Skip to content

Pull requests: vllm-project/vllm

New pull request New
1,198 Open 14,192 Closed
1,198 Open 14,192 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

Install pre-built xformers-0.0.32.post2 built with pt-2.9.0 ci/build
#27598 opened Oct 27, 2025 by huydhn Loading… v0.11.1
4
[Model] fix glm4_moe_mtp load weights with GLM-4.6 checkpoint.
#27597 opened Oct 27, 2025 by wuyaoxuehun Loading…
6
[Bugfix][Frontend] validate arg priority in frontend LLM class before add request frontend
#27596 opened Oct 27, 2025 by junpuf Loading…
3 tasks
4
Fix intermediatetensors spawn error #27591 qwen Related to Qwen models
#27594 opened Oct 27, 2025 by baonudesifeizhai Loading…
5 tasks
2
Remove deprecated fields from CompilationConfig documentation Improvements or additions to documentation llama Related to Llama models v1
#27593 opened Oct 27, 2025 by hmellor Loading…
3
[Stability fix] turn off HMA allocator when connector is set ready ONLY add when PR is ready to merge/full CI is needed
#27592 opened Oct 27, 2025 by KuntaiDu Loading…
5 tasks
v0.11.1
9
[Bug] Fix shape issue for eplb expert weights ready ONLY add when PR is ready to merge/full CI is needed
#27589 opened Oct 27, 2025 by yewentao256 Loading…
2
[Misc] Separate out utils.counter and move utils.Device to engine frontend ready ONLY add when PR is ready to merge/full CI is needed v1
#27588 opened Oct 27, 2025 by DarkLight1337 Loading…
5 tasks
5
[Build] Optimize docker layers for better caching ci/build ready ONLY add when PR is ready to merge/full CI is needed
#27585 opened Oct 27, 2025 by rzabarazesh Loading…
5 tasks
3
Rename clashing method names for vLLM model protocol deepseek Related to DeepSeek models documentation Improvements or additions to documentation gpt-oss Related to GPT-OSS models llama Related to Llama models multi-modality Related to multi-modality (#4194) qwen Related to Qwen models speculative-decoding tpu Related to Google TPUs v1
#27583 opened Oct 27, 2025 by hmellor Loading…
4
[ROCm][GEMM] update aiter fp8 linear check for different platforms rocm Related to AMD ROCm
#27579 opened Oct 27, 2025 by zhuyuhua-v Loading…
5
[perf] Enable concurrent execution of "shared_experts" and "selected_experts" qwen Related to Qwen models
#27578 opened Oct 27, 2025 by ZJY0516 Loading…
5 tasks
1
[Prefix Cache] Include lora_name in BlockStored event for deterministic KV-cache reconstruction documentation Improvements or additions to documentation kv-connector v1
#27577 opened Oct 27, 2025 by sagiahrac Loading…
5
[CI][Bugfix] Fix triton import check error
#27574 opened Oct 27, 2025 by MengqingCao • Draft
5 tasks
4
[Bugfix][P/D] Fix throughput stats in disaggregated setup kv-connector v1
#27569 opened Oct 27, 2025 by NickLucche Loading…
9
[DeepSeek v3.2] Make top-k work for any logit values. deepseek Related to DeepSeek models
#27568 opened Oct 27, 2025 by dcampora Loading…
5 tasks
16
Fix a robust parsing issue in KimiK2ToolParser that causes IndexError frontend ready ONLY add when PR is ready to merge/full CI is needed tool-calling
#27565 opened Oct 27, 2025 by wangln19 Loading…
3 of 10 tasks
@chaunceyjiang
5
[Misc] Replace CUDA_VISIBLE_DEVICES in DP with torch.cuda.set_device for device selection on cuda-like devices kv-connector v1
#27564 opened Oct 27, 2025 by ilmarkov Loading…
5 tasks
4
[Bugfix][Rocm] Fix shared expert weight loading failure in DeepSeek-MTP deepseek Related to DeepSeek models rocm Related to AMD ROCm
#27563 opened Oct 27, 2025 by zhyajie Loading…
5
[Bugfix] Validate custom logits processor xargs for online serving deepseek Related to DeepSeek models documentation Improvements or additions to documentation frontend ready ONLY add when PR is ready to merge/full CI is needed v1
#27560 opened Oct 27, 2025 by Isotr0py • Draft
1 of 5 tasks
6
[Bugfix] fixed inconsistent finish_reason handling between V0 and V1 engines ready ONLY add when PR is ready to merge/full CI is needed v1
#27555 opened Oct 27, 2025 by chaunceyjiang Loading…
5 tasks
1
@DarkLight1337
4
covt_e4m3_bf16
#27553 opened Oct 27, 2025 by wangyxbh Loading…
5 tasks
4
[CI/Build] Test torchrun with 8 cards ci/build documentation Improvements or additions to documentation ready ONLY add when PR is ready to merge/full CI is needed
#27548 opened Oct 27, 2025 by 22quinn Loading…
3 of 5 tasks
7
[BUG] Fix hybrid kvcache kernel page size issue v1
#27547 opened Oct 27, 2025 by vadiklyutiy Loading…
4
[Docs] add Shanghai Meetup - 2025/10 documentation Improvements or additions to documentation
#27545 opened Oct 27, 2025 by kebe7jun Loading…
5 tasks
@youkaichao
3
ProTip! Type g i on any issue or pull request to go back to the issue listing page.