Skip to content

Pull requests: NVIDIA/TensorRT-LLM

New pull request New
260 Open 1,966 Closed
260 Open 1,966 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[doc] Add speculative decoding PyTorch docs
#4962 opened Jun 5, 2025 by mikeiovine Loading…
Fix Llama-3_3-Nemotron-Super-49B-v1 FP8 accuracy threshold configs
#4961 opened Jun 5, 2025 by moraxu Loading…
2
Doc: Add info about stop words appearing in output
#4956 opened Jun 5, 2025 by Linda-Stadter Loading…
Add customized renormalized moe routing kernel for moe cutlass backend
#4955 opened Jun 5, 2025 by ChristinaZ Loading…
@ChristinaZ
3
[nvbug 5325284][fix] Increase Nemotron-H warmup request robustness
#4954 opened Jun 5, 2025 by tomeras91 Loading…
5
[fix] Add MPI barrier for proper NVLS multicast team initialization
#4952 opened Jun 5, 2025 by ikryukov Loading…
ci: [nvbugs/5280806] Unwaive unittests/_torch.
#4951 opened Jun 5, 2025 by yuxianq Loading…
12
[WIP] Introduce Flux MoE operator
#4948 opened Jun 5, 2025 by lancelly • Draft
[fix] Fix illegal mem access and possible accuracy lose
#4943 opened Jun 5, 2025 by liji-nv Loading…
8
[TRTLLM-5512] - Move part of tests from A100X to A100_80GB_PCIE multi…
#4942 opened Jun 5, 2025 by yiqingy0 Loading…
6
fix: Refactor the first token response in PD (#4692)
#4941 opened Jun 5, 2025 by Shunkangz Loading…
@Shunkangz
6
[TRTLLM-5692][tests] Add speculative decoding test cases on torch flow
#4940 opened Jun 5, 2025 by crazydemo Loading…
3
User/zhanruis/0605 test use build stage wheels for release
#4939 opened Jun 5, 2025 by ZhanruiSunCh Loading…
43
Perf: cache tokens in Python side to reduce reading overhead of pybind
#4938 opened Jun 5, 2025 by QiJune Loading…
6
Raise shut down error for each request
#4936 opened Jun 5, 2025 by Shunkangz • Draft
@Shunkangz
2
fix: Mapping rank boundary check bug
#4935 opened Jun 5, 2025 by venkywonka Loading…
4
[TRTLLM-4932] Add Llama-3.1-Nemotron-Nano-8B-v1-FP8 accuracy tests
#4933 opened Jun 5, 2025 by moraxu Loading…
fix: [nvbugs/5324229] Fix broken WInt4AFP8FusedMoEMethod since FusedMoE refactor.
#4930 opened Jun 5, 2025 by yuxianq Loading…
5
Kv cache transfer support duplicate heads
#4929 opened Jun 5, 2025 by chuangz0 Loading…
9
chore: cleanup GDS Cmake interface
#4928 opened Jun 5, 2025 by achartier Loading…
7
fix:https://nvbugs/5324252
#4925 opened Jun 5, 2025 by nv-guomingz Loading…
6
fix: trtllm-bench --dataset required=True
#4924 opened Jun 5, 2025 by jasonqinzhou Loading…
@jasonqinzhou
Coalesce text diffs in streaming requests.
#4923 opened Jun 5, 2025 by pathorn Loading…
fix: [nvbug 5321627] handle cases when TRT backend return more logits than output tokens
#4921 opened Jun 4, 2025 by hchings Loading…
@hchings
8
chore: Refactor apply_rope.
#4918 opened Jun 4, 2025 by bobboli Loading…
31
ProTip! Updated in the last three days: updated:>2025-06-02.