWd1krTI&T7QG+&@VyF6G!>1x_UgR0GR{;Ur#eS+IirL$-M) z%!B!IZ7fNmw7g-K?J*C&LS1xPF3%E{e3@mkY2S%x z4i;Sll0<1KWEKrCCqu9}?Y)yPEQ#5Yy0ovyj$@Wu_wjA03_lVtw^-(}H&q&~Sc=z{ z^#e!}rKM&?yoRyutBp?IJ?KF(Bv56#JZW+~X0M;e8 zVM}wU%cC@sL}_WyEW@4_G=sVrxb>_eEPkIQme?I^zEtddzdu-#_L6wHWyNAP_Z^e)#gHUQ%M4~&98j_jSi*ZvY9lOJGK=GLb_j|u*zF5-@fu9x<(6YDI^el-X6P6@ zu*8fcNtBkg%(C5l3x2y}W7@K=!V;S+sY|UGwhlAwQLSlUsq~4&%PpNX+3PM^+v?yR zMb>#FiPGZEES_%P=RjQ=Z42)wEJ0r-mPRGmd}&rI6pzUV-ATOMa_S{}dR=b2Pbk#I z)SDzxS^}75WPrC9SVC8Z4Hp*AZxTz}(rk=P&baj!>N3@i#LF$0y6wQ@So{cEd?TWV z6G@`9L@-N_uJ7?&nbON?lCTuWEd3s^`SfGd3mdncAuojl^e2K*~lWkvjx6JsiU|Cy*#LF#vO%t##Z8qwnWms*JL}{r~ z3D5h>o(_Egbt$Z1x?Wh0f0tN3Rbl7-J=foN1k39IBwlXmF~bQh=_6e5I(g>`l0<1~ z#VoG#BeC1VJY}bousHvaSbR#eF*c}Oi@{(y@SeoWEscwXp=GX@_gR=PhrW;`N=tuc znf5T%6zbw{Wwc9JvSpTG_N*=^zKp_WeDBtfc)6v{fl+8VX1N*fD@$x5NtBk!%;FN( z3$MopX5BgI##1d)Cj>(pj`E()b%hc5Qw&2Fsd45-+zz-Sk0=L9v;}P?x4<9{>IkO3N{3S&%z? zAy}#&ntWAQYzrlpAy(}9%iM?w+F%)ZlElj`?zPx_$?4L&Jy`mkBT1B&P-Zb$VUh)w zZ>!&g3rm{J64IWn!Ol;+5(Sp_n@GIe@~kI&h%;t_B|b|s-$s%sEzg)GrGh?o6Bf1@ z8zU@{MUuK?KVow~eN-=;`x82nc)4Y2XV&^wrBm<~s7pr!l0<3w%q)7E*R-H6k4*+A z2#de1)}K8jF&fM=#dPgssLOOS5-+zjwP*MH&AjJ!1xvr_B#F{et}>qYCm;112aQhW zAAHXgmRtX`{LS8k5kF$F*)1b?Y!r0hB_@zeiVnObx;62(>`tgHeoZ?*nPmB=4!ruW zgYZZBKNwgNI`B?qkaU07ffqWGHK?_=Ubz`M@Dj60s{d05p5CN7UeJLzBcG&Hci?sA zjfRh_#U{guYE|sV?<$HB^+yL@;$jB8SN`Hs=)fyqh9voq9eDm>ZOowqulP@r?9UFo z+S?~$-(1D0ORc;s%w@2|wIcCyOV+B< zSmTQu&g6h4t0PIGv=}l=zq`74b#Bz=h(@EoR(=yUO6pR*D(jAN4oJQSmMUFIyxelW z8+(^@aQhj!l2yGoNusojW0nK$dw9YayWq66vap0~l31Q?NORmlW+^b;ZD~ra$ve{l@nX{f*o*5^#gE3~8LE`0>LEhEz z5~F9uSK454$|gyamP5?aJZdWLvpBx*06k&JkXdw0*(-Y4V;|xZ-z&x>UTz7S+7>N? z2GyAYmM~M2L}|IgEMZ&kHG%muasB9S!V>KusY_4+TQQk>*bU$F>*`42<(4aHR%q!K zlaT?I!>%NW((;&DydTXP36|LYQT>G_aEruJeIL6zUwS-zG>kFN;Ur#e39Nbx=S!_h zL-0=H%UF^`X~|-i#z9tWwRDxsNMUhzlvwijvH9{rYXhDuA8aJ?a!a|99%z|k(rgga zdNL}@8q1@ANhzxTs0w|Xy5og^&zGE1vyW{J@q9szYJvzo-qE$4KS&~o116Zc(c zXh)JLEe)8ZMa|h(Fvc>y8q5`zL?=mI0z0$$(ln|XuF9P_OycF173WUjeA&>vbS~6o z`6-e_Y3a%=9%;$nz*1CFZ;7ykZk1RD>akUZl`C3DLS2k1ka)ReOKv{arOO-Nb6{y) zgCtQ}Mlws+NK@?aE+}4Ot+4oQlUVL&GfUsBL--`#upEh(TOOLTRk^#>mec^tgK8v+ z(lU=(YPL3tgSrfL`D8CFj?NNGqtWa-`RK8s&QKR$PZBS;RBysIHttn>&?c}f_901> zmW|Bf{-Z8_S+C?YuU*2DC9^~(v$=nuSL{@n<(A67y5bl+H2UOBu)JSFk|-_v zn5E95b_<{`ojQCvAS|)lC3Werf}QtE&bBKBmTLwiUT#SqRto2S{7Ii;U^(B9BvD#| zn8m4JybV}1brzixmY^LHi^iIj_%B_2)G7%qg)K?E+|u&GX-v{T&-g1?>U1PYl$QI< z(sON6CRjG*ZMi5co-PtggE_1&tu6iVN&Lv-j=$sOmcyxRzP$JmkNaMa)*?ximbc8Z zsQyOWpR8;Nl0<1KVix5~ZaBv-I~b z{|YQ-#hN7v%kiBOOHnH8kT}x2@o}(NX_0ujrAIE?)corlKYUVBvkFO~v=}o>)!vWr zet+&XtGB}9yh~zn4Pa{*G2@r!gC*F4#LF#(=0P~dHY{I`mW5U%iPAEYS$g;<;yrR+ zqmWO+k}b2WsKd_teeG{*fhDIkiI-c7&034R;xJ!&zO*dW z_^hO&n;NT>IKurSeGT~BwlU_(q-SfsH1O%`;_Q>B1x2%)6CN9p*@bVQMq?2 z2#dFe#8TUi&6gd?2J>Nz?fOOH<(A2QPjJ30)>?53>QYZ5^7oHWTJA7QveyOt5@nb{ zyPCpcyGLRvNVmbC(I>T&DOle0C-HKN+t~n2V({}(6|h_%N|Goo$;|S~VLE<|u4dd~ z9brk6Spw78F*#|K{e7^kO(OAf%i5vtXmNZQ9|o3LZ%7iQimUFr6O_gSeiD_WzFoh&hT3R#9$h8x32ZTiqwmpT# zb+5$IE0?X0IpwZQhq~N#A@Op{#b%i}#!4oP#xJ(l@+3)=mI2H%Y{@LVwrqazzLBux z$}BS?*mbhQx8ycp>2ZR@%PqS?W}+pf+CAKp_+cPPqO?q5meSeYFJX+0toh4CSQ7S0 z>e5_`ohu*tx5Um!GaV8yw-nqQgO=zMUidcDvKAzX(z1eCn%z1d4VIPpea(a=#9Lzd zv6QXD41Te!Ce-DA@qWMK<(3p5)`7QJf9406FBV!PiPGZ2ES2`^;M*O#_dLvn#b>|7 zvi2jJwoP>^27zVSV-hd7Jl8kHay^mu-X)={0QCgle%Y%1b&R}_0%G_31q7O>y5_6rs zoxUz|^g^(dt4QMImS>x0qoth9hZRtl#@Zx_((;8_mNm_*0+#r?Pn?7$&_`mi%w%<0 zQ_*$;Sah0_c)7*AJbOF+?DC~PV5zG|k|-_ZtK)h9n|{dvu(WNcxl35w4@oR8A?#BX zt_iHuE3*ZOms{5M-HLVTm)*G!SZ=l_NtBkR%<@w|I|VG84|F>qEcr4^7N7fP=iY4& zbul|m;^mg5Ti&C^clWJCu;>MnBuYy!X1VVgz8)-n=IuKnEQ!97y3}dF?q4oVK9&oX zI*ue>ZW)&G5G^&DoKFUe(>anvX&KKfJ*sxu1D3;?lP?HM=wXTFjtRSlxwpj_pNnaH zA@OpHr`|ragzf8&`^BdIAW4)KD`pup>DD%|EB}-;$Z^+L36^tv#fW_C7#LF!q zHCS=$4~)m}^|||yBudLwW~r3!c^eu77e5}DE-cMzZ`<$Er&~82+dgmqI5)$Er>iN%wahtAn4g&zDzip+Q4_C-|F=)h({GG1Hfx zd&8k)Rm+5A`nMgcfBe8rXInOhj@2>aNZNnySdE%D3_DhD%y+s49jnGuNWy>BvHD>g zJ2el`)!Pk6yj!zKPUTVJ4m(OTxPHY9I#xH#Cu#m$$7|dVI%s;8~JHY0;>GcN*vWRl>bJ zHoWLqURVlbmK$qWhlKH&9u6?Z{$FQz;?~pp{SCaxOk^fxo`npNDMRKVjv*8>MKUF1 zEGdOjh)N+9icl0Ggd~))kd%;Q4k7b%&i%WdtMy*)+c~~}ziypI7>CvzN`qHh>=4!%{6(;G+t4#Jj3>!Z~obvSii`2_4 zKBL&>u&AbcUV>$cB?(blIx$N?#?6ah$!&GIzOcmalURIvvv+!RHdtR1`eJ5C>gASU z7L%|qpSB%{1IyMPBt&T$&Mc>Q%*4B-(O(X<5|+^Y63dPG>@S^BKcNa(25B8Ete0D~ z7IZ|*`YS7+fn{`A5~8$NGs}%8$$OwLC#K!)BrNXX63fxWY)~UL-tGoiy38Z>a!a4i zMz~F0@aZ0&nOCqUAxg_?X4#~7?ig6Ys_6F-mOPo|$y0VkzQ|r*>`OmAQZKi>aPYvs z?9L0oDbIIJNQlxBz%2E0i!}$!i2mD$2}{ZWNnaj?vL_-I^?0^~V0|Nr=*Nm07-KEW-)BM>!L$geBmR#9|X+jen`JOP3t5*yoXYxg{ux^<{v; z&0f%#IRzv{X?e;lS}tjKVHx}Iq1ggqag2~yu5M;$v(4Vy;uRB9jpl{*a?9zzFTxEh zQ@T{?5(}nS+F05y2p{t;5G;pEk`SduqblzA2gEEo4wfQKpRW{_Y?-C~d}di&V2M*W z{o9gyx#dcG_CAb{N!y-)Woc&;qO{atmQELA@x2Qbj%;2hEQyhlzL;CGedWiRrSK~o z_hU)D+;V;mdzYt1SNjoQiHs*9N{a!rEL!s&Ek<=uY!#O9!xD@8dp3n*G$7I*)|W5M zNWI+hG1__JT5sAgL3_DIvx@n8&FIq*O z71qlwHFDWIigZ8wEQG!+&?X^D%T#7@NN~pEG1GPK2ZbeHW~s-&KpkK@qB8WQvNx%h zTV|xO>8SblUg8sXK>;L0X<5N6io2^AjxkTIq#9LTNd~EN6EPDhhpRv1aoPVF^4cu{3sVrGW0I3JT(Y8+w5FhvetvH;7|_M~2J z>0r7E`|`p_=M(fLrz;6jS{^cs&nx!SP4r(~p9xE@%u>G|o4{+-;`My67+H{dxn*}w zGFm1jOj!$-1=C1~((;8_CNvsu1w@B`HSI7e`Ze*JrVVVMDN#eM0KxmgCQ> zpv5)gDqd?_`-+4pEtRU_G0fK2&s&2fD8NCp(LdK=B92QeJD0FKzRR^WCWEE_UQ#c& zXqQVz%d2Z|N1;yVNT~H zi6pe-yUv>d7Vk>gARX9jc<` zs`X6|unh7dAxg^xW*IW~jT>0}Qj0Ydmis3qed*Mb9Vbt3w(>4mj31DCxh2bd1X`A7 ze0vWTqo*W9X<5uH{+4@hgC%;nQ5#{2iLIP+G#6<&o38qhP7Ea&KQ@aXT%s{ETGBFf(U9z+?XJuB2XW=~W{V|B#fN z7PFu)$(|%cX*tI%+b(VD3>KSGi$(~`OPOWsBDPIF;OB`?bhoHP>gAT{P1p#p)hB}~ zV7aYJLX?(tW~o^2X*O7NYI#f$mg{FEeVL%m*8MJ*{o2Da)~Ye7ms@%bWXss+BIEXe zrGY*PQCjkuW!uN8(O^0D`n9#NM8!!gooBIYu$#kOXM!c+6RDS5G{&+6)U|D@TnEeT zpCm+SDOVl$`^~IfTY+WRftHJe#rLel62R|^#l}n?0+#DXNxj@M%OxMzmkDhh7K6q2 zGzn2!8Z*nSatDThCFaR9Ct>*}vkbLk=g93&``LnJ&l^%Nw~X8}3@y|367We#>rW&^ zY3ac%`8Bfeoy1uKhi(v-taFmS*d1WUmNBln55ZzMlGMvBj!*Q_k`(c75cFldISEl( z%$Vis=4~gSFYb5jw+T!9d5I;d61(?0U|4!CSe|B)dby?NTX(c{{IawJST;Q)Axg`9 zW?A;Ubud_R4qe?VETQodOBeq2`O__C8~{tB=cHb4X|{pwUk0ai$LsvjHLn%^2&H8m zv$XEJ58r+g7~k`Vu()54ScZLL`kQ`owHt;T>3U|IB))XOb*GoRuzHv4r8e7Ym2s7K+CP+DS{ zWooUHU%@hRRLjf4l9C|li{pE?e@Tkjg!}!zlSsYXa^(g)(e63gq%ZWvWCjUQT2h&1 zNphF|V43~>N2;*IT$EVM$Fp^Rs=o_PtM4oOv#?%n@o=t!-yi#Ayb#|@x~L`zQCeOz z%eA75!eBV4;+KSMVafd4@|*W8-gZsE0kNFC*ugNh7E|}Bz|GJwoOs$zzB5}W)S{r(g4RyUL9cBzuYlbyR)c-WKR{X-(VKB9pR*yug zr`GoP$K!wBsnisFhO=Qa5+s(WU#8YNe`dgrFPmz^)S6x!67q*rYl+&{PB67rY<6Pd z((vol+LH#>IBRy#F1rRSn11~)7Y6)hYHjtJ*_brcC9@n%t=SGEN&j_f?Ux_;;ZRy> zm|9yhhQ$8+)S7PgXq;NRY+K(Drq(`Mkih>mwU*nMJ;#3J!h1v5;*Fa|GF4j%$4_i4 zQM%E=g)p^NFpI?e+0>-keF^@Z5t1SCqfo`asSqrf!{s*Z%cTIcelUcUA8 zJHf8VSE<+{3I3Nw)kuiyf6*vW1J5+v{Uh+MkJrpGABDy7lEhN44zpbH-t`kK1`9~N z+|q4QcicnIu&Oy4EajJx5T&Iqvp9}P-3EP`U#q)j*JP#LGAZsn7Dxuu8M z6I}NXTb#h>F!H{U5T#{1vy6Cq3eU_Rdl(oA%iUy2UwYSJ7t}VHd0<~seMr6B;yKw5 zE$Yn&_q~buSzU-A#BpC+>hW&u#Cmnk$Sl$CBPCb zYqaN;hh?np3KF8UTw@m7qb=Tn<&?)AQ(+0bCb1kbpM^i;@ydnZ^G`&^NH0{?Ni)Se&m*EDgW1DV!E{U*laLv!A41ZYjFuI$Gjv z-Wfn&+LmZp_#>1S&6>F1f2_-r@ma}0bL}{tXEOYIG@QVD_-ez{fl5|7Tm#%4S|1!_zWHea3hmd-?<$TvO*q6Qo zinxR2@hB3aw6tRuJBwX7QPexUrK7M!+>}@+=bm)w2qkT0?XLvHl$u|*~h0ny6i350xbHS zNr=+2l3Cic8|49(Wr2ft3(Nf!Nnh?+v+bD4A=f;xgocrNxy9L^?JJ|JH46aCkw_Av zwD>TK*|hSP!1A!EZKSZorAjQh8`-+==9q|A7+V{Vdb#C}W*uC{l1EhR2^K9Q5~8$3 zFiYNsRqesz`eoE9VF|e{vz%j35WPP$UDKpmTLV-h|+S2 zSsJyRj8}d?mN2;_EN*us7Kd}}c&yr}-ebTrONZ3UEnP2%Gs|nu(qLIxgM=t8kC>(G z#^C+Xm)`CTQibKE%%XRT9rM?`b=(aswYHLax#dZuCt6x|_s6T)**i#x((;vA9t>QI z-y$g0uKGh^xqesDm$i4?`7bVD(h#H5T&KEE*`^_ z^(?CkeeplE_LZgARzK?Bf|ZR9x-ma)h<5~8%U;+B&3 zc$c*M!O~xa#rK}XQlvgxUt%V^+y~3p@EtAHEvw->??=ujAxg_WX6X{V^$u88 zY|Jqf7WYhvC4s+N{b28M&!8`k{-j=R@o1WWmJXHs$HFppCzymNE$5l#L*>DT!E!Hg zLvLZplUbVhu5D#p4_2J^z@uQrkBd750(b$v&IWc zOqRrAKA(N8(0>1>zF>KjP3q;ADyzM58A~4Of=_i?y&@q>i*_yC@9(dh9S4^7E?!o` z67W!B`O=Y{Bagm00I&7++d%5&makV%pyh+9Rukw;g{>q+X=%bN^GB@-1WVB}8y5(R z<0FZsS5>x6PP)?I3iQRb3aOV{n*Lz-*tSJFWP-)KHVIK$jG1NN>V#@wIp*W&AS~H3 z%k;tQp2fMa8q2|QIFQuKE#1d9!M?<5Rl}##Jwr)|(lVA=RwTst1y!!N=pc{*^R)gATuPTFW`XxbmoUy@dk5T)fdv*et9pbdTL(0lqtVTpYzu~d4_?u)hS z`~csr{&_H|ms^ex%fe->q*?EK(3evqNr=+&hFPxOeu&q{MwDN2OIQMPB^Ld4Y#9sQ zGYj8MSgZ!Ams`TxpGV8yX2gxc{&<-5nrKN<{%z9hR@x=}QL`;m-C)v7b0&%V z-K3ZMW;O`rSm|>fOnTYQBT@g;q*s{|CF5bztD`-M{DVoaoz}bF!kRwCfdq*q>X%6` ztK$sVVr`Kg8L{`{&q)yGM^cvK00NwR1>_>VMIw$Nrb; zYcJ!GQ@3Uo`NEPbvs~-P?)b)Syo2xd@HHUya*KW61K8uzPw(b{MXL)5QCf_cWqAIu z63~~=)_-X<{pSi}(sM~)N*J)cTJ0P?JaUS)BlU92y`x2OemQhgHd{-$cH)?f*Wz49ES7E)}@?_F5JaXz?F0%;q#i}$3QCe)6 zrSF!c>R>67Hn*yUrWj@Q%GDAxg^*W-%GI z58oWF`|gasu-tzo=}Uu9wu~M7vGX%nJin59xy8FVn~dzaB^%#`v!Q5Q;g3*SjxkG- z#U}WipYC)YBVmbqEwOCTXYcbXTJEkkSQ6%vdby>G-Emw{H;Rzhd3H){Hm}M)uw22-B6aXg zW8$q|eqiylGM*qTFJ+djQtU5n(&4~f=!@rkQZKhOo9v9s*q4CMreN8>goG$9y3Eq= z>1muOdKEL*T3D{XmGos#4R$5V_nR~B_btzndby?Um!oL;>t!!3ur$9+LX?*F%<}%6 zMIY!(`AJn52}{&FiDmwGwoUHy*XkY6m(|v!UT&GF&z@OyFi613$Yfg*qO=TQmW$Qi z;F;+`l7#zy{q7`iQa3l*tR3}a!cd5{%Eml*B-ByHtIq`lonrRiHeKEyVue4 zO6(Js_zx0GQ*(Cr+G9-cX|QOoA@y?0qDi~Z@-h4}-X%TcM?#dANM>oj=Jb77_iwdp zb5vMDKT0gqJF#QSO0Aci2g`#me7D>o^>Rzycs3cS^--%8SbAiW5T)fYvouMXw-GD}9e*YZ zOP0_vSWi#aU0~VLpM)qa-^-%rKO4<9>dJ9`d2CFi|M0# zPlP4rv&2$xoZX3ERAbX_urz!_>gATBleVH||B*fTKEIa*Bt&V^XBMZ{_wo7uC@-_O z!V>UBVyV4_?JGCT>ox~0ZMTtnx#i>j!DzXC;fE9SMK6eiC@uY%rPcRl_#WQ_pSyn- z7RRp=OM@Hi>5guL;_&H?L0UZu>*bbFefy#1)6u=1z>-pygeWbOm}UB$nIm8sTeI(b zsb=DdJDFu#Z+6UY<+(ov`eGAJ>gARlJ=&w?jC-Ygu)mm7* zen>3w9oXc(j>9!v_qA%1dby=@&U0MGnlJu=*T>p6A|XmkJhS{ zC2*C*vP6sZ#j4k&JJ6TD%SpZ5((`dVTAHl-hF_yA?LtD7mZr=S8Wdv$7R}bVvxUXk zNn%Od&9-ChPJSB)mYi;+UT*nZyb)SbR-3MYzL@kQAxcY6X34$Q$Q~@MmOQZ+mRyRzV33ecKXWHWzV7Zh;LX?&T%+juAHcqY8Ss&;nED_EUOF|yI z*0&|^`a$SRtu3TpZb^E`zCLeudFUgsEZ9y$looepxvDn>r`DFgY91gg-Yyc0R&{p% zqPb#gELcW;BK2~M#hh`te<^65_6aPFf07WTWjC{!M=rsyod0OvCRA7oWR|iPY+pI+ zcAG4)yy!^k$bx8 z`Mn%U>gATQReaE5+ddihFEhuG5T&J5J>2g%xZ<<|hEQJZ^m-sHpZ~V}-w=we<;*#) zEvHObXlprTNw3l+H7jb=*Gl+u9S4ziudd!6rlSUjktD@*lzw+M9cAJ(AP1(Senya# zKc0??YJ6b;Oh=74PU3zy9rYuc4G(pFSFJNlM~#XjQUB9)lwSVjvoIYMdXYr_!F1H( zVG*NY?U-?m1c@c;m+7c|_ZhH?)&3hW9o6(U3Hih6s3jQ=`1QszuQnDgSHDh2eLk@S zXPOL87jFp*rrE~Lg#rJ!V9vF(w3#^d@5Rz?=G-~|f5F@~e+4EzetEMDOh?)Ikfi@Q z9reo(Y?c{R4W^?$>?E=OJ{{F~%`BXbsy|D!1WZRQ-9rNZ({z;o>=4{abY0zKF>LX^ z9w3>jt%PwMwv~7m@Gu#sqqatom_M72dOSXR6C4G;+R^)=*y6aYm25pLZD)Ie$R>IV zU|ByKO6uiX&--&a;s0{xUBwXiUt%Lji0XgQsL%cvPq$e7nnRjz^h;rRDYJB3$lg;D zaBSRRu)N$w>gAT;;*-%5ty`9D_}qt;6-N!!?w z(`qBf_Xjn=a=Rv}ms@I_U`HF%(&BLz`(1q!qO`1GmJoaUuVB%-WKc_3zR4`dDzLNJ z$rXCwR6*%Gq+V`0v7rq1#l3YKd?tQm770;Wb~4MN?mO`vMHe(GG!>RC4@qB+ykrwa z!-u6mg}!8zCiQYlg=&Lv-5=Uwaw05a7b}qvr6q=0{N^=22YtCay?lFNiQgcxyy258 zZ=dd*4VH5ANxj^X+-C<`?2VRaf~E2j5~8%+V3xRKTYN6ISE*9Q!V>B!vG_G+>wfP*C2t+GlzM4} z&t8AAB=vHOXLYs#=^fPK2UtGLBq2&m2WHt-+w?710@^y*2}{f-iN)EKP2oJUxHAa) zV%CM!%Pj{!vz^q7g)hc{WkhchqO=TU7Blao_?5u%QQ3~d60lifY5r&(E*`UH2jg>o zr_Ydjxy9k`WehnuZBtpWgeQ^^rNxR_5>{$eguV=q?dc&bj@}YWl}T*zNFQ{!JM`uD z1yV1!3@FLA$>vY0lmts+G6_*y9GT^z-jN+(IbN-^pRi=hEPBP*Hu>;}g>Aqxs6MHe zTN)*8z`k7cbe#{D4=qTD(&EP~E+x(zgXNL=`Y>Th+#>1A$Q)*Q68tz3EJ41cUT(R! zgFSogKHJh7Ec>sM5T)fXv$PD!tN@m!9bBV@C48&IVqA_Lk7>{Ae;X`^dXRd#W%4U_ z2TR9s8tz~A^dlik%N1ss^L^+mupEp{jT06xABknu2xc*}Y3%?O{{&Jmw=65t8<(*S z=Wl62Uv69_AxcX&vm6WY9RrqjbH-i~mVBAzogTY;U99y}d~cl1Vp1=+Oy4>IErS=f z-2s;64kScr`OYkFJI)LPOaGw(w}s`ducR+dMr<g5*q>QQJ}8QyOR zSb{7`h|*HE0UpCty>RI{SSoATJQ9{zKZ(Wd9($M}?c7v+ug|e^q+V`G+j0`mU!H!B zJP4KsRY{1_(wbSG-j81hme`FYUkgj%Hi>2RQ?|Yg-;sxR;(u%-^>RykpdMOG8a$p2 zmh>PJqO|m97R!0wQD7-^!|;o+IQvU1nf$BGhpk_1hIK#h9;ugGK2K)*O3k^G2Y}_x zV-ljYSTalArztjIu{<}ic#D6oUF6CvwIbO>QGCOUdSJ=CLh9v~C%LY;zC0KWMENo*NAyw4c##NP`eAxg^y zX1QYNh2K^1s+wphECn)4(_`#gARiMQ-By(y#dLJ777|kc22L6&m7x zzf49YyyJUge2TfSgzS=78kS*0Cu>V|T?EV6@@b@AZkal)K3XUOV@4$1fdsm_q91mgDK{UC~dUY}*TcaoR{il$Ku1 zqHk4XBv?9c>%UM~Udk*nsjM$172o6im#|<`FSn$;a>l-xk2rM%ESvX}5T#{2vy>QK z2Cpd@XRcW#EZ2i2eL2#J_2uZ?J9y0h{3)rITWW4&_r+4yubB##`}rh9X|ZD#r_m{q z&=-5Z;7!636(X_xXw9}`UzUYw!TQqwJgJvkqB8@qFYBf+*aVgmuStl~;=wHWi%)L{ zOV;3`JA}n|x5P3aXeIuPx`~bOUD20DlX|&jaAS7SAaQUxoQ`TXk%TBMdzdAsVf-Ml ztbb5^pRjzBS!QLk*KwH|I}CxoT*@W&a*O4DZ~Q}2uf2N*eR-HiLX?&>%rd$|lnYp{ ze29w{maILJzKnRnCcXR*o%jS6&!41TZgGg}iI&KXr|?P0@g>d{{s^VzF0-`gvTOrb z?zirHPFUjiN-WFG*y2&@weJP63^F41a*I_fc1`L2?ic-_FXelY5T)fUvvldX(HSh7 z?H?u!OK7OX(#(`y8^8Y`8t?H%bR+e0i}%BG*q8W6?eSXQS`!kYw3Kdy`+fI`Kk*ek z$#vRg2+Os?fia9eHrlVT4Au0bh9aZhv z4K}hheP{$sN4+p4alf06T77{H4;AY%(*&lY8ucJi|I>6-q}j8XFda3ZABp^f=_qTn z8aOc$J(L8ACF+;ysF|M_Fk+KccbJY^IEsY);dE5r+Wt}S?kKaT?F*NyU#Fw$op;2U zrX5DbhQM^xwpSer1Aa3dRWHg3lZM-Bb%g1tx9>^Pf1Qr{^#8QLVp`*ngjn zy54*dPDg2;ZRiZsQL#lj7K;DVbkw4X?5bc+PUA4x;?*rpGF4lNxJqm*;k)nBL70v@ zUxCE@*>se<%_p3Wihfi#TWoRM!z5eJyZqbg7jKuqy9-@Ekb3#n)A_bLZaojY9y%HR zmo7g^i0XgQXu$rLQciw2A+a-Q(;H#QlUcsjWJiI~$-y;Yul6jJ)XObFT|Lnb9E#4GZ{+b+j5 zz~o#KqO=4uOT_G3)?lggz`B{RWXmj${EJj!iH(NB`f?+e)XOa{GtIFt4&6UwL0^vK zkr1WjIJ0Qfnd$|W5&iab5SGM)lD^n%oq<0?dtd4ou-KZCdbwq#er*h~PISgEg{?3r zAxg_lX3;fkj?a@tnnv{$mheLo%ju)+&f>AE*YO=iHseUW+|pv&W3)^kZ;w}h{3eqS zrR4>)Y?(H+G4$n1`71+&#VbN$F+6I6Kcn}Z{qLYJ=`~2b+|ps5DTXxJHVCiEW!57h zN{d!wJkzMOsY(y96x3=pQCRY2mf$efmk$m;HehjVM(X93F%dy%@mMn37S@;P1|&pj zsm(0s&(*;%xEQwzV0GF->gAT8Iqb@>NBd9aVA0t_LX?(a z%rbl6%oAXV*x=wKEP+QPmL21nrPp=uYS5P{zNB7miN45Q3t#W&1^iN&b07&(T4pdy z**yWD!Se2THBVu2j*?jP?y{?;ORUd~1tYVf0 zJzlm1OJ18j+k_=oW*NSj-6bto)$AHrlDm+4xh1yeAzWV$xRt@l$W^^bh|;o+S$yxAS!P8pfUp)dOaNWI*$Rz9ksq!vGkZM3^`+#-F@+&Y%MWHLJM}Rh z^M9~%P8XK@F_OLvEzP!L4xO`b3g@ORsh3-l{U@TO@3X60pfB;uNQlx>tqC5(?6Y`? z=amQFzkMn!amOW=!1-)D_Hb0)!O)ld@1$OCdA*`DT1qqVmai>*8MsOZ;yc`wU};Uz1%Y2flaJ_wRYM7mbuy_L}?koEP5}S z&jrif71uRd{d2zWc2Z(l6Uv6OPn=ncUkPmIM(X93?Je085oNUpH3iE`FA}1(OlFp8 zCK)*Cb~(j`_nf;$}i$yw{R?xn*_W1zg5HJS%z&EE_hF5T#`qvsB-A z3->SjV<+ee%k@}EUy_EfjI|q>_5MW!*V;9cJ^3@*dEayi5|Jv>afTlQULb z1WQ+gZwA8Rdszy1Jx;BqG;v)m zEGg$DeYw|(oxhyDG8Mmf5p6)~<(5;{><@7*Ic*_WzIGuYO3OlK*>AjB7nZT?T@IeY z5)&`6T%W^^ljrQ|>J7`-sr{s0ZuwHcHXt{PX5m-P%S4e7rDX%N_^jxf2$nAs_V^1+ zzy*n=dOG_VU4y%w1Iw}*q+V`mSE>!JFZ-`9#QS2=^GJx&vX@!r)tQtJ7Tt1Idxga@ zL1LK|!WNIO+lJy4j&(IsFSj(Ue-|wt4O|<*GUlU4LX?&`W=Z{QxEU-{{KJn3OSa6C zFoP{)E4?~zguVpsCG~R4+d&EETqp5T)fEvrG?2)dNdVAJ;3w5}qirw0Xz+ za&*@2gD?dq}v)Td|7bHU<2nuI7VWt!oBe|O&a4={w%aMhlB!jkg07sC91|D5+s(W zU#6obSu&tm^E3TnIw~oeg#6)j)X{gBZo+g_@k*hE%hj*bQC-8Fai%HRaK9!jnDdH< z6$bofI_m4eHJH@qK+7>O9o4xUN&2tTQNR2^lOgkV!E{ugDkS#br=!NUcEsr@&(WP{ z!E{vpS|sp4O-G%1z^**!Tkfw7TZz66NTzZtQH5f94AW7SnvV(Iz2E%V=hK z<#4=l-9nmJ}M&cF5(yd9o+)|}#e_UVQpLIC_eaY`gLX;La zX347l4&NcMv+L=a!s47Pu{c#?XBs=xN@PM`too9Ax#g4LL+r~FovPcw;yaXtC@s5~ zJEGB8BUT!HmJQ6LsZiDchkroe0h|*%nEVEA4 z!uOL^(zc%?EO9p_mUwM8?NQD+csf`PR3r6r%PAZ7ha8x82cN$zS(k(;EvC%Uw0+PD z=!@T-PD_O)UGLu=Bgk8P~ z7CVc%YlOuuMPm8uBD+(PTyzmWM}BxZsh3*}m$2q${A`bRN_xyCAxeuAv$V@zh0l{U z-g|VTu)LI6T3=^ZOlFljg?Hj(B1pa5a=2wLTwnTLev8k>O^G2PN{c_U)cjFyC-kN3 z?it&K<$9{5FXnIA`(yf)|A1HITOA_xa?7}~gK%HDZ=dCR*p5{@MnaU9C}t@=Zovq! zXj?~y2}{&%iRF_%d(D8(xT7m!-M_hj)XOc8&8A~t20!eHQ#e-3Nr=*t%q-(7I-`N2Bq8ohP0W2nONr=)?y#*e_TpiQ_-y3&R z_gb2;#HUFtrz^0&=o>8950-6>NWI)LeA6hje}vL9kXc?G`!WnHjW=w2 zD=hBm5=-`GcK2GpfB83HN%bN1a?7H7zsh3+W*I@6$X&;$w0e#sxg@hC5d~Y`<^#VTlGH1Qte0Dg7_{+@hQ7h3iZ2V~_BOySal%h|+SASz?;R{shamrK?*C zi{k@{#d`ufPS)F9suc8PlryQ9Tl6+;N6VJ8EAiMez@3CBEm_QxeY_!ipI^5vorEP@ zW^wG!*8Q}R>+z|ZE&WKn-16Ks11)|z2aCWm*4mVWC@r6vrDx*P>d==N4X5`Jmc%Sc zU!3#VJMik?+-(SbIjm3W<(79h`{2Ivb6x)-U@2utLX?(@EpfkJ`;-lClN~MZ4-=N~ zhZ0K%M>bK^@6+>Su-uwO>gATX_qSkQoX=+9-RmlLBt&Ux!7O$ibnr{ZHRo0|7Z$Ha z63g)s>>N2^O5O~x99mB5<(6(yzG%s?9oq-`GRlpFC@p=M#iaJ)iLi_vsFZ6ZEcr6a zi!ip|*FDiD2`pxHNWI)*p~aq1^_a8q09f`kCLv0T1+zrYw!mu_?&S|H5SF`-C4F(+ z&iZ1o+&mR5t~E%#+)^=NEB2*~{>!3ZIaH5?C@qVaMXzFN9Q5VZhipe-iOrT+-cMxv z{jPn3T7t#5#*@N&xn(7M7%^lD_!rvi<&> z^Z|HYxp)n!ms{R!XDeOGi*$Ul%iNQMC@t@q<>=C9)4`(iX#8nmiO7{$bh@&>Ol*B< z2lQpfOHwbllssGl`%>*xYn*%V$|oU8OW9Vq-yc!@@F%b=`W|*kSiGM}EE=y_Ut%r7 zuY%=9$$f?Oa?82l?EPdf8n(v$ewGdiQCb=?OW7|09buqmkY0tm!gBm?%Wo#VYQ8v+ zLsb_}dTxYCuhC6Ol48@`ut?Zdjkq}SFSBf^qP3zX(vp2P3cD>|6tNf*R9-HSfaiUB|&0|`eo9q zZB6#q2Yb2O!=#t~XcF>=lU}-av|hlZ*8rW0h0E2ilU@xx-EcmuZk}cjSTLU*t5g{9 zn@O+KuyvR;^v3etFzMxRnk4<#Nv~gi;Ptk%*TJOMy#x~b?~`6F{&K-duNezX8^NSk zMluQfPm^A&L$Bc$??k1ac*nVTD#=uBC5-jiR-$83+v_mtm7Pvv{%q3gV|mR;*m~ad zXpkwkI0dq;XG{}zv~j%B?p9jp3``+Bzuz(rDYzo7}{O$ z34Qs{T2HghKUWw-UP>$}E7(EL%I~}JnRxTgq+V{Rf4@EU<*|(m-ocvDlY}TOYndg~ z*%GfZ=C`?_FSpDb)ELj2x@xz^D~xuBNr=)C#4K+% z6uS+5xsx`nhOoSpSwwd`W+^Mj>OnXe~Ryo|5UUDwi}@b%UE@95~8#e*T*vruR&Mv zO~_FZa|a5`H<_jHE_Ro6%ivZ|!Ls-ysh3-va{8cU#^PqX!O}RMgeWa~%+j>6$s<_C zJjxi37M85HlD?!oX75_N+PyDM-)2=N^>T}OdNAWd(0W3>vlMtn)6SK5TZiUbI zR<)TjSygATURoFq@5B)v(M8u0HBt&TmV3rw|Cg9x#y*0DE zg(c;Kq%Sch>^S*)dP@tiZ1yDea?7G8W3eyJ#*t%S89U}nLX?(hX4!rt3csd3Z;O4P zu*7_nSe#0-N#byiGH1ZzwUgA#Epg$y(DFHZPhGI|4<#W=%T;Dc={~y|touH;h6jWt z;FH9Xc8wjw=pU?96f7CbNWI+Ry66;I`oBzE50+c2Nr=+&lv%2+9)goAEx!637Z%3? ziAAFcyL-LJ@Z4^&lq}|3STDD%EO#3%zLB%>YU#VOBt&V^XpQ@Q?*qQ=p)cbt9WMw= zw#-t}j@>2g{_e^cu#~bP^>Rzg$Sr7D=5p^2^rev<2~k>VFiYLjzBsLJ?E2=0uq1w# z^ySuFwtuN{H51QYrfwwla?9yW>>1zPK?Z%nvcr#rC@lue65eU>Ea*#u^{`A~3I8Io zRzgQtUW+RMLa9V9{GcLX?)N%<}xvvt3{@KUCtQu;j}uNf+2a zbM0mi+kwU81gV!>TxYS#d(YXGqQRnho`fhZE0|?T8;i1F*)`LoNZWs|UEKX9>C5B` zY<)R+ZuKUx?6)WNa*LfUyVxCcv<*IS*JKR|QChY#OXRT21Hh6uoP~Bq2)6A!cb@J0IVF8S8ybS6BjnNGzLQv*WQ) zmnHb6u$DVWz1-6KBVu?v+Ukmct z`|2euW9C0ez1-681)IFT7~FagSgO}cEBp~k%R^>4Zl8@)o~d2cg1+3;c-mW7k{ru4uUA^Hw3f}> z1q*tWE~!~jqrTR?0c@M>yv-ZGLaX1E)XOb{W6tCHGVPJ(A+R*JBOyvlr8amBb3OZW zGFZ~Kju{~=5vwE?ZT@seXB|ttKDK)osh3+08%N@K<(HfLkAvkx7zt5YS~5$W*kawm zl6$J@L}BrEl309qvGt{Sr;*EG8M~Z8>gAS6m)2unjLTW-gC$&}S>caRT1=Q_LUc2n z@+|qZ)f{0dkXdZ!vGv7tV#y$|G<{3z<(786*~b}*#k`FGOV`gNL}{78EOVBJU4_2* zJnFwhSnjWu^rfyDo5JaTd=!2mw9i*kFSoRE$icpp@2-!>Fl$Q>DEtvh%VK6R>HMSz zSPBxhI15Xhv&6D$7CT2idbU9$=u5*Jq+V`WYRA_7%lh?qKwoyJlMtoFi&@S--k}GU zSvo_!geAm9VllbSZUZ?yX>|%L^>s`p6+A!UxF7T;kP@6R3`Oui~ktU!D|Y?^yk?IV2k_8zM=)+*13*TC@ybtC<3po+n6%(vr_CeuqwKf~9L^-!sAz zee znh-$Zem7CnvOD`_9^K0Lz(i44Fp2t~CW;K2rg8L{DvLJYp5S25d!u0M8RSnwRR4=c zWA?u|)cx)VmYjx7ABwH#H<{&KNp|Hodg5g(SZ_z^lX|(OY5!YjsWE$#HdtDBBq2&m z4`!L3GbIo#O=b>#DJ)s*Bz>`c#IBb5wAaGDT0dJ-FSp#Z*ok|!kUdALgJthh5~8%2 zG0O*o^Z>A&7}&HxSmM`9ET@|@%hUNr`=Bo&BT2p7aw%v4_Qk*|1n-izvLGQ!%Y0_J z-{vO1LDZ$kVl9Jzt}uqWODtP1v5Or`?*-xWG0*ywdbwqEyYgsB?a^p7^yRTB2~k?s zF-!evowT4YKDu`+2#dRi#4_&|I|_W@eya-f<#i)cFSo3@5rP)00ymr{tfo&wl$Ky- zDO++W-YKbE{Ch27$&*=XK4X0u?lz?ZShC8Kdb#CCcR#c^f9;5ON{p+M5TzxSS^i3Y zfPI;qf4!-&q->D%rRFsL{N;qaR$ytYaig$aZV8`19xd({gwmKaZo#l0+Jf1b9g-T6{v{K)T^8u&jGb>gAR^XMgOAxxU_Hu#|X7LX?&g26(2S9d&m+ zSjr6_I9OO5H%crO*0Ocq$GKPvSd7Mzdb!26&2zMrIx_lJq6;9J|8UGSR0m^kw~LQZKjo6z_)?(~mOg{om8=KdPT+LZgnUvjx0rac@t0Gpm*Rwk={FLhv}|VXc^*kS&34Pf)pVZ4OK_h%{eVO98Cl~tSOT%yMJjf~8FbQZKjYykr~ps|M9@(yMWG5~8$RW0nomym~=jmY3fbE-ZmQ z63g;yY#DP|Y<(Fl->Q>(xy34J6!zuBidm-6mwNR`h|-eFEM*eg;FM=I9jh2&arTv1 z+&;2ptYCaa{8E@}O;RtnXxwD`{o0*MF96HV`XoeY(QJqN{SU=_dP84Ej=Gm1EV(jE zT_bieJw9)rKJ;biL{cxe*iALYzRWsr9|e}-RwP7esmUzYN4%^LmKsk>-4vE2KS^I6 zrm))l6twN_4^pKc)ZAW153y#5~8$BW0uzg_a6XDuZZfOg{44d88Mr!``@-b z!za5s#gclt#k}Y~v@{7w#%Yg`coL$ttYntaWvbyZ|5M!wTJ8S1c5y#I(ie?r_L1W5 zXKLXxR=F{$ms^(DmqW{|!(nBiFAelbh|=Q2EUlx{ahu#LI-$I<#BG;YR&He1l-|rZ z9|OzS(KDo8ZuxG!@C1g~c~WVsYBco>E(WN)z8x zvg{10ms`eeXE#1-78|hvEE8SRP*> z^>WMh;qS38E81`D0L$3dWD=sZY-E>B^@(D?FQT4t+Urj?~L7 z`Kz>X8PmSda|u{FULhe$%XwxgTdW&Coxb7V;vK@0C$rStZjXPde|60gU}+jd>gASQ zwQHiKPOrImA36OD2~k?^GfTxLohrjJ*5kmFFkwmAE9r|7pC$}h_g5rXx*a3+a!Vc8 z=RF3NDP5{`i3L+EZ7gjU)SBo$4f?Y93<*(MJ}^t*hxc#5((~@AXkm#7l~{_LV12Ra z@gf{75syf{+>#m1rf-kiM{Wm8vlk>pY0>V0`~8YbFRlhl)sG+Jge4$MVp*ugc5@~N zzv8>2W@~&Zte0EN`ml=zktT2Nf@NSy5~8#;VV0;3c@{8G)1gH1o5GUzx8*nQj+!v^ zEKYje*Qt>KlU{QxkR-*V*U9f}+-Pl`67^uxE3i6A`L{_g+jjTyM{PJ?z7|Y+nbjq6 zznk=`c8>kBvR$-0!lc*uCM4>An)I4#ux=epdbMv&BGr>#>-a0O8hU%-9lcV9BuFe# zzf5{*7cgM;ccblK(#xPb3HifGuPU?leuGIbZPyZo%hj)wUasT3a6aq)cC8w)V4l(~ zT^R72Nw3@6H(^qLhwL9P={2c7N&2soUcdanL#3|1f=REI%}MOPPkK!lUkgyHVSY zu=Uj2VRBn+aUAzaww^{u*_p^E$sQwoX zWA?x7jIMz1TbjCg?n7b8mRX!uu_M|w6EjV~(sVnims>nl|G8S~bwFa7J%-&#J2ZPbt}l7RNWI*$%cln((XK5pKLVDLW+X&u z31OC^w@)nvOZ=Bf{NMiXG&CX^!HoD>peM!$G^>T~OJU0HaVbJz^V7ZV(LX?&_%rY>2 z?tQRyZFQrQumna(ELuC+)gQ|@Pw;!>Jq$^`+|nzqG4^F%yu*61Z0JEkl$Mem@l2yr zDJ>1K4F2icM_8OAC6-;&*!p7I@;Ah{nQx&lyN8kxrNxL@%C%Z; z0ez|4^xk-3Njfa)%hdYpIltHamnT49h8vK2g(ck!`;xhE#bdBEHzFZQ%U{eAxGXvs zEET$)uo9MtBN9tPM>bL9@YGljENA+Xdbwp&g_UT@d2s4GSiFXk5T(V2S*jUci2%#G z;%1A5#XCx33AJG#l*w#sfzMt~u0ZPLmh<}C(DL?G|4U#AtWH9d78hn&(8};ISmq3A z=_D)#GRyjR?3qRPpu6~-U+5T8FSqpVdDrgeWb?nB~*`g?P+g_pQ6H zu*4meSbi2`+p%~)<+RSp zP+JH{foJOF5ZcMxPpWzEkz7*zkjWA^Y>u!ZQJXN zu)LI6dPcEh{=6L?2f-3*MC#?1uTMO1-9J3~oj3Hwz848mT6CFZ<#MBjU@;s0?6R<2 zKQ8Hu`2=>gwE3VFTfx$B5~-J4%5-8=I8WZz!>+?X7$^f@NS7sh3-trMh5Wtj_jc0e$)Sf`lk7LzpF4JJb~{(`P)+5f*FIEAhUTaUfbsl)qRJEUqCWL}|IqEU61xj0VfXV(0XP#r=%LVlb5L zU(UWRumsC9ZBj3{xa6?+E|g3-h}*H&)kuia@|anUKI-iWmI?=}wGfs(nMFH^og+V= z)EVC)p?#9n%PnQ!vjf!ncOKz;7fxLuAxg_PW|`Q|=r~x8yfN%3EGco4z9c4D;o^~Z z*)9P3GWj#9ms`GV_8Pm}0RQMy5mMWd_7)D1sBN8kJQ@ZsOmYA~=OQ!+s zc zm3!ut-3^wyD@eWEvTUOvHaBjd7xtxz3kgwL`Z3GHp4&=5UwW1PK1Ns^&r2-M1?(8c z{l?0tV9~5h>gATc8t!Nbb(n)wo^NzXh|)5NS^BrQflmZorDX}T?3*yC67;3wy@0vGk{B=P%U}7dFLC-y z@jbpfACr2y`V9Wz2d-P^MZsZEt{C-&57^$9<0HSB3B4Y_yvijnm0Q~&JSK; z3w^n$bGWcxZgHC#j(vH1aVTz+Z*(UiO3Qv`d9f`l5&E*=ZpJ!c@k)?bifS@TrQ=WV zy9zfSl6tx2j^P-z>|A?yIrJs|83|EZ;+dsuPZxa0YPmhDw+KtV%;M6Tor_r;+Z#e( z-X9|Ma?8V?YzoKDNCV%smUN7SC@mSx;`8MazF%3(>_eci+`TC2ON2MO&OfXBetbL5 z7E4kux4iFTj(s`SvbrYp<=jjXqO^QumJ2=Z;IU=n_*UV<5}PQoWF2M}kNOAjDYfUx zq+V{(ThkdW53-*h1gATDDQv^8J+T+w<6F0%geWaNndNYoDsNyS=gu(Q3&P^}x8>J~93NwEoTC|0y!lR8 lKEB!3DhyEl4&%P;chsD{xHim^Ok74{{%n@y(5UUz{~u?7b)WzM diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/events.out.tfevents.1788167878.0ab01cb300d6.13.0.v2 b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/events.out.tfevents.1788167878.0ab01cb300d6.13.0.v2 new file mode 100644 index 0000000000000000000000000000000000000000..8b3fe7f28f055b4724658d0d7cbff09c13a7d375 GIT binary patch literal 516862 zcmbTl&#iV$T?luCvsm3fLn2u)IoLKGDtBq0e=Dsu>- z$ecu(+RxeRaop>(zH9As&Ts!~U!T|g?ppV}*8LqS_y7NUaLf3WikzJ9u?YtrnzU^= z+sk2|doQ!a&h9>g`b_Ms+R4Y+-P^-!j;qI#UXC7K&b=1-%yaGK{P(|m&syR&&&Sy- zWX@5Q9>3G7j^vrCXnRGT|2wUlrmDKCqN|6ux2dYKd^5RDiX;6W8;`E9s-mjo?dag@ z;N{)At(si#-?h8_{9db!TlKqETeY*Qii?+n(=2xnFSmaOEPdF07zB9Ao9!|99nc_N zW3tMrhra`KRJB#>ySRGHc5t2L?c?nE?})$uXNYKhrP*S`-x2$lt(v0Z-LxkE@9l3Q z->ti{r%u;v3Vna)%rs~{LuG2aKRI1hyQr%E z{~Y@7;bOr2J$@QE95MQxVQ=y2or<;ApA3D~UaHL(JGjnsa`2hwA-h3JrhRaO?)=s8 z)WF|iTa*TTQkmZUPgoDtZmNx`FnX(6^<%b5>$~bHlixvdhq4k?Rv7#V>Z{sEwKWyw zzQ}Es^HK*lPgmzZZI`neJ_`O+rTU#`5BMGUs{Hyil`(_=1a?!^QElks?Cm4-7gswy z0fC=%f{LK|BZvQqA>)U{(6nv+8QN=I75~A%x1_CB(hQYPqyI$oQthGIltge*#_dM} zO4qoui{9@jH@gm2Dz7H|iRz@Pp<37b?;Ysh!$&dR+7!x1k2X0D!|>YFKjr^_7<%~p zYbZPYYY_h3{(tBA&3uD7IqhR-nEl?2w=d_WsC=0DC&y5=ziPX`2eR8DR|g+wr&->M zX1mSv_WoDiET4I9;*E>EJ$(1D-=Pzn7dxm#&iNDCM^#_7B^~;&e|Lqr>dn5mjkR|j z-kSc7TlU%gu*xmBKXLt34OG?WI9CrBmwE0kkob2k{Pl(!M40sZop|rE-Ak1=i~b~v zj|9GWFiD(P{BY3kkY+pHJXh(t@=r)l)$XcIJiVNq-8}#P^UrehSmf^W?&YLem z7km9qx|}=ZyNc2W(+A|3;v{=ite5=&kBYHT15;pJ->?2t%|9-h+V()o-i6j$@b}37 z>raIE?~#LTQai4?7)Er))Sab7ks~k5NPAv@jMU8QZQ+3U$w?~bJ5)YfqhoEutd;$L z{~^au3H+#1klkR%0Jh>yh;gnYf@n+lMKzEU|KGHvuK zgnTotTm+DqAyS4I5{{4~3M&f%(zWT`?*tN5M-~!v?>s_UEeM(n5Vv+xIUjOf<1<3$ zd0kTk$bKy;Lk!75NO`-2D*#b^)vVsczfZbV9Ee)t8H8*yKXe`-13yXSe2Ck0hM`F~k8OKJ71V1c;rQy%B+wb0BSP4&X*;oesVOkiRsgaz3PD_({yjZl}`#AVs=T zh8VIOA$}))UjXFE!pY_YlBX=&k&628>hqQgQ0W++-m#F|G`J0p> zgvf;>L~lUGDu6V#*=0*0VXCr_pffvhBi4-f&IgFDom9?;n7c9=u|Aax0dmku$`C?) z_aS6u_m8sxGHH|JYyzp_KuX<$5i-;I;bnltgh=Ilhdd&sl(s-X&!SA#z6% z(lTLC06+$B?=z1;itEdEWK+y`+=xe>3GM(1j*!av5ZmtxCU7bny3ivIAYJ!L8A6Ef zS%jot)w2Ugd$S)N1d`c67NYTsxx6sCHLfQ>+{a1fe27YHzPBCXl#>vXDCMw&F%yh}84~h<=q+&WHS{AB7n!l@>1nNW~8+LkRJ` zgOGDCepUfwY50y+1QO6l7Gj?nfE!U#^0^iudt;<>KBR}02WE_#^L!RS2E8C@{R*p>9-y?B0fDp5g<3@?f#VWAq9{2W5(skonrx_p(14nA-?Ys zVpRUE8X)S;x^5wm!p5>4DO|&>a9-+)gUzLll}bw*sLisTlz2sU>9yA#&dl zGUkc51wd+}s)7k5wTUcbf#VW{R5kB=03a{oq;fvwqtXJ*aA?2a06;X8r3@j&xQ-bd z`@$Vp2LnXo-L(h;iD)VdY1?xdZiH6nY0m-jtx78ALsnKZ2aLsU-75eR`cujfLVQ&a z($U%wFB;2c2geY|PY&e5@TCZu75nuOK++dT<$TCPqq&&z;FL=^K-|2g3?U@5F+$W2 z4}S$6>6>nMkU&bB$#!J25i=~VG*tczAm*2(az5mP#W;i*WEETo$mi=)h7cmx3L%Yc zg9-p*s?;@^K(d?5Le|usj~lV6?~ASg(QhJ^^C5n#r(;HYx5@4R`Po{^5JHUGBV_#e zih2Nf_VL3B0!e5g3wb`o0XIT(s-8DM!fK>)KBQ_Pa~_#&ne_`G_HrNpY=RKt+X*4M z9W_k>Qb*}rI)MbXl!ZJV=!TGni?me%a<#Ek&WD&dFUE|R594kDWP58VLkP*#K}erd zf9V25dHec{1X96)?AXf;i|4y9#sTDOgjCLlSk*2>$dPsF9{^GuD`f~Fa(W0EbU$(e zKu#R9%ps6FtzLAaMtzaz5m+0`oZNWbi^mfY>BU8A3>=2|{u@G{pyW znq{Yp3BLI!N`dIKHl zYPJ3afs}F}u4QJp5${JO;0dwKG^v~qsckwBGr}DWrU4||Udj+cjK?A5w9EMo06Ao7 z`G!Dp+RAq1u*DeMh@0O&jDwEUOp?m^ki>(bm@(#rg*QMl?W7DL#McHP!`9b61Bj(s z>rVud)J_&sa(OszL}=|@Yk)k_l*;*#(H6|kzF|&!3_x6Tr3@h?b2>tDmCpqN#HLMI zHGu@RmxbiX4?{?K>C5^6`JpM5^C64tGv}g&l_#?SQm89s2qAI~2+`fVBNiaaibrb+ zq>2MkwlhJ9kI&*60MThEmGdFfn$N_H_v;TP0Hje%DMJV`c0tIbRzG?JM8DzEy4HWc z=qc(T+mYk7gK#6hXKY*#kom2oaz5n8$TgS|pmC-*Kr}l_8A6D!8$zyM%hdo#%WGy0 z2qdGUEX3D*C~m~imRYv|^5Ldb&WG5V>tcpm?|xCiHc?>{&&o*vO zAhGJQkkQMA;6@A#ZSM?_pV?A5A9Cm6Sj_O5Qid;jLh_^xAw+IDLdwTx;;n8_&4V@s z;;$hKNpIC3H=^^aKns8*$v^y4&W9}3vA~Qgo@eu*BLh{W3?W3r7a_ORn{I=SJTpD0 zK_KNENc*DBxDjXj*2Dwk_!z024{17*$yj;7qcK1>T1y#1h_N3+6kL4T0%T<5nyv(r zrzzW!M+XfNVx}__-&jrCCzbOdOByV~41=npKLN7gu#_Q$*l$3{>r38R09n{!bT0x) z=_CuWmhXleF)YIU8+0U3TPo*6+J~rPM&;&TNdQsRlQM)5-%SWn+_Od#Ad${(`VmN& zmMo;p+wQm#24CcT0TSmSmGdFdtC`#7n>!{n1jzm+Qic!`8;FpPu`yEslJmZFFoD!? zAl9x-M-tS3JO)VVSE-y2Y0^v${}_kkTbTglMXi(}gk)|<$m))Dl>pLtM6xM?6nB>G zNFy_5M>Q!N@c|%TETwWjq=P?m-|yLF;uwJZw2?A|kdhFD6s$7A%XnqmWflaI*+mv| zY?L$sylCe^T5Mm#LkVh{Z>;Y1LW{eYo6za%!_5eZ>$_|D=M{Ko5c@aoNH(AJ;qpc7! z_uj+=fb{4gmGdDnPRw~kwY3_aN2&~@3?U>l0U<^`PtOKO&mB#d5y(#t#HK+zg!Jy$ zZYw|{_ekY@$kytwI2CZh$-#4>Bg+p;8A3=&B0^ejuJ{5Ff8F9W1X9vnwj(8)%wKbJ zfj&4qoOnsAfpCJ8A6EMQG}>O| zJY55j?pabfAJTTuNX&Tn?DAcJe7+`S2qE@q2zfNsVkAJ+jFqAYq=ExE(@hySBH&nh z8-RQ7bZ=Qm#%`t~r~98U z21tCERL+M?iD`%skMN!K0FoOmWe6c9*$7elc6AhV#3otq6oL5mk%h!MF&*g_y%|?4 z_?(i;`H*&l8Hi{9uT=o)eqPEDLgaD~vckd87a)tjDW4;dQVt}uwh{g{r(Js<3mw_1 zpz^1j4~g!lgBf$$kM;yeb5$ur2+_zz$dFgjqhSGB)sx;g3FL^uKP|1x)9r*88hff= z4T04xeafXE!P>gXEweGi^Zl5HLw_$NA*&wGew0G~)8e`n7R@nj{J{a!V6D*1YANlX zR@bGkd5uZCJ2d(Ti=hl^rKJDU^17A2?>@j%vb%Nr{Q0|!*VlD;`+O+OB|0{`-B1`w zJ`JTDGLmE#*tOiq+#P?_wI4C$_g_8d^!|w|D_Tey|Kkcf^QTE(`oFVwd?}oz;@nQk zl3ilAu%R4|$A#l>421F6KufCshc$LXpDAIGLhO%yBY$s*SHCj{RF3FMLI3R{JDJT} za_;$S= zd|>BLI4)QlOOe9kg69qV^WEH8`~wzUt{f$${BH~Go(3Gc0>{dtHA=V1AtuK__Mq?H zk2!@pOkEiQBm7Z-RL(zER@gCb_hUMJ#TofigQN`M-^4f{|0WY!oWv&uS`Ts`5J*yA zS;(s>=Ik{gHX5f*yFHi6`H*}5MtIO~ykg`vXs+%XDMJXczlo6Pi`wJ!%?&StO9&*W zpDbioO9oPC`nCsj#QL37&W9YD#vF4lER8Y%NTVt#LkRJ`gOI0V^;KYKmF}DQia@G3 zkiz!Ng^^KI${T=ePLj&`kS`aQoAj^)2HyeFK2^#PLSpYDa9{aA2MbQbIrZ>ct{{XdhU=igpkaK2(k9xxB?(|hkdFdkc-IxOdq}}aLy8y&tf>h3j*y}O-$aY)x<^WkWP0A2Lj9(#Ss>N))ox8V9y$ONj4U+AM zxn?~a7OyKKj{_uer&P{|=+t2j7}fJ{Hv!0!7%4*tv46urY;G@yj(je!X+7`~ z#Ox!}w;f4|iWAlfga3?U@80wMPg=s5ty;XrU_0;%CZ65lY_ zTF#?JGy+JABT_jZGWXPEgvh6lNd-vWNhw1J$^49v-Cfon07!=IjP3+dY$V$eojS}u zVxl+gDL?|9q;fvQ{5`fO-vG#y!%~J2BKI93mqTB<0%Z8$Plf~%H%u0CrHFy_ zzi+k(Ai*Q0az4Z^ojG6(8}k5Hb~PF&We6b}zYuajHB${B{S(uM5lDcsEX2X(8+N4R zhQ2O9bW^2rK4d{}W~zhs#oPGo=IU7~Lky9_ui3PFH)J?KmUUh`l0e>ZAo>>72r1oY zI~5@5J*9F!oVAtXf}A%H4sy#cIo&8B` z1^hrm!XT-f57CHgjYqAMecs`E)bFNJh7e+02O+O}>>mM;#DZ2f1d?hhJG6TBW5)N; z6Z4CqBZrElaz4bc&;TKAwL0M8*Q!Ly5JCzS5t2CPiy1)1C%vCeAQ8i5Ay2kd;?Pns zU9t`!N++dqK4knT=0WY%F!doYw7Q;?GK3KOdI&jsrdkIe)dSNV2;?UR5+3p#A?w!N z2nNWM;Zivt(lMYXcErZr8XrEa86#x~Ar(pp`RZq^2auPq{pJ!#$q3nwTG1B7g9^z0-++S`||B#^*SvXE_h%+4OV@A?CP zXw8$#`4ES6=IS?5PkA^%&Ui{0LWq7Nge05F;gnF#LCJmuQo(_^ZvKuPd2()95I~j= zm&*B&^?}UW{Rah=cp0BNM#>OEG8-f0#6ugru{zRy`DOySV=mi~#ZJtbUFXq-H2_)M zS1RX23X6729TzG8%7XF^k`YgtVzrlxlBX+@NrNV zBdMGZ>3o;D)>1R-y%hG5vu09;5aQVaAv3euj0Z^D2@m%Wi0>F#h+B`3*pZEs^(vtw z)zhSMKBSd4bNFrO?^yv5li5;+5F*zKAt|0+w*#bL?TG^fQp$mhvt?fC4Ia_I6d+GG zOXYlskrH!>qB3C@&JvpuB4r36L9G$8TyZ+SOa8TSWfFnpSju+9ust)hqOBuT0P-YA zD(6E!uVJRlt-Cr4_mPPaQic$s(FP&gx_4?09Vx09b%H>W#>zsr+<%Ef%QY>eIY6?$ zN#%S1vpErZ1=rP0;%FaT9rLT$QS1|gP|h{ z2c>d8Wbi0OgiN&DYYH7%c0|e$LJB)zM-IAus|OwFzIXps0x23V+mR>hnVtR3{4{%j z+&d(d^C9o&GLM7um$bsC@!O6`8A6D?Izr@kY2b^V*U_%|1d?GT3mNi*xoAw%nu_m7 z-W`z2`4B%_=BTx8kfA+vBq>?S5JD<65Tc!Q0&lGHfAzadAh8oL~U9Q_hFX-qRe1R*LekN`RbdBxML8zMT-VJ80=T7+MR1>J$@*|3q1cS~q5BITgRG z3y@inQaK;;vn4ZcuhrM$?f`igD`f~FO*$h)am)KyfE+8ydrlzb9LU<+3`B46{PzHH zeJz#qA^TjJGy8xAP4UdG^GV7OLSnlhq>jrgeDyYGT=;7O$+MR2NUb+>hxmSGeK~+k zog$UpF;lNQMXk;SNYDc*LkP*#K}g(l>qLNrs}8Itkg!RzkXvn;^N6>%P8xJ1 zy}4A*hoq@6@7X3u2mNJBpAbo`7=_zjjh>t_}HUttdO%|d(;4u!ZX%2Dt zfNsKkshkfv8N+)>jbnRmu=T3i~5uTFlf6=*Z|utG)yhF+&!zI`$4i7B;iSIgR)C zN#%UV<%wS~qjvpW{P3v5VJSlhu{T7B+nKyHfE;w_I+#Fyav(pOmElIr4{w0Cx@+c3 z<$TD3Z06aozg9y{=!lhH?YvxZ9aJX`2e}RNGj(; zBIhyt$Yw8%FVKd+UG0}fLOhdGK3J%kq9YhP|yh=U7HSDN+7esKjr#?F<>`H(TW z%y$jEOrB^1WU!}{A%p~3AjIw2B13?bXFgd^AUV#mLn~h)4?D7}&WF?t zW%iL$A5S^xhG7mETJsMRR34GTI3MNHBC{)Rp{r0?C*s3u)wj7dw(abpI`Y)G3t8`H*K% zm=`^9y}sh?Am95^h7eLQ5h3ENSx8C*F#%iXz5*&GK3Ic8-z5_uEdMRH}Oehe7IE3hg>LP zZme1cYv9Y!v3sNpAw+)~LaI*=T@8zUZynCRMj$1w|FrmA&es%o`_^5Ltc1nq)~QmE zVDWih4KEDwa=bVT7N38~kV5{`;`6_*cf+*7`?kJ-#pn03rL=!qd~RQT6qADY{QLxq z&mZJTN&lzC=SJ7^+rr{=-#b#Gc=7p*tID`#CeNHlz`i}CSjr(INoMi+$F<7%V6jI2 z<$PFtp87({_#YRauhz?N2aC^b-$+@qi_iV?SK`WmzV6lbFdp?kN%jA*_i5?-fIPC2%K4DY;mi&8nKRoa0A%4bDMJWJaYo2FgZy*Q5x1?$ zbtnHl-=ff4wj=AOG0!*W#1FwA#26ngmGdD=7Ywl@FV?sm1jyOlQic#>?1B(o9j~zf zS-r)l0fD6Y$U-ukGf$k$41WbcN0x<2<$Or^t+ohJoc75NAn&833?Za&EM%*w77qdzqTNJMifLkO{-j}Vh*{cb`> zye(R{C6J#S$l?>XacA$}Cn^j&QZrF1=R;awU>*y4#QEUls*veYh7eM*03mKuf0P1b zy~$Gz0x4N6+mT8G=8=a*htIgyZt`fUoDWH?&m6U0$c2yEw|79;0%y&&QgXD(!>KH1GgWX z4;``nxVRUABrKJMluuzEok#lQ!~o=2ja1Hu%ne|cI?u18hcjS)$}Rb`2|`G$Cqllo zxq-j^*j?YKAAtlelZBj|!yL7qlup7Y&JPWxaz12YQ|4N0=EBGAp(7#2Qic$s?}d;U zbr<{{i5W*#h7d>v2l8B%`9R*|s*pnHh*6qU&WBvx)DI88ch1Z@1d!YdQic$c>5Y)w zu@_QdXcZ2>X-Xh>mdkd;Xv78FNA5Z6L<1zkQYz;|yt*+T!*W%P$M1h{w~;c05UWK9 zX{9u$3qXFTM_Uj`+6q~SQ|pTeQQxg@0gxR!QaK-zeVe(lYB0VH?(CoRqzoaXWHCZ= z?_{n4Nb`?#tq3G~r7UDXSO!8?7iQphTkFxBD0U|)w_$Il}Y(a$`C?4mm*|U zk#B#1`1dlHLLk1YWFh-EpFv2)vkZI(U^PH0=R?l5WtMT4IT_-#>5QRLh7cmR93f8* zraT6S)&k|31X9X@Y*S~hex0}J;2mOrAE}%V@mR)82_387F%fq5(<`J5AtY!8LK^n# zTnms;!@EudlCxU2Bfja(B}($XkeRTL3^kR?`H<6YBk`yeIzbJ8q9oly$`C>{Rw3l0 zLQyV2)GzE_Kp;tLWFc?nF#AZ9e*v@B?y6MIhukS;F5~O1?vK+HJPV`@AtYrrLQc6H z`U#IZ8;@}DB9I_od0B~XHFL2Q5&q~dJTX@tOV#r!70btBPYy1;yJf)duUEb*zdlW6 zxeXN}v}pLEB`9$EP_TrJ>AQ?rs(6-^!%SDEDtP9CCC-Mb=UZBiWv=CSTR6S}i|!06 zL})2zEFE)HcY`%7VuH?P1l!;7ltTef90W2~rEoH1A)ub@JN z7G;05ocfr)2`v7jw4;cnoM$oZ!dyqD##AbTCDw|n=UdimFh8R>qWqf%SYoD9Awo+b zV>z$0q&`?$kN&imSn}3$`ZD)9bK$b2=j1(LS!Yev^DPA?wz%&*_BXEu%S&4-L}-av zk9~RQ-Z&O4jk{e=AeNL39E;h)Ox$0t{AgnZ7CQr~o^LsmH3=;{=f0i{mLG$t5TV6> z16r<*m&Y--xMbT=VhP*GvD911OyC_R-|-$;7QLqG`Igp)EYY$d`s@_2wEIMb2rW7r z(b8m#3M^BHjxtZNE<9gr>T0rCB5q`v~+e|b`dP6E>IytO9f-e zSJ#RL%l@fa=ZK{^fYX=D2hZUcd!Eu*11#elsCvGo$@y_;xqY?A53t;xON9t6X#v=m zwH{aAf#v$ax0i?|a}&qXI+VGM+^)E51XvP}QuTbxR5#{_u$#_r(G4s+PE#R5i|;11 z9P0FQ23QI|oxV;iaho}o)lW}hU$%|kcN;9R->G`OrR@{u*U4X0J{bX)E_KqRjSyPQ zH=|`xmzrv@^vUtPNh|?dIF?pB&Z8yNJbW%#KHQ}0`4(G8CmdtlZ@KIT%iaf6h|tnx z3tCJDEOG$Ln^zV^#PW`3sq(vumR09%ag3RKqU!k;gJsOV|H#Cq9$22%P$5D~5o1}Q z^kFtwcAszYlvoM_IemFQI}0tAs^vGKFE&@HdcI}%TxaZyOKkFGuqYK!Awo-RAok_q zvMK6dX;)QTMl7jYIhMn^%r(re!Ly9PqWz1i=UcoMZ9q$uPS;yt*{1kd+6bX#?pCza zFWIOCmK>e9cf=C0jbkabU@Qy9R^I`OULUHSZ&~iW5-n}?M%sbp?LaC-Xwlz>7MGn{ z%D{58>zdER@{?!TmciVM{TMNB8CV|mrt0~Y9`e3u2`Y{`2bQFPREW@0!&vgNuj_-w z&u!FqVkz0q>5KLUX8uQT-Ho@vV!xQG=UXi2Y(mS!6RO+6vUD{SBD7>~$G+HBoyP02 ztB#H3r~LgcH9Lr7d9ynm_x+X~_sWCiq$^d=w}i`Cw_%i3$-~tb)<9e@!a> zc1!Z61C59!Foa`?aAl6krse9f(3j64R6XBfzGN+0vg1BZ154R1Dnw{e3qebV9c2~J z7rV_KEs3RqXPF_-9LLfXT6YCYmoTcHZ~4p44=n@ozxjgYY7`YBw3ING*AI_g0L!Vl zgWD6!ols6+77b&rlRu^V=7HtyLaLr``H;60E#HS7hyzQ=GAcx9NeIQh1V4AS2TR8v zidw{y7RIrB`^b#3_~lz8z@n^JSE}b*nl0Immi4vYkAmeyeJVs~@eD)Dqh~>KVCk-( zt4l1=J2;k;iOe(DyO=F=iW9sykHbQ6#+KGKp*gfkwSQ<5xHzSs$ z2#%#|r!+iQ?u)!R6f8~KQT2RF{S(X>`xLcmG+5SmqC$k0sS#*Neon@MEbqorAwr8rBwBV&8HV@!qfdlPB$g_k z<7Ie*=m?@GZ^nnBg`Ez@m}qa|N)&JD06WK$tROF3gXG5^pH=*tkjnbU}+Xcwn1 zrdiBPktxOKv-@O<;LBj0zE2Qg&fqx{uBJ43_;%wC#x{BZ^~5 zYRtSpUiNKn3|KDtQuTa`>*X70$u9|v1WR-P6(Y1OjY7*N|Jsvaxpno!9Ab%$=2%S2 zndg>wj{n5xmi2E^^?b{v26xbsIIJcfEJuo{5TV658Z8+r3U$EJYhbz?vH0)iS?V+Q zVtONA;B~T^DOJz6^fxU;OJPn|JSH!&phAQe<=tp;alDJ~D?^5@^C6aUo~6li=3Z=| zcOMTNW9C#n-!eX+1TFF*@6JPC46LXSp{0{-Wynmd*8Eqow)bf=OU;Qq+|;LTIs%LCdDdANUNm*1j@;Si)jCmc7@QYs=+p zmOKH={drV9-_j`g6Iyl~)x&dTuZ2{I(4rHImhr~LxzLxb4rjI#OAXJmEswb#8_}Vq zDp<_tQ}uj{{@x#GiQ0K?BJ}0+LMlXPsbDM`u@9BOvN~sDII$G(<@DuLAoHC6_Rcn5 zVA*t&s^?qO0)C;TuASj3uzV|`LWGvIz1WvPvy};8F^?O+n^-dUaV!d78B0)$ot9wf zpG?*BE$6)DOyH+ir@wus1(wQGDnw}U-G`QSE9NAFCI66G9I?dh=U7HOWacV#x)C!P zEaAtgdcMW&T357$H(H4A_bt+?5TV6L?3UllD;SL#_nUk-euLWGtkacDWc@Qfo^Hor_cMlA1m7NZW#JH8oRZOy?_ z^pUFPTilj3Ld&)jjV^%Yel-;$v=lKGvpubDg5}4MC27P`c!1NF)wh}VV%`R=i_v06 z)$=WPqS`T*-8tT1SuvIh5n5soU|*UAet!x}ts~ERXAw)tLHU1LN4|B@I2>(thAhNo z?cFX@NrHvs8D&E;rG3-ep2L1mi<{uQ*g@sN4Jze7tt9WVs1L?DH(#Cwt6>%HQE~sY zl)V0tJs9;({$op6@f-1kiuymTB~SGGuo>1&zbvOB#f!9GyrP5lL3jD9ulxy@;#h7=RxdFr;Gsw^ijgZiT&qqvvUp$JykFllreCj;5-*s-mlh zx3{UPvV1eSPKrJHFi%VtJ-OQl#=6I7s-A!J%vswCkDjsqts27LrR_v2MEG}^%luvZ z>g*T*mZO^cbBQG)o@04va|n02x^8RT!E(GCRnNEdsoMlCtMxBE0Ez8vbiG_BDB;nmW)?w z_n|Lqs)s)ymXZWcU)pqK-e9Tex?Tj!Lb(xAJ>Sy$mL^(!LhglvWv((6BD7>CU|$@! zU&TAXaW@;35KHzUj-~1x^F-d&*1iK+f*w-!e9O=|eb7?2^^6)=+eEe9N$wRLW^1=T5be7 zUI$C)^95DJQo*yh7cy^24=&f;2V<=LTdJOKX*P@*V+)+>m4an&B^4sHlrWZ@)2|D` z(zdO^4`R8K#OX`&TIT#RcUAjju$<7M>iL%P)dR6FADg9i1xw%VREW@$kc54S%JMS> z%kz+0`Kf=;_DDS z^ku})kjBK~dz51__Ge~C?YtB<82a*1F;uGOTXM9SegCoCJba36+kgrYTI7zRMg4fI zdtez`?b?c1N_m!Wcjk#nZRaIl!6GN`Ce`yT$?t|?UlbL$;|ii@%2bHZlEYXQt}ECG z7O#H&IuJ`v3a2ld^qExzKa#e-1IwtPR6XC~vuYGt?$wcVf-!cN+A+n&db z2Frq~pIXF{bc|y$SHFSRmiNNdafZa|TB@FJId#(#Eq3eFPJqR;-Y;n*gqEqt(DGs8 zs|{e;6n?!Mu>>9GSbol7-ug_eeYO<(av+AP=Udtx9EX;gzmDz&i*`H}BD826M~ltU zk;}nSy(pwNu~hLajnbH>Kg%9{#aW*1)=>3)OV7&jXxZ#yiK_^*Hc%l#OF3i7-L6pz z7PBVP42h-a1g9^xH<>qBb=D=@LSNk5Q1yI^zZJ8;d}{s-$JiK6Dnw{WIe~q#*wkJD zEF0%{8AdD_sT@nf7v>40*|kl0Om_07>iL#$XRNU=pQnGsw?4C0P$5Fg(p0o8i0)Yh zmSYJY%!noSB*)@s!dM0;n_h>$Bp##c`4;p3%)Xyue{w18FRjl|AwrAsNwj=@dOr>< z{S(hv5{v&SjwL&hnN?8yyz3^g>`A5S`IfrD%>HsDs|@dvn`KZTLW}Y#v^_3Q}uj{R?jgw#$N7wnFf7nvy}=FS_&D9pPkcp=!;9G z<#b}nJI(3Kl@H8yazKw9d$3sjpz8S+uS8q4lpf7|36}UeW2KD{S|UzkU-CUxdTC4+dozW6Zg$OP7X=u^k>VWqzOB;Q1 zA(pT+9Lw{y%=x8tbxUQisHjo(d`s2X(P-&)`zEd&-lk542rW8i&@$Fl{SAz<5gXIo ziKT{T*)@T=uUwm>-UuvBgQ$AGrPH(tXj!g15oh0?-ARQAEftLAW=YZ{u*~!HT|_L! z>72ea_`u96Fuogl1}t}%QT2R_OiHJiPK@QAq5lA|bbe2T2rcFrXqmry%VDte8T@<`u>_ps zSWX>2fqjYmQhy3qE{ve+`4(@rq1cznaYufFWr-yfBD6F)hnBhRm*exX*vrX5#PY8X zYrgDH- zA~HFawsOpUW#i-3kH8Xqi>l{aS`RQq%gL9Y--G4b11dyl(a%K7e%d;A-3Yj)#vEOUXq} zUrI`t`RlM1IEJ`YnZs?46$Tq zaV#Gkn3<6sw;8!YU+Sb%^?XY}xIS7&eEnJumIG(05TPX?3oX8XMMQ$7cY$#xu_RpL zSjx9B_xm^6J$(w6W<^vz-{RrX6D=KF0~Uj2@lz^9XtBD4mcyRQ?}Fv?OQoyC5}3`g zjGn+e8xPB{!TZYBAE|o2<@@P=XsIZ(#jiL`s-{AO7PV}&Z1!yV5-fSE?%W`j3ZCUw zG&9R{ZG3eJ^ySb7s-AC|UeF6IwHGhpYTKsUs1Tv0gt4TaUz!1yI@(cph~>^@PG2VJ zGsm${x6QJ_(p8VD=Ua9))kaG}z3My|V_W)DAwo;SW$a6g!}VaW#B0rdL@a4nIF^hW zW`EhZ`m;J%UiYBt`IbWq7>m-Fjdj72+?NUwT0F0yf!ycdgX)ayC1L|^4td}8A9 z+A_GqD_jY=yfszNw@e+*%)1y{wY3;5a_Urw&|-KME%7HjadpSpQ`P0f;+w;q3ZdTTiVQQ!px3~+Jfc8Nh(BW$zd!_htBT~mLb)VUx_8>8mBM2^H1X# z>o}tfR}U_f+a=ZWE&CdGz`nFuv?u^9O3GA-&=PbF`*QV7+)S{fZJ+&%Sdy-DET=9r z?~)vepfHOZOJJtH7e{P1W-)o1QmCOV(Ozf9Q+Pawc2omP;C-r|Zz(&{5iM0-_d9^)-4iNAXennb?R%6v!5ACc^juS7DazyY#Yg@i z_N7grk2hGf+fwy>i{9DBXc;`srxjTGbfQ9pmXti~OYPXEIJYOit)CjPWZd9b3S*Md zGAcfAEm+dDsCvF7yt4{gF6BIR0gFj@Dnw{mdIK#lO-pd!|I2ioIwDP%0u3tr%Las%VT2E>wA$mvUHPsWnDb%q}F zWwr%X&$p~x*#O6wVvn>vV7YHag$OMXg-l<@rQ&&i;4I%k#FBE8V;K?0yce6exK12c zN?oXWzU9LVCA6$L-3-UrY7Z(zXtBSEmVO-vK7}zhEYRG9Si){`EPa%icdxsjc;L$6 z_B*J0zQwUn5iRG_&frRwfM_a2XwkWamN`Se+y;yHz2@e`Qp2-6c+Nb#=UCPXdJN=-{$mX+)?Jem|w%@_|@k1 z^{INk#iv1C?90n#n)t=T^UbLcp(X7$_T@y{jeTGlaAof#V#&P2u}n5%&y`O)<5!$c z6j1ei%P;G{aP^>GW#Jg;i`jiDL}>B7gO<{-P4NA`?>+Y!#1eOxV_CPHx!<2_s!{=C ztZoTa&$o2D^&KtE`~1SMYsJ5!LWCCcyJ$%;I+6o@u~jm3B$j}C97}}~^Uh+fWf4Ay zY2JjY=Ua@oDB~ErpgI)aiCd{rAwo-&duWOKyb-_PmN7(rKC!&xS<;p;_m%F>tD~VW zC(NjNzGc!;d9-+RzFY!h?7&zmL})2uEUnjEvVp$D&&c;Amcsj-zKlM~ys!MeKyx5i zikzu>zGYw7TO4DnYBt;e%R4tJL}-btFzu#-=-$5KC$i$8w8Z)v06C{~B1l z$58cr%Zhzp(P9@K-v=yDCsH9o%iJQgcs13-H6v-mX00ZchzA_YkQ}BjNpA1ggQbxL zRnNCn-F}ajO!s^A%ct-tL_Umb&w)5TT`pv2;C=Vg_UEQI>olv6MXI^ku_h=6-+oz^(J4FQf9P zdcNghX*pV^R=sTpmd?Y)f5KHzWj^+J)W}Suk^JHJJC{L&A z`4&aH3bY*Fw`2}j2F|8JgqDCuXxa4o#|p5Nt=Sn#ED6ON%S|8Vy;$e6NIWJV3ZUxw zmgtv{(Ndm07We%@K~#v)VpWWmCq2TKf+cgvtXN_Re9W;lOg@17OYY<*E1)k+&8T|5 zC0Dr!Ek7-_S3_S^$5J6ei`rwfZ1+0y4J<>>>%OE` z!F>Is;&%RVu)N8qLWGtQ#-e}m10ItLeXEj*<<1jMUplU5j>(o!F5~><=uK2T-}0B; zBkW67*R>14GANh|5n2+SU|&??_V_?we)d0?N-SwlIhI*=j3q508h;hw!!W9zZ<(yh zd|@Z`VmEuRY#l{~2rZsZ(Q-ZQ!4z0&fACawI%_Kb_+6qonN*TsrG2NtT9{J2dt?__Y5(ymmGYlf+S@A(z&O_=qvf#D zer^F3_fISBXKdMlQ4@RBQ-GEBCikeQ|IyN{6vNJcp|1A3cMHmf@Me!_g@m z{w|pXREY5JVpxJh%j$do&0sM)RdAJ9e4lYFxd)EoE_b+-5k4)Qc7dwrTa1S>-zHqN zvEFO2C|{*QgciAHXn8(z6~0}#u67`wSW0=8y*HSrKPmTLdq7`$cB1O}mQ=5N>`U$O zR`_Lz6S`E0(2~Pg7HN9ojMZk}-0l)f&T~#*T6AV+6FxX{9A`#0P;i&(`IfRpH__to zJ|DjwwMmr<5n6(tV_#M^PSJodHsGRRF|j1Q;8<2{VZIPk($nJ>^yTt(s-AC|WOE5E zHnH9Cw{C22Q6WOh)E8)}-Iab6EXC{OUJy%ADaX>}ZUXKvn;zc#0hUgIR6XC4bto4t z3)_VGRuz2;p@|swxc$R>H`_ZyEdOH4=+O|PdJ>Qc1>O5Lz z9a)X5I^UR3Awo+zW9jwY4Bsx4IR;k{OHmo8FX0(XU$mP(8Vh~Na;56|mc5x-Xz^a# zIvDyg#)k?KT2jieFa4&!QOCa2POT=EjF%isTmo}UcKVfJ29}pf)1-R7X@g*5(>FyM? z4=knnREW@G{0c3OSFb+?%ksYO6{r6_+rz({W2yI^d4p9oeya-frI!&^&$kTFzJ`{1 zeOG3J#m9^a5n7bX(c(0w)(mU5m&`z>Z;rIL&XSLBFmaw-R%cJScSDYG**@j;W3dyDF`Ia?jQ_!N= zya-oxkGxHV2rW8q(el7D09OuchCJ6MmKvTVxr}-8duzwE4=~2YMN##9OR*wze%aFB zY6JA;{5~o~XsKWrx>?OWHf^OX#tK?ZJ}nx418{WWMKE>_#$UEcQu@{m_>efmA);GROD~T23#& zV-9^8A4Y`;Exzy3@+!0he@Ag(^w7b?68C{)nRkh~hB>EIrwv&24p8-c%dnbLXnAlt zc_&z+lBp1(#ry+W5~IK3%#}lrl}(8y;3LQK;3+c$&wksbC1C0QimK;ZrrbD-mQT;) z7J;S5M=C^UY4Q;*^Y4}F!We6Ly}+DU-tjDM*32v%&y?uBV5z)K)$=Xil9@5KZ^(H3 z{p2-|s1Tv0h_TFU)eEna_fL%&Pb`HMoW2-VFcu}Vzm9?>*nz6&TVfV6`~G&9p0Usu z<9Sqw&=OmLec7Jb)EN3Qql43AVoCkPu?(HXocANkuZ{xCNNuW~Z>c+<*iHIvg$L2n+&vjrsK!sH zLWGtY#&T`gluWP;Ka@Y8SV}%~`tr3mGtQb2c?d0&Pwz+qOXhMa zL};=4f)?)+erR#(*li851b*dM4$ffik+)C!jBn$DyHfRhi_+!8Xh~_18VHsHdQ^ze zqV^RnZ_9hmg}ywk`Lv!`DtMMhxy*4)(epa~zGaGhrBu(itV`dI7A4gk_|>z~DpZKj zQo>l0^PHc7<*QbDAhF!3=JaLDJZ5I3!-jV!V1Fs@LDlmu=JlETm(W(#h0vD~eW?(k zC7~MoGO0^CuC)KUYC|Zoq}6aNnVpz7zORq$#8ql5Hc<6^OW0xNH@hZ2jmA|3y|z&y zLW^e&TAF_AI2HO*<2G&=u|$94SbUo@$FY;!`{RA(jm13*=B`>P_JSW0=8DdU-Avb*D)HDIw-{v_4&EwN9S-?(dbx-EX`_()?a zL}klH5DSX z1pUCi+%M0;S!+I>{Z0}~(oc>>V?8IbbI1T1lVsSu%M z>QBaEWP#_Gr2bNs)teae9Jv2=KN9-@B+`3WuvGNp+(~t zT3VVm#@|uYpWXTru~hLaVeB=`>zK@uU@=}m)$=Xdd&6;z1)o2FGl>REYn=6dcNhD0n?W*y+-1@*D*d+h|rQ!i+%C! zJKGkHV=a~^6%tFvUmVLkW#&C{>h$pAU^(PR)$=V+KkdT4v~M3%4f{){EmVlmvh**s zES>P+99Ue9eC`uVtgnLX*MdHbWM<&m?(eV_EXPMu^?b{dXN<)^<54MC=8vO7gcc3C zQSehnPkfv%f@RBuAy0_Ke=Wxn8pF&a*0Dc}$7D@gs-AE0*%gg_86LOK3M{p=sSu&X zpRtS>k%+Sill!TZ5lcDGGQThL-14L4*@MuR>Y-FU-?HW=^Sh%j-sa*A?*${N5TT`< zv25QE-W>W8o_gyovE;4e^rdef=2_nsuif{-;(d;)=UdjfFxN1*_dUa}Y&5(~g$ON{ z@;Js;nXJU;Fvp%HR1!;yAICD|a2(!O`u1_04HnP0R6XCaU3UkLv14XkZD5Qke5OK# zmK4TPu<09qmB#hNqHn|!=FhPVUCUgLO{wqf0+#6+R6XC4emx#7 zP{6(%vid3{$xdmdSgldcI|03e%SkM~Z#GQgeU`5n6sSmZr{iab=h7 zyS^icr6iEkm%w8?ag0@@zkd#v+%Hr;-_l8U1NLS3hOhXGQ{8@1Awr9NJse|Onq}hr ziHJ!M@(O;4hO|UU(#k!Awo+wV{w_Pd<(|d z*AF=ph$Ueg$C47uJm)vKI1_(CC$O5T=UYPdF~{TqCk?-XrPp6nh|r>=gne-_u*2U@ zcWxa#l~@9|b1cPX%=0mW^;__r_(Bb;o^P4pwhhPF@~~3;z47l|sSu$hkgiL$5@BPuzS$jb~So+&iAwo+jV|ldU?FX>* zUZxX3EIDDEzO0+b?E6#NIQ0XI);g-5Z&7T&2`$S)la0V~auXFIw3w^n7?ba=JQOTb z@;_}OmZTjV%d+;d*p~{gQTR^W@IF<~w>(7u>)y_$d`4v?MW>bh&54z!I;J zv4dEG!Z{Y}b1`U<^KdGIzMQR3)$=Xc;mmO?*r~7#EOVMuAwo-&`q-DE9!;{ra@=xL zG_h3iEH73u*U2Lq-otCllu1-Q-=gQv>@RBq9G{V^&;XenYWzH1aZgQd1no5RGC z5y7#v?7>`H1~=KL1(v~TR6XDFttE5bKen(2=kwoFr$U65Aq{Yhec7|W7#0D)o*i_8 zSgQYJ`G?i%b@is>o>R8Jp&hJF-?@-V60A=5{;Yy26Ptz9gVpKr%czw9v^sr(%0P_E zzxiqdtWH0*mWunQ)#+`<1Ywj~|A?WmI(@_@D(e5VI(>a{#w1vszB!nR6t7OtYtkCG zt=`d+lQ5!oMN&azM9Hj9H=Ot$LsAYmGKJOY>9JJE|F}9mA?o`cSe@=3Pld^@PXFst zdwek%Zup`C2IkvjD&QYhr>l0%X^pUAl(to=;U1kG6pA9|>tJ4)SsM!C! zI=%brNw_-Q-F>MFtWH0dMFsv>tJBT9MdRx9kMkp5z|s5jH7Zkhlvu(XB{sY%`U0!d zGYhGh|7~^ph8yefJHE9a7M>zUoY+Xt(evCx<|#w!(Ck4l)(=!u_57pfxTDOOU~aK0al5_-fxxVo8bSSd!i_SLf%{gTKK3;!{D@^DUdzn5**& z-2wH%(zS*P5nA*c;TSWFd|D62m~w~tMZ^-en`1fLj=2h)(sq_2Sl)D@>iHIr?aY&` zy@fk*1yQ43REW?L##qjLxZ4pdUBV2X5=#xw;=Y}^3anYV`!-k>?567Z7KcV_ag4QH zvIoBv{V0wK5n5^(%SXk^La;bS%Dp6(;ys+cT#0Ai_*T9>84i}c`>1-pWs>cBv|L_y zLIK9uy2Dh6&@!VjjmAA9`|?%ec9ZJs^?qQ6fjR2syiRg zgT7dFqe6t1&P}i{eP5n30Za52hwsD^u$N4N>PN6M1vxP5JHlK z5JD1?