Pull requests: NVIDIA/TensorRT-LLM
Pull requests list
[None][feat] Multi-K (512/1024/2048) and Multi-dtype (fp32/bf16/fp16) GVR Top-K
#13948
opened May 9, 2026 by
longcheng-nv
Collaborator
Loading…
4 tasks done
[None][doc] Gemma 4 support & eval task updates
#13947
opened May 9, 2026 by
Hudayday
Collaborator
Loading…
[https://nvbugs/6100102][fix] Fix cutlass grouped gemm launcher EpilogueScalars construction
#13945
opened May 9, 2026 by
yifeizhang-c
Collaborator
Loading…
1 task done
[Draft][For pre-codereview only]User/shreyasm/attn2d ulysses
#13944
opened May 9, 2026 by
juney-nvidia
Collaborator
•
Draft
1 task
[https://nvbugs/6160629][fix] Add both test filenames to
EXCLUDE_TEST_FILES in `examples/auto_deploy/llmc/cr
#13943
opened May 9, 2026 by
tensorrt-cicd
Collaborator
Loading…
2 tasks done
[None][feat] Add --use-3rdparty-cache to accelerate cmake configuration of clean build
#13942
opened May 9, 2026 by
yuantailing
Member
Loading…
1 task done
[None][infra] Enable draco-oci-iad A100 cluster for CI tests
#13940
opened May 9, 2026 by
ZhanruiSunCh
Collaborator
Loading…
1 task
[None][feat] Keep DSv4 o_a_proj as FP8, and port vLLM's fused_inv_rope_fp8_quant
deepseek-v4
#13938
opened May 9, 2026 by
lishicheng1996-nv
Collaborator
Loading…
5 tasks done
[None][refactor] Decouple cached prefix from KVSlice token_range
#13937
opened May 9, 2026 by
Shixiaowei02
Collaborator
•
Draft
1 task done
[None][test] Add dsv4 dis-agg module-level unit tests
deepseek-v4
#13936
opened May 9, 2026 by
Shixiaowei02
Collaborator
•
Draft
1 task done
[None][test] Add checkpoint_format / load_format keys to test_features_contract
#13933
opened May 9, 2026 by
chienchunhung
Collaborator
Loading…
2 tasks done
[TRTLLM-35237][feat] Add cute dsl FP4 paged MQA logits decode kernel
#13929
opened May 9, 2026 by
limin2021
Collaborator
Loading…
1 task
[None][chore] Add long seq test for DSV4.
deepseek-v4
#13928
opened May 9, 2026 by
Tracin
Collaborator
Loading…
1 task done
[TRTLLM-12440][feat] Add GMS-only weight sharing support
#13926
opened May 9, 2026 by
chienchunhung
Collaborator
•
Draft
1 task done
[None][fix] Fix and unwaive AutoDeploy accuracy tests
#13925
opened May 8, 2026 by
bmarimuthu-nv
Collaborator
Loading…
1 task done
[None][fix] Fix accracy regression in DeepSeek models
#13924
opened May 8, 2026 by
taylor-yb-lee
Collaborator
Loading…
1 task done
[https://nvbugs/6159129][fix] Added an FP8_BLOCK_SCALES + extra_acc_spec=tp_attn reference entry (accuracy 92.
#13923
opened May 8, 2026 by
tensorrt-cicd
Collaborator
Loading…
2 tasks done
[https://nvbugs/6159132][fix] Differentiate the two paths via extra_acc_spec="tp_attn" when attention_dp=False
#13922
opened May 8, 2026 by
tensorrt-cicd
Collaborator
Loading…
2 tasks done
[None][infra] Check license with both isPermissive and isProprietary flags
#13921
opened May 8, 2026 by
yuanjingx87
Collaborator
Loading…
1 task
[#13909][fix] Reuse hidden_states buffer across CUDA graph captures in Eagle3
Community want to contribute
PRs initiated from Community
#13920
opened May 8, 2026 by
ml-inference
Loading…
[TRTLLM-12339][feat] Support T5 encoder-decoder models in the PyTorch backend
#13919
opened May 8, 2026 by
cascade812
Collaborator
Loading…
1 task done
[None][fix] Make SleepConfig picklable by replacing closure lambda in defaultdict
Community want to contribute
PRs initiated from Community
#13918
opened May 8, 2026 by
hhzhang16
Loading…
1 task
[None][doc] Add guide for integrating custom kernels in PyTorch backend
#13917
opened May 8, 2026 by
chang-l
Collaborator
Loading…
5 tasks done
[https://nvbugs/6157892] [fix] MistralCommonImageProcessor text-only path
#13916
opened May 8, 2026 by
evezhier
Collaborator
Loading…
1 task
ProTip!
Find all pull requests that aren't related to any open issues with -linked:issue.