Pull requests: NVIDIA/TensorRT-LLM
Pull requests list
[None][test] Waive 1 failed cases for main in QA CI
#15930
opened Jul 3, 2026 by
trtllm-agent
Collaborator
•
Draft
[None][fix] Enable cuda_scaled_mm fast path for FP8 linear on SM121
#15928
opened Jul 3, 2026 by
souvikDevloper
Loading…
4 tasks done
[https://nvbugs/6337231][fix] Unskip and fix iter-stats unit tests' fake-self predicate
#15927
opened Jul 3, 2026 by
YihuiLu512
Collaborator
Loading…
1 task done
[https://nvbugs/6410928][fix] Keep
TestGPTE2E::test_check_gpt_e2e as a no-op stub (empty method body) so…
#15926
opened Jul 3, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[https://nvbugs/6341070][fix] Fix scaffolding MajorityVoteController output handling
#15925
opened Jul 3, 2026 by
KleinBlueC
Loading…
1 task
[https://nvbugs/6411931][fix] Append
,fo to the ignore-words-list in pyproject.toml…
#15924
opened Jul 3, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[None][fix] Enable MiniMax M3 piecewise CUDA graphs
#15923
opened Jul 3, 2026 by
liji-nv
Collaborator
Loading…
1 task done
[https://nvbugs/6412108][fix] Restore original order —
all_reduce the routed partial first, then add the…
#15922
opened Jul 3, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[https://nvbugs/6412133][fix] Only populate
self.all_weights[self.device_id] at construction; lazily…
#15921
opened Jul 3, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[None][perf] Move greedy stop checks to host
#15920
opened Jul 3, 2026 by
mingyangHao
Collaborator
Loading…
1 task
[TRTLLM-14022][feat] Remove legacy TensorRT Python backend
#15918
opened Jul 3, 2026 by
Wanli-Jiang
Collaborator
•
Draft
1 task done
[https://nvbugs/6405665][test] Disable block reuse for KV cache comparison
#15917
opened Jul 3, 2026 by
jiaganc
Collaborator
Loading…
1 task done
[https://nvbugs/6410093][fix] Revert the per-batch cross-attn slicing (drop real_text_lens plumbing through…
#15915
opened Jul 3, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[https://nvbugs/6337051][fix] Refresh inputs on graph capture to avoid unintended modification
#15913
opened Jul 3, 2026 by
pengbowang-nv
Collaborator
•
Draft
1 task done
[None][test] Dump sibling worker logs from disagg BENCHMARK pytest
#15912
opened Jul 3, 2026 by
chenfeiz0326
Collaborator
Loading…
4 tasks done
[https://nvbugs/6410336][fix] Bump WAN_LPIPS_THRESHOLD from 0.05 to 0.10 (with explanatory comment) and…
#15911
opened Jul 3, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[None][perf] serve: opt-in msgspec msgpack transport for disagg orchestrator->worker request body
#15910
opened Jul 3, 2026 by
Tabrizian
Member
Loading…
[TRTLLM-14024][feat] Prune CuTe DSL NVFP4 GEMM autotuner tactics with nvMatmulHeuristics
#15909
opened Jul 3, 2026 by
peaceh-nv
Collaborator
Loading…
1 task
[None][test] Add opt-in background prefetch of test MPI sessions and model page cache
#15908
opened Jul 3, 2026 by
sunnyqgg
Collaborator
Loading…
3 tasks done
[TRTLLM-13784][chore] Remove legacy TensorRT-engine Triton backend
#15907
opened Jul 3, 2026 by
Wanli-Jiang
Collaborator
Loading…
1 task done
[None][Perf] Update CuTeDSL MegaMoE kernels
#15906
opened Jul 3, 2026 by
Barry-Delaney
Collaborator
•
Draft
[None][feat] Disagg coordinator + orchestrator fleet
#15905
opened Jul 3, 2026 by
reasonsolo
Collaborator
•
Draft
1 task done
ProTip!
Filter pull requests by the default branch with base:main.