%prvET-Sb%<7d6swf1%2_t*bx9lfl*&tac??T;g~EJ(`v3M>Ib>gBZfd$-3M z$#QeX4!rNXwt6ra>axp%#LJfFrP;mm^hD>|VCgu4B+**lFw3ePnS;RM^!!UcWr_C{ z>axj(-7B3VKH%z4{5ujaTOL$m7I*tHCBb6!l_b$x1{TLL<~rlGKUjv9I9o(n!uAT5 z4p-Q2!i-Pm8=x*5Dw25Fl2w5{k8O0>ieKVgjYtx$C5&07J5O5lzl>7dZw}G z* z=e4^YgXQf<5-(c@<#9xdL!O~~!Se7sNusp`FiZHT-Wy@=uX?twwX!$|2$m2l_ME&u z<{SRL<-#{4Ubajf!=CvKa@M*8ma3T~iPmCR3hVMS`0^02n12}1Nm()^%e2Gnd8}if z20NiH9ln!z*>bA``+{}n*^fD33CKC{&mW<+*fPuDwAr=5veTzh4`oRR6zcL|E9>l< z7GsQG7GDICc-ivPjom9(UVVn|VuucsBw9-Xvv}Tqho@{jG<(rkS;B(^%h9K7C-JQg zhdM)DRt+QZvgKBD_NC<2gU`5|Fv^l7(OTM+#xZs?f1g}Xmt4gU4OSNK1A=AY3Rai6 zy@hLlr9cG|FIzf3V$b|0t$S~TF?OdqNusrQGs~DsQ(VDP?8vfF%3>cZSejR5UwvGn z8z+P1?I{v3TSgCN_m^UY2jEvy?+YY}){>(P*5$loI(9=HPB9s)EUA*kb>$X(PPVVS zWF=VAUXpm(@;Ph?UX$Yso0W#T9Q;6%Xe|?%<%&xXuEX@$Q)#lY#DoZSi7UmPlP|96 zv;!=ygGjt=>EO-Y_pkTfdIT(`j*ui;OANC#G|IOg##p}+=`)li_@H31IK}FcD|ccc zu$WaO@v^1Wi1}C-W1mK78EZt6Xe|xP;uxD)?`9ZSdgTh8t1NDZ1dDYs{+!&Q^%}4& zEkxpFOTa63-``gt!T{>BrzAn1H^;Sy!Zlq`l7*)xpCelwgeyN;81 z*)rFiJtxmi+3yAxld~j=){@CAyUGv3br|C&rmK`CF;u8a@y2Y7jad@-9qKZ>4T+a6 z)^_aK@@}*P{_g14&LoM}GO`?wv5#iI7Qq;MnNV!KvP2vfEcFVo_ZRasskqLc^@zmF zmU9)@{pCTyU>svppOYk7O9ZocnvJ;)mLtXQZB!PYBZ4JQ4|dvV{5h>*2T_qB(C+n58f57ELoDppxS1vORILzPQe&^a*V{wmRXTgalSOH zb!jnJhMgfvw3da;GG_UT-%ywFCezO;OVSCUF4I=9XUl5FCC`FoPCpVaTONBXLW{{g z)9zsTG?XOKT9TM0WMsf|u$1l7JX%>I!v#xL7jLYKf9eGM##dnoiI**JudwIjoQ*c) zx1?lql0<9iQVGXcwHZylz*4c-_c&$oJ1JQ5d}sF;vn^|0Kwah!B=NFkLkzRjpS?T- z>e6^5Nuss*G0TFslRkpwfM@h=WpO+uSUw+O&tq#wKEsn-+UzFrvgJa_IXK3`%Qd_P zmYn`1iPn<8GS=nyo%8sdT*KA%fwH`jEFqTcThjd)pS4hzZd*ybYzgyW^CcnN6W?DN z?j}jJ78_=n`>}Kq)TO}r;ZKw${`wZt<^WXmDRU@3LIM!K?uMF^I%R_y-b+kJe0uvpzA@v>!Q4))EW^Kx4}Wuwev zl0<81Q3c1?gISkx9p>fjm+zFt>x^LWtis0F36Gk%J{Frq;$_RgDeT#D$iu~1P?ykD zl0<9qVwT8jV|PJa+EqXMSy{3r%P>=BY2bMTy9n&+lX%&(#*U3K>)itif@M%Ml0<9C zW)`y+dvOi+tj+ptWl1?J)Wxb!DBfQt)k?u%(U?7t#LJe!=h!v5c|?oKU@5nZB+**N zR>d(^@%fG_P?ulh2j{l@`#W;fIl&SsWES0VBZTv!BGvmXp2MeDT@8cm>pD(Ls_# zYpGie>vB5cy(w5m)P7M+S)9)cmP+~9vt@@h%O-%OO%oC?TY?|4XUp~-7hC{K`PL+f z*5b@8mtTyS3zqfA4wY4w49PP5E3-uTy~B0OVP{FaZ1E0ZmY}2ocfgV(nk3O$GMFXf z$z@!F{b9MVin1hJ5bAQl+Y9IZst?z3O(|s(iI*)|vzOs~asO2qPX<3clO)kvhE~Th z7L^@c2<+H7XsP`(v+;P1hvt|alY<n(%C!mkQEPB!+dEskHpKC2kuL;E}4s);&m+FGLl4VsZ;~& zvhv{EvtZfU{$4X>vA-l(j-O!H$&$Q`SsqpUQ4K8XcaV75vLTXLUY<Da$>mtM*ed|9x(Uc%P-??+r70Cll9CGoPQ)2HQVsg)U%02b4MB#G7%%q%~9J1M92>HG zw6Y{#5$a-Cfz5rNGoA62z*iO|UbcL5pN^Jor);n{!E*vhqO~M4%QL5Mc3`<`I(?k7 zM8pb~2R7`K4Tpv=@ie-VgGjt=ac;jDEfc)j;Ciy_D3U~L=}-&Dn74gDJc%-~ck`*r z;&WB7Bqy@9KD$pgU0{rz8BF43%fyQ8UOA?4=R+`G{LM)ct;L5~?#FG%HQ49&zh)_m zL!4kKSA?BNRU+TP&0xvtMB-)3=CL!eE{8q@;V*Emawkc&7K7SYm%Dd|4ummws@T=} z%91Ww>K|n9$P4cO>JFC16-c~niQQ|1mf{ax+Cp6pRVPWbmMP5g%h6~sSdw;mEK`=a zYeHRuzOr|*URQVH_m`9YBwn_>pE3m24(TM zAy_&cVtaemc1>&mmiP%IUbdJ#V)x3gzYR8l#m9yu(ONv1<=Vxc*l8bIF56vMvLs8f zX>7iXDwe(tEZ3@&c-eBnnyvNydh!x~tzmjSl0<9CVwM-Hm(z%a(h6+1mK@5yqFHF2_HUBwC9_T^wWIhgO;gbHDxux82GTc~h{& zEoARv?<#D=_m{mMBwn@*?8oZT>7-RVumpLNBw9-(v-B@DV=Y)(>>jyaS^RDZmZwix zT}~ZH!&5aa14z7VF)*Hm*RhG2QMe8hd59#@T58tAx>&ZE`wSYISKEXfP?o^IEdSDJ z?_X^!u40@zw!0p5+FzPSGHE*Pch!4?t-)OzuJwja`}NC6mjBdgpT4Rc{!!n{f5-uy z_VqWAbpO_Af31ourutmr<|^p4|FMar`agBr&-neOJ9OH6ZYL@Ao%ZIR%j4gcr;(`> z%=DbTB#|0X|LC;u{PqA&r?|HmKru>9|n&}qN!5=r}?JMAso zO~X!mbDveSq0@fRHIneZ>aQdq|TlqcNx?dDnJXex<*;3^qv)qpDQ3vXhzddcs@6QM4F(@4B*S@ns%X`FW}if2c)pF@&p zE#(?uT{@N+TMEY5ntZJilqFlTSS7Qk^T!ol`h#U#F%mCZg3hxQlPN>;;EMd>awLh? zvXoi8<5u7oU*8o!A1F)8J)thQY}xA1rk9&bfaRwniI**{i_ONTz~~1Kc;>x@8%d(I zq%cdnG1KlqU2=_&eWEN;iGt;Z1$&!)wPorxumnyg@vNE?r??;&1c~J_eSrZX}7;V%P}l z(myb%EYxLC@w#7>B}1}AHDxQmV;#TU1WWx2Bwn`I*J5Mr&gQ|>!P2eNxm%I96SvLrkd>N3oNoflgs+Z}h|ybdJsvSnK|8)I|J)^q{Oqr)VL){?+1IihM{ zS7$!2u-sOE@Ae2!5-eZFZp8b`(S<>HHer!HBwn_Ruwq~0U#@E90hUz(B#GA2rZJAO zcD-(`f-&|!+P^>eco#+fJk;+e#`<4L@1`8=LolLzmsH5@FS_ec`0#hY2` z9B8;3EYp7UF0L&0$%1A7Pj-L#VrsPl=1X^jK7Zn6%l@eGcz-GOeR3JF1QjAlw3ZxA zurBQf9JB??`a&hkDNCwk=@G`R$yUqD1cPN*DH1PRsx)D%*se`W{QyhtN+gNaGJ#pb zW)wLBma{%@swhj$W1%jl>)3Ph{V=i`x^y(lf*fuVb|;Ez1R#WOouTTTCjkx@>>Frvg|GZ6isv7B^-Ywzz#8 zsLQ$ceVQvvrerC+z!@!;L0|fU#qt}8*I4-OsA)4^<8MOV&C&PIAEC8mGRwN-*0_q@ zD6~X7Wl4N0)aBg`w(=YGd5kMqj8Bqy*^+L>zAl8`J(mya;uA@dXe}d~;ussTr7Qk6 z)Y>YKx+qJ;Gr^LU&Q7N2yK4tNTQ;pr;$_QBGxjcK`S4m^u$*c_l4vav%wl`m(;dc` zfAs^sl*K1iu#CvdzW7!d7m2?M7gV3b%a)7)whQN{?_)gqs7f=EL~E(m4C`VPesw%p z9!{UzUs)WU3l^VN>^hcav;ybLfO#Zdw$z`^zAW}C(ESnAC2lE6qO~|M%lhLR@pnkB zUuZu}S<)rT#kQ;?qI-v2_?)~bj>OBBbHms(jOpYIe1G|Vn1Hm3$G2l= z*V(nRZIvZUvRJ=l&-{C)ULFFLYa2i@|U<#@5Wf-xMsbI+7$>OP7{7#zt9Zod=8C z>39ca@p~;;JkGLrv9i{kePE1Tk0SB1C5=DxU;6EW>-bwe#|*Cx z%HsG&uq=AQI=YMRKaKDEkH3?6*)lA229B|Bt}F4>fC;&`{rMxbmi(=-E{iSuV8>mp zmt!_5%Nxm3BaGemD}On75b6^2k;KcE<0)+J*M0QW3&z;7?<9%VV#6#!i<_T@G3K0E zf19$zzZL4T`3$p6S`ym`EM`ARylmNYiCKhW>=22UEj=o-@0E+be=Y)+gySTM*5bu1qjKjHKHNJUWu$+%3Nwk)1X6ZgR4!`lu^0o+5mXr*k zF4y+5cd{)Ca-x zcs2VKh&$VQECkDsgCt(IRI19}E02CYiSI9ikCP-?OBAzAzTIm$jIlel-kn#LfRBP@ z`#RRyRm8UdK95~(OyXtBzANngW$EPvyuVCrMUrSOb=zWHPBqBY5-jHfj>jmA^C!X5 zu?Sn|AHI46J|}-YOyXrrkU85G{bKBz@i4}cPLd>Ai!-xK_Ip1REG-(WxS=c=k|lZ! z+ZAnluxJMuW3eqryli>n$#z9=SQ_#PEIT@oBw9-bvkZy+_!BJ0_M6>RmV`{9F0Y!i z9ji^=?ZnPu|3f5Rwz%=rPA7NEzZNV_Pmm;9%g}Z>#u`SKEe@6|&r3d1mhjJl#bz@b zV?CzkYYcU%`JTkfmi`mib!_~8H+Bwn^O+0AyW-Ye@jAL{ZjiX_omDz(SDMEmwx26gGU#!?lY#ek5MDWNu+=N~?`S@VzpBFiE1d*fUF1^Zj@V?V4>4ACx6kvOL+#*80}J zd5UvC#h1j(mW~tHTHlKJp=+Qn1A|BsttFLN9^IaX-^FI0>zk!4FyLvZiPjR#EH4*3;o8Og3Xk%P`TIMv+jqe-{1#h> z>27{!Ce$U(h{Vg5qpjKgvFK?tPeNUs8;~SgORf0q^9h9<<1`!8-V3oKaxainaeEgCJe@P%fO_WC6y)dhftSh-Prqn z@4e@-gJ|A!5-(e34Q1~yuAT3e0!!PsB#G9N$Sm$hQ}aPxD%EXOL0KZQ1de3s+%|k8NwgM&&RCaOgU(HWG3LI*v$3+IOO_qG+1}Bk z-Dgh%%VR$hFI%>MW9u*nBOc=YWpgk|qP0w6me;lNJOIn0TvjcWCGNLSmv%*5@H_Ih z*-a~hrKll^mo0bZu`#wPAP_rM9+o6Yw3aw#*|m0?BUn1mY|}wmLhWtr0_*tQ?gh<;$=(umu&9G%zT4gYAsigBwCAc7aU_rr7W9* zCHqH??#kk^Qn2`xWNS)Q*YCwsHk=G{{)v|@H=S5rR<||z0CT^oAxWaOcrZ)h&3#6L zrQN33-pZ0CS!$1C^QGC~Ewx~bjoCrsWy}1w?7i~!z=L?-f9Ok+Xf0XHvSDe|W3V`N z-a1fOl2!?Ix!}g0VfMW4V|x{5{%p|>EDa4wyllz6oZa_hgG19`jHQ<(Nwk(X%(CQdk56DZc-&!*vc#_y z>e9iO?dw~WP&*JTCXpmwwj3YG};fAGJOXvQkJlF zg2l{+J@YTW)$=M?o(&=KvZZ<+c1=DrweSY8d>l=ZXf0vPGTqz)Pdv=iyq3MPcsUA| zQeD|I|EH;+Gr>~21BsU{wf3`p7jM4KI}etn-ANLyrJM=YW#XzbcpWP<;LTcP$(Ae* zr`R_OgFM&Lz|w3aiI*+++q1r;DIxXo`^)w*B#G9tlv!SX4sd}n_F?iV7iCFVFVrP0 zmOb--HVl0SmKn22ylg2^m^~+NHM>{?>au1LNusr+Fw2WppR&NRe3re3vP3xvmSvsT zb8@Zgb8&qv-y#w(TPmhn<9)yW{Nwm*b9VM5iPmD$1IJi5zu^zTQYpUo4rK}0AXqYc zv$^lktX?wIW!!xdFIz^BXX`L8OI%(Fbvc$ol4vaf%#u3P`5QC=pId%nud*clW%-xx z^dTA6_$KqS%ihh{kPre|M&xow2iid?(}ZgNZSA0ogPwu zId-St>;L2;bf*X1CJFzm?sT6znK+$lFLSknt3<{Fl2dz?NMctB>t~JbLU;P&CnU}P z)}20f@}S&s^{i`M-d|mDoSlWM=ceZDOZ>5+1@FOFPnk;M<*R4zcy?FYakxPU{J$)o zO_FH;UkrQVouI&mX%oRxWKD92vSdh>gw5E)@oo zBw9-Xvm6*T^eb5QR-S!YS;AcfOL}{@3ukce9(excnA#*>wgly6UrB#mHNOj%yp2f` zt))#b9AnEaqm8fmo1I*yW*46r(%}pz_RfP zNusrQGfR%IVfZ`n4>#nvqAd1qg2ikWd((K|KX)$}WBwi_UbYOn%f2j@?A9g|EW^D? z60IeNDb^+E@m|~=)otnh8_JR@S(;n0`BHJo@MT~LSwrGwOV?m_?sZDXTgAXq-jyWL zS|%`yTYx2Y4hK%#b5~hn+=aS04qzwu{C<{j2`piDBwn`6IKWnZe=m57UrC=YAxX5B z7-lif)8HA*m$uudJW`h6O@bw?Ci}9uJat$Ou*B{q@v_Bk4_h%AZ`TID`m8!Yl4var zd*c`zwl?PguvmR)k*X|in+3~gGq!VOaP4Y5XYqYV)CXlr^bqROA(Pc*+U4W8!@JN45-(d?hqLS0Z=*-pp}OZRNusrk z?1N*hd?CvW7-QAqH)bhIgr{IhdCRWJ&ujI?ld;O)C-Jhy`#$SVuR1Tn80vB?g(T5h zBACT^5*>JX?|fICNn+%$MIaNW5$@D#|PcKN)|Ay7<;7 zNwgLRW?46PLS-0ZjpE-LDoeU#nN*%x{O1nGvpjz$l6cv&s}(z$;_IR}xMFf9g(T5h z(wU{FcL|*PXIe#?D^yp=tf0E)ktcylhE~WM@UsD7XoCNVMohl4vbv zW;n)N57xrn9&vRxR#2AEZGz=XW!BN%Z2FsVVDU00@v>z?7~6r@$|mVOjImyAND{3j zlv$4Z%;*4P?0u|xb!G9`E?7obvabu(niyw+rC|bzmn~@q?CV00$R_Jwe;xA#>$ekL#T_tnC&&zuzu`#LJez0QNO*M$kt*`N%AaB+*)um}Of) z7rc(e-9FY*St7j!i}MQhzF)U|Rh%#1l1RL4$-K_iF1EKbtqNmoUMfkVwRGu^V=RA{ zJr!ZTRI^;(L0SBE3YJAH*nDYzVd-xeV{Impc-b<&09(aA^SL5+5OtnOl4vb{%wlF# z{v22g=l1HZERMSb%l1a>Xeo)8Ek(Ytx!*sx(4g<8D@`_Q2^9s{%;a5TYkH+?hfk;6WfEOWxi5>{s^tbhFRuM zZ-Tp`4Msj3s4Vfjg}VG2!oK=6T>Z8wSPEH_c-b;(4YN3mZe;_O1ye{8ttFmWvik%_ zz!)oAZ|?|Y3G)#wzLnU{mGw^fGhyn788V;A<4c-b=l%X)m~e^_4nMp;sPg}O|w%)Va7TaGOS7T?<>Ubggj$}G)`EbIi9A`eLtt!3;W9Ah?ZqYT0F zF7N$0$`Z9#uw44cEOzGt=Yge2F%mCZYL;SStk{{Q)4?*QEJ>oZL@`Udz}Vkl>2!Ga zB4r8i6D)TZvH4;(;J6o9tWJ@5*^+w&+f8_Q_ROMSxp$r<(OT*b#=1Cu$&Wi%n$MVQ zuPn~{1k2LuY(2R^{2uJIFF&8e%a*_8j5v!&~-}?ShhM_`znil zpkQfvn!StJrljLe;=P_EUbZ-8u(AuBwC9-vzUJE^$jc)TGS0xmQ=~o ztS0q%zBbYGp5iC4a|Hhm<8INT|!= z@Rj(?U+&k|JbH=~sAv>3EXF%a#ycm-CL2lBx-;wPz`bQ~CrerZcz}Ci} z%yz|Rm`UGAylknyf$ddDt~$0M%ooF)1OEIGTFYE!dH2pa9_CBjl2TWdB{4*(OT*l3 zj8)zlj=$4;&w<3tma6U9+_#T$+X0q|&LoM}lE^HZ>uq@lmbUMbZYoQ}LBVqO8oMU1 zNNIH!=DyKa5-(eFB(t^gi0zkOfMxkFl0<9iFapQe{)AC8!15_)K%%nv91<)IC$Y8h zd|n?Of~9LA5-(eN9(TZdWl_)dBfxUC1WBT`_%O?kh`;u5<@o>qd47%`Q(#Ek%*V>& z5Gq*g53v1$jZZZF1eP9mNW5$r;KkPZ{6~aGgXOQ*(m#KM)?zRc>(XfJt*KxsQlfR5 zvZPBEn`>;oj6YkfAy`^hB=NGPkPVwJ?`yuF1D2jfB#G8Cg;{nbufp|YzgflJDofm9 zp)MPfnWfe43){gmZy|}7E$@!7G1k0);j3WzvVtVhTH=@`VeQe`P?z295*8TXl4vc)qi~Eh>t>1FgQdFqd{-8aqk_fT zn(ZMeuxNIBuypN1;$_Q6Q#Qu@LJS8(T{;gYNwgLZW=W~u^Cr|K*AAPU)_;F{&5|s4 zeAp=)P4Csm@0ExBNW5&BvW@K_>FRbE*ONB~lO$S87PGX>(GY*zcH88Z`IRLpOsLBj z3-$~%a)l>aPR=Ir8cQv9`bk{SEIj#W$|90PYq2oLG4^fGiC0jUE?={YC`;rq!IHNs zTk8w%-T}KhU;ifYvgO2B_FXJ@a-SF&WAF0){PRa>Es@O9yMGV-Ub$jTOlf8DJ1$rp zBiJ{-hY=q+vSnizwhrUeyHEzyB`laE(OTXxi%EA!Kd4KYzP0NsOMJLcm+vdt z7%LL_8c#pjI+VoAmQKys`q;&iLr;U{z6D95wG6btF?RDr0Pa&t4$U@Jmavn8Oaa-AaavgOgswfJmV!6x!2SUk>?Bw9-tvux?Oc?#6!(XlkzE|dqd^cEG0?r7QP**lztTsJA z2bOtPNW5%u?#Sj#zRcozz*6ZZNusp`Fw2|oH*rmA@X2GNl*Re1VA*WW?k_XTM^*q! z<^>WjTdE9VztA>wSx-EfA~=>L(OL}0U|nj?OT&{~-kK~Qt1KCk#cB@QcX8PH_y@3@ z?MLEe%cT?SzCXLU3GTUSJd7mKT5Or+`|dTVFkkX7=sj6k63z*A$$HAZS%lTuhVL&s zSCDww^1d*eFP2NZj=~rlwvHsxS`wJ$%7@YQ!E*m*iJ8h09w}IQ9$ta#$zML2Uxd2s zd_dx5OYy7hzVB*04zJ0xpOPe6OPjGc#+uig^b;&$F%Rb|i}!iK(tkCZ`$HDqS^}1s znYVgq^i+B#G9N!y4=2G%XNMqP)Ig`6^{el`P$Fv%Njn1{vXX?3D+J zmo1O)vN3kge*kuLPxdBBw3Z3X()Vr1LKtJF$vxI9OUy-~E=|MO9{4IzS&w0iS-c?e zvZaitBR*SB-eCF(##p8IB#G7%!z?e4$KiaLKcIx0vIIv7mR5%Bn|O+k1%Bh3F_Of~ zmcnn@^O#-NHS54~ZY)WnwKN=uV=QFx%OGe1Za*YzhqAa``q%Ds-@_B|iZ<4$>Qd-V zj~qiXX}Z(bZ@q-=v?jAWQlLA1#T1g|KXs@3zU+X1REH)-;-NeJ)<%-<-@4P2Zm_R2 z54-I<1l{RD+eoVaQ+N8aH4fI$oj%N$q||q(SB$KPOmiEjM3@~pf=D7YqW;mH{^L38 zPM>(fH3YiT#~vn0{>Scg>jsGy(4GG41WESK?sT7_i?OGCXH(PnFfgy4B?xjTKzvh~=V{^R%@E%rVyQ6$-4q6Oj{pdmxFJC>w9N3*;%nVoDQRHY(lKj{IFY8A=y9$=dG0DEl zlKI!w^B<_P?x4RNxW=vZpYr5E?sO>4=gtlNfNDPceY2#URFzzOtbcQ6+S|XU`Mwg~@pf2;SwmG3J zJ~4u&eIfSNRNK!OSAQH5NW5%WH(?=KKE@^Cn?}JTl0<8%HUaC>&iic@u=rokb5>a# zt_YUA4cR%14`K%3eEEEj#LJeT&CY0Pm-wm>jImpfNfNEafmvD_)x)no8B=dWDNDL! zDSCpf{`g(%Y6WBL{2UT5TbvHEH;sg*-|;DM*%#JuzaA?~mSlOC z#SDcrd&jIjxwNW5&R*`MvacQ2jw5bE-z2T7u}EMyj&OZ~C4YjexkG-XM;Ce-C& zCAMO+yU4y5U>VSv#LJeVb=a!h(8W(if@ObCl0<7sVwN1kPT&sje(N^BRhG!>f~AN* z>$q!Qbc{Dx4BSY(Y$>?Y1+U3{o@~IcK9O5V60N1nWE^9*mD=Qox!*n5B2!uXZU~mP z3)u6R=|!#LJe{t*kER+l<%(^X2PCl0<8h$aqIE?$`XH5s7no3 zw!%0&BcU-^_U$3@vL$Z|wmZstY=%2n>Iae}T1z~$^r+AWPuSU;dt*^$3A-g&^77M9 z){LH*0hW%1NxW>yvS43)S1uVg7AyrylO$S8i>WxqyxguIfV$-J8D2(NylxAYd->Qi z%%Sk=#leypK;mVKLu+<_F&}Zp6fAv1NfNEai&<)X2`&MaiytagR+enZlD`t$9aX;5 z*r8y#@SViVmTk@17)#61(*`W>U3_Y@`EtBw zlOJHI){Vr=mLUV#JM!pv4#8k~+nXfOTE+Z0^T>D2rkw{SqwY59~2hmhk(MB?sGu^SaOp>{J;Y zM&f15`(*YEQ|Nf;BA72rB1jUgC7fA$1#iV2MSX*(4pA2G2ZAO457u#4tJd!FP?w^| zNW5$*7|iC&mZnkZUQMz4QwgE6-0I*FGp zsgdlyzuCFA3s}tWk|bJ-=}a7B*=>S-z~cIR+e~E%ek53GM6+j@v^?$bUK#X`#LJcj zpV{0$I=fvISXz7`Nwk(=W~sBO815wYE|?rAXsY6C-Jgnx(B;g zejoY(zp zyTxEhDngQIEe5l(E>U@nUBJ@eTh(34k}g@s9AnRx%eH&sH;b=1tNn?WE&V;%I)C$G zIo5;aXaSN$Ynj3<`Saz<0hV65zW6Ci+%usrAMDuNuVyl@71SkPArdcJwoGDkKi<#- z*ZD7(AW5{AIA-ZoIFB1x@)?~upe&)Og2g)z`;L4y!5e>Rpl4vc)b8w8!yEy$6SjLZWJfbWf&jm|%Yj%HeOsj#X27H@B;$_R6C+s>_Z_BWr zV7Y5cl4va+%4M{Hw`?gELoDp{xqBWC7+GK9jjw@ka*d0U^si`pBcOyciw;V zB}ufFEM|FGcEtj)L@g{AsVqrpLS14PvuA#ro@P-nUnV(_c-gYfkX@78=Xx><>axz6 zB+*(d>~M^wZ#Vt|mcH#$qm?D{g<$D9pIK(Dcv%%J{kxNR*;4ffdw)4}_NgaWru8LB zw3bL_v8_2{Ggt~vI&e){{L%%>1!HziuJUn?J6MX(BJr|ibqK4A#Uxuih1O{SNussX zoQrh{SeA&VpG=)QH$hn(Uka9PN7%`&@2-``-ypJGM&f0QWo|b2KQ-Quox^q3kR)1* zBeU#u9peslNov&Pp|ZS@ERzx)ajkD;?bi4%X4#s=%a-lg?7m+=Wm+VRv3i|J60PM8 zvjh+9ycH}%f98LxEb*^|y3~2Xc0~ss{X7xI*qXcr{>00cYY*9V?ELVY`N2}R2uY%~ z44j8!%*3cuPq57T_$Xaj!d?rOQX|=XDY&R0jB2 zWht|4i#yyHEW6#?|5TQgw?bW%Wf8M{2IK$!jvVz)utYy%&&dWe)(inl)=&~JTh{I7{V97t9#LJe_;p`cv`N_tCFvf~5CrPvx z!-ZIvjps_81ezWK}rcqU8N%$btA@Q>1z?d-lwK=Ka3CU`cIAl4vbC7Gqrs zZE8~v>T);Os-3c=N|xri*xGoFUlVZMGX5Bemo1knvukq9$+_8J>2QW5(OM=j%e(n2 zYJ#P3O5Lu?67yN8OQqrLn}tchycnp<=$9m3wk$jBh4+_DW0P@@qVWfkL~Ds*mU5e8 zc7kP&|NCCb68uH5IDcZNY()4Bt__xMp(I|mJUh?kezEI@yP+lrQs4BV>6l_ z!}pi0@~8SMi`!ShGWIvKoLV~^El+EZc-b<|-yQ3cquYXNV6m@Hl4vb%%wn3n@if#W zHhIl3WyzE*6a3gdrLWHA)1fX0UXXa%viTc(PR`M?R2o>uzb8qwmP}^x8F6V7SSH-) zXQ3>KSwdYBQ<+7S{jTPElB!TGYYW0Q?gmzO6=60N1$GOWw{>?FL7&0G^? zt1J#b1k0!P?3!F>hP1WlQu2wim8^`{LK3E`tk`Bw9;4v#fnIAP-myS2A0! zEOFUFUHW^ly$W&97e56{y;>w*wrmV!*D;IJp?knGy%9;GwU{l(F}C!>3vaOWUR7$1 zvV{H=Ecu_XU(sk+F&6ib%-KNVWy_-l-Z)>1Tpu_eEcrK+Bw9--v-GvxUl1(0M<;Di z7LQ+o#bh^o9{ZTrr5)78cmj!+E#>yJ`~E%u71*0_VLC~owUkknD8Csbks49|=(KN=nT||bIRb6qiq^qadPPURYeowYJjP)#M5-(pp13Iw| zqABNl<2hLKJV+Al|I0%5|KjxdHtzJg^Qy-#Wl8!i+zA#MvsLVFNiT;rnO%ImIgCP60OCISz4YvhbJIB8RwIrEZLH!^<{Q{v52)>0~W98Bwn^u^J8xs z4F^Xb086MHNusqZXBMmNxw4=xy{=Avs4OXKg}MyC%66@7-;^77c-KuJ@v>#rCRP{6 z8Z#<^CH)~uqP2uF%f_nRo`9um!=_J_C2E~unLLVpU065FyE@cmV@nb*TM7nj#p~EP ztNr-fgTFhFBwEW;W?9;`;v%rr9`-X`SppmdOTn5Bc&|(_>W`oJKW(OOEa!nvPqU2+>(hF!S)URj*i3zkeLc7MtBsFg9)W!_a1FI$d{X7|cI zT@$dgtLSZ#L~H5EEZbxIo&Za;G`BCxk|9~jK4ZV57@z%Y9#{syC-JhSQVVu}2}$bEHn=jR_w73Lw zzsW$7L~99RmK=F6V;4czteOTB{@(2ozCo}glx905c9v+@1}q~Mk$BnS-;!A-pG!Xo zmXa$;60IePS)SF;9tU-a3Hw%1S-hPE%Yh>7I@W*I*}7o)c80{umW|QdaPC{bxzPbE zv6n~^t)<9noclveOp?G7xcX9YWwCb=EPZU)my!vEAK;n90nQ{|w)7~n9W6yJG`QYWwQYFi-`E0GK;glKEVm;`ylg4Fn9Y~&#rBkfx`f7% zBwC9fvvdsp{Taqs9h(}pl_hwiVENvF?T}coxWj3%7|bK_vZd-;W^wAYegRmHE+t8{ zmb=Vy|Ndy~w(WELRYPTQa}z8F2C@0F-al#`SSFn#@v`OjJa&IsY0z^eSdybj60Ie_ z1J3<#W$WSj%Fi4RH&>QS$r9I=?N)!jYz=m)4R9y%vZYvIW@)y~u?kqec#$MpOFL${ zer`ku7-JisEN!POiS9yO+)dehDeYJaPubY=ip0y7s6On=Vr2Mkocjeok|bJ-4YSOf z?l}uAEB+*)SF-yTy zD_eo(Nmvn6W%1c8Sd!D&I?Uz{d%nSZxq61g%a&*NccaCrWY}F8W8E*2Bw9;6vt)Q# z<8O5D`EwlqJqnsLRyu>^fHU%URs*adIGumo2@Quy^EwZq08%T?ULKNwk(p%(5%c zyb;WoTKP@JD@*7W!P0mT+ZkDCm;>(g%3GYo%a+sm+4ETMALS;4C8QilqP2K2OXJZ7 z`2ONs#BiFjcx)9c1vA)ntnQnGcpb}Gj>OBBZV_z0`ALC zr!7~O$Zdk9x)=K`X!~9}DuLz72of(_?4sGTrL*m$lVIsTh9uEi#xhG^o9YH&nf|fq z8fEd@E?6R#u(k2``PU2r%Z48$UbciUXV>I+53iR5i$U&YfBp!qWfQZQwJH(|7V|y7 zHYkhZ4#9G51zQ^*&||g>SjHY8@vhuQl}|HCmo!1ARliI*)~>azLr;_wf=CR_C;NwgMcW@$0d_bpi5R(}ssmTbwA z+LNt~*YF-Y8Z0NBNxW=1S(!b<6fG8p^QDw0Nusr!WtO?iob13--u7~svZU-5>N2nn zdxmlPVZIbBcNUU(*)lnhSq{8t9S8HJtvyMiwPY~Mz7^eYo!@Mq>uF_)@)0b{UD@0( z-5_QvSQZ^5@v@~{N%p>9esMMYjk__&NfNE4h9k~>|Hu*Fp)N(cjkusJ0eb|?rmpOJ zaz4bNusrcGt2#YZ}D4_$Lb?@l_g=XP?zeh*mW%5 z48u}Tm*OrYUbdWh!7Tg!`t3!qtnna8w3b)Q689`~F4V>J+loiZ67DBh-d$#6ERXHz z0I)davi}n=Ti&OzF*dxN*=(?MD@c-PEtS^e+#lD{y$e{XFBq1pEZ+MBOSc4eO>TFt z(R#4-uqW}dCA&SljwMyK#FLMXts_aamj2B0#3=*2pzGzW{90M;_Y0OtFV+RUcU{sa zu(T*j;$_RMHSE5B{dJ$#P?wWsND{5Zo>@%FE?fe2F^qfpL0M8I%X(Y(UU{?pw3Z{xa<;`A+zV%X`f!%A#P|z!vE0kLZXR3uW`o5t zip0wncP}>gO~(Je0+v@-NfNCkm04=FnL8iqlKy4!Z)FJ%5G-?!vbjHXT@4$s6wa~a zPrPjT_?5j^w(tnUj=R12ND{53j1$iN;E@^Y!7}B#Y2Jx{e@Au;6fEn?vOS8Ymz;SF z7W;c7UbZaF!@ltqDinw(wzqyvl4vca%<}Mg+n->`Q@TW9WyzE*zoXb1?5pRM-+<*> z6%wzp9NB~OWnygmCoski)h0=_mc`7{bbMEQuS}kwSW;OMgM_+7__MjctZ`x7tI)!p z#LJc|BiT2L22HNxFM_moBuTWEU}kX|H3Lr#NZR99L0KXW2o?{1TF~yf_G6$fg%6N; z*|O&vTeoyr;*DKu{zpg>ttFXRoP&x4!x$U4a(Z=T@d*|zu^#N1fBd61IA01ok$Bm1 zgm)0#N`HrAEZm(W(OQaaz`1X8dy2|1ZB3R~{v-xu9o>>E^OL#{Tudz&H z?_!%1i(p-9_8>{LmTt^){6WrdP?s*H3O7-fbjgyqn7!|}_ZzVrEEbhXylk1Xn9cnZ z%im35?%%0Nl4vb+ndRm9TeZMq7TR7 z(OPCQ%U9dcYoIQMuP%*HmdGQ5<=sZMx939Nonv9Xbo@l(Ws98^`<7IzSpFAaiT+NK zXe~a>^4N4FK95-kx>zZT-%-JG-jRK?=wV)_C0LdnA@Q;$?H*ejZ&7S0uEBOZMUrSO zx0uCtU0Hvqi$n3@6P3j=Ot5Td!cL#xye(w{SejQL@v`O1-~b$BOZ!>Zfx3LEPLgOX z1}-@Fou)Nj02ZI-HEfjSjbv%jk=-k&`M<|sY1+|&#LJekx%Q!D&N+W?u(+F$Bw9-= zW?5jr2)m&FI+S~kvcw+~>XPx5Sweg(41v15ZcXB4O9OXyP0qI8Mxrcs+x;UOD@v`OVd$yk3Jji|z)aAtml0<8XW0q&@ zYC1p@aMiE=8L12_Bgy{Ro&I)|H};fwoprAi49tyfNrHdrPQQL*7iJnde$aL3PLJ$NGX1yR>Hqix zP10Xgf$sG9UL@^*?oMBI-5a~p7ahnE3*G6*`jdqJRd;%7|87{*4(A74gsa5KVI-&a zD)EC|B?2EOm_v7ZQwx&jf9pJi~iWKaxaiF?Pi#rzMZ;CWGZ-;`wdL5_M9rcy(r7H=zS#nu2B3 z6A~|5u8#}FCntmF)>xOlFG&)uWgN3CELgP?Sgh`?^HG+7Q-WnwPj;;z67U1Br_m`S zUbeie$KK3423xm-xwkT%B+*(tnB_`?+qgp_V|{;rWpO?&SSAf(JBoI^J!J`XDYu!# z%a#q_{IM>N{8L+iW!H9+L~Ds=7K1A5u?xD*gNg^0B}1~5xWvX-(}8yQ)o0=}5-(fg z_p!IxIG5(Q(<|*2Nusr6F-zX_MRLIy+u7u4n6e~92z9Y;%jSN!faG^j7mGF|Ubd_l z!z^`2W&MJ>gm)oHw3hlCaqc(va3~3uzN3RqD@*tp!Q%Im?O45j!wkbFYcZ0oYxD<@+19jPdi^R*8b{*N9 zhUKn$SeHo;ND{5Zm07No8tnyj8Mm(A6=ktMCs=lkV=E?)Zv039%WO9iFI#dIWM6!b zRJexc`Mugol4vcF%(8Asp*XOV?O8frSyClSr8lfDZC19!-5#q4ka*c*)Sa!!ceYC{ z0%PpfaFRr8`N%9EZ7buc0bLtExvMNOkwRU%@N=&nD>U$iy4-3_;$_PgXSTvvyYeXf z9g=RHNfNE4rW?+E_eR&y(qU4-BV`FbFIc=Qvt0$DOVfverO7N3FIz4@Wnbf_bt|$Q z>at@#Nusq3XO`Lvv)jWM>*6pgRax9F2$o50nZ+^m8Gcy|a3%4w<@8}@nf zamTq|ty~2>zyGSe=Wk{4xg=N$1hD&y>FITNe=&0+@v`OH4K`nr?)&`!OXE!>iPkcZ zS+YwtSOImpI(1s!Nq_J5aEKNx<@gCZ=8xxk!F6n;9f_AMQya0JUhU`jVwc+bB_xT~ zvWi)}oX+5H4~EWZR#;imB}=sj?0IaK-I^s}iCaSAWlQp1wjy8u_GoLU%b;~6iPjRv zEK%KaRfD>eukyR3vcz2$>eBro{~A}~`(Uv2sYBvri|2i|Ln32o_p)FKX+n}{Eoscs z=9Kjgu*|iHuAnTTF@k0IIJP_LO+xFdVA(d2#LJes_Uzg6=TENyuxyw?l4vdEHsRb~ zmf?oG3Cq=Tsje&@R|Lz#W^5OZsZp;2V9C{(#LJeQkJv8H3bxm;g5_>Yl0<9i!z@!= zmp_KO3>-MJuCioF7RRg1;<59}0#+oG2T9z_PY^e=F zU~ySiy@|3U#R_#XU&hAR?rCT6f+fBkdrq!Z%>_>gZDvm5Wy_(#?0L-cWA}Ps`DIO#Xe}np(!RH~2h`=K znMogIc_Ud`_OsX%e)vi_h03GR2?jX`jI4B%L8U{`Be;``HyxEvQieW zc)_x~y)VxF_sL6xz_Q1b#LJd@M(q9)Th0%6-nSk|l4vc49ys@n{L*m8YVNSP6O|=f zvRp6Ap82m4(ONn&OR&=;T-$ z$#IMC!nvFG7u2QfZ4xhACPuO`Haf?gdtfnoNRntRw#+iqpf2txT3tG?ow7vT5-cm< zv$?;a*IPrd?5#rLWy}4!Z0`48_;Dy$PS++$v=(1x2|D|&7tH;2Ha8b3OTcZxVtA3g z?|(fu73a&cc_d!8?0(6v$&WMZJ^;)2r6h^glE5sHNym4CW#X^x_R8X%AXv&TV(a{k zLOS4YLq+7<@h4uk?0RqjpDmMC4tWX|^P(h)){@r~=e~bzj)!1zsWon$vSdh>=EvB( zSY2PM3>ahk8k2b0lG=$q!{j-7)7EurzjS=&CFUcZ9l>?9JYh zjRL>oNtCwRNW5%ueHx5)DH+=@2`nFcND{4O8nZ+;*jWfHRh&P0C`uRB=W zdy#nAGUo`htnd5)zwhT6K$2)JIkw>3|82E!2GnKQNUK0)NtG-YIka~sio&}d>^}2Gz;$}4jtU~ zg|LLZVJuxYk-B{EJ;D`Pu8jooyyaO6`Sz+kteZ2k44D9uNG(Rhl4GNTe@jjKd{&CE zc&0NJODA$4Q_%|^fh-3DKs;|TcOvt#PT-lMv7`3{muEs?}B)4o+08ZT$}?a2|AT+R}6 zgFGw!4W@KKmV=HUp10^%lKYs|B)=47DRKcxq?Tf0*`hZj6It%9HTWznv6)O=Ui~C< z%bhj*;7-WX+dw>T8AR9mF0SeBg)D}9KoY5?vl|}!RZq9KK$eqg&5DI3FpIH_DJSn@ z`#yZcCs8VN1M$3NXbu@KW5;Zoge=cAK@zED4zVnM<%mxTtLczhE-cP(8Oxb?@-F5U z^S}#TV^dr~Ja75rNA~rVo;_`kEUiw1BvQ*kVsZJt?-Z)byZ&C)P5=7#TEtKBoWO393@w{cnTT+*dJ-bdK%RB>+L~02m7WEdTlaXcOLyNk?lJ<_N%hPG3F1pt{ z<{^uFJc#El2j7tA%eMt~xPxe53P>Wg6cEdzbuQzO<+80#V_}KNW-RNMko#E9yQCV( zQgsBx^Oix|$R3j58rf4(U0hCpBvOm|2|V`eY?+6D5yZ%`&sp>hEb-d-@y5!#kfpW~ zi03UKlgO!>{fxJ^LzaYAAc@oxLM*PX>+m=69+OPF3rqY5rY@7Nkg?zE=1Xm4DVqi2 zc}w6ZGEZKmzB&w9VitlVQcEtelvgR6p}J_M_U$Vy!MTiOpSl&Ui(=58amcd&9f;>G z=~u{l@|VzvL&!4Y6G$Spv~kB{f8ytBACTohbol^bam!;Yb$^jLOyl(LyOHJGQxMNv zw&{`g{r<+P_*r=*2_%tPCKAi4p_O=_(mT6*!-S=Rv#d-fby*QTADn$mM1=> z3wr!M2R!!8-9ZwmWf!r`Q{5Ga>e6w?q0z#U`H`uMl?^#zXI?!8-2bx12E_9gvnFKE z)rPs6c%EE*2qckO0*R$`aJwtWaxG}aL}7{k#8`ZWk+I+Oz+U{0Y;+sM^A_`cWDc{; zHf{(SFFm3_5~<}avCO{JWjL}ZELE8yEWY`SWnSJnJoZyxwC#q*OTsD;&s&~1xQ^fV zdubf(g)B=qf+SK)^OJb&d#ki9LKcS?ALa^+{b$C~u$;_cLNcd>A&bRp5YJomUC5rR z`}OAGx}13jl1MFMh^6VY;g!fTeMx|^uoQBZTn#dZG1$1i9kLX!1o6B@!-DMR->~F3 z{#|SD^&pAVvXxkBUO99WS&Te4nF>o{0aF(XPg0j}7Ml2^<9ZiCJa0KUnap9rkA69Z zt}(MfkVI;^L@Z70Qg9z%#5R2kVF~@hSR#IrQv#nK_|g&8W!_#8&s+K*Bc1k#uXfsq zEV_q45~(GPSVH#eXogW*ZYwr+vB_NGb2M@AjE=+FwgNl7y}({VpJpxT5~(v~M%-11_MtdW%TZY5%zw zNb(;$?Qfi2=!QD&AM^vs{_M0@D)Phq@qqy{8yd!b9T(ZHNB9s ztOvSFSkD1DWp@dyS>!J9OWV*0b=t381k(JsPWwmSHix1o^t+$t2gDsGo4b43-yxmD z{a35hLf87+nyY@t^LNh~XUJ2n$-fuJXdXx+wImbEb=$)YktO3t+z##qJYGBDJg{mbU$ObU+p(Q`JCW3H#1i5?Ygv zyNswB_)E#uwIH6icOoosZdB3+A+mJ$H zGO|ROf_UC?Xf|0b9q`BpznOck0ZF8me#G)AS$!F@933*{iLk`}VCvGKiahtfTYKY^ zTrP|T@w_GL7TLA7RV5gAceI=gl1MGfiRIp@fp}GJb49Bc!V*}{Si*miFC_sJ9;c(a zXw+8z9nV`DKOtj(#`@=-(KS}P5lAAnxDm_Nr^`2>x~!g^ogysGKN-sqTXKz+_8Id9 zS&kTkc;52PlDy3h2-U+oB*vM6BvQ*`V#)W`8;C55R{j~nQpQ=XY$BGqwl6m#OM5#I z&s)L=kg;!Ay9+*FIm`(pky(5trtilwe&uN?_;|%&GFoFx7qN|!V>X|vD6tvcHw;M`g{zsv|A72 zc}rL3N9T#k-QRyqHVsvNUzl5o9U31>$+j z-ks#Dk4Kv|n#l4v5+spYjuMNmdm6qc7kfV|7Z%$p#aL-vrdV^;nhhKD73)@STCIJ} z1AEEbGT`a1d*~W-)CKXpCGi!Rj|~a4ZH_FiBS8|WrK=|%`&aAv;;zm%rGpv^OYk1X zGQNU*!P-&T9Pjq1TW#>~c;52-5!vCr<4cq&vRtbJl1MG{h-G!~9Q@nqPnym&JMQZ{YX+&4)k|sbway6t6kC99dp#eIFn!_WK!2 zhy{7CY&-XU5VG_i0^)f~kqVi^T-)0ncc~SP1WBZp{lsG5-X{@R61*da2}>boSxUc> z_6f`ZJ*m~wfVF|TmEOyOy;^qD; zPjsuHx;$+U;(1HmcrqXB>-8M3_4Vrkl1MEoUU=*eP@jaqULVfZoFOb8c8q0|61gXv z&s|)EEDoPRJa4hSMy|2y?*j1iWkd-`BDG8=mZPa_g3vX#|8dz|VX->MSOV&h=l)}Z z84Z!eFagB#mgEiOo~)PPI2w(Y@KlgQYOx}gmDO``hiYE^d&a_&%~@Rbk>~#5nh6h( zrD+Qg&s$peAZz1>wyF5L)EO0!L~045mJ#YpQC*rfa4;2?m_tlmLK>4!`z7n6@cWBf za}dv49OH;($`rfms4k0DKoY4Xn^^p<=iWk=vaFdF!s2hwSc+k2STL*%({|!#!|%x7)Y@$l7AFVBGFgr6CQMx0*8p8(UR^;v zZ+W|e)Wxn(9sDinmnKLewHOeKPmu8rRF_hx+#SMF%vsU{$=XFNy->V1ZsHE&dCTqZ z1cOSMU4Zuzj$qtmD^4)GwKw|Ezm z^|9;2KjE>TkOGoOE&jwZ<6_)XWQnb_)lOK#4l|b8&B*VNJoE6cjVw`0AfC50J3z)u z_;qq3)njFlL~2PVmSZdK|3nt`5u==h#nXwg{LmuL{c%BxcH{wMCYD{ve6e zVnHnPM%Tcnoa-BTJP?)$SH^PDk31_cs&&O(1RYF4Ja1Xhmb@cJMSr}CEKSycBvOkf zvBW4H!@qMix@h@RVevlBSmNH0cd_}5>sB;gl;V0N3X829W2w?6rv^NpGk-j~j|C)yc;52(IT1Zf@=xlp5Ls?Ufq351xf40D{n3>pcy+ zkflML&>AcL`i>lZim}98Aal#9_1^g*i;)tD=PmXV$hV|1BTVp`(z{k5iPW-$SbnG- z@IaQYp4N)O;(MC0SnVS3`#pUQ;eV! zWZ80dd=p`@Kf_oCMv&8jrnIcS4b^4YB@oYB##fO!%&6%;_#Jsq5J)1mJRp`9qq>J6 zOG2w=%ED5}S-PAc*VvEiYR8c!M;*lTmK&O6yetTuGaHQ;m2M!3)KW$)dbv|Ikj1ZF zrmCMe;4zw?)y8Qw-k3KmZb_hTX0?Kfh1DP0%Ganm*a#iT85TA zg~h{*vAkCzmeAt8c)w-yMIfHHOqx$lK-dudX$rFRGzCed7Dr;~;MQg#vShS1&=wY} zbBtx+J#vzZf@82D8ZVPOfq32$ZA9k$#*^DRA&X)!kVI;UAeJhXS$4>>^K_%Z!jjEd zhD{=t=bgJ9MV320AfC6lE+g021MfQcd*#?GAc@pcL@d36qw#m-9tV^3geAtCsmm)( z@_cEU`MoK!C?5myyd|PHdA{gftBy}|xa1C!NG%;N;<3N>U2;cMmsRaO#|n%8dB&pG zgxtq|6h9h?ET?*ac;1rGl05ep>u=nIuCZ`!kVI-RB$iGSCTvC)hx`?jg~jOtW0|8* z=KLvJhL<4A$`TOITh{B6d2*Sm(lcbqr~pZ%7F%Lj=^EsKEUK%tW(i9%XQ_9F%#$;! zCng|EgH0fwx9na?#{SH}LThBXy%QvnT5b`GxzRRdWI5*ZYre3g_%L-jev#DWm&N@i z$TB|=#PgQBkH~(Q+Y?&s)SZ5#9!M$gPPZ~p!^}PD1=(nHjBWxFi7V=l z?(~v1A8?X|E!-+mcltzSkmNshr%zqoM;Udee^Ld>{_IZAu)B?W$}4IQ-Hy7`S9Jyn z{-!&<*Tis~=~(^nO{hCPV=c(^Z@bg~_yf;(FmpoP=?*(U+JEj&zk2@)?oNME(6ktJ zr+>Bv3IA1hdhyuRxTf33kL`r+5>pR>oU*%w)fIA=n0?vgE$U7`a}=ccZ{6wozFWSc zyJxk^kN(y zWyz==8F)UK*ew`Y-fRW&yd`NMnVq*>v{nsSY^*>MsbxH|#MSq~ zs|?+v8yy#xGR|Ull03IOO$x1%WzSO(&s!Q5kyW|Z_KV3^pCph(YS~FFWlLV+-?p8( zG4+(Nqy;c_**NbwzMXZ-2#!UTt(G93w@h>+vp}7!pAS%7@@+s8sU?6|vO{ZcL*vEe zWGoj7$ZF|}?WOxrUCOjTJa6fdMZS_M4V^I#)#cn^kVI-}>W|0%tOmZ-P+dke z85|@mwpSU;>C)}^8cVl0i{H#!w*~RMMJtQ!^!n0II~ZA%JAfoo%P4BmHGYOHdEFIm z3QIm`(MTuvvDLr4R-n2hXoGm(GJP1SONqx>yrXE7E=VG^Y$leZsyivjvaZ&XyTTG5 z#MC9*itJn|jp%U`S*>sj#>OGZxLm z5*RRP+dOh)Osx}nb(=RB>Iu_ zl{K=~KStNs+hHJ{x3tqCt8(fmf0ZK3_%R@f)UuXXx-V{Tf-J{WmYX*`?Es4ZZqTQ+ovQ)(`{UR*(p^Qc0D)|kfti>;I z_h4EQi03Vb`;e~-+grL0MRnQy1|*SMY6sx4KeCg}0%WN@uIG1QDda5shwsAA{Ualj zu%*^?5YJmWdyzSR!IJ*?D{0PLkVI-3NG$vAox<SBEaB#~NH5=-Zalk!nr?C*xxSozm(kI-9;Wu`g#f;B<8 z5TAPXA_m0smSg&4&Oc4V)&g19#)BkMiwCh}l@HjCEWu946otj(He=BWAP8Vfi$>uD(`jcIrkIs#qiRz-+ z1jO@}3xT9AzPjOSk)=s%kVI{$)3QNo#rY;$)*W>$`wM9L=J1V%= z*WdBHCGY^*Rgk+mav`z|Y6y}@ElY_dNJV2mvbgjL?IbMz;fy7vFZmMRc*q2NdZTWC z5YJnN7Lt=GZlAWSfh-?~f+SLlE3q^h6n6|+dT8176c(qujK%pgxsSCwAB(@neH#qo zd5h{XvSW2mR1N$ack}f@5~(GISQ^BPzl|(0ohE1tOEG5|bD#XCeX}~<@tV@D-XNa0 zEZj_XMz&}YZiDKQ+#e*7TFQxK=X>gtol;MiUeh2-@cs;rIOc2jo zE_Wp7_xG$78iFj(=Yu3tixIKZ@3W#kvV7creyp%~MlzPuF=TET8TSr9Ury`=@w{b~ z4H+-dma2o1rSKp~BDFXXOPs>gC&)7S#OleyViUz!YAz+u{r~k4;M29Hd4YJ|;@ zlQnwXszR1K7eNxKC6ZVYYdPWHQfoVI>?~o)>0ZF8mIm9wWNdeE3(|UQW5EkbLjOF`FGG2O>&cq#e*8M>|Z%OM)c6yCe?VgM3 za%C7uBDEYO76acI1*k5Tfh*PuOBrV|d{4$povCSGQC*B~f_UCir9tXa{fMCtvSdVn zBvMNlv6Rg##dChg%UWB6CG8vRsJ;Nu(C_ zAUyUPJQ_a*U1RrVJ+cuN??;TKu@$+Gt*W_b9BKTf+2;bXEcbRgBrLX%8B6P>#Nri_b`4qTUIy{J<>NfE*0(CQE?yt=3kFG~78_!j zlrZ5uvLw%1cvM*OIg6`1884mu>UKkxk47M#x17)+-z<7n|Eh}W($o|rky=8CWoqM2 z_%yn1b+k_iOZ*e2F5AM%cjSq8=i+cWROH^nMf=XKQzZ3 zs*NT&UKW-L&hoh@sY|W4P4LOBa|}T|Z@HQ6hkt*p{aWu+$WpWrB#~Nn5z9t{oFY`0 zXQi`(g(WkNsmqlla$1mHgCYe~m-K3ezvFpJd~b3eOKae@5?RdZf+SK)AhG;fyc54y z+T7H*EiBQ`7|RZIa!($1a^yo~xf2QEdCT3|Q6mpirIpqB%z4+oeWEojw*Y9}VV&+BWmaFRStBEYv6hRWHWh=34>~nSw8vB}N zW0Qp?@dZ;CyPafhyt2>3MaXi0Er{nW+MS5yK<9S26LQ;DkVI;^L@YBcln)?Fuj5VL z2uo-JV+p=Q?#W-?jl$z)wUj4`~lamYmw7f5-Ed&ufXrGHWNE7el1MGy#A0CH6L&&(ncSdCSo~iymJ374 zeJtX~f&s|VVm659EsC*Z&L4g|Vk5E?7=a{GOER&Hx-|48s!NakFKU_p^&QzMg|X~a zu)u5MYd$aUjVzxIgLvNJw498Wk!HUdBa7v6kVI;!8-mAvo%|Csk!9kvGYy2Ln6v2l zlJPR3dQ1+oES(SHd5eQ0`TnwMVZc6Qsciz1NG*ej#k%y$5oGBTzPy>Rq`YG4^4gf} zyT~#h*%Dc5J_PZ+rP={v`FuR;8nR@?fh1DPDq?B;RI?7UxN7ujBP?O5jAce`aF)1e>g!s7Xwv3x&9=E>1@ z&O{^26=x97TS9h{y>P4U_V7h@S>pkcNGkxSPN zLv<;>2;zB*c`#Y)J882Le}Cy41d>QC{fNbP?i*EPaSU9nBP_9Rn7Y*7NY=)i*)_-C z#kSrA@w}yN99bXpKd=q&PyTfeB#~N{6N|k@YrLj($Fb`WVF^rUEc5G;?(~OgaR*Rc zt{eyPyv1ZbnOkld`+fo*FK0j!sl|<0Of}veM0IJmp-^8~oHH0p=WFB|v!7h|BeKNr z0PzybKym`Y%8$q6kmc51kVI;EOe_|bGx{M*FPE@!!cxXr0u#vC|JWv}Cb9&a0r9*= z`y|XRM#DZJOV6T=rNUzSj9IUNu6U_8!q7-QRWE_je>;nhtHjY0*ZQGI5YOK|TeKl}&v|2apF{tboW~%E?Eg}&EBU{y^luh} zo&-~MDlLU2_&sC!KHnV=xx4nNqmX6BFc8mM&YJk+C-lzHLHMiB&@mv1)H07)>O9U~ zfh^S?McW9ATMlEHrAj)8f)}r!jVwAZKs;~x)``qE?A6ZUvk8B_0!gG6dtyJXKB}z%r=f&HClx%GlD=oZwVhsb`xfga_)()v6P!2iPREKEbZL=%u!te zhZr6emdp=KT^f8LGuqphP4GF39mjxp-r|@^-pubV(KJVvkVznk)KW++7WU_(k)>9K z$_ZhK&SfkXjmS0Dp-U}uRF|p`AfC5GDv)ldJ{#-bMHZJYAc@q{{uaI`J~KUzzhJ3s z&pjh7zIlwru{HUEwfu<9a%35|7sT_Hdr!!Htj53`yld^nL6AghnMo{>y9RARb!o9K z@Pe?|e`G8Qx5zzt*X|LwkR^5-i03UWhLQ0S(5F8B5^uT(B#~P76U&5V7qpROkk#hP z!cxdt_G*#3I0nssi7cviAfC4z(+b4*b|MT=2g8fJlb-cq9`S;e;Ky3_;JWoaHrBDGi%OH9d^smsul01P0csL;$O&ET10Nf^W;`* zwOmkLT(^RF-co#>eDzUzHLM6(6s$lJsikEY9{Yh>zeb{KELAP`ov=84Wh`m_bhqS6h-^Ol)57xD9DYElO7=yuWsNu(A7Vo5sj8sEo8pFQ?bSc*A|N*39H*EmKaIZTU3w_YMkk2@fV)Z$Mpjka#bJMVP|_AV2auy2e-xi+~cyBt;j zfh_tXKs;}GbB@gU9VSGiAdAsBkVI-pCl;?=%if~#qV@QDrLcGwGnSa2q%N!7m*Gy8 z@)sbUw;YQjU*ZSod!0v?-Kijn)S`3;kNp8Y>iEq2%H1z&t@>-XhfN7%aW^5B+9ktn zP+j!wKs;}WZB530u<^-AWV!DIl1MH3#PYu1R6|skbiWe~ge8}=^czRUe&*HOLCBI7 z4B~l9*IHyg7JN(@f4yE429iiE8;RxaaxMJ6U*5>LnXtruXX?^mE}17Eyk2l0S^6l0 zc;4b?Nyh%g`d9GjT3^&a5~<|^vAmi5^cbp3$K&1G2uoloV`&;g-uF|xr{oSRaV}#lKds0&g;|SN^h1{9OCX-N zw9_GTn2BHgFCfd5Adp0AQ4Ggp|4Z1YDrDJZal4DKlyMf%Y2?sTTl*d}kR@UZi03WU zj7axj>*_@tkmb~FkVI-3LM$WS*23>%(?bsQ7M8RhOkK|RA#<1-;STtFL;hV5&szej zljqCnfZp4XW%WalL~5}hmYbU*$O)l)rs@bwL^)$ws72=d?}zu>jp|bGAc*HJPN`(< z7kBbXLzbi?Ac@rCNi09(KVCuCn2uiSA;RMQle6fM4%J@<9r1~Wm&Sp3-g4?D8T%(| ztK#ksgJ~d%)bfH@-c6Zhk1XF*Kj;gKZ3SalI-gh``YpwEY1I|P^OleXWVgCOq#0i8 zE9nE0NG&z*;<2w-y*FNmIg}naPFV6eOPeXgVqq114_#xSA3!{BslAC@V@2y+6VNrL z@&zQ3TKW^q#PFuL<4#-az*J$0|Hag0jXv2K`J_bxUbmc}3*vc8QXsh}udF}07qUd? zgCtUm8L^}+8HIO9Og}kmwy*?OGM34E$ePml;qlng!yCl&mVrv-K2~pakv6j2_XA0! z7I$K4r_&YxuEM=bDn`QMR>fEr+><=FBO#PgO*`DB;pxGB5vuC=tUAc@rC zOe|w3*TKI-;=H<%m9W_FWh_UBlC{2-$;q9O<@sh1&s)k0$@8UWpL6(k6gTbyNu-ts z#PWH6leMTWbyZ*P7nVZKGNdzkNA`Q1ig&CI?E>O?OKv%-%idWfS?C&z>J5@eEoH<~ zPmV$NHcVX%ev;oEeXebc$BWWt5YJlzj*{m~yYUfD$YQb!B#~OW-@{|y zY;C0tvM6oe;Vdkn`xwjjCFH%bT{q1vbd8zs1@XLP$47FF9ez6&&y$M}f+SMQ0%9rX zb$A@IC=`!6DJ&lQ8O!j7WPQw7yA$5As{0GX^OpSUq%IylGsttl=IGylgw*0lEbH~% z{gEYfP9ra2u{yw5o}3`>FEjme@xF`7#UP%ysJM_0qLP@dkI*%?dj&`$wL}n$LCcT0 zb9hXr*S^A%%~|%eA$94Xv=E=>FsnC+=Ph}y$Tik6rylN5eXIkLNG(OgqV#6&a8#Eb z&AhJ&ON=d3mv8mRJ-Ky`mlYZ>Wo66*sl||3tS6i*Ll*VL z{qG2i(?Q1a$d^1{!gLyyB1`p^AfC4*1(R=lE2eFKj4b!ofh1CkEwQ+FY2O=JuK)NI zB`n39rA@V+cy4LE%E|y)o+?!Sj^{0F?vn3fmlqAeU7dpzK@zFu7P0i+oIMv=`h`7s zEG#L9n7TahCSzag;Yd$pc^wVndCSX5q%I3~pPq#3;u8y!NGCu)8K+6=;*so&PowL#6U6hD3E#+jJWs~UCDoUBeM}=8B#~OIiKSzm{9DM9GrRJ=u;g+U zvzFvtY=v6$d}L|77{v3IBfCQJH5M78;f3llX$43kwFDDO+@*j+$kOKS{d{4GJK*xx+;kl2?l6l7Nu-t>Vwsx182`dS`?n6?geB04v3Q%1 zXJvSwXV;NM=@W?OE&FqcrRAYE__WizMIedPq8yFK{_$C}@!G}AtU2Yv;(Ua$gu0Qj zzbAeP-app&BZ%iM55|&juMG>IyP#`qXc0&vwTvg0XICC;qigK`TJ`Fy|N8b?##s)} zB9>n5g7EKKnrsL0yyb-QWjtQo&UVFfOEqhdL~7YdETh)n)kPKy^V~YZl6I7-OC3vM zap}7-4P9f^Hi3BFGWk&uwrq09!TXdp?gUAsmH=XzaiZoJWI3a7y^*j)9Ahj+3gkZK zu`_-IvTUpm;(1HgujIKOUS<`5uCef@Ac@qHNh~9S1MuludKo)g2#dEfV|lUl7@m(M z=(feA4rg z^OmX#Vln7g-x67l9Rx|FmIe>-*k3++_I_j;K5Nu4VX5FO2DTgVeJrxc>jY#`F#_?t zrC$PBQ_4CT@*G+6mVqQv%P?Y@G`+DVvP^5)V6?Dgo?z;dGKq9SJ4QUmCm#M(1@XMa zvNQQ6e&*w61!U3h0+L8AYl$V!tQOweQ+)U31YwDGXDpj1krNQc*N?)#)bb$)#PgQj z>xkvr(CYTcvML@Vky^ZnW!U9Jd?MA>nBJewG1SdX0=nNqb6Xh9)>H0<^HMKfA>4^VQFM1S?7rwJyCbMb3c$t(w&}G z-xhbLpRj1AiMrE|4gp#Ib9efz&mo4WJ6(MgNcVT$=^4k!TW5sg^a#|Qo;Dt&`agB2 z=QYU4K;7vRr-PL8?(~E;q&wYbUu}HXGM)<(i7V=l?(}QXx%iP3e{6bV)SW)r7$o_R z-RYLiBL|`G^r+<^*`M9%ZZ}hLPx*O`rUTH0X|Nh3_?zzZe)r$tOxh!w+(X^z3pRjE z|F%2*k3Vqji^plGJN?KGkoKRu(?d({;qG+JfQ=TYJH4MZNcgY1(*s-Bk9tO;Sq~LJa0L%oSa9VQ?do`TFW#8Nu-uIVrgIAu?@1=CA`}$ELNV3 z#daU*PTz3m6@Efnq=9(evfhErHa?xZgwKp@m<^IhE!AT1Y~$=Kp8#aZUwwJMuw-)< zi@oHz<@rtne~qiV1;q1~6;@=toK4onzo`-AR~5YJl-9}GrW@3*BQj~7LQP}ihasq7XF>y=^h}7)DlB1*8|IEAWL_(#IwRu%vqj{AeND< zX1_soDenT}d5ijRQkP!?qHw3mSxt~cYAGj{Juj?$k!7{2hp(`tcr$gGokYIIz20-U zJ+dTr0r9+LZ6#SP)tU6GHoC?x_W?Bo{Q!V-3#vCLaUEUNA) z?~tW`4U^yTyk$#Eavy7Mc0vtV-Y9}3Qi~C>_$`lXhAb;>`-BLK=LN=cS&htLLOM+8 zg)9q>f_UCizksa#Uc5D;3RzmYgCtUm6R~LS)QCftg;Awp!eZmYSZ04+hoAc?b&vHx zmZz~Gp0`v4kbCmMdrEjW;pjw=L~4m7meyDE@%fjV=EOw_OD<=znMXK{- z;(1H#D>7b^d)&vr1HaV>B#~M=KgMIf(aA=!$nxy0@pEAbyvSIBQ^_iJdck;nYC!NM z5YJn3ACTQqNxshbH-`&?KoY5C4zZY(E8>-3YqhQ^!s2|1v9!o1yFBmon1(y;Z+8Xp zyd~!)xhIF5{`wSMV}+U^iPUnCSX{KK?L~Dt5LuiqEM=U8VToYCvLs4gw|aA31_}-qOj3 z%waZ9slEeQwz`5OQj7W%JofjU$-}2Og)Nx!O<24yGZueq@{ZiBrfD*&i`sG!&s!cU zlP`;2IleEDMb`o(ky@q`i+;vfe2#BN{Z{3|VjI9%g3`%vbdNH+o`@{D<3T)cF+NJ} zV=wMB%|RB6=^%;JVnZx7P1{(Z@uIsXqngEEyFKzbOGR}uPjKYbvS?JlRYzFjuP}8PRBr{I!#J2;9g8e|6+t|2SrtR- zvU5|X>c}##DM%u<T>e_VsO&(1+rMWfOy^#H=W#*r#)zd zzxdWZ1(HZDyNIRpvhTP{?boD?_QH}G#MEV6IeCBCRbO`ivb38G;(5!eAaYNR(YL{4 zKgI|oky-+YBe zL~3dN6pwwiks-J{-TmXh0m5Q`ow0OiO|G$koX@y>FzOqK=Pd<= zNG)TCWkJ>U`=~B;K35wiEQOrqz8kUhG2UZ~uCd7PAfC7EZ%Ee1{L&xc--y^;36e-H zTZzT5x4#mytm+gyN>~y@n7YL6C-2DKeI0S9ihXYo&s$pTBfFzkCp^a|>VDDzNu-ub z#8T(n^BiPJALuqgSVC_wmhJR+tyk>na2eIbT_42r7NfP~o~(Id#A0-f%^MGrNG)l^ zk{-LM1+v&ZUpZY^JVF_Z%_H(&Ieom|PGkvB1@XMaL4&Ma+--A04_Ow!1xciqMsaxT zZ#Q2w8(HR$>_1mntZp(E+h@eGAS5pqSu`htc;2$=H5vQ;8`heGEU7a<5~)RxSZcMc zhEH=aR;^ehEZLl;aaZzPY_O;9IArPC6~yzF+Z)OI{*07ismOA%4@e@ltS1)hb>Ux7 zT~gmhE*F-VTTEU2+w8#eyK7n}Na_0l-9$cxl|1z?ee*;OR7H?vShW;s? zE-cUABP=O*n7SyRC+lOQ!^Vw9mgY4-|BmM^Lk-AIukQ^{=o)i51LAqh271EI=0R3?XQY7-NFuc;#N)BA^f(`%7IbKp z>M3E#me=Pg}Vllj<)#~bdTdvZrhkVI0i1J$K&?W@242&u)5SPYtN?}EmDpQx=@ zg~d6Fv6LSn*O zCDNFD<7*Q2r9ZNC8v){ZOWOlv4s*BfIsCrgY8*%+we)_D?_*CqP9`GD*dvMeg(c!X zW7(@suCYFgz3>U4D()bjx7mij15jN;3_v_@*?5}F`L72YK8q~vr-CF>%TZ#9Qde<9mV-x36NJU~A!AX! zOHQM+{kasMfMDnj;(5!RaB?48=hvGlyw<_EI8)asimEcu+pt213w z3j2iDEuZUxc;4dfLb^K)^Auu`<;F;mL~8j?Easl?aYw|0CyH6Z5+B3VMazwR7gKP^ zdxEYpokS4NTb7xU`PdCbRXn$Bng)_cEnQ#WvA<;ZW86KM)A8j8VF`Z3SejQOWB<(c zhxmk0wca3}x6Bwx)?hP2cG#f0T-E_eq?UQa68B3FpX5@#_2~j(aeK^IB0CUE@T}yG zs4m|$Ks;~R=1j(msqGW|eLub%NFue^6N{?jv*u{LOmAdXA}kf0Ws)fwFWqK*^G23; zEkQhQdHRUl#~x~a*^cVss0xxuE#bs6>3jSzWLa}lt3p^ZpD=Za@g~p8M{A$sZ{q7C zK|F6+(S$rJ_r1-HM3%RYKoY5?kXRliCDlWg^4&ib*8KGyIXafH%$Y*E)VBM_o~f`442JoB`k%UB`JVB_Yaw=Z9x`a3lPs+ ze43NB@!A%_rpThT86=TfLW$*Yy>z^u+-Y-rRbfed#?+;K4%u&6eZLp}F6Olx#PgP| zSI9cdgsT;MkY&^XkVIu!iL=-iPWN!h{t~SvN8}g34=#2^f6%F$fAfC58 z2_fGswA*Qhqq^jKf+SMQWMa8B{24xdK4FEumatg8U@SA~i6K`Nx8V1E>-8X>xA;!F zg2#R>W3wN~GG;qSBDGi%i$X;w{7w8o|Jj3tC7ZKUSdewgnIp18&^1;y9K`dMJ3(Z; z?0Nhce;4~Y79^2cf`~;^RjCddFFCGidcqQuz|_U=3Ax4&CU*=%mhdwmp0`Y|BF`6} zF6o}AE)#t~5~(GdSOx^uaz~cpw45=*;-AP^I$tDn827t&*~sGf9>nt&?JvaAu2vFW zhnf5tB#~NLCgHJv&T-xZWbw%hoFptxNsMKxH(5`9z9N1YvSb#4c;1p4MBev%jBH$t z>ax2WB#~MSh()=9UteU|T)AbYuoQC^m4)OQ%iGa36j`e3Z2cY2Tc)%mbIV_=M%O@= zAB{m0sbxE{EHR84jVwM3hR+w4lw_tZ<%z`dM)5QLC9u!SKs;}0+0qr?$6i0x(m|G) z79fe#;!iARN=+K0hGy5?qQ$~e_CL%2Z)je>c!fou#U>`3)-N{MrmtAHdeds{Ynt34 zzjM;7(@$&EY2UU8WRi5+54yY?ciLN5n|=s&+BYo+S^lZhK6CVNJY95*{Q3@c+BZ=6 z_M7*2o%Y?HksT&#w$G2CPJ8`&Al3h=)84ai;B3@suh|5oly}-69HEYxYX;wMjz&kB zGDsw@s6RUG`kigJgep+V7n95%{s+V`;jj5Ce?cH zzJ+mOGU~K%HV!2GSDp4@4sN)n<4(t_qPxU^X&|TcE)hoV5?>zg>x(+=E$4tV|E<$L z{?o}kboU(H$8Wi~@Tv2tvZ-AD1Ac^$< zl1~0FHedYl?%OV_)~yy6&lJXzKAWsCrUnhh=ft<(1mbzi2Tju5k9$|@1W$JRYoSd{hvF|KEi*toZ!vvB>aw^-Mkum;$^l8FmK0(c zFgpf+^|@`l;;^tdr!f|TXXH)eg#Q^!R2RLKAfC6>9!fe@G?d!o?({F~KoY4%@g*Mn zXBKASbK)nc_jMJPGS0H!lyu#g=ebrx7L#rup0`xAB&*nwA@RddU263ONu-t`#FAS2 zt0fvQv-L|ogeC0_QVHxNu(A}VsX~U zxqvL+Y8<*GEZ!N6MSBSO>QiIcA#+rhyaphiw^Ys`Z{|Z(+xJJ7tIa_YspSQ+1l%0* z5?R#7PP-~Bwwa9Or7>Ay{9$+vugE7%0r9-$=mfIzJGEXrYhqrX zAhK-ep%N-A`J5%_Ecv>Sa6xwws>{TFAfC7En?#;3o6R>2K$iQuAc@q{pI97B%#xAi z=C9A;!V;gw)a6jB1D-DT{?PXovLw|m_#Mw%-tHv#3WKq-v@w{bx0U7(}7< zOcj>QY^E-$1ITmVYV`@cLt;Zu5YJokj*)k<^gHwLmy!qCAc@qX`3m31%zqf+-|ex? zOUV?L==Y38Czi~|B4(I%LUq|$1mby1h6Aw#XDH$GmHU)~BvQ)~Vlm0{U4-i5b<*>L zu=wUMmTHlti@@nb3%pzXwKs_8EhECo?_0G;$@R&2(J^a@yP#`-1M$42-(B+C>AspR&moIPIY=V4JRlb3h{Bo3qN=4`A}ocR zMTK^#p4w_dxOBu1;+i(S+Mwj~j*H2+d%w_5l zRk{p6E3KN{#5*MFe*y8l<`}n4ZzPz=kFki)Y3f_kNt)|<2#_bd}y1fu=cOr z9-(=RCHV1H1zmR43UXVm;aU_<&H|z0uscdMfBrH~+ z7>oKgGEWX$vL1hlzuXGM^A?vAWKC(~rhky;{%C1b;mIjAlZhW2hHEZLl8 z=T6e4_T|`t?Wit^V?aD_>C}bnDk!dV5kFscPXKu zku{~^Cq7#sON}}ezvFq!fJtP$SnqFw=VRuLKoY5?<7+(jCu+sEMwUxoqdE$U|7XV1 z`=lAZk1d@&5WgcYEC=ztMPVv=uN?Kt{1v*!j1{7P{}ED)A+b2D%)s5Y0pkw!5EiEb z#`2c_-uUe+HS^7VaeqzZx52SzEZE| zt5IFtyMTD!Qt40DF5HZ*e37L~ACN?92_}}dcD>J_YwWCn+I(S&{l?TqXD69kIxcoy zi7W=wKs;}m{*YLfS=!A7Nu-t>VtHb{Edg1om**@NmcU}ha^9W1BQG`|fv+*G z8Xtbg^On)5*7_<-)GEjNJs{ zdCTZ5@~l*PUEcs%@^*qGQcKfxJoayn(!lp*mqd?)!eaY_vAiuK7Qe2JcpauzI*8{j zFW-|cg7Eo<_&2*Yy$4C8mQloFbv<7Zjs2r_mL3t7e9ki9HhC9|s?vIn>XN1e;(1HW ziR4{u!DH_hs4iEPK@zEDGqFU4^jwZC&FT+v6PEaLrY<+KN$2p70ef+0S9UWH&s&~l zk!NKui<$U4ay1o@L~8LRme?b|u%*(n#u;G={>fMxJCM3O)cKKw>T=)!i03U`R*+|9 zUX5BGP+d|FgCtT*DzOwtY2uSxd$)UXURc~J7|V}aWY5)&_sg3hOY_bkp0|9{A^R>y zU)z$6EKhoYBvMO*3_SKXxG8#~Yivo2_Y9E2S)aGP&z9lTtm5k*=40#t@I^$d=vZ#Ls@w{c*CSvLNy%%0j zPAdjUq!uq?DPB^#5nW@m_7z14i*FTUspd>hJMGl|atX5BbOiCdrH&GL-_P!I>mI7h z7*~))YDpxP;idgQB1`I#+Yg1s-m1=@ziiaoj?DRUO)vFDmUs^k&s#Kmk#(559_gCM z;^_^NNG-K9@!0R2n^%M^^+(#o2}>boaZMu@L+c{Ehs1s&i03TzR#-gtGM4cg^Ok<^$s9)Oehs{L^w1lSL~4m6 zmXuzr#v@D0@|$_WVr9cvx^^PZ$_;KCXP~O=`1ApPdo9xf@w~;zfUNa}tvHWQqdPblB#~N{5{qj_ z{S~M#y9PI{6c+#ejOFcjGWHL?Jcaj=tQra8d5iTLax#VEg^jCGU5qAxBvOkjv2?bY zh~LFV&P=Jf?yv92P6rrE-d(Z=yCBjv7>$>~3qd?@8S|5T-=FZXWDK&TECorVmKb7* z*W3|>>eBY5XMJHQ<}98Y$S;A}X9YMTi*6=}=Pje=UBKV>$86BkK$f>3KoY5?oLF+F zJjTB_UNgYFsj#HjGIi;gKu*#hbIuZpwj+@D-y-ZlHSM0JVF1o6CO*ETY@G#VP-6Iov6f+SLl5wWNr zJh=*8V|!QrR1+4@gN&tTS7Mo3WpE8yj%@((yyeGkGG3B`F5wf~b#{OxQi~I@)X#cR zj4bsN5;_Zu%^}7T@sd1W?i8tKAGw^oz!5-=zLcXY2i0+L8Ao!{ZHKi#PhUgy`F zsI4n3fewsi&Hys!Z+Ifl9a+ZC0`a^hWeS;F4%^gt6{?Hh0+2*%nL{k~-C9mXb+PVW zF+y0J9U05&R5FJ->V9e(vaA{o;(1GGTQXiwJ74XCEd9oUBvQ*kV$myYi{~(=1ACKF4f$Cc^GvbPk(^Op6;$eydvX)Vo>rAZY?BDI7OOJP;7uBa|2)s9RNmbAl6 zUA~VZC%NocYupc6Zg&OoyrroRdEeiy?${Su`e=eAQcD4`)E#v-3^f6F8a`{Tuv~Vk z^LO3peSF9lvi3n&3Q%`?0=|bgiMkaHb14f+|pVdZ8xB^l!V<|M&wZ9De!=b*CE-0%`xb zJAKFd1l*l&d2iMc)Sccy4}LP@%-Jh(+~1g z8&gdkuQJq#14(56mul+d|FZXZ>MdkxIP2hIVevk~SZ)TAGm12)=?+Gg8b%t)`TBT_Y^{oF!mCsf%rAFJn}fSXU6wTeK|6+ic~m z?K6>O;AxOVY6&5huNsSNkwt4u(I#PuKgQH0){FekRZ`*AXUJk`1>$*2+m2)<%lLg? zd?s-hTaZL*$t9NLJYC!g+30TAPGJdlW-Mt%WS6IY$VdJu`bcK;ez4X$YoGf*-#=ij*XR9N*RZdB?X|B-=ol;VnC2ub zm7L|56L}lAQZJ-FvZVNdc;4c;nXFtH{o!>0vJ?b^BvQ*xVrf0)`C?>QGQZg&VR`Gx z)J46o5x!*()JZc%mKTdbJa1WRL5{I`7ZY(uM7vcWiPREKEH;bG@lGyg`7gbMCEkm% zD7PeUN(vm;Yaz@2FCd<`JfM4e^?iOt57njn50FG^$src)SElb!UEEzx`U^{lH)C0n zKxP^8UzhjC7B3LbTdeKKO0QC*fJkH+5de}%EiLl#xqtCM0REnmvj2u)VR1gnSjuH z<2zn~x8V(l=PmWOlkUO5kEwXKpylsD5~*byv8=F9!CeI3Cskh(mNXxxF3mrZvE|4S zTKIfF3&Y$7tf-DwaK@zDYlUS5rjleTmqgS7Z6&7D##?sn`Sd5(m9-?!< z*G3S}TLu)9j)>OrKkzrZ61IROQj79?eD15Y_QI=gf1F!;S6FQQ7)#7HGQ(IF(QPoY zR7HSz-s1n5T*sClvav*USsDY9NG;=urPKYq;pp7odwJL+VaekxNjap8V7{M{AF_m0Fp>8=EPE%-|q^t+?iaRCM-$*OkH-Uug3TN#4iS~kmcG; z5YJnB=R4!KaZZ=J6(URc0+2*%2_lw)?KANVqki+duY@HufUz`rXM!yU&e$A4mZ0_^ zp0~WceGpsR?_F_67OgHIiPVx#EEflN{(sVIrk55rujNXBG z-m>;RS>4-K$aP!K|F7< zs!6V6BfD7OZxFqn1d>QC{=_mgNpT&r+}t1eTUY{5GM1dFWOuE}tcc77cO@w{bq zGU@0(P&CaO*Tn}Uky=uS@o(3OI{?Co)bBZC;5xAq#ecc-}HD zi>&tWSZ$4W*;sV|B#~N%5KCfe-(kq2>#x>KSdxR8x~RsIc^HHGI;&7!#+?T7yv0kO ztR}oXvAH@r#tvTqNu-w5)RJC=-z+K}Eodz);ini&Z@NNa-bx+3lKA&=5YJmWDv@U~ z<=va{jv=LIKoY6Nn^=yn)WzRb@ZTAsCM+JO8B4dtWHsT1C1t^=E_wSvJa0Kbzj2r9 zJE<3{i-RjjBDLHnmivz{k448=p@U5~VJYV4db3bHe-g zue}G7NG<(7;_KKYhjz)xvaO`a5MhZu!&vfnk-3YFZEBrEbs1g(;(5!~TjYG%eC`H* zFIB5X(?35#YFSP!=hsFiAj_`7FGmVX&{@VZ$B@h^%{55Ht8Y_BgLvMuegLsVe_S~Q zA7hh25~<}dvCQ5-#tK>9_C7H|SRBtW7FF{lc+7vo$)*I=rT&#WNC2`#PgO*;iN7HI&CaMmW|OM ziPZ9)Sf&Qe?}+Ns?cl`O!jclo)FrPr`9{PuB{O?u(ewoIyv64PnTK(?n1jCy*T5en zky?6u!smYQq?7m^-^qvdb%iA&jIr1%k#`nmQ+#S7%bj!(&s&18lQI9ZR$p-!!T2nY zL~2{ZfjwX42NzQTmiq03MdLW6^ za+6pd26p|vVHJeR~M90|ti6EZ0*dHa&`vZq;$LGtX=^%;J zQcNs$b`R>0EbGepm`_Qn4QAne2KBtaVGO)*Zhy+^JQqg8-L|`+@+TGt>mb%_(n38mHWuKZ!|4)C$fzC4dQuAUKkm}EGZ1YYp$Yeef{$zq?TF4 z(tk^Q5<14V42=#H7TYMs;ysJZ!)W>_;QPzTu^^teJT)ZiVdf-WOF(t`t_hMzEw;om z)O87-gPrlf_O!6%ahAnn$x7n5-^qB*m2Mh{=PiX{E_j~*o^@>_WO@DyB#~M!5zF|Y zez?2iVByU3!jg1_smqiZWbR_zmXRCLG1lNai03UQPmy`c0W0;EBTL+GkVI<9BbLd{ zU$;ib*x%!u8m9f06N$xLiyY$dUmax4bp0||sCuU$l>`)=H zv>m+hp0H$GW9ssDDVZA|ea{!4`}w0lJa4&Bi_9nQfAtl=OL{#CB#~M|iKVpsz+Gg~ zsT`IfEHT#^i|=AGpIpyy1O6stPVM1;;(5#ZM&!O9n$h(*s*6@5kVI<9B9_3&g?Pu1 z6@3~%7nZ;n#uBMW#+C|As_>d#?V%u^w|x6V&V7w-PfL(x#u$)9YH9KXpZlk~E8@An zX^t7Mg=NnT#$x${jA6$2)xvX1Z4QHY-m-}98c;Xx%SdEV@c~JsmWjmTtfhn3j^1<( z%oCOZ&f?LO9AoD;48>hH3lD>M-cq`Q9AgcfE;d1yw4)%2)M802UhVsNq4ULNvgs#b zNxsR{MO~9z$0nV>jdus`JQKw8mRo1ZHM!mqYdlVFvj8NKT22wmcLkGA$Wm1@?5nVZ z-(oCd)5u)k)~D;Ap=jcmRTq$@YXL|iwTvc~6<2fc&fsrM-0N8U z^&UAhj;V{4F*(LYsp}-6x^&nK;(5#ZzjfTTXgUm8j@W=CQp*NnY4~0VukGow(x8#B z#NK8sU*41X&Y&GKQv~iBTL){5YJoYv>33*F{ZZ#B#~O45KCcguWDq8uU*nsSRCUS zi(##e`2EXswUXAza!MJ*^A^(_vh&fmZl@g4F?O*PNFue=D#7RemHHldPDy2IQb%Da z_5FDwy>jKwj8 zT$6k4zJ=eve3}p9dCO-7GI#OKENCOLL>qu4Qi~6<)E=4=g)ChccN!)v-gg;G&!c2) znX!Apa#WWpBM{G9dKr;j1EMz1)Ib(rQ;7bfr20Q~+AFrc`UZ8{Z@+(2l9d+8UHxG5%>sWwH|7)lH-+rKS!q$1H)4shmNc-PA?G0QrNvFMEoyDlr z{;?fM_&;^pr|X}?HTADC2*25U?g(nMvH~T`2k}YwVtesb_~_RD~h@g0P(zKw-;GWs5*JiW@LFk93+ui5{aeH zgofSFUG0Wb@0G$*%vq}5l2M@55}o#_F6Z1qJa3t`gN&SZT^yE+EEPT=iPTa~ET*Sb zmLf}LY3W*FNlRwxvbqs@(s;ft8}AkU<~NAvEf4+4-h?)hKk@f1^=g;>`4Lh}uW$JB z)$!)JSI9Cu`TizhiGIjf#(p3(jQTG^;(5!*L~@MXcFYV%$5@*h9)Es>)Z#)c z-`#C-7xYozr8|Yi_Az4_wun4w3?Ju?SK*ikgLvMuxenBVJ`?pVC&qv3Y%?=RHTTBjB3@v+o zN5@#zx6cQJB`Jle%j(8tRxYiD0^VD_+iwuhTSn8>9=XolTcW!3tzGozM@TJQO7Xd` zIPIqnvh-Yh^{}vnrZN_t3FK|ug>C+$QC<410`a`%_y;nRRobE1ky;#yW%c8j z0%Yl0M=MZRN;%8*$^&><*6PmY3#cw?jX^wbX`@Z*a{KJ-y2$dV6-XkrTqBl4L!$8h z{Y&CnoEDahr%YYCl#{1ftKB8rkY)LO5YJmQyOG(Si)-KERcqH%K@zE@kXQn3Q}FI5 zCec}8!V;6lSfWpmbN^RlZ~?L`i39Pxvj-;(3e0M{=)BnNWTg9b?)qAc@p6hggimv_B%t z%7*sWg=J4VV;QMUX0oyutE8d2D7OUhyk+=Ma<5eFuov$ao7)~Fky`ADMgPJXO=PK8 zL+iG%6mS;9?c{CT1r0?nWU*BO@x0}#G8vCe%j|<^^A@c|QC=ZU4d<1UL6MFc}rSL(y3Cr+!%jX!E`K0BDK6Emgq*S(~u=$^#}I_2=WMhf!Trhkztf%O+w8*KS6Sv35O#B|e9-j2l4e z(mXQgH?mBa4B~l9@G4T5&-oLtBTLB)kVI-ZK`cjnn?FQ#xn7vlPgp{78OyI8WTn@4 zqjvba3P;O8Ja17LLdMC%!I5~-!mPkipLt-TfR+WaJ8^Jrly<}77h$edEc-QV`3 zW6bjri03U{mgE>aW?L1DEGFq7iPSQJSl$#hz$=N19mY-)mb7oI5}#Oy%nAfC6}yhq;g zjdxURh%C0NKoY4XiC7-CJYR||Pd0jO7nV}a(tV2+wj>;NIF9P_a2tr{E$X49Lv>-R z57o$0Webu>EfvJl<5VZy33)kksjaYN6f$+GG+u=*jgwotAhwrH>!&d3QJ%SV;OOfJXd~QKkOx{%awX1f8u$IiZOX-F=wSF z-d(Fl6Ocq|aU&M>RTkl>E{CoqdI-y&FN|fD51H%R5a*2N`CadTc-}I14B0iH-qz3$ zs4fd1fFx4OZDKJjQ3yj8m1&2L2}=QIsW*?@Uvw-cJRfq33B@d6oJ-ZXo84OxO$ zgCtT*w@Q5OAH7izpZlicI-U`h@Dj$ddj|OqeC@MlgOFwVG!V~Q_Ggg!u{n;1@jHva z*&vD3vXEG=Dw*P4HnyDld_h<|zA~2AL&-IHifYnUWU;#d;(5z-Lt?S;y@{`5pQAt$ zsl|y{_SF1{Eq-gFqJ*WKvzYHP#n)tAtuXwh7Q;;-p0^zNMaKNq+8^!FF?MbnNFuew z5KHgA(Y4UI?=xw~4PnXr#?+;~2^lBPQdv@t>JqFC;(1GBWip@K>D53yZ#i5CB#~N* zh{fqnUw33VV=_HnSYk^VOLQ?=ccDJOB?4IvTY`Ar@-mxPZa16r1l47r4M-xjsQ&=mGWwJIi^bPBHIQY@F%ZvN+Ut}1%kh<)4Uy&536Ml;nMW+& zAGTCL7Q65lslwv;ov|2Xk}?0oFA*8SQpj0~D#^Wa^NY>+UU|G6#Pb%fb!2_AQ9%e{KUV*d04{-*tbHz1z3G@clY=f{kUU*TOg2E7MKq?Vn;Qgqn2BeD$2 z3#bs5w-roX^4gPK&)!e(d=Xufl|F-b-tzGdImRsOHs6lwGNlY8ky^ruC8u8@ewU=# z*;H}!U+-Sye=(L`Gs*kPb+fVzk!5Bd5YJoc7Lj?F7I~fVF0^k3gCtT*4zcXLbS@3m zWmET|^@Js)lCfyUl6Og(b7F&$rPV+X&s%O^AoF8hhxBKnx3gzZuK*4&=VSqgOsY_rvyqc-}I0A-N_`n3OsRSc-Wb zg(bS0vCOe1>&KRK`hj;m)Y=Z>c}vwnGC!6&r>iQm+_D8pq?Sx#u{2B=h%5~n5AG!_ zzP9z4ugofu`~Lklariy*BNY(OTXHqXnyZ;rTDHiNt_G4wEp-*9X(=_R-=MC=N<&ja za|^94SLz~*=fqzFgvHj5v9xGQp7)z}?-GiRu>mtcJa3t6Kz62Bdi2E!WVtaPB#~Ow ziDiLWqikd;UmrhQSn@c_%i&}v{iHRCD#$W%D~RVUN!