CodeQL

134 workflow runs

api: bro.tensor.openSafetensors and saveSafetensors ask the host's on…
completed successfully:
CodeQL#134:Commitfb8f819pushedbywlejon
main
Oct 26, 20262h 30m 18s
CodeQL
completed successfully:
CodeQL#133:Scheduled
main
Oct 26, 202645m 7s
verify: lift shipped .spv backward shaders directly and expand multi-…
completed successfully:
CodeQL#132:Commit0d092fcpushedbywlejon
main
Oct 24, 202635m 8s
Add backward-pass formal proofs, SPIR-V/PTX lifter audit, and gradien…
completed successfully:
CodeQL#131:Commitd903fb3pushedbywlejon
main
Oct 23, 202643m 16s
feat(jit): add verified fused backward TraceDAG::backward() autodiff …
cancelled:
CodeQL#130:Commit3631acepushedbywlejon
main
Oct 23, 202626m 35s
cuda: build again after the kernel split. elementwise.cu's activation…
completed successfully:
CodeQL#129:Commit68bd428pushedbywlejon
main
Oct 23, 202640m 42s
fix(jit): enable BROTENSOR_HAS_BRASS_JIT in kernel_disk_cache.h when …
completed successfully:
CodeQL#128:Commitf52c9f7pushedbywlejon
main
Oct 23, 202631m 3s
feat(jit): persistent on-disk superoptimized ML kernel cache & CPU SI…
completed successfully:
CodeQL#127:Commit875a21fpushedbywlejon
main
Oct 23, 202642m 12s
Decompose oversized CUDA kernels under 1,000 lines and specialize int…
cancelled:
CodeQL#126:Commit2c6071cpushedbywlejon
main
Oct 23, 20266m 37s
Enforce C++ host-boundary rooting and std::isfinite domain gates (#84)
completed successfully:
CodeQL#125:Commitc18b565pushedbywlejon
main
Oct 23, 202638m 20s
Verify CPU reference vs brass JIT and fused kernel equivalence (#80)
completed successfully:
CodeQL#124:Commitaad414bpushedbywlejon
main
Oct 23, 202629m 31s
Verify & harden untrusted byte parsers against 64-bit overflow and tr…
completed successfully:
CodeQL#123:Commit33e5681pushedbywlejon
main
Oct 23, 202624m 29s
Include omitted operands in op dispatchers and harden safetensors hea…
completed successfully:
CodeQL#122:Commitda554e3pushedbywlejon
main
Oct 22, 202637m 27s
Brand GpuTensor and SafetensorsFile handles against cross-type casts …
completed successfully:
CodeQL#121:Commitcbc906epushedbywlejon
main
Oct 22, 202638m 17s
CodeQL
completed successfully:
CodeQL#120:Scheduled
main
Oct 19, 202638m 0s
api: under Visual Studio the native manifest rule is a target of its …
completed successfully:
CodeQL#119:Commit0bbf4abpushedbywlejon
main
Oct 17, 202628m 19s
metal: implement dequantize_rows for INT8 and FP8_E4M3 weights
completed successfully:
CodeQL#118:Commite19ab1bpushedbywlejon
main
Oct 16, 202657m 27s
Allow degenerate N/H/W == 0 in pixel_shuffle_upsample_2x_forward wrapper
completed successfully:
CodeQL#117:Commit160ae1bpushedbywlejon
main
Oct 16, 202637m 28s
cpu: shutdown stops the thread pool only if one was made
completed successfully:
CodeQL#116:Commit0b5c6fdpushedbywlejon
main
Oct 16, 202624m 12s
cuda: FP32 implicit-GEMM conv2d, and conv_transpose2d as GEMM + col2im
completed successfully:
CodeQL#115:Commit2c272dbpushedbywlejon
main
Oct 14, 202621m 55s
dispatch: the spatial guards check the element count, not the row spl…
failed:
CodeQL#114:Commit6012b83pushedbywlejon
main
Oct 14, 20261m 17s
Expand symbolic Trace JIT with Where, Clamp, Cast, and column broadcasts
completed successfully:
CodeQL#113:Commit448edcepushedbywlejon
main
Oct 13, 202628m 11s
runtime: idle(device), a non-blocking check that the queued work has …
completed successfully:
CodeQL#112:Commit69427a7pushedbywlejon
main
Oct 13, 202624m 21s
quant: dequantize_rows(Wq, scales, out, W), quantize_rows' inverse (W…
completed successfully:
CodeQL#111:Commit7c835d5pushedbywlejon
main
Oct 12, 202636m 48s
metal: the weight-only INT8 / FP8 linear reads eight weights a lane a…
completed successfully:
CodeQL#110:Commitd4af8f6pushedbywlejon
main
Oct 12, 202634m 43s