Pull requests: NVIDIA/TensorRT-LLM
Pull requests list
[None][fix] Auto-select reasoning_at_start=True for Qwen3.5-style hybrid templates
#18098
opened Aug 22, 2026 by
yenhao-huang
Loading…
4 of 9 tasks
[None][fix] Map qwen3_5/qwen3_5_moe tool-parser auto-selection to qwen3_coder
#18097
opened Aug 22, 2026 by
yenhao-huang
Loading…
4 of 9 tasks
[None][fix] Normalize ruff-legacy baseline paths to POSIX on Windows
#18096
opened Aug 22, 2026 by
yenhao-huang
Loading…
9 tasks done
[https://nvbugs/6625710][fix] Re-attach radix-tree blocks detached under a live request
#18095
opened Aug 22, 2026 by
lowsfer
Member
Loading…
1 task done
[None][fix] CuTe DSL GVR top-K decode: repair the non-converged threshold search
#18094
opened Aug 22, 2026 by
longcheng-nv
Collaborator
Loading…
[https://nvbugs/6621358][fix] Enable draft KV reuse in cache manager V2
#18093
opened Aug 22, 2026 by
yizhang-nv
Member
•
Draft
1 task done
[None][fix] Synchronize KV cache V2 host fallback across ranks
#18092
opened Aug 22, 2026 by
yizhang-nv
Member
Loading…
1 task done
[None][feat] Add NVFP4 as a cold-page KV Cache Compression Method
api-compatible
Accepted LLM API contract change that is backwards-compatible
[None][fix] preserve streaming SSE event boundaries
#18090
opened Aug 22, 2026 by
reasonsolo
Collaborator
Loading…
[https://nvbugs/6571410][fix] Move Gemma4 perf test to post-merge
#18089
opened Aug 21, 2026 by
2ez4bz
Collaborator
Loading…
1 task done
[https://nvbugs/6572838][fix] Fix LLM-only behavior for Mistral large
api-compatible
Accepted LLM API contract change that is backwards-compatible
ci: full pre-merge approved
#18086
opened Aug 21, 2026 by
2ez4bz
Collaborator
Loading…
1 task done
[None][perf] Fuse relu2 into a single elementwise kernel
#18082
opened Aug 21, 2026 by
nv-guomingz
Collaborator
Loading…
3 tasks done
[None][perf] Convolve the Mamba2 prefill on a channel-last projection
#18081
opened Aug 21, 2026 by
nv-guomingz
Collaborator
Loading…
3 tasks done
[None][feat] Add cuDNN attention backend
VisualGen
#18075
opened Aug 21, 2026 by
xrq-phys
Collaborator
Loading…
1 task done
[#17916][fix] Honor per-request enable_thinking in DeepSeekR1Parser
#18074
opened Aug 21, 2026 by
NishchalTendulkar
Loading…
6 tasks done
[TRTLLMINF-160][fix] Support CBTS test result reuse
#18073
opened Aug 21, 2026 by
hanjingtian
•
Draft
[TRTLLM-15719][test] add Qwen3.8-2.4T-A95B accuracy guards
#18071
opened Aug 21, 2026 by
Wanli-Jiang
Collaborator
Loading…
1 task done
[TRTLLMINF-331][infra] Conservatively enable multi-GPU tests when GitHub API fails
#18070
opened Aug 21, 2026 by
ZhanruiSunCh
Collaborator
Loading…
1 task done
[None][feat] Optimize human trigger replan for modeling agent
#18069
opened Aug 21, 2026 by
WeiHaocheng
Collaborator
Loading…
[TRTLLM-15649][feat] Forward MegaMoE SiTU kwargs to CuteDSL.
#18067
opened Aug 21, 2026 by
guqiqi
Contributor
Loading…
ProTip!
Updated in the last three days: updated:>2026-08-19.