>`l*mL{I_}rhe3nYGNs z%>lW{GB4F-tgs~6Gj(}aLf*apoKdqmvOIqY;(1Hn4q|!KE5{erWo-^fBDKsTmXNR$ zhfx#ogWfZy35)0MdjHg&9{QL($vpbBVLa+i4;=Q1je3#9r#b*CFR*>phN>5CVD zl=ANM^BHQGIWKHR7CNFl7lTCNi27T1`uRbhaFS2mZ=XWl>DAjnl7HEqKH*H;7pObE zu`NjU_wMvq=Z1LT^JH`E4|HJm-Ukx=LwEYtManpnlhWm8xI6tI$n?KD4Y|jHLLNPc;oXRiPREBED`5EJVq8plNzgpCB}iVn9j1smjIuKTk$H-_ZA?Y zw^%i@h0$=PhG5 zlBe18PJZ|a@ar~^L~2n~#7`P=qk7}L3BQjw*)A*voJDmksmsXD{`d*->~s*%TW;+q z*Rk*0UOJ(=1kMFXq?WeC5;`O0Iy%Po%^qqiEXn(sx_JH~PZ~|!bZ?=$oLmUvd5cmS zdFx};>;vwC-oG3qky-{5%LlF6HIQZQiC+%F67IxU@)wX*IC0bG9zd3z=RiDf$$w5( zd8#$OfxicveF-FyTC|B}>Ac_g$$Z?f+Xsck;{apPZA`DpU&l{DmS5vQJa4hzLEd1c zpX!CbH@;{pNFueYA{M<v0OQthZm!O5D>*kI2SZib{*azZyi{le=ue8?C zJBlm;t{{oj;zcacx)+}y%Mru&Cxj*TAY5|%>F(snetCRb0&!aK=r_yOX1%TNV!e`%w!Yyq;|uL4P= zmS4o8-^CSocl1)6eM4AM+?cvpEhS@^=r(uo%KN%XAfC7M-$!PCPsjDf9d~z|fh1B( ziyC+gV|`+=13Jb6=c>dDOT;0@vbdUbbw0b>P8}U%&lZ7r-g5RWS+&;kRk=2*%YYRi ziPX}USUfhW&PJ9=30e1r#rrU0*`z|c({pd8;#D}Q4?#R{>Gp{nWAW}g6p+R78Au|v zXc9~BtNB}y#pz*4s<2dYmJJc)EotGQ$ZM!B=hHzvZ>f_{?l0Mgtne|`DGMZ#T9y&Z zk0$pgBa7i+^9*5m>(120we=$WEVjwt7VqJ0G#bS7mi`OL`Lbl??;=!}nv+2isbw3n zWR6YUjVu9r8gGOp-h;918%tJ4EjeHJAgarl zosmVc|KV@KQp{N@^2l@L2TybSCVuA#5YJmY!pU=GK#SRU<^A^YAc@qHNi2H(#^aqS z><_N55SBDArY?(=$ZGY};#Ak+bW5R2OA`kVI-JBbK6C#(1}& z8w-XgZux7qN3=I%xqpZ36}@M|C_Lt0xfI0nmTg1Gb7k+!tF4ixY!yf%wJ6ub_x*-R z9(Xlje%;@7g~j(MWAUs(R(n{EtI$Suxz+^4^A;;zGH-d`>gH5r8QdBqky^SFi+ZD{ zebM<+7?-3hEVe$3rTs>7jGg<^19uUG-Uso#CAb|K!xUGT??o1qCm@N`GM-qhjDJNT z%kimhErlhIv(#5M#`j8>V~x9@x;&Z#;(5!a=j846dYv>p2b;AJB#~P5iKYJqw|xBN&U_x(vjYU0&VnQK52sl}XFR&~C26ka&T8&${v&*Gs=RfhhWkpXiZ>j#~sslR4CMbaMCuQ~Nr3 ztgyrcFc$yF)%X}QtXQmz>SFi{#PgQCCS(kwSeVxfS!&gs^5;iLEnkSm`SN-jWO?V+ zdWx_F1~L}gVsfu+@x9_YvUEQV;(5!im*igAz9)=3tBgBUnw|ymyybcya*X{-D_M^$4=;iwQj00E9Q<*n8LEr3QSCLt;t|AH z_8XA7zF+Ts4xw{@-FgtuTRI;h^Dq}jy}{odO}7L|q?WzJvZ&HM7FkxTeYjp&$~lX> z7nx6fGV2E+AmCuHP&Lu+A)J;hk;wa6O6 zD?Umds4kw1K|F8C`#|Sm?w-Q)mSc=S5~(GHSRRZSy$;o7#Py!`!V+|vu`H$U#Rj;N zZ`<~Y2l2e6$qRB#p73JZWKm@_|@1=Z?ZVQGU~_vtL*oLl}!oFd4%fX#Z^v zs!Q5(5YJn}tzGeWEU4!T{Jrt<#vqB*QlmD$?_0MVj<3mz>mR!cOCe|RHz0F;U3|Z~ zB1^3;AfC6FWs`G%P`V4A=RdLoB#~O$5zGAKXZh$DE6+LNDJ&^xn7Zt-Bl9p>i|*lX z+NXX3@x0~sH*$>4{Z=vp9b?l=K@zED2(jqazVr-P)+;Xe6PAdxjHP>BGS_E2u>yZP zeO?0)&s%2DdCSPehq}m;(G(<+T4oZ9(OOmfjNIO)Tad7LpJOZ@^+}i7?A_|PF3x*F zJa5^un9RdmC`=DRmT?C`5~*c1v0Qp|26qr`d;ICFuvBuE5r$+(Du>D?cpfGy1H|)| zR1;zuS>qD!>YV%*B#~Nn5ldE!&bXsHtX1?yVR;+M)FpfJ2K+3hXc+eko%;{fKs;|5 z97*P3j2hp?V@reXAc@rCO)NJZzI;RH%Pd3NXkm#DV=O<-$$PQ5VbT4O<@Z_;&s(0% zAoqQ{)k}AwW9)_*NFuddBo-S>J=~!>M`zkCVF@|USTZ^ji;LRHcgSM21jO@}qq|7g zjdxBM-YLv^6-Xkr+$R>ZZaTG5T^=`Sl_)IE7Z{83L9#yCv)p<$vKY1j@w}ztEqO*( zQ>@5Gb@`@i|VQ5`TF zSsvU3Nu-uaVwuqG2Y&Z@df}y~!jcxw)Wxm^xmW%;STPIL<+CA(=Pmo6lINGw#FKN7 z#ceG}BDJ)vgU2v6Y<@IBbqTn?^QEvvUt}yA9^^V!QW||0SuQ^X@w{cj5i-~JKC2Y( zN;BjYNFufLBbK@CSKzVb&HYodg~j(0W9d8J5MPsxbuBj{i>W(^=PlZ`$k?*hK=Yoc zF3pdDBvQ*%Vwss09g6Cben7cESZpsdmbivw_xbnn`WDENd&aVaQVb2_%tP77>en z;r6A-V*TUhAYn%3M*b=jrXsyY1NgiA8S9l-$rB^Sq~(US}cfV zXvXCtWVx&Ds39!jF^t7ti9Cy$^<2;nSvu;1c;1q#OU5uw5~|lB%Y`K%iPYjuEN?!V zy+94k$wlwR{r|fB$MO%~f$v#PW-zzJTRumf_6y2ECP}Bgal``LX+I!y`Bl_uUtSHe z{I^bfmxH9!K0zn95$d!bSU3L<@1HvDYjh{0=;f322cS;-&y7H;|5K-ZXiVd4sMB7# z6-X)Xv>zWtI_*zwU)&8HQ92z!B5_3ht<&BxxDXGm>pAzyN1gT)yMQGBveUkO!r?Ed z)BZ{?knHcB_L&P?;Qn~iFeNW^U``td68uA_efEe}IFp6xL^srFKYTdI^uKo6|Lq5M z+`3E~b=upF18M(zr+w+oBHU>oUFK$uI_=wOf`tE5r@dCxHC)pZajGlORig7ukW+e< za3@!Zwp%A0MVN))E+sWdJ<8gzzdfv%ciAi^>n|?vZ`pU~7 zp1*p2(I=z8T?;j8p#R13I!Ge@UxLa1VlqMLF0xDy2$>-)<(#GBBUvGFc55&mo!<`u z@x0|jF_~dJsJjQRSoH}9Nu-vW#IkQ`{Z7b|x6W*yuw>q3>XO}-i~>VzFWigj5@G@3 zdCR^7WR_u~RX_YDUfmibky@S;%kNn4<;bGAVYI%m#NJ{o|M?6BKE@0uU&kFpv0Feq zZ+TioMuAHPx#7;PfjdDGsil}$5{o|KZw_}DQ){`f1jRCza1S!W7{5OI45~{VO%Tsp z?%yD@rLQiyrK9sjaTZ7-wbZYNpES1CtbsdJt$(~&Ei8_4jK$ix+%Q; z6VF@TEGHwU_Onjn@AO`%3zA4JortA_a)U4E7<11KUN0<#oTaA~>AE>?J6;J{Mt%hG zyrqF1*2>@Bc&7v;ky=I(ONNeaV`TYJdxNF0q}*oevc;0@d)@fHC4SPVQU>w7 zrOPs6sY<^Qh>o!>tw0i~MTc0*N14w>mc~zpS_@0W9mb+Ty9nmB8Q_QN()JaI=Ph$> z$o=Kw`bRFv5}pH+NGpkOa7-tmlO)hF_#u`l!1bY%JX5XAEq{jtPi zsaWcaERN4W5~;<3Sdu3nK8Y;Px+XaaOC@L7Xl{b9$yzrr;Z-=l`+<1gVzG~0$7T*b zg*%Ashk+zgO8~KSifW7ZSzL3?#Z_3|CNOoGc!|8dcKPL)j_R`H4v6P1CP&EqCFanV z)u=8G4?q&BC7M`T^)WONdl1MG7#9~;tYX@X09j5LlEFpIp%Oef)Ej7Da9eN{6^*9jETWZ9S?c z8jg;!X8l1Dsii%!*i3!i2w65oIbIZ&w0lfl+}n^{Hi`_!<9!x4+ye2u{#5AfC7MRv|0B6!)n% zL6(a{K@zED7O_mv3B_-)8uw~@OIUm#FqU3J$<7pK2AwuRmV!Ybp0}*LMLLJSD7L{J zMDIs}BvQ*7VmV-@i@y)!J2X2%SZtFSOG8D^+C49}D7n%g%H%PTst%$`RG2 z##E3*YOy7jo?m<`&@tBKX-KlLI^(wsk8(j0 zspTlK=+&7u4q5!`SUwe&q=!sh^5V#heBGp;c-ON!O+Y+vDbqiN-^QIu@7@K~<*Euu zBDGv1mZvp_eMT0~x}#qTOXwrU(rg`B8M#*{dI@=5|9Vtd5d*4c}qHDtS6u-BIH%AfC5OY(=hPt1Tu(pmYCtACN?9 zF(j6qijlg=;#~fqzOWQQa9YnYWz2a{*o%`E3k{=Pi*%WHn*k-LLPVx~!Z6l1MG? z#4;=7WG=Fl%~+-)EaA@>%RC#>)!E!z$r9DYT?fSTmSY>poA?vI9q|h9*NZ_CsU?(H zsynyXj4VU6yQ>O|M>=B(uORpRzP2rHB1@JIi03VxYLM&Lm`jJhqq-C=21%rr1ZufF zvo*3j8~?3~u#|Ha=LO^#(;nFh&yAm-4&r&s&J)B^l&{?#Sys&jNu-u6Vkv*3eHmFM zc8cjOESb-lx;%>_&&YaK?YpD8H17@KdCRgllx8+*@Bmr9+JJc8vZD>@w!Q9NoPaFKdqEPZrHK;0@7MZieh^tS zyk=<#OVA6((qt>?PJiRqDhOGu13^4*8GVe*UFd(fi=V}ogn%ScOHX3a^v+gBmJdC; zju#fkmy9KFF&V>jd++%SS@sPC@w_FsjND)D>I@iz>hf&_NFudNB$np=>f$-2cTt}- zg{6?Qn71VN%3G^`yhoPATOgjdXc&-lUs18fW@OoU7bKBd42Y$w(ON51m$_@N%oLWC zS4>^}50SMg1@CWlLY8CZAfC6J?|%-D$6nrShIb9nvjRz^7E5B8RAg?9EM)`ib%Z4% zld;UaP3B=rbC+8_#rrj5 z88Mr5+Ix5ou11ztZ$Lb6Q7<9SVz=B*^hK7qe2_$HIYlfbPSf!Uyyo9ptPqw;&eG40 zd>hK|Ri}H%l3oeodCPDsG9EkQcxe=}nAU3f=SN5_w}>U&r`If0mr1p8&-dF9WEtENB#~M&h$XmYKn$`t83b(bmOKi!V>b9u`Es{mJ#MY zYmp^?D~RVUo41np$b0py@Gi86c7Y^POM`~^zF%A~5_fd#IBV<>7UyinQq^V!eqR}P zz3&TT8S4q+c}vF*IGQ)}Wc96ay?Nu-v}#Imr_p!KLO3(XYv2um?%32skzruda; zGZtA+TY-4q()j?{F~l=o=QTQCHrj(EQp;#!89l{Z4OtHLe(5AEX*oi;@^NLZqC8H@HtvQEjm^$A~8 zm)W&W{E6o+xxdNW#aX3gJ&>hJ2_%tPHV})kfod?S%aW68y@ka$kFjLcAv;j(mU!ZK z7CE6Hp0@;DA@|BU8k%@KW*GsJNG{>ZZ7#NYs7v3@Cb?T1422|hxl1MFAi6z^uiz2FvSzcVIuq5R(b=jgw){ogXaBPJv$)7^ql4q2REf_UDt zz6&{D^d8?ih|c|9IUtGD@`+f!1fIk@xD~&ge@$523K&btbkf;%sW{pPSt@fuJZ~Ad zhFH|EYs^QMfrTK6)KaSvzVCNj8Q&7sC8l1xIAJN}EMY6jx!-G>`Zi=)&JloI>XL$M4a_&tlPwKs;|ri=ku7 z$tw8V-@gJRky?fkOM~VM7b45J(Cd$cCFUbz(ceJsFP{GEaJTKmND$9kMlT>^{%8Fi z?a?uo5d)G)EwhQ`cej}X9~-nLdK$XoqXH2j_ZdoWSN@@;(3edda@R-Y-e-) zjxRhHB#~O|h~?|R^S;QEp4Tc@SPD4HoxWt8tk!25erFN;8N~CJ<(@5IoPX8hoZ$Si*}KOOB-xeiqAVtckB<^Us5L-qP2PT*n^PdS;0%?V~^v zsU?|Ms++B?gO0J@FE;)V7LPBCWm-?NT;~_=3Hcn{{WIm zE$@iMq~##|PJGevVO7FX&ROzO$lSQ&u`0Z)X2>}Z&s%b?k}hbK%i~`mOVVYKL~5xb zmi__b@mHDx-YeAJ{?~it%wnc4G0|jBX+&A;+o&#$OF%qtnQ2MhSq!^1cN((H{0Wjs zEh@@*45M#UfzSQR+wLj}OKb^a`EZq-FH66>86e9YI}p!X>ZOwN#doO--pygjevm|J z89*!!9%d%!eCd+n)=XG}zA_e@VDhC&-MZ};Aj|3(AfC6l9VYKy$C~w=it3X479^2c zrW4B|g7#c%K0Hp1fgjj?p`AYV0WF+UYwldW7pJa3utncQFM+go%(mK~lTiPW-! zSn8{Jmm!OJe1COeDda3V^T<4Zx93}NU7{a?c-~^&nXI)==wBlgSx!6yNu(BQVo~UE z3Gdvh^yNo4VM!@v>Jl@Mtb=`eZS*5_jCub6@w~}=nVAmac}v1O@*deFLANuiOX6#gL~02mmYUOd z8=xlO!6&_j3yb~tf9y`*(Sy9{SN)z9kGj*tLqR4Q28B0kZtJ z?)0-hQ3Jp1JeCdcY1{tc>_4R&_)?`r3{oyaqaG0 zMBV9kOF`QI-ksj6Mmg?I_n58L5_P9f{sj{LPu=O!`)}i#s#n)9LsyB8HNF3w-mw`4^t~*_^b2;A2bBOOi4ROWs{?8TXZ{LAWI!aa(uA8!>0Xo)Q zgF!rh^$ed$z6HHG*UlCFFH^%n650QvkVyWQHxC|aqdRo{oYHZ^Qps7qd?!zUW>H^@ zkwxhfi03Wpw&Y18ab#i$vZ#FnNu-uH#PYuORQ#qS?b}UFVR`$5sf+3|^43S=R(Je1 zuKsTj&s%bmNvHj)oBIlo#i`bpTa>jD&E*Wifozigz zs>>o@5YJmC#FAO;k!ydRMwZf(Ac@q{v@w3tcy#sc4P;r`WR{Mwg#2VIRU^q0VB*D= zW00j=8xYT1M&BgoOMBg?9?du_nj{!&`wM-(G86!7ppt>A+_SsNaiaE>lL1g8<-u_NIk;Nkm z#PgQoe6sRhW0k=VWNDcXl1MF!iRHb8cTHs3dM(OWSkitmbunv5RuxQo=z_0f_2WT2 zZ|TsB%=|X~V&{e|Cm(<$Qp;vyIXl%o6Ipg^+iw(>=t{;?xPgrGdhP2v5!I#J0uawz zEWVKI*xvcd1Ciy~Vvs~?aUqtXqwDa!(l>j?7Gd%I%~&QJB&!Ldg4*D{)|P(+@w}zX zklbIIIMl#zi;| zaYv5~bc|g+2$D!GFNmdm%Xg~C@}l6BldvRJGj$oSPiB8~X4Q&E=gYwo5YJnho+7h9 zp69dfAj^uMAc@rSl~}qY?CFXu@m)6_5|&Wg`hOo=7G{yR*D==v@XT*nKM>Db8ebr9 zeJV%ooQN!;!$1GQ2KeoQ>0*#XYUx5Omz@HyBTJ(L3IW1W%2^&*k=5#*wI&ZmmXGcrp0_OQLB=p2 zrWoPxkgPcdl1MFMh-Kn~=Hrp&dZVOZVac#(>Z0AByuG&6H@%N6S6hL2-jZBG>auUZ zj!1pfNG-a=a=3fbN63=A%s*6EVs<5xaE!T*px~5J9vYauwc~@AH9hkb@SVG=lrK(=VtD*xeK|F8ysZ6@lKPsI_Lzeb7 zAc@rSlvphD$ThjAr^6#*3E#(9612(pV85(BjaL+v1%PwpUfkY!>QkVI-( zODw^@Q-V-kbe~WEAuNv0j3vO1jQLA8crHVADSiXud5fYwIrp8%FT(fz)9*nNsl}dH zR7|%|K^CQMEvtm3khApABP&*W4-Llmm!7_!;@kAP~=6a$6G1!qG1H+qU;cf+SK)1hE+X z9ubEucVC`R5|#)T#uAuF=EpoEm*VaYhg=ZPTR!e1D^}fSEPjU0m-e4P5~<}OvE&|Y zz6e=fs;_S*EZ(k+<;OcRPJU3gs{%S-ZuJE5yyePh(xKX5@qWB=#d{z~BDLfb%N&PR zlaOW9++l5mrINGkP$1`i>*H5(XIFwgi03V-^~rp4gypRpQauL7x1e~7W9xs$HWoUZNhZgeZwfOy{WE`?aKO$^r~i`PbwL~0pGESul-Rzj9p zwyu4JCFC$;=~hHm;aqg;>W$79@BJX2w=~{GcGXn8>+lF!@(+O|Qi~R`9GDh`S9z9Y z=nfGUXLrUj=onc^oYrh8-YM*2Hi+jfUs{s7tY*3_l}3Hv#dy#rP)~!)RK)$M0W$Sb!u_OS9(qzJIIT>ZYhJ-CAE-CoG{y8B0+% zndeUleTOYh9YH*A>Dr$hV+Lx*c%1yMCrBc-^d^>CRu1_4mJWBV%!I|whp{;NkU7}) zJG$0D$Jn`45YJmG&ys8Mn@_#kp<^uX1xO;bOePky!M&4ET^z4#ZWESL&SH~Gu4BWq zRk|RHY66JoEdz&=V=UVJFy2Z3VKPV}wJaf)6HoP)Bg?XSs=I_G;}}yHpZ?_i{;eCG zw;;>grXZfT%-?VZciSG$EQ&&PS<(h1ky^G8%l^U=+}Raep1)UEVtg4(?<_Lcm%p$O z&nY=5zW)=?TOLM`XXGNUZVJev)c_=sT3m_cNN_2hw`{3#*;!Zu{TRy%CGsBGGN3=+ zN#B1Wi03VRClX6)vrPP5>pof_iPRE8ECVWY_n>3!UQHW!VcFx)Sk9V|cNPubPqanH z*tUisp0}(YN8XD~ob86cIo!VmNFuew5zD~oMR?b<^)3^Agr$J94ALe$w=T9iw;oxv zyg)o}@$n$qCYspTcHY+pCl3Dw2mVZ-CXk{rO)Wy%J!9>#gnQvA;1 z^D_|7Tb^$wbA3mAR-8i?@7ExS)bfp3<~C^c4Oy;#Njohp;em{0@Otu}-sz=r z4Ovnqf_UEYHl7?~4@Ya_^<#coAc@q{l~^3k+u(Va=+0{*g{7RcWV|Q$%Cc@t@jBS! zeLy^K8PtYc#|%FQoJYr);t-HTY8gu`cSAd#Lv=Z)-Zw^AGEXpd`S4{CeiqBP*$2<{ z9qI|6N%aVa0iPWM;EK@qZ!fQxI4=BAYEU_mU%jp}Wvn#E13hshl zs{!J9OI{fn!wkH70`EkbKOQ8JS~d~Oi6R!|PvpMeIRhk-TJ{smwVXxC=op(h!y`pl9D^B)-8M2eKDBVC6FT=hECBJm zrSUtm9_CcUY`nJTp|JL~5zi65sbv?_Ky7Su$4IeifEV&a(UhnNRjeU4hRRD^n29TaIRs>sXbC z3VwbuvII$_7Byng?_e5*ESttp|0yhQ&oFiIHz4oC#~%BI`z(%32l2e6$2c-~G4osY zZd8{>b3qcRWdyN2omB_#;&k1xnS%9S?_T52GL|i)$Sxb(Tixo4&X*-7AfC5mT_$tm zxkgoC$nxGCB#~O?5=+10r+Cb-{^U&^VF@|MSnijQy3A=e>maghcm?8li|!+`J~?Lp z!CT1EC>JD=TGkOuo9(;sT5JCvryB{2b0}kRIY8FKSZw{2fGo8Kfq35X?i@K^nwaeE ziY!+(KoY5CH?gc+Zh_w;_cpa@AuPq5<*f!8Cx6)Dmxn9?n?XEp@&C^j5?V?P>Nlus zvC`1g(A;8K=C&Tl(tHOnUo!gegExfRvrMG=VS zEf3p}xeK-69ko$iR(}Ucq?Sj-QeAiwujxHsoYq5Fd@nE-hwa3Y9hP(jS$6#d@w{cz z9Wu6@QLL~GS^6nH`|~5DmiNSRu+{+!)X*GXRCl1TT=|dXAHD-$e2I)^7Q5XUfI96r zl!Hu?PW#>mr{PZfiH4pNQK!9*V!c1V=U;c)ZyGw!40YNcst?lrQ>XotHDrc1@4%&o zsMG#t_(boZo$Ov>qo$Le>i6iQ7o%X55q|?6O zZU;NmX|L1^B>9(}_R;TqTcb|u)j{}+h*G~Jt{lIpr&5ol^d%vk5?SJpIKeDODWOUaU`)OER)M;Nj3ncuX zI_(z~JirHrajsn+x=Og~f}FCeMC@mBl~`3A`2uy?Z(0J<{JT#3CuJ+~Xv2Nj{9)pX zV;jy~J&hH}=)A4E5xx_g8wcX~t7nD^SxIc5U^^E5FD6q#650QvpwJ3GX#}pFiAS^- z9=0AMEP0&8={ETm!LG~3$IyA(=Q)VyE#p^_yV{MB^R$p9G7BV;TG|kc&-(+l&>eb4 z=eLuDCFvqlmr--b$f?7ay?DRaA$>qRZ&{&DW*FTbT)_L1P96f1NG*ei#b*62Y}vZ{ zoR+YJUScdN=gG*)C#%+G&D*gESr)bf@x0};E3q`p?u}RB#B>Hpq!uG$=@UQd9y-Q+E^90l zmQv0VltjiD-m1pq50XeNJBa1({yq3j$+(>6%Y-E( zf~m{RS!9=uOSR(gcYCTofq33hlul;lZrxMGE4=r714*P7PhzRJbj=V{mvGmYtAr&c zlCi|jBwaVLsV^R*W30j(#Pb%jGP0^5w3%ZBs!N^YAc@p+fmr${ei?);+oA`U2}@uU zV+pMw_ZNpdr&=J({Otq(#PgQohvdmTz}F9VswnRUNu-t}VkucwZHX+=dTp(RWzQAH zvZO9q9W_wZsy4FhT@K=TOMWcbHNful2t0;pXAF`^E!o7<(&Z~&vD&1|TRUMX;4Bl4 zkrhRU4$s3gzkT|Gc-}JNPAGmdH=TVa79C^5hk_(hO9in!xi}Hm#bdxJM`1~hX6mx< zl{p@kEt@(`57os%1H|(dvnS-59IQOz98%~kLiW~E}*ZFGzcjkItTmhh{L z<*O~ZjxGAp6QBD-Zi0B;qP35VVXCz%@ckt>0VI)H`Vh+^-%j_DWpTkMPhs)6##sDH z$ZC(`YYM|rUFwv8c-~T_NUq68yn|CvU9OdbBvQ*1VrlcHp&hc6DA(~5mU7P0D}t;f zUQy8ocMk-J$5d z7u6-%2gLIhlc!|&lbvH);WzQ4Pk4jfk`ja^$2!^mQ{ z6~yzFT{X#g%zVjay!v*`E|5fODJ7P%-!9!mb!j-#Jy}>HVi`-lGP1&ZUX9&&M&4i& zi03T>j+2#M&MWTYw+l08fFx2&qt^JozpU{qe7|j5Ftr!9!(9ZY?}H>#OE+S|_#U0(`IC1*L6Lsom(r(fKNEa4+T zJa0L!MxI|>mjrx5=gaZO5Qr?dysy3&miRl2#jlc#lP@@Z#w&{M9tZKf zC2Sly#*X@qH$j%bGa!l7Vn!@Qoos3#OW&imMZyvi&sc0ul2vQ=4bQ(rb+I@B;(1G~ zi7S47*_-hWuUx4-1CmHBPQ)^INH;xXSvFa}Ojw)~7|UEovZCnytC@KBlj4~mp0`wn zkh(ma-ysgwC2IjlBDI_(mUh2X@Eq)w9UUu$rI@qy??~ogcm7U%i0bm(0mSo`F|lOc zva+KoUU|RD1tgJLVu;0IOyMb1mk-JxYuNm?LLx1ZsY}ILvZ_Ekqm>G>Olu9|d5fkt zS?#gje&{z;ml!pWL~40PENhP3#PcvG)vnYRmgu{TWsx#jNqqTSQWUb>=>g(-%PbW# zw!D415uQ_;J^&<ApsDk>!a6i03UUJjklG z+m`-QQC+TEgCtT*-8T5X-)EbeA+khnnyex$w)YrIh⪚JMFrUzunQ=3dHl4P!%#B zTivK1zK;3Xfh1CkIp?J^_5Ljg(Z)(WLlF>mDKO1MW`+dOhG(vS-FRd$2J+& z^+k30W&x5&EgHnKPQ%9oS)T8E)y`v(zJOn7NhMTiPREB zEU|Gj+at>($H$t&67z_$sQQ!V%8mtS>y8VE}PXKCX@-X+buu)H0zR5b?iyrm?Fbao9?nmZ1iFZWeI5~-yvv6#oq ze~v7V?Y|fbOL8hx7j;82&%bfVw0LC6Jpkf)OOXY6-nVy+#5*3=^8iVtmchg_z2ivy z&cbd?l(DdcKVd8p8l>xH$%{hV*|kv<#PgQj7sxw4&DO2oqq<5eb(OZ2aj$c%lDoj zp11T;Cc8V-=ak^x>1Xax?2{KCQYexBcuCFB)| z=Pl<=$Ty4fD2r97F6Qq*5~;<9SX$Vd;?shvxV_sdEU^z6OQ9|q$J+JE!Y3c4`G9!d zqEkTLUmR-~cSe?jhd~mlC6ri#&&|Re-rk{MJB1}MgRvZ#PrfC2cRu8g>SB8d#PgPj z0&?$#d)bfg0JUbLQpnE^nb)>Jb6mpi?nWR^tZHI3`$ntIN0ARG5%iJdGcX=GWcv|5YJnj9+K1N z!=_%xrv{WZ2T7!sCav)t=Iiz250ItD<-5m)CE_t->D7;{DQ)P~WFD%^sCW?1TaG4? z=f1(#yC+axJ|=@CQcF)_(X%n>j4UZS2SS9!_X%U!dxFgQM;vU4cMsms2l2e+_EWNM zxyk(XbYyut6(o^b#t_TWxmD*N%Oxw5^TJZjStgDqW3q{ziyo@Wq}3pvw@f!7-z(>} zYugH0$~J%`Qp*Bj$y@Kz4_VCDsYeS-_EV-VQ;>)-LX> zI)mR|zKsM)q!xQ(iBnK7M3(AdzhZ?YK8vv&9!*XR(b^XJ0$G%zK|F8iv4xDuKb}9t zr)+GF14*Qoy~N_*v?2a37V3WUwy*>}V=RM}$=ot{Q(aYLu{;3cdCP<=WK0g4JGw5q z#zq|lNu-w3#FDa8#}(D3dz1aC!s7azu{6&jJBWPU^XeeWmo^}tw;b$8-jNIRALCzX zO6draNG&&rrGH>NJ~g2AqQwt|rHHe7 zwR9mCy~V9@AM6L!okhap@QSg7G$HrPvU-b)k!62(5YJnlZzIqBBa4?jMV1$RK@zD& zn^;_CPcuZ<*z2x_CBl-+S-wvq@5mh!-Ed#uP)!ieTka{7y4Y_$wiH>u3<61{mRZE2 zquA;evLrjVuMn1m9HuVE`;(I?X2qt*pt=~kgLvLj;z;^nKfb((dq{fvf+SLl4Y6b> ze#GCzyT8a&*!cH%V+)B^Fcgs(YsHsv59kQ;dRT2MIedPvWr;4 z3oqc`v^TjLT31-yUo#dDHS!Aw3d=%jB8&5F5YJm4TqAS-`ZbnZM%UP?yC8|w5=bnb zIi1YVH5T)3S0iC5<}9`5k>`H@_AM46OTS_e&s)B1Aa(imyk|DDTq*}iq?Rkh(zd!k zKEX}ns9{TC$;@Nw5*k7Fj40cl!u?|-fHf5%TQ!F4=w7kB|7oVMkN{c75-UuZ}7-pwGT zd`GwbK^4r59{;!vx}s*efJEYo`fW$|cl{4I$rPt5+Gt1jk0g-fKkn$BZ{pD$?dZPs zv#|2I`h7?D)tK&hAGZGIU!&25`B?FNWr9EK=>C$`6KC4~X4iPMqkCU{km=v<=>F{w z{IxaC0qy8+&oh>uSg+k6p0~WOPVW7_ z?HZaP%LgxzL~4m77N=sr{>Y+vJI`ELVm~mJXEVu){4({l#>k?d2I6^3mBVCJ&Mzed zcdWjB1d>QCX~Z(kVZ1G}MtKF&Y5kdCT}8WE^|aAoKyU zXpI0#q?T&!@N+*wJ+d~k%>1xvm#`FamiuMoWQw_Cr&u6MbP0&(EvnbN@a~RhwTJgb z7QSmb6byT@=rf`(^v*Z(+!?`V5HYEuOAqK2~i^NinjF zxd4(#E&Yh4U&93%$Wqii{D`nbd}b`&#*mfYMm3M&?%Q*!AfC50en;+?eu+CCBg+bP zkVI-RAeJNA<_D2w@{61hVe$RKSh~(2JG;&<+MtE%VqXa2dCMjZvikGYGy%_H?v{Wg zQp;juX;je@?;bodRwYVU$~jBBbL7k7v9ouKktJsei03UzC&)axTX;x+RF}E4K@zED zGqE(Q_IMDg%RH?fH-sg-n5m1W3Hg1P38!-7kmZLqi03T_RLS$j>53!nR&Ov4B#~Nt ziN)x<3GO5|+H*fuSmM7jmewuE%I|X3dydF5#|OmomK&EzUAlG~Q3utf^{4)E!W$FcjnK|F7HG??5k5gLc^I>?88Ac@pcPAp%x8sL*D zo*NpM2}^n@QVb4wKcP4W5tr{s8g3MZt-5tyvF$^Bh@PR=27ABcv8RVwqp7 zXFRg_Hq&n-EDk>yOMnr%_ZKK$a7T6d(jLU~mZoXc(&e-cvaIR?l1MGa#M1eBSG?AD zEKH}1u;g-<4#{MfTF)Bi(vT%92E_B0)>`B%sbbd)YmjB+O^`%tSxYSHZTI4nkJhzb zuPH1E*e!>CG!_k zm+Mo=+C{|bDBKl2XFZ7LEyvx+`+i)XW^Tx$xD_OkTD}oWNG(b# z__?1k{Na66mkyKfuM(DkEsW*VQ8JEs=bM(IYs_dCi03Vj>F--U(RIi`bveHfB#~OW z5=+uGB|L|jWmU3SSezUgOV>AKZt2~pizTw0Tm#~Hi^@|nj?Js7jMtOXH-RKliw?0^ zFY?CkFE<{Ar1c9F2b;=FV%7L~5B$EOu+U z;4%53&*Xi=lCqVlOAAjjAA7j+13vMv{Sgq)Tc))mJBKg!nQDlxu}wiBiPU0CEY+Gv z+oAg<`>A_?u!K4>mZ$`BjqSZ`x)#+%^)-m+Ee#)#_ZPQM?eRMQ$wH7sYH=l&ZFl`} z$7*r>`e0%4+QwLXHj?+3A#J8?N0x_oKs;}`F^#;xq_-bwgz94X5G0XWjuT6fcU~!~ z%dd`YFA7U3X9=}ifS>!(2``M0MX?u%=Pkd=$i4r-c@N$VwXQ!%BDGv4mIVf7=E&l$ z{XJG#vYeT^q%|RH;{)Hl#2wx%^g%ptF)1VSJn)R;(5#H&E&h7>czEV&^6{|1(HZDMZ|JEYNr*dOU&NF z3}Fd$VJxj0llPYw{>PsqOZov2&s*$%lJAuVN9}lxET@iwBvMOlRs7r!m~DgC#t%ih zz7!Uh9gL-aFd4@l|J25v_knRBp0^k@BXgKX7lZKbj*>)>L~7|kEP8cPx}v%i##$E& zOCe`5uOPpny41lPzgIR%2JyUQ_Bpb4kyG>xufu%150XeNLx^SS)jHjgW#+oL5@AW( z$Wx{;G#8wFiQD-qL#l*+JAvJ7xf?%i|FsiPUnKSY|KYJrFepF3^eVA}j&B z>-?!>)$tLTWwwr+wH9@(KAHeBNjg?_+YQGZs~hY>olwVWx9K3uKXt6GxJ)`$2bQ)- zM;)tEO+dOob*wJEN;+1LW@|^Ij#c-iAl3iWv3lWsQZ(vVtzr#Q$~#tvzG#nUUt6!f z!e=*pum_3674=)kYE~)fSbf>!Z6fMeeY+VX`Hvl|1A{X>P{-;;XOQgoj@6egeQ}?u z=f_%_=)%00dYxXFD^@METCixf&B9enS1kYk%LUgS^~agMb{TLTb*zrg0Ga-6$Leo? zplQSSQ>bHgzzdM}pF36yzcj)ft0SI9Iirr%=dVG+f7P*iX!|RCVeDz1fIDaAyazd@ zqeKuHC30^^7od*S2E`!Ff9qH+)OtMtJ)y(OF8358PWE3T&Tp$f_o|aGB{j|8jYXrU zu3}7OJU@Ci%_S>Yg=!x^q0zIq5=bKbzeJJ$OU331?U3bKkct*2c+ z{=JLtWgwon+;2>t1aXFO-;kwOHQmZTLTb53EFV2L%tRKAeoaOSOOP95aj78P9>-^! z;4|+RX@hv)Qs)7A13a7Bzz|nBTKc# zAfC6}_(f*4z8~^-p}P1l14*QoiNw-XCm46&eX%uKD=hw=jAcw+(ye}~!B_mN&5t*L zc;2$L9$78z)p5%{Wa+*QB#~N{5X;n&4LhK^JU+L@Nmv}b7>movx%iP;k+Ke3PDFut z-m+~M`GVCvP;Vu&6vcuhQi}tzB-cK68d(${tGNqHE@$zcLY^;Wm-G&zYi!di5YJm) zeIu4Ez0ntuMe7|%BDL%%mXxfpU}S04a;%@QBDgsDXIia;Y}? z;+qjQ68{cKo$ers)DliC`-fkdg)E*oy-o;Ah&N-&|3Kbmd$x+e_x{9m5YJmwT98!+ zH`|*vP+fjzfh1B(GO?I9i^sp$=iuKdOjz7~7|Yn%$b3D`Z9G1OC`_B#XpWTLKfu|5YJnhcOuWqCUzT^ zAONu{Jm&?|qE4rmkN;71+R|~}RmJsNOR+-oNmQ3tZ%x_>ODSi$_laC%leAB~Ll(md5YJmSA0zk6TF032$TCdHwDOOT zTAmZjq+J6vkY$R6p}MeS9cJp%F>oTTOKeSt!N@Ym55)79wL{47_NX3FtB));$3POP zrG!{ct=sV(SyuG&>?bU-0gUC+I%jNoySh^#vM3A$@w}z8BbkpC-PX}WmY5MBiPWOp z5kL1k6|4(GmbOX9MhHvb5yn!rAvvRHzoIKX+oO3pi03WYx#WzZTZ+e$k)=TvNFueU z6N^i}Q9iP)_IRW(EG|bGi&|asJBl8OFYp=ODSjZHw`iXw@B5p6XydMeQO7_MsYRDq zhG^T5N0#d^ZcZ1LLe8S^K<@o+gU91B`Q2L(&s(}TCF9tBj|V=eF0DUf(5s%qs68-r{zH%*SpFu7^)R_~ZnVNG+>~C0S=W z{(YFq))AJ%5)sH)w1TGKd;e&?y?8fN{|z9Xw-^m0yKe3rsQwVurGXPjBDL%$mj3sh zSEIUk&vITPEWXDXOZUs4ku})a*8}kR{Zki!c;4c-fUGH*PZ~K4Sw=4bNu-vc#Bw{*z8G2Z9VaFWOZsW1 zE}gEBZo<|{9lN4ytoCjY&s$O^lMcKmM{N!uOTB#{iPSQkSd0>yK17z?b(=pHmZ)IH zGT(umd%dsoLvv&~(*nfvmM0xU@w?cGByS63iR}!MNG;2W#iCIRzQ!u*H^~zg{}9Hq zp*ETG>s}~TK^D(eAfC4*_6Ww7i4(P7A&WsDkVIT+lSNFuddB9@^G-0__MLyhpdTmD{$NjS^YC13)1 z7xNudhId0A>3@?Cpa(@Wuv$EdSJj_au>Y;GJ19w-M^VD+mVZ z{?vhYzm&te5RGcwR`KVAv9hy#hl z74=&O-l6s%@oR!jy_#202cBjENb(;$@Iv_i=SN@GF7f7O9^`UtbyRA@Nel`a2yhUAuTw`B7r>i4NNNbQpYB494 z`)}>%Bg=;mdZxnS6v0@u_K`2XnT_@hLzbcv5YJn3eaJP|E3{$?vc&xYNu-u7#8Ts` z&KhKSaqPx&Vaew#iyX-t;C82^xyW+xDv0MT?dp(mY)M)Po&|oq4U$ML2Z-gMsuF(F z@NHRhov@@_VCr&aIoU@(c&RB~VYJl%@w`R%5&0U|W^2+-RF@AUKoY6tJh3Qvsp0*~ z1rgeCMMV+l_uohv3eGd`obY^z;f8P8ji$CIx1WBZp6k=&otlAA- zWAiKu-Gs#}lCdPjkQK(nvx&H?Akh}Y^A^j~f>(pR5RU0k<-BvQ+3V!3*`Zwaz2 zobu#=u#|F^20>(xZw>3&uTfnl%?9zj#n6Ym&A#j1!V+0}n1CcwO9in6#|0fjmi0@5 zjtfgx6jPUO(c}%#=hKp=$a2#f#PgOPR^<6Iac>?z1MlTpkVI-}-UZKL_T7Gt@0T@} zE}_B_dx^0a+mK&xbZ#+fG_tH+1>$+j#7^Y-vfZ!#2vnDs>p>E!r8lu?dgtJ7LgTX? zFAGaxG-Fxup3GskOKzj*NFueE5X+>hGu@G8!Gr4e zgr$(PtV}0+;?;g$=zuH%SsJQbF- z7^W`q3FK^#;cc$t-*-rE0OEOzjU&1DzsD`b=f~`729iiEUc^%1{k1cy%YdK7dBPHL zg|TRsl6yaB&+Zab7u`BvE8}^~k{{vt9rgOTxp?JwU?Y%3YB@tJO9pG=v!YL~e)&mQ ze6KQ=$JfaYqLDdo@i`^-n?XEpsd|fyV~O_KDd>LLz8xfyT5b`GOYp_Ls4k&Hn*0=& za?WCLpS&Y~=w~|=)n(-<5YJoe9+7$Sy)ENkB1@MEAc@rSf>=ghEyKS;JJRrFO~=1i zezUJJbqNVyf|sTAhs6v*mZS4PJa3sglURHsBJtT#`^`WSsil-yG&XijM0M#<{;iR) z#K$rg4}bF9pKRX+@=PeDklW{El$a#Fe^1|gHiPX}KSSG#e(+bt)O!IqP zg~jzcW3fpj>oA)Rw!&vfD64{a-ZH)_dG1fHw-a|2ob3vdNG&6YMQK>VVPpww71~!= zia5*E#^n8_YE;I2bia70fq35XZUw2!#KoO(hj;t#Ac@p6mssxm-oj@Se$qcQTv*a? zFm)-PIt}0Z5qmVPk!6G-i03VhzwE`&m&a?jc0u>chdCgL)Uujb+T|2iLv?ZL)mUFx zqT(4#z*y4Ztzcg34YFts1o6D(RZH4YRKIpPvh>jgNu-uN#A21v<|eW%jQcWOSp07? zmQ|_b47~PjmgBS4>m?Ug#`Bgpf#kii+q%K7$kH_(B#~N95{vfL@zan+wOz5XusGae zECJQY+Qou)YbK()9Owe#dCR8yY1OLcFJ$TOt7s)GxtwK1 zAer+Q-x-d-EZUp^@w}zOL-J+u?#v*3mZy0bNFud7CKh$GX)Tdu*Y~|^geBoNQVPCt%V%Pla;8}jve;P8-6AX@35;ct z2RX;L!?~$Pk!9H`5YJmYBgoptTJ@x0}HZL-#P?S|axge(iF8?bxVJXdS`?sCYiCMrIGJ1 z&P@;CcVuf@5YJmW1d)5c>ei0<_hAmL2T7!s6~toNyBXf47UxzaN>~C?7>njEGEYue zTQCh>W4gW|p0_Avkk0$6BeOfBy0i`eNu(ARV!7_t2Y-L@dYXD&Se))KmKLwc`^!P? z^KFsECl|!?mPw1w;`vy1&9@cEGN=$Fky?%r%fUr+&!GF|&4AQ9!jjKfuIiF!rE!&t zv&hn42gLK14Vq-0+-7plTx1zG4kVFUqKPH)=AB2#((2jJ$HJ15%G4$4CfU*Lywe2l zBG_06;(5#cT4WpxtiKEI?CSRwB#~Mk5KB>~o0pK~(c}HO!V-Fyv0NHMzVUrhnDi9Y zMKcG)^Ojd5$f-2*N7lrB59$RViPTa^EGu_g;7+f%dYy}e#Vd`mTpLDC3u>^b=lSdB#~Ms6N_G>*oUZTtE{t4OJTYHm*o%LghxM-70e|D zbGD*x!mk=2lcbw4yL2Ev&+lusPD4;PVaNcG<)69ptDD#~V?e6^she<;!lzoOo6z3?q?C6Pp0B5hXOgMDjdRf5kzoiDi7V>2 zZbH+qA8?YW2#=YloABfukmNsh6W(lLT#C90ryQ}Zysm!lCM;+_688Y!KV`EPU6{Vh zS5+qXLpP!0q0u-~(D%;@sGIQEDv;^lb`$>g2by$SGXQlH+HU}9|GAs+mvcMZO_-l= zw-R*|W;ud{|EilX=bZ+A63tZ&j6|b^#!iq^HcB`cl2PJhBjpXKn=sTJr1@{%gi~w0 zdx@R|uM4ZHiV-L90W*3gKOwV?gB9j#=vwdc4aD=K=aF7yc0SE=|4#IOnNR_e$o?-? zmXiNV$noTw$g;?zYY$;@NoOpfJ;-XQwa;$+#rM4*i03U`YLoMpwlJncYB#~Mo zh^5T+2w7!N?=VqVA~G0@%WJX+%degB2xNKM7sT_HyPL@hW4oVEeS7k$V79RMK4L5uwa7}=+>WcSAWO{)AfC5O=t-`z@BKgEH^A^1kVI<9BbJ21 zo)O5RtGaHnu#|I_&;#TvY1)&-kI3TE4aD;nyP9Mc_%dVzUj3<|36e-Hzli04^~c@F zlJ4-sR#>t#nYxrdARQ9Vs@%YPukY^w@w{b_D(PHt3)5(duCbWCAc@q{qC0+TYGB~@1k1f#k%-U8xz%h49(`SKv4-x*}Ni zc|NH8xt`LUp-65XAGA*@MV_vW8i{n~}v^8zhlhOo`=-&IcoOznq&L zdsJA8ILp~bWK0h9JrRv8mrX%DZ<%+G{E~}mVK;4L(Y6Fhq?QfD5_PLRKFf2jfBmz< zlAguXWp)j+DtFn^8UMo0*5e?aw_K@5?)|qhzj`5y-WiZYYS~9DjYIa~v(^;NRicF@ z>KS9%Q$fDOSDQMi4ArHnHHhae5p+!c{hTG4*DJ=fa z8B4vzWS*?xmH!r56t{tR-qLkC*%x~yz%CD2ieG>vQp;^(*?TQ+AhIkukd`Ja4lfu> zg)><(si_}>_xQSA0r9-WIGB9BKHqvmI}baU^KB`mp|rPopN zrKFivkqN3xi~S&;w^%!qRk^*kr?QbHHvlA&TD}v@nkDlyk){2W!Eb~mA)Beos>kG6 zsi9MVzxaCI0`a_MQcJQb_q1VlcVtmb1xciqMm_Lze{Z#Z_&h(26Q-YqCFCVzDN`q1 z1?EYfc%B@84#e}8ektS{vu#)M3|Ur1f+SK)cVaO}x`@B}6gF0>5El1Wj79ql*+H~_ zRae}tzN{^X=PeVW$k*$@E;**iqS6T@ky=I(%R8M;@6k1OXrXm&r@vQzi#dz;FLJ-^ zt3J^m-7lvsKs;|TyHDz({Jz;pWKpmMNu-u}#Ih*zEdGTZB_T^;V~|8@Swk!vPe+)ex+rcfY%eS^xs1hw{xbNE6*X`t zajiWdp0|9zPAt}L3-FGJcl$sRsl|<03L7N3AR;5B18QjhHHnmn!C6V>H$ zClJqD#?~Qo%c*bs+o8I&m<^IhEkVTMao7@fuC(74(qCAd@)(Pi3HhZkoetG~kVRn+ zi03Wg^~gK&wbCYAkVV-WB#~O;h$St(<7rfvtJgMa3rjv{*;GNgJ$5Ty$wZdW1Q5?# z3fqvmWwj_chvwMcHgyu7r>$YSarq_bx9>{Wf6Nu+6z2A~| zo_;Wef3#eoZj#p^9&SxvjutrooWMwW~fAfC5WF(8&a z)6{5WiLwJpq?S&^GUon;{irSN^vQyQY$g;_>WVf&c7BZF-`Q#mW=8Ru>*G|Rut(rFVb<9UmcAz2?A<~vTT_L zl1ME_iKW|~mmbKnYGZ7muoQBZ<VKTeE$Iq9}dq5JY zUavesv_*8s2chgpDl-eNe0+%H#kE%E#Q2pf<@YI#pA zqYWnBMV5V^``;24-y+7+VmW!=pKmv76sk-44-n5=Mp=?^EPk{>6;zjd)wL@B2&qM} z7k=(9-nj^$6R&!w(>-A+=PZrKl75vwNyd0L)QJob&s&boAot5I%VNBCVeuR!ky_dl zi|?3$A5dMQV>)CBOZF$GF5^y--B5-bZE*MPnp+^Ax0p62md4SK52Cu9O9e@!mVw0b z&Dc5>S!^5L%M+IP&y3|<30b$SQ|ydSKqzek;(1GxS7e_2_CZH?WKr(~l1MF6h-JVt zi+$Llcm9*G1btyFUlYmMgqONaF+=yug()DOw>+>R^RYA6l)fU%&RHOd)M7y_nd$TK zX$}boT+4;UwV1JF_9nllbfDzUB4kNA2I6_krTt`0Y0&0I__UzJV30&=*-9*X9(}-; zd$TNRZu@&psfe=_#*z2R&S}~BSI#X>Ks;~pXhg0tg{p-Gs4i+2Ac@rCPb?Wfs%E1` zrx*RkG!vFXU;o&F=MhLoG%fw?;iv=8V;IOJ>A(v)Pc9E zrVgI^RvD@H4RzohnE=xLsRPgN8|lC+yxThrb>MkU1F8O}4!lkZkxHlo@8KMfQr>|# zI+L8|w?=nA-jg=W6eJQ?)NdVl->Z`I{2nhh^FSSV*O!7M|FHv4v)bDar~|K}``F6s z>h}&j{Q`a5H`na$=B4Pu+_imNWr9C+;N9D6fHSEDWFAEwc%HjKrhnUk_uC&ht7?B& z)PXm1A4vPp9e62*U2z9q^Lv{IqYk_jKalWWb>Nv)nSeX+hQHnL1C0{VM?p^6DB*CD zj1s-|pLn1Syqr@Y&425_`*L?r40@{7RI_R=Mx6A&Mx5V%QMbA$Id@^j$^!$@lc04E z5YLaED{hdz*Sk!WdZPc!*1jN#?Eg~b0{Oqp9qX5eEZdB$cM+DTZ;WNW4%zEdeRoc8 zWHDI^;(5!F738UwIplP8WT|Zpl1MG7#1gCC9iJhgRilZfu=tm7miwgJqjrtpBgitY z0f^@(>1C$`l1MF@#PZr$1^=2uO2L$Q!V*%(Sh}?!T?Jn@d~S&>xnDs% zZ`pW>d@1=js^cM4m+gJ0R{jxE%Xng0tQpr8)#cKjw@ZY@{Rd+iYee2OY+_OuCjC9C)ui03U6 z=`TTV9+jOwyttKK1Dar(trMutzqmV0$p)JB#Ut{|SbEb2q{ zU-~t$ia{1{FOWoPc}Xnk@e}a*WZSw%of4LO&Jxs(yu<#?8;83+F7^TOyd|;~8Iu!# z4#F#pdV@d`sillq8fqWItJvKt)x%Sx~AXOSi2Wn#9lWNl^YV!3E89>>CsZ1680?kNEAyk*>La<0$K zll#4pMdu4hBDL5NOR8ZO-gVP0H?2TeVx1UEL@w#{I$Bp3pW(gaJ&5Nmr?!y}ZzJ7m zKFHGhD@Y=>xD!h|gRQGcUEX{XmcVU{CH5oP*)_t^Srgqa3GYBWZ>iFfJYQ@+pBaoS zwx2-~spS;0Y-=;(DY9H|q+ZSW@0DK{XU3w_oP53h`PC8cQd4LH;(5#I{p1YqZh6VL zJ1Rj9B#~OK6H9P+FMNLgls>oW2}>boIr@Q|eY?5CD@#-tt-&Ckw=9St&&tmAtEr*7 zB$y~-9{#_QU@RBJW~av-;!vrQTi;&s$EUle&Zt_rP5RCwxH? zsbvJQBsaM~0$CpYh#M&^**lrKXlasr|9afM_2?S=p>Vx2p10_4BI_^@CSAkN%7l6# ziPU05EJGhZ_<<~0pVv(kmiS$aMNNU!<(d62d=6IO4-n5=PMMPDe&fjV2Iv~wQ+<8q zA0f3^6U)Pb+B1-)VodE>!V=`lSPtltx#dgCV`ETVMoj_nyd`x}2>xcUI6V9{vY5{T zNu-vY#A3C;7M~OUY+k;pu(Im$sjl<2lUT?;wfP za*SAhnY!R#qFmR;ZMm=%ah9-w)A5+B&^kH~)n$b#i03UmExd4D92;$Uhw9R98Au|v z#1M=1(@#gyHKy`u%UWSc-^0|U=Uj4)#TlHz9o`elKs;}Gbc;OqEl;e#zYk+w&9U;2 zkXjxRi$TH${Ci6MeG0Y;OOzX92?`>+JKmUC<8ytw9|7^a#cw~ECqGLaioe9i1%V_| z%Lih~S(dd4)n&thg?og>-<`2!^den2N>voMpli(c7>MUB-4~MYVx~j);?79VV30&= zsi{ex`^stfOs~!@TkRJXhrNuYRa;Yhzr1X-9RJ#>dJ2f=E!_&pH8v?_-%@nHY|$K!Lb-p%T9PFNB=n7VY#A#1P)%Q`wE%ULrJ&s)YnA-~vuGek2DSprvpBvQ*% zV%e_u@D;jWI{I5h3rmP6WARQRYkgroYYjq{pMfBrw@maWT{y{iNBbd*UkFGdwOA61 z>C1-ek>%c*<~N1K-HWk=_9Y!flY2K+MV5*v5YJl%J|y>xgT-dtwRZmoNFud35li9O z-T3$VX67rT2}?0&3DRAN@BR4a?&`?m@(INAmgHEnZW-ykY(A>XsP7<&)N+tmjEm>t zZ{oH~lAZ`l=02t_U9-tJHp%!9-fbJd7R2+G$L6HNdq>{oI>^%85hRgXE)q-Pt;K86 zHD;D;|5{jLycx^3IMNlpBcmstTP}GI;(1G}KBUWYM@d#qWV!nlB#~O~5=+?m5d2+i zL`r{f*y(@$U3{9Y8#9IX{y;Uo3C-u8-<6qZ>#fwN&kkpZnv6M|?+? zDO=4IxBtC9md{z-l!#^RwWapx-cND`@x0}xAF0cyUIF;`lumkqBvMN&Vlj(+wgxqz zt@w7ov9LJouluL73FCH=RdNU0_s3CZr1LM3NzxfvwM}o_8M$K9jH9SC@}1&?%D?BI zIwQ5JkWmy> zonBeT-#a6{E>6Y$j20R5aA)Lz9T}Ag{?Hkz;4&R&a*W+@40T3Mb_1FIZD-_ff8fN# z*L6{6q`41B`_G+`9banV&d8B9>u8|P$jpNv;lJvP+;6`WzfG@x=d&M;cpZ*|oYGN3 zg^UvEkNQ_fosprzAkBa4jO^$%2cLtL<=CjT7;#em8gYK>jQlo@%r+7%TfRWoda)~r z=SNQ!7cvW6H=@f&^nY3F1(HbrFMY}XC1*p2(dbFg>geRo!V-Fbu|#N)mET7T?H?nH z&lwQUTWoEEagE!(i@`IegbN^v)G~osj0Ro0j4UZd*ZT;ImmgypvXPvEzQJGz{`Hfp zKR`Thu`eO}mf8(nJr&iZm%@$8KSF9TBNoN+ZeGZ8X~)N*!cxjvLbsAVSVI;(oQ>*Y zOOZaAofp|uX@o2eejtg|vWZxhn+D=NzK@QN8!IeX{!CqF1d})O8K=G7kmcSv z5YJn*a*5^Vr6Rn-_$d-3ky?C+#kpcM?7_?KEH_G`^%FOTHzD=Pf;E zlWsz%hu=mci=!<_BDMS=mYO*R_}qoA8#T5FOIiR^m!XfySD!q;rn8X6&Kbn>mgzl7 zUFvOrzXMrr?g2@pmL~o1Ta()lYy9iLVN+FIg(cz$V`<_?R!riyRp4Jei=PAHc}x4I zgj*%8F^mM`1M zvocQG=o+fa)E*#-)G~%xu03gjpZm|eLyihdIcI63MRpPRXWLanmSbTcp0^k@CU3K6 zem;sq7NbOvL~2<;EM-=^^U=LO-(k*KVaYzm)aCPg@(w$+K{UR`=IsFSyrq>TSuOo_ z&_o|u>g@$dq!xQ(*?Fm!3bHtCRE`#w_&~;T?5;T;laICQfzORQ=?&s}i}OTpJh$}q z*oQkK4_p9Aq?WzJ^8DB3J;*X;QpQbT2|CVLb~hmVVn62C~@NG+#{#Zq^68)RwJ`0+hqaXrCUdaWj3u+04Q1|W-38xYT1tQL}cKQnG! z7P2UJ1WBZp8^rRc`F-5&@%?e1=fYCNS)8*;_wCdhZSdaf@ozyqZ%N-m?){LQKKOjH z^_qPv{|Koii&%yZP}+g&qI!1ZJ7Gyb$<)QF8@XTno9FwYx}-&cc-~SsjdVxN-M66_ z)ulKVB#~Oa63f_yP4L&ah|_((2}@KEV{tl4o|PBv*5TbyDd#~vZ<)7;+%H#3#^Dvl z%*!B&)Y70oe(qO^jxj@Zxl*oI&E@aaQvXwoWyKrP<+-EyMmAc@q{g;+*qXxk%;GbmU z+jluM0Qp+r2c@We}8(BVw9qcSD3BgQVW?Us*(Nh}jYlbWbFN1jAQvU(@y70l$75}=H z=M9iVYOx`fS&EG&Aj_p2QJTUM62e$Kmy>HO&qwbhvMA?)c;2F5O~$cJH3IOCJJSy! ziPW--Sd7-cTZ1e)fz3t;i~AYIa@S@up2IjE{)Kl!s-Fe%yrr(semsYH;NoJ6?w9ln zAc@oxNGvzYCgHOQV|9m(7nWkqvil*aODiA$OQM9FS$+3Xnu< z$sm@NPG|A?%AIFaUnDFs=NOB3eNvamz@eXzC1C-G=Pezy$o&%4dO3c7S-b=!ky<_y z%aC1~`0U#z=WnhQmVi*k(%~W5Y2W|n?Ze3O=`x7tEpK~;;ku~Cuf)H1vGfK=BDK`g z!q5F0t&ih5Oz0rhO~T?7##lxlC3DO7&4;f>me;x(2H^*O&Z)&A9;l=9r}WJ-m+F?yrrElvD|U9#J>;IVk1Z*wL}t2%)XN&k>$SY zv?sz6dy%oE=#lR)ot_PEiR!Y-8^rULO#{e!@`ziH@Gnqj9|TFHmNa5fY217(vJC8* z{6<&;BN>aE@(g^xD10))bC|)yKs;|b>q6#ZLxz39@0Bygfh1DPTVk2o>HzL2vRdo( zRaji27)y-@ zs*Op?!m{JiAG@OljV5o)dr}tOM%_`foIxf@chu0Enz%b^Q={`2P^Js`_Jbw^ce zK)R#u&d{rmx}$XWfpmZBj+)zTC%&8F7d6a4-BE-5L8||$JL*xA{u$IAHT@V!DesON zI+1ioz4%yd7`mc-PJ=|^iu$cPD$#;;M~(f`*cNq1-3tRr{$qF4%nk0(QFm1D&&ie7 z)$iR=AGgfLJxz=J60e{Ov%|}j$^?Jtj=H*j9?nz|+cOn)N9}n7GX2}`sNeoT0|nhM z)E(9MBS`zt-BHt$hT`rhtJ9NrqV6c$uOQ*S>W&I|xD~%0RPS7>j7GetyAoaO7<=+)Vg&XUF$1G zgLr=Qd>2n<=W2!LaL=!r0Z1bIzf|c*{x5O6m*CU3YgirXEG!YxjAdFm*}Jes|C152 z)LRYWB^LUoF?#ce;i$$58$c4N#ei5Yl{XDRPqp(4&-D`)-^+|8;yqcBSG;=@ci}|t z2l2e+Q6JJHKkK`v4zirT1CmHBi-{#aN#!xJG&p3VD=g)lMYG8kJk}d~w!mMn)6+pb zZ?ViE`^n7fw%vv-Yirt8{t;5kW@0HH`3|2=c&+Q{Ny3sH!_+13AXzbKxo|fAB}$uq zAfC5mI+DGlJzMNgK$b1TK@zFOmsm=-xM-ldjJrByp0LDUVJw$*NWXu4;6;4zTi*up zyk+1LG7EgZe9~oPiBAPdq?RyZiL0^WJhJTCUd2*ag03-*qm<@|9Vp0~6vAlKNv z_dRToC2SfSHrt{w!pPZZfCx zkC0m8iN*fYw;{+9KCYx(Sc*BzxQC?sHuL_cw&;GTVhiGVOKvaHA+dd5I{u}w`2iq_ z)bfm25;|%+Axp1;J?rfJdsQy;HdB|02IT#vp_Ar!WO=Cr;(5!%{^Y%K(ze+zQC-fC z14*QoZ^Tj%_3RF^M1>hwDUROX8 zsYPiJe(vXWe~8Z|?d3j8O;`dF8H-bAvX|6!+Vyf|`E(G(^A;6t(iJ^3cr-p^H7E!q zky^SE%gn;a`lv1q``79%EKW&`W&Bz)w^a1Hy&G9p4+rtQX^3bCyq<&v=le(73^6?i)BDKsW zmLGdh-$HfKvT>UxEGfxMU7FE%u^flSQK&BATR=Q-iL6iF#TK25sD>=Pc7r5Ri!HI_ zr`NBAEW^xh8VgHk3S+4~o9rGusdcdvvg9lW@x0|!Avr(Btoux3WZAO@B#~NNiDmfx zUbw?sbx7bcVez`dSUl>H_sVwVMfj{W%dsGyx188bp8Kk0AIBg|xWxSym{u_wrEk`nmC9LHU{3}iYKS2_yzmaa#PgO0{m64aZ0Gn^$fD35B#~Nb55~`Z{Vq+^kfr_N zW#@#&EOd>%T?FEJOS_(A-SS(RBVOlUYz2}?EklUqOH{+U$g=OUeWI|W z-Dm3Z`5?LXhiuk;gDfxNK|F8SGmOk(?$2yI7hPl4DIkf|Vn{4`vGs80y=$N58Nw3r zfU%seN`8mrpxbo(3)HGZK|F8C{z!I1t!!6g7_v+q1(HZD%ZO#k!+HZyU2J>y$PpIb zbjDKcEtw}9yfk`=>f)mX;(1F*Q*w>jd0SLRmf+zaiPYjuES0D_im511}3;iqh(3G zNOWNu8y&Aq@P}@XL3U<1lR>9nQ&G1^*?f@c-*$Wa_6KTpb1p&M9{U!9wEx`gvD|$u z?)EU+u+Rl{dyHKM68@`hkGwK+vh0Qp)A0$Wi&lZ0(orIwj1mJQ>#Cw|5B2pR&426m zxVZC*0(z=_zR;kl7;)nN8gYK>_L%BK#+i|YFCL(4{rgoA&ySw1a!5bsne%Nsq0#fo zEs#X|f4N8gFQw5dO^_wBV0H!|j>>Sd9Y-?pPae?XQ--#|QXDR@Gj&?+;N3X!F6 z1xO;byd#z$CHZ)7+#(g_-ooOV$yiL)$uET^T#CeBN%wXK@w{cM0_l*@ZelzKSvL0r zNu-wQL-3o%@f~_kP+e|c89hu`ia5*qrR2-v)X=aws4fM!Ks;}`s!Ue0p8e{00a?za zf+SK)8)CWi`rK(`nOwd_Us%!~Gj%ypjl9iH=y1IzvW#fsSQ*b-5|YSlqlc2wJ!G-c z0ZF8m{>0LDW+dLbkhl8jOks(7!dT9|BX3Pw?Y!}d$@eWFp0`w0C)Ze*a@liaN!R^$YS+=k*lyIJY(w8$ds(eD>Zw7_r$MIGOLW|E$cGL*K5^&Z}Fb^);&NHsU@6P zHby7l&XvARYws79kmro0?r_o>X}xSEK575G9*E~H&X#1AA!e8vKGSQS6-XkrBooU) zw|RD`E<1{D9v2q(7mP)rg1njM#yIvt_sg2^AfC6Z|4R1v&ph{28(m{_tKO>oBczsG zVsR=SUmICIswkcpmSWEG%beW%RrPf6Z{C}B1M$3Nei!nUG|;)|8L~9e1WBZppTuHQ z)tbu6_(6wrY0c;52ek&MZUv(xaLyYnlML~3a^6whJO zp1kjl>XJI>?j2!?dC6GptC4GLp>ZhgChRj3#PgQYS!8bMpmGJDP1tNaNFudp5X+jX zSHh9yX0_8#geBk=W6AtMRv60zw^^Y3YokVI-3ODxeV zAK*RlJsPgh7Z#@+#M3l z>as{R@~g1qa~Ad4WNw*!Wn2Za7|sOoyrs@0@|E;#21%qAPhz>7wSOhD zSZ-2R7M9T0jKy1@{Hmt;wpj(p;=3Kh^A^V*WK33^+7EZF8M%QZQcDQ2jNkZq8nRfu zQ*A3OUU`fqqb2Dmy0N-o8M3GqgLvMustI}DuV$@+-z&G5gCtVRO=8LWz7o%q%QpT{ z7nV}aa@L-Fy|&JIS%xgvcYt`_@-v)_V|}eI-9gt_tOrOUwLB*l^*)nYqPkplQWzjC zS#OxSY*HX!<9v(W<4!L<4-n5=7XBn(NxOZ_>5D8)Ix1HF5mHMDvFJXH*^ex4*&jy= zOKd)4DXT~3FpGO^$K6qr-hz1EvN@C7FGrO(wm}w^PauiZqC5;g_rH8OV2vye51LI8 zmcX}+#eFmx$D)eIJol+nX|A& z6f%~s)5y-Qirm$Gk!52Fi03V-qt0T>fvI=!-s`g&Ac@qnn^=CiZ2E*Oy0tt#g~j(h zV>zor&W&5J_d^S0>DLg%^OoolWFPs&H+iL~F1K2MBvQ)>VkuB~x)51%vsMHMOF3sb z-;g}_b(b8ifh>pfKs;~hXhEJYwFZ81FBv`b$J{0`j)T+6)_gI2-1BU z-axf6s>_AzAfC5GwI_4_=1F$J$g(;KB#~O`49CxXd!@$_$a3wIa=Ng%eqt=WQ%L8^ zteoRhk>z?Bi03VX14xI&9-~z=kY!J`hLwMW)S^Z#&dClhkR@cm(`;cW;w z{?hIQi03WI>&P|sBJ{HcvfMiZl1MGfi6x?7^+nVmkX)KwclY1#`z3!_{;xs6X5rH1 z8VgNLSKBN!U8`HCR@FvTIw-V%LcYXkK7Q~4b*_A=`m!>Uq;qBW_FlMirO&-q?x=HR zWo?kZszUNzzO8E9mZM(itQYSJZEvD+?!*&XrpI5*DG(m5V({AEvyk+`Da=qT#)&%ceINk>&kywcIC`YZJ_HU-{rww{(Riq=d1o${;IQk&$Qc8(_7~ zAfC7QH79R?*G%5kLY65mAc@pcKrBi+HQ%FaOlAJ#iNfMu%2-lQkQwdGd$xGL^2n7S zp0|weMOI9*XXlzA%jm5jiPWMn0>5e0?H{!sS$a3zG)Gv9IZMz(vM;u3wLb3X8XGne z#Pb#{Q!=A{9G#(yEMu)e5~-yvu{?L*mxL^}6FzJ)n)z`5YJnZ zcanR*VC?3_=o%|@0ZF75b7FC+AAcUzCE;U+tFSngGnTDFMXumhO+p zm-z6jXU8E+**%a%YB^6VMV9w`kfmXZ6=A{>TESQ%x06-uchUZ1k>!;Ni03Wob;vkY zd#1h?vh42)l1MEn#L~#JeGsy!wlBXTEMC7DOPkB&mnrstJ3RWoS|tffDQD4sMb{SWy5o+bl^sDmZy7UztYjH_)EbT~ zvwMIfQcDG~tiLmK8?v0M+ayC+vK;FD-g!T<9+{8TiEV3->e8+yi03U^dXstbMXMk9 z+=WAGAc@q{TpQ0}CNBA!gDl%>FL)&^v0E66t|93ta{Ygu-S=P5@Ao)xBiS=El9iF2 z>`}IcNEt;aTgm8EWEGO4EeSd-}meK{?Iwcxx8Ml&mVBk z?RkHm$Gom{U5}riqLIb36^Q382?I!ncUz^cxMOuf7m!41=|?O@#{=AvBXzlOH3jeL{A(RZBDLrc%fz7dkCCNS!&Mc+;xuqb9)Y5b`zVC1E-82lX58Ig7WmopGmEs|6s5)Z#%bjcb(8LzY=>kNFErfIDN+Gbbxli+a_q zhAhQeAfC6pTu z-V%S2jKSXiprwK=PalIMQcD@Jj7(hg9a$P^houNhxEEtdS0Zy3p+7gaL+8s%9T3l3 zI?f>DmXU1(gOO#hAxI*%)E|TI`}%2)mdKJbC^1J^yu2C9kK3eMefQhecol*BSP;)! zBD2YRrD2a*_#OG!WROH^=|U_K%@@r=mb#A{76{8H&hnCekt$JtBJT3++_X_uJa5st zOYSd`fzkLYfhWg-BvQ*rVp$rSyc1dK>7;!SmRujEF2%3OILzjn@9-Gx3qugkTdp{h z`^&RRHFtE39Wnt)q?Q@P(qoVM1k|`;Qog#L=il%9=XW>wQ%6yLAsO`x+#gebI*KM% zUsIJy(oq!X)f0CV+51}hqmH7oIv~qGbrdzgZlHI*Jm`*i;=2zjqYPDp-yCfZF8v?M4UY5NF$}1b^r# zvaqnlnWh|>-UD?My>SDX{%uFmZ$Hp6<^EyRQIzZp(*AQtQD1E{+))&nb|4&e6pcOr z68@`>BA=IM@eJAq=aFXUDiM1G|wPYsH3Rs36SQ$bre-B?u2Kr@9#+a z|AmeF-xcS#y>PZvlCL@JJk$ZNi+%3};`ys*_vvILtLMPb*XVz-@&`$z|I149zl>X6 zcOtqI9Pbj}Qd~WQd>PA==VW$aNt2CF(0O~fH;Csgbsmy8z`3DM@Ef4XK#)XgaUmAp zy0e^-Wqh;Z&cfofhq2_Rktf=52d%dti@GX^=Pd?-WR$^EJ*Y0KOLjMqL~1!qEVH&L zj6{}s4af8smLkqFy_kIWy2X;dczw)nWf0F>%6F3|r&uqIe#ny41tgJLE)vUkyIOeH z&DO7PjS`kLKc+61SCad3x=P$HR2R!K5YJm~+$8(Qn0B$n>&XRIHP-kBvQ*@V!3qRArYM~ z9&bvv2}{C$#?sD9)w(4_V%6 zjrS3j-~)^$tQ+~_;k!pEJ5gO4>413N@@N9Nzx3E%4bNb8G6YGamSx29G2R>R>U?eg zpu@uA62Mpn%_1YeX^Q))Oh7zudEiTC7PbD0#&y|a4w6VMF~o9a+O|q$8S1bj zRajyUGL~8WNL?JuEY_pC^o|1Yye0G$x$l27z1ocj(hJ)@E3whu@mwfrKM^_tgnkmb#BpAW*4AH>u} zqZ!#Va?X>^cwOw6(;%L=geQ=ZtT!c}Ya`2t^B{@T(s~?zhABw1`hhG)wIhEDOVVM+ zk{eFy66tS;S9dIK0^)f~&t;@8mYJ{dO#IVUAc@q{pIA(i*Wh>Lzz^Q_z5X7R3q8VE z=JX*m@h_X|^&uaAzgfzcp|)UtqBRt8kycjSoHE7XLgl(VE4 zk-czCie{}v=l=dqAfC4@I!x}Bcjs^4fh^s7f+SMQMq(LP{;f8u%g76M8p4tt%+%#* zXGi@0Qhq(`B(e-h1o6CO%_g!g)~Nq=yg%8S6p%z}*+VQLGX~oui-&KDmav>Z##oel zk#U%s`Lh=w%e6od&s(-clVdE?q5)px`|ub@BDF*k%gj#k4Uom$s=;Jo2?${j3xD<`sw zKyj&IS5%jFZ$Uh72`nHpzNHQJl_HDgXOKi{$tRY@ros3+w)^SV#llj+Szcz5Yx1`3 z;dqatvJw!_TdWU|Yx45mHQyl1q;il%YWYSi&s8S|qGQbAyTxC^k{rs^CFBNqwwyIy za}%md$65=j;(1F<1(_vXu61E8vOH@5l1MGh#^d|`hnTa$$kL+isV%}1euA-7)F9`6 zgz5Vs$dc9r#Pb&YgXGz=u(j<`WO1Gal1ME*iKXm8%lpXk;(C;auy~zhENu$NjIaNR zZ+Ip?E*-@4ma$Hx+vB}q;y`4naUUd+TE-EJ_VG>{$gm0>e?;2Gc1 zIv}36%x*!hV@2)C@w4UWsUV5eVn-|q>J#0NCCD`{Mp)v)8B6*(G7gh%RQn>TOP&vi z=PiBj@4@p@AL6p~(J|(}A0&}lJc&i^#-TgNGGXtRtHKf#!C2mgk(qdAo(_wUr9~)6BDI7Pi=zDzJZ`C4{vusioFW;^J~y%=qQl!bOJoVHUQiX! zTXX}+`@Y7B-uRinLp_j0YPm`*TGPT#p}NdGTk%j>ia5(JU9zvwfWjYmZ*}WyAfC4< zUm|0DXPx%pFC9P50!gHnJYq?0s*3kkj|{AQEi7rLn7Y`0An%n2$DDCQb?JK?#Pb%7 z2IT%S{%l-Rbd1$G1(HZDABp9|`|M7rE|$F%zY0rK6k`dPYlNTqmp*EO=U;x!0`a`X zNtvvIwr#k3A+jhtgCtT*g9-S)zdohB05u4V=v}k6_uucx&;Dom!(KQ&){&=~tg;#! zsB@+6AdpGYxni&18F#J}Khz(HI#=RGf-L{kxl&$9I#-g8MjN5dm3`wux<7TU9HD#R zB<}DTjyhMo^g*isQ|HQ$XFgX^=Srm!NGb1J`E4&8bK4nm6@b- z<%&jXQ`EUq+X5u{kDV(?tF-2$&Xs~~3RTC|@0}~pPub&slXKVJA4CUc?2%ek3I5Qz za@Keg&J=5XH~@97Tnz=8{%z;VZ$D5;uQTpk*&hYc{&VNb<4cQi=SsD$zB#CKcr!k>$ig5YJn_rIWjwO4n9K$g<)ENFue2B9@~LBk-EV zn6NT6VR>|hsf&JJ@)X#x;0oSD;?!9X&s%)Plez@|wAzO(HkUyXsbwayv7NNy&ctM-WU+iTLMeSNLGrPPAz0Pp$n2oEvty-?bbDT#3b^7 zskX2L$1s+XUu48&>BHP9$f6zw;(5y`CsLPqt@}yH^6V@~BDJ^@i?h-6eW)%E6GlxI z7MHV(MInoHd)(bII|5nyXMuR$(wxp-cdj)SpZf!zfFx4O5n{RAz-d3SbR9C>R9H$l zi**QDe`#wyIv-h1EC=ztrPduXOS&dvDBf+`;V?)dwOk?=XX`O1kfr6<4oie3GnT1~ zN_TP{yH*^zA6YzVb*zf#Et~I=Ss$y^9e5|?{AM7D)RIjsDG_V&eoH0aFWLx8%sIxg zOqq=ST<+EN2eMeVyi^s>TNH}P`Jy}U8s5=8z7I$uwY(*kTXxa-IyT=c-BDQl&oh?M zt;lZMn>$nsM3#>?K|F6!e{%{yLv%Wl@e`fj{-@gmLbG)d|+4$WLbaw{&8VRy2w~Q zy(jxF^vVy%Gg#kGgLvN3x-Pj_-r9XZ6Ip6}07;}4Lt?pNH98Ji&Kg~g5th(PjODL( zq$|4P;bY3^7}Kx;@w}zNn9MBJu7SDR5E{di)cvP++4a1o$D9F4q!tZgDd;d6-z&AId36_-@I=Nkgmw}~cy3Nd zb?IXV;(3c|2sy@vTbN!z=gXuJkVIRLJir=Rj#z*sQp-YO*>T)d9i1HL4l6hkt-aBehTaZL**+eYM?fc>m?{B$7rwB`K5>uB&*U58ovrBV~kwxhP zi03U^&YZ*VFZNFo@NC@iI>V}dgw*0kEb}*3n}q7JL1)VxVTrrWSg!Tofse6u@sCoF zWy}T;&s!4wNVmuRJL>p3xzh=dL~1!jEbhUB`ytDxq}9uWCFllY$q6E}aToSH3q_WW z=^&oBtkfo51rKZ9oQcjC#T<}CYDp%RR{finB1?FiQFg-Obd#|pD3fuRwmL5^Bg^+r zAfC4@Eg<7CGp$clB1?^)Ac@rSj95ypTkJ%Z%7GuZ3QG}ZY1eByKE|q>^v3f*#iKwx zZ)p^;7oRUa^*7XPxb@VKz}rZSc-1NE`xk##>j*4Lu? z%c^+Z5;BII`}do82O-PBIv|PEGM-qPbza;USrk87#t2I}XBoAKj9om>?{VZZUNk^K}ZYi}lJGcwWq^ z?zO6T-g09u*)wveR?-Aim#mf`iPW-|SW+9#!(VM)`m+6PVM$11EJ=xLvBkYs%o0?W zm8Bq_x15?z-o+NwRl}Vt+sZ)_sl|&}UiQtz@5p&GG9L&_@NLGjU>fNpK3h)(pD$(e zKs;|b^N{R+S#4e{zE>Vz0+L8ACx~Td!yDe{d>LGA!AoIrNoOn@&uzzbnQ81_57i~D zE{NwXeH)QZucMiJi_kH4rYT4wwImYDq-C}Jk;TYv_D5kU;VhjWkZ1m2Z>KxRqU!+S zdCNIFvft9yrYc^@(rpJwBDFjumR%YN^-!~e)Bb5ScK`jpe&VJ2N z7tXPxAd{pEC-GHR+=cVNM!PfW!kHfiviwsQ&Vu@*@nLf~Zu2D6g>&uAY`57E9Z_SGK_YQP{nmw3 z!=7~EwCLdLin?(6XMiOCu?xp;b+7~K!fBjiRCQea-h~r!*%9~7X#9E|j}FX3mz z_(K=Yx1g;!(|nH$A5a(0?9m|8zwN^L?FYuRG1!Xsk9j=_r2XeEoQUt%xC^JuW9wh2 z3nyR-NcgY1a89qw!B5nC0_*HVSBbqdKu+0J!m)^4C9dvNQAAxhx@I8Ff9t~e)q0Z? zx_a)PyRdlgHoTLVy$^`zubzntWPOZgtl3!fzoeW3No4;^ zwNK=Kxxf80ULVsuT2)C{Vlo*^t9@iW*&@@qcvt6dmq0vkSv84RPLC{CLl*sOAc@q{ zP#3>xSeiKEx7i_EpQ{Rs{~g9snn5~O2Ah=Pu7dWvKs;|r${}w}+gw-lL6)OFAc@q{ zm00c;H=K`-v5Bc&dI?JIT>#RVQ#D^e>)S^W! zF_E3xBg?ntNkfDs|1MLPYv;%)!|?YD3(&n`%vuo7TUz`iZ%uEC8sHf$%Pk;@)G~`$ znuTmHM3w-}e&dBDDT}dq4kS-b-N)XkKo<8gAfC76`jEFK&rh5EP+fwifFx4OYGP^P zIpGSjJR7ZVC@i7bjHP`J*{fi2)?YTrvOy2T^A`PeWE8tn;Q{WDXlewKNG&^w#qY^_ zTo>OR>1M*>c8{^VOCw!4JyXVaL6!qvAfC4$GZADyhN5DSCB+%xkoH>s`pJmmQc?T z+l1x(ea7Pd(HKADI(=P?*OXlG2l2dRekobEls5ftdt@1L7$lKe-VsaTwYYA`(!XAu zr?3R%GM0V|$h*>(r&%pgUA`%Sc-|5eMy_Ky>vRK^)@w{cB6FK+WO<#BbS-O4)Nu-tz#1go7{5@nbPVzV| zECrlJHGVTb##ZiC9EmK?Z-aQ=qUlP$8c;ON7q9b+ya$p3^B(Rl^Y%L%wnVti03UG&B(~_@Jsrcs4m)uAc@p6jac@lm*6#69|9*N2ut`w z#^TV3jAYGS@R@wY={ShzEf+VEXUjuc}@Bt%T?9)xx(^^vshmv z&&fHdR(O{5uO1+tw_Kb?u45KgE9{V^Wq*)FY6&Ekz3z9fBg>PTTMLCH_c2qKjXlXX z@6WW(!fQz@2ZMOt@^T8Xj4pnMJMdaA1xciqSYpw(>x}oon%+{eOjzQcFqS=?NnI9L z4Va9sV>7ygc;52VhCK6IeARz|>N0CENFudl5X=7ATKEeH^SX5YDJ(%x8OyxxmViMZiPTa=ESqvnJkT-be!f9H-@ivoo$?vW(d}ei zY({ZkGi2$M2;zB5YFlz0+o^sx8ChDSf+SK)b$xu_UtRZuGP3+qz1UJ%ia5*Gkz~Kv z1#zo(Aj>Xe5YJmwW5`~fDKBR(5~iltPgtU!GZyX5WcK>J zt1<4rE$IW|d5dW$@;sKQW!(hTrF0NTBDLrfi;kIbGi2$Q>^nkOd|xn@iEqe?yU#gA z^H5#t-vRNwC2uTw7dyZ4oI0|^=7J-gyf0%+0yMl5mHMEvAoIMiuWXTi+{UbSX^E)7LA6a`*xGP0`4lXO$PD2 z<=Z53jMeCKYzwN(&`gj-YI#m9!$V)-RRrGq)prO>31`_EOI8uQ&Gpnn=ZjY`i03V} zt;oCBtl0gFkwrZMB#~Oa6HCctn^<(dJp69HTUauSn7ZUGA>(6j>MMRimdSlBs^WRe z^1)=>(mwj48?soMgCtT*^U3%bW}T%k?kEaQdU{Y;VqPq67O!2g6fjAyIYvB_!l!4opYo!vP*vLhNv!qK_H&D z#4RVC!~>0YIitEf4+BZ0mI=gC?roKfEd6v#VuhuWv)o)qp2t>ij==jY-kk>Gd5iWH z@-DXj*8vON7Hg{G6=XC_PbF^4~CZIW&W;RGIm_HSSnVZ3W_a%h_wB zF87X&#$N`1t_qS!E$fKII_{P&s>|drFVlr3sf4j?)+WbT`z<;8=oqV0{Z&;wZ#nRY ztn9MQaVbN`*n+wsiPYjvEU7(@;4g$G&6%GkETL~1%hm)k4s&{*20q3t7lU}-GOQMP zP99gd8qd9|uLMb?mXpNtVECyLRF{KhLtYAt+dIb6!kWw@I~%`=M0FXz7{v1ys|Ms4 z+mT=M6spUs6(EV!a*bH*XT)?t7W*zYJ_t)GXVF|pcJnFza+>Vr*&D?3miA`kdF+(h zBiwx(Fc2h>S{@Ng!iw?uBS!w#^M478$NN8aN98Xfs^K$^{6gJPJ=TFtlI|$Cj%v6& zYVu)gL)0Cm>Iky@Q+L$%nxs4GW{8Fp>W)%&0qOqK9aX>F6_08fyr_K{bw|ze1gZW{ z-BFD#E7qXys3bp-Qr;bvYD~JLw4J^@MrViPL6ArsQNMLZg`1G>D9>ejdZ;@}`xr>_ zAG@QXHs#Gm-BEofn^zrIzjsHO9CgAyO>YkCHAUS~)B9OeCHO;kRI>xlIMc<(_xqym zsL(+m)4%PG`t1kaUw^3<>W+Fa3Z(t#?x=^=>~VKgGyVK>)E%{PJV^Mjx}&u2zrs&$ zMkh4!+d+mt$SJ!@IKCiPiDNTtd!g6Rmp>kVI-}WPsl^GTwj3Pfo4J zxu^(Bzz4?SRGYk+Pr1|?cSkjv0^)hgZ40$>quRI|H)#XqJ5YJma(EXNf&1$qAJ#EbE4w6VM+Qc&T zrmz%nmWS=g6K&bo>pIAiTMFWN%Y$(8rs4WI?>Vvrl!GKviwUvROVf)-7L~%a+QO3j zk*UjwdE^N~Y)`l#eNMt?eYRa8fHX+9go^Okg7GFoc8$zvn3I4uB4q?TR8vfO8%DymD;7?)YX;`N!a z+|wf+ctzt?w2@_qK8WWndk>Oh?09hn{<2=U5lAAn1QW}QuVaglW#?^|g~IZQv-pi7 z9YxzBF5y`pms}9fTYP4b`^($+_wkx|vwV<5YPn1-ng%uz$fAENah0&-eqrj8dYX)6 zjW&M!71gEADG<+FhIc1xd|wzW-HMK}v{;ZtYRMs%_8Knzk>z<P2Dc$~iJZF?uf2A!qh9qRq?Y%@qWErmLu3hb%y$u%pmN5NZAOl< z0{vTf54^fNK|F5>xlPWOGb?qMqPl$d21%rrI#cj{|K8Oa+Q^cZy1`dioW3!ZyLHJ- z{KeDJTTxy9$_DYgrAY+2jy3)0jIU$09)l!OOGjd{d{cn;jC_4=K#;H$ahA=G$WHr5 zjy=ABEG3B`p10I=Aa&`rc0Hc;xttlGdaPCum=>Ke{I0UkZ{)E$4{k{E*L0ktJoMMWL`H{A4Wa*V|&t*~7{6QC&7p z1o6B@cQcud8+`q=GO|QW0ZF8mOkz<=tF;SRyqc7j3QO=W#&Y-uSw*0~YoR@|v`_}| zyyd7FIbXb+)lNZ{QyoDPspU1X7;1aql`7{(J5~ydi(|vz$1cWhB9@&KPvF_OW4a)o zwEiNky7xoYq|80zAOFuGpp_&>x8d=JXK|F8y zHG`b{LHS{gP+dIcfFx4OVq(#>czO=irNlWs3EOYVjwQ^FOo3A&Xt} zPpeJ_-%(vAjsQudmea&?*1M%IvYeb~K3!Nsof%81J$bKet=SQGdO1u5@w{c_CUU;G zrq=(6EY)U#BvMN%u?&1&?Ig0a8@FV>u(<7DECw@4H(^>wEj*{S)&RuwmbIB=W!H!3 znc>LNbOuNwwY(sfkk_q@kR^P?YD-}$(S+ESm^Ol><$=)9Q(v~+z z7Q59TiPX}<5I@5VnWYzo>awKgqAkL5-j%Va<&c$KeruEQzSqH;AfC633ntew^`FOb zk>$<^kVI>3==S%JQ*aql2w(vBF=Pjr2ldghLcXGZ& zUI0m?mWjlo@$FT2WV!z&+D}*JuXpJoGgZsq_OSn5@xwwFQ@i1^*cf4|VQv`_TE%6V@`SLYU_b{?dI}4IXEn&ox zK3W&=PiE3DEJ0YjJQz#K6EgpDa`bs!bd1#*2I6^3sw&xM@vB~2{5J~v`Dp=?NG*?vt437L2?Vk!uoEKxMG$!wrGaLf&e&qHuK|F7nG@QJ@^vKh`jOy}i9!Mg!G@gd<`)T7c zzN5Oh9Z@V6mLPA&vT!%qYpwjb_fT|uF;?ng9)PZNX9b}Sp;9X7Xh&%8S+zSSy4!jobAj>~> z;Qh8g*{QAx#i#?X`5ut&PaSym-jEKw9#*eDqYk`l0U*`?sRJ)I?pFrtz|#%}Ddin_ zxy?G^C;TMyNBC_eDhwnNN7Qc}c)l%22i{oQYd=s2-j*{U$$#v?oBL*LD(b*{cjtK3 zarJu#UVB#;+&7n4eDVc4Fh9qKRwejD2j1a5J8`D*AH1Vc2j1B0Ak)9?!29h7mc>L& zK^=H{w?W!}?!arce+%xwJ2l!qrc$}Jdb;?@yw>?Ly%K;mCzne zt`ar#mZhN%JcZ{V&425_+ZdaVciNvBt@~44ah!aZtLLoKWRxNK=wnZG^>m8^@%+`Z zMJ$<#_die-&b~D z58`>tyZ2<)r}3iw)sSWFR**z$F(sBcH#Vjs%js4jO2U%1o2kq4ePkqSRlg1sP+eAc z1o6DZZWDQOI)CgU-U->D2S_5d*bs|x*Hc50Wuf0U6=8|;Wh`~~l9>{nro}^$Or35*#di;5v7SwyoZ6=Q;oVTXj|1_% zrQuXED%bkg2>itm<>erW)N+hi)}4H-fh_B-mJbk?a?UdT8`(o*orx#@!cIGT5YJn# z)F-n(%JWX*nG!cAkVI;UCziFLi}B2&mW}c#VR_`o)aA?w^43(XaXDV&>%9@g^A=TO zGJCC*ehi=cL!3YospUSg%IlCYPtC{NJA&mSvWKfQs@m-<~mJa1{( zihM07Gp7~q77y$Rl1MG3#PX%i5Ij?IVQSViVF~tUEKk~z+3WM8EbvHHo(YKOEu;F8 z(bB*?4LpOj&m1I?TIw3%H;u3dJrhw~^g8O835&}<#NFud#A{Nc9)0~mzXooLLgr$VDbd4u6vykh*Wy2utRErY?hx$aQSap~92MqNxPpdCLe5@(i=* z<#xQM*PON>iPU0DEM0GJT!1WDPbxMEOUwbr610?bM~#gi8;&e$MIfHHgf%0hKaF~Q z!)vc!zXwUAmKDUZ^31+l$Z~7R4HseY4`3{F+LEsZY_Xhh8Cg76fq34ca*K>Gc5v>z zADu5&>p&8z#hF;vzwh$~S)R7tx?5N(IZMA{@;p{BQ#TMYI60j4*^A?+ie;A_qJV+EIBVhJa1V&hjibn7Y<&8>XP;jB#~Nb z8sq!^h_4-np<}Feoq8$4Qp#EEZ;|t*`emP9$l}`+#PgQCab&c#u6IikEsn%u zp|Pw5vM6r9@>5s}I7{nw7mKmP$UGE9vc}u+y4&#Kmgj{ap0}7CCfDRyYeO$0%iH%LiPTa+EDM8+ z@l5>E$i3Z#46`J=Pj+;ll%V1Lnn72ORqy9iPX|^I(~-n?;MQZ zU;Ii!M+!