Skip to content

Pull requests: vllm-project/vllm

New pull request New
1,273 Open 16,019 Closed
1,273 Open 16,019 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

Upstream fp8 with static scales gpt oss gpt-oss Related to GPT-OSS models needs-rebase
#30357 opened Dec 9, 2025 by maleksan85 • Draft
9
[CI][DeepSeek] Add nightly DeepSeek R1 lm_eval tests on H200 ci/build deepseek Related to DeepSeek models ready ONLY add when PR is ready to merge/full CI is needed
#30356 opened Dec 9, 2025 by MatthewBonanni Loading…
2 of 5 tasks
3
[Fix] Handle multiple tool calls in Qwen3-MTP tool parser frontend qwen Related to Qwen models tool-calling
#30353 opened Dec 9, 2025 by ArkVex Loading…
1
9
[Bugfix] Cache added_vocab to avoid per-token overhead
#30351 opened Dec 9, 2025 by scratch-ml Loading…
5 tasks
3
Remove virtual engine handling codex kv-connector needs-rebase qwen Related to Qwen models tpu Related to Google TPUs v1
#30350 opened Dec 9, 2025 by WoosukKwon Loading…
2
[BugFix] Fix minimax m2 model rope_parameters
#30349 opened Dec 9, 2025 by esmeetu Loading…
5 tasks
@hmellor
5
[Docs]: adds a new metric vllm:request_prefill_kv_computed_tokens in docs documentation Improvements or additions to documentation
#30348 opened Dec 9, 2025 by googs1025 Loading…
5 tasks
3
[cpu][ci] Add CPU Attention Tests for Neon Backend
#30347 opened Dec 9, 2025 by fadara01 Loading…
2 tasks
9
[Core] Major fix catch backend grammar exceptions (xgrammar, outlines, etc) in scheduler v1
#30346 opened Dec 9, 2025 by blancsw Loading…
8
Fix typos in comments across multiple files documentation Improvements or additions to documentation ready ONLY add when PR is ready to merge/full CI is needed v1
#30345 opened Dec 9, 2025 by wilsonwu Loading…
5 tasks
4
[Bugfix] Fix HunyuanOCR cross-image contamination in