Pull requests: NVIDIA/TensorRT-LLM
Pull requests list
[https://nvbugs/6422339][test] Unwaive 3 disagg-perf-sanity GB200 tests after #15961 revert
#16220
opened Jul 10, 2026 by
chenfeiz0326
Collaborator
Loading…
2 tasks done
[None][fix] Derive DSA indexer tokens-per-block from the active KV cache manager
#16219
opened Jul 10, 2026 by
qiaoxj07
Collaborator
Loading…
3 tasks done
[None][chore] KVCacheManagerV2: backend-neutral Python and test preparation
#16218
opened Jul 10, 2026 by
lowsfer
Member
Loading…
3 tasks done
[None][fix] Bump DeepGEMM to nv_dev HEAD for SM120 attention support
#16217
opened Jul 10, 2026 by
lucifer1004
Collaborator
Loading…
[None][fix] Propagate NCCL_INCLUDE_DIR to common header dirs
#16216
opened Jul 10, 2026 by
lucifer1004
Collaborator
Loading…
[None][fix] Generate CUTLASS kernel instantiations without installing cutlass_library
#16215
opened Jul 10, 2026 by
lucifer1004
Collaborator
Loading…
[TRTLLM-14022][feat] BREAKING: DON'T REVIEW Remove legacy TensorRT Python backend
api-breaking
Accepted LLM API contract change that is backwards-incompatible
#16213
opened Jul 10, 2026 by
Wanli-Jiang
Collaborator
•
Draft
1 task done
[https://nvbugs/6127669][fix] Fix layer-wise benchmarks performance alignment test
#16212
opened Jul 10, 2026 by
kaiyux
Member
Loading…
1 task done
[None][feat] Add fused T5 encoder self-attention path in bertAttentionPlugin
#16211
opened Jul 10, 2026 by
zhanghuanzj
Loading…
[None][test] Fix the multinode test case on DGX-Spark(perf skipped and func hang)
#16209
opened Jul 10, 2026 by
JennyLiu-nv
Collaborator
Loading…
1 task done
[https://nvbugs/6428113][fix] Replace the assertion +
[0] indexing with max(all_rank_num_tokens)…
#16207
opened Jul 9, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[None][chore] Remove pre-installed opencv
#16206
opened Jul 9, 2026 by
tburt-nv
Collaborator
Loading…
1 task done
[https://nvbugs/6432948][fix] Add per-engine flag
_trtllm_gen_jit_warmup_done; skip…
#16205
opened Jul 9, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[TRTLLMINF-95][infra] Add LLVM BOLT profile-generation engine and pipeline
#16204
opened Jul 9, 2026 by
mlefeb01
Collaborator
Loading…
1 task
[https://nvbugs/6433376][fix] Update the Dense test to mirror the MoE sibling — assert
bfloat16 under…
#16203
opened Jul 9, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[https://nvbugs/6435100][fix] Replace the hard assertion with a branch that accepts both shapes: length-1…
#16202
opened Jul 9, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[https://nvbugs/6430702][perf] Restore NVLink one-sided A2A fast path
#16200
opened Jul 9, 2026 by
chienchunhung
Collaborator
Loading…
[https://nvbugs/6434572][fix] Add a new
pytorch_model_config.py pattern block covering the six 4-GPU…
#16199
opened Jul 9, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[None][fix] Fix indexer-k-cache transfer for disagg
#16197
opened Jul 9, 2026 by
Tabrizian
Member
Loading…
1 task
[https://nvbugs/6402018][fix] Defer DeepGEMM PDL init to MoE construction to preserve KV cache
#16195
opened Jul 9, 2026 by
trtllm-agent
Collaborator
Loading…
2 tasks done
[None][infra] Upgrade NIXL to v1.3.1
#16194
opened Jul 9, 2026 by
zackyoray
Collaborator
Loading…
1 task done
nano v3 disable piecewise
#16193
opened Jul 9, 2026 by
tcherckez-nvidia
Collaborator
Loading…
1 task done
ProTip!
Exclude everything labeled
bug with -label:bug.