^a3C42sHrdz5aKo~j=onl18N~CJPv+$Qk})lF5VExY0g^~9eTikDuaY7< zU(StIo+KOtB@r*E81LG z(!!a#=va`|wyR$D!h3n9J_hl;<<)g!nf}2HTS5y#5~*c3v3N95KaVUQ7Cc=hEKw1R z<>oT-oUE}qAqUkZVg-ojElcK+)wba$i)_#_=3xtxNG;*SvQ1+=-ora*m%PDXBf*+ZF0UG zeq@=C>QdPj#PgOPS>&4hG5Q!D>r3tel1MFMh@~QZ0-m#&o3r+)u#|9?(tWe=yV&Hu zZSY*u)2ASww=~iq&oCVuH!MVTne_@Jky>UGOPhY1N1d91 zE=WUli5?B&c}s#lIrkT3+`#ipm>SCI`C`ni)SHR#`}dx;#k~qAg6_Q(7PkwG<)9(iD_UvWKs+`+ z@g#`nEmPi+v5Q>GGADGtG>HaDq?XRaa(2^)8L0WR)3Zt6gyrr3EPv>Vj(bEpFvj(3 z;EB4TPrU(|Bwf*wx$SUQbbZGYHmEDw`yCh>+(%u}3+jTD@~&ue<4%~_v+WPOe*0t-kVqU+ zzjZ|~*hspf^;@c}KwZ(>T7e}0u`627z1d^b72W(5NcMYI^i^v&+{-+>V$myfU>bYs zRsG$6=!$;h;ej({WMqV(uIQvaAk)9?ivH~f_U*RoEb59L9|+R^b62#sr8Dk|{x$n# zSJV|uE0P zioSgsr1@`M(S|oGWs2P zo7G6}{u)^h+yqIamg&S2>RC-6)#Zr8n7YDp{t{!E-i_R$-EAvySM);{5YJmI$CA2a z7=)!Ei-I>uBDGi(i^HtYYshlQr;n1b1jI3xso%*SMJ9n$aVPP}01(ex7Q7~JvxAnk z!(Z0x8w`?2Ejx(i(zLB(kwv3@xr(qjUS=#SHj@tTg4uP=(J}T{HxSQTw#*|>v~%>; zaR;7ZUywv<2_lxWO(yFi%d4CIU4^B9vy9qDo}6Zz-oy8~+F2l;x6G&el=T1J$Q~VI zYafCnQp*KmvGW;Gj4bmNyYv;7Qzhs?D0;)@HV-U|4A9Ol0JjXE|?mx!7GUB z8iOQKiwd#CY&D2Pmalf6Q-vk>DpQv)g=F-{Fy!oNbc~JM0^)hgfhaO6cdpL4^QbPK zt{{ojGKg3V+v`3+mOU0NO@t*bk+Gb!BQw6I`&W7+OQ#hep0^mhCL>v64{RTfEE={T ziPSQMSnNih!+RCH>^$6DSc0xGmY-9|9!2&$)bL&fk9vW4-tzGPd7HglHhCAav{4vY z^&_N~rNr{2@aj5Lmwk>qmI;ef5@UJs7g-apWvz(smBV*}c-~SLOgis-O<%kkS-yIM zBvQ*3Vwr8-^C`0Q>2P7SuoQ8YiZ5i<&9$6sxbwb-I*8{jEmo6bEW>*@zQ5Gc07;~l z{lxNN<*8_7DUSSMFDz--nYvU{BlpU&+LJQS`BMEdi03VjW|HURN{2ElRF|E}Ac@ox zO)OoPHN`W&?_Z7EDlAbq7)yyc`C3qQ({=ZerLYvl^A_6-a*TC|dpHVN63amnsU?kA zK3dGgBUws~cI*@u-$+jU#X-nKVOU+i7X8|fFx2&A+hY( z`xTFtrayh`BP`{dW&SNPOFA%O*cw!qLB$}Rw^Y6*&&g||dgIQ>@ns;1)bf*9daZ0^ zht3z5yiNy%x9l|`7K>2dmZ&Zv??4i%rIiVOhFLaD z1MgE(5o>l-SQ1hgOX5V*?J+s?C7vaSEH2JdfGrJTF4${)q$-&s#3tB>OD--SVx8>e4z5B#~NtiDgyZJY`fDt&~G) z!V+_vu{?<+*JR)NhwwLuR^0{hyk$=%+1F>I!edWl8J`D|NG%b>665!560+pjBxVbX ze>!9Nb$>QK#^!$*>x$~K&jG~qmWwNiCE{p9U1T}710<1JZV-#=oQ3O=#pcNCN5WFc zSw@{F*W@-4i#H(4pl}e+ThiW>Xa1cL)2xxDEE*({TAmWii_C%6$WkY=L7}kZXE1g7 zGJwop&+}|?8(BJ>1M$3NfDb?mZ(N-W8*#9z%7B83K|>En|tLMfUb@=olMS+qu^MzvnDU zIm?KVwt#BlAXoW zW&PXD_0DW^>m7wGAKgJbZ`s?5T$6)3Jt{|aY2pWxNG->SWmlkWbySz8)|Op`#W9DmXoiq3 z&*u-F7a+@%wIH6iJeo(wVVn=$Ux_UH96%DOC4pFW+q}W=FYoL(^bwW<&eDnQ-@mFt zp(C<%^ab&}W#)V`cA>WIsRpuiKM0aYEf0ugPN0-@NFue=pN;SPJ5*Hhs+&_WFDD3#*8|29TbI1Q^!$0e zDY7h&2l2dRg(W#(nvA=*4_VsW07;~lF2u6(<#K$!q>WaeA}pUc%i;}N@crd+=?x2H z(bfa;yye?*GAAC^aZ4$(^f3ZSq?VDy;#Ow~-T`64T+NxnlAFiWWpNA{>vL?WumoA$ zmVkKPV!NG;DQ$Av-3M9LuLMb?mKnt2Q2FH}I$sRdPnjnyaSs_w>OQhk#Y#m3_mGU= z2I6^3Q3aV78`jihIkKql0!gHnmBixw$qMg!_A+AE5@88?#8?s{$hp5RqY$rB^F9aS zc}rb2(iv%RYV%xVId%mkky>1crQ57o_?bUr`?ZzA;`ErY7)6qC{$Djr+M;7@w^OkoC z#9}*C?IEhm%yf`MYRMv&W>Zg|MwZD<$GQkh)KkW?)r-s{tM9jHk1Un#Ks;|r9YT(= zZspM#$l|UBl1MEj#1hd&H5gge|6J!KEWY`SC1e}fnd1FPXZ(H3pN;IR;(5!KB63Y` zZC8ND`Jc81Nu-uqbMSrta77Kga(IyU8Gm6Z=PdpT_AUG<^{yERUWsbs5l=?0ghi*=;zoTv-I-c}rs(^8MuR zAFtteBDG8;l&s*x9CihCs zqb>2yM~hZ~BvQ*VVlh(8z~lU9a@SuF7MB9XQd~gl(*MT57<7zzJ_GT*<-~nbmyWvQ z7NWW|e+`mIEnA5tcF5E@s4lbT`&|{563*hINxpRx)-SIIvUn{6@w~-h61gUqpSgtR zUxHVIBvQ)(Vli#C)D>Bds9#7Cmdrw?E{c!Hb!>gjlgY>uqzU4A%hFHeI`-MW{~S~o zpAjI5)DlB1iuTp;w>ob%$-W~jF)ta*aAz{k-=N}&6SAl)gLvMOet^vJB^VF)LzV*_ zK@zFuHnDj3%*NwmGd^_56BhqhjOEBk^8V7ZNf$f@`>+DU^A@{KWK5|>n;*}RWkn6O zsvjY>yd;*np=LJd80#8R_FPyhIm?7dVzIEAhx_`5odWT^#dtNDOVSLt9*-=~VnGtA z{472f-FNzK@zE@ zwHbbf+0dolH?-)ar0?p_!t(NemOre9ve`*iL!D5|aYd`4HV*}vB&(s0x9osdLp9$x z>=0TF)qf1g@=vRwZhh3ofAuoY<@eBPsCv2}-Je!N+1c*IsYbOZGDNGPObkJ)|I=!y z2SYCvq18}&CLpDJHPj|^vKlJ>tUq2Sd}=;OB#x-xRzo$)C99#X)nC62t%ka}1SI*7 ztD)S@c3Gm;P?Po6R2^5puZHq4_QGqJ^1emspaXM&l}%NGKdgr8yTb=(x-;?Xb+j7F zb`8k%Z&yS8_5)p_gQla^Q0+E=wEw&sO7+7|yc+7$=C$+DYN*&PAmP7S4fSqI1H2k) zYQmwu=qjP_3UbP>5{_-iRbqmd>O!;{s;L)9^WRoOU9Ge;M0bKZc1tV86({L6bM;SSq@0!WQ@M9(W|{UKbG0TQupHC>uJj!riwu`hX--i!QM|?4W-ES*m$WX&@|a zZx~CNg8{bqzbz?6bs4%3#PgPFA!PP?_=B@})+av*B#~Mc5KEmAaVL?*q~(I&Px+kVItGBQOykjgDHOSj+ z`+!2c>SmD#i03Vho{~3>k~yh(FPw#bAc@p+lUP1^zTSh5vAS111__Jfd&Y8k3z=Em zT<134%d^cb5YJnj%*ZTh(1<~A(J|(o1(HZD`NZNmusU9$TDzp9mar6X7S|*)6Q8K! zg!he`{29damQJ~3&--hwv~gGT(;py-)bfp3UTo69UD0z!`b-p-d&s&1j$Virdu*wHy>DL$}ky^$P%l4D^@t48-E!|))ET1^bHVb2X-;Yx0 zt$^whyc)#wmf&_|AK!}(WAQ$|qwGNvsl|*~6mAT`-?|x}yYgEPamM?Qi*V^D&BRZmEtV4Z}L~8LQmLFDnEs!O((rbgT1bt>K!*j{3PamaA zxC_Uxj(1f&Z+SS7jQ;p^JAqer&20*jNG+knlK#Ny46>B@S9cN?r!S1<$raLpr_<{K z?pkYq6vXqE)$hpY&&TY>xGTCc3?z|St`bX)=lk#~f?JvV*JP`cJmsWej-cb z_aL6PBqfmN{ zGBOLq^Oh^q$V_~fWjimSx)?qLNu(AvVmaF=2k%LI?!8TO2Nu-t|#L{8YXWSL7w136}VJYD(x_!wGI}akP@jJ5hQ4r5t1~w=6 z%IMsm_?wUo!$1o6)rgs>|NCH=hYh=1-Ohc0YRM*+GdjERDz(H!?P6hx`Ndf3Xp%ke8$4Tc9UWsy=^&oB>@_0S zvDbzL3Br3u*6R=Ac}unyS=nXZqz)c~4IT=TNG+yXgKs;|bVMU%{Qf3F@_x&fHAc@p6gjkN`C*f=Ixq3hA3ro^g#*&jy zu46@^8u%?M;~Fy-p-51IZG__6?O4G@!`)51`Er1XU6i*-vnRB z7G!#{2kceg;XTmULq2ef$&t_MnQ&AZ=j@*uhwib|kww z^j~)OJhB*c1@XM)NPBXO-ALD}jq38bH%KD2ydswQ^^T>ZV{FsyPCCNk=)zdW-6X3! zboOUgBFl{XAfC7M+7pFWL7N7&zm6<*pMoS(OEq(R-(S&719wIio;zVEECrm!^(dKt zd9=r{8nUz=cd{y;x0n==du7+b_ckKSq-h|D)Y67nEMuxqMRmF2-`rGKl3kg)oHQbJ z*)^^e-Z7-_5fIN?W<(H+R-45)kY)EtkVIY$BE^1+A~5x?Hc>W}~p=x-oTG|AWj|=9|s_jx0)z zKs;}`af|ej)ZCT32Gymc1xO;b_z_ENOk4bHnIFH$Nm%0C8B6yBSA4Im+aTk^#561F%KbcGFzA0@Rs!RO| zAfC6Vts^~Gx=|YV8D@k5NFuc)6N`TJFx(Zb)#lM2VR7G?CspT256f`hZLv;zv4Lv9kZW>+<^}HA)ky)3H&N%LarqJM?#roA#o!3tS6HU{y$rKb~F zLG;P>g9@sPNh^>1Ut$oPSxetivEs0rV zJbCufoz0Nt#z2rnYUxER^&Ow;AIU{3V7P36C0r9*g%9QMW;<~Bs z8Dy!w5hRgX<`RqJ6fL|8+UvoM6k&O^o2g4kBeJtz*X`4zkmd3V5YJn@>1wFg!wXIz zi%$tiBDJg~mLo5Q;OFH18(U-vOM)+B*`q}6FNL&dn_!$dl-x96w*6tv&KWF&;SHj}r$5_VHA$2J*-Glos9LIrp-lBY!^!B{F;$e;~7WyEG z)RIUnuco@~Mk@-MzcMTm7LWf~eqT}0f0!>`2k_Bv_F(ioZl9i3l|c489Iuhzk@%qG UBI<{9 literal 0 HcmV?d00001 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/flags_0.json b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/flags_0.json index a1deccf7d..22669d92b 100644 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/flags_0.json +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/flags_0.json @@ -27,7 +27,7 @@ "xml_output_file": "", "experimental_orbax_use_distributed_process_id": false, "experimental_orbax_use_distributed_barrier": false, - "submission_path": "submissions_algorithms/submissions/self_tuning/schedule_free_adamw_jax_v2/submission.py", + "submission_path": "submissions_algorithms/submissions/self_tuning/schedule_free_adamw_jax_v2_bn_fix/submission.py", "workload": "imagenet_resnet", "tuning_ruleset": "self", "tuning_search_space": null, @@ -38,7 +38,7 @@ "framework": "jax", "torch_compile": true, "experiment_dir": "/experiment_runs", - "experiment_name": "submissions_a100/schedule_free_adamw_jax_v2/study_2", + "experiment_name": "submissions_a100/schedule_free_adamw_jax_v2_bn_fix/study_2", "save_checkpoints": false, "save_intermediate_checkpoints": true, "resume_last_run": null, @@ -59,6 +59,7 @@ "helpshort": false, "helpfull": false, "helpxml": false, + "only_check_flags": false, "chex_n_cpu_devices": 1, "chex_assert_multiple_cpu_devices": false, "chex_skip_pmap_variant_if_single_device": true, diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/measurements.csv b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/measurements.csv index 8548286bc..e3b362ab3 100644 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/measurements.csv +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/measurements.csv @@ -1,1938 +1,2363 @@ global_step,grad_norm,loss,train/accuracy,train/loss,validation/accuracy,validation/loss,validation/num_examples,test/accuracy,test/loss,test/num_examples,score,total_duration,accumulated_submission_time,accumulated_eval_time,accumulated_logging_time,preemption_count -0,0.32613337,6.908206,,,,,,,,,,,,,, -1,,,0.0007573342,6.9125247,0.00068,6.91273,50000.0,0.00080000004,6.912909,10000.0,55.403265714645386,113.45354461669922,55.403265714645386,58.0493426322937,0.0,0.0 -100,0.8996247,6.7688417,,,,,,,,,,,,,, -200,1.6886598,6.4462175,,,,,,,,,,,,,, -300,2.236852,6.2716646,,,,,,,,,,,,,, -400,2.128366,6.2247353,,,,,,,,,,,,,, -500,2.709414,6.159617,,,,,,,,,,,,,, -600,2.2997937,6.027183,,,,,,,,,,,,,, -700,2.8295984,5.9304576,,,,,,,,,,,,,, -800,4.8218546,5.9338236,,,,,,,,,,,,,, -900,6.284852,5.9143395,,,,,,,,,,,,,, -1000,1.5892961,5.952158,,,,,,,,,,,,,, -1100,4.0788145,5.83061,,,,,,,,,,,,,, -1200,3.3639123,5.8046236,,,,,,,,,,,,,, -1300,2.4896843,5.8066854,,,,,,,,,,,,,, -1400,2.7103987,5.822195,,,,,,,,,,,,,, -1500,2.2846446,5.713479,,,,,,,,,,,,,, -1600,4.15943,5.6355734,,,,,,,,,,,,,, -1700,3.9693167,5.7512646,,,,,,,,,,,,,, -1800,2.3531916,5.68324,,,,,,,,,,,,,, -1900,5.8729978,5.6698713,,,,,,,,,,,,,, -2000,2.6857984,5.570158,,,,,,,,,,,,,, -2100,2.9501588,5.5314846,,,,,,,,,,,,,, -2200,4.772345,5.6718407,,,,,,,,,,,,,, -2300,2.3999035,5.5249157,,,,,,,,,,,,,, -2400,5.980454,5.530423,,,,,,,,,,,,,, -2500,5.5302067,5.5369496,,,,,,,,,,,,,, -2600,2.6095183,5.4367037,,,,,,,,,,,,,, -2700,2.5007555,5.5355754,,,,,,,,,,,,,, -2800,2.123375,5.3722706,,,,,,,,,,,,,, -2900,3.7864027,5.4772644,,,,,,,,,,,,,, -3000,3.2771642,5.368007,,,,,,,,,,,,,, -3100,3.2992141,5.498305,,,,,,,,,,,,,, -3200,2.212985,5.261932,,,,,,,,,,,,,, -3300,2.0509152,5.237334,,,,,,,,,,,,,, -3400,2.3782806,5.332192,,,,,,,,,,,,,, -3500,2.6957636,5.3275557,,,,,,,,,,,,,, -3600,1.5491158,5.1319933,,,,,,,,,,,,,, -3700,1.9965632,5.165702,,,,,,,,,,,,,, -3800,2.4652886,5.186155,,,,,,,,,,,,,, -3900,3.4159768,5.3259,,,,,,,,,,,,,, -4000,2.584018,5.209312,,,,,,,,,,,,,, -4100,7.1126304,5.181421,,,,,,,,,,,,,, -4200,2.1089644,5.03139,,,,,,,,,,,,,, -4300,3.3495324,5.318861,,,,,,,,,,,,,, -4400,3.0301814,5.1459312,,,,,,,,,,,,,, -4500,5.94406,5.0516005,,,,,,,,,,,,,, -4600,6.2873073,5.0627317,,,,,,,,,,,,,, -4700,2.7101483,5.037991,,,,,,,,,,,,,, -4800,5.4784217,5.1593738,,,,,,,,,,,,,, -4900,2.5405636,4.9272532,,,,,,,,,,,,,, -5000,2.5973322,4.854315,,,,,,,,,,,,,, -5091,,,0.03683036,6.0105863,0.03464,6.073887,50000.0,0.025600001,6.2957478,10000.0,2051.405519247055,2132.3507702350616,2051.405519247055,80.8735921382904,0.0282888412475585,0.0 -5100,3.0536687,4.9539075,,,,,,,,,,,,,, -5200,2.6351094,4.916395,,,,,,,,,,,,,, -5300,4.073905,4.9842744,,,,,,,,,,,,,, -5400,7.66666,5.0104446,,,,,,,,,,,,,, -5500,2.1717532,4.946067,,,,,,,,,,,,,, -5600,4.941291,5.053197,,,,,,,,,,,,,, -5700,3.2240264,4.700118,,,,,,,,,,,,,, -5800,5.332956,5.017934,,,,,,,,,,,,,, -5900,2.5677135,4.9340773,,,,,,,,,,,,,, -6000,3.4718938,4.8420305,,,,,,,,,,,,,, -6100,3.498859,4.71889,,,,,,,,,,,,,, -6200,8.484103,4.963932,,,,,,,,,,,,,, -6300,2.9837537,4.788044,,,,,,,,,,,,,, -6400,3.5700173,4.8348823,,,,,,,,,,,,,, -6500,8.076867,5.0427775,,,,,,,,,,,,,, -6600,5.3547416,4.7983637,,,,,,,,,,,,,, -6700,8.914992,4.9072895,,,,,,,,,,,,,, -6800,5.695804,4.936513,,,,,,,,,,,,,, -6900,5.53057,4.873515,,,,,,,,,,,,,, -7000,4.144065,4.7157636,,,,,,,,,,,,,, -7100,5.426704,4.743924,,,,,,,,,,,,,, -7200,3.953298,4.5941815,,,,,,,,,,,,,, -7300,5.157229,4.5932217,,,,,,,,,,,,,, -7400,4.61665,4.816322,,,,,,,,,,,,,, -7500,2.9376328,4.6537323,,,,,,,,,,,,,, -7600,3.5215342,4.5661073,,,,,,,,,,,,,, -7700,2.8087149,4.520059,,,,,,,,,,,,,, -7800,3.9979968,4.7383738,,,,,,,,,,,,,, -7900,3.816926,4.6109176,,,,,,,,,,,,,, -8000,6.0515995,4.6081705,,,,,,,,,,,,,, -8100,5.4585614,4.780594,,,,,,,,,,,,,, -8200,2.866084,4.5090456,,,,,,,,,,,,,, -8300,4.061979,4.5646,,,,,,,,,,,,,, -8400,4.8484974,4.8089147,,,,,,,,,,,,,, -8500,5.6259613,4.656275,,,,,,,,,,,,,, -8600,2.7164025,4.465431,,,,,,,,,,,,,, -8700,3.2639744,4.4934897,,,,,,,,,,,,,, -8800,5.209244,4.5432205,,,,,,,,,,,,,, -8900,7.47742,4.4246397,,,,,,,,,,,,,, -9000,7.9363103,4.509776,,,,,,,,,,,,,, -9100,3.1809134,4.5368423,,,,,,,,,,,,,, -9200,4.791094,4.713248,,,,,,,,,,,,,, -9300,4.619687,4.5691395,,,,,,,,,,,,,, -9400,3.1089106,4.4528265,,,,,,,,,,,,,, -9500,4.3220406,4.496364,,,,,,,,,,,,,, -9600,3.6211421,4.5407896,,,,,,,,,,,,,, -9700,5.0317063,4.562549,,,,,,,,,,,,,, -9800,3.3404558,4.507644,,,,,,,,,,,,,, -9900,6.4189534,5.3375816,,,,,,,,,,,,,, -10000,5.718392,4.4105597,,,,,,,,,,,,,, -10100,4.658959,4.3569145,,,,,,,,,,,,,, -10200,5.685758,4.766016,,,,,,,,,,,,,, -10300,3.6148455,4.4646816,,,,,,,,,,,,,, -10400,3.3866925,4.4377003,,,,,,,,,,,,,, -10483,,,0.011838329,7.4746013,0.01248,7.483133,50000.0,0.0092,7.565653,10000.0,4047.7705268859854,4152.775046825409,4047.7705268859854,104.81943488121031,0.0968890190124511,0.0 -10500,4.742617,4.4082994,,,,,,,,,,,,,, -10600,6.1021595,4.772485,,,,,,,,,,,,,, -10700,3.5745132,4.3019457,,,,,,,,,,,,,, -10800,4.0482135,4.5007987,,,,,,,,,,,,,, -10900,4.2445846,4.3842764,,,,,,,,,,,,,, -11000,2.615108,4.353362,,,,,,,,,,,,,, -11100,6.9523845,4.649722,,,,,,,,,,,,,, -11200,4.475399,4.4851074,,,,,,,,,,,,,, -11300,4.4156737,4.444968,,,,,,,,,,,,,, -11400,4.174511,4.2522373,,,,,,,,,,,,,, -11500,6.8448663,4.4615555,,,,,,,,,,,,,, -11600,3.8904195,4.323244,,,,,,,,,,,,,, -11700,4.4755354,4.2502112,,,,,,,,,,,,,, -11800,4.6716733,4.2932577,,,,,,,,,,,,,, -11900,5.4300494,4.391985,,,,,,,,,,,,,, -12000,5.5662775,4.557373,,,,,,,,,,,,,, -12100,4.708589,4.373385,,,,,,,,,,,,,, -12200,4.1173472,4.291561,,,,,,,,,,,,,, -12300,5.649754,4.2744226,,,,,,,,,,,,,, -12400,5.8242426,4.297608,,,,,,,,,,,,,, -12500,6.6620073,4.3694286,,,,,,,,,,,,,, -12600,10.304541,4.356823,,,,,,,,,,,,,, -12700,2.5696533,4.292733,,,,,,,,,,,,,, -12800,4.5342135,4.3642783,,,,,,,,,,,,,, -12900,5.4623165,4.389805,,,,,,,,,,,,,, -13000,4.455353,4.215511,,,,,,,,,,,,,, -13100,4.262497,4.3168926,,,,,,,,,,,,,, -13200,4.836943,4.2082024,,,,,,,,,,,,,, -13300,4.4469323,4.165616,,,,,,,,,,,,,, -13400,5.05738,4.377295,,,,,,,,,,,,,, -13500,4.3487725,4.240519,,,,,,,,,,,,,, -13600,3.4954212,4.212987,,,,,,,,,,,,,, -13700,19.33727,4.730032,,,,,,,,,,,,,, -13800,3.9991996,4.1616254,,,,,,,,,,,,,, -13900,3.0594757,4.121607,,,,,,,,,,,,,, -14000,13.660304,4.312894,,,,,,,,,,,,,, -14100,4.6917377,4.197096,,,,,,,,,,,,,, -14200,6.4460773,4.2874684,,,,,,,,,,,,,, -14300,5.8789973,4.3151584,,,,,,,,,,,,,, -14400,5.3040423,4.2579317,,,,,,,,,,,,,, -14500,3.7781277,4.2022085,,,,,,,,,,,,,, -14600,7.5626345,4.3698053,,,,,,,,,,,,,, -14700,4.7572074,4.1840825,,,,,,,,,,,,,, -14800,3.5479312,4.183678,,,,,,,,,,,,,, -14900,5.753579,4.2376366,,,,,,,,,,,,,, -15000,7.2250323,4.329081,,,,,,,,,,,,,, -15100,4.46364,4.182172,,,,,,,,,,,,,, -15200,5.9894333,4.148809,,,,,,,,,,,,,, -15300,9.251641,4.281979,,,,,,,,,,,,,, -15400,6.962396,4.2101684,,,,,,,,,,,,,, -15500,5.154573,4.1486616,,,,,,,,,,,,,, -15600,4.016614,4.1505713,,,,,,,,,,,,,, -15700,6.1276026,4.092343,,,,,,,,,,,,,, -15800,5.0641165,4.119713,,,,,,,,,,,,,, -15877,,,0.004803093,8.611359,0.00446,8.60309,50000.0,0.0046,8.634853,10000.0,6044.017327547073,6172.654530525208,6044.017327547073,128.34734845161438,0.155996561050415,0.0 -15900,4.57888,4.093354,,,,,,,,,,,,,, -16000,13.225956,4.05182,,,,,,,,,,,,,, -16100,7.334598,4.4720497,,,,,,,,,,,,,, -16200,3.6273224,4.2317495,,,,,,,,,,,,,, -16300,8.078614,4.173712,,,,,,,,,,,,,, -16400,5.320896,4.243307,,,,,,,,,,,,,, -16500,10.412425,4.2549644,,,,,,,,,,,,,, -16600,10.703931,4.3110123,,,,,,,,,,,,,, -16700,3.8985214,4.1608133,,,,,,,,,,,,,, -16800,3.5429342,4.151753,,,,,,,,,,,,,, -16900,6.4630256,4.278426,,,,,,,,,,,,,, -17000,4.048737,4.1002846,,,,,,,,,,,,,, -17100,6.7914953,3.9721212,,,,,,,,,,,,,, -17200,3.6593626,4.050123,,,,,,,,,,,,,, -17300,6.6902194,4.2633195,,,,,,,,,,,,,, -17400,4.60932,4.1178985,,,,,,,,,,,,,, -17500,7.9708548,4.1050916,,,,,,,,,,,,,, -17600,2.6216917,4.146805,,,,,,,,,,,,,, -17700,4.7296653,4.0564594,,,,,,,,,,,,,, -17800,7.799707,4.2269287,,,,,,,,,,,,,, -17900,6.8645077,4.1073165,,,,,,,,,,,,,, -18000,9.228012,4.161897,,,,,,,,,,,,,, -18100,8.892206,4.0767446,,,,,,,,,,,,,, -18200,4.7332573,4.035946,,,,,,,,,,,,,, -18300,6.9426317,4.379856,,,,,,,,,,,,,, -18400,5.192965,4.246373,,,,,,,,,,,,,, -18500,3.7718272,3.9473422,,,,,,,,,,,,,, -18600,6.6334357,4.1849785,,,,,,,,,,,,,, -18700,4.7322025,4.0966463,,,,,,,,,,,,,, -18800,9.549347,4.1870756,,,,,,,,,,,,,, -18900,9.255634,4.0793915,,,,,,,,,,,,,, -19000,10.306611,4.3056154,,,,,,,,,,,,,, -19100,3.4269578,4.068286,,,,,,,,,,,,,, -19200,4.3178926,4.030745,,,,,,,,,,,,,, -19300,7.502645,4.1645145,,,,,,,,,,,,,, -19400,3.6044953,4.057675,,,,,,,,,,,,,, -19500,4.1716957,3.9916868,,,,,,,,,,,,,, -19600,4.710085,4.284005,,,,,,,,,,,,,, -19700,7.094948,4.128718,,,,,,,,,,,,,, -19800,4.996074,4.214055,,,,,,,,,,,,,, -19900,6.261025,4.188875,,,,,,,,,,,,,, -20000,5.478778,4.1562486,,,,,,,,,,,,,, -20100,4.217375,4.108989,,,,,,,,,,,,,, -20200,5.179275,4.0980144,,,,,,,,,,,,,, -20300,7.8845077,4.1060314,,,,,,,,,,,,,, -20400,6.4547405,4.133496,,,,,,,,,,,,,, -20500,6.6166973,4.1593704,,,,,,,,,,,,,, -20600,9.413151,4.2928805,,,,,,,,,,,,,, -20700,7.022496,4.159736,,,,,,,,,,,,,, -20800,6.854642,3.9935193,,,,,,,,,,,,,, -20900,15.363938,4.427829,,,,,,,,,,,,,, -21000,4.5074787,4.078317,,,,,,,,,,,,,, -21100,172.12183,4.684217,,,,,,,,,,,,,, -21200,8.730419,4.1503983,,,,,,,,,,,,,, -21268,,,0.0022919322,9.054386,0.00214,9.063879,50000.0,0.0025000002,9.078246,10000.0,8040.030968904495,8194.285801887512,8040.030968904495,153.84260439872742,0.2327916622161865,0.0 -21300,4.0549254,3.9688008,,,,,,,,,,,,,, -21400,5.245344,3.922998,,,,,,,,,,,,,, -21500,3.6690187,3.9799995,,,,,,,,,,,,,, -21600,4.365008,4.003008,,,,,,,,,,,,,, -21700,8.622546,4.019605,,,,,,,,,,,,,, -21800,6.297261,4.079359,,,,,,,,,,,,,, -21900,4.2562895,4.0091815,,,,,,,,,,,,,, -22000,5.0948405,4.0710354,,,,,,,,,,,,,, -22100,5.627423,4.087027,,,,,,,,,,,,,, -22200,9.705976,3.9641938,,,,,,,,,,,,,, -22300,11.566601,3.898779,,,,,,,,,,,,,, -22400,14.85371,4.2494497,,,,,,,,,,,,,, -22500,5.3872824,3.9543471,,,,,,,,,,,,,, -22600,3.6187618,4.0621257,,,,,,,,,,,,,, -22700,5.3349485,4.110099,,,,,,,,,,,,,, -22800,5.688113,4.12165,,,,,,,,,,,,,, -22900,5.4531026,4.049833,,,,,,,,,,,,,, -23000,10.370789,3.9534934,,,,,,,,,,,,,, -23100,6.2461042,4.078623,,,,,,,,,,,,,, -23200,2.4730885,3.982605,,,,,,,,,,,,,, -23300,4.9338408,4.0624886,,,,,,,,,,,,,, -23400,4.4797587,3.8812568,,,,,,,,,,,,,, -23500,8.2697735,4.055298,,,,,,,,,,,,,, -23600,4.2814116,4.0046673,,,,,,,,,,,,,, -23700,4.901803,3.9208598,,,,,,,,,,,,,, -23800,3.2840915,3.8979301,,,,,,,,,,,,,, -23900,5.047675,4.082336,,,,,,,,,,,,,, -24000,7.71506,4.0892963,,,,,,,,,,,,,, -24100,5.444098,4.10045,,,,,,,,,,,,,, -24200,12.715812,4.072267,,,,,,,,,,,,,, -24300,7.617429,3.9664364,,,,,,,,,,,,,, -24400,34.862453,4.144719,,,,,,,,,,,,,, -24500,3.3269737,3.8677778,,,,,,,,,,,,,, -24600,6.338803,3.9111123,,,,,,,,,,,,,, -24700,4.7011805,3.908792,,,,,,,,,,,,,, -24800,4.3293614,3.8776958,,,,,,,,,,,,,, -24900,6.150139,3.9630978,,,,,,,,,,,,,, -25000,5.3020363,3.8640473,,,,,,,,,,,,,, -25100,12.331409,4.2634387,,,,,,,,,,,,,, -25200,6.8357325,4.014176,,,,,,,,,,,,,, -25300,8.95919,3.94396,,,,,,,,,,,,,, -25400,7.7545238,3.9830778,,,,,,,,,,,,,, -25500,2.9759076,3.884789,,,,,,,,,,,,,, -25600,5.631503,4.1341662,,,,,,,,,,,,,, -25700,5.4408426,3.9994335,,,,,,,,,,,,,, -25800,9.098928,3.789277,,,,,,,,,,,,,, -25900,5.686289,3.9443994,,,,,,,,,,,,,, -26000,12.45071,3.8306904,,,,,,,,,,,,,, -26100,6.9194365,4.0217752,,,,,,,,,,,,,, -26200,5.154389,3.9351559,,,,,,,,,,,,,, -26300,3.6307793,3.8739924,,,,,,,,,,,,,, -26400,10.270405,4.105071,,,,,,,,,,,,,, -26500,4.810779,3.888793,,,,,,,,,,,,,, -26600,5.1272016,3.8464904,,,,,,,,,,,,,, -26661,,,0.0014748087,9.214938,0.00156,9.203479,50000.0,0.0015,9.203677,10000.0,10036.045359134674,10213.59507727623,10036.045359134674,177.0388777256012,0.2861685752868652,0.0 -26700,5.959622,4.092478,,,,,,,,,,,,,, -26800,5.150821,3.9019198,,,,,,,,,,,,,, -26900,6.1790857,3.9832623,,,,,,,,,,,,,, -27000,5.6589074,3.9820309,,,,,,,,,,,,,, -27100,8.318817,4.0195036,,,,,,,,,,,,,, -27200,6.7748747,4.122509,,,,,,,,,,,,,, -27300,4.271657,3.9145846,,,,,,,,,,,,,, -27400,7.361725,4.19463,,,,,,,,,,,,,, -27500,5.072948,3.8473644,,,,,,,,,,,,,, -27600,4.327602,3.9479809,,,,,,,,,,,,,, -27700,4.496597,4.02079,,,,,,,,,,,,,, -27800,3.7771697,3.8987079,,,,,,,,,,,,,, -27900,4.4086103,3.7915576,,,,,,,,,,,,,, -28000,8.001785,3.8547556,,,,,,,,,,,,,, -28100,4.0981445,4.063959,,,,,,,,,,,,,, -28200,5.188049,3.9182615,,,,,,,,,,,,,, -28300,6.277266,3.9122481,,,,,,,,,,,,,, -28400,5.167067,3.8868423,,,,,,,,,,,,,, -28500,5.7057633,3.7813003,,,,,,,,,,,,,, -28600,10.778226,4.39167,,,,,,,,,,,,,, -28700,9.745854,4.073748,,,,,,,,,,,,,, -28800,5.0420227,3.945804,,,,,,,,,,,,,, -28900,5.4345355,4.017138,,,,,,,,,,,,,, -29000,5.016843,3.8265827,,,,,,,,,,,,,, -29100,11.05119,3.8809946,,,,,,,,,,,,,, -29200,4.1003666,3.8374457,,,,,,,,,,,,,, -29300,4.0710993,3.8358033,,,,,,,,,,,,,, -29400,3.4880676,3.8371868,,,,,,,,,,,,,, -29500,9.529087,4.1671247,,,,,,,,,,,,,, -29600,7.605472,4.1196823,,,,,,,,,,,,,, -29700,4.778376,4.0277963,,,,,,,,,,,,,, -29800,3.392955,3.9394233,,,,,,,,,,,,,, -29900,3.678391,3.9557762,,,,,,,,,,,,,, -30000,3.345279,3.8883355,,,,,,,,,,,,,, -30100,6.906122,3.958286,,,,,,,,,,,,,, -30200,4.6704025,3.8556445,,,,,,,,,,,,,, -30300,5.8046246,3.8618999,,,,,,,,,,,,,, -30400,5.695255,3.7948425,,,,,,,,,,,,,, -30500,6.758835,3.8469481,,,,,,,,,,,,,, -30600,8.103054,4.009036,,,,,,,,,,,,,, -30700,4.641742,3.8836963,,,,,,,,,,,,,, -30800,6.360194,3.9035304,,,,,,,,,,,,,, -30900,13.155696,3.9695382,,,,,,,,,,,,,, -31000,6.139472,3.9184349,,,,,,,,,,,,,, -31100,6.638228,3.9322586,,,,,,,,,,,,,, -31200,5.009262,3.8185034,,,,,,,,,,,,,, -31300,7.49821,3.8807251,,,,,,,,,,,,,, -31400,5.5738688,3.795884,,,,,,,,,,,,,, -31500,3.5929406,3.8185697,,,,,,,,,,,,,, -31600,10.701888,3.8235369,,,,,,,,,,,,,, -31700,4.7191963,3.9521375,,,,,,,,,,,,,, -31800,135.08714,4.2451086,,,,,,,,,,,,,, -31900,8.264283,3.883503,,,,,,,,,,,,,, -32000,5.289908,3.954288,,,,,,,,,,,,,, -32053,,,0.001434949,8.892026,0.00172,8.847008,50000.0,0.0014000001,8.848667,10000.0,12032.575147390366,12233.807253837584,12032.575147390366,200.61437606811523,0.3487226963043213,0.0 -32100,7.591611,3.8695662,,,,,,,,,,,,,, -32200,6.857138,3.9086103,,,,,,,,,,,,,, -32300,5.195123,3.9733908,,,,,,,,,,,,,, -32400,4.338388,3.8967555,,,,,,,,,,,,,, -32500,5.395411,3.8643744,,,,,,,,,,,,,, -32600,10.14653,3.9275818,,,,,,,,,,,,,, -32700,4.221452,3.7896986,,,,,,,,,,,,,, -32800,3.6692972,3.7686641,,,,,,,,,,,,,, -32900,33.26421,3.8806403,,,,,,,,,,,,,, -33000,3.8690147,3.7900949,,,,,,,,,,,,,, -33100,5.1266336,3.7427845,,,,,,,,,,,,,, -33200,3.9918678,3.7234888,,,,,,,,,,,,,, -33300,3.6903114,3.862156,,,,,,,,,,,,,, -33400,5.708204,3.8487425,,,,,,,,,,,,,, -33500,8.163433,3.9448986,,,,,,,,,,,,,, -33600,6.8696256,3.960354,,,,,,,,,,,,,, -33700,7.287049,3.941011,,,,,,,,,,,,,, -33800,3.1191926,3.7862105,,,,,,,,,,,,,, -33900,2.6103685,3.7794127,,,,,,,,,,,,,, -34000,7.537171,3.9351223,,,,,,,,,,,,,, -34100,4.395468,3.6783266,,,,,,,,,,,,,, -34200,4.2732544,3.9681234,,,,,,,,,,,,,, -34300,5.3038006,3.870523,,,,,,,,,,,,,, -34400,3.0871537,3.6702046,,,,,,,,,,,,,, -34500,4.371456,3.7387981,,,,,,,,,,,,,, -34600,4.37798,3.8974183,,,,,,,,,,,,,, -34700,6.7170835,3.9198442,,,,,,,,,,,,,, -34800,5.078334,3.7783456,,,,,,,,,,,,,, -34900,6.9958487,3.915125,,,,,,,,,,,,,, -35000,6.5691805,3.9535413,,,,,,,,,,,,,, -35100,5.5531554,3.7505808,,,,,,,,,,,,,, -35200,4.052001,3.7976549,,,,,,,,,,,,,, -35300,8.910617,3.861353,,,,,,,,,,,,,, -35400,5.3544693,3.7187104,,,,,,,,,,,,,, -35500,5.268887,3.851135,,,,,,,,,,,,,, -35600,6.603231,3.7419064,,,,,,,,,,,,,, -35700,8.0082245,3.911096,,,,,,,,,,,,,, -35800,5.1276345,3.8945794,,,,,,,,,,,,,, -35900,4.42219,3.9026978,,,,,,,,,,,,,, -36000,6.4024253,3.9707599,,,,,,,,,,,,,, -36100,5.630109,3.8156404,,,,,,,,,,,,,, -36200,3.0079281,3.8993645,,,,,,,,,,,,,, -36300,10.671001,3.8400176,,,,,,,,,,,,,, -36400,3.921409,3.7437782,,,,,,,,,,,,,, -36500,6.9502754,3.8418236,,,,,,,,,,,,,, -36600,7.770076,3.9480374,,,,,,,,,,,,,, -36700,4.5850897,3.8245914,,,,,,,,,,,,,, -36800,6.174434,3.887665,,,,,,,,,,,,,, -36900,6.2235684,3.8175545,,,,,,,,,,,,,, -37000,5.6397552,3.7557416,,,,,,,,,,,,,, -37100,5.1756544,3.845253,,,,,,,,,,,,,, -37200,5.509775,3.7919765,,,,,,,,,,,,,, -37300,5.1527834,3.7999766,,,,,,,,,,,,,, -37400,5.3228908,3.8286974,,,,,,,,,,,,,, -37430,,,0.0015345982,8.616239,0.0015199999,8.605657,50000.0,0.0015,8.609212,10000.0,14028.56815481186,14252.74825811386,14028.56815481186,223.4686014652252,0.3977727890014648,0.0 -37500,5.6014605,3.8313503,,,,,,,,,,,,,, -37600,4.705033,3.8481631,,,,,,,,,,,,,, -37700,6.892469,3.645044,,,,,,,,,,,,,, -37800,4.984472,3.7173934,,,,,,,,,,,,,, -37900,7.03322,3.9073634,,,,,,,,,,,,,, -38000,6.568992,3.9831567,,,,,,,,,,,,,, -38100,5.034747,3.749309,,,,,,,,,,,,,, -38200,6.766976,3.845902,,,,,,,,,,,,,, -38300,3.7803938,3.7108011,,,,,,,,,,,,,, -38400,4.603743,3.7370963,,,,,,,,,,,,,, -38500,3.1511352,3.5548015,,,,,,,,,,,,,, -38600,4.937529,3.9110606,,,,,,,,,,,,,, -38700,5.0627303,3.7136924,,,,,,,,,,,,,, -38800,3.4589257,3.7391706,,,,,,,,,,,,,, -38900,6.5525513,3.8391783,,,,,,,,,,,,,, -39000,4.7897716,3.7391198,,,,,,,,,,,,,, -39100,4.0563245,3.7620628,,,,,,,,,,,,,, -39200,14.808853,3.6549141,,,,,,,,,,,,,, -39300,3.6112406,3.8304412,,,,,,,,,,,,,, -39400,23.429544,3.8279743,,,,,,,,,,,,,, -39500,3.5358815,3.6174603,,,,,,,,,,,,,, -39600,4.38195,3.746718,,,,,,,,,,,,,, -39700,3.6870372,3.753665,,,,,,,,,,,,,, -39800,14.768377,3.8182902,,,,,,,,,,,,,, -39900,4.571193,3.784555,,,,,,,,,,,,,, -40000,5.682659,3.8361135,,,,,,,,,,,,,, -40100,2.5467467,3.7267065,,,,,,,,,,,,,, -40200,4.4758973,3.7318711,,,,,,,,,,,,,, -40300,27.17862,3.9489746,,,,,,,,,,,,,, -40400,5.570227,3.945172,,,,,,,,,,,,,, -40500,5.4100623,3.8554316,,,,,,,,,,,,,, -40600,8.718293,3.6932645,,,,,,,,,,,,,, -40700,5.832206,3.7260723,,,,,,,,,,,,,, -40800,4.978761,3.89892,,,,,,,,,,,,,, -40900,5.1433806,3.9319139,,,,,,,,,,,,,, -41000,3.373328,3.7544048,,,,,,,,,,,,,, -41100,4.638525,3.748985,,,,,,,,,,,,,, -41200,3.9721177,3.6815019,,,,,,,,,,,,,, -41300,4.114133,3.7549343,,,,,,,,,,,,,, -41400,3.9547038,3.680047,,,,,,,,,,,,,, -41500,3.8075297,3.8150434,,,,,,,,,,,,,, -41600,5.3538637,3.7212415,,,,,,,,,,,,,, -41700,8.781359,4.0569987,,,,,,,,,,,,,, -41800,3.967826,3.9165444,,,,,,,,,,,,,, -41900,5.200409,3.8966637,,,,,,,,,,,,,, -42000,5.7449675,3.762511,,,,,,,,,,,,,, -42100,4.745189,3.6860383,,,,,,,,,,,,,, -42200,4.5447683,3.7291493,,,,,,,,,,,,,, -42300,4.931208,3.7591326,,,,,,,,,,,,,, -42400,3.714072,3.584532,,,,,,,,,,,,,, -42500,4.775662,3.6616747,,,,,,,,,,,,,, -42600,3.6071515,3.5559769,,,,,,,,,,,,,, -42700,6.350673,3.9236867,,,,,,,,,,,,,, -42800,4.080056,3.7406719,,,,,,,,,,,,,, -42810,,,0.0016143175,8.523104,0.0016399999,8.457493,50000.0,0.0013000001,8.466099,10000.0,16024.628784179688,16271.091526985168,16024.628784179688,245.62259697914124,0.481757640838623,0.0 -42900,3.233359,3.791287,,,,,,,,,,,,,, -43000,8.144944,3.7797966,,,,,,,,,,,,,, -43100,6.860063,3.804646,,,,,,,,,,,,,, -43200,5.361058,3.7727442,,,,,,,,,,,,,, -43300,7.4865403,3.7987003,,,,,,,,,,,,,, -43400,4.0709343,3.7418737,,,,,,,,,,,,,, -43500,5.784103,3.8518972,,,,,,,,,,,,,, -43600,6.7711973,3.8813503,,,,,,,,,,,,,, -43700,4.172581,3.6968985,,,,,,,,,,,,,, -43800,6.01177,3.7050116,,,,,,,,,,,,,, -43900,6.121245,3.774964,,,,,,,,,,,,,, -44000,2.8626075,3.7536752,,,,,,,,,,,,,, -44100,3.6845698,3.812346,,,,,,,,,,,,,, -44200,4.8793197,3.7568653,,,,,,,,,,,,,, -44300,6.3242903,3.8674452,,,,,,,,,,,,,, -44400,3.4310787,3.7691636,,,,,,,,,,,,,, -44500,6.1671224,3.7819505,,,,,,,,,,,,,, -44600,6.7096405,3.6179826,,,,,,,,,,,,,, -44700,4.3758807,3.7704706,,,,,,,,,,,,,, -44800,6.2746224,3.771899,,,,,,,,,,,,,, -44900,5.318879,3.9112592,,,,,,,,,,,,,, -45000,4.737287,3.6933684,,,,,,,,,,,,,, -45100,5.210596,3.843363,,,,,,,,,,,,,, -45200,5.88041,3.6363156,,,,,,,,,,,,,, -45300,4.370427,3.7357852,,,,,,,,,,,,,, -45400,3.740309,3.7522786,,,,,,,,,,,,,, -45500,7.4854712,3.8026457,,,,,,,,,,,,,, -45600,6.4128413,3.7992973,,,,,,,,,,,,,, -45700,5.7779484,3.8500192,,,,,,,,,,,,,, -45800,3.4584332,3.6915455,,,,,,,,,,,,,, -45900,4.392711,3.7696352,,,,,,,,,,,,,, -46000,3.6281676,3.6314585,,,,,,,,,,,,,, -46100,4.8753405,3.9160802,,,,,,,,,,,,,, -46200,8.391311,3.8987525,,,,,,,,,,,,,, -46300,7.641426,3.8653035,,,,,,,,,,,,,, -46400,2.728656,3.7048788,,,,,,,,,,,,,, -46500,3.2073116,3.5706532,,,,,,,,,,,,,, -46600,7.3689556,3.9729738,,,,,,,,,,,,,, -46700,3.8383431,3.6635785,,,,,,,,,,,,,, -46800,4.5454264,3.7427726,,,,,,,,,,,,,, -46900,4.464068,3.812859,,,,,,,,,,,,,, -47000,4.4735637,3.7726932,,,,,,,,,,,,,, -47100,4.298506,3.765564,,,,,,,,,,,,,, -47200,3.928154,3.78124,,,,,,,,,,,,,, -47300,4.241397,3.6908484,,,,,,,,,,,,,, -47400,4.854414,3.7291965,,,,,,,,,,,,,, -47500,9.73472,3.884656,,,,,,,,,,,,,, -47600,6.3949823,3.7754965,,,,,,,,,,,,,, -47700,2.9654615,3.7779522,,,,,,,,,,,,,, -47800,4.183248,3.7936106,,,,,,,,,,,,,, -47900,4.5232987,3.8574052,,,,,,,,,,,,,, -48000,5.149483,3.505331,,,,,,,,,,,,,, -48100,6.391189,3.8798933,,,,,,,,,,,,,, -48190,,,0.0017737563,8.355035,0.0013799999,8.319946,50000.0,0.0015,8.331606,10000.0,18021.00621008873,18289.564914941788,18021.00621008873,267.5842914581299,0.5713906288146973,0.0 -48200,3.9567606,3.73705,,,,,,,,,,,,,, -48300,4.7399845,3.9334893,,,,,,,,,,,,,, -48400,4.133525,3.6690865,,,,,,,,,,,,,, -48500,4.1901546,3.6990604,,,,,,,,,,,,,, -48600,5.445069,3.5764153,,,,,,,,,,,,,, -48700,5.51976,3.7048001,,,,,,,,,,,,,, -48800,4.202644,3.6550016,,,,,,,,,,,,,, -48900,4.585452,3.66802,,,,,,,,,,,,,, -49000,5.417252,3.813417,,,,,,,,,,,,,, -49100,3.8677702,3.7248406,,,,,,,,,,,,,, -49200,6.149771,3.9360223,,,,,,,,,,,,,, -49300,3.756167,3.6336176,,,,,,,,,,,,,, -49400,3.7987118,3.7790928,,,,,,,,,,,,,, -49500,3.8745131,3.5722938,,,,,,,,,,,,,, -49600,4.035454,3.5715647,,,,,,,,,,,,,, -49700,7.4084954,3.8669395,,,,,,,,,,,,,, -49800,6.31718,3.955976,,,,,,,,,,,,,, -49900,4.681115,3.795253,,,,,,,,,,,,,, -50000,4.70121,3.7560732,,,,,,,,,,,,,, -50100,5.1367226,3.6229777,,,,,,,,,,,,,, -50200,3.34185,3.6207418,,,,,,,,,,,,,, -50300,6.235125,3.8704543,,,,,,,,,,,,,, -50400,4.1189423,3.7074592,,,,,,,,,,,,,, -50500,4.6139836,3.649982,,,,,,,,,,,,,, -50600,3.2703905,3.6830068,,,,,,,,,,,,,, -50700,3.6553051,3.5095954,,,,,,,,,,,,,, -50800,3.1986117,3.6470408,,,,,,,,,,,,,, -50900,2.6846032,3.5843654,,,,,,,,,,,,,, -51000,3.213009,3.6980884,,,,,,,,,,,,,, -51100,5.2780604,3.7652197,,,,,,,,,,,,,, -51200,6.7457833,3.774393,,,,,,,,,,,,,, -51300,3.9143586,3.7409496,,,,,,,,,,,,,, -51400,4.522526,3.6831398,,,,,,,,,,,,,, -51500,5.5410733,3.6375947,,,,,,,,,,,,,, -51600,4.519454,3.7371824,,,,,,,,,,,,,, -51700,5.549562,3.875846,,,,,,,,,,,,,, -51800,5.521986,3.7139416,,,,,,,,,,,,,, -51900,11.28245,3.696845,,,,,,,,,,,,,, -52000,4.926649,3.6357481,,,,,,,,,,,,,, -52100,4.0208497,3.691379,,,,,,,,,,,,,, -52200,4.3511,3.8008099,,,,,,,,,,,,,, -52300,3.7141848,3.800838,,,,,,,,,,,,,, -52400,5.326634,3.6528478,,,,,,,,,,,,,, -52500,4.705577,3.8540375,,,,,,,,,,,,,, -52600,5.3616586,3.6842418,,,,,,,,,,,,,, -52700,4.8636875,3.9323952,,,,,,,,,,,,,, -52800,6.67047,3.7469368,,,,,,,,,,,,,, -52900,4.01569,3.6574667,,,,,,,,,,,,,, -53000,9.742966,4.1271515,,,,,,,,,,,,,, -53100,5.422812,3.622478,,,,,,,,,,,,,, -53200,5.299853,3.6691437,,,,,,,,,,,,,, -53300,3.746199,3.795694,,,,,,,,,,,,,, -53400,10.378281,4.048086,,,,,,,,,,,,,, -53500,4.1919894,3.8442283,,,,,,,,,,,,,, -53567,,,0.0015744579,8.254096,0.0013799999,8.238354,50000.0,0.0011,8.248967,10000.0,20017.06850886345,20307.42605495453,20017.06850886345,289.2678849697113,0.6422865390777588,0.0 -53600,5.102597,3.6819746,,,,,,,,,,,,,, -53700,4.544042,3.730662,,,,,,,,,,,,,, -53800,4.3056383,3.7614489,,,,,,,,,,,,,, -53900,4.1343365,3.753978,,,,,,,,,,,,,, -54000,3.4224353,3.7242408,,,,,,,,,,,,,, -54100,4.3304024,3.8296933,,,,,,,,,,,,,, -54200,4.1718082,3.522817,,,,,,,,,,,,,, -54300,4.583864,3.9063482,,,,,,,,,,,,,, -54400,3.7564192,3.7491283,,,,,,,,,,,,,, -54500,3.2643561,3.6584184,,,,,,,,,,,,,, -54600,4.095115,3.531508,,,,,,,,,,,,,, -54700,3.138389,3.7429168,,,,,,,,,,,,,, -54800,3.5899172,3.6408021,,,,,,,,,,,,,, -54900,5.7007656,3.7326455,,,,,,,,,,,,,, -55000,5.843623,3.7705183,,,,,,,,,,,,,, -55100,5.7329383,3.7339067,,,,,,,,,,,,,, -55200,3.5471032,3.7092164,,,,,,,,,,,,,, -55300,5.172371,3.7323318,,,,,,,,,,,,,, -55400,4.113261,3.784714,,,,,,,,,,,,,, -55500,3.97589,3.5700655,,,,,,,,,,,,,, -55600,3.0925424,3.637457,,,,,,,,,,,,,, -55700,2.8553777,3.634065,,,,,,,,,,,,,, -55800,3.6194642,3.6808505,,,,,,,,,,,,,, -55900,4.6797686,3.570249,,,,,,,,,,,,,, -56000,8.178003,3.9456992,,,,,,,,,,,,,, -56100,5.8561854,3.7673132,,,,,,,,,,,,,, -56200,2.7545424,3.663545,,,,,,,,,,,,,, -56300,6.551484,3.7315106,,,,,,,,,,,,,, -56400,3.697322,3.7253242,,,,,,,,,,,,,, -56500,4.882788,3.852465,,,,,,,,,,,,,, -56600,2.703216,3.7038207,,,,,,,,,,,,,, -56700,4.222612,3.8228204,,,,,,,,,,,,,, -56800,5.5856647,3.6046033,,,,,,,,,,,,,, -56900,4.4930363,3.7248306,,,,,,,,,,,,,, -57000,5.1847363,3.7492206,,,,,,,,,,,,,, -57100,6.6305685,3.6776583,,,,,,,,,,,,,, -57200,4.7986283,3.8181348,,,,,,,,,,,,,, -57300,4.7771945,3.8020163,,,,,,,,,,,,,, -57400,5.665223,3.5636992,,,,,,,,,,,,,, -57500,3.2533004,3.6566176,,,,,,,,,,,,,, -57600,7.712425,3.7218313,,,,,,,,,,,,,, -57700,3.7068686,3.5647128,,,,,,,,,,,,,, -57800,4.308841,3.6987472,,,,,,,,,,,,,, -57900,4.6167293,3.6289463,,,,,,,,,,,,,, -58000,3.591323,3.809054,,,,,,,,,,,,,, -58100,3.6830928,3.7053158,,,,,,,,,,,,,, -58200,2.936018,3.6747675,,,,,,,,,,,,,, -58300,2.8499029,3.6576042,,,,,,,,,,,,,, -58400,5.002469,3.612557,,,,,,,,,,,,,, -58500,7.891636,3.9228632,,,,,,,,,,,,,, -58600,5.153987,3.7773485,,,,,,,,,,,,,, -58700,4.5825753,3.7219775,,,,,,,,,,,,,, -58800,3.5390065,3.4685004,,,,,,,,,,,,,, -58900,4.233904,3.6904335,,,,,,,,,,,,,, -58943,,,0.0011360012,8.292785,0.0013799999,8.236893,50000.0,0.0014000001,8.245119,10000.0,22013.158291101456,22325.183495998383,22013.158291101456,310.8434724807739,0.6898555755615234,0.0 -59000,3.996221,3.659425,,,,,,,,,,,,,, -59100,6.342341,3.790187,,,,,,,,,,,,,, -59200,2.9211614,3.706529,,,,,,,,,,,,,, -59300,3.9754884,3.794591,,,,,,,,,,,,,, -59400,8.345152,3.88928,,,,,,,,,,,,,, -59500,3.6168916,3.6039095,,,,,,,,,,,,,, -59600,3.8348293,3.648517,,,,,,,,,,,,,, -59700,4.245219,3.758367,,,,,,,,,,,,,, -59800,4.286621,3.661335,,,,,,,,,,,,,, -59900,5.903714,3.6392508,,,,,,,,,,,,,, -60000,3.795126,3.710505,,,,,,,,,,,,,, -60100,3.2061152,3.6317549,,,,,,,,,,,,,, -60200,3.8096795,3.5454235,,,,,,,,,,,,,, -60300,4.2136674,3.536181,,,,,,,,,,,,,, -60400,5.51497,3.680822,,,,,,,,,,,,,, -60500,3.899279,3.6859455,,,,,,,,,,,,,, -60600,5.11689,3.7149167,,,,,,,,,,,,,, -60700,5.6317697,3.7579787,,,,,,,,,,,,,, -60800,4.0378428,3.8505619,,,,,,,,,,,,,, -60900,5.8288727,3.7450297,,,,,,,,,,,,,, -61000,5.5185375,3.8577082,,,,,,,,,,,,,, -61100,3.3579602,3.6726496,,,,,,,,,,,,,, -61200,4.966517,3.656342,,,,,,,,,,,,,, -61300,6.5440702,3.8240097,,,,,,,,,,,,,, -61400,3.3949702,3.6574292,,,,,,,,,,,,,, -61500,2.8325977,3.8428855,,,,,,,,,,,,,, -61600,7.366558,3.6677353,,,,,,,,,,,,,, -61700,5.0595117,3.7454073,,,,,,,,,,,,,, -61800,5.0647974,3.7430792,,,,,,,,,,,,,, -61900,4.390219,3.7563982,,,,,,,,,,,,,, -62000,6.8711886,3.8388832,,,,,,,,,,,,,, -62100,3.0926113,3.6251266,,,,,,,,,,,,,, -62200,4.685038,3.693298,,,,,,,,,,,,,, -62300,3.8061483,3.8138037,,,,,,,,,,,,,, -62400,3.7118244,3.7823977,,,,,,,,,,,,,, -62500,4.3120894,3.738719,,,,,,,,,,,,,, -62600,3.6746478,3.5901456,,,,,,,,,,,,,, -62700,4.3442497,3.637821,,,,,,,,,,,,,, -62800,5.202395,3.7120612,,,,,,,,,,,,,, -62900,3.2613354,3.60037,,,,,,,,,,,,,, -63000,4.5929174,3.7763655,,,,,,,,,,,,,, -63100,6.534228,3.8574407,,,,,,,,,,,,,, -63200,6.0517735,3.7247052,,,,,,,,,,,,,, -63300,3.8001313,3.7156796,,,,,,,,,,,,,, -63400,4.3997927,3.7067313,,,,,,,,,,,,,, -63500,4.888543,3.5608168,,,,,,,,,,,,,, -63600,4.5800996,3.7411032,,,,,,,,,,,,,, -63700,3.9519184,3.8315284,,,,,,,,,,,,,, -63800,3.5821202,3.6953397,,,,,,,,,,,,,, -63900,6.2379885,3.5600772,,,,,,,,,,,,,, -64000,5.7882566,3.7112365,,,,,,,,,,,,,, -64100,3.393917,3.6345448,,,,,,,,,,,,,, -64200,6.212124,3.9386034,,,,,,,,,,,,,, -64300,4.902136,3.9601932,,,,,,,,,,,,,, -64321,,,0.0013751594,8.30454,0.00134,8.2691765,50000.0,0.0012,8.277351,10000.0,24009.70991158485,24343.664563417435,24009.70991158485,332.6784863471985,0.7396748065948486,0.0 -64400,3.3284185,3.5913029,,,,,,,,,,,,,, -64500,3.373733,3.7671032,,,,,,,,,,,,,, -64600,4.0269394,3.9272354,,,,,,,,,,,,,, -64700,3.6824327,3.5889583,,,,,,,,,,,,,, -64800,5.3890615,3.9841387,,,,,,,,,,,,,, -64900,5.435727,3.7903485,,,,,,,,,,,,,, -65000,4.8724546,3.6676724,,,,,,,,,,,,,, -65100,6.054438,3.674529,,,,,,,,,,,,,, -65200,4.0000663,3.7813356,,,,,,,,,,,,,, -65300,5.1036963,3.6610203,,,,,,,,,,,,,, -65400,5.4136233,3.7009604,,,,,,,,,,,,,, -65500,6.212788,3.68082,,,,,,,,,,,,,, -65600,4.302271,3.6404734,,,,,,,,,,,,,, -65700,5.2580285,3.6237094,,,,,,,,,,,,,, -65800,6.0123653,3.9236746,,,,,,,,,,,,,, -65900,4.494995,3.894881,,,,,,,,,,,,,, -66000,5.08434,3.7999926,,,,,,,,,,,,,, -66100,4.5031667,3.7028728,,,,,,,,,,,,,, -66200,3.7355683,3.806026,,,,,,,,,,,,,, -66300,5.2257714,3.7539878,,,,,,,,,,,,,, -66400,4.777587,3.785008,,,,,,,,,,,,,, -66500,7.3730216,3.9139276,,,,,,,,,,,,,, -66600,4.196311,3.6402617,,,,,,,,,,,,,, -66700,4.0377474,3.7641482,,,,,,,,,,,,,, -66800,3.261865,3.7893577,,,,,,,,,,,,,, -66900,4.287854,3.7291353,,,,,,,,,,,,,, -67000,5.231855,3.9797254,,,,,,,,,,,,,, -67100,5.759413,3.6006603,,,,,,,,,,,,,, -67200,3.8676271,3.7739062,,,,,,,,,,,,,, -67300,4.3175097,3.5438597,,,,,,,,,,,,,, -67400,5.7390943,3.747023,,,,,,,,,,,,,, -67500,5.1437693,3.5731215,,,,,,,,,,,,,, -67600,4.1467834,3.6335893,,,,,,,,,,,,,, -67700,4.805644,3.7446074,,,,,,,,,,,,,, -67800,7.3551054,3.8195052,,,,,,,,,,,,,, -67900,3.6350331,3.7038608,,,,,,,,,,,,,, -68000,3.7977474,3.7210813,,,,,,,,,,,,,, -68100,4.364474,3.6158082,,,,,,,,,,,,,, -68200,3.6107056,3.7774186,,,,,,,,,,,,,, -68300,5.445451,3.7837834,,,,,,,,,,,,,, -68400,6.0258555,3.797514,,,,,,,,,,,,,, -68500,4.596245,3.6736765,,,,,,,,,,,,,, -68600,2.983429,3.6206512,,,,,,,,,,,,,, -68700,4.1114054,3.58536,,,,,,,,,,,,,, -68800,4.6494203,3.6546912,,,,,,,,,,,,,, -68900,4.2319546,3.7652044,,,,,,,,,,,,,, -69000,3.7322397,3.6137373,,,,,,,,,,,,,, -69100,4.546603,3.8685598,,,,,,,,,,,,,, -69200,3.3252,3.6184258,,,,,,,,,,,,,, -69300,5.5717545,3.7034116,,,,,,,,,,,,,, -69400,4.3447266,3.7390714,,,,,,,,,,,,,, -69500,3.9627001,3.6972017,,,,,,,,,,,,,, -69600,3.5485718,3.9074125,,,,,,,,,,,,,, -69699,,,0.0016541772,8.343106,0.0013799999,8.300176,50000.0,0.0013000001,8.307022,10000.0,26005.58596253395,26361.27652549744,26005.58596253395,354.2806496620178,0.8288896083831787,0.0 -69700,3.0763597,3.6340947,,,,,,,,,,,,,, -69800,3.2726645,3.7183995,,,,,,,,,,,,,, -69900,3.7268963,3.706842,,,,,,,,,,,,,, -70000,4.355092,3.6223803,,,,,,,,,,,,,, -70100,4.1190324,3.72391,,,,,,,,,,,,,, -70200,4.109454,3.7333221,,,,,,,,,,,,,, -70300,3.9767313,3.6054943,,,,,,,,,,,,,, -70400,3.7855008,3.6247163,,,,,,,,,,,,,, -70500,4.403283,3.591683,,,,,,,,,,,,,, -70600,4.0885725,3.6123512,,,,,,,,,,,,,, -70700,3.9812024,3.6567602,,,,,,,,,,,,,, -70800,4.877323,3.7366748,,,,,,,,,,,,,, -70900,5.061778,3.6422448,,,,,,,,,,,,,, -71000,5.0544257,3.7712371,,,,,,,,,,,,,, -71100,4.6260424,3.643805,,,,,,,,,,,,,, -71200,3.695007,3.6665251,,,,,,,,,,,,,, -71300,4.119859,3.734808,,,,,,,,,,,,,, -71400,5.016264,3.8335338,,,,,,,,,,,,,, -71500,3.450362,3.7055655,,,,,,,,,,,,,, -71600,4.1726995,3.6913748,,,,,,,,,,,,,, -71700,3.6885657,3.7924755,,,,,,,,,,,,,, -71800,5.224075,3.6851997,,,,,,,,,,,,,, -71900,4.763123,3.7300482,,,,,,,,,,,,,, -72000,4.5464926,3.638176,,,,,,,,,,,,,, -72100,4.336789,3.634328,,,,,,,,,,,,,, -72200,5.337066,3.7263005,,,,,,,,,,,,,, -72300,4.312495,3.7705479,,,,,,,,,,,,,, -72400,5.2340174,3.7015874,,,,,,,,,,,,,, -72500,16.160307,3.6449723,,,,,,,,,,,,,, -72600,4.1231313,3.782938,,,,,,,,,,,,,, -72700,3.792717,3.5429924,,,,,,,,,,,,,, -72800,3.6102245,3.7499983,,,,,,,,,,,,,, -72900,5.9347835,3.8544497,,,,,,,,,,,,,, -73000,3.99551,3.6853924,,,,,,,,,,,,,, -73100,3.9729502,3.7176409,,,,,,,,,,,,,, -73200,5.499223,3.723164,,,,,,,,,,,,,, -73300,3.713052,3.6006546,,,,,,,,,,,,,, -73400,3.92067,3.5712893,,,,,,,,,,,,,, -73500,6.6406074,3.6906126,,,,,,,,,,,,,, -73600,3.4652507,3.6673465,,,,,,,,,,,,,, -73700,5.465667,3.7879238,,,,,,,,,,,,,, -73800,5.2051764,3.807529,,,,,,,,,,,,,, -73900,4.6203413,3.7000291,,,,,,,,,,,,,, -74000,16.703053,3.889381,,,,,,,,,,,,,, -74100,3.995649,3.7222993,,,,,,,,,,,,,, -74200,5.209008,3.7081943,,,,,,,,,,,,,, -74300,3.2752733,3.574119,,,,,,,,,,,,,, -74400,4.0519595,3.8073206,,,,,,,,,,,,,, -74500,4.077787,3.6988323,,,,,,,,,,,,,, -74600,6.8197,4.015129,,,,,,,,,,,,,, -74700,5.7109823,3.8277967,,,,,,,,,,,,,, -74800,10.868199,3.9117017,,,,,,,,,,,,,, -74900,4.0015106,3.7430801,,,,,,,,,,,,,, -75000,4.182785,3.8202,,,,,,,,,,,,,, -75079,,,0.0012755102,8.378688,0.0015,8.342849,50000.0,0.0015,8.351302,10000.0,28001.82784414292,28378.600981235504,28001.82784414292,375.2546329498291,0.8931126594543457,0.0 -75100,4.6242394,3.8771315,,,,,,,,,,,,,, -75200,4.6780167,3.7693422,,,,,,,,,,,,,, -75300,4.21187,3.634479,,,,,,,,,,,,,, -75400,8.4983425,3.8081367,,,,,,,,,,,,,, -75500,5.156115,3.8330145,,,,,,,,,,,,,, -75600,4.7705503,3.8850787,,,,,,,,,,,,,, -75700,6.722166,3.8560786,,,,,,,,,,,,,, -75800,13.116931,4.072983,,,,,,,,,,,,,, -75900,4.106633,3.7265596,,,,,,,,,,,,,, -76000,10.040435,3.7526584,,,,,,,,,,,,,, -76100,4.963212,3.6611538,,,,,,,,,,,,,, -76200,4.213085,3.8204396,,,,,,,,,,,,,, -76300,4.4077096,3.74506,,,,,,,,,,,,,, -76400,4.204818,3.7657986,,,,,,,,,,,,,, -76500,4.797326,3.652558,,,,,,,,,,,,,, -76600,9.651666,3.9150932,,,,,,,,,,,,,, -76700,4.23635,3.6699443,,,,,,,,,,,,,, -76800,8.727786,3.8864272,,,,,,,,,,,,,, -76900,4.4273987,3.6828876,,,,,,,,,,,,,, -77000,3.7011256,3.7921543,,,,,,,,,,,,,, -77100,3.8807344,3.7487175,,,,,,,,,,,,,, -77200,4.657239,3.7088683,,,,,,,,,,,,,, -77300,4.568222,3.7467177,,,,,,,,,,,,,, -77400,3.5019302,3.7731433,,,,,,,,,,,,,, -77500,5.8318963,3.6897573,,,,,,,,,,,,,, -77600,5.9768066,3.6930666,,,,,,,,,,,,,, -77700,5.286203,3.5832589,,,,,,,,,,,,,, -77800,6.9916763,3.71275,,,,,,,,,,,,,, -77900,7.630866,3.9102893,,,,,,,,,,,,,, -78000,3.68612,3.6625662,,,,,,,,,,,,,, -78100,3.6448247,3.7135496,,,,,,,,,,,,,, -78200,4.2014594,3.5891323,,,,,,,,,,,,,, -78300,6.301098,3.8720932,,,,,,,,,,,,,, -78400,3.6370265,3.7082758,,,,,,,,,,,,,, -78500,5.1371684,3.8048735,,,,,,,,,,,,,, -78600,4.721362,3.6827984,,,,,,,,,,,,,, -78700,4.5497026,3.8184388,,,,,,,,,,,,,, -78800,6.933742,3.7381797,,,,,,,,,,,,,, -78900,2.6900876,3.5264869,,,,,,,,,,,,,, -79000,5.1847777,3.7332544,,,,,,,,,,,,,, -79100,5.10089,3.7583928,,,,,,,,,,,,,, -79200,4.841084,3.7307878,,,,,,,,,,,,,, -79300,8.546919,3.6919632,,,,,,,,,,,,,, -79400,22.877117,3.9550338,,,,,,,,,,,,,, -79500,3.3079047,3.6714902,,,,,,,,,,,,,, -79600,3.2795196,3.663868,,,,,,,,,,,,,, -79700,3.969151,3.7125008,,,,,,,,,,,,,, -79800,4.833071,3.743774,,,,,,,,,,,,,, -79900,4.423922,3.646618,,,,,,,,,,,,,, -80000,28.221586,3.8373697,,,,,,,,,,,,,, -80100,3.8819625,3.6733217,,,,,,,,,,,,,, -80200,4.2936563,3.6577265,,,,,,,,,,,,,, -80300,4.4273825,3.768567,,,,,,,,,,,,,, -80400,4.206446,3.610547,,,,,,,,,,,,,, -80458,,,0.0012555803,8.448164,0.00148,8.408929,50000.0,0.0017000001,8.418403,10000.0,29998.31781888008,30396.9351439476,29998.31781888008,396.9780945777893,0.9690923690795898,0.0 -80500,4.3604817,3.6706944,,,,,,,,,,,,,, -80600,3.7497292,3.5451212,,,,,,,,,,,,,, -80700,4.4072013,3.612948,,,,,,,,,,,,,, -80800,9.745925,4.086563,,,,,,,,,,,,,, -80900,5.2949805,3.892077,,,,,,,,,,,,,, -81000,5.3424587,3.8992424,,,,,,,,,,,,,, -81100,4.0884304,3.8405552,,,,,,,,,,,,,, -81200,7.48768,3.7788117,,,,,,,,,,,,,, -81300,3.3334348,3.6949148,,,,,,,,,,,,,, -81400,4.1387525,3.7388055,,,,,,,,,,,,,, -81500,3.3276384,3.6580677,,,,,,,,,,,,,, -81600,5.316972,3.781006,,,,,,,,,,,,,, -81700,4.3671255,3.615456,,,,,,,,,,,,,, -81800,4.582855,3.6619382,,,,,,,,,,,,,, -81900,4.655253,3.7641082,,,,,,,,,,,,,, -82000,4.6851897,3.7448974,,,,,,,,,,,,,, -82100,3.6709967,3.6616235,,,,,,,,,,,,,, -82200,6.2347946,3.8002195,,,,,,,,,,,,,, -82300,13.550301,3.621792,,,,,,,,,,,,,, -82400,3.8587499,3.6314955,,,,,,,,,,,,,, -82500,4.2534103,3.762947,,,,,,,,,,,,,, -82600,5.122409,3.8258846,,,,,,,,,,,,,, -82700,3.880105,3.8083062,,,,,,,,,,,,,, -82800,6.0307746,3.8386478,,,,,,,,,,,,,, -82900,4.6310673,3.6410317,,,,,,,,,,,,,, -83000,13.350249,3.655267,,,,,,,,,,,,,, -83100,4.3594584,3.6365476,,,,,,,,,,,,,, -83200,5.2150393,3.7967262,,,,,,,,,,,,,, -83300,3.7474353,3.715391,,,,,,,,,,,,,, -83400,3.8922644,3.6570034,,,,,,,,,,,,,, -83500,3.5506115,3.7588706,,,,,,,,,,,,,, -83600,8.767225,3.6229532,,,,,,,,,,,,,, -83700,4.5360184,3.7456276,,,,,,,,,,,,,, -83800,4.077093,3.6660898,,,,,,,,,,,,,, -83900,3.4527814,3.6512082,,,,,,,,,,,,,, -84000,5.3602085,3.6860776,,,,,,,,,,,,,, -84100,7.207712,3.952118,,,,,,,,,,,,,, -84200,3.9100406,3.7119308,,,,,,,,,,,,,, -84300,4.4492936,3.6870944,,,,,,,,,,,,,, -84400,8.033104,3.807269,,,,,,,,,,,,,, -84500,3.5776265,3.8489366,,,,,,,,,,,,,, -84600,3.4804292,3.6723943,,,,,,,,,,,,,, -84700,6.8095436,3.898816,,,,,,,,,,,,,, -84800,4.1812754,3.7341142,,,,,,,,,,,,,, -84900,4.142295,3.6755126,,,,,,,,,,,,,, -85000,4.370268,3.7085047,,,,,,,,,,,,,, -85100,8.09583,3.6847107,,,,,,,,,,,,,, -85200,5.955877,3.6833243,,,,,,,,,,,,,, -85300,4.2946515,3.8861446,,,,,,,,,,,,,, -85400,4.476201,3.7305276,,,,,,,,,,,,,, -85500,3.462543,3.8077376,,,,,,,,,,,,,, -85600,3.1786368,3.7279918,,,,,,,,,,,,,, -85700,5.984392,3.6953201,,,,,,,,,,,,,, -85800,4.936541,3.9158366,,,,,,,,,,,,,, -85840,,,0.0013950892,8.513527,0.0014599999,8.475016,50000.0,0.0019,8.485265,10000.0,31994.878091335297,32414.399368047714,31994.878091335297,417.75972962379456,1.046638011932373,0.0 -85900,3.8085,3.8174293,,,,,,,,,,,,,, -86000,5.710397,3.7118623,,,,,,,,,,,,,, -86100,4.4302754,3.6620054,,,,,,,,,,,,,, -86200,5.1661234,3.7931142,,,,,,,,,,,,,, -86300,4.067063,3.7048576,,,,,,,,,,,,,, -86400,10.74344,3.7350957,,,,,,,,,,,,,, -86500,2.915202,3.6271505,,,,,,,,,,,,,, -86600,4.3705215,3.6843042,,,,,,,,,,,,,, -86700,10.0713005,3.87465,,,,,,,,,,,,,, -86800,3.3466582,3.7135398,,,,,,,,,,,,,, -86900,4.721815,3.794293,,,,,,,,,,,,,, -87000,4.1976066,3.6491814,,,,,,,,,,,,,, -87100,3.2356842,3.6637003,,,,,,,,,,,,,, -87200,5.8492913,3.6640506,,,,,,,,,,,,,, -87300,2.703364,3.8005834,,,,,,,,,,,,,, -87400,3.140761,3.6342027,,,,,,,,,,,,,, -87500,3.3822546,3.5684772,,,,,,,,,,,,,, -87600,10.273377,3.8324249,,,,,,,,,,,,,, -87700,4.6839213,3.6868467,,,,,,,,,,,,,, -87800,5.802632,3.76715,,,,,,,,,,,,,, -87900,4.292172,3.8844566,,,,,,,,,,,,,, -88000,5.0629,3.8519406,,,,,,,,,,,,,, -88100,17.83724,3.8209224,,,,,,,,,,,,,, -88200,4.969198,3.7306075,,,,,,,,,,,,,, -88300,4.503082,3.7727485,,,,,,,,,,,,,, -88400,4.101709,3.6961248,,,,,,,,,,,,,, -88500,5.9288325,3.7562146,,,,,,,,,,,,,, -88600,3.810259,3.6426651,,,,,,,,,,,,,, -88700,6.5538235,3.8502505,,,,,,,,,,,,,, -88800,3.7759695,3.664488,,,,,,,,,,,,,, -88900,3.841606,3.7958798,,,,,,,,,,,,,, -89000,5.151311,3.7295485,,,,,,,,,,,,,, -89100,6.675685,3.8797095,,,,,,,,,,,,,, -89200,9.65925,4.1251707,,,,,,,,,,,,,, -89300,9.926823,4.1228523,,,,,,,,,,,,,, -89400,3.5175087,3.7726688,,,,,,,,,,,,,, -89500,5.9058647,3.7012491,,,,,,,,,,,,,, -89600,5.3827825,3.789619,,,,,,,,,,,,,, -89700,5.0719433,3.6829362,,,,,,,,,,,,,, -89800,4.976498,3.8719993,,,,,,,,,,,,,, -89900,5.642222,3.6912549,,,,,,,,,,,,,, -90000,3.95415,3.653887,,,,,,,,,,,,,, -90100,14.245801,3.664567,,,,,,,,,,,,,, -90200,3.993743,3.65514,,,,,,,,,,,,,, -90300,3.3769164,3.8674796,,,,,,,,,,,,,, -90400,4.024641,3.8290832,,,,,,,,,,,,,, -90500,4.10543,3.8367589,,,,,,,,,,,,,, -90600,5.207498,3.7125068,,,,,,,,,,,,,, -90700,4.796744,3.7784548,,,,,,,,,,,,,, -90800,6.2102795,3.6026187,,,,,,,,,,,,,, -90900,4.541249,3.6748042,,,,,,,,,,,,,, -91000,3.6567078,3.6694112,,,,,,,,,,,,,, -91100,3.091129,3.651654,,,,,,,,,,,,,, -91200,5.0620646,3.7096143,,,,,,,,,,,,,, -91222,,,0.0014548788,8.590439,0.00144,8.561899,50000.0,0.0017000001,8.573786,10000.0,33991.13047719002,34432.37569451332,33991.13047719002,439.3616163730621,1.124169111251831,0.0 -91300,2.547638,3.7277615,,,,,,,,,,,,,, -91400,4.4510517,3.743557,,,,,,,,,,,,,, -91500,3.278894,3.730298,,,,,,,,,,,,,, -91600,3.7270136,3.7888942,,,,,,,,,,,,,, -91700,4.723735,3.8680854,,,,,,,,,,,,,, -91800,5.5849357,3.7523327,,,,,,,,,,,,,, -91900,6.9652386,3.8298204,,,,,,,,,,,,,, -92000,3.2906754,3.8351812,,,,,,,,,,,,,, -92100,4.310475,3.8761106,,,,,,,,,,,,,, -92200,4.5705404,3.684867,,,,,,,,,,,,,, -92300,5.217611,3.8911352,,,,,,,,,,,,,, -92400,3.68007,3.6798558,,,,,,,,,,,,,, -92500,5.9094834,3.7619095,,,,,,,,,,,,,, -92600,10.775208,3.7354012,,,,,,,,,,,,,, -92700,3.2637532,3.6805398,,,,,,,,,,,,,, -92800,4.919008,3.8519971,,,,,,,,,,,,,, -92900,32.958885,3.6928515,,,,,,,,,,,,,, -93000,4.0208936,3.813185,,,,,,,,,,,,,, -93100,4.2847958,3.804291,,,,,,,,,,,,,, -93200,11.706284,3.729376,,,,,,,,,,,,,, -93300,3.4159865,3.648869,,,,,,,,,,,,,, -93400,4.351746,3.7468314,,,,,,,,,,,,,, -93500,4.4123454,3.9402564,,,,,,,,,,,,,, -93600,3.1445572,3.700891,,,,,,,,,,,,,, -93700,4.3464503,3.7459462,,,,,,,,,,,,,, -93800,2.4149592,3.771294,,,,,,,,,,,,,, -93900,4.031586,3.7586553,,,,,,,,,,,,,, -94000,3.651843,3.7202234,,,,,,,,,,,,,, -94100,5.830987,4.0011973,,,,,,,,,,,,,, -94200,8.656314,3.7309766,,,,,,,,,,,,,, -94300,3.5552535,3.7525475,,,,,,,,,,,,,, -94400,9.551026,3.8588562,,,,,,,,,,,,,, -94500,4.0110264,3.7127454,,,,,,,,,,,,,, -94600,4.6652427,3.8556402,,,,,,,,,,,,,, -94700,7.016716,3.9069908,,,,,,,,,,,,,, -94800,7.016181,3.711336,,,,,,,,,,,,,, -94900,8.367496,3.9818115,,,,,,,,,,,,,, -95000,4.5873175,3.5731997,,,,,,,,,,,,,, -95100,3.1140895,3.6465018,,,,,,,,,,,,,, -95200,3.7257106,3.6965134,,,,,,,,,,,,,, -95300,2.9257567,3.7988722,,,,,,,,,,,,,, -95400,4.1518655,3.860529,,,,,,,,,,,,,, -95500,3.6626282,3.7341032,,,,,,,,,,,,,, -95600,4.4380097,3.7850697,,,,,,,,,,,,,, -95700,4.7390103,3.829414,,,,,,,,,,,,,, -95800,7.633068,4.06688,,,,,,,,,,,,,, -95900,5.583686,3.861206,,,,,,,,,,,,,, -96000,9.669243,3.9009948,,,,,,,,,,,,,, -96100,5.557235,3.8858442,,,,,,,,,,,,,, -96200,3.537206,3.7434409,,,,,,,,,,,,,, -96300,6.238057,3.9103022,,,,,,,,,,,,,, -96400,5.585465,3.8446937,,,,,,,,,,,,,, -96500,3.8151786,3.783318,,,,,,,,,,,,,, -96600,7.198297,3.8589225,,,,,,,,,,,,,, -96604,,,0.0013751594,8.670943,0.00134,8.643026,50000.0,0.0016000001,8.656847,10000.0,35987.429832696915,36450.645072460175,35987.429832696915,461.238737821579,1.1705234050750732,0.0 -96700,10.02115,4.164205,,,,,,,,,,,,,, -96800,4.1400533,3.8750541,,,,,,,,,,,,,, -96900,6.066524,3.7650132,,,,,,,,,,,,,, -97000,22.11784,3.867489,,,,,,,,,,,,,, -97100,3.964484,3.7056916,,,,,,,,,,,,,, -97200,11.715841,3.734234,,,,,,,,,,,,,, -97300,9.084395,3.5978394,,,,,,,,,,,,,, -97400,2.9926429,3.611564,,,,,,,,,,,,,, -97500,4.084071,3.8763776,,,,,,,,,,,,,, -97600,3.447087,3.6784024,,,,,,,,,,,,,, -97700,2.899346,3.7545362,,,,,,,,,,,,,, -97800,5.158788,3.7822745,,,,,,,,,,,,,, -97900,2.6765318,3.603287,,,,,,,,,,,,,, -98000,19.483349,3.7687798,,,,,,,,,,,,,, -98100,4.1607404,3.7848563,,,,,,,,,,,,,, -98200,5.433202,3.8007228,,,,,,,,,,,,,, -98300,6.1155796,3.640069,,,,,,,,,,,,,, -98400,20.02986,3.7205515,,,,,,,,,,,,,, -98500,3.7253966,3.7649407,,,,,,,,,,,,,, -98600,14.664683,3.7501874,,,,,,,,,,,,,, -98700,4.1926465,3.6741219,,,,,,,,,,,,,, -98800,5.4085283,3.8038263,,,,,,,,,,,,,, -98900,5.3144426,3.753775,,,,,,,,,,,,,, -99000,5.616575,3.9307652,,,,,,,,,,,,,, -99100,2.1500576,3.6343374,,,,,,,,,,,,,, -99200,4.622518,3.7905889,,,,,,,,,,,,,, -99300,9.947974,3.8257842,,,,,,,,,,,,,, -99400,29.408968,3.8729877,,,,,,,,,,,,,, -99500,3.7627568,3.9777927,,,,,,,,,,,,,, -99600,6.8490314,3.868311,,,,,,,,,,,,,, -99700,2.9914587,3.7217634,,,,,,,,,,,,,, -99800,3.6575682,3.9404228,,,,,,,,,,,,,, -99900,3.743743,3.8181705,,,,,,,,,,,,,, -100000,5.213537,3.7580345,,,,,,,,,,,,,, -100100,5.7681136,3.817658,,,,,,,,,,,,,, -100200,3.1379638,3.7662516,,,,,,,,,,,,,, -100300,3.5252156,3.702547,,,,,,,,,,,,,, -100400,5.0905404,3.6574957,,,,,,,,,,,,,, -100500,2.6065629,3.6551356,,,,,,,,,,,,,, -100600,6.835861,3.848422,,,,,,,,,,,,,, -100700,4.0247374,3.7707098,,,,,,,,,,,,,, -100800,3.2217374,3.7488935,,,,,,,,,,,,,, -100900,3.2105563,3.8217697,,,,,,,,,,,,,, -101000,3.1922183,3.7300534,,,,,,,,,,,,,, -101100,4.2420454,3.7483969,,,,,,,,,,,,,, -101200,3.324865,3.7612886,,,,,,,,,,,,,, -101300,7.927113,3.8030047,,,,,,,,,,,,,, -101400,4.21424,3.6951756,,,,,,,,,,,,,, -101500,5.226307,3.6563315,,,,,,,,,,,,,, -101600,4.4793873,3.6535587,,,,,,,,,,,,,, -101700,3.8735588,3.8200595,,,,,,,,,,,,,, -101800,8.283504,3.7654133,,,,,,,,,,,,,, -101900,3.9880593,3.7171433,,,,,,,,,,,,,, -101983,,,0.0013153699,8.769457,0.0013799999,8.745812,50000.0,0.0016000001,8.761139,10000.0,37983.7277469635,38468.81688165665,37983.7277469635,483.01501297950745,1.223264217376709,0.0 -102000,3.1575124,3.657662,,,,,,,,,,,,,, -102100,4.628154,3.7482252,,,,,,,,,,,,,, -102200,2.7923036,3.7876067,,,,,,,,,,,,,, -102300,6.8345127,4.025429,,,,,,,,,,,,,, -102400,5.147302,3.829444,,,,,,,,,,,,,, -102500,3.5125365,3.90688,,,,,,,,,,,,,, -102600,4.61977,3.7374935,,,,,,,,,,,,,, -102700,3.0486178,3.717242,,,,,,,,,,,,,, -102800,4.0725813,3.8248727,,,,,,,,,,,,,, -102900,5.263498,3.8242505,,,,,,,,,,,,,, -103000,4.610568,3.8710418,,,,,,,,,,,,,, -103100,4.167436,3.8502474,,,,,,,,,,,,,, -103200,7.477087,3.8740845,,,,,,,,,,,,,, -103300,4.102345,3.7404408,,,,,,,,,,,,,, -103400,11.0699005,3.737256,,,,,,,,,,,,,, -103500,4.9087443,3.7872577,,,,,,,,,,,,,, -103600,3.2853172,3.682421,,,,,,,,,,,,,, -103700,6.016867,3.9951997,,,,,,,,,,,,,, -103800,4.769753,3.8905568,,,,,,,,,,,,,, -103900,7.3593907,3.7720356,,,,,,,,,,,,,, -104000,4.55443,3.8462288,,,,,,,,,,,,,, -104100,4.299364,3.9009738,,,,,,,,,,,,,, -104200,3.7769523,3.8696814,,,,,,,,,,,,,, -104300,5.869291,3.740556,,,,,,,,,,,,,, -104400,3.3916917,3.6460438,,,,,,,,,,,,,, -104500,2.806473,3.7474964,,,,,,,,,,,,,, -104600,3.9873009,3.789916,,,,,,,,,,,,,, -104700,5.763408,3.7551293,,,,,,,,,,,,,, -104800,4.924016,3.6920986,,,,,,,,,,,,,, -104900,3.2009275,3.7568874,,,,,,,,,,,,,, -105000,4.498651,3.8319554,,,,,,,,,,,,,, -105100,3.7132545,3.7167592,,,,,,,,,,,,,, -105200,3.3872197,3.732009,,,,,,,,,,,,,, -105300,7.5692163,3.9716988,,,,,,,,,,,,,, -105400,5.086343,3.8396227,,,,,,,,,,,,,, -105500,7.8396807,3.8229988,,,,,,,,,,,,,, -105600,5.0206213,3.665444,,,,,,,,,,,,,, -105700,5.567786,3.7844899,,,,,,,,,,,,,, -105800,5.0032883,3.8728948,,,,,,,,,,,,,, -105900,6.9761887,3.8329275,,,,,,,,,,,,,, -106000,2.7433336,3.7737575,,,,,,,,,,,,,, -106100,3.8133948,3.7213864,,,,,,,,,,,,,, -106200,18.875595,3.7386448,,,,,,,,,,,,,, -106300,7.8654795,3.9503183,,,,,,,,,,,,,, -106400,7.276898,3.7091265,,,,,,,,,,,,,, -106500,3.8480132,3.6295211,,,,,,,,,,,,,, -106600,5.056129,3.7617226,,,,,,,,,,,,,, -106700,6.690023,3.9038954,,,,,,,,,,,,,, -106800,4.426622,3.8032303,,,,,,,,,,,,,, -106900,4.5926538,3.7447698,,,,,,,,,,,,,, -107000,9.12772,3.9798007,,,,,,,,,,,,,, -107100,4.2719564,3.7565327,,,,,,,,,,,,,, -107200,4.9716687,3.6858864,,,,,,,,,,,,,, -107300,3.9434478,3.8999798,,,,,,,,,,,,,, -107362,,,0.0008569834,8.899464,0.00134,8.870979,50000.0,0.0014000001,8.886436,10000.0,39979.80666279793,40485.68164682388,39979.80666279793,503.7032513618469,1.2757625579833984,0.0 -107400,4.839107,3.8362489,,,,,,,,,,,,,, -107500,9.340968,4.1777916,,,,,,,,,,,,,, -107600,3.2588742,3.6523137,,,,,,,,,,,,,, -107700,10.935608,3.8340027,,,,,,,,,,,,,, -107800,4.2553315,3.815528,,,,,,,,,,,,,, -107900,3.2306237,3.6731112,,,,,,,,,,,,,, -108000,2.6548827,3.8348122,,,,,,,,,,,,,, -108100,3.2797813,3.72851,,,,,,,,,,,,,, -108200,3.315504,3.8000576,,,,,,,,,,,,,, -108300,5.638857,3.7928681,,,,,,,,,,,,,, -108400,3.7285564,3.6927736,,,,,,,,,,,,,, -108500,4.6391463,3.640729,,,,,,,,,,,,,, -108600,4.036069,3.6203818,,,,,,,,,,,,,, -108700,5.321406,3.7284646,,,,,,,,,,,,,, -108800,3.8763962,3.7474918,,,,,,,,,,,,,, -108900,4.12656,3.7598104,,,,,,,,,,,,,, -109000,2.948451,3.7390714,,,,,,,,,,,,,, -109100,7.1217937,3.8328485,,,,,,,,,,,,,, -109200,14.449773,3.7786837,,,,,,,,,,,,,, -109300,6.579891,3.998945,,,,,,,,,,,,,, -109400,3.9652913,3.8671813,,,,,,,,,,,,,, -109500,7.098851,3.7011862,,,,,,,,,,,,,, -109600,4.696351,3.8378808,,,,,,,,,,,,,, -109700,4.2092304,3.7194402,,,,,,,,,,,,,, -109800,3.586736,3.6804829,,,,,,,,,,,,,, -109900,4.03536,3.6561754,,,,,,,,,,,,,, -110000,3.2376575,3.6879425,,,,,,,,,,,,,, -110100,4.1240325,3.879743,,,,,,,,,,,,,, -110200,6.4197617,3.765854,,,,,,,,,,,,,, -110300,3.5797093,3.5959458,,,,,,,,,,,,,, -110400,2.766726,3.66345,,,,,,,,,,,,,, -110500,3.3109481,3.8067126,,,,,,,,,,,,,, -110600,4.097724,3.761745,,,,,,,,,,,,,, -110700,3.3275802,3.7649403,,,,,,,,,,,,,, -110800,4.8435144,3.763934,,,,,,,,,,,,,, -110900,5.9221096,3.9234076,,,,,,,,,,,,,, -111000,4.8845477,3.772979,,,,,,,,,,,,,, -111100,4.478643,3.749793,,,,,,,,,,,,,, -111200,4.743607,3.6786394,,,,,,,,,,,,,, -111300,4.344918,3.840508,,,,,,,,,,,,,, -111400,6.9159026,3.7263045,,,,,,,,,,,,,, -111500,4.796621,3.7728038,,,,,,,,,,,,,, -111600,3.7223108,3.6754766,,,,,,,,,,,,,, -111700,4.905661,3.685032,,,,,,,,,,,,,, -111800,5.36471,3.90521,,,,,,,,,,,,,, -111900,3.8588936,3.7621896,,,,,,,,,,,,,, -112000,3.2335482,3.6835656,,,,,,,,,,,,,, -112100,6.745647,3.9613497,,,,,,,,,,,,,, -112200,4.04233,3.73661,,,,,,,,,,,,,, -112300,4.8823876,3.6574185,,,,,,,,,,,,,, -112400,3.0430808,3.8217812,,,,,,,,,,,,,, -112500,3.423074,3.7725227,,,,,,,,,,,,,, -112600,3.0485926,3.7058446,,,,,,,,,,,,,, -112700,2.6994367,3.8776584,,,,,,,,,,,,,, -112744,,,0.0012157207,9.044412,0.0013199999,9.01385,50000.0,0.0013000001,9.029619,10000.0,41976.64593648911,42503.76131486893,41976.64593648911,524.8504509925842,1.324159860610962,0.0 -112800,3.157272,3.7129812,,,,,,,,,,,,,, -112900,3.7611427,3.659523,,,,,,,,,,,,,, -113000,6.650254,3.8312447,,,,,,,,,,,,,, -113100,3.5521686,3.8325486,,,,,,,,,,,,,, -113200,3.8724706,3.8451252,,,,,,,,,,,,,, -113300,4.374045,3.7740822,,,,,,,,,,,,,, -113400,8.170891,3.6765049,,,,,,,,,,,,,, -113500,3.7763908,3.6576133,,,,,,,,,,,,,, -113600,3.0424829,3.830046,,,,,,,,,,,,,, -113700,3.7534602,3.6690378,,,,,,,,,,,,,, -113800,11.98796,3.762815,,,,,,,,,,,,,, -113900,2.735117,3.7645063,,,,,,,,,,,,,, -114000,5.0754642,3.8721952,,,,,,,,,,,,,, -114100,3.0092566,3.7166271,,,,,,,,,,,,,, -114200,10.224397,3.8240023,,,,,,,,,,,,,, -114300,4.7924104,3.7556272,,,,,,,,,,,,,, -114400,2.961219,3.803156,,,,,,,,,,,,,, -114500,6.412239,3.9888124,,,,,,,,,,,,,, -114600,3.6869993,3.7691545,,,,,,,,,,,,,, -114700,3.7884545,3.8960135,,,,,,,,,,,,,, -114800,23.12278,3.9975963,,,,,,,,,,,,,, -114900,2.6148098,3.7914047,,,,,,,,,,,,,, -115000,6.2055526,3.8870473,,,,,,,,,,,,,, -115100,21.316307,3.8755717,,,,,,,,,,,,,, -115200,8.036631,4.068156,,,,,,,,,,,,,, -115300,13.400393,3.944716,,,,,,,,,,,,,, -115400,3.1694944,3.906153,,,,,,,,,,,,,, -115500,3.6735964,3.9165168,,,,,,,,,,,,,, -115600,3.6033463,3.8713717,,,,,,,,,,,,,, -115700,5.875718,4.098073,,,,,,,,,,,,,, -115800,6.8495913,3.860514,,,,,,,,,,,,,, -115900,3.280614,3.7344713,,,,,,,,,,,,,, -116000,5.065333,3.8415809,,,,,,,,,,,,,, -116100,6.075076,3.8946724,,,,,,,,,,,,,, -116200,5.4394517,3.8947587,,,,,,,,,,,,,, -116300,2.5137222,3.7099814,,,,,,,,,,,,,, -116400,8.748765,3.8077517,,,,,,,,,,,,,, -116500,4.779448,3.8911273,,,,,,,,,,,,,, -116600,6.2974052,3.9183784,,,,,,,,,,,,,, -116700,22.318897,3.8897395,,,,,,,,,,,,,, -116800,4.0187354,3.7296443,,,,,,,,,,,,,, -116900,3.1749613,3.7554755,,,,,,,,,,,,,, -117000,8.572327,3.6103604,,,,,,,,,,,,,, -117100,6.9260435,3.833392,,,,,,,,,,,,,, -117200,3.3317425,3.7946835,,,,,,,,,,,,,, -117300,3.7748964,3.865444,,,,,,,,,,,,,, -117400,3.7021601,3.773489,,,,,,,,,,,,,, -117500,3.6013558,3.799815,,,,,,,,,,,,,, -117600,6.136844,3.8516488,,,,,,,,,,,,,, -117700,4.409923,3.7458818,,,,,,,,,,,,,, -117800,3.4777715,3.8337727,,,,,,,,,,,,,, -117900,6.718494,3.694402,,,,,,,,,,,,,, -118000,3.2834194,3.8344197,,,,,,,,,,,,,, -118100,4.201541,3.7724104,,,,,,,,,,,,,, -118126,,,0.0013153699,9.158456,0.0013,9.12994,50000.0,0.0012,9.145188,10000.0,43972.96816420555,44522.39081430435,43972.96816420555,547.0459926128387,1.3902931213378906,0.0 -118200,3.5223742,3.894187,,,,,,,,,,,,,, -118300,3.4716122,3.6754012,,,,,,,,,,,,,, -118400,4.132032,3.9381073,,,,,,,,,,,,,, -118500,5.380265,3.8550587,,,,,,,,,,,,,, -118600,5.3321266,3.7392998,,,,,,,,,,,,,, -118700,4.5536766,3.8869376,,,,,,,,,,,,,, -118800,5.2444134,3.7666993,,,,,,,,,,,,,, -118900,3.6728542,3.815713,,,,,,,,,,,,,, -119000,4.874372,3.8619142,,,,,,,,,,,,,, -119100,11.357101,3.7488103,,,,,,,,,,,,,, -119200,20.027647,3.9073179,,,,,,,,,,,,,, -119300,6.8306055,3.94164,,,,,,,,,,,,,, -119400,3.2365978,3.7957928,,,,,,,,,,,,,, -119500,8.095415,3.7938917,,,,,,,,,,,,,, -119600,5.3202357,3.836544,,,,,,,,,,,,,, -119700,34.193073,4.0908093,,,,,,,,,,,,,, -119800,8.464476,4.021948,,,,,,,,,,,,,, -119900,4.1358953,3.9543302,,,,,,,,,,,,,, -120000,6.8605833,3.8960545,,,,,,,,,,,,,, -120100,5.276606,3.867667,,,,,,,,,,,,,, -120200,7.493173,3.7432568,,,,,,,,,,,,,, -120300,3.5400198,3.8375542,,,,,,,,,,,,,, -120400,3.9103856,3.8674383,,,,,,,,,,,,,, -120500,3.7919848,3.7902513,,,,,,,,,,,,,, -120600,5.063653,3.6972976,,,,,,,,,,,,,, -120700,11.456695,3.8790824,,,,,,,,,,,,,, -120800,4.384346,3.9022663,,,,,,,,,,,,,, -120900,3.0391445,3.6180804,,,,,,,,,,,,,, -121000,6.029488,3.8754377,,,,,,,,,,,,,, -121100,4.942533,3.79937,,,,,,,,,,,,,, -121200,19.822945,3.9309742,,,,,,,,,,,,,, -121300,3.5402656,3.815024,,,,,,,,,,,,,, -121400,4.3590097,3.8298397,,,,,,,,,,,,,, -121500,5.4364505,3.9049067,,,,,,,,,,,,,, -121600,5.0002785,3.8602164,,,,,,,,,,,,,, -121700,5.150326,3.838139,,,,,,,,,,,,,, -121800,4.9703364,3.7232947,,,,,,,,,,,,,, -121900,3.70392,3.672414,,,,,,,,,,,,,, -122000,7.048591,3.9133964,,,,,,,,,,,,,, -122100,4.068853,3.9169888,,,,,,,,,,,,,, -122200,3.967652,3.779396,,,,,,,,,,,,,, -122300,10.353907,3.8772187,,,,,,,,,,,,,, -122400,3.6035285,3.7279267,,,,,,,,,,,,,, -122500,3.9310503,3.9201198,,,,,,,,,,,,,, -122600,8.407293,3.924677,,,,,,,,,,,,,, -122700,13.110547,3.9743264,,,,,,,,,,,,,, -122800,4.8760576,3.8353102,,,,,,,,,,,,,, -122900,3.495871,3.8266702,,,,,,,,,,,,,, -123000,7.4928946,3.9273524,,,,,,,,,,,,,, -123100,4.180525,3.8668952,,,,,,,,,,,,,, -123200,3.418832,3.729317,,,,,,,,,,,,,, -123300,5.2138267,3.8891559,,,,,,,,,,,,,, -123400,3.4474113,3.6259475,,,,,,,,,,,,,, -123500,5.185448,3.8281808,,,,,,,,,,,,,, -123505,,,0.0012755102,9.268085,0.00128,9.239881,50000.0,0.0012,9.25315,10000.0,45969.52896118164,46542.27374839783,45969.52896118164,570.2733144760132,1.4398610591888428,0.0 -123600,3.77449,3.7466297,,,,,,,,,,,,,, -123700,4.574336,3.7731404,,,,,,,,,,,,,, -123800,3.6368647,3.8211827,,,,,,,,,,,,,, -123900,4.785826,3.7183294,,,,,,,,,,,,,, -124000,5.1142583,3.7770953,,,,,,,,,,,,,, -124100,3.9237792,3.8446398,,,,,,,,,,,,,, -124200,9.965909,3.7087083,,,,,,,,,,,,,, -124300,11.482971,3.7525687,,,,,,,,,,,,,, -124400,4.8776355,3.781496,,,,,,,,,,,,,, -124500,3.4105527,3.6849234,,,,,,,,,,,,,, -124600,5.5804815,3.829556,,,,,,,,,,,,,, -124700,11.496072,3.9199498,,,,,,,,,,,,,, -124800,2.9043767,3.8523471,,,,,,,,,,,,,, -124900,5.7016454,3.972629,,,,,,,,,,,,,, -125000,4.525282,3.873972,,,,,,,,,,,,,, -125100,4.9940395,3.8486784,,,,,,,,,,,,,, -125200,6.0591154,3.7156446,,,,,,,,,,,,,, -125300,3.8805113,3.8669333,,,,,,,,,,,,,, -125400,5.980544,3.883273,,,,,,,,,,,,,, -125500,33.823524,4.0418396,,,,,,,,,,,,,, -125600,5.1693516,3.764006,,,,,,,,,,,,,, -125700,4.4557433,3.8555107,,,,,,,,,,,,,, -125800,2.555472,3.8115363,,,,,,,,,,,,,, -125900,3.4355767,3.9966335,,,,,,,,,,,,,, -126000,2.8973572,3.751844,,,,,,,,,,,,,, -126100,4.036828,3.726747,,,,,,,,,,,,,, -126200,4.3349543,3.7973719,,,,,,,,,,,,,, -126300,3.0264647,3.9016478,,,,,,,,,,,,,, -126400,3.5046937,3.843339,,,,,,,,,,,,,, -126500,6.3990088,4.0184813,,,,,,,,,,,,,, -126600,24.828491,3.960604,,,,,,,,,,,,,, -126700,6.7600293,3.9224925,,,,,,,,,,,,,, -126800,5.6091313,3.8782248,,,,,,,,,,,,,, -126900,4.1106153,3.8601446,,,,,,,,,,,,,, -127000,4.831293,3.8351736,,,,,,,,,,,,,, -127100,5.19489,3.8942163,,,,,,,,,,,,,, -127200,5.8501697,3.7962632,,,,,,,,,,,,,, -127300,5.93777,3.9155009,,,,,,,,,,,,,, -127400,4.625712,3.831863,,,,,,,,,,,,,, -127500,3.6980538,3.967484,,,,,,,,,,,,,, -127600,3.8263996,3.9507976,,,,,,,,,,,,,, -127700,4.245906,3.82312,,,,,,,,,,,,,, -127800,4.6896095,3.83208,,,,,,,,,,,,,, -127900,2.7255135,3.7265549,,,,,,,,,,,,,, -128000,3.479273,3.9693122,,,,,,,,,,,,,, -128100,3.9251673,3.876958,,,,,,,,,,,,,, -128200,5.952557,3.742822,,,,,,,,,,,,,, -128300,4.997305,3.752056,,,,,,,,,,,,,, -128400,15.753588,3.9468553,,,,,,,,,,,,,, -128500,4.140839,3.8287106,,,,,,,,,,,,,, -128600,5.913165,3.6912713,,,,,,,,,,,,,, -128650,,,0.001434949,9.360594,0.0013,9.334594,50000.0,0.0011,9.347084,10000.0,47966.240132808685,48577.46176099777,47966.240132808685,608.6497070789337,1.4954533576965332,0.0 -128700,4.1361065,3.7888737,,,,,,,,,,,,,, -128800,4.131704,3.771037,,,,,,,,,,,,,, -128900,10.327168,3.8868606,,,,,,,,,,,,,, -129000,5.0123954,3.8645692,,,,,,,,,,,,,, -129100,6.1088643,3.9626913,,,,,,,,,,,,,, -129200,6.0015864,3.8596525,,,,,,,,,,,,,, -129300,8.474434,3.8456106,,,,,,,,,,,,,, -129400,6.763921,3.8906226,,,,,,,,,,,,,, -129500,7.8828416,3.8710666,,,,,,,,,,,,,, -129600,4.612991,3.7496066,,,,,,,,,,,,,, -129700,7.094853,3.9525528,,,,,,,,,,,,,, -129800,2.6987665,3.764192,,,,,,,,,,,,,, -129900,6.6447706,3.780396,,,,,,,,,,,,,, -130000,4.265842,3.8424835,,,,,,,,,,,,,, -130100,4.4824114,3.8164,,,,,,,,,,,,,, -130200,3.8789942,3.9313097,,,,,,,,,,,,,, -130300,15.351172,3.877748,,,,,,,,,,,,,, -130400,5.872207,3.8162217,,,,,,,,,,,,,, -130500,5.0853257,3.7430346,,,,,,,,,,,,,, -130600,5.332602,3.7384691,,,,,,,,,,,,,, -130700,6.817777,3.834875,,,,,,,,,,,,,, -130800,3.183642,3.8912535,,,,,,,,,,,,,, -130900,14.383554,3.7520123,,,,,,,,,,,,,, -131000,8.499101,4.028434,,,,,,,,,,,,,, -131100,5.491158,3.8884406,,,,,,,,,,,,,, -131200,7.538159,3.8498774,,,,,,,,,,,,,, -131300,4.6072536,3.8249419,,,,,,,,,,,,,, -131400,7.248749,4.164377,,,,,,,,,,,,,, -131500,3.100877,3.6405263,,,,,,,,,,,,,, -131600,3.633017,3.8352199,,,,,,,,,,,,,, -131700,3.6362364,3.8972,,,,,,,,,,,,,, -131800,5.7197514,3.7907624,,,,,,,,,,,,,, -131900,4.534899,3.768252,,,,,,,,,,,,,, -132000,3.7257755,3.7793033,,,,,,,,,,,,,, -132100,24.01521,3.9003732,,,,,,,,,,,,,, -132200,8.4832535,3.8734608,,,,,,,,,,,,,, -132300,3.9109676,3.8937817,,,,,,,,,,,,,, -132400,8.857049,4.115003,,,,,,,,,,,,,, -132500,17.30652,3.9737992,,,,,,,,,,,,,, -132600,4.5008163,4.027688,,,,,,,,,,,,,, -132700,4.9666276,3.7501674,,,,,,,,,,,,,, -132800,3.8770273,3.8106906,,,,,,,,,,,,,, -132900,7.12603,3.9820058,,,,,,,,,,,,,, -133000,4.1196795,3.7295024,,,,,,,,,,,,,, -133100,5.309162,3.8222263,,,,,,,,,,,,,, -133200,3.5104556,3.794701,,,,,,,,,,,,,, -133300,4.731496,3.8684788,,,,,,,,,,,,,, -133400,3.7211657,3.9165468,,,,,,,,,,,,,, -133500,5.4945173,3.8485303,,,,,,,,,,,,,, -133600,4.174463,3.798112,,,,,,,,,,,,,, -133700,7.9361978,3.8797352,,,,,,,,,,,,,, -133726,,,0.0012157207,9.420927,0.00136,9.384761,50000.0,0.0012,9.396325,10000.0,49962.06882119179,50611.829721450806,49962.06882119179,646.7725491523743,1.700319766998291,0.0 -133800,6.5851173,3.9891148,,,,,,,,,,,,,, -133900,3.309292,3.9502783,,,,,,,,,,,,,, -134000,4.1263843,3.8912814,,,,,,,,,,,,,, -134100,3.0703616,3.933331,,,,,,,,,,,,,, -134200,5.7076354,3.8480613,,,,,,,,,,,,,, -134300,4.2127285,3.8783822,,,,,,,,,,,,,, -134400,3.9787827,3.9837298,,,,,,,,,,,,,, -134500,5.9550333,3.8338687,,,,,,,,,,,,,, -134600,8.219698,4.0145917,,,,,,,,,,,,,, -134700,10.423573,3.6400797,,,,,,,,,,,,,, -134800,4.2272205,3.8627605,,,,,,,,,,,,,, -134900,4.825908,3.8977382,,,,,,,,,,,,,, -135000,3.4322166,3.935482,,,,,,,,,,,,,, -135100,10.54115,3.7081017,,,,,,,,,,,,,, -135200,3.7590482,3.751068,,,,,,,,,,,,,, -135300,5.547554,3.7563553,,,,,,,,,,,,,, -135400,5.192935,3.9318242,,,,,,,,,,,,,, -135500,3.7250822,3.6914709,,,,,,,,,,,,,, -135600,2.4564395,3.7507126,,,,,,,,,,,,,, -135700,3.1030455,3.7164574,,,,,,,,,,,,,, -135800,5.200261,3.9016142,,,,,,,,,,,,,, -135900,3.7498894,3.9070694,,,,,,,,,,,,,, -136000,91.960075,3.9426038,,,,,,,,,,,,,, -136100,6.81938,3.8811378,,,,,,,,,,,,,, -136200,5.027086,3.6630566,,,,,,,,,,,,,, -136300,6.2092404,3.8189926,,,,,,,,,,,,,, -136400,4.024073,3.778483,,,,,,,,,,,,,, -136500,3.324407,3.8160572,,,,,,,,,,,,,, -136600,8.178828,4.015041,,,,,,,,,,,,,, -136700,4.0948396,3.8216214,,,,,,,,,,,,,, -136800,5.6500287,3.766555,,,,,,,,,,,,,, -136900,5.2095127,3.8094542,,,,,,,,,,,,,, -137000,3.2579467,3.6130233,,,,,,,,,,,,,, -137100,5.8184648,3.8766913,,,,,,,,,,,,,, -137200,4.040083,3.761336,,,,,,,,,,,,,, -137300,4.8838506,3.775906,,,,,,,,,,,,,, -137400,3.4770412,3.761632,,,,,,,,,,,,,, -137500,5.116904,3.7801676,,,,,,,,,,,,,, -137600,3.8256075,3.8200767,,,,,,,,,,,,,, -137700,3.9379554,3.8451314,,,,,,,,,,,,,, -137800,3.9868882,3.7340758,,,,,,,,,,,,,, -137900,4.596782,3.9094348,,,,,,,,,,,,,, -138000,2.9868076,3.8406253,,,,,,,,,,,,,, -138100,2.9404895,3.7078092,,,,,,,,,,,,,, -138200,7.182351,3.8980658,,,,,,,,,,,,,, -138300,11.149739,3.7440794,,,,,,,,,,,,,, -138400,6.5421333,3.9067407,,,,,,,,,,,,,, -138500,9.3442335,4.246112,,,,,,,,,,,,,, -138508,,,0.0012755102,9.419978,0.00136,9.398717,50000.0,0.0013000001,9.4097805,10000.0,51957.18427658081,52638.24155449867,51957.18427658081,676.9540636539459,2.354547500610352,0.0 -138600,3.663202,3.8909578,,,,,,,,,,,,,, -138700,64.224686,4.1004596,,,,,,,,,,,,,, -138800,3.2570677,3.7858155,,,,,,,,,,,,,, -138900,4.353622,3.78552,,,,,,,,,,,,,, -139000,5.263773,4.0954127,,,,,,,,,,,,,, -139100,5.2988005,3.9184816,,,,,,,,,,,,,, -139200,3.9885302,3.8498178,,,,,,,,,,,,,, -139300,76.36329,4.218534,,,,,,,,,,,,,, -139400,4.249964,3.8577638,,,,,,,,,,,,,, -139500,3.4573672,3.7508802,,,,,,,,,,,,,, -139600,7.177361,3.8358512,,,,,,,,,,,,,, -139700,4.9277725,3.8184347,,,,,,,,,,,,,, -139800,7.2055054,3.8267255,,,,,,,,,,,,,, -139900,12.950655,4.0530205,,,,,,,,,,,,,, -140000,4.7352667,3.823207,,,,,,,,,,,,,, -140100,5.064902,3.8864157,,,,,,,,,,,,,, -140200,5.458665,3.9222054,,,,,,,,,,,,,, -140300,3.3594112,3.720769,,,,,,,,,,,,,, -140400,3.6445327,3.8913088,,,,,,,,,,,,,, -140500,4.1317406,3.764555,,,,,,,,,,,,,, -140600,6.494982,3.8728871,,,,,,,,,,,,,, -140700,4.3905473,3.8126903,,,,,,,,,,,,,, -140800,7.2763743,3.9773364,,,,,,,,,,,,,, -140900,4.154356,3.8054717,,,,,,,,,,,,,, -141000,24.050682,4.03772,,,,,,,,,,,,,, -141100,4.127982,3.7962687,,,,,,,,,,,,,, -141200,4.220572,3.7542224,,,,,,,,,,,,,, -141300,3.7134833,3.7549622,,,,,,,,,,,,,, -141400,3.1702204,3.727015,,,,,,,,,,,,,, -141500,5.3342853,3.8579397,,,,,,,,,,,,,, -141600,4.222396,3.8799963,,,,,,,,,,,,,, -141700,6.0724382,3.8960862,,,,,,,,,,,,,, -141800,6.571535,3.8929424,,,,,,,,,,,,,, -141900,4.063754,3.7750857,,,,,,,,,,,,,, -142000,4.3810306,3.8099744,,,,,,,,,,,,,, -142100,5.392723,3.8004358,,,,,,,,,,,,,, -142200,8.1363,3.8270135,,,,,,,,,,,,,, -142300,4.029725,3.784245,,,,,,,,,,,,,, -142400,6.9994965,3.8289876,,,,,,,,,,,,,, -142500,7.584088,3.7875366,,,,,,,,,,,,,, -142600,7.731447,3.9578493,,,,,,,,,,,,,, -142700,3.3745327,3.6836023,,,,,,,,,,,,,, -142800,3.4365313,3.8794684,,,,,,,,,,,,,, -142900,8.509372,3.8916903,,,,,,,,,,,,,, -143000,4.1331477,3.91869,,,,,,,,,,,,,, -143100,7.0392065,3.7184043,,,,,,,,,,,,,, -143180,,,0.0013153699,9.403687,0.0014,9.406859,50000.0,0.0012,9.415645,10000.0,53952.60812115669,54674.6584277153,53952.60812115669,717.1750359535217,2.830092191696167,0.0 -143200,8.417012,3.8002343,,,,,,,,,,,,,, -143300,3.8859582,3.7243056,,,,,,,,,,,,,, -143400,3.8062036,3.7753544,,,,,,,,,,,,,, -143500,5.6619096,3.819959,,,,,,,,,,,,,, -143600,17.891428,3.99231,,,,,,,,,,,,,, -143700,2.8821466,3.7749252,,,,,,,,,,,,,, -143800,3.601325,3.8266885,,,,,,,,,,,,,, -143900,3.353802,3.8357353,,,,,,,,,,,,,, -144000,5.2488403,3.8046536,,,,,,,,,,,,,, -144100,67.02205,3.9896965,,,,,,,,,,,,,, -144200,3.637189,3.693881,,,,,,,,,,,,,, -144300,8.611149,3.8990107,,,,,,,,,,,,,, -144400,5.8657193,3.8856676,,,,,,,,,,,,,, -144500,3.7960792,3.83992,,,,,,,,,,,,,, -144600,5.471855,3.8827894,,,,,,,,,,,,,, -144700,6.017971,3.973042,,,,,,,,,,,,,, -144800,3.7230933,3.7869058,,,,,,,,,,,,,, -144900,4.17113,3.9195085,,,,,,,,,,,,,, -145000,26.115973,3.9542093,,,,,,,,,,,,,, -145100,2.7483776,3.7512102,,,,,,,,,,,,,, -145200,5.9460835,3.8501487,,,,,,,,,,,,,, -145300,3.850978,3.7350805,,,,,,,,,,,,,, -145400,6.2755084,4.062192,,,,,,,,,,,,,, -145500,4.936452,3.9079704,,,,,,,,,,,,,, -145600,3.065388,3.729074,,,,,,,,,,,,,, -145700,5.951106,3.7430916,,,,,,,,,,,,,, -145800,24.099552,3.8994055,,,,,,,,,,,,,, -145900,11.072425,3.857088,,,,,,,,,,,,,, -146000,3.922376,3.825079,,,,,,,,,,,,,, -146100,3.4354656,3.7345052,,,,,,,,,,,,,, -146200,4.4444237,3.8440661,,,,,,,,,,,,,, -146300,7.0046325,3.9814,,,,,,,,,,,,,, -146400,5.335746,3.7620268,,,,,,,,,,,,,, -146500,6.760083,3.9517574,,,,,,,,,,,,,, -146600,4.293241,3.7198446,,,,,,,,,,,,,, -146700,6.0312667,3.7503433,,,,,,,,,,,,,, -146800,3.2752748,3.7268858,,,,,,,,,,,,,, -146900,5.1298113,3.792056,,,,,,,,,,,,,, -147000,4.258085,3.9766102,,,,,,,,,,,,,, -147100,5.0237784,3.796319,,,,,,,,,,,,,, -147200,3.2671251,3.8037283,,,,,,,,,,,,,, -147300,3.9880888,3.8779745,,,,,,,,,,,,,, -147400,3.7063823,3.7941246,,,,,,,,,,,,,, -147500,4.922503,3.7816653,,,,,,,,,,,,,, -147523,,,0.0015943877,9.445664,0.0013799999,9.394714,50000.0,0.0014000001,9.402309,10000.0,55949.09929037094,56731.34460759163,55949.09929037094,776.5972204208374,3.354834079742432,0.0 -147600,3.1716354,3.7378397,,,,,,,,,,,,,, -147700,3.3042555,3.720696,,,,,,,,,,,,,, -147800,4.2256713,3.7667284,,,,,,,,,,,,,, -147900,4.7392054,3.9379175,,,,,,,,,,,,,, -148000,3.572447,3.900601,,,,,,,,,,,,,, -148100,9.288154,3.8983397,,,,,,,,,,,,,, -148200,10.898133,4.010025,,,,,,,,,,,,,, -148300,15.055841,3.7654238,,,,,,,,,,,,,, -148400,6.096031,3.7907152,,,,,,,,,,,,,, -148500,4.697271,3.9001446,,,,,,,,,,,,,, -148600,6.910241,3.9030862,,,,,,,,,,,,,, -148700,21.46214,3.7532768,,,,,,,,,,,,,, -148800,7.7798495,3.8777633,,,,,,,,,,,,,, -148900,4.2619405,3.8896973,,,,,,,,,,,,,, -149000,4.098154,3.739449,,,,,,,,,,,,,, -149100,2.8748207,3.9043446,,,,,,,,,,,,,, -149200,8.268768,4.0144277,,,,,,,,,,,,,, -149300,5.4596143,3.8668227,,,,,,,,,,,,,, -149400,4.1215377,4.0124655,,,,,,,,,,,,,, -149500,3.0758722,3.761863,,,,,,,,,,,,,, -149600,5.5063562,3.799037,,,,,,,,,,,,,, -149700,5.03024,3.7995045,,,,,,,,,,,,,, -149800,4.2173133,3.8576658,,,,,,,,,,,,,, -149900,5.636962,3.7762184,,,,,,,,,,,,,, -150000,4.918457,3.73599,,,,,,,,,,,,,, -150100,11.616698,3.7973948,,,,,,,,,,,,,, -150200,4.347864,4.036248,,,,,,,,,,,,,, -150300,6.6763134,3.7735045,,,,,,,,,,,,,, -150400,5.4279256,3.9672842,,,,,,,,,,,,,, -150500,4.634214,3.857708,,,,,,,,,,,,,, -150600,15.116459,4.0284414,,,,,,,,,,,,,, -150700,4.18612,3.867536,,,,,,,,,,,,,, -150800,7.285976,3.7961574,,,,,,,,,,,,,, -150900,3.5583847,3.7621083,,,,,,,,,,,,,, -151000,3.324401,3.7974336,,,,,,,,,,,,,, -151100,6.050889,3.8820176,,,,,,,,,,,,,, -151200,13.137903,4.1853614,,,,,,,,,,,,,, -151300,2.7950597,3.8225584,,,,,,,,,,,,,, -151400,3.147219,3.8597286,,,,,,,,,,,,,, -151500,3.5861723,3.6991458,,,,,,,,,,,,,, -151600,8.565522,3.8736215,,,,,,,,,,,,,, -151700,4.320493,3.6521337,,,,,,,,,,,,,, -151800,7.360284,3.822566,,,,,,,,,,,,,, -151900,6.5596066,3.7674623,,,,,,,,,,,,,, -152000,4.8493156,3.7548082,,,,,,,,,,,,,, -152100,7.4942713,3.9481244,,,,,,,,,,,,,, -152200,4.505962,3.6923234,,,,,,,,,,,,,, -152300,3.9388266,3.8162346,,,,,,,,,,,,,, -152348,,,0.0014548788,9.38589,0.00148,9.364571,50000.0,0.0014000001,9.372682,10000.0,57944.53552174568,58768.525552749634,57944.53552174568,817.6529867649078,3.770930528640747,0.0 -152400,17.20805,3.855658,,,,,,,,,,,,,, -152500,6.974955,3.951393,,,,,,,,,,,,,, -152600,5.717931,3.7809281,,,,,,,,,,,,,, -152700,5.2584343,3.8981168,,,,,,,,,,,,,, -152800,3.802099,3.9198563,,,,,,,,,,,,,, -152900,3.505695,3.8322506,,,,,,,,,,,,,, -153000,2.85077,3.8038218,,,,,,,,,,,,,, -153100,11.638981,4.108363,,,,,,,,,,,,,, -153200,4.421865,3.7954037,,,,,,,,,,,,,, -153300,2.5327616,3.7290578,,,,,,,,,,,,,, -153400,4.076623,3.9194632,,,,,,,,,,,,,, -153500,2.5437398,3.667018,,,,,,,,,,,,,, -153600,2.8428187,3.8215213,,,,,,,,,,,,,, -153700,5.3189445,3.8194094,,,,,,,,,,,,,, -153800,5.8602004,3.8340673,,,,,,,,,,,,,, -153900,3.5451646,3.74924,,,,,,,,,,,,,, -154000,3.936041,3.7663155,,,,,,,,,,,,,, -154100,4.932789,3.8913043,,,,,,,,,,,,,, -154200,5.0930037,3.9221964,,,,,,,,,,,,,, -154300,5.6015525,3.8808422,,,,,,,,,,,,,, -154400,11.102789,3.967843,,,,,,,,,,,,,, -154500,4.572291,3.9171813,,,,,,,,,,,,,, -154600,6.434405,4.0352945,,,,,,,,,,,,,, -154700,2.3424513,3.7549133,,,,,,,,,,,,,, -154800,6.8453436,3.8667982,,,,,,,,,,,,,, -154900,6.3977723,3.7872944,,,,,,,,,,,,,, -155000,2.3470433,3.793666,,,,,,,,,,,,,, -155100,3.2869186,3.8164806,,,,,,,,,,,,,, -155200,2.8233383,3.7084625,,,,,,,,,,,,,, -155300,4.066618,3.6179955,,,,,,,,,,,,,, -155400,5.534333,3.885864,,,,,,,,,,,,,, -155500,4.0048094,3.866613,,,,,,,,,,,,,, -155600,4.0715647,3.813377,,,,,,,,,,,,,, -155700,3.9585571,3.8034194,,,,,,,,,,,,,, -155800,7.6412864,3.855689,,,,,,,,,,,,,, -155900,3.2434893,3.7727237,,,,,,,,,,,,,, -156000,2.7327576,3.7605867,,,,,,,,,,,,,, -156100,4.40445,3.8449583,,,,,,,,,,,,,, -156200,5.5486035,3.8099506,,,,,,,,,,,,,, -156300,4.0794883,3.8747747,,,,,,,,,,,,,, -156400,7.2459526,3.7785864,,,,,,,,,,,,,, -156500,4.0568595,3.811737,,,,,,,,,,,,,, -156600,37.379444,3.8493261,,,,,,,,,,,,,, -156638,,,0.0015146683,9.388604,0.00148,9.365797,50000.0,0.0015,9.373345,10000.0,59939.89631223679,60806.5135269165,59939.89631223679,859.5641634464264,4.192249059677124,0.0 -156700,5.616528,3.763995,,,,,,,,,,,,,, -156800,2.974624,3.751839,,,,,,,,,,,,,, -156900,4.0366583,3.8593993,,,,,,,,,,,,,, -157000,3.9090736,3.7433,,,,,,,,,,,,,, -157100,16.009733,3.8563066,,,,,,,,,,,,,, -157200,9.098629,3.9111164,,,,,,,,,,,,,, -157300,7.167022,3.9568849,,,,,,,,,,,,,, -157400,4.793281,4.003694,,,,,,,,,,,,,, -157500,3.525197,3.7287874,,,,,,,,,,,,,, -157600,3.3508904,3.677533,,,,,,,,,,,,,, -157700,6.51819,3.8434534,,,,,,,,,,,,,, -157800,4.421385,3.820516,,,,,,,,,,,,,, -157900,5.976167,3.857009,,,,,,,,,,,,,, -158000,4.10925,3.7644696,,,,,,,,,,,,,, -158100,6.240832,3.9307413,,,,,,,,,,,,,, -158200,4.5723467,3.7067308,,,,,,,,,,,,,, -158300,4.800234,3.8346467,,,,,,,,,,,,,, -158400,2.4688623,3.8526554,,,,,,,,,,,,,, -158500,7.53108,3.9493961,,,,,,,,,,,,,, -158600,5.399687,4.023742,,,,,,,,,,,,,, -158700,4.4904656,3.9191358,,,,,,,,,,,,,, -158800,10.111828,3.8100913,,,,,,,,,,,,,, -158900,12.602244,3.8489337,,,,,,,,,,,,,, -159000,9.391828,4.0041513,,,,,,,,,,,,,, -159100,3.2428465,3.855812,,,,,,,,,,,,,, -159200,4.2909927,3.8429728,,,,,,,,,,,,,, -159300,4.977233,3.802751,,,,,,,,,,,,,, -159400,7.9178305,3.6169186,,,,,,,,,,,,,, -159500,6.19245,3.8896956,,,,,,,,,,,,,, -159600,10.224848,4.018451,,,,,,,,,,,,,, -159700,6.392645,3.8596911,,,,,,,,,,,,,, -159800,4.641001,3.991294,,,,,,,,,,,,,, -159900,5.8374,3.7533555,,,,,,,,,,,,,, -160000,4.218525,3.7982335,,,,,,,,,,,,,, -160100,2.614822,3.739038,,,,,,,,,,,,,, -160200,3.5965292,3.861714,,,,,,,,,,,,,, -160300,3.645774,3.591772,,,,,,,,,,,,,, -160400,2.6256206,3.769012,,,,,,,,,,,,,, -160500,14.687573,3.9166553,,,,,,,,,,,,,, -160600,8.888832,3.6242583,,,,,,,,,,,,,, -160700,7.009732,3.9674118,,,,,,,,,,,,,, -160800,6.1001782,3.9146838,,,,,,,,,,,,,, -160837,,,0.0013950892,9.382287,0.00154,9.348522,50000.0,0.0015,9.357764,10000.0,61935.455322265625,62844.32478642464,61935.455322265625,901.0739798545836,4.68764328956604,0.0 -160900,9.485073,3.8564272,,,,,,,,,,,,,, -161000,5.414852,3.8168828,,,,,,,,,,,,,, -161100,32.928566,3.9093034,,,,,,,,,,,,,, -161200,12.961156,3.7882524,,,,,,,,,,,,,, -161300,9.303776,3.7016978,,,,,,,,,,,,,, -161400,5.413611,3.7554555,,,,,,,,,,,,,, -161500,4.0364475,3.7915123,,,,,,,,,,,,,, -161600,5.7954383,3.8934681,,,,,,,,,,,,,, -161700,3.7775707,3.9295907,,,,,,,,,,,,,, -161800,8.024208,3.6629934,,,,,,,,,,,,,, -161900,4.1369824,3.7662277,,,,,,,,,,,,,, -162000,5.718033,3.8609405,,,,,,,,,,,,,, -162100,4.3806605,3.7654688,,,,,,,,,,,,,, -162200,2.761912,3.814372,,,,,,,,,,,,,, -162300,6.5889945,3.7841563,,,,,,,,,,,,,, -162400,3.1367278,3.8828845,,,,,,,,,,,,,, -162500,5.043025,3.8337498,,,,,,,,,,,,,, -162600,3.9429064,3.6909423,,,,,,,,,,,,,, -162700,6.242691,3.737183,,,,,,,,,,,,,, -162800,3.5594177,3.8462024,,,,,,,,,,,,,, -162900,3.3521237,3.9334826,,,,,,,,,,,,,, -163000,4.6993055,4.0227013,,,,,,,,,,,,,, -163100,3.1979957,3.8216012,,,,,,,,,,,,,, -163200,3.1136677,3.7315147,,,,,,,,,,,,,, -163300,2.816383,3.6603634,,,,,,,,,,,,,, -163400,3.9927511,3.7463508,,,,,,,,,,,,,, -163500,3.8040173,3.6602168,,,,,,,,,,,,,, -163600,7.3406034,3.7064066,,,,,,,,,,,,,, -163700,14.617626,3.7977931,,,,,,,,,,,,,, -163800,2.0753503,3.8294225,,,,,,,,,,,,,, -163900,5.589823,4.0049148,,,,,,,,,,,,,, -164000,6.152741,3.828398,,,,,,,,,,,,,, -164100,2.5785506,3.6663685,,,,,,,,,,,,,, -164200,6.4942,3.81677,,,,,,,,,,,,,, -164300,3.7301397,3.8629954,,,,,,,,,,,,,, -164400,4.371313,3.7803206,,,,,,,,,,,,,, -164500,13.651944,3.9233408,,,,,,,,,,,,,, -164600,6.033597,3.8025541,,,,,,,,,,,,,, -164700,7.616769,3.8281505,,,,,,,,,,,,,, -164800,6.9350615,3.8666878,,,,,,,,,,,,,, -164900,4.721319,3.7539568,,,,,,,,,,,,,, -164985,,,0.0015943877,9.390181,0.00154,9.356932,50000.0,0.0015,9.366373,10000.0,63930.97182655335,64881.60176515579,63930.97182655335,942.0960912704468,5.1383795738220215,0.0 -165000,2.8817663,3.868962,,,,,,,,,,,,,, -165100,3.531655,3.679029,,,,,,,,,,,,,, -165200,5.2550144,3.7692611,,,,,,,,,,,,,, -165300,8.497959,3.837573,,,,,,,,,,,,,, -165400,4.741467,3.8573174,,,,,,,,,,,,,, -165500,8.722345,3.8135202,,,,,,,,,,,,,, -165600,3.1732826,3.905072,,,,,,,,,,,,,, -165700,4.9458985,3.7662313,,,,,,,,,,,,,, -165800,3.989323,3.8304808,,,,,,,,,,,,,, -165900,3.9185348,3.774608,,,,,,,,,,,,,, -166000,3.4116223,3.7802181,,,,,,,,,,,,,, -166100,3.892124,3.774105,,,,,,,,,,,,,, -166200,3.858989,3.7895212,,,,,,,,,,,,,, -166300,6.3389277,4.340108,,,,,,,,,,,,,, -166400,4.9385114,3.8964894,,,,,,,,,,,,,, -166500,4.46242,3.8284376,,,,,,,,,,,,,, -166600,4.206891,3.842183,,,,,,,,,,,,,, -166700,5.5914774,3.8133087,,,,,,,,,,,,,, -166800,3.0470264,3.8175251,,,,,,,,,,,,,, -166900,3.4779296,3.7050748,,,,,,,,,,,,,, -167000,5.6279964,3.878571,,,,,,,,,,,,,, -167100,23.448168,4.057926,,,,,,,,,,,,,, -167200,9.976771,3.8266304,,,,,,,,,,,,,, -167300,4.9495816,3.81945,,,,,,,,,,,,,, -167400,4.2471795,3.745185,,,,,,,,,,,,,, -167500,4.1822224,3.7951603,,,,,,,,,,,,,, -167600,6.384552,3.876464,,,,,,,,,,,,,, -167700,4.230561,3.899108,,,,,,,,,,,,,, -167800,4.3187604,3.870554,,,,,,,,,,,,,, -167900,3.6675365,3.6831293,,,,,,,,,,,,,, -168000,6.562756,3.906938,,,,,,,,,,,,,, -168100,2.5360956,3.71943,,,,,,,,,,,,,, -168200,4.1205716,3.7521975,,,,,,,,,,,,,, -168300,4.1540318,3.7317944,,,,,,,,,,,,,, -168400,2.8844056,3.7534606,,,,,,,,,,,,,, -168500,4.902186,3.6748042,,,,,,,,,,,,,, -168600,3.8304625,3.978655,,,,,,,,,,,,,, -168700,4.6459203,3.6545665,,,,,,,,,,,,,, -168800,3.0266416,3.8108716,,,,,,,,,,,,,, -168900,4.1930423,3.8335257,,,,,,,,,,,,,, -169000,6.0900364,3.7705176,,,,,,,,,,,,,, -169039,,,0.0015345982,9.38863,0.0016,9.360721,50000.0,0.0015,9.372903,10000.0,65926.25481128693,66920.25790929794,65926.25481128693,984.6689186096193,5.627253293991089,0.0 -169100,4.401186,3.813406,,,,,,,,,,,,,, -169200,4.4761515,3.6735218,,,,,,,,,,,,,, -169300,6.976401,3.9594014,,,,,,,,,,,,,, -169400,4.407276,3.8219943,,,,,,,,,,,,,, -169500,5.1843834,4.018228,,,,,,,,,,,,,, -169600,8.079224,3.8805811,,,,,,,,,,,,,, -169700,3.4439404,3.7644958,,,,,,,,,,,,,, -169800,2.4559627,3.634181,,,,,,,,,,,,,, -169900,7.1060996,3.7142787,,,,,,,,,,,,,, -170000,3.2568486,3.6986215,,,,,,,,,,,,,, -170100,5.084026,3.7134788,,,,,,,,,,,,,, -170200,5.328367,3.8214717,,,,,,,,,,,,,, -170300,2.5380025,3.8716407,,,,,,,,,,,,,, -170400,3.5903141,3.8649669,,,,,,,,,,,,,, -170500,3.546907,3.763191,,,,,,,,,,,,,, -170600,5.476774,3.7902415,,,,,,,,,,,,,, -170700,8.3212185,3.7168767,,,,,,,,,,,,,, -170800,6.068613,3.768695,,,,,,,,,,,,,, -170900,4.212216,3.8996959,,,,,,,,,,,,,, -171000,4.7884316,3.7263882,,,,,,,,,,,,,, -171100,3.4938843,3.750629,,,,,,,,,,,,,, -171200,2.8983111,3.7881851,,,,,,,,,,,,,, -171300,4.453057,3.7739065,,,,,,,,,,,,,, -171400,3.987098,3.8540359,,,,,,,,,,,,,, -171500,6.1758437,3.6884186,,,,,,,,,,,,,, -171600,13.986029,3.8641233,,,,,,,,,,,,,, -171700,5.1799974,3.8561058,,,,,,,,,,,,,, -171800,4.1569843,3.7561307,,,,,,,,,,,,,, -171900,6.451967,3.8361535,,,,,,,,,,,,,, -172000,4.898999,3.7665448,,,,,,,,,,,,,, -172100,22.533875,3.8522623,,,,,,,,,,,,,, -172200,4.209978,3.812691,,,,,,,,,,,,,, -172300,2.4928014,3.7273822,,,,,,,,,,,,,, -172400,4.214897,3.8088326,,,,,,,,,,,,,, -172500,3.8033898,3.894095,,,,,,,,,,,,,, -172600,3.8589463,3.7880135,,,,,,,,,,,,,, -172700,2.7505615,3.8571224,,,,,,,,,,,,,, -172800,6.7541695,3.8069103,,,,,,,,,,,,,, -172900,2.973385,3.7876449,,,,,,,,,,,,,, -173000,4.425632,3.8298557,,,,,,,,,,,,,, -173100,3.7519653,3.796032,,,,,,,,,,,,,, -173140,,,0.0012356505,9.44169,0.00156,9.386361,50000.0,0.0015,9.398773,10000.0,67922.46532917023,68957.95928025246,67922.46532917023,1025.479611635208,6.089812755584717,0.0 -173200,9.121915,3.8934317,,,,,,,,,,,,,, -173300,4.1409454,3.760781,,,,,,,,,,,,,, -173400,5.6834025,3.7004266,,,,,,,,,,,,,, -173500,5.4645824,3.9960523,,,,,,,,,,,,,, -173600,12.505391,3.7997086,,,,,,,,,,,,,, -173700,7.0380816,3.8449407,,,,,,,,,,,,,, -173800,4.5463605,3.807358,,,,,,,,,,,,,, -173900,63.091724,3.9510756,,,,,,,,,,,,,, -174000,5.9390526,3.9618783,,,,,,,,,,,,,, -174100,2.5319636,3.8388221,,,,,,,,,,,,,, -174200,6.4054694,3.754303,,,,,,,,,,,,,, -174300,5.541139,3.9109914,,,,,,,,,,,,,, -174400,3.9128869,3.7040262,,,,,,,,,,,,,, -174500,3.4187372,3.8474798,,,,,,,,,,,,,, -174600,3.7105234,3.7777908,,,,,,,,,,,,,, -174700,3.033978,3.7939565,,,,,,,,,,,,,, -174800,50.762062,4.060713,,,,,,,,,,,,,, -174900,3.5991752,3.6591165,,,,,,,,,,,,,, -175000,10.934308,3.9261436,,,,,,,,,,,,,, -175100,3.1374035,3.8422709,,,,,,,,,,,,,, -175200,10.22655,3.7531035,,,,,,,,,,,,,, -175300,13.348222,3.8291469,,,,,,,,,,,,,, -175400,7.1445017,3.9900038,,,,,,,,,,,,,, -175500,7.457942,3.8215265,,,,,,,,,,,,,, -175600,4.315648,3.834619,,,,,,,,,,,,,, -175700,3.5976567,3.7941794,,,,,,,,,,,,,, -175800,6.780549,3.9405642,,,,,,,,,,,,,, -175900,4.528333,3.8013558,,,,,,,,,,,,,, -176000,4.923599,3.7713976,,,,,,,,,,,,,, -176100,6.916127,3.8319397,,,,,,,,,,,,,, -176200,12.3015,3.749299,,,,,,,,,,,,,, -176300,4.6831584,3.7265751,,,,,,,,,,,,,, -176400,4.7720017,3.848799,,,,,,,,,,,,,, -176500,3.4377663,3.822871,,,,,,,,,,,,,, -176600,4.7116785,3.853363,,,,,,,,,,,,,, -176700,17.05834,3.7871063,,,,,,,,,,,,,, -176800,3.2346313,3.8516927,,,,,,,,,,,,,, -176900,3.7220078,3.817111,,,,,,,,,,,,,, -177000,3.3779304,3.6953158,,,,,,,,,,,,,, -177100,3.7908895,3.8298092,,,,,,,,,,,,,, -177200,2.1202412,3.6911662,,,,,,,,,,,,,, -177300,3.020615,3.8426514,,,,,,,,,,,,,, -177337,,,0.0014947385,9.390521,0.0016999999,9.368401,50000.0,0.0015,9.381394,10000.0,69918.275775671,70995.66643381119,69918.275775671,1066.5990326404572,6.489320278167725,0.0 -177400,3.4462967,3.8149242,,,,,,,,,,,,,, -177500,2.8018847,3.6784847,,,,,,,,,,,,,, -177600,3.7320805,3.8316085,,,,,,,,,,,,,, -177700,2.8259358,3.78893,,,,,,,,,,,,,, -177800,10.393411,3.6931221,,,,,,,,,,,,,, -177900,4.2672763,3.8128867,,,,,,,,,,,,,, -178000,10.409967,3.9147904,,,,,,,,,,,,,, -178100,3.4862409,3.7953467,,,,,,,,,,,,,, -178200,6.6651554,3.8257987,,,,,,,,,,,,,, -178300,3.076569,3.7598433,,,,,,,,,,,,,, -178400,11.866055,3.6389055,,,,,,,,,,,,,, -178500,7.75984,4.072914,,,,,,,,,,,,,, -178600,26.591286,3.9246545,,,,,,,,,,,,,, -178700,5.452311,3.7823546,,,,,,,,,,,,,, -178800,3.8294184,3.8211622,,,,,,,,,,,,,, -178900,3.4046302,3.6761303,,,,,,,,,,,,,, -179000,5.0523376,3.9536443,,,,,,,,,,,,,, -179100,5.9207187,3.918936,,,,,,,,,,,,,, -179200,6.689147,3.7992554,,,,,,,,,,,,,, -179300,4.6778774,3.741457,,,,,,,,,,,,,, -179400,4.424163,3.935109,,,,,,,,,,,,,, -179500,5.863819,3.7922301,,,,,,,,,,,,,, -179600,6.6573696,3.8373995,,,,,,,,,,,,,, -179700,5.544376,3.7151976,,,,,,,,,,,,,, -179800,9.340443,3.9339223,,,,,,,,,,,,,, -179900,5.6051636,3.707269,,,,,,,,,,,,,, -180000,9.358218,3.8270614,,,,,,,,,,,,,, -180100,3.0456102,3.653181,,,,,,,,,,,,,, -180200,3.6660714,3.7167213,,,,,,,,,,,,,, -180300,6.116247,3.8326325,,,,,,,,,,,,,, -180400,22.15659,3.7939487,,,,,,,,,,,,,, -180500,4.368332,3.650488,,,,,,,,,,,,,, -180600,4.008203,3.709988,,,,,,,,,,,,,, -180700,7.2030244,3.8491287,,,,,,,,,,,,,, -180800,6.1079683,3.6784365,,,,,,,,,,,,,, -180900,11.550073,3.7314618,,,,,,,,,,,,,, -181000,3.6507938,3.697931,,,,,,,,,,,,,, -181100,5.409558,3.8581476,,,,,,,,,,,,,, -181200,4.303969,3.8596127,,,,,,,,,,,,,, -181300,8.622207,3.7247763,,,,,,,,,,,,,, -181398,,,0.0013352997,9.425164,0.0016999999,9.390179,50000.0,0.0015,9.403217,10000.0,71913.75967979431,73033.81380438805,71913.75967979431,1108.525652885437,6.931232213973999,0.0 -181400,4.4519143,3.8184469,,,,,,,,,,,,,, -181500,6.602558,3.734921,,,,,,,,,,,,,, -181600,19.92081,3.7950711,,,,,,,,,,,,,, -181700,3.1505606,3.7838588,,,,,,,,,,,,,, -181800,2.2375083,3.6685536,,,,,,,,,,,,,, -181900,5.1505766,3.8722801,,,,,,,,,,,,,, -182000,4.7889395,3.6704774,,,,,,,,,,,,,, -182100,3.428813,3.7717957,,,,,,,,,,,,,, -182200,4.052313,3.940524,,,,,,,,,,,,,, -182300,4.8599777,3.7823496,,,,,,,,,,,,,, -182400,7.3184137,3.8677406,,,,,,,,,,,,,, -182500,3.2476482,3.7229154,,,,,,,,,,,,,, -182600,3.1097772,3.715004,,,,,,,,,,,,,, -182700,5.1287107,3.8141983,,,,,,,,,,,,,, -182800,9.308673,3.715695,,,,,,,,,,,,,, -182900,3.9149945,3.848405,,,,,,,,,,,,,, -183000,4.728837,3.813557,,,,,,,,,,,,,, -183100,3.8737833,3.7999496,,,,,,,,,,,,,, -183200,3.3406007,3.8564138,,,,,,,,,,,,,, -183300,4.869385,3.827547,,,,,,,,,,,,,, -183400,3.0061312,3.6441727,,,,,,,,,,,,,, -183500,3.0475783,3.6695483,,,,,,,,,,,,,, -183600,4.89072,3.9064274,,,,,,,,,,,,,, -183700,5.7796855,3.9008932,,,,,,,,,,,,,, -183800,5.31291,3.8191724,,,,,,,,,,,,,, -183900,2.210478,3.5891724,,,,,,,,,,,,,, -184000,3.0383904,3.7901301,,,,,,,,,,,,,, -184100,7.502666,3.7773857,,,,,,,,,,,,,, -184200,5.701761,4.0971427,,,,,,,,,,,,,, -184300,3.3446157,3.8189998,,,,,,,,,,,,,, -184400,6.1622677,3.8467674,,,,,,,,,,,,,, -184500,3.4236681,3.6349926,,,,,,,,,,,,,, -184600,3.7439787,3.7590032,,,,,,,,,,,,,, -184700,4.5416536,3.8524115,,,,,,,,,,,,,, -184800,20.175745,3.9131007,,,,,,,,,,,,,, -184900,7.6037126,3.826201,,,,,,,,,,,,,, -185000,3.275352,3.766748,,,,,,,,,,,,,, -185100,8.906727,3.7939596,,,,,,,,,,,,,, -185200,9.632234,3.8845887,,,,,,,,,,,,,, -185300,1.8322484,3.6784196,,,,,,,,,,,,,, -185400,4.3659506,3.886353,,,,,,,,,,,,,, -185500,3.6604214,3.734647,,,,,,,,,,,,,, -185561,,,0.0014748087,9.436946,0.00168,9.422385,50000.0,0.0017000001,9.43556,10000.0,73909.05753827095,75067.66927337646,73909.05753827095,1146.3552420139313,7.326283931732178,0.0 -185600,5.7987547,3.7611222,,,,,,,,,,,,,, -185700,3.8410375,3.8204045,,,,,,,,,,,,,, -185800,4.265686,3.6613536,,,,,,,,,,,,,, -185900,4.474493,3.7961617,,,,,,,,,,,,,, -186000,11.974519,3.864711,,,,,,,,,,,,,, -186100,15.0586605,3.80512,,,,,,,,,,,,,, -186200,3.464011,3.7960343,,,,,,,,,,,,,, -186300,3.72559,3.7767324,,,,,,,,,,,,,, -186400,2.6054845,3.8534555,,,,,,,,,,,,,, -186500,2.7246368,3.803078,,,,,,,,,,,,,, -186600,17.855177,3.8306804,,,,,,,,,,,,,, -186700,4.5267825,3.673974,,,,,,,,,,,,,, -186800,3.9697614,3.710946,,,,,,,,,,,,,, -186900,4.818208,3.6003819,,,,,,,,,,,,,, -187000,6.2882414,3.7023036,,,,,,,,,,,,,, -187100,60.914364,3.9375296,,,,,,,,,,,,,, -187200,3.9198246,3.828491,,,,,,,,,,,,,, -187300,3.7276983,3.8319478,,,,,,,,,,,,,, -187400,11.482843,3.7292852,,,,,,,,,,,,,, -187500,4.434324,3.8703768,,,,,,,,,,,,,, -187600,5.756841,3.7401342,,,,,,,,,,,,,, -187700,4.1822705,3.7143059,,,,,,,,,,,,,, -187800,6.965905,3.7779682,,,,,,,,,,,,,, -187900,4.112825,3.7807398,,,,,,,,,,,,,, -188000,6.2004013,3.9138074,,,,,,,,,,,,,, -188100,5.002247,3.7694488,,,,,,,,,,,,,, -188200,3.162055,3.7219973,,,,,,,,,,,,,, -188300,4.3093896,3.8641477,,,,,,,,,,,,,, -188400,3.9523344,3.8093295,,,,,,,,,,,,,, -188500,3.5791008,3.6218739,,,,,,,,,,,,,, -188600,13.275085,3.723219,,,,,,,,,,,,,, -188700,2.8682544,3.606424,,,,,,,,,,,,,, -188800,4.3750334,3.7182932,,,,,,,,,,,,,, -188900,8.633361,3.781147,,,,,,,,,,,,,, -189000,5.2409577,3.7180374,,,,,,,,,,,,,, -189100,21.551529,3.8371637,,,,,,,,,,,,,, -189200,13.451133,3.8281868,,,,,,,,,,,,,, -189300,4.374985,3.7372754,,,,,,,,,,,,,, -189400,13.766833,3.8528252,,,,,,,,,,,,,, -189500,11.595244,3.8546677,,,,,,,,,,,,,, -189600,3.7334602,3.780143,,,,,,,,,,,,,, -189700,3.7160935,3.7408772,,,,,,,,,,,,,, -189721,,,,,,,,,,,75905.17727708817,,,,,0.0 +0,0.6713023,6.9223967,,,,,,,,,,,,,, +1,,,0.0007573342,6.9125247,0.00068,6.91273,50000.0,0.00080000004,6.9129095,10000.0,52.91309928894043,117.18459677696228,52.91309928894043,64.27141284942627,0.0,0.0 +100,0.6732675,6.8051786,,,,,,,,,,,,,, +200,0.8279272,6.4964256,,,,,,,,,,,,,, +300,0.9894157,6.1964664,,,,,,,,,,,,,, +400,1.8994367,5.9636917,,,,,,,,,,,,,, +500,4.463675,5.8021355,,,,,,,,,,,,,, +600,3.448649,5.549911,,,,,,,,,,,,,, +700,4.1367326,5.3575125,,,,,,,,,,,,,, +800,3.4241135,5.240439,,,,,,,,,,,,,, +900,6.0213556,4.997196,,,,,,,,,,,,,, +1000,5.664543,5.1665163,,,,,,,,,,,,,, +1100,3.5806499,4.8967133,,,,,,,,,,,,,, +1200,4.2157106,4.769206,,,,,,,,,,,,,, +1300,4.731968,4.7414017,,,,,,,,,,,,,, +1400,6.556045,4.5446644,,,,,,,,,,,,,, +1500,8.186122,4.4882245,,,,,,,,,,,,,, +1600,6.157769,4.468399,,,,,,,,,,,,,, +1700,4.8452125,4.4095473,,,,,,,,,,,,,, +1800,3.7392683,4.2734833,,,,,,,,,,,,,, +1900,5.120582,4.1723003,,,,,,,,,,,,,, +2000,5.7835703,3.9605649,,,,,,,,,,,,,, +2100,3.6723592,3.865443,,,,,,,,,,,,,, +2200,4.4364605,3.8933778,,,,,,,,,,,,,, +2300,3.6287751,3.879135,,,,,,,,,,,,,, +2400,3.2232623,3.6756186,,,,,,,,,,,,,, +2500,3.4139717,3.6565855,,,,,,,,,,,,,, +2600,3.7413301,3.519341,,,,,,,,,,,,,, +2700,3.84993,3.7066567,,,,,,,,,,,,,, +2800,3.781147,3.4161122,,,,,,,,,,,,,, +2900,4.345715,3.3421917,,,,,,,,,,,,,, +3000,3.900597,3.3178554,,,,,,,,,,,,,, +3100,3.4220376,3.2843199,,,,,,,,,,,,,, +3200,3.2616594,3.1931348,,,,,,,,,,,,,, +3300,3.6417232,3.0921278,,,,,,,,,,,,,, +3400,4.614095,3.176371,,,,,,,,,,,,,, +3500,3.8950982,3.0589523,,,,,,,,,,,,,, +3600,3.4069495,2.9844847,,,,,,,,,,,,,, +3700,3.469931,3.0213032,,,,,,,,,,,,,, +3800,3.9887073,2.9217954,,,,,,,,,,,,,, +3900,2.6450412,2.973175,,,,,,,,,,,,,, +4000,3.2164812,2.944903,,,,,,,,,,,,,, +4100,3.143095,2.8451371,,,,,,,,,,,,,, +4200,2.5667765,2.5756903,,,,,,,,,,,,,, +4300,3.8288858,2.885184,,,,,,,,,,,,,, +4400,2.4932928,2.6563308,,,,,,,,,,,,,, +4500,3.5233526,2.686227,,,,,,,,,,,,,, +4600,2.3761024,2.6628616,,,,,,,,,,,,,, +4700,3.7111886,2.5667148,,,,,,,,,,,,,, +4800,2.7228427,2.5981257,,,,,,,,,,,,,, +4900,2.3265884,2.5728264,,,,,,,,,,,,,, +5000,3.0561707,2.4604645,,,,,,,,,,,,,, +5100,4.2081323,2.5181365,,,,,,,,,,,,,, +5200,2.750637,2.4802196,,,,,,,,,,,,,, +5300,3.0783362,2.5320334,,,,,,,,,,,,,, +5400,2.3811414,2.3544753,,,,,,,,,,,,,, +5500,1.9294297,2.4226274,,,,,,,,,,,,,, +5600,2.0577,2.3770676,,,,,,,,,,,,,, +5700,2.6448102,2.3234134,,,,,,,,,,,,,, +5800,2.4552207,2.5097256,,,,,,,,,,,,,, +5900,3.1033998,2.396461,,,,,,,,,,,,,, +6000,2.5513463,2.1276076,,,,,,,,,,,,,, +6100,2.332324,2.1998985,,,,,,,,,,,,,, +6200,2.1485863,2.1842122,,,,,,,,,,,,,, +6300,2.2128503,2.1971273,,,,,,,,,,,,,, +6400,2.2067325,2.1780167,,,,,,,,,,,,,, +6500,2.628272,2.2342987,,,,,,,,,,,,,, +6574,,,0.63345027,1.5319011,0.57534,1.8242104,50000.0,0.44320002,2.5707448,10000.0,2048.840027332306,2174.077612400055,2048.840027332306,125.14888834953308,0.0383615493774414,0.0 +6600,3.4276085,2.143886,,,,,,,,,,,,,, +6700,2.9612327,2.1813905,,,,,,,,,,,,,, +6800,2.5679724,2.1731935,,,,,,,,,,,,,, +6900,2.4335911,2.1892586,,,,,,,,,,,,,, +7000,2.3358943,2.1699443,,,,,,,,,,,,,, +7100,2.1723597,2.0352473,,,,,,,,,,,,,, +7200,2.4455323,2.0193636,,,,,,,,,,,,,, +7300,2.4453838,1.9815785,,,,,,,,,,,,,, +7400,2.3310165,2.04994,,,,,,,,,,,,,, +7500,2.1564426,2.0957813,,,,,,,,,,,,,, +7600,2.7363708,2.03057,,,,,,,,,,,,,, +7700,2.2771633,1.9784484,,,,,,,,,,,,,, +7800,3.3548832,2.154948,,,,,,,,,,,,,, +7900,2.4959888,2.0939426,,,,,,,,,,,,,, +8000,2.0605214,1.9716021,,,,,,,,,,,,,, +8100,2.0269997,1.9753654,,,,,,,,,,,,,, +8200,2.6592662,1.9484674,,,,,,,,,,,,,, +8300,2.3306727,1.931921,,,,,,,,,,,,,, +8400,2.3693924,2.103004,,,,,,,,,,,,,, +8500,2.8095837,2.0188708,,,,,,,,,,,,,, +8600,2.2074773,1.894985,,,,,,,,,,,,,, +8700,2.711433,1.7700984,,,,,,,,,,,,,, +8800,2.2592149,1.9621513,,,,,,,,,,,,,, +8900,2.6085415,1.7523129,,,,,,,,,,,,,, +9000,2.1136386,1.8664043,,,,,,,,,,,,,, +9100,2.2770007,1.8920763,,,,,,,,,,,,,, +9200,2.6492016,1.9573493,,,,,,,,,,,,,, +9300,3.0132332,1.9327533,,,,,,,,,,,,,, +9400,2.2009568,1.8772725,,,,,,,,,,,,,, +9500,2.341134,1.9010428,,,,,,,,,,,,,, +9600,2.5960746,1.8685433,,,,,,,,,,,,,, +9700,2.2035668,1.8230989,,,,,,,,,,,,,, +9800,2.3515594,1.8992987,,,,,,,,,,,,,, +9900,1.9531469,1.8034528,,,,,,,,,,,,,, +10000,2.0118968,1.7611663,,,,,,,,,,,,,, +10100,2.0668476,1.7491786,,,,,,,,,,,,,, +10200,2.4907792,1.8243978,,,,,,,,,,,,,, +10300,2.2087104,1.8670421,,,,,,,,,,,,,, +10400,2.306542,1.9684355,,,,,,,,,,,,,, +10500,2.5482168,1.8215971,,,,,,,,,,,,,, +10600,2.6003804,1.8267671,,,,,,,,,,,,,, +10700,2.5026364,1.7535398,,,,,,,,,,,,,, +10800,2.3807228,1.9236634,,,,,,,,,,,,,, +10900,2.4224682,1.743006,,,,,,,,,,,,,, +11000,2.4657652,1.8093528,,,,,,,,,,,,,, +11100,2.2118576,1.8176498,,,,,,,,,,,,,, +11200,2.6209748,1.8674273,,,,,,,,,,,,,, +11300,2.8695307,1.8229274,,,,,,,,,,,,,, +11400,3.2597477,1.7282455,,,,,,,,,,,,,, +11500,2.199509,1.683197,,,,,,,,,,,,,, +11600,2.242317,1.7156677,,,,,,,,,,,,,, +11700,2.298526,1.6748354,,,,,,,,,,,,,, +11800,1.9165884,1.6611478,,,,,,,,,,,,,, +11900,2.2776134,1.7896419,,,,,,,,,,,,,, +12000,2.513702,1.7087758,,,,,,,,,,,,,, +12100,2.6357732,1.7630031,,,,,,,,,,,,,, +12200,2.3500385,1.6332464,,,,,,,,,,,,,, +12300,2.1895428,1.5877569,,,,,,,,,,,,,, +12400,2.1415784,1.6569583,,,,,,,,,,,,,, +12500,2.6314452,1.5640115,,,,,,,,,,,,,, +12600,2.172481,1.6280878,,,,,,,,,,,,,, +12700,2.608717,1.7626035,,,,,,,,,,,,,, +12800,2.3205695,1.6876832,,,,,,,,,,,,,, +12900,2.762408,1.8627491,,,,,,,,,,,,,, +13000,2.7997494,1.5938356,,,,,,,,,,,,,, +13100,2.5975933,1.6629461,,,,,,,,,,,,,, +13200,2.4437668,1.5824903,,,,,,,,,,,,,, +13300,2.269699,1.5391333,,,,,,,,,,,,,, +13400,2.3729563,1.6627873,,,,,,,,,,,,,, +13500,2.4979784,1.6674211,,,,,,,,,,,,,, +13600,2.3331096,1.5826015,,,,,,,,,,,,,, +13700,2.6142168,1.6689107,,,,,,,,,,,,,, +13800,2.6323695,1.6393216,,,,,,,,,,,,,, +13900,2.3569791,1.6604097,,,,,,,,,,,,,, +13920,,,0.7597656,0.9458039,0.66716,1.3747368,50000.0,0.5343,2.1329305,10000.0,4044.9069986343375,4192.527692556381,4044.9069986343375,147.4297194480896,0.0868222713470459,0.0 +14000,3.0350115,1.6969239,,,,,,,,,,,,,, +14100,2.621231,1.5532659,,,,,,,,,,,,,, +14200,2.974387,1.6971188,,,,,,,,,,,,,, +14300,2.7357543,1.7021216,,,,,,,,,,,,,, +14400,2.780167,1.6691524,,,,,,,,,,,,,, +14500,2.1496155,1.617668,,,,,,,,,,,,,, +14600,2.5855052,1.6095719,,,,,,,,,,,,,, +14700,2.2697601,1.6139101,,,,,,,,,,,,,, +14800,2.8866594,1.7605444,,,,,,,,,,,,,, +14900,2.4191031,1.6256027,,,,,,,,,,,,,, +15000,2.7074242,1.6248866,,,,,,,,,,,,,, +15100,2.6054132,1.6539142,,,,,,,,,,,,,, +15200,2.5788372,1.5797102,,,,,,,,,,,,,, +15300,2.3922675,1.5725656,,,,,,,,,,,,,, +15400,2.6645756,1.5964327,,,,,,,,,,,,,, +15500,2.7354321,1.6281344,,,,,,,,,,,,,, +15600,2.8365343,1.5315821,,,,,,,,,,,,,, +15700,2.9557157,1.49427,,,,,,,,,,,,,, +15800,2.6223001,1.5085263,,,,,,,,,,,,,, +15900,2.7346373,1.4732732,,,,,,,,,,,,,, +16000,2.642257,1.4495434,,,,,,,,,,,,,, +16100,2.8667867,1.610338,,,,,,,,,,,,,, +16200,3.1004465,1.654129,,,,,,,,,,,,,, +16300,2.8889666,1.4718907,,,,,,,,,,,,,, +16400,3.2106285,1.5856781,,,,,,,,,,,,,, +16500,2.7078297,1.473959,,,,,,,,,,,,,, +16600,2.9253182,1.5339279,,,,,,,,,,,,,, +16700,2.7744844,1.5396224,,,,,,,,,,,,,, +16800,3.092213,1.5274389,,,,,,,,,,,,,, +16900,2.9338717,1.4752641,,,,,,,,,,,,,, +17000,3.4707348,1.4933245,,,,,,,,,,,,,, +17100,3.4258916,1.4522219,,,,,,,,,,,,,, +17200,3.34001,1.5034567,,,,,,,,,,,,,, +17300,3.1978204,1.5897224,,,,,,,,,,,,,, +17400,3.4867997,1.576574,,,,,,,,,,,,,, +17500,3.2303505,1.4816519,,,,,,,,,,,,,, +17600,3.2879238,1.5123013,,,,,,,,,,,,,, +17700,3.326056,1.4286954,,,,,,,,,,,,,, +17800,3.5685642,1.4889247,,,,,,,,,,,,,, +17900,3.4917724,1.5587144,,,,,,,,,,,,,, +18000,3.58451,1.518644,,,,,,,,,,,,,, +18100,4.2018366,1.493459,,,,,,,,,,,,,, +18200,3.9445481,1.4944425,,,,,,,,,,,,,, +18300,4.000143,1.5370042,,,,,,,,,,,,,, +18400,4.2937374,1.5541427,,,,,,,,,,,,,, +18500,3.8750963,1.4579442,,,,,,,,,,,,,, +18600,4.5487585,1.3954844,,,,,,,,,,,,,, +18700,4.0458107,1.518939,,,,,,,,,,,,,, +18800,4.105224,1.492229,,,,,,,,,,,,,, +18900,4.743766,1.478499,,,,,,,,,,,,,, +19000,4.483285,1.5229484,,,,,,,,,,,,,, +19100,4.7624435,1.5833651,,,,,,,,,,,,,, +19200,5.0520024,1.5220892,,,,,,,,,,,,,, +19300,5.0908437,1.4739071,,,,,,,,,,,,,, +19400,5.638278,1.5797584,,,,,,,,,,,,,, +19500,5.053106,1.4385422,,,,,,,,,,,,,, +19600,5.3574443,1.5606914,,,,,,,,,,,,,, +19700,5.5736513,1.5450962,,,,,,,,,,,,,, +19800,6.7417264,1.6309283,,,,,,,,,,,,,, +19900,6.4303493,1.6098533,,,,,,,,,,,,,, +20000,6.626558,1.5774004,,,,,,,,,,,,,, +20100,6.5080824,1.5472186,,,,,,,,,,,,,, +20200,7.016639,1.4600239,,,,,,,,,,,,,, +20300,6.5893774,1.4151797,,,,,,,,,,,,,, +20400,7.282048,1.5558981,,,,,,,,,,,,,, +20500,7.552375,1.5217782,,,,,,,,,,,,,, +20600,7.863426,1.5843675,,,,,,,,,,,,,, +20700,7.922863,1.4261692,,,,,,,,,,,,,, +20800,8.028256,1.4293826,,,,,,,,,,,,,, +20900,10.486676,1.5300984,,,,,,,,,,,,,, +21000,8.677633,1.4619032,,,,,,,,,,,,,, +21100,8.904157,1.5087048,,,,,,,,,,,,,, +21200,9.725544,1.4908292,,,,,,,,,,,,,, +21266,,,0.8064812,0.74569297,0.6971,1.2396864,50000.0,0.5623,1.972755,10000.0,6040.991535186768,6211.1035261154175,6040.991535186768,169.818678855896,0.1349682807922363,0.0 +21300,9.406685,1.5405936,,,,,,,,,,,,,, +21400,9.001268,1.322902,,,,,,,,,,,,,, +21500,10.26239,1.4281567,,,,,,,,,,,,,, +21600,10.113791,1.4933807,,,,,,,,,,,,,, +21700,11.149961,1.4745469,,,,,,,,,,,,,, +21800,10.801688,1.4619765,,,,,,,,,,,,,, +21900,11.783837,1.5049852,,,,,,,,,,,,,, +22000,11.140398,1.4829233,,,,,,,,,,,,,, +22100,11.593483,1.5430199,,,,,,,,,,,,,, +22200,11.263743,1.4632571,,,,,,,,,,,,,, +22300,12.335799,1.4707428,,,,,,,,,,,,,, +22400,11.759468,1.535286,,,,,,,,,,,,,, +22500,11.995282,1.5268123,,,,,,,,,,,,,, +22600,12.468128,1.4467937,,,,,,,,,,,,,, +22700,12.801199,1.6025171,,,,,,,,,,,,,, +22800,12.915553,1.5500178,,,,,,,,,,,,,, +22900,12.996625,1.4899919,,,,,,,,,,,,,, +23000,12.457157,1.4595335,,,,,,,,,,,,,, +23100,13.2329035,1.5153093,,,,,,,,,,,,,, +23200,12.763252,1.5098118,,,,,,,,,,,,,, +23300,12.972905,1.6302419,,,,,,,,,,,,,, +23400,12.69852,1.4136146,,,,,,,,,,,,,, +23500,13.729357,1.5463493,,,,,,,,,,,,,, +23600,13.272187,1.5285623,,,,,,,,,,,,,, +23700,12.644032,1.4883387,,,,,,,,,,,,,, +23800,13.084261,1.4790528,,,,,,,,,,,,,, +23900,12.575275,1.5619521,,,,,,,,,,,,,, +24000,12.304837,1.390585,,,,,,,,,,,,,, +24100,12.258388,1.525451,,,,,,,,,,,,,, +24200,11.880969,1.4036472,,,,,,,,,,,,,, +24300,12.103781,1.4992353,,,,,,,,,,,,,, +24400,11.928204,1.3931465,,,,,,,,,,,,,, +24500,11.7056465,1.4694211,,,,,,,,,,,,,, +24600,12.13845,1.5468781,,,,,,,,,,,,,, +24700,11.193844,1.4395385,,,,,,,,,,,,,, +24800,11.531634,1.4053001,,,,,,,,,,,,,, +24900,11.036751,1.4170511,,,,,,,,,,,,,, +25000,10.839705,1.4573946,,,,,,,,,,,,,, +25100,11.357416,1.4634182,,,,,,,,,,,,,, +25200,10.925623,1.5204108,,,,,,,,,,,,,, +25300,10.379297,1.438975,,,,,,,,,,,,,, +25400,10.401284,1.4607158,,,,,,,,,,,,,, +25500,9.391441,1.3095536,,,,,,,,,,,,,, +25600,10.705259,1.5459126,,,,,,,,,,,,,, +25700,9.696978,1.3967657,,,,,,,,,,,,,, +25800,9.816317,1.4359928,,,,,,,,,,,,,, +25900,9.6503315,1.4513462,,,,,,,,,,,,,, +26000,9.117806,1.335875,,,,,,,,,,,,,, +26100,9.917699,1.4318799,,,,,,,,,,,,,, +26200,8.983199,1.2969193,,,,,,,,,,,,,, +26300,8.571432,1.3564396,,,,,,,,,,,,,, +26400,9.13045,1.4363668,,,,,,,,,,,,,, +26500,8.884771,1.4640765,,,,,,,,,,,,,, +26600,8.978347,1.4533213,,,,,,,,,,,,,, +26700,8.780392,1.4545158,,,,,,,,,,,,,, +26800,8.460188,1.411263,,,,,,,,,,,,,, +26900,8.300876,1.5080321,,,,,,,,,,,,,, +27000,8.671027,1.4209853,,,,,,,,,,,,,, +27100,8.31554,1.4897249,,,,,,,,,,,,,, +27200,7.9790025,1.3534343,,,,,,,,,,,,,, +27300,7.838155,1.4412386,,,,,,,,,,,,,, +27400,8.533997,1.4924297,,,,,,,,,,,,,, +27500,8.105306,1.3942785,,,,,,,,,,,,,, +27600,7.3618937,1.3881619,,,,,,,,,,,,,, +27700,7.707334,1.4619086,,,,,,,,,,,,,, +27800,7.432123,1.3889232,,,,,,,,,,,,,, +27900,7.2372155,1.3477175,,,,,,,,,,,,,, +28000,7.277936,1.331742,,,,,,,,,,,,,, +28100,7.5113544,1.5032631,,,,,,,,,,,,,, +28200,6.9344006,1.4020025,,,,,,,,,,,,,, +28300,6.7510734,1.2943436,,,,,,,,,,,,,, +28400,6.5761485,1.3428833,,,,,,,,,,,,,, +28500,6.855224,1.3868749,,,,,,,,,,,,,, +28600,7.0012617,1.385662,,,,,,,,,,,,,, +28612,,,0.83091515,0.63762105,0.71488,1.1683154,50000.0,0.58250004,1.8919845,10000.0,8036.958885192871,8229.120635509491,8036.958885192871,191.76151204109192,0.1867382526397705,0.0 +28700,7.163887,1.4423733,,,,,,,,,,,,,, +28800,6.726471,1.3722471,,,,,,,,,,,,,, +28900,6.451502,1.3578378,,,,,,,,,,,,,, +29000,6.5456414,1.469851,,,,,,,,,,,,,, +29100,6.1040444,1.2386316,,,,,,,,,,,,,, +29200,6.3644724,1.3190022,,,,,,,,,,,,,, +29300,5.965795,1.2751932,,,,,,,,,,,,,, +29400,6.0860143,1.3182552,,,,,,,,,,,,,, +29500,6.659902,1.372334,,,,,,,,,,,,,, +29600,5.9023814,1.3997685,,,,,,,,,,,,,, +29700,6.285915,1.5150459,,,,,,,,,,,,,, +29800,6.3181524,1.4524887,,,,,,,,,,,,,, +29900,6.7743955,1.3603165,,,,,,,,,,,,,, +30000,6.032089,1.4418997,,,,,,,,,,,,,, +30100,5.5228877,1.3230852,,,,,,,,,,,,,, +30200,6.063093,1.3351341,,,,,,,,,,,,,, +30300,5.768406,1.3954833,,,,,,,,,,,,,, +30400,5.789111,1.3224598,,,,,,,,,,,,,, +30500,5.7178035,1.3184634,,,,,,,,,,,,,, +30600,5.868816,1.3037063,,,,,,,,,,,,,, +30700,5.802487,1.408796,,,,,,,,,,,,,, +30800,5.76387,1.4135809,,,,,,,,,,,,,, +30900,5.700442,1.3622816,,,,,,,,,,,,,, +31000,5.4504642,1.3133823,,,,,,,,,,,,,, +31100,5.201325,1.3084106,,,,,,,,,,,,,, +31200,5.6939,1.4155872,,,,,,,,,,,,,, +31300,5.0067983,1.3106652,,,,,,,,,,,,,, +31400,5.2853756,1.2621176,,,,,,,,,,,,,, +31500,5.336551,1.4012649,,,,,,,,,,,,,, +31600,5.116517,1.3187212,,,,,,,,,,,,,, +31700,5.501377,1.4875959,,,,,,,,,,,,,, +31800,5.6458225,1.383873,,,,,,,,,,,,,, +31900,5.158446,1.3052216,,,,,,,,,,,,,, +32000,5.2547126,1.3737957,,,,,,,,,,,,,, +32100,5.089676,1.2602338,,,,,,,,,,,,,, +32200,5.141697,1.3746885,,,,,,,,,,,,,, +32300,5.07589,1.331913,,,,,,,,,,,,,, +32400,5.106059,1.3676746,,,,,,,,,,,,,, +32500,4.7071714,1.2621331,,,,,,,,,,,,,, +32600,5.10006,1.2911916,,,,,,,,,,,,,, +32700,4.8420253,1.3518192,,,,,,,,,,,,,, +32800,4.8400717,1.3381021,,,,,,,,,,,,,, +32900,4.893854,1.3486198,,,,,,,,,,,,,, +33000,5.0568233,1.321912,,,,,,,,,,,,,, +33100,5.262635,1.2984071,,,,,,,,,,,,,, +33200,4.661323,1.286449,,,,,,,,,,,,,, +33300,5.243497,1.2617567,,,,,,,,,,,,,, +33400,4.586353,1.3256364,,,,,,,,,,,,,, +33500,4.5998816,1.3848494,,,,,,,,,,,,,, +33600,4.79894,1.3619969,,,,,,,,,,,,,, +33700,4.520998,1.1590207,,,,,,,,,,,,,, +33800,4.625613,1.3215344,,,,,,,,,,,,,, +33900,4.527818,1.325979,,,,,,,,,,,,,, +34000,4.7602363,1.3617405,,,,,,,,,,,,,, +34100,4.5494733,1.2068985,,,,,,,,,,,,,, +34200,4.7353053,1.3046237,,,,,,,,,,,,,, +34300,4.5527883,1.2981448,,,,,,,,,,,,,, +34400,4.192521,1.1851058,,,,,,,,,,,,,, +34500,4.5580015,1.2942646,,,,,,,,,,,,,, +34600,4.519845,1.3174038,,,,,,,,,,,,,, +34700,4.4508443,1.3302569,,,,,,,,,,,,,, +34800,4.205676,1.1698258,,,,,,,,,,,,,, +34900,4.5804257,1.3509094,,,,,,,,,,,,,, +35000,4.4362273,1.3542262,,,,,,,,,,,,,, +35100,4.1063433,1.2658732,,,,,,,,,,,,,, +35200,4.472426,1.3692381,,,,,,,,,,,,,, +35300,4.15397,1.1984208,,,,,,,,,,,,,, +35400,4.28467,1.3462679,,,,,,,,,,,,,, +35500,4.36009,1.2508689,,,,,,,,,,,,,, +35600,4.424654,1.2479652,,,,,,,,,,,,,, +35700,4.13845,1.3007168,,,,,,,,,,,,,, +35800,4.1900067,1.2789454,,,,,,,,,,,,,, +35900,4.565182,1.3138453,,,,,,,,,,,,,, +35959,,,0.85227996,0.538871,0.72279996,1.1272918,50000.0,0.59010005,1.8540716,10000.0,10032.998893737791,10247.509325027466,10032.998893737791,213.9108157157898,0.3050100803375244,0.0 +36000,4.2088995,1.2957379,,,,,,,,,,,,,, +36100,4.1479445,1.2280351,,,,,,,,,,,,,, +36200,4.187688,1.2583427,,,,,,,,,,,,,, +36300,4.332422,1.2669443,,,,,,,,,,,,,, +36400,4.3467455,1.2571616,,,,,,,,,,,,,, +36500,4.455009,1.2742817,,,,,,,,,,,,,, +36600,4.033005,1.3168771,,,,,,,,,,,,,, +36700,4.167399,1.3879325,,,,,,,,,,,,,, +36800,4.1825514,1.3181162,,,,,,,,,,,,,, +36900,4.306579,1.3641237,,,,,,,,,,,,,, +37000,3.7651088,1.204195,,,,,,,,,,,,,, +37100,4.1901727,1.2142888,,,,,,,,,,,,,, +37200,4.191022,1.2689383,,,,,,,,,,,,,, +37300,4.128292,1.3420787,,,,,,,,,,,,,, +37400,4.099293,1.2561657,,,,,,,,,,,,,, +37500,4.1285844,1.3287587,,,,,,,,,,,,,, +37600,4.063677,1.2648399,,,,,,,,,,,,,, +37700,4.0302086,1.2339768,,,,,,,,,,,,,, +37800,4.250649,1.2914381,,,,,,,,,,,,,, +37900,4.3553967,1.2763268,,,,,,,,,,,,,, +38000,4.4761744,1.3179142,,,,,,,,,,,,,, +38100,4.205199,1.3093216,,,,,,,,,,,,,, +38200,4.4125757,1.2881683,,,,,,,,,,,,,, +38300,4.1832566,1.3085775,,,,,,,,,,,,,, +38400,4.072553,1.2978271,,,,,,,,,,,,,, +38500,3.9329197,1.1758287,,,,,,,,,,,,,, +38600,4.0503216,1.2662597,,,,,,,,,,,,,, +38700,3.849073,1.1775622,,,,,,,,,,,,,, +38800,4.2786446,1.2504998,,,,,,,,,,,,,, +38900,4.1427555,1.2898879,,,,,,,,,,,,,, +39000,4.0140843,1.1732808,,,,,,,,,,,,,, +39100,4.159515,1.3284045,,,,,,,,,,,,,, +39200,4.136724,1.2364813,,,,,,,,,,,,,, +39300,3.8808453,1.2370865,,,,,,,,,,,,,, +39400,3.809262,1.1478584,,,,,,,,,,,,,, +39500,4.2138376,1.2179761,,,,,,,,,,,,,, +39600,4.0153995,1.2585366,,,,,,,,,,,,,, +39700,4.0058923,1.2478075,,,,,,,,,,,,,, +39800,3.6523702,1.1494145,,,,,,,,,,,,,, +39900,3.8997114,1.2695134,,,,,,,,,,,,,, +40000,4.055899,1.2207327,,,,,,,,,,,,,, +40100,4.0557303,1.242005,,,,,,,,,,,,,, +40200,3.782456,1.1772448,,,,,,,,,,,,,, +40300,3.9004335,1.2059838,,,,,,,,,,,,,, +40400,3.8352633,1.2575256,,,,,,,,,,,,,, +40500,4.0938115,1.2450589,,,,,,,,,,,,,, +40600,3.8813667,1.2040583,,,,,,,,,,,,,, +40700,3.8351758,1.2053324,,,,,,,,,,,,,, +40800,4.0546045,1.3762014,,,,,,,,,,,,,, +40900,3.9939475,1.244741,,,,,,,,,,,,,, +41000,3.8793776,1.256114,,,,,,,,,,,,,, +41100,3.6541548,1.1486545,,,,,,,,,,,,,, +41200,3.6512213,1.1559856,,,,,,,,,,,,,, +41300,3.960081,1.1922839,,,,,,,,,,,,,, +41400,3.8036969,1.1998053,,,,,,,,,,,,,, +41500,3.946402,1.2649186,,,,,,,,,,,,,, +41600,3.968401,1.2477939,,,,,,,,,,,,,, +41700,3.964877,1.2460287,,,,,,,,,,,,,, +41800,3.8166442,1.3088869,,,,,,,,,,,,,, +41900,3.9835963,1.2928442,,,,,,,,,,,,,, +42000,4.0475826,1.2919239,,,,,,,,,,,,,, +42100,3.8659348,1.1910262,,,,,,,,,,,,,, +42200,3.6807845,1.2429824,,,,,,,,,,,,,, +42300,3.9098845,1.1830094,,,,,,,,,,,,,, +42400,3.939172,1.2722433,,,,,,,,,,,,,, +42500,3.7246444,1.1566801,,,,,,,,,,,,,, +42600,3.697497,1.2084401,,,,,,,,,,,,,, +42700,3.6626546,1.2575581,,,,,,,,,,,,,, +42800,3.8989973,1.2413719,,,,,,,,,,,,,, +42900,4.05826,1.302223,,,,,,,,,,,,,, +43000,4.1056833,1.2063618,,,,,,,,,,,,,, +43100,3.785879,1.2155445,,,,,,,,,,,,,, +43200,4.1031876,1.2614989,,,,,,,,,,,,,, +43300,3.7373056,1.2178309,,,,,,,,,,,,,, +43305,,,0.87549824,0.461255,0.72856,1.110315,50000.0,0.6014,1.8353652,10000.0,12029.003738880156,12263.979522943497,12029.003738880156,234.2279772758484,0.3990471363067627,0.0 +43400,3.758775,1.2211493,,,,,,,,,,,,,, +43500,3.838369,1.3141534,,,,,,,,,,,,,, +43600,3.840034,1.2013876,,,,,,,,,,,,,, +43700,3.8022676,1.2862813,,,,,,,,,,,,,, +43800,3.9054875,1.139348,,,,,,,,,,,,,, +43900,3.9180703,1.2099111,,,,,,,,,,,,,, +44000,3.7666147,1.1831305,,,,,,,,,,,,,, +44100,3.9341269,1.2372073,,,,,,,,,,,,,, +44200,3.6843376,1.2083907,,,,,,,,,,,,,, +44300,3.99669,1.2787347,,,,,,,,,,,,,, +44400,3.9241638,1.2437062,,,,,,,,,,,,,, +44500,3.6839964,1.1668802,,,,,,,,,,,,,, +44600,3.646383,1.1201465,,,,,,,,,,,,,, +44700,3.771726,1.2500966,,,,,,,,,,,,,, +44800,3.5588775,1.1874282,,,,,,,,,,,,,, +44900,3.6903758,1.2086264,,,,,,,,,,,,,, +45000,3.8533034,1.199441,,,,,,,,,,,,,, +45100,3.7426927,1.2166877,,,,,,,,,,,,,, +45200,3.8454752,1.0834129,,,,,,,,,,,,,, +45300,3.6512003,1.0973831,,,,,,,,,,,,,, +45400,4.042835,1.264158,,,,,,,,,,,,,, +45500,3.89529,1.23367,,,,,,,,,,,,,, +45600,3.9634361,1.2658185,,,,,,,,,,,,,, +45700,3.663157,1.2126045,,,,,,,,,,,,,, +45800,3.7708585,1.1736377,,,,,,,,,,,,,, +45900,4.1615806,1.2981557,,,,,,,,,,,,,, +46000,3.9752054,1.2058008,,,,,,,,,,,,,, +46100,3.8168378,1.2613497,,,,,,,,,,,,,, +46200,3.8483484,1.2571222,,,,,,,,,,,,,, +46300,3.633871,1.2187614,,,,,,,,,,,,,, +46400,3.81775,1.1392074,,,,,,,,,,,,,, +46500,3.8533263,1.1855035,,,,,,,,,,,,,, +46600,3.7138875,1.1475059,,,,,,,,,,,,,, +46700,3.8574653,1.2164347,,,,,,,,,,,,,, +46800,3.8558087,1.2167228,,,,,,,,,,,,,, +46900,3.8240185,1.1806303,,,,,,,,,,,,,, +47000,3.7516644,1.1926588,,,,,,,,,,,,,, +47100,3.5360892,1.1235671,,,,,,,,,,,,,, +47200,3.7130394,1.2315613,,,,,,,,,,,,,, +47300,3.8925238,1.1290244,,,,,,,,,,,,,, +47400,3.7779324,1.1525757,,,,,,,,,,,,,, +47500,3.9430337,1.164497,,,,,,,,,,,,,, +47600,3.7279005,1.1591728,,,,,,,,,,,,,, +47700,4.0636673,1.2984593,,,,,,,,,,,,,, +47800,3.709247,1.1685615,,,,,,,,,,,,,, +47900,3.7909658,1.1385629,,,,,,,,,,,,,, +48000,3.6654634,1.100507,,,,,,,,,,,,,, +48100,3.959243,1.2706262,,,,,,,,,,,,,, +48200,3.683822,1.2227424,,,,,,,,,,,,,, +48300,3.951264,1.2738168,,,,,,,,,,,,,, +48400,3.8315246,1.256419,,,,,,,,,,,,,, +48500,3.690515,1.2040054,,,,,,,,,,,,,, +48600,3.7551534,1.1426451,,,,,,,,,,,,,, +48700,3.5489802,1.1523899,,,,,,,,,,,,,, +48800,4.0646996,1.1820878,,,,,,,,,,,,,, +48900,3.8195705,1.2134461,,,,,,,,,,,,,, +49000,3.8589845,1.2861208,,,,,,,,,,,,,, +49100,4.076884,1.1019967,,,,,,,,,,,,,, +49200,3.5611022,1.1566298,,,,,,,,,,,,,, +49300,4.0794454,1.1880989,,,,,,,,,,,,,, +49400,3.7222571,1.2842453,,,,,,,,,,,,,, +49500,3.619926,1.2544177,,,,,,,,,,,,,, +49600,3.610679,1.1426619,,,,,,,,,,,,,, +49700,3.7005517,1.1931249,,,,,,,,,,,,,, +49800,4.162051,1.2766393,,,,,,,,,,,,,, +49900,3.8201911,1.1742187,,,,,,,,,,,,,, +50000,4.033286,1.1967845,,,,,,,,,,,,,, +50100,3.719063,1.14393,,,,,,,,,,,,,, +50200,3.9788907,1.1504235,,,,,,,,,,,,,, +50300,3.8005278,1.2322545,,,,,,,,,,,,,, +50400,3.8445477,1.2171235,,,,,,,,,,,,,, +50500,3.5295475,1.1034695,,,,,,,,,,,,,, +50600,3.8341892,1.1428819,,,,,,,,,,,,,, +50652,,,0.8825733,0.42156616,0.73304,1.099628,50000.0,0.6049,1.8242571,10000.0,14025.104043245316,14280.152136325836,14025.104043245316,254.1667010784149,0.4768648147583008,0.0 +50700,3.856446,1.1265738,,,,,,,,,,,,,, +50800,3.7209387,1.2232614,,,,,,,,,,,,,, +50900,3.9475174,1.1787422,,,,,,,,,,,,,, +51000,3.7816567,1.1190157,,,,,,,,,,,,,, +51100,4.438498,1.2693188,,,,,,,,,,,,,, +51200,3.6849558,1.151834,,,,,,,,,,,,,, +51300,3.8855135,1.1746887,,,,,,,,,,,,,, +51400,3.557338,1.1118021,,,,,,,,,,,,,, +51500,3.822354,1.2042451,,,,,,,,,,,,,, +51600,3.8021438,1.1663927,,,,,,,,,,,,,, +51700,3.7539885,1.143114,,,,,,,,,,,,,, +51800,3.6831186,1.171516,,,,,,,,,,,,,, +51900,3.6539068,1.1556969,,,,,,,,,,,,,, +52000,4.1219535,1.0937388,,,,,,,,,,,,,, +52100,4.0955977,1.1905463,,,,,,,,,,,,,, +52200,3.9252086,1.1882751,,,,,,,,,,,,,, +52300,3.8711452,1.2405598,,,,,,,,,,,,,, +52400,3.6672602,1.0790393,,,,,,,,,,,,,, +52500,4.031848,1.198161,,,,,,,,,,,,,, +52600,3.8052483,1.168052,,,,,,,,,,,,,, +52700,4.1768384,1.210728,,,,,,,,,,,,,, +52800,3.8563814,1.2226424,,,,,,,,,,,,,, +52900,3.8286998,1.1992027,,,,,,,,,,,,,, +53000,4.061286,1.2435791,,,,,,,,,,,,,, +53100,3.7742918,1.0716408,,,,,,,,,,,,,, +53200,3.773767,1.1041257,,,,,,,,,,,,,, +53300,4.0174336,1.2508738,,,,,,,,,,,,,, +53400,3.9715,1.1899488,,,,,,,,,,,,,, +53500,4.3113356,1.202406,,,,,,,,,,,,,, +53600,4.308871,1.22746,,,,,,,,,,,,,, +53700,3.9699793,1.2339712,,,,,,,,,,,,,, +53800,4.0780044,1.1912181,,,,,,,,,,,,,, +53900,3.831034,1.2666469,,,,,,,,,,,,,, +54000,4.0323863,1.1796027,,,,,,,,,,,,,, +54100,4.352496,1.2586602,,,,,,,,,,,,,, +54200,3.6723557,1.1189687,,,,,,,,,,,,,, +54300,4.552175,1.2961984,,,,,,,,,,,,,, +54400,4.0918627,1.2088381,,,,,,,,,,,,,, +54500,3.8744678,1.1262267,,,,,,,,,,,,,, +54600,3.6525435,1.0134423,,,,,,,,,,,,,, +54700,3.8955545,1.1288979,,,,,,,,,,,,,, +54800,3.652568,1.0719738,,,,,,,,,,,,,, +54900,3.7272418,1.0911667,,,,,,,,,,,,,, +55000,3.8629835,1.1250968,,,,,,,,,,,,,, +55100,3.9354248,1.2094066,,,,,,,,,,,,,, +55200,4.14725,1.237493,,,,,,,,,,,,,, +55300,3.8530462,1.0989838,,,,,,,,,,,,,, +55400,3.8659694,1.131269,,,,,,,,,,,,,, +55500,4.146842,1.131792,,,,,,,,,,,,,, +55600,4.1163673,1.2117164,,,,,,,,,,,,,, +55700,3.8638272,1.1123538,,,,,,,,,,,,,, +55800,3.8218632,1.1377207,,,,,,,,,,,,,, +55900,3.705408,1.0731494,,,,,,,,,,,,,, +56000,3.802778,1.1962593,,,,,,,,,,,,,, +56100,4.0114374,1.169816,,,,,,,,,,,,,, +56200,3.9915,1.1629682,,,,,,,,,,,,,, +56300,4.3516307,1.1437631,,,,,,,,,,,,,, +56400,4.037383,1.1680291,,,,,,,,,,,,,, +56500,4.216048,1.3193591,,,,,,,,,,,,,, +56600,4.009959,1.1605302,,,,,,,,,,,,,, +56700,4.074022,1.1144714,,,,,,,,,,,,,, +56800,3.7664495,1.0946993,,,,,,,,,,,,,, +56900,3.9570668,1.093939,,,,,,,,,,,,,, +57000,4.007513,1.1606045,,,,,,,,,,,,,, +57100,3.911201,1.1433891,,,,,,,,,,,,,, +57200,3.9955914,1.2456777,,,,,,,,,,,,,, +57300,3.9539754,1.124114,,,,,,,,,,,,,, +57400,4.0560384,0.9987987,,,,,,,,,,,,,, +57500,4.183344,1.1602571,,,,,,,,,,,,,, +57600,4.0076694,1.1833366,,,,,,,,,,,,,, +57700,4.0574794,1.1080097,,,,,,,,,,,,,, +57800,4.268199,1.1539365,,,,,,,,,,,,,, +57900,4.091437,1.1722476,,,,,,,,,,,,,, +58000,,,0.89333546,0.38592175,0.73793995,1.0922809,50000.0,0.60840005,1.827331,10000.0,16021.15643143654,16297.145352840424,16021.15643143654,274.9797277450561,0.5500438213348389,0.0 +58000,4.266199,1.2664751,,,,,,,,,,,,,, +58100,4.271528,1.1736848,,,,,,,,,,,,,, +58200,3.9098234,1.1469874,,,,,,,,,,,,,, +58300,4.2364326,1.2047976,,,,,,,,,,,,,, +58400,3.8811357,1.1175302,,,,,,,,,,,,,, +58500,4.3930254,1.2825935,,,,,,,,,,,,,, +58600,4.021306,1.0993365,,,,,,,,,,,,,, +58700,4.3637843,1.1538273,,,,,,,,,,,,,, +58800,3.8454103,1.0427529,,,,,,,,,,,,,, +58900,3.9080584,1.1944274,,,,,,,,,,,,,, +59000,4.2907887,1.1320474,,,,,,,,,,,,,, +59100,3.986759,1.1571306,,,,,,,,,,,,,, +59200,4.42542,1.2238481,,,,,,,,,,,,,, +59300,4.1663775,1.1755134,,,,,,,,,,,,,, +59400,4.1101885,1.1590779,,,,,,,,,,,,,, +59500,4.1747913,1.1312954,,,,,,,,,,,,,, +59600,4.1313505,1.184025,,,,,,,,,,,,,, +59700,4.1988487,1.1663826,,,,,,,,,,,,,, +59800,4.288027,1.1486768,,,,,,,,,,,,,, +59900,4.1312985,1.133668,,,,,,,,,,,,,, +60000,4.259805,1.1808662,,,,,,,,,,,,,, +60100,4.3595853,1.1758353,,,,,,,,,,,,,, +60200,4.0811357,1.0540582,,,,,,,,,,,,,, +60300,4.3406944,1.1885872,,,,,,,,,,,,,, +60400,4.1503606,1.1544098,,,,,,,,,,,,,, +60500,4.0770993,1.086525,,,,,,,,,,,,,, +60600,4.1834636,1.1093826,,,,,,,,,,,,,, +60700,4.333501,1.2967234,,,,,,,,,,,,,, +60800,4.3776274,1.1511333,,,,,,,,,,,,,, +60900,4.6207724,1.2068123,,,,,,,,,,,,,, +61000,4.273546,1.0745435,,,,,,,,,,,,,, +61100,4.1861777,1.185734,,,,,,,,,,,,,, +61200,4.2603564,1.129649,,,,,,,,,,,,,, +61300,4.1431174,1.1200348,,,,,,,,,,,,,, +61400,4.191992,1.1138744,,,,,,,,,,,,,, +61500,4.136183,1.2025254,,,,,,,,,,,,,, +61600,4.4478493,1.2000651,,,,,,,,,,,,,, +61700,4.166732,1.0973763,,,,,,,,,,,,,, +61800,3.8528018,1.0914725,,,,,,,,,,,,,, +61900,4.2000227,1.1129133,,,,,,,,,,,,,, +62000,3.960378,1.1407962,,,,,,,,,,,,,, +62100,4.15252,1.118896,,,,,,,,,,,,,, +62200,4.3230796,1.099472,,,,,,,,,,,,,, +62300,4.439029,1.1521847,,,,,,,,,,,,,, +62400,4.4056544,1.1520469,,,,,,,,,,,,,, +62500,4.4586806,1.0874255,,,,,,,,,,,,,, +62600,4.1254106,1.083565,,,,,,,,,,,,,, +62700,4.187085,1.069104,,,,,,,,,,,,,, +62800,4.339331,1.1084336,,,,,,,,,,,,,, +62900,4.2554927,1.0886254,,,,,,,,,,,,,, +63000,4.6627398,1.2226794,,,,,,,,,,,,,, +63100,4.64983,1.2214763,,,,,,,,,,,,,, +63200,4.228373,1.1480267,,,,,,,,,,,,,, +63300,4.24091,1.116143,,,,,,,,,,,,,, +63400,4.505836,1.2003531,,,,,,,,,,,,,, +63500,4.397471,1.0809308,,,,,,,,,,,,,, +63600,4.3889093,1.0458391,,,,,,,,,,,,,, +63700,4.265614,1.1787379,,,,,,,,,,,,,, +63800,4.3987827,1.0236881,,,,,,,,,,,,,, +63900,4.1457057,1.053657,,,,,,,,,,,,,, +64000,4.44475,1.1091942,,,,,,,,,,,,,, +64100,4.503634,1.1250671,,,,,,,,,,,,,, +64200,4.771666,1.1823153,,,,,,,,,,,,,, +64300,4.398668,1.171495,,,,,,,,,,,,,, +64400,4.5516863,1.1718559,,,,,,,,,,,,,, +64500,4.6554656,1.1378293,,,,,,,,,,,,,, +64600,4.5468926,1.1397963,,,,,,,,,,,,,, +64700,4.4327855,1.0918393,,,,,,,,,,,,,, +64800,5.0004153,1.2811095,,,,,,,,,,,,,, +64900,4.546041,1.181517,,,,,,,,,,,,,, +65000,4.3732634,1.0954612,,,,,,,,,,,,,, +65100,4.645698,1.0748384,,,,,,,,,,,,,, +65200,4.8896317,1.2568674,,,,,,,,,,,,,, +65300,4.395275,1.1298604,,,,,,,,,,,,,, +65347,,,0.902264,0.35404882,0.74158,1.082871,50000.0,0.60940003,1.8199519,10000.0,18017.071206331253,18312.510173797607,18017.071206331253,294.29212617874146,0.6333343982696533,0.0 +65400,4.1013985,1.0438628,,,,,,,,,,,,,, +65500,4.575092,1.1421041,,,,,,,,,,,,,, +65600,4.8488116,1.1906431,,,,,,,,,,,,,, +65700,4.292194,1.0560154,,,,,,,,,,,,,, +65800,4.673631,1.186329,,,,,,,,,,,,,, +65900,4.711935,1.1794851,,,,,,,,,,,,,, +66000,4.860057,1.1710649,,,,,,,,,,,,,, +66100,4.5957174,1.0941316,,,,,,,,,,,,,, +66200,4.7702646,1.1939869,,,,,,,,,,,,,, +66300,4.395035,1.1467046,,,,,,,,,,,,,, +66400,4.7678843,1.1233928,,,,,,,,,,,,,, +66500,4.7392626,1.1042908,,,,,,,,,,,,,, +66600,4.4891205,1.0323498,,,,,,,,,,,,,, +66700,4.6674447,1.1761155,,,,,,,,,,,,,, +66800,4.850111,1.2334927,,,,,,,,,,,,,, +66900,4.7207894,1.122287,,,,,,,,,,,,,, +67000,4.7252994,1.1959786,,,,,,,,,,,,,, +67100,4.805488,1.1020101,,,,,,,,,,,,,, +67200,4.814242,1.1361172,,,,,,,,,,,,,, +67300,4.710191,1.0817449,,,,,,,,,,,,,, +67400,4.733649,1.1257764,,,,,,,,,,,,,, +67500,4.735122,1.048217,,,,,,,,,,,,,, +67600,4.858154,1.1353048,,,,,,,,,,,,,, +67700,4.8453445,1.1502013,,,,,,,,,,,,,, +67800,5.0091267,1.1272146,,,,,,,,,,,,,, +67900,5.078135,1.1565523,,,,,,,,,,,,,, +68000,4.9953475,1.186733,,,,,,,,,,,,,, +68100,4.884612,1.1477438,,,,,,,,,,,,,, +68200,5.104713,1.0627395,,,,,,,,,,,,,, +68300,4.7650495,1.1504947,,,,,,,,,,,,,, +68400,4.8135657,1.0259149,,,,,,,,,,,,,, +68500,4.8979807,1.1635754,,,,,,,,,,,,,, +68600,5.1913557,1.1521981,,,,,,,,,,,,,, +68700,4.766917,1.1064397,,,,,,,,,,,,,, +68800,4.7429748,1.0124551,,,,,,,,,,,,,, +68900,5.1092186,1.1678386,,,,,,,,,,,,,, +69000,4.915888,1.0976502,,,,,,,,,,,,,, +69100,4.9939275,1.1646918,,,,,,,,,,,,,, +69200,4.7816005,1.1242425,,,,,,,,,,,,,, +69300,4.622041,1.0229255,,,,,,,,,,,,,, +69400,4.9112043,1.1304281,,,,,,,,,,,,,, +69500,4.8161283,1.0334134,,,,,,,,,,,,,, +69600,4.833179,1.1574497,,,,,,,,,,,,,, +69700,4.846081,1.121536,,,,,,,,,,,,,, +69800,4.948748,1.189501,,,,,,,,,,,,,, +69900,4.989412,1.1351153,,,,,,,,,,,,,, +70000,5.0187893,1.1443202,,,,,,,,,,,,,, +70100,5.0282884,1.0329019,,,,,,,,,,,,,, +70200,4.9962077,1.1389139,,,,,,,,,,,,,, +70300,4.8510723,1.1627276,,,,,,,,,,,,,, +70400,4.9225383,1.1140366,,,,,,,,,,,,,, +70500,5.1522565,1.1049968,,,,,,,,,,,,,, +70600,5.018624,1.090923,,,,,,,,,,,,,, +70700,5.124017,1.1382887,,,,,,,,,,,,,, +70800,4.6881695,1.0554055,,,,,,,,,,,,,, +70900,5.0530376,1.0685118,,,,,,,,,,,,,, +71000,5.073338,1.1066813,,,,,,,,,,,,,, +71100,4.900101,1.0835748,,,,,,,,,,,,,, +71200,4.8896112,1.1217282,,,,,,,,,,,,,, +71300,4.8255243,1.0652604,,,,,,,,,,,,,, +71400,5.046197,1.1287489,,,,,,,,,,,,,, +71500,5.1316175,1.1914294,,,,,,,,,,,,,, +71600,5.0314293,1.1474926,,,,,,,,,,,,,, +71700,5.113085,1.1210182,,,,,,,,,,,,,, +71800,5.015654,1.079128,,,,,,,,,,,,,, +71900,5.3342195,1.1214805,,,,,,,,,,,,,, +72000,4.7615294,1.0512674,,,,,,,,,,,,,, +72100,4.807605,1.0979234,,,,,,,,,,,,,, +72200,5.260127,1.2045759,,,,,,,,,,,,,, +72300,5.15654,1.1268678,,,,,,,,,,,,,, +72400,5.1114707,1.1325665,,,,,,,,,,,,,, +72500,5.1741796,1.0892267,,,,,,,,,,,,,, +72600,4.9667826,1.1475439,,,,,,,,,,,,,, +72694,,,0.90927935,0.32634285,0.74307996,1.0753398,50000.0,0.6147,1.8146323,10000.0,20012.96212363243,20327.902194976807,20012.96212363243,313.6133460998535,0.7590162754058838,0.0 +72700,4.6899843,0.97615135,,,,,,,,,,,,,, +72800,4.9431405,1.0665078,,,,,,,,,,,,,, +72900,5.512061,1.1972532,,,,,,,,,,,,,, +73000,5.0363684,1.0724593,,,,,,,,,,,,,, +73100,5.249212,1.1665466,,,,,,,,,,,,,, +73200,5.332741,1.2013266,,,,,,,,,,,,,, +73300,5.324157,1.0642511,,,,,,,,,,,,,, +73400,4.9939175,1.156178,,,,,,,,,,,,,, +73500,5.163724,1.0922663,,,,,,,,,,,,,, +73600,4.9074717,1.0498846,,,,,,,,,,,,,, +73700,4.8206954,1.093143,,,,,,,,,,,,,, +73800,5.0171614,1.0809097,,,,,,,,,,,,,, +73900,5.090183,1.1102786,,,,,,,,,,,,,, +74000,5.0841956,1.1111267,,,,,,,,,,,,,, +74100,5.1485434,1.0917758,,,,,,,,,,,,,, +74200,5.2152987,1.1212313,,,,,,,,,,,,,, +74300,4.7904363,1.0373541,,,,,,,,,,,,,, +74400,5.0878873,1.052003,,,,,,,,,,,,,, +74500,4.849293,1.0941191,,,,,,,,,,,,,, +74600,5.276229,1.1550941,,,,,,,,,,,,,, +74700,4.9015603,1.0848163,,,,,,,,,,,,,, +74800,5.1719303,1.1358135,,,,,,,,,,,,,, +74900,4.9958024,1.1031027,,,,,,,,,,,,,, +75000,5.4661593,1.1077738,,,,,,,,,,,,,, +75100,5.0615077,1.1366379,,,,,,,,,,,,,, +75200,5.077212,1.0915116,,,,,,,,,,,,,, +75300,4.9992113,0.9825882,,,,,,,,,,,,,, +75400,5.197593,1.1111698,,,,,,,,,,,,,, +75500,5.3314877,1.1875978,,,,,,,,,,,,,, +75600,5.0602856,1.1027426,,,,,,,,,,,,,, +75700,5.08697,1.192348,,,,,,,,,,,,,, +75800,5.441875,1.2674406,,,,,,,,,,,,,, +75900,4.949965,1.0476015,,,,,,,,,,,,,, +76000,5.0859857,1.1001587,,,,,,,,,,,,,, +76100,5.0455136,1.1088716,,,,,,,,,,,,,, +76200,5.341018,1.1231399,,,,,,,,,,,,,, +76300,5.3704367,1.0661956,,,,,,,,,,,,,, +76400,4.976427,1.0836548,,,,,,,,,,,,,, +76500,5.4026556,1.1506675,,,,,,,,,,,,,, +76600,4.9655247,1.0757575,,,,,,,,,,,,,, +76700,5.2035623,1.1152128,,,,,,,,,,,,,, +76800,5.209283,1.1114662,,,,,,,,,,,,,, +76900,5.270101,1.1034777,,,,,,,,,,,,,, +77000,4.8308296,1.0300521,,,,,,,,,,,,,, +77100,5.3809524,1.1462505,,,,,,,,,,,,,, +77200,5.091733,1.113556,,,,,,,,,,,,,, +77300,5.3890595,1.1136119,,,,,,,,,,,,,, +77400,5.3050623,1.1488097,,,,,,,,,,,,,, +77500,5.026466,1.1129503,,,,,,,,,,,,,, +77600,4.986379,1.0278469,,,,,,,,,,,,,, +77700,5.2701235,1.0422487,,,,,,,,,,,,,, +77800,5.0935373,1.107171,,,,,,,,,,,,,, +77900,5.197537,1.0779983,,,,,,,,,,,,,, +78000,4.9968553,1.0701377,,,,,,,,,,,,,, +78100,5.239736,1.0844916,,,,,,,,,,,,,, +78200,4.9963965,1.1201118,,,,,,,,,,,,,, +78300,5.0270433,1.0806214,,,,,,,,,,,,,, +78400,5.1397977,1.1045668,,,,,,,,,,,,,, +78500,5.1425457,1.0673243,,,,,,,,,,,,,, +78600,5.02767,1.0249481,,,,,,,,,,,,,, +78700,4.728236,0.9893838,,,,,,,,,,,,,, +78800,5.2896347,1.0834748,,,,,,,,,,,,,, +78900,5.052718,1.0009924,,,,,,,,,,,,,, +79000,4.82037,1.0780344,,,,,,,,,,,,,, +79100,4.7704825,1.0067875,,,,,,,,,,,,,, +79200,4.845875,1.0587311,,,,,,,,,,,,,, +79300,4.9850364,1.080037,,,,,,,,,,,,,, +79400,5.6578145,1.2805588,,,,,,,,,,,,,, +79500,5.17447,1.1091934,,,,,,,,,,,,,, +79600,4.9196935,1.0635347,,,,,,,,,,,,,, +79700,5.0095954,1.1064584,,,,,,,,,,,,,, +79800,5.11172,1.1639938,,,,,,,,,,,,,, +79900,4.948285,1.05159,,,,,,,,,,,,,, +80000,5.069772,1.0621607,,,,,,,,,,,,,, +80043,,,0.91222894,0.31514737,0.74571997,1.0688671,50000.0,0.6169,1.8074372,10000.0,22009.03094482422,22343.13378739357,22009.03094482422,332.6312868595124,0.8456320762634277,0.0 +80100,5.1131988,1.169847,,,,,,,,,,,,,, +80200,4.946498,1.0154576,,,,,,,,,,,,,, +80300,5.2878795,1.133779,,,,,,,,,,,,,, +80400,5.182885,1.0644695,,,,,,,,,,,,,, +80500,4.8877487,1.0537732,,,,,,,,,,,,,, +80600,5.0263906,1.0620787,,,,,,,,,,,,,, +80700,4.8828907,1.0320381,,,,,,,,,,,,,, +80800,4.9460044,1.0686345,,,,,,,,,,,,,, +80900,5.1323714,1.1387948,,,,,,,,,,,,,, +81000,5.0419364,1.0906144,,,,,,,,,,,,,, +81100,5.2103124,1.1718459,,,,,,,,,,,,,, +81200,4.8997874,1.0311811,,,,,,,,,,,,,, +81300,5.077867,1.090095,,,,,,,,,,,,,, +81400,5.1318045,1.0895289,,,,,,,,,,,,,, +81500,5.088109,1.103672,,,,,,,,,,,,,, +81600,5.099239,1.0408028,,,,,,,,,,,,,, +81700,4.827699,1.1043036,,,,,,,,,,,,,, +81800,4.971064,1.0910473,,,,,,,,,,,,,, +81900,5.0135436,1.0581582,,,,,,,,,,,,,, +82000,4.9640903,1.0352639,,,,,,,,,,,,,, +82100,5.274636,1.0858243,,,,,,,,,,,,,, +82200,5.006178,1.024553,,,,,,,,,,,,,, +82300,5.0700665,1.0593002,,,,,,,,,,,,,, +82400,4.902981,1.0498555,,,,,,,,,,,,,, +82500,5.0452766,1.0761722,,,,,,,,,,,,,, +82600,4.914962,1.0793989,,,,,,,,,,,,,, +82700,5.0972834,1.1361911,,,,,,,,,,,,,, +82800,4.956486,1.1064572,,,,,,,,,,,,,, +82900,5.2352395,1.0173517,,,,,,,,,,,,,, +83000,4.9941444,0.9846731,,,,,,,,,,,,,, +83100,4.9594126,1.0473714,,,,,,,,,,,,,, +83200,5.04539,1.0236719,,,,,,,,,,,,,, +83300,5.161511,1.0587641,,,,,,,,,,,,,, +83400,5.181464,1.0942295,,,,,,,,,,,,,, +83500,5.074194,1.0586226,,,,,,,,,,,,,, +83600,5.1713724,1.0009553,,,,,,,,,,,,,, +83700,5.1238685,1.04446,,,,,,,,,,,,,, +83800,5.1548634,1.0726624,,,,,,,,,,,,,, +83900,5.0789886,1.1255805,,,,,,,,,,,,,, +84000,4.7159834,0.9851528,,,,,,,,,,,,,, +84100,5.438589,1.1654451,,,,,,,,,,,,,, +84200,5.2021155,1.0195572,,,,,,,,,,,,,, +84300,5.058208,1.1060691,,,,,,,,,,,,,, +84400,4.837041,1.0481673,,,,,,,,,,,,,, +84500,5.12096,1.1078612,,,,,,,,,,,,,, +84600,4.86999,1.0453006,,,,,,,,,,,,,, +84700,5.0442452,1.0835396,,,,,,,,,,,,,, +84800,5.22209,1.1302387,,,,,,,,,,,,,, +84900,4.9576373,1.0302064,,,,,,,,,,,,,, +85000,4.869359,1.1109618,,,,,,,,,,,,,, +85100,5.034587,1.0882887,,,,,,,,,,,,,, +85200,5.167698,1.0896208,,,,,,,,,,,,,, +85300,5.0002513,1.0972663,,,,,,,,,,,,,, +85400,4.837368,0.9755047,,,,,,,,,,,,,, +85500,4.7857847,1.0627222,,,,,,,,,,,,,, +85600,4.9846253,1.1040779,,,,,,,,,,,,,, +85700,4.884222,1.0519648,,,,,,,,,,,,,, +85800,5.123118,1.074173,,,,,,,,,,,,,, +85900,5.249913,1.1543036,,,,,,,,,,,,,, +86000,4.9746966,1.0515488,,,,,,,,,,,,,, +86100,5.0895796,1.0050457,,,,,,,,,,,,,, +86200,5.067255,1.0853183,,,,,,,,,,,,,, +86300,4.729895,1.0786406,,,,,,,,,,,,,, +86400,4.949068,1.1230841,,,,,,,,,,,,,, +86500,5.0749006,1.1195447,,,,,,,,,,,,,, +86600,5.099457,1.0718032,,,,,,,,,,,,,, +86700,4.877786,1.0005264,,,,,,,,,,,,,, +86800,4.8197565,1.0957747,,,,,,,,,,,,,, +86900,4.9505,1.0673534,,,,,,,,,,,,,, +87000,4.930877,1.0989255,,,,,,,,,,,,,, +87100,4.9982147,1.0931754,,,,,,,,,,,,,, +87200,5.0601563,1.0923957,,,,,,,,,,,,,, +87300,5.031198,1.1320168,,,,,,,,,,,,,, +87391,,,0.91685265,0.29532543,0.74799997,1.0682663,50000.0,0.6222,1.8081042,10000.0,24004.935247659683,24358.087627649307,24004.935247659683,351.5444669723511,0.9268791675567628,0.0 +87400,5.015869,1.0852727,,,,,,,,,,,,,, +87500,4.9341426,1.012235,,,,,,,,,,,,,, +87600,4.7858152,1.0094504,,,,,,,,,,,,,, +87700,4.8194375,1.0377021,,,,,,,,,,,,,, +87800,5.2482977,1.0672426,,,,,,,,,,,,,, +87900,5.049841,1.1824479,,,,,,,,,,,,,, +88000,4.90905,1.1345601,,,,,,,,,,,,,, +88100,4.8712363,1.054158,,,,,,,,,,,,,, +88200,4.7149897,0.98034084,,,,,,,,,,,,,, +88300,4.8513575,1.0380288,,,,,,,,,,,,,, +88400,5.0067377,1.0429959,,,,,,,,,,,,,, +88500,4.8720307,1.0394173,,,,,,,,,,,,,, +88600,4.863335,1.1059322,,,,,,,,,,,,,, +88700,4.7515917,1.0520637,,,,,,,,,,,,,, +88800,4.9439836,0.99090314,,,,,,,,,,,,,, +88900,4.9882364,1.0989187,,,,,,,,,,,,,, +89000,5.2238436,1.010062,,,,,,,,,,,,,, +89100,5.0995555,1.1460268,,,,,,,,,,,,,, +89200,4.8322935,0.9931568,,,,,,,,,,,,,, +89300,4.909796,1.0210639,,,,,,,,,,,,,, +89400,5.0393767,1.0735879,,,,,,,,,,,,,, +89500,4.947811,1.0446227,,,,,,,,,,,,,, +89600,5.052219,1.1364977,,,,,,,,,,,,,, +89700,4.9590635,1.111038,,,,,,,,,,,,,, +89800,4.7353134,0.946989,,,,,,,,,,,,,, +89900,5.279326,1.1187086,,,,,,,,,,,,,, +90000,5.1742086,1.0775628,,,,,,,,,,,,,, +90100,4.7956986,0.9834789,,,,,,,,,,,,,, +90200,5.272673,1.11917,,,,,,,,,,,,,, +90300,5.660727,1.1786872,,,,,,,,,,,,,, +90400,5.172415,1.1467474,,,,,,,,,,,,,, +90500,5.1060367,1.0421147,,,,,,,,,,,,,, +90600,4.9556656,1.0542107,,,,,,,,,,,,,, +90700,4.9489884,1.0779397,,,,,,,,,,,,,, +90800,4.924531,1.1202385,,,,,,,,,,,,,, +90900,4.8799877,1.0456531,,,,,,,,,,,,,, +91000,4.9104,1.0538738,,,,,,,,,,,,,, +91100,5.1652822,1.0837772,,,,,,,,,,,,,, +91200,4.861031,0.9770347,,,,,,,,,,,,,, +91300,4.9131556,1.0742205,,,,,,,,,,,,,, +91400,4.979848,1.0733672,,,,,,,,,,,,,, +91500,5.0065427,1.0828426,,,,,,,,,,,,,, +91600,5.27891,1.0683128,,,,,,,,,,,,,, +91700,5.118815,1.0788077,,,,,,,,,,,,,, +91800,5.1175914,1.0877798,,,,,,,,,,,,,, +91900,5.3716702,1.079496,,,,,,,,,,,,,, +92000,4.9744644,1.0612086,,,,,,,,,,,,,, +92100,5.135711,1.0713037,,,,,,,,,,,,,, +92200,5.0397167,1.0348458,,,,,,,,,,,,,, +92300,5.326661,1.1817472,,,,,,,,,,,,,, +92400,5.176075,1.1199017,,,,,,,,,,,,,, +92500,5.160791,1.136613,,,,,,,,,,,,,, +92600,4.847156,1.0833344,,,,,,,,,,,,,, +92700,4.866338,1.0173717,,,,,,,,,,,,,, +92800,4.7310104,0.97787917,,,,,,,,,,,,,, +92900,5.4378824,1.0558954,,,,,,,,,,,,,, +93000,5.451285,1.1221316,,,,,,,,,,,,,, +93100,5.274907,1.1346939,,,,,,,,,,,,,, +93200,4.73127,0.9297258,,,,,,,,,,,,,, +93300,4.782538,1.0598221,,,,,,,,,,,,,, +93400,4.7977295,0.9936654,,,,,,,,,,,,,, +93500,5.5949597,1.1642555,,,,,,,,,,,,,, +93600,4.7652235,1.0357654,,,,,,,,,,,,,, +93700,5.343917,1.1298788,,,,,,,,,,,,,, +93800,4.9282737,1.0203506,,,,,,,,,,,,,, +93900,5.0251594,1.0504158,,,,,,,,,,,,,, +94000,4.8549056,1.0002855,,,,,,,,,,,,,, +94100,5.069383,1.0210922,,,,,,,,,,,,,, +94200,5.237695,1.1314602,,,,,,,,,,,,,, +94300,5.2064524,1.1304226,,,,,,,,,,,,,, +94400,5.077344,1.056291,,,,,,,,,,,,,, +94500,4.610357,0.97220606,,,,,,,,,,,,,, +94600,5.03667,1.1185843,,,,,,,,,,,,,, +94700,4.9727693,1.0761825,,,,,,,,,,,,,, +94739,,,0.9225127,0.27492654,0.74825996,1.0668399,50000.0,0.6223,1.8116835,10000.0,26000.92212820053,26372.754496335983,26000.92212820053,370.096216917038,1.0008885860443115,0.0 +94800,4.971934,1.029256,,,,,,,,,,,,,, +94900,5.3871617,1.2059314,,,,,,,,,,,,,, +95000,4.5983825,1.0087382,,,,,,,,,,,,,, +95100,4.9457693,1.1177403,,,,,,,,,,,,,, +95200,4.9336615,1.0619826,,,,,,,,,,,,,, +95300,4.8128366,1.0082529,,,,,,,,,,,,,, +95400,4.9936223,1.0789049,,,,,,,,,,,,,, +95500,4.9450784,1.029114,,,,,,,,,,,,,, +95600,5.5061736,1.0796665,,,,,,,,,,,,,, +95700,4.8125243,1.0255804,,,,,,,,,,,,,, +95800,5.0795546,1.0806682,,,,,,,,,,,,,, +95900,4.8437567,1.0355265,,,,,,,,,,,,,, +96000,5.3442397,1.037286,,,,,,,,,,,,,, +96100,5.0466065,1.0690119,,,,,,,,,,,,,, +96200,4.8625293,1.01255,,,,,,,,,,,,,, +96300,4.772211,0.96389896,,,,,,,,,,,,,, +96400,5.2974033,1.0052752,,,,,,,,,,,,,, +96500,4.8313074,1.0771823,,,,,,,,,,,,,, +96600,4.9377627,1.0503266,,,,,,,,,,,,,, +96700,4.926253,1.031436,,,,,,,,,,,,,, +96800,4.8064666,1.0545588,,,,,,,,,,,,,, +96900,5.284399,1.0499146,,,,,,,,,,,,,, +97000,5.0631294,1.0725465,,,,,,,,,,,,,, +97100,4.9528403,1.026155,,,,,,,,,,,,,, +97200,4.983527,1.0607932,,,,,,,,,,,,,, +97300,4.8502336,0.9238431,,,,,,,,,,,,,, +97400,4.8864746,0.92702675,,,,,,,,,,,,,, +97500,4.906455,1.001454,,,,,,,,,,,,,, +97600,4.96418,1.064037,,,,,,,,,,,,,, +97700,4.991408,1.040691,,,,,,,,,,,,,, +97800,5.1578016,1.0418515,,,,,,,,,,,,,, +97900,4.951785,1.0414765,,,,,,,,,,,,,, +98000,4.865336,0.9932012,,,,,,,,,,,,,, +98100,4.90823,1.0820711,,,,,,,,,,,,,, +98200,5.233084,1.1257625,,,,,,,,,,,,,, +98300,4.576355,0.9375049,,,,,,,,,,,,,, +98400,4.6483183,0.9683882,,,,,,,,,,,,,, +98500,5.208369,1.0753921,,,,,,,,,,,,,, +98600,5.0288897,1.039562,,,,,,,,,,,,,, +98700,4.966915,1.0096915,,,,,,,,,,,,,, +98800,5.030628,1.0480418,,,,,,,,,,,,,, +98900,4.924155,0.9764905,,,,,,,,,,,,,, +99000,4.966123,1.0577075,,,,,,,,,,,,,, +99100,4.9284573,1.0316542,,,,,,,,,,,,,, +99200,4.736952,1.0671934,,,,,,,,,,,,,, +99300,4.9404187,1.0320588,,,,,,,,,,,,,, +99400,4.9105377,1.0081161,,,,,,,,,,,,,, +99500,4.896153,1.0679349,,,,,,,,,,,,,, +99600,5.053324,1.0520259,,,,,,,,,,,,,, +99700,4.7733684,0.98437655,,,,,,,,,,,,,, +99800,5.040752,1.0362785,,,,,,,,,,,,,, +99900,4.678988,1.0400318,,,,,,,,,,,,,, +100000,5.248449,1.0327235,,,,,,,,,,,,,, +100100,4.7701435,1.0261389,,,,,,,,,,,,,, +100200,5.76545,1.0709963,,,,,,,,,,,,,, +100300,4.8403316,0.9635717,,,,,,,,,,,,,, +100400,4.840884,1.0655408,,,,,,,,,,,,,, +100500,4.7900114,1.0359106,,,,,,,,,,,,,, +100600,5.4092526,1.033861,,,,,,,,,,,,,, +100700,5.5870733,1.0350655,,,,,,,,,,,,,, +100800,5.060688,1.0704024,,,,,,,,,,,,,, +100900,5.1270385,1.1521246,,,,,,,,,,,,,, +101000,4.9267173,1.0340974,,,,,,,,,,,,,, +101100,5.0273223,1.0406004,,,,,,,,,,,,,, +101200,5.0330763,1.0904372,,,,,,,,,,,,,, +101300,4.825054,1.040417,,,,,,,,,,,,,, +101400,4.8433833,1.0063362,,,,,,,,,,,,,, +101500,4.7445316,0.9552045,,,,,,,,,,,,,, +101600,4.9734483,0.9699378,,,,,,,,,,,,,, +101700,4.9826045,1.0179889,,,,,,,,,,,,,, +101800,4.846098,1.0159227,,,,,,,,,,,,,, +101900,5.068577,1.0338603,,,,,,,,,,,,,, +102000,4.6018443,1.0067147,,,,,,,,,,,,,, +102085,,,0.92524314,0.26571548,0.74825996,1.0670127,50000.0,0.62280005,1.8185028,10000.0,27996.79708981514,28386.60117483139,27996.79708981514,387.9196887016296,1.094411849975586,0.0 +102100,4.8165536,1.0145237,,,,,,,,,,,,,, +102200,4.984593,1.1294681,,,,,,,,,,,,,, +102300,5.106274,0.99789655,,,,,,,,,,,,,, +102400,5.0797663,1.1212626,,,,,,,,,,,,,, +102500,4.828175,1.0665109,,,,,,,,,,,,,, +102600,4.6693797,0.9746636,,,,,,,,,,,,,, +102700,4.991742,1.1489108,,,,,,,,,,,,,, +102800,4.8037853,1.0352353,,,,,,,,,,,,,, +102900,4.9764557,1.0332578,,,,,,,,,,,,,, +103000,5.1560593,1.0299481,,,,,,,,,,,,,, +103100,4.8152423,1.0263474,,,,,,,,,,,,,, +103200,4.935308,1.0087792,,,,,,,,,,,,,, +103300,4.6776104,0.9450888,,,,,,,,,,,,,, +103400,5.112031,1.0652318,,,,,,,,,,,,,, +103500,4.8390865,1.0880541,,,,,,,,,,,,,, +103600,4.8992805,1.0457368,,,,,,,,,,,,,, +103700,5.151355,1.0618922,,,,,,,,,,,,,, +103800,4.6631045,0.9681028,,,,,,,,,,,,,, +103900,4.6520686,1.0008343,,,,,,,,,,,,,, +104000,4.9891534,1.1246748,,,,,,,,,,,,,, +104100,5.0030246,1.1289878,,,,,,,,,,,,,, +104200,4.892868,0.94871056,,,,,,,,,,,,,, +104300,4.9158015,1.0313106,,,,,,,,,,,,,, +104400,4.8994656,0.910297,,,,,,,,,,,,,, +104500,5.3346868,1.1130015,,,,,,,,,,,,,, +104600,5.1613374,1.093242,,,,,,,,,,,,,, +104700,4.978858,1.0265714,,,,,,,,,,,,,, +104800,5.02178,1.0154506,,,,,,,,,,,,,, +104900,5.09,1.1012554,,,,,,,,,,,,,, +105000,5.0728946,1.1202216,,,,,,,,,,,,,, +105100,4.9044104,1.0085796,,,,,,,,,,,,,, +105200,4.782901,1.0124747,,,,,,,,,,,,,, +105300,5.0474954,1.0666139,,,,,,,,,,,,,, +105400,4.9824667,1.0837853,,,,,,,,,,,,,, +105500,5.0978456,1.0608373,,,,,,,,,,,,,, +105600,4.901785,1.0610209,,,,,,,,,,,,,, +105700,5.1392493,0.996705,,,,,,,,,,,,,, +105800,4.9657884,0.98087573,,,,,,,,,,,,,, +105900,4.8126035,0.9211607,,,,,,,,,,,,,, +106000,5.02951,1.1282569,,,,,,,,,,,,,, +106100,4.822311,0.97867465,,,,,,,,,,,,,, +106200,4.602568,0.9673629,,,,,,,,,,,,,, +106300,4.7574406,0.9747516,,,,,,,,,,,,,, +106400,4.7410483,0.99326664,,,,,,,,,,,,,, +106500,4.908258,0.9690332,,,,,,,,,,,,,, +106600,4.9820747,1.04942,,,,,,,,,,,,,, +106700,5.0539837,0.97016275,,,,,,,,,,,,,, +106800,4.9958615,1.0309873,,,,,,,,,,,,,, +106900,4.7925887,1.0513299,,,,,,,,,,,,,, +107000,4.823419,1.0618563,,,,,,,,,,,,,, +107100,5.0164924,1.0420153,,,,,,,,,,,,,, +107200,4.6824603,1.048447,,,,,,,,,,,,,, +107300,4.9568424,0.98710984,,,,,,,,,,,,,, +107400,4.9920335,1.0839574,,,,,,,,,,,,,, +107500,4.862012,1.0089915,,,,,,,,,,,,,, +107600,4.7313714,0.9476348,,,,,,,,,,,,,, +107700,5.0707183,1.0468774,,,,,,,,,,,,,, +107800,4.986874,1.0479662,,,,,,,,,,,,,, +107900,4.962134,1.010965,,,,,,,,,,,,,, +108000,5.0021253,1.0440912,,,,,,,,,,,,,, +108100,4.84435,1.0180464,,,,,,,,,,,,,, +108200,4.861069,1.014976,,,,,,,,,,,,,, +108300,4.550824,0.9304216,,,,,,,,,,,,,, +108400,5.096271,1.0834605,,,,,,,,,,,,,, +108500,4.7714195,0.9449178,,,,,,,,,,,,,, +108600,4.995847,1.022461,,,,,,,,,,,,,, +108700,4.8973136,1.027491,,,,,,,,,,,,,, +108800,5.4782295,1.0703835,,,,,,,,,,,,,, +108900,4.7718186,0.97434974,,,,,,,,,,,,,, +109000,4.874917,1.0296489,,,,,,,,,,,,,, +109100,4.872949,1.0442245,,,,,,,,,,,,,, +109200,4.6780825,0.9946947,,,,,,,,,,,,,, +109300,5.002723,1.026279,,,,,,,,,,,,,, +109400,4.690196,1.029937,,,,,,,,,,,,,, +109432,,,0.92867106,0.24968423,0.74715996,1.068822,50000.0,0.62600005,1.8233396,10000.0,29992.7430062294,30401.084956884384,29992.7430062294,406.3058910369873,1.1918253898620603,0.0 +109500,4.801522,0.96706516,,,,,,,,,,,,,, +109600,4.6309695,0.94665325,,,,,,,,,,,,,, +109700,4.709745,0.926046,,,,,,,,,,,,,, +109800,4.7704463,0.933454,,,,,,,,,,,,,, +109900,4.7483864,0.9647609,,,,,,,,,,,,,, +110000,4.7748294,1.0267948,,,,,,,,,,,,,, +110100,4.913699,1.0453882,,,,,,,,,,,,,, +110200,4.9403057,1.0344093,,,,,,,,,,,,,, +110300,5.0147204,0.9563101,,,,,,,,,,,,,, +110400,4.837789,0.9863468,,,,,,,,,,,,,, +110500,4.854003,1.0487121,,,,,,,,,,,,,, +110600,5.102975,1.0185803,,,,,,,,,,,,,, +110700,4.728213,1.012747,,,,,,,,,,,,,, +110800,4.9214945,1.1082319,,,,,,,,,,,,,, +110900,4.841588,1.0290319,,,,,,,,,,,,,, +111000,4.92126,0.98809916,,,,,,,,,,,,,, +111100,5.6315446,1.1268932,,,,,,,,,,,,,, +111200,4.7993054,1.0329545,,,,,,,,,,,,,, +111300,4.7985682,1.0452944,,,,,,,,,,,,,, +111400,4.5857244,0.92092896,,,,,,,,,,,,,, +111500,5.30459,1.125345,,,,,,,,,,,,,, +111600,4.8758044,1.0154452,,,,,,,,,,,,,, +111700,4.5307283,0.8983118,,,,,,,,,,,,,, +111800,4.7871943,1.0687206,,,,,,,,,,,,,, +111900,5.047826,1.0879747,,,,,,,,,,,,,, +112000,4.706248,1.0193214,,,,,,,,,,,,,, +112100,4.9735284,0.98236823,,,,,,,,,,,,,, +112200,4.7906756,0.96368307,,,,,,,,,,,,,, +112300,4.8535166,0.90861785,,,,,,,,,,,,,, +112400,4.9116173,1.0415003,,,,,,,,,,,,,, +112500,4.8590045,1.0765696,,,,,,,,,,,,,, +112600,4.956579,0.9893778,,,,,,,,,,,,,, +112700,5.241381,1.119113,,,,,,,,,,,,,, +112800,5.1168976,0.929006,,,,,,,,,,,,,, +112900,4.9995403,1.0130777,,,,,,,,,,,,,, +113000,4.833617,1.0511134,,,,,,,,,,,,,, +113100,4.836831,1.0662905,,,,,,,,,,,,,, +113200,4.9219503,1.0437819,,,,,,,,,,,,,, +113300,5.148913,1.065239,,,,,,,,,,,,,, +113400,4.6620884,0.92997336,,,,,,,,,,,,,, +113500,4.667789,0.9499159,,,,,,,,,,,,,, +113600,5.010042,1.1198801,,,,,,,,,,,,,, +113700,5.1212487,0.9405273,,,,,,,,,,,,,, +113800,4.5593066,0.94129765,,,,,,,,,,,,,, +113900,4.7905293,0.9637942,,,,,,,,,,,,,, +114000,4.8400645,1.0616673,,,,,,,,,,,,,, +114100,4.919086,0.9630899,,,,,,,,,,,,,, +114200,4.7565856,0.9676967,,,,,,,,,,,,,, +114300,4.6868496,0.9639516,,,,,,,,,,,,,, +114400,4.7950845,1.0859941,,,,,,,,,,,,,, +114500,4.79559,0.9288429,,,,,,,,,,,,,, +114600,4.9394565,1.1138916,,,,,,,,,,,,,, +114700,4.849266,1.0432489,,,,,,,,,,,,,, +114800,4.51229,1.0457942,,,,,,,,,,,,,, +114900,4.9129457,1.0573994,,,,,,,,,,,,,, +115000,4.9358788,1.0909123,,,,,,,,,,,,,, +115100,4.840479,1.0400784,,,,,,,,,,,,,, +115200,4.7978096,0.96150446,,,,,,,,,,,,,, +115300,5.111965,1.0326382,,,,,,,,,,,,,, +115400,4.977584,1.0777724,,,,,,,,,,,,,, +115500,5.076112,1.0934718,,,,,,,,,,,,,, +115600,4.905673,1.0136691,,,,,,,,,,,,,, +115700,5.1985188,1.1519661,,,,,,,,,,,,,, +115800,4.736073,0.913378,,,,,,,,,,,,,, +115900,4.9523306,1.0094622,,,,,,,,,,,,,, +116000,4.8906217,1.0090384,,,,,,,,,,,,,, +116100,4.8917546,1.0378342,,,,,,,,,,,,,, +116200,4.793153,0.95777047,,,,,,,,,,,,,, +116300,4.9508944,1.021409,,,,,,,,,,,,,, +116400,4.8222466,1.0563765,,,,,,,,,,,,,, +116500,4.8783836,1.0769783,,,,,,,,,,,,,, +116600,5.0492487,1.0727092,,,,,,,,,,,,,, +116700,4.8641443,0.9424585,,,,,,,,,,,,,, +116781,,,0.9322385,0.23868681,0.75016,1.0682027,50000.0,0.62670004,1.8252075,10000.0,31988.6567633152,32414.598189353943,31988.6567633152,423.79282784461975,1.2480707168579102,0.0 +116800,4.905413,1.0330223,,,,,,,,,,,,,, +116900,5.0512547,0.9753528,,,,,,,,,,,,,, +117000,4.620399,0.93169135,,,,,,,,,,,,,, +117100,4.7340326,0.9747113,,,,,,,,,,,,,, +117200,4.9731574,1.0987182,,,,,,,,,,,,,, +117300,4.537647,0.991987,,,,,,,,,,,,,, +117400,4.7273664,1.0061153,,,,,,,,,,,,,, +117500,5.108097,1.1136163,,,,,,,,,,,,,, +117600,4.545338,0.89131355,,,,,,,,,,,,,, +117700,5.0027823,1.039705,,,,,,,,,,,,,, +117800,5.0158057,1.021648,,,,,,,,,,,,,, +117900,4.866134,0.9511409,,,,,,,,,,,,,, +118000,5.296025,1.0942793,,,,,,,,,,,,,, +118100,4.7428536,1.0383934,,,,,,,,,,,,,, +118200,4.8019223,1.0082182,,,,,,,,,,,,,, +118300,5.24807,1.0115995,,,,,,,,,,,,,, +118400,5.025344,1.058372,,,,,,,,,,,,,, +118500,4.903811,1.0537958,,,,,,,,,,,,,, +118600,4.4030967,0.89893323,,,,,,,,,,,,,, +118700,4.4300385,0.92146397,,,,,,,,,,,,,, +118800,5.0304103,0.9861905,,,,,,,,,,,,,, +118900,4.8253593,0.9991658,,,,,,,,,,,,,, +119000,4.6844983,1.060196,,,,,,,,,,,,,, +119100,4.8754244,0.998723,,,,,,,,,,,,,, +119200,4.712451,0.95422184,,,,,,,,,,,,,, +119300,4.8261538,1.0757599,,,,,,,,,,,,,, +119400,5.2146297,1.0947982,,,,,,,,,,,,,, +119500,4.9171863,0.9912127,,,,,,,,,,,,,, +119600,4.8451085,1.0090383,,,,,,,,,,,,,, +119700,4.7925467,0.9623436,,,,,,,,,,,,,, +119800,4.827485,1.0105479,,,,,,,,,,,,,, +119900,5.138244,1.0616235,,,,,,,,,,,,,, +120000,4.867122,0.9511703,,,,,,,,,,,,,, +120100,4.8496614,1.0185947,,,,,,,,,,,,,, +120200,4.7504973,0.86502254,,,,,,,,,,,,,, +120300,4.8653193,1.0140812,,,,,,,,,,,,,, +120400,4.801099,0.9922774,,,,,,,,,,,,,, +120500,4.679156,1.000351,,,,,,,,,,,,,, +120600,4.931273,0.9591998,,,,,,,,,,,,,, +120700,4.883638,1.0076534,,,,,,,,,,,,,, +120800,5.0187387,1.0324793,,,,,,,,,,,,,, +120900,4.772478,0.9745577,,,,,,,,,,,,,, +121000,4.859785,1.0034049,,,,,,,,,,,,,, +121100,4.847606,0.92010534,,,,,,,,,,,,,, +121200,4.77775,1.0270114,,,,,,,,,,,,,, +121300,4.882304,0.96712697,,,,,,,,,,,,,, +121400,5.024407,1.0458624,,,,,,,,,,,,,, +121500,4.802477,1.0041317,,,,,,,,,,,,,, +121600,4.9806914,1.0961485,,,,,,,,,,,,,, +121700,4.4587307,0.9423578,,,,,,,,,,,,,, +121800,4.7078414,0.962086,,,,,,,,,,,,,, +121900,4.5587187,0.9423975,,,,,,,,,,,,,, +122000,5.037104,1.023555,,,,,,,,,,,,,, +122100,4.7199235,1.0190829,,,,,,,,,,,,,, +122200,5.1880717,1.0732079,,,,,,,,,,,,,, +122300,4.876639,0.98969436,,,,,,,,,,,,,, +122400,4.5600524,0.9179948,,,,,,,,,,,,,, +122500,4.935834,1.0466142,,,,,,,,,,,,,, +122600,4.890375,0.9885495,,,,,,,,,,,,,, +122700,5.052093,1.017765,,,,,,,,,,,,,, +122800,5.1417108,1.0142896,,,,,,,,,,,,,, +122900,4.769492,1.0279737,,,,,,,,,,,,,, +123000,4.72066,1.0053718,,,,,,,,,,,,,, +123100,4.8641877,1.0110009,,,,,,,,,,,,,, +123200,4.77547,0.9957265,,,,,,,,,,,,,, +123300,4.9087296,0.9391662,,,,,,,,,,,,,, +123400,4.5276394,0.93366057,,,,,,,,,,,,,, +123500,4.878313,1.032884,,,,,,,,,,,,,, +123600,4.772636,0.93949854,,,,,,,,,,,,,, +123700,4.968863,1.0051432,,,,,,,,,,,,,, +123800,4.949727,1.0437366,,,,,,,,,,,,,, +123900,4.7780323,0.97549486,,,,,,,,,,,,,, +124000,4.721355,0.94223523,,,,,,,,,,,,,, +124100,4.710768,1.0220839,,,,,,,,,,,,,, +124129,,,0.934949,0.22692154,0.75034,1.0701694,50000.0,0.62640005,1.8303214,10000.0,33984.70328807831,34428.11667752266,33984.70328807831,441.1199595928192,1.3386790752410889,0.0 +124200,4.5737886,0.9209982,,,,,,,,,,,,,, +124300,4.830995,0.9547917,,,,,,,,,,,,,, +124400,4.952796,1.0068662,,,,,,,,,,,,,, +124500,5.081885,1.0316589,,,,,,,,,,,,,, +124600,4.9026585,0.99184024,,,,,,,,,,,,,, +124700,4.9617815,1.0519023,,,,,,,,,,,,,, +124800,5.219538,1.0418155,,,,,,,,,,,,,, +124900,4.7678733,1.0012627,,,,,,,,,,,,,, +125000,4.8822308,1.0082824,,,,,,,,,,,,,, +125100,4.934769,1.0723062,,,,,,,,,,,,,, +125200,4.677311,0.95438564,,,,,,,,,,,,,, +125300,4.9155393,1.0936502,,,,,,,,,,,,,, +125400,4.973422,1.0231977,,,,,,,,,,,,,, +125500,5.413673,1.0127783,,,,,,,,,,,,,, +125600,4.9660645,1.0411808,,,,,,,,,,,,,, +125700,5.225246,0.9738194,,,,,,,,,,,,,, +125800,4.876797,1.0271417,,,,,,,,,,,,,, +125900,5.0422716,1.0656103,,,,,,,,,,,,,, +126000,4.799026,0.9911194,,,,,,,,,,,,,, +126100,5.1100097,1.0665646,,,,,,,,,,,,,, +126200,4.877458,0.9479408,,,,,,,,,,,,,, +126300,4.9459853,1.0705506,,,,,,,,,,,,,, +126400,5.1366663,1.0094185,,,,,,,,,,,,,, +126500,5.316727,1.1089749,,,,,,,,,,,,,, +126600,5.015868,1.0500666,,,,,,,,,,,,,, +126700,5.037516,1.042836,,,,,,,,,,,,,, +126800,4.710676,0.9986215,,,,,,,,,,,,,, +126900,4.8464146,0.95550066,,,,,,,,,,,,,, +127000,5.0985045,1.0855991,,,,,,,,,,,,,, +127100,4.994109,0.9869741,,,,,,,,,,,,,, +127200,4.615044,0.9512423,,,,,,,,,,,,,, +127300,4.981158,0.9851112,,,,,,,,,,,,,, +127400,4.7875977,0.9928128,,,,,,,,,,,,,, +127500,5.2898817,1.0777745,,,,,,,,,,,,,, +127600,4.982982,1.0870653,,,,,,,,,,,,,, +127700,4.9591517,1.0438778,,,,,,,,,,,,,, +127800,4.9833703,1.0370998,,,,,,,,,,,,,, +127900,4.6482096,0.9618728,,,,,,,,,,,,,, +128000,4.9036665,1.0325817,,,,,,,,,,,,,, +128100,5.3171353,1.0063818,,,,,,,,,,,,,, +128200,4.8507967,0.9372589,,,,,,,,,,,,,, +128300,4.6848807,0.9390083,,,,,,,,,,,,,, +128400,4.7746267,0.9705465,,,,,,,,,,,,,, +128500,4.4585114,0.97836536,,,,,,,,,,,,,, +128600,4.7465024,0.8961332,,,,,,,,,,,,,, +128700,4.7546206,0.91653085,,,,,,,,,,,,,, +128800,5.2314563,1.0824347,,,,,,,,,,,,,, +128900,5.007485,1.0125215,,,,,,,,,,,,,, +129000,5.1197333,0.93484545,,,,,,,,,,,,,, +129100,4.990144,1.0237384,,,,,,,,,,,,,, +129200,4.7139325,0.92693347,,,,,,,,,,,,,, +129300,4.6993756,0.89535314,,,,,,,,,,,,,, +129400,4.6557107,0.92381966,,,,,,,,,,,,,, +129500,5.032966,1.0032687,,,,,,,,,,,,,, +129600,5.05882,1.0229712,,,,,,,,,,,,,, +129700,5.0108776,1.0092726,,,,,,,,,,,,,, +129800,4.770065,0.94086236,,,,,,,,,,,,,, +129900,4.8255115,0.93745315,,,,,,,,,,,,,, +130000,4.924546,1.0300646,,,,,,,,,,,,,, +130100,4.893179,0.9906937,,,,,,,,,,,,,, +130200,4.788145,1.0364263,,,,,,,,,,,,,, +130300,5.164314,1.0315181,,,,,,,,,,,,,, +130400,4.8989153,0.9652846,,,,,,,,,,,,,, +130500,4.685306,0.95430243,,,,,,,,,,,,,, +130600,4.865205,0.98730886,,,,,,,,,,,,,, +130700,4.9170218,1.0087198,,,,,,,,,,,,,, +130800,4.7693725,1.0094461,,,,,,,,,,,,,, +130900,4.579858,0.86528456,,,,,,,,,,,,,, +131000,4.736083,0.9471941,,,,,,,,,,,,,, +131100,5.1320662,0.9884502,,,,,,,,,,,,,, +131200,4.511147,0.9261792,,,,,,,,,,,,,, +131300,5.067234,1.056392,,,,,,,,,,,,,, +131400,5.0778675,1.0464445,,,,,,,,,,,,,, +131478,,,0.9367626,0.21946916,0.75009996,1.0705539,50000.0,0.6244,1.8374841,10000.0,35980.75158596039,36442.4638531208,35980.75158596039,459.3026838302612,1.3988158702850342,0.0 +131500,4.693057,0.9457774,,,,,,,,,,,,,, +131600,4.896812,1.0125635,,,,,,,,,,,,,, +131700,4.8093534,1.0111111,,,,,,,,,,,,,, +131800,4.8714867,0.9539485,,,,,,,,,,,,,, +131900,4.6823096,0.9816654,,,,,,,,,,,,,, +132000,4.694303,1.0098255,,,,,,,,,,,,,, +132100,4.7635794,0.9284907,,,,,,,,,,,,,, +132200,4.8918147,1.0097082,,,,,,,,,,,,,, +132300,5.057134,1.0623653,,,,,,,,,,,,,, +132400,4.963855,0.9778732,,,,,,,,,,,,,, +132500,5.015321,0.953359,,,,,,,,,,,,,, +132600,5.1587334,0.9580051,,,,,,,,,,,,,, +132700,4.8694496,0.9974463,,,,,,,,,,,,,, +132800,4.914293,0.99360865,,,,,,,,,,,,,, +132900,5.0209994,1.0403692,,,,,,,,,,,,,, +133000,5.026879,1.0304294,,,,,,,,,,,,,, +133100,5.0312753,0.9774796,,,,,,,,,,,,,, +133200,4.735252,0.92722285,,,,,,,,,,,,,, +133300,5.121175,1.054668,,,,,,,,,,,,,, +133400,5.0599546,1.0632153,,,,,,,,,,,,,, +133500,4.871829,0.9675969,,,,,,,,,,,,,, +133600,4.7295146,0.85193723,,,,,,,,,,,,,, +133700,5.232988,1.0175332,,,,,,,,,,,,,, +133800,4.597335,0.9407187,,,,,,,,,,,,,, +133900,5.1849265,1.1002558,,,,,,,,,,,,,, +134000,4.896555,1.0148003,,,,,,,,,,,,,, +134100,5.0786433,1.0784714,,,,,,,,,,,,,, +134200,4.943905,0.94439375,,,,,,,,,,,,,, +134300,4.791806,0.98441446,,,,,,,,,,,,,, +134400,5.0971246,1.0815158,,,,,,,,,,,,,, +134500,4.5262985,0.90646416,,,,,,,,,,,,,, +134600,4.7442584,0.9701177,,,,,,,,,,,,,, +134700,4.7550516,0.90710294,,,,,,,,,,,,,, +134800,5.0671053,1.124654,,,,,,,,,,,,,, +134900,4.553476,0.92281955,,,,,,,,,,,,,, +135000,5.556941,1.0210674,,,,,,,,,,,,,, +135100,4.877062,0.90120876,,,,,,,,,,,,,, +135200,4.807922,0.98804474,,,,,,,,,,,,,, +135300,4.762678,0.9933232,,,,,,,,,,,,,, +135400,5.029172,1.0584276,,,,,,,,,,,,,, +135500,4.9691143,0.98440814,,,,,,,,,,,,,, +135600,4.7554603,0.9155531,,,,,,,,,,,,,, +135700,4.861955,0.92717797,,,,,,,,,,,,,, +135800,4.8152366,1.0280632,,,,,,,,,,,,,, +135900,4.9665065,1.062847,,,,,,,,,,,,,, +136000,4.622301,0.93473065,,,,,,,,,,,,,, +136100,5.026497,0.9974421,,,,,,,,,,,,,, +136200,4.4838486,0.90096366,,,,,,,,,,,,,, +136300,4.8400197,0.93889797,,,,,,,,,,,,,, +136400,5.0297556,0.9433825,,,,,,,,,,,,,, +136500,5.18485,1.0080308,,,,,,,,,,,,,, +136600,5.1050367,1.0315423,,,,,,,,,,,,,, +136700,5.106911,0.96242094,,,,,,,,,,,,,, +136800,4.740869,0.9307385,,,,,,,,,,,,,, +136900,4.9467554,0.9665981,,,,,,,,,,,,,, +137000,5.140576,0.9653907,,,,,,,,,,,,,, +137100,4.808713,0.9474648,,,,,,,,,,,,,, +137200,4.8002005,0.96579736,,,,,,,,,,,,,, +137300,4.869572,0.9417191,,,,,,,,,,,,,, +137400,4.549491,0.9319492,,,,,,,,,,,,,, +137500,4.4692745,0.8801754,,,,,,,,,,,,,, +137600,4.8442917,1.0707017,,,,,,,,,,,,,, +137700,4.740719,0.86784375,,,,,,,,,,,,,, +137800,5.0386043,1.0009265,,,,,,,,,,,,,, +137900,4.915534,1.0007018,,,,,,,,,,,,,, +138000,5.0469537,1.0209492,,,,,,,,,,,,,, +138100,5.2657957,0.97326815,,,,,,,,,,,,,, +138200,4.8287034,0.9235097,,,,,,,,,,,,,, +138300,5.062838,1.0544198,,,,,,,,,,,,,, +138400,4.9225893,1.0045519,,,,,,,,,,,,,, +138500,4.8909693,1.0194856,,,,,,,,,,,,,, +138600,5.104599,1.0369521,,,,,,,,,,,,,, +138700,5.160777,1.0477265,,,,,,,,,,,,,, +138800,4.7293477,0.9454881,,,,,,,,,,,,,, +138828,,,0.9412866,0.20869601,0.75215995,1.0715029,50000.0,0.6236,1.848797,10000.0,37976.86044549942,38456.877252578735,37976.86044549942,477.4918093681336,1.4586880207061768,0.0 +138900,4.823478,0.8900263,,,,,,,,,,,,,, +139000,5.154408,1.0618584,,,,,,,,,,,,,, +139100,4.860782,1.0008203,,,,,,,,,,,,,, +139200,4.9041533,0.954359,,,,,,,,,,,,,, +139300,4.8044176,0.94367504,,,,,,,,,,,,,, +139400,4.730575,0.9402377,,,,,,,,,,,,,, +139500,4.640752,0.8534265,,,,,,,,,,,,,, +139600,4.636614,0.8963083,,,,,,,,,,,,,, +139700,4.8276753,0.9443851,,,,,,,,,,,,,, +139800,4.820111,0.94890434,,,,,,,,,,,,,, +139900,4.7977843,0.91971487,,,,,,,,,,,,,, +140000,4.8932395,0.9636575,,,,,,,,,,,,,, +140100,4.616307,0.96703756,,,,,,,,,,,,,, +140200,5.3360605,1.0877016,,,,,,,,,,,,,, +140300,5.16262,1.0028671,,,,,,,,,,,,,, +140400,4.8315725,1.0165404,,,,,,,,,,,,,, +140500,5.1604714,0.9574215,,,,,,,,,,,,,, +140600,5.002828,0.95097005,,,,,,,,,,,,,, +140700,5.2421126,1.0406849,,,,,,,,,,,,,, +140800,4.744933,0.9733248,,,,,,,,,,,,,, +140900,5.228757,0.9783738,,,,,,,,,,,,,, +141000,5.198481,1.0371671,,,,,,,,,,,,,, +141100,4.8509974,1.0110495,,,,,,,,,,,,,, +141200,4.384256,0.8997439,,,,,,,,,,,,,, +141300,4.955876,0.9624778,,,,,,,,,,,,,, +141400,5.027757,1.0539001,,,,,,,,,,,,,, +141500,4.9287696,0.959399,,,,,,,,,,,,,, +141600,5.5817266,1.0072464,,,,,,,,,,,,,, +141700,4.846467,0.9202703,,,,,,,,,,,,,, +141800,5.16503,0.97033846,,,,,,,,,,,,,, +141900,4.560389,0.9421697,,,,,,,,,,,,,, +142000,5.1072516,0.9940845,,,,,,,,,,,,,, +142100,4.9065104,1.0154629,,,,,,,,,,,,,, +142200,4.9260855,0.9427116,,,,,,,,,,,,,, +142300,4.589767,1.0212046,,,,,,,,,,,,,, +142400,4.9718614,0.96758616,,,,,,,,,,,,,, +142500,4.8314495,0.98619175,,,,,,,,,,,,,, +142600,5.1862655,0.9584446,,,,,,,,,,,,,, +142700,4.9173856,0.9690772,,,,,,,,,,,,,, +142800,5.3495703,1.0747564,,,,,,,,,,,,,, +142900,4.9213305,1.0449921,,,,,,,,,,,,,, +143000,4.966934,0.9152713,,,,,,,,,,,,,, +143100,4.546169,0.8652524,,,,,,,,,,,,,, +143200,5.0474353,1.0079899,,,,,,,,,,,,,, +143300,4.836947,0.96403027,,,,,,,,,,,,,, +143400,4.639582,0.9641321,,,,,,,,,,,,,, +143500,4.659806,0.8832091,,,,,,,,,,,,,, +143600,4.7669845,1.0243117,,,,,,,,,,,,,, +143700,4.656814,1.020699,,,,,,,,,,,,,, +143800,4.874742,0.95482326,,,,,,,,,,,,,, +143900,4.959917,0.9766673,,,,,,,,,,,,,, +144000,4.9218946,0.9540019,,,,,,,,,,,,,, +144100,5.4242244,1.0936083,,,,,,,,,,,,,, +144200,4.77081,0.9375223,,,,,,,,,,,,,, +144300,4.9419465,0.99447036,,,,,,,,,,,,,, +144400,4.9501033,0.92300326,,,,,,,,,,,,,, +144500,4.603264,0.95464593,,,,,,,,,,,,,, +144600,4.8355503,1.0063673,,,,,,,,,,,,,, +144700,5.0416546,0.9782463,,,,,,,,,,,,,, +144800,4.7405543,0.9348578,,,,,,,,,,,,,, +144900,4.888357,0.99113905,,,,,,,,,,,,,, +145000,4.848068,1.024834,,,,,,,,,,,,,, +145100,4.89332,1.0364821,,,,,,,,,,,,,, +145200,4.8316865,0.8924085,,,,,,,,,,,,,, +145300,4.896423,0.92641115,,,,,,,,,,,,,, +145400,5.085191,0.9875624,,,,,,,,,,,,,, +145500,4.982947,1.0134501,,,,,,,,,,,,,, +145600,4.8380165,0.89346176,,,,,,,,,,,,,, +145700,5.111641,0.98299384,,,,,,,,,,,,,, +145800,4.8727164,0.9983542,,,,,,,,,,,,,, +145900,5.316831,1.041575,,,,,,,,,,,,,, +146000,4.8642044,0.94997764,,,,,,,,,,,,,, +146100,4.590548,0.8766248,,,,,,,,,,,,,, +146178,,,0.94202405,0.20745388,0.75236,1.0740654,50000.0,0.62520003,1.8540839,10000.0,39972.98211836815,40471.64096331597,39972.98211836815,496.0162832736969,1.5198278427124023,0.0 +146200,4.6969833,0.92031527,,,,,,,,,,,,,, +146300,4.8906403,1.0387858,,,,,,,,,,,,,, +146400,5.15001,1.0185277,,,,,,,,,,,,,, +146500,4.7734685,0.95226663,,,,,,,,,,,,,, +146600,4.710722,0.98527515,,,,,,,,,,,,,, +146700,4.975688,0.92005306,,,,,,,,,,,,,, +146800,5.0281906,1.0484096,,,,,,,,,,,,,, +146900,5.036391,0.96112734,,,,,,,,,,,,,, +147000,5.1542296,1.0736976,,,,,,,,,,,,,, +147100,4.8869576,0.9591866,,,,,,,,,,,,,, +147200,5.267994,0.977334,,,,,,,,,,,,,, +147300,4.9312377,1.0043814,,,,,,,,,,,,,, +147400,5.114022,0.9818306,,,,,,,,,,,,,, +147500,4.7401195,0.9409015,,,,,,,,,,,,,, +147600,4.8191404,0.95162797,,,,,,,,,,,,,, +147700,4.692425,0.9667174,,,,,,,,,,,,,, +147800,4.842016,0.9538467,,,,,,,,,,,,,, +147900,5.2242155,1.0499055,,,,,,,,,,,,,, +148000,5.0023737,1.0649664,,,,,,,,,,,,,, +148100,4.8628645,0.9025545,,,,,,,,,,,,,, +148200,4.8496375,0.9863834,,,,,,,,,,,,,, +148300,5.1893277,1.0106732,,,,,,,,,,,,,, +148400,4.7538347,0.9074214,,,,,,,,,,,,,, +148500,5.022422,0.89793575,,,,,,,,,,,,,, +148600,5.051462,0.9741717,,,,,,,,,,,,,, +148700,4.870256,0.97674674,,,,,,,,,,,,,, +148800,4.6715775,0.93143505,,,,,,,,,,,,,, +148900,5.204407,1.0205133,,,,,,,,,,,,,, +149000,5.0158696,0.93947786,,,,,,,,,,,,,, +149100,4.9571204,0.9935174,,,,,,,,,,,,,, +149200,4.9401717,0.9533123,,,,,,,,,,,,,, +149300,4.610341,0.9639683,,,,,,,,,,,,,, +149400,5.1636615,1.069518,,,,,,,,,,,,,, +149500,4.9699483,0.93750745,,,,,,,,,,,,,, +149600,4.9002366,1.0006431,,,,,,,,,,,,,, +149700,4.9005,0.9163182,,,,,,,,,,,,,, +149800,4.79676,1.0463302,,,,,,,,,,,,,, +149900,4.8776774,0.99064934,,,,,,,,,,,,,, +150000,4.8536406,0.9414015,,,,,,,,,,,,,, +150100,4.88117,0.9281985,,,,,,,,,,,,,, +150200,5.326756,1.0813929,,,,,,,,,,,,,, +150300,5.198439,1.0039771,,,,,,,,,,,,,, +150400,5.1613784,1.0408866,,,,,,,,,,,,,, +150500,5.2111654,0.9533516,,,,,,,,,,,,,, +150600,4.999195,1.0082045,,,,,,,,,,,,,, +150700,5.162005,0.96829116,,,,,,,,,,,,,, +150800,4.895872,0.9047017,,,,,,,,,,,,,, +150900,4.975243,1.0069423,,,,,,,,,,,,,, +151000,5.2947173,1.0717652,,,,,,,,,,,,,, +151100,4.9595704,0.97604626,,,,,,,,,,,,,, +151200,4.945578,0.98064244,,,,,,,,,,,,,, +151300,5.2264924,0.9860178,,,,,,,,,,,,,, +151400,4.70564,0.89960015,,,,,,,,,,,,,, +151500,5.0143847,0.9587138,,,,,,,,,,,,,, +151600,5.427043,0.888385,,,,,,,,,,,,,, +151700,4.7115326,0.86747134,,,,,,,,,,,,,, +151800,4.9405756,1.0070753,,,,,,,,,,,,,, +151900,4.5315704,0.85522324,,,,,,,,,,,,,, +152000,4.998766,0.9989876,,,,,,,,,,,,,, +152100,4.9909697,0.93614614,,,,,,,,,,,,,, +152200,4.8404613,0.96277624,,,,,,,,,,,,,, +152300,5.073439,0.9943139,,,,,,,,,,,,,, +152400,5.188105,0.9988728,,,,,,,,,,,,,, +152500,4.6870823,0.9415919,,,,,,,,,,,,,, +152600,4.698452,0.8820457,,,,,,,,,,,,,, +152700,4.99519,0.99095726,,,,,,,,,,,,,, +152800,5.1639595,1.0184209,,,,,,,,,,,,,, +152900,4.8657064,1.0252521,,,,,,,,,,,,,, +153000,5.1189375,1.0230038,,,,,,,,,,,,,, +153100,4.989733,0.97869194,,,,,,,,,,,,,, +153200,4.9745626,1.0017662,,,,,,,,,,,,,, +153300,4.8572936,0.976642,,,,,,,,,,,,,, +153400,5.2218995,0.9248753,,,,,,,,,,,,,, +153500,4.836357,0.9237293,,,,,,,,,,,,,, +153527,,,0.9441366,0.19585101,0.75236,1.0734808,50000.0,0.62710005,1.8547612,10000.0,41968.85789465904,42493.38979125023,41968.85789465904,521.7591128349304,1.5930283069610596,0.0 +153600,5.0341606,0.98828745,,,,,,,,,,,,,, +153700,5.2311096,0.97171843,,,,,,,,,,,,,, +153800,4.946824,1.0024484,,,,,,,,,,,,,, +153900,4.8576274,0.9792189,,,,,,,,,,,,,, +154000,4.861319,0.91757536,,,,,,,,,,,,,, +154100,4.7946486,0.9444678,,,,,,,,,,,,,, +154200,5.0887365,1.033653,,,,,,,,,,,,,, +154300,4.904961,0.9074364,,,,,,,,,,,,,, +154400,5.118889,0.9921062,,,,,,,,,,,,,, +154500,5.148547,1.0392872,,,,,,,,,,,,,, +154600,5.3988304,1.0125505,,,,,,,,,,,,,, +154700,4.8694386,0.9486622,,,,,,,,,,,,,, +154800,5.235265,1.0351851,,,,,,,,,,,,,, +154900,5.062743,0.9841555,,,,,,,,,,,,,, +155000,4.780767,0.9402491,,,,,,,,,,,,,, +155100,4.7579393,0.91493523,,,,,,,,,,,,,, +155200,4.800048,0.9138137,,,,,,,,,,,,,, +155300,4.852859,0.90383625,,,,,,,,,,,,,, +155400,4.932058,0.89570653,,,,,,,,,,,,,, +155500,4.96959,0.96247596,,,,,,,,,,,,,, +155600,5.094177,0.9887114,,,,,,,,,,,,,, +155700,5.019676,0.9568639,,,,,,,,,,,,,, +155800,5.0482183,1.00141,,,,,,,,,,,,,, +155900,4.878861,0.9643959,,,,,,,,,,,,,, +156000,4.5849037,0.89951456,,,,,,,,,,,,,, +156100,5.039023,0.9491671,,,,,,,,,,,,,, +156200,4.481903,0.8540571,,,,,,,,,,,,,, +156300,5.051878,0.9942675,,,,,,,,,,,,,, +156400,4.9609666,0.96925634,,,,,,,,,,,,,, +156500,5.1655607,1.0188591,,,,,,,,,,,,,, +156600,4.846595,0.88693774,,,,,,,,,,,,,, +156700,5.205613,1.0169598,,,,,,,,,,,,,, +156800,5.0966396,0.993937,,,,,,,,,,,,,, +156900,5.0161605,0.9745462,,,,,,,,,,,,,, +157000,4.833003,0.93891704,,,,,,,,,,,,,, +157100,5.043244,0.92591655,,,,,,,,,,,,,, +157200,4.8439455,0.90342593,,,,,,,,,,,,,, +157300,5.170467,0.9466367,,,,,,,,,,,,,, +157400,4.8604774,1.0301799,,,,,,,,,,,,,, +157500,4.928747,0.9555377,,,,,,,,,,,,,, +157600,4.8625093,0.92928016,,,,,,,,,,,,,, +157700,5.0635324,0.9759159,,,,,,,,,,,,,, +157800,5.4263844,1.0437652,,,,,,,,,,,,,, +157900,4.8891516,0.98195076,,,,,,,,,,,,,, +158000,5.146425,0.93447423,,,,,,,,,,,,,, +158100,4.619809,0.89140415,,,,,,,,,,,,,, +158200,4.658225,0.8950545,,,,,,,,,,,,,, +158300,4.7932982,0.9192587,,,,,,,,,,,,,, +158400,5.1224675,1.0411048,,,,,,,,,,,,,, +158500,5.1537724,0.91159976,,,,,,,,,,,,,, +158600,5.215651,0.9954379,,,,,,,,,,,,,, +158700,5.0085964,1.014415,,,,,,,,,,,,,, +158800,4.744814,0.87692666,,,,,,,,,,,,,, +158900,4.8997355,0.88751495,,,,,,,,,,,,,, +159000,4.9643283,0.9556666,,,,,,,,,,,,,, +159100,4.9807534,1.0411775,,,,,,,,,,,,,, +159200,5.035143,1.0239049,,,,,,,,,,,,,, +159300,5.203268,0.98543733,,,,,,,,,,,,,, +159400,5.0231075,0.8908726,,,,,,,,,,,,,, +159500,4.8501725,0.9252333,,,,,,,,,,,,,, +159600,4.7738843,0.9374221,,,,,,,,,,,,,, +159700,5.113865,1.0642116,,,,,,,,,,,,,, +159800,5.1255393,1.0054386,,,,,,,,,,,,,, +159900,4.8125935,0.9136924,,,,,,,,,,,,,, +160000,4.953841,0.9339092,,,,,,,,,,,,,, +160100,4.814553,0.89575154,,,,,,,,,,,,,, +160200,5.234737,0.9969123,,,,,,,,,,,,,, +160300,4.9631686,0.9264186,,,,,,,,,,,,,, +160400,4.898416,0.93023026,,,,,,,,,,,,,, +160500,4.9527526,0.92329454,,,,,,,,,,,,,, +160600,4.999945,0.9090071,,,,,,,,,,,,,, +160700,4.993601,1.006202,,,,,,,,,,,,,, +160759,,,0.9484614,0.18218228,0.75246,1.0760537,50000.0,0.62740004,1.861663,10000.0,43964.741797447205,44508.94835686684,43964.741797447205,541.2974145412445,1.670964002609253,0.0 +160800,5.261658,1.0051761,,,,,,,,,,,,,, +160900,4.993445,0.9471278,,,,,,,,,,,,,, +161000,5.0144043,0.9628086,,,,,,,,,,,,,, +161100,5.0714884,0.9343177,,,,,,,,,,,,,, +161200,5.153485,0.935333,,,,,,,,,,,,,, +161300,4.770609,0.864174,,,,,,,,,,,,,, +161400,4.9316564,0.95114326,,,,,,,,,,,,,, +161500,5.000479,0.9763637,,,,,,,,,,,,,, +161600,5.113159,1.0261189,,,,,,,,,,,,,, +161700,5.074276,1.0335367,,,,,,,,,,,,,, +161800,5.1556754,0.951219,,,,,,,,,,,,,, +161900,4.648331,0.94523346,,,,,,,,,,,,,, +162000,5.26716,1.0287483,,,,,,,,,,,,,, +162100,5.0910735,0.91122687,,,,,,,,,,,,,, +162200,4.9432974,0.89975214,,,,,,,,,,,,,, +162300,4.989021,0.9870471,,,,,,,,,,,,,, +162400,5.0978518,0.9663479,,,,,,,,,,,,,, +162500,4.99167,0.9261491,,,,,,,,,,,,,, +162600,4.667116,0.8542503,,,,,,,,,,,,,, +162700,4.596889,0.86612165,,,,,,,,,,,,,, +162800,5.1009326,0.9826566,,,,,,,,,,,,,, +162900,5.2621636,0.9855632,,,,,,,,,,,,,, +163000,5.0757747,1.0296882,,,,,,,,,,,,,, +163100,4.734865,0.9463732,,,,,,,,,,,,,, +163200,4.7441506,0.86463356,,,,,,,,,,,,,, +163300,5.2253547,0.94059956,,,,,,,,,,,,,, +163400,4.965241,0.94290864,,,,,,,,,,,,,, +163500,4.9924636,0.92253816,,,,,,,,,,,,,, +163600,4.7466793,0.8413927,,,,,,,,,,,,,, +163700,5.0604186,0.919266,,,,,,,,,,,,,, +163800,5.1425347,1.033745,,,,,,,,,,,,,, +163900,5.0433736,0.9723869,,,,,,,,,,,,,, +164000,5.162091,0.94388825,,,,,,,,,,,,,, +164100,4.9954987,0.9839959,,,,,,,,,,,,,, +164200,5.195619,0.9774797,,,,,,,,,,,,,, +164300,5.1599517,0.947557,,,,,,,,,,,,,, +164400,5.119515,0.9835315,,,,,,,,,,,,,, +164500,5.260225,0.98629135,,,,,,,,,,,,,, +164600,4.999389,0.87292004,,,,,,,,,,,,,, +164700,5.2493143,0.9495363,,,,,,,,,,,,,, +164800,4.921786,0.9014333,,,,,,,,,,,,,, +164900,4.7809405,0.93779147,,,,,,,,,,,,,, +165000,5.0339723,0.98308116,,,,,,,,,,,,,, +165100,5.083905,0.9609132,,,,,,,,,,,,,, +165200,5.067041,1.0165966,,,,,,,,,,,,,, +165300,5.3332787,0.95642424,,,,,,,,,,,,,, +165400,5.0396905,0.92955697,,,,,,,,,,,,,, +165500,4.85305,0.9068086,,,,,,,,,,,,,, +165536,,,0.9486009,0.1856623,0.75286,1.0761234,50000.0,0.62450004,1.8669696,10000.0,45960.87983632088,46551.746829509735,45960.87983632088,587.8112554550171,1.769763708114624,0.0 +165600,5.152842,1.023554,,,,,,,,,,,,,, +165700,4.8461695,0.94055194,,,,,,,,,,,,,, +165800,4.9945116,0.9140401,,,,,,,,,,,,,, +165900,4.9166875,0.9318278,,,,,,,,,,,,,, +166000,5.2307158,0.90936655,,,,,,,,,,,,,, +166100,4.9255576,0.93304724,,,,,,,,,,,,,, +166200,5.062471,0.9562727,,,,,,,,,,,,,, +166300,5.1890273,0.9611095,,,,,,,,,,,,,, +166400,4.735781,0.8619313,,,,,,,,,,,,,, +166500,4.748887,0.9387449,,,,,,,,,,,,,, +166600,4.938687,0.92040443,,,,,,,,,,,,,, +166700,5.065445,0.92942667,,,,,,,,,,,,,, +166800,5.3260508,0.98965883,,,,,,,,,,,,,, +166900,5.0000443,0.9614711,,,,,,,,,,,,,, +167000,4.9107776,0.93284136,,,,,,,,,,,,,, +167100,5.043282,0.9362128,,,,,,,,,,,,,, +167200,4.9232755,0.8459238,,,,,,,,,,,,,, +167300,5.0962386,1.011502,,,,,,,,,,,,,, +167400,5.08624,0.9223484,,,,,,,,,,,,,, +167500,5.230426,0.96657914,,,,,,,,,,,,,, +167600,5.1910257,0.97638035,,,,,,,,,,,,,, +167700,5.307725,1.0581448,,,,,,,,,,,,,, +167800,5.082458,0.9685336,,,,,,,,,,,,,, +167900,4.56976,0.8071921,,,,,,,,,,,,,, +168000,5.3283873,1.0506202,,,,,,,,,,,,,, +168100,4.840439,0.8947907,,,,,,,,,,,,,, +168200,5.0525208,0.91423243,,,,,,,,,,,,,, +168300,5.283834,0.9668013,,,,,,,,,,,,,, +168400,4.930122,0.9361955,,,,,,,,,,,,,, +168500,4.6637383,0.81897295,,,,,,,,,,,,,, +168600,4.893497,0.9989497,,,,,,,,,,,,,, +168700,4.958419,0.88324016,,,,,,,,,,,,,, +168800,4.9155364,0.9655736,,,,,,,,,,,,,, +168900,5.1232796,0.9623852,,,,,,,,,,,,,, +169000,5.337968,0.967808,,,,,,,,,,,,,, +169100,4.7867913,0.9171232,,,,,,,,,,,,,, +169200,5.0116706,0.93903106,,,,,,,,,,,,,, +169300,4.759863,0.9013236,,,,,,,,,,,,,, +169400,5.5044985,1.04298,,,,,,,,,,,,,, +169500,5.349217,1.0471644,,,,,,,,,,,,,, +169600,5.331132,0.95801604,,,,,,,,,,,,,, +169700,4.9265146,0.93870765,,,,,,,,,,,,,, +169800,4.9497027,0.89570045,,,,,,,,,,,,,, +169900,5.2354465,0.9655469,,,,,,,,,,,,,, +169928,,,0.9453723,0.18652251,0.75308,1.0758554,50000.0,0.62670004,1.8681524,10000.0,47956.30036401749,48568.78353047371,47956.30036401749,608.7469418048859,2.263536214828491,0.0 +170000,4.9984455,0.9480269,,,,,,,,,,,,,, +170100,5.008411,0.88444054,,,,,,,,,,,,,, +170200,5.1033134,0.93548775,,,,,,,,,,,,,, +170300,5.1742597,0.9878,,,,,,,,,,,,,, +170400,5.005757,0.94108033,,,,,,,,,,,,,, +170500,5.1122913,1.012366,,,,,,,,,,,,,, +170600,4.969985,1.0430417,,,,,,,,,,,,,, +170700,5.0887766,0.943095,,,,,,,,,,,,,, +170800,4.821752,0.932933,,,,,,,,,,,,,, +170900,5.208303,0.9638181,,,,,,,,,,,,,, +171000,4.9247227,0.87365806,,,,,,,,,,,,,, +171100,5.1695533,0.9676844,,,,,,,,,,,,,, +171200,5.004554,0.9939327,,,,,,,,,,,,,, +171300,4.776684,0.9372167,,,,,,,,,,,,,, +171400,5.1402273,0.9885636,,,,,,,,,,,,,, +171500,5.2846427,0.9524603,,,,,,,,,,,,,, +171600,5.1721735,0.97962594,,,,,,,,,,,,,, +171700,5.593189,0.99597466,,,,,,,,,,,,,, +171800,5.0241704,0.9131294,,,,,,,,,,,,,, +171900,5.120052,0.94852316,,,,,,,,,,,,,, +172000,5.104409,0.90507007,,,,,,,,,,,,,, +172100,5.3043904,0.9672328,,,,,,,,,,,,,, +172200,5.172736,0.9962969,,,,,,,,,,,,,, +172300,4.750158,0.8858863,,,,,,,,,,,,,, +172400,5.1699114,0.9299823,,,,,,,,,,,,,, +172500,4.560922,0.825485,,,,,,,,,,,,,, +172600,4.8781595,0.92298317,,,,,,,,,,,,,, +172700,5.2076297,0.9719002,,,,,,,,,,,,,, +172800,4.814147,0.94146156,,,,,,,,,,,,,, +172900,5.244767,0.9907903,,,,,,,,,,,,,, +173000,5.095126,0.91785175,,,,,,,,,,,,,, +173100,4.9282928,0.865515,,,,,,,,,,,,,, +173200,5.063684,0.9688129,,,,,,,,,,,,,, +173300,4.9021363,0.9553601,,,,,,,,,,,,,, +173400,4.786781,0.84942925,,,,,,,,,,,,,, +173500,5.114274,0.9875612,,,,,,,,,,,,,, +173600,5.05738,0.9982177,,,,,,,,,,,,,, +173700,5.4497447,0.96833944,,,,,,,,,,,,,, +173800,5.1624126,0.98929965,,,,,,,,,,,,,, +173900,4.8760767,0.9287993,,,,,,,,,,,,,, +174000,4.851548,1.0099708,,,,,,,,,,,,,, +174100,5.410206,1.0253218,,,,,,,,,,,,,, +174200,5.183514,0.9238368,,,,,,,,,,,,,, +174300,5.0809116,0.995248,,,,,,,,,,,,,, +174400,5.0131035,0.92520773,,,,,,,,,,,,,, +174418,,,0.9489995,0.18063688,0.75373995,1.076455,50000.0,0.6291,1.8668636,10000.0,49951.91783928871,50590.50589442253,49951.91783928871,634.1093051433563,2.748711347579956,0.0 +174500,5.036942,1.0144895,,,,,,,,,,,,,, +174600,4.719867,0.8249006,,,,,,,,,,,,,, +174700,5.120421,0.8937489,,,,,,,,,,,,,, +174800,4.8922763,0.9498565,,,,,,,,,,,,,, +174900,5.2343516,0.90193987,,,,,,,,,,,,,, +175000,5.0992794,0.9688784,,,,,,,,,,,,,, +175100,5.000911,0.9047041,,,,,,,,,,,,,, +175200,5.1290555,0.9797969,,,,,,,,,,,,,, +175300,5.034257,0.9143141,,,,,,,,,,,,,, +175400,4.902943,0.95259476,,,,,,,,,,,,,, +175500,4.944501,0.8661345,,,,,,,,,,,,,, +175600,5.275557,0.9017285,,,,,,,,,,,,,, +175700,5.126555,0.9883629,,,,,,,,,,,,,, +175800,5.0009403,0.9335717,,,,,,,,,,,,,, +175900,5.322895,1.0109267,,,,,,,,,,,,,, +176000,5.20842,0.9239348,,,,,,,,,,,,,, +176100,5.1726613,0.98684967,,,,,,,,,,,,,, +176200,5.0655894,0.9173806,,,,,,,,,,,,,, +176300,5.212427,0.94204676,,,,,,,,,,,,,, +176400,4.928747,0.9401436,,,,,,,,,,,,,, +176500,5.213183,0.98827493,,,,,,,,,,,,,, +176600,5.190755,0.9286671,,,,,,,,,,,,,, +176700,5.2113004,0.98265946,,,,,,,,,,,,,, +176800,4.938755,1.0394479,,,,,,,,,,,,,, +176900,5.0952997,0.98197055,,,,,,,,,,,,,, +177000,5.1285305,0.95700824,,,,,,,,,,,,,, +177100,5.148158,0.9816646,,,,,,,,,,,,,, +177200,5.0597153,0.9705163,,,,,,,,,,,,,, +177300,5.0406494,0.9050668,,,,,,,,,,,,,, +177400,5.232636,1.0322815,,,,,,,,,,,,,, +177500,5.0346665,0.9409811,,,,,,,,,,,,,, +177600,5.083625,1.0042052,,,,,,,,,,,,,, +177700,5.054672,0.93323743,,,,,,,,,,,,,, +177800,4.885261,0.8499916,,,,,,,,,,,,,, +177900,4.9462905,0.9739598,,,,,,,,,,,,,, +178000,4.735991,0.86484694,,,,,,,,,,,,,, +178100,5.039835,0.9000094,,,,,,,,,,,,,, +178200,5.075594,0.9157424,,,,,,,,,,,,,, +178300,4.948076,0.94641227,,,,,,,,,,,,,, +178400,4.9398665,0.8853956,,,,,,,,,,,,,, +178500,5.1525974,1.0153712,,,,,,,,,,,,,, +178600,5.2327666,0.96011287,,,,,,,,,,,,,, +178700,4.925015,0.91871405,,,,,,,,,,,,,, +178800,5.127267,0.956084,,,,,,,,,,,,,, +178846,,,0.94782364,0.1783816,0.75402,1.0778383,50000.0,0.6283,1.8672477,10000.0,51947.29002213478,52625.81468272209,51947.29002213478,673.3249371051788,3.187523365020752,0.0 +178900,4.7738566,0.9123995,,,,,,,,,,,,,, +179000,4.9198766,0.9970912,,,,,,,,,,,,,, +179100,5.0532913,0.9524287,,,,,,,,,,,,,, +179200,5.3425074,0.96323335,,,,,,,,,,,,,, +179300,5.1845365,0.949697,,,,,,,,,,,,,, +179400,5.214295,0.9824343,,,,,,,,,,,,,, +179500,5.1163926,0.9222437,,,,,,,,,,,,,, +179600,5.161383,0.9583783,,,,,,,,,,,,,, +179700,5.0293884,0.8476839,,,,,,,,,,,,,, +179800,5.0402894,0.90521246,,,,,,,,,,,,,, +179900,5.062415,0.85462457,,,,,,,,,,,,,, +180000,5.2312965,0.8575758,,,,,,,,,,,,,, +180100,4.849779,0.8113138,,,,,,,,,,,,,, +180200,5.097011,0.9459634,,,,,,,,,,,,,, +180300,4.910517,0.944772,,,,,,,,,,,,,, +180400,5.0186453,0.9116261,,,,,,,,,,,,,, +180500,5.5011196,0.97363204,,,,,,,,,,,,,, +180600,5.0210023,0.9508739,,,,,,,,,,,,,, +180700,4.840489,0.86090726,,,,,,,,,,,,,, +180800,5.266316,0.929271,,,,,,,,,,,,,, +180900,4.8468018,0.8942833,,,,,,,,,,,,,, +181000,4.987508,0.9300423,,,,,,,,,,,,,, +181100,5.115768,0.937235,,,,,,,,,,,,,, +181200,5.002644,0.8996254,,,,,,,,,,,,,, +181300,4.8690004,0.8720096,,,,,,,,,,,,,, +181400,5.159334,0.9714869,,,,,,,,,,,,,, +181500,4.869708,0.9027443,,,,,,,,,,,,,, +181600,5.202916,0.9343599,,,,,,,,,,,,,, +181700,4.8246655,0.9024069,,,,,,,,,,,,,, +181800,5.3873053,0.9736725,,,,,,,,,,,,,, +181900,4.835203,0.9001156,,,,,,,,,,,,,, +182000,4.637961,0.8328963,,,,,,,,,,,,,, +182100,5.1180224,0.97406274,,,,,,,,,,,,,, +182200,5.2687507,0.9869485,,,,,,,,,,,,,, +182300,4.7618613,0.88352716,,,,,,,,,,,,,, +182400,5.1224566,0.8871628,,,,,,,,,,,,,, +182500,4.992258,0.89370286,,,,,,,,,,,,,, +182600,4.682897,0.8550284,,,,,,,,,,,,,, +182700,5.14408,0.9484037,,,,,,,,,,,,,, +182800,5.1197777,0.9539137,,,,,,,,,,,,,, +182900,4.9297843,0.87992334,,,,,,,,,,,,,, +183000,4.887799,0.8901391,,,,,,,,,,,,,, +183100,4.943494,0.8662341,,,,,,,,,,,,,, +183200,5.190832,1.0011495,,,,,,,,,,,,,, +183300,4.9794555,0.95684373,,,,,,,,,,,,,, +183400,5.206515,0.9533624,,,,,,,,,,,,,, +183424,,,0.9513512,0.17198183,0.75394,1.0774896,50000.0,0.629,1.8643882,10000.0,53942.5781993866,54664.01664042473,53942.5781993866,715.5178511142731,3.6479856967926025,0.0 +183500,5.110367,0.8879999,,,,,,,,,,,,,, +183600,5.357843,0.9694779,,,,,,,,,,,,,, +183700,5.0927086,0.9667546,,,,,,,,,,,,,, +183800,4.9153976,0.9160719,,,,,,,,,,,,,, +183900,4.934108,0.8932394,,,,,,,,,,,,,, +184000,4.7495494,0.8645725,,,,,,,,,,,,,, +184100,4.746131,0.89900196,,,,,,,,,,,,,, +184200,4.960619,1.0215876,,,,,,,,,,,,,, +184300,5.0422573,0.9675811,,,,,,,,,,,,,, +184400,4.9430337,0.90562797,,,,,,,,,,,,,, +184500,4.8463283,0.88127255,,,,,,,,,,,,,, +184600,5.0790954,0.92456114,,,,,,,,,,,,,, +184700,5.189856,0.96176314,,,,,,,,,,,,,, +184800,5.138074,0.9528495,,,,,,,,,,,,,, +184900,4.6970854,0.9094183,,,,,,,,,,,,,, +185000,5.1036005,0.9181908,,,,,,,,,,,,,, +185100,5.0743923,0.8846005,,,,,,,,,,,,,, +185200,5.203348,0.9448322,,,,,,,,,,,,,, +185300,5.025396,0.9068973,,,,,,,,,,,,,, +185400,5.1111283,0.88138604,,,,,,,,,,,,,, +185500,4.922669,0.92931265,,,,,,,,,,,,,, +185600,5.185923,0.9751317,,,,,,,,,,,,,, +185700,5.0903134,0.95289946,,,,,,,,,,,,,, +185800,4.994513,0.8622793,,,,,,,,,,,,,, +185900,5.013796,0.9743435,,,,,,,,,,,,,, +186000,4.799688,0.81425184,,,,,,,,,,,,,, +186100,4.993161,0.9759148,,,,,,,,,,,,,, +186200,4.572606,0.82138264,,,,,,,,,,,,,, +186300,5.006608,0.97286355,,,,,,,,,,,,,, +186400,5.5369763,0.9993459,,,,,,,,,,,,,, +186500,5.3034134,0.9438823,,,,,,,,,,,,,, +186600,5.533377,0.9145151,,,,,,,,,,,,,, +186700,5.0441008,0.8958327,,,,,,,,,,,,,, +186800,5.1234336,0.9352436,,,,,,,,,,,,,, +186900,5.0969844,0.8700875,,,,,,,,,,,,,, +187000,5.1776156,0.9319978,,,,,,,,,,,,,, +187100,5.174819,0.93573713,,,,,,,,,,,,,, +187200,4.944274,0.8643768,,,,,,,,,,,,,, +187300,4.9854345,0.83619016,,,,,,,,,,,,,, +187400,5.2047105,0.8938449,,,,,,,,,,,,,, +187500,4.7936473,0.9224609,,,,,,,,,,,,,, +187529,,,0.950275,0.17458051,0.75412,1.0794177,50000.0,0.6301,1.8626338,10000.0,55937.934849500656,56699.74064064026,55937.934849500656,755.1613154411316,4.11983323097229,0.0 +187600,5.0552874,0.8531773,,,,,,,,,,,,,, +187700,5.044179,0.94916266,,,,,,,,,,,,,, +187800,4.881084,0.8825333,,,,,,,,,,,,,, +187900,5.276174,1.0084288,,,,,,,,,,,,,, +188000,5.390972,0.9985031,,,,,,,,,,,,,, +188100,5.2001,1.0020059,,,,,,,,,,,,,, +188200,5.1662536,0.9224262,,,,,,,,,,,,,, +188300,5.234635,0.9185755,,,,,,,,,,,,,, +188400,5.326428,0.9326956,,,,,,,,,,,,,, +188500,5.12849,0.85715365,,,,,,,,,,,,,, +188600,5.2474494,0.98667973,,,,,,,,,,,,,, +188700,4.7367125,0.8138906,,,,,,,,,,,,,, +188800,5.008528,0.8929577,,,,,,,,,,,,,, +188900,5.399804,0.9717091,,,,,,,,,,,,,, +189000,5.02671,0.8971951,,,,,,,,,,,,,, +189100,4.8860674,0.8346412,,,,,,,,,,,,,, +189200,5.2883186,0.91543674,,,,,,,,,,,,,, +189300,5.068224,0.9055982,,,,,,,,,,,,,, +189400,5.0378213,0.8800858,,,,,,,,,,,,,, +189500,5.3147793,0.96141523,,,,,,,,,,,,,, +189600,4.8515778,0.9537127,,,,,,,,,,,,,, +189700,5.221658,0.9315383,,,,,,,,,,,,,, +189800,5.2847457,1.0015546,,,,,,,,,,,,,, +189900,5.157924,0.9224256,,,,,,,,,,,,,, +190000,4.778889,0.8880926,,,,,,,,,,,,,, +190100,4.892291,0.91816056,,,,,,,,,,,,,, +190200,4.901444,0.8633422,,,,,,,,,,,,,, +190300,4.895757,0.9246034,,,,,,,,,,,,,, +190400,4.9971046,0.94478106,,,,,,,,,,,,,, +190500,5.3188543,0.9533683,,,,,,,,,,,,,, +190600,4.854382,0.88718146,,,,,,,,,,,,,, +190700,5.1930075,0.9621534,,,,,,,,,,,,,, +190800,5.226264,0.92428505,,,,,,,,,,,,,, +190900,5.1542816,0.98049414,,,,,,,,,,,,,, +191000,5.1901455,0.9566804,,,,,,,,,,,,,, +191100,5.3373475,0.9686191,,,,,,,,,,,,,, +191200,5.126772,0.8767175,,,,,,,,,,,,,, +191248,,,0.9526666,0.16960119,0.75324,1.0773498,50000.0,0.62750006,1.8607931,10000.0,57933.47280240059,58743.22166013718,57933.47280240059,802.2567172050476,4.655627489089966,0.0 +191300,4.9652224,0.9512378,,,,,,,,,,,,,, +191400,5.2142234,0.91742015,,,,,,,,,,,,,, +191500,5.2505226,0.95766276,,,,,,,,,,,,,, +191600,5.2832937,0.9859845,,,,,,,,,,,,,, +191700,5.1497536,0.94144666,,,,,,,,,,,,,, +191800,5.2550054,0.9645185,,,,,,,,,,,,,, +191900,5.094381,0.95265037,,,,,,,,,,,,,, +192000,4.8804154,0.9083675,,,,,,,,,,,,,, +192100,4.588733,0.84352493,,,,,,,,,,,,,, +192200,5.4014254,1.0030234,,,,,,,,,,,,,, +192300,5.2318754,0.9462592,,,,,,,,,,,,,, +192400,4.936216,0.957001,,,,,,,,,,,,,, +192500,5.0360556,0.9078521,,,,,,,,,,,,,, +192600,5.109738,0.8998678,,,,,,,,,,,,,, +192700,5.1984825,0.95316267,,,,,,,,,,,,,, +192800,5.020749,0.92571783,,,,,,,,,,,,,, +192900,5.0785937,0.89888847,,,,,,,,,,,,,, +193000,5.1658983,0.942782,,,,,,,,,,,,,, +193100,5.1011734,0.89900315,,,,,,,,,,,,,, +193200,5.036683,0.91129434,,,,,,,,,,,,,, +193300,4.9399967,0.902738,,,,,,,,,,,,,, +193400,5.021849,0.9198102,,,,,,,,,,,,,, +193500,4.99968,0.8867873,,,,,,,,,,,,,, +193600,5.0576777,0.906448,,,,,,,,,,,,,, +193700,5.371496,0.89949894,,,,,,,,,,,,,, +193800,4.8367596,0.8914553,,,,,,,,,,,,,, +193900,5.1628833,0.8961816,,,,,,,,,,,,,, +194000,5.589412,1.0284178,,,,,,,,,,,,,, +194100,5.445899,1.0103502,,,,,,,,,,,,,, +194200,4.789029,0.9262799,,,,,,,,,,,,,, +194300,4.9817066,0.9026138,,,,,,,,,,,,,, +194400,5.0112696,0.89669216,,,,,,,,,,,,,, +194500,5.1814523,0.9410533,,,,,,,,,,,,,, +194600,5.097622,0.93243617,,,,,,,,,,,,,, +194700,5.332059,0.9835844,,,,,,,,,,,,,, +194800,5.2144628,0.95469207,,,,,,,,,,,,,, +194900,4.890124,0.86769426,,,,,,,,,,,,,, +195000,5.3274894,1.0178063,,,,,,,,,,,,,, +195063,,,0.952547,0.17043896,0.75376,1.0782413,50000.0,0.62990004,1.8617669,10000.0,59929.071627378464,60776.140051841736,59929.071627378464,838.8255090713501,5.14893364906311,0.0 +195100,4.9622316,0.9323058,,,,,,,,,,,,,, +195200,5.3553753,1.0366966,,,,,,,,,,,,,, +195300,5.1472573,0.9325548,,,,,,,,,,,,,, +195400,5.1162066,0.9177793,,,,,,,,,,,,,, +195500,4.9716315,0.9017098,,,,,,,,,,,,,, +195600,5.568151,0.95567024,,,,,,,,,,,,,, +195700,4.8871107,0.8821664,,,,,,,,,,,,,, +195800,5.168012,0.9795944,,,,,,,,,,,,,, +195900,4.8276925,0.9166574,,,,,,,,,,,,,, +196000,4.9856925,0.9455525,,,,,,,,,,,,,, +196100,5.427358,0.93924296,,,,,,,,,,,,,, +196200,5.0155396,0.9030016,,,,,,,,,,,,,, +196300,5.0887656,0.8802263,,,,,,,,,,,,,, +196400,5.157572,1.0119542,,,,,,,,,,,,,, +196500,5.0355244,0.86223006,,,,,,,,,,,,,, +196600,5.1968107,0.95267105,,,,,,,,,,,,,, +196700,5.130921,0.89862525,,,,,,,,,,,,,, +196800,5.287567,0.9231607,,,,,,,,,,,,,, +196900,5.1379232,0.8977867,,,,,,,,,,,,,, +197000,4.969734,0.9156674,,,,,,,,,,,,,, +197100,4.644208,0.8736707,,,,,,,,,,,,,, +197200,5.3574615,0.9754261,,,,,,,,,,,,,, +197300,4.856748,0.8386726,,,,,,,,,,,,,, +197400,4.91269,0.89179945,,,,,,,,,,,,,, +197500,5.24994,0.9795364,,,,,,,,,,,,,, +197600,5.090019,0.87519664,,,,,,,,,,,,,, +197700,5.37724,0.9122578,,,,,,,,,,,,,, +197800,5.2348776,0.93376935,,,,,,,,,,,,,, +197900,5.293559,1.0175369,,,,,,,,,,,,,, +198000,5.093357,0.93677104,,,,,,,,,,,,,, +198100,5.0232315,0.9631995,,,,,,,,,,,,,, +198200,4.9585567,0.9069748,,,,,,,,,,,,,, +198300,5.0462523,0.9585158,,,,,,,,,,,,,, +198400,5.1283355,0.92988074,,,,,,,,,,,,,, +198500,5.2117214,0.9638483,,,,,,,,,,,,,, +198600,4.8352957,0.90543664,,,,,,,,,,,,,, +198700,5.416764,0.9534272,,,,,,,,,,,,,, +198800,4.994712,0.8898619,,,,,,,,,,,,,, +198900,5.0256696,0.86437833,,,,,,,,,,,,,, +199000,4.9539485,0.92544425,,,,,,,,,,,,,, +199096,,,0.95272636,0.16754384,0.7528,1.078183,50000.0,0.62920004,1.8614823,10000.0,61924.34237098694,62818.38122630119,61924.34237098694,885.0028548240662,5.684395551681519,0.0 +199100,4.854145,0.85932165,,,,,,,,,,,,,, +199200,5.089237,0.9018992,,,,,,,,,,,,,, +199300,5.237952,0.96406364,,,,,,,,,,,,,, +199400,4.9868126,0.8238881,,,,,,,,,,,,,, +199500,5.2743864,0.95092374,,,,,,,,,,,,,, +199600,5.4552116,0.998711,,,,,,,,,,,,,, +199700,5.572746,0.992167,,,,,,,,,,,,,, +199800,5.0806108,0.8806011,,,,,,,,,,,,,, +199900,5.1981287,0.9039405,,,,,,,,,,,,,, +200000,4.808098,0.8420427,,,,,,,,,,,,,, +200100,5.2274,0.8521678,,,,,,,,,,,,,, +200200,4.9777207,0.93493974,,,,,,,,,,,,,, +200300,5.1950974,0.9569918,,,,,,,,,,,,,, +200400,4.92442,0.9084672,,,,,,,,,,,,,, +200500,5.2714424,0.95386267,,,,,,,,,,,,,, +200600,5.0813384,0.9512837,,,,,,,,,,,,,, +200700,4.992594,0.87364995,,,,,,,,,,,,,, +200800,4.9610195,0.98465264,,,,,,,,,,,,,, +200900,4.981588,0.87373114,,,,,,,,,,,,,, +201000,4.7942686,0.8757074,,,,,,,,,,,,,, +201100,4.943863,0.93708885,,,,,,,,,,,,,, +201200,5.2072015,0.92152107,,,,,,,,,,,,,, +201300,5.327813,1.0010875,,,,,,,,,,,,,, +201400,4.959412,0.89911044,,,,,,,,,,,,,, +201500,5.096189,0.94648874,,,,,,,,,,,,,, +201600,5.152907,0.9395279,,,,,,,,,,,,,, +201700,4.8002234,0.8782333,,,,,,,,,,,,,, +201800,5.4368534,0.95615613,,,,,,,,,,,,,, +201900,5.0818233,0.8984829,,,,,,,,,,,,,, +202000,5.047395,0.89019924,,,,,,,,,,,,,, +202100,5.0435443,0.9282651,,,,,,,,,,,,,, +202200,5.127607,0.9304385,,,,,,,,,,,,,, +202300,4.6825175,0.85436714,,,,,,,,,,,,,, +202400,5.577588,0.9244984,,,,,,,,,,,,,, +202500,5.2690783,0.8966766,,,,,,,,,,,,,, +202600,5.2415895,0.9816131,,,,,,,,,,,,,, +202700,5.0639887,0.8747781,,,,,,,,,,,,,, +202800,4.878237,0.8256624,,,,,,,,,,,,,, +202849,,,0.9549585,0.15947141,0.75386,1.0782359,50000.0,0.62990004,1.8620335,10000.0,63919.883076667786,64852.70789504051,63919.883076667786,922.935383796692,6.243690252304077,0.0 +202900,5.142001,1.0034465,,,,,,,,,,,,,, +203000,5.087765,0.9284964,,,,,,,,,,,,,, +203100,5.1334925,0.9225967,,,,,,,,,,,,,, +203200,4.9663835,0.94086874,,,,,,,,,,,,,, +203300,5.141143,0.87890506,,,,,,,,,,,,,, +203400,5.2273045,0.9915011,,,,,,,,,,,,,, +203500,4.931765,0.8931645,,,,,,,,,,,,,, +203600,4.994811,0.92639333,,,,,,,,,,,,,, +203700,5.1535664,0.88712174,,,,,,,,,,,,,, +203800,5.4974566,1.0241472,,,,,,,,,,,,,, +203900,5.0240636,0.9431415,,,,,,,,,,,,,, +204000,5.485109,1.0129191,,,,,,,,,,,,,, +204100,4.93861,0.9177239,,,,,,,,,,,,,, +204200,5.197873,0.8559029,,,,,,,,,,,,,, +204300,4.938789,0.8780633,,,,,,,,,,,,,, +204400,5.1637387,0.862895,,,,,,,,,,,,,, +204500,5.1901245,0.99145365,,,,,,,,,,,,,, +204600,4.852474,0.86464053,,,,,,,,,,,,,, +204700,5.131825,0.91076183,,,,,,,,,,,,,, +204800,5.0810146,0.9338769,,,,,,,,,,,,,, +204900,5.310151,0.90579027,,,,,,,,,,,,,, +205000,4.974202,0.8556853,,,,,,,,,,,,,, +205100,5.0827785,0.85216093,,,,,,,,,,,,,, +205200,5.02273,0.849154,,,,,,,,,,,,,, +205300,5.306622,0.9881296,,,,,,,,,,,,,, +205400,5.74054,1.0138365,,,,,,,,,,,,,, +205500,5.424652,1.0562061,,,,,,,,,,,,,, +205600,5.0391154,0.8735479,,,,,,,,,,,,,, +205700,4.9341316,0.8910332,,,,,,,,,,,,,, +205800,4.7377996,0.8327043,,,,,,,,,,,,,, +205900,5.2400203,0.98556167,,,,,,,,,,,,,, +206000,5.2157583,0.99706566,,,,,,,,,,,,,, +206100,5.039386,0.7976914,,,,,,,,,,,,,, +206200,4.9913826,0.91560733,,,,,,,,,,,,,, +206300,4.636569,0.8123666,,,,,,,,,,,,,, +206400,5.1354594,0.96781564,,,,,,,,,,,,,, +206500,5.3705683,0.9465696,,,,,,,,,,,,,, +206600,5.5642314,1.044662,,,,,,,,,,,,,, +206700,4.8269916,0.8509695,,,,,,,,,,,,,, +206754,,,0.9528858,0.16299346,0.75408,1.0789212,50000.0,0.629,1.8630679,10000.0,65915.35211348534,66881.94443774223,65915.35211348534,955.9324560165404,6.702232837677002,0.0 +206800,5.094331,0.8758286,,,,,,,,,,,,,, +206900,5.1025925,0.9203047,,,,,,,,,,,,,, +207000,5.2449665,0.9406729,,,,,,,,,,,,,, +207100,5.2751245,0.8883793,,,,,,,,,,,,,, +207200,5.062266,0.893429,,,,,,,,,,,,,, +207300,4.9866753,0.8484735,,,,,,,,,,,,,, +207400,5.0771847,0.90951717,,,,,,,,,,,,,, +207500,5.1799335,0.9289082,,,,,,,,,,,,,, +207600,5.2986317,0.94220084,,,,,,,,,,,,,, +207700,5.2514644,0.921932,,,,,,,,,,,,,, +207800,4.6855927,0.83184195,,,,,,,,,,,,,, +207900,4.9897003,0.95073223,,,,,,,,,,,,,, +208000,5.172411,0.89750963,,,,,,,,,,,,,, +208100,5.3917484,0.9950885,,,,,,,,,,,,,, +208200,5.1988683,0.87977004,,,,,,,,,,,,,, +208300,5.2339754,0.9455337,,,,,,,,,,,,,, +208400,5.496594,0.9707696,,,,,,,,,,,,,, +208500,4.9788485,0.8897008,,,,,,,,,,,,,, +208600,4.978503,0.867851,,,,,,,,,,,,,, +208700,5.0453944,0.8872617,,,,,,,,,,,,,, +208800,5.4021854,0.97327226,,,,,,,,,,,,,, +208900,5.222142,0.9825079,,,,,,,,,,,,,, +209000,5.4594874,0.96187603,,,,,,,,,,,,,, +209100,5.219306,0.9284543,,,,,,,,,,,,,, +209200,5.1604433,0.9863014,,,,,,,,,,,,,, +209300,5.2895327,1.0071801,,,,,,,,,,,,,, +209400,5.095505,0.9366816,,,,,,,,,,,,,, +209500,5.1500916,0.968586,,,,,,,,,,,,,, +209600,5.2893643,0.9760083,,,,,,,,,,,,,, +209700,5.325281,0.96253854,,,,,,,,,,,,,, +209800,5.133908,0.9025451,,,,,,,,,,,,,, +209900,5.200687,0.8621603,,,,,,,,,,,,,, +210000,5.0692644,0.9429909,,,,,,,,,,,,,, +210100,5.5488267,0.91372585,,,,,,,,,,,,,, +210200,5.013029,0.87389046,,,,,,,,,,,,,, +210300,4.992096,0.92762417,,,,,,,,,,,,,, +210400,5.02389,0.864603,,,,,,,,,,,,,, +210500,4.9644294,0.8720778,,,,,,,,,,,,,, +210600,5.156888,0.97168744,,,,,,,,,,,,,, +210700,5.0015163,0.91964847,,,,,,,,,,,,,, +210800,5.292617,0.954654,,,,,,,,,,,,,, +210830,,,0.9546795,0.16126078,0.75395995,1.0778822,50000.0,0.6302,1.8621116,10000.0,67910.6963338852,68918.14360046387,67910.6963338852,996.0133640766144,7.245025157928467,0.0 +210900,5.153116,0.8613507,,,,,,,,,,,,,, +211000,5.315665,0.98848414,,,,,,,,,,,,,, +211100,5.3485475,0.9610522,,,,,,,,,,,,,, +211200,5.133538,0.9285419,,,,,,,,,,,,,, +211300,4.9898496,0.87153405,,,,,,,,,,,,,, +211400,5.457027,0.95457816,,,,,,,,,,,,,, +211500,5.1028814,0.90436715,,,,,,,,,,,,,, +211600,4.988109,0.8917821,,,,,,,,,,,,,, +211700,5.135812,0.92283,,,,,,,,,,,,,, +211800,5.723661,0.9441191,,,,,,,,,,,,,, +211900,5.128078,0.8854506,,,,,,,,,,,,,, +212000,4.9938536,0.9233067,,,,,,,,,,,,,, +212100,5.114876,0.8868686,,,,,,,,,,,,,, +212200,4.9842544,0.90180194,,,,,,,,,,,,,, +212300,5.001011,0.87477577,,,,,,,,,,,,,, +212400,4.961604,0.80205905,,,,,,,,,,,,,, +212500,5.018125,0.95575035,,,,,,,,,,,,,, +212600,4.8936033,0.8804634,,,,,,,,,,,,,, +212700,5.377979,0.9779346,,,,,,,,,,,,,, +212800,5.202916,0.9589001,,,,,,,,,,,,,, +212900,5.0114303,0.90928084,,,,,,,,,,,,,, +213000,5.072605,0.90816134,,,,,,,,,,,,,, +213100,4.945306,0.8509699,,,,,,,,,,,,,, +213200,5.0665,0.89907587,,,,,,,,,,,,,, +213300,5.114423,0.88169324,,,,,,,,,,,,,, +213400,5.0106196,0.87183386,,,,,,,,,,,,,, +213500,5.0433083,0.84040165,,,,,,,,,,,,,, +213600,5.0990915,0.8774149,,,,,,,,,,,,,, +213700,4.7428236,0.8831363,,,,,,,,,,,,,, +213800,5.0153117,0.9237356,,,,,,,,,,,,,, +213900,5.2754474,0.9349259,,,,,,,,,,,,,, +214000,5.3656344,1.0174567,,,,,,,,,,,,,, +214100,4.8150363,0.8611651,,,,,,,,,,,,,, +214200,5.113802,0.89779884,,,,,,,,,,,,,, +214300,5.3075795,0.93612075,,,,,,,,,,,,,, +214400,5.079487,0.8900384,,,,,,,,,,,,,, +214500,5.2254066,0.96503127,,,,,,,,,,,,,, +214600,5.327312,1.002486,,,,,,,,,,,,,, +214700,5.308074,0.9312092,,,,,,,,,,,,,, +214800,4.9101305,0.8940265,,,,,,,,,,,,,, +214900,5.282629,0.87170416,,,,,,,,,,,,,, +215000,5.073806,0.8984159,,,,,,,,,,,,,, +215100,5.0384865,0.880203,,,,,,,,,,,,,, +215200,5.7225347,1.0435605,,,,,,,,,,,,,, +215277,,,0.9546795,0.15797712,0.75408,1.0774853,50000.0,0.63310003,1.8600335,10000.0,69906.23444628716,70955.25825119019,69906.23444628716,1036.815337896347,7.803174257278442,0.0 +215300,5.209727,0.92458105,,,,,,,,,,,,,, +215400,5.4042883,0.953952,,,,,,,,,,,,,, +215500,5.296565,0.8851534,,,,,,,,,,,,,, +215600,5.240959,0.9085546,,,,,,,,,,,,,, +215700,5.009772,0.88030136,,,,,,,,,,,,,, +215800,4.938934,0.87099105,,,,,,,,,,,,,, +215900,4.8852687,0.89155114,,,,,,,,,,,,,, +216000,4.846564,0.7862259,,,,,,,,,,,,,, +216100,5.349639,0.9574717,,,,,,,,,,,,,, +216200,5.2707314,0.95424485,,,,,,,,,,,,,, +216300,5.2665176,0.9599327,,,,,,,,,,,,,, +216400,5.193632,0.86845726,,,,,,,,,,,,,, +216500,4.908529,0.8023777,,,,,,,,,,,,,, +216600,5.337294,0.91448,,,,,,,,,,,,,, +216700,5.105943,0.8863859,,,,,,,,,,,,,, +216800,5.0270524,0.9209032,,,,,,,,,,,,,, +216900,5.284718,0.9968642,,,,,,,,,,,,,, +217000,5.2203918,0.9313672,,,,,,,,,,,,,, +217100,4.9800344,0.8858732,,,,,,,,,,,,,, +217200,5.0396547,0.9210735,,,,,,,,,,,,,, +217300,5.3347816,0.97480357,,,,,,,,,,,,,, +217400,5.1645746,0.9219173,,,,,,,,,,,,,, +217500,4.841171,0.8522967,,,,,,,,,,,,,, +217600,5.1531954,0.96086353,,,,,,,,,,,,,, +217700,5.0294237,0.884645,,,,,,,,,,,,,, +217800,4.8027625,0.7918198,,,,,,,,,,,,,, +217900,5.2266183,0.86897826,,,,,,,,,,,,,, +218000,5.3321977,0.98133004,,,,,,,,,,,,,, +218100,5.283939,0.96068203,,,,,,,,,,,,,, +218200,5.034923,0.8902592,,,,,,,,,,,,,, +218300,4.824457,0.86642116,,,,,,,,,,,,,, +218400,4.8722277,0.91594744,,,,,,,,,,,,,, +218500,5.4252605,0.9741702,,,,,,,,,,,,,, +218600,5.2142477,0.92636025,,,,,,,,,,,,,, +218700,5.211658,0.94290596,,,,,,,,,,,,,, +218800,5.136646,0.92348903,,,,,,,,,,,,,, +218900,5.3196874,0.9385422,,,,,,,,,,,,,, +219000,5.1108956,0.89516,,,,,,,,,,,,,, +219100,5.0208263,0.92416054,,,,,,,,,,,,,, +219200,5.103062,0.9026141,,,,,,,,,,,,,, +219300,5.1832533,0.888795,,,,,,,,,,,,,, +219400,5.2133775,0.8955112,,,,,,,,,,,,,, +219500,4.84986,0.8460989,,,,,,,,,,,,,, +219600,5.014862,0.9206774,,,,,,,,,,,,,, +219700,4.9568667,0.8829387,,,,,,,,,,,,,, +219800,5.3214383,0.9632787,,,,,,,,,,,,,, +219900,5.4161167,0.92981344,,,,,,,,,,,,,, +220000,4.997305,0.89931095,,,,,,,,,,,,,, +220100,5.5423603,0.9184139,,,,,,,,,,,,,, +220200,5.02984,0.902712,,,,,,,,,,,,,, +220300,5.0603404,0.9114016,,,,,,,,,,,,,, +220397,,,0.95649314,0.15317032,0.75404,1.080798,50000.0,0.6341,1.863171,10000.0,71901.50822019577,72989.8373169899,71901.50822019577,1075.351613998413,8.280278444290161,0.0 +220400,5.2279987,1.0017228,,,,,,,,,,,,,, +220500,5.1228337,0.88086134,,,,,,,,,,,,,, +220600,5.2269797,0.9029439,,,,,,,,,,,,,, +220700,5.2159896,0.9803972,,,,,,,,,,,,,, +220800,4.980408,0.9081791,,,,,,,,,,,,,, +220900,5.010544,0.91249454,,,,,,,,,,,,,, +221000,4.892558,0.88128614,,,,,,,,,,,,,, +221100,5.1677184,0.9017781,,,,,,,,,,,,,, +221200,5.595472,1.0283965,,,,,,,,,,,,,, +221300,5.1355963,0.9290332,,,,,,,,,,,,,, +221400,4.8653083,0.8863214,,,,,,,,,,,,,, +221500,4.9659796,0.8650927,,,,,,,,,,,,,, +221600,5.050287,0.8762206,,,,,,,,,,,,,, +221700,5.270135,0.8728773,,,,,,,,,,,,,, +221800,4.951304,0.89439356,,,,,,,,,,,,,, +221900,5.104475,0.8794708,,,,,,,,,,,,,, +222000,5.2753243,0.97024417,,,,,,,,,,,,,, +222100,5.1184635,0.89265454,,,,,,,,,,,,,, +222200,4.997801,0.92497826,,,,,,,,,,,,,, +222300,4.9796863,0.84022677,,,,,,,,,,,,,, +222400,5.28201,0.9425012,,,,,,,,,,,,,, +222500,4.853951,0.8962872,,,,,,,,,,,,,, +222600,5.3078256,0.92769766,,,,,,,,,,,,,, +222700,5.405352,0.9499087,,,,,,,,,,,,,, +222800,4.743131,0.87527996,,,,,,,,,,,,,, +222900,5.093333,0.92454326,,,,,,,,,,,,,, +223000,5.0761385,0.8587646,,,,,,,,,,,,,, +223100,5.060522,0.94421756,,,,,,,,,,,,,, +223200,5.331088,0.93048036,,,,,,,,,,,,,, +223300,4.6283484,0.83763206,,,,,,,,,,,,,, +223400,5.115885,0.96212316,,,,,,,,,,,,,, +223500,4.8400674,0.8676809,,,,,,,,,,,,,, +223600,5.2006187,0.92280936,,,,,,,,,,,,,, +223700,5.3853393,0.9958111,,,,,,,,,,,,,, +223800,5.0345364,0.91473186,,,,,,,,,,,,,, +223900,5.3153796,0.9217161,,,,,,,,,,,,,, +224000,4.9492016,0.8892548,,,,,,,,,,,,,, +224100,5.26176,0.89906514,,,,,,,,,,,,,, +224200,5.170938,0.90937364,,,,,,,,,,,,,, +224300,5.0455155,0.9224608,,,,,,,,,,,,,, +224400,4.9506135,0.8751482,,,,,,,,,,,,,, +224500,5.2392545,1.0115783,,,,,,,,,,,,,, +224600,4.985041,0.89337766,,,,,,,,,,,,,, +224700,5.155628,0.93984663,,,,,,,,,,,,,, +224800,5.004521,0.9245262,,,,,,,,,,,,,, +224900,4.967423,0.8815662,,,,,,,,,,,,,, +225000,5.35936,0.8634929,,,,,,,,,,,,,, +225100,5.344352,0.9146207,,,,,,,,,,,,,, +225200,4.9444323,0.8850175,,,,,,,,,,,,,, +225300,4.997132,0.8535713,,,,,,,,,,,,,, +225400,5.1853743,0.9141326,,,,,,,,,,,,,, +225500,5.4515452,0.9966766,,,,,,,,,,,,,, +225600,5.127328,0.9304979,,,,,,,,,,,,,, +225700,5.021611,0.79902285,,,,,,,,,,,,,, +225800,5.1513257,0.93318045,,,,,,,,,,,,,, +225900,4.9463496,0.86017716,,,,,,,,,,,,,, +226000,5.1860847,0.9174198,,,,,,,,,,,,,, +226100,5.300269,0.94345444,,,,,,,,,,,,,, +226200,5.3167067,0.89156544,,,,,,,,,,,,,, +226300,5.076334,0.8932842,,,,,,,,,,,,,, +226400,5.329573,0.98484296,,,,,,,,,,,,,, +226406,,,0.95523757,0.15633641,0.75372,1.0817096,50000.0,0.63430005,1.8659685,10000.0,73897.11191749573,75022.37075853348,73897.11191749573,1111.535054922104,8.781720638275146,0.0 +226500,4.998298,0.81913865,,,,,,,,,,,,,, +226600,5.2754154,0.98889124,,,,,,,,,,,,,, +226700,5.1140137,0.8066398,,,,,,,,,,,,,, +226800,5.5442715,0.89443034,,,,,,,,,,,,,, +226900,5.2435827,0.8191415,,,,,,,,,,,,,, +227000,5.2343655,0.8590567,,,,,,,,,,,,,, +227100,5.3084784,0.92972386,,,,,,,,,,,,,, +227200,5.2614155,0.90822965,,,,,,,,,,,,,, +227300,4.9326334,0.92096764,,,,,,,,,,,,,, +227400,5.036497,0.8722205,,,,,,,,,,,,,, +227500,5.2687807,0.8036066,,,,,,,,,,,,,, +227600,5.129593,0.9697646,,,,,,,,,,,,,, +227700,4.784746,0.82962817,,,,,,,,,,,,,, +227800,5.042297,0.8889498,,,,,,,,,,,,,, +227900,5.198848,0.8784989,,,,,,,,,,,,,, +228000,5.069074,0.8548619,,,,,,,,,,,,,, +228100,5.3416066,0.99511516,,,,,,,,,,,,,, +228200,5.152013,0.8384189,,,,,,,,,,,,,, +228300,5.269827,1.0044866,,,,,,,,,,,,,, +228400,5.242281,0.9437058,,,,,,,,,,,,,, +228500,5.353851,0.9914513,,,,,,,,,,,,,, +228600,5.1957216,0.95207655,,,,,,,,,,,,,, +228700,5.550254,0.94268113,,,,,,,,,,,,,, +228800,5.2874527,0.94571614,,,,,,,,,,,,,, +228900,5.613104,1.0503197,,,,,,,,,,,,,, +229000,5.0465045,0.9111926,,,,,,,,,,,,,, +229100,5.1274667,0.9772841,,,,,,,,,,,,,, +229200,5.1616993,0.94459105,,,,,,,,,,,,,, +229300,5.541454,0.9734502,,,,,,,,,,,,,, +229400,4.9639497,0.85980624,,,,,,,,,,,,,, +229500,4.6327915,0.7670916,,,,,,,,,,,,,, +229600,5.357014,0.9371649,,,,,,,,,,,,,, +229700,5.409813,0.9799932,,,,,,,,,,,,,, +229800,4.934324,0.9098384,,,,,,,,,,,,,, +229900,5.275178,0.8990069,,,,,,,,,,,,,, +230000,5.2524166,0.9163827,,,,,,,,,,,,,, +230100,5.5426807,0.9463514,,,,,,,,,,,,,, +230200,4.8132014,0.8499898,,,,,,,,,,,,,, +230300,5.2537394,0.9262945,,,,,,,,,,,,,, +230400,5.5182095,0.94054055,,,,,,,,,,,,,, +230500,5.3716245,0.9204641,,,,,,,,,,,,,, +230600,5.364909,0.9387766,,,,,,,,,,,,,, +230700,4.926332,0.9076877,,,,,,,,,,,,,, +230800,5.2558475,0.9364003,,,,,,,,,,,,,, +230900,4.909439,0.81914365,,,,,,,,,,,,,, +231000,5.1186247,0.8749615,,,,,,,,,,,,,, +231100,5.4946256,1.0761766,,,,,,,,,,,,,, +231200,4.983416,0.877363,,,,,,,,,,,,,, +231300,4.956263,0.87382066,,,,,,,,,,,,,, +231400,5.057019,0.9189538,,,,,,,,,,,,,, +231500,5.453634,0.9052627,,,,,,,,,,,,,, +231600,5.3070993,0.9718408,,,,,,,,,,,,,, +231700,5.044973,0.89939964,,,,,,,,,,,,,, +231800,5.3860703,0.85518587,,,,,,,,,,,,,, +231900,5.1513143,0.89475477,,,,,,,,,,,,,, +232000,5.191963,0.8752988,,,,,,,,,,,,,, +232100,5.2257843,0.8918358,,,,,,,,,,,,,, +232200,4.804822,0.875126,,,,,,,,,,,,,, +232206,,,,,,,,,,,75892.38308477402,,,,,0.0 diff --git a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/meta_data_0.json b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/meta_data_0.json index 5d6616b74..d027b7903 100644 --- a/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/meta_data_0.json +++ b/logs/self_tuning/schedule_free_adamw_jax_v2/study_2/imagenet_resnet_jax/trial_1/meta_data_0.json @@ -15,47 +15,47 @@ "workload.use_gelu": false, "workload.use_silu": false, "workload.validation_target_value": 0.77431, - "cpu.util.avg_percent_since_last": 2.5, - "cpu.freq.current": 2200.207999999999, + "cpu.util.avg_percent_since_last": 4.8, + "cpu.freq.current": 2200.216, "mem.total": 359053524992, - "mem.available": 348623962112, - "mem.used": 7263256576, + "mem.available": 348540833792, + "mem.used": 7263576064, "mem.percent_used": 2.9, - "mem.read_bytes_since_boot": 9218609028608, - "mem.write_bytes_since_boot": 41883461120, - "net.bytes_sent_since_boot": 15207, - "net.bytes_recv_since_boot": 707261, + "mem.read_bytes_since_boot": 1130566144, + "mem.write_bytes_since_boot": 26298594816, + "net.bytes_sent_since_boot": 14691, + "net.bytes_recv_since_boot": 740619, "gpu.count": 4, - "gpu.0.compute.util": 0.0, + "gpu.0.compute.util": 0.01, "gpu.0.mem.util": 0.7490478515625, "gpu.0.mem.total": 40960.0, "gpu.0.mem.used": 30681.0, "gpu.0.mem.free": 9646.0, - "gpu.0.temp.current": 34.0, + "gpu.0.temp.current": 39.0, "gpu.1.compute.util": 0.0, "gpu.1.mem.util": 0.7488525390625, "gpu.1.mem.total": 40960.0, "gpu.1.mem.used": 30673.0, "gpu.1.mem.free": 9654.0, - "gpu.1.temp.current": 33.0, + "gpu.1.temp.current": 36.0, "gpu.2.compute.util": 0.0, "gpu.2.mem.util": 0.7488525390625, "gpu.2.mem.total": 40960.0, "gpu.2.mem.used": 30673.0, "gpu.2.mem.free": 9654.0, - "gpu.2.temp.current": 32.0, + "gpu.2.temp.current": 37.0, "gpu.3.compute.util": 0.0, "gpu.3.mem.util": 0.7488525390625, "gpu.3.mem.total": 40960.0, "gpu.3.mem.used": 30673.0, "gpu.3.mem.free": 9654.0, - "gpu.3.temp.current": 35.0, - "gpu.avg.compute.util": 0.0, + "gpu.3.temp.current": 38.0, + "gpu.avg.compute.util": 0.0025, "gpu.avg.mem.util": 0.7489013671875, "gpu.avg.mem.total": 40960.0, "gpu.avg.mem.used": 30675.0, "gpu.avg.mem.free": 9652.0, - "gpu.avg.temp.current": 33.5, + "gpu.avg.temp.current": 37.5, "os_platform": "Linux-6.1.0-44-cloud-amd64-x86_64-with-glibc2.31", "python_version": "3.11.10", "python_compiler": "GCC 9.4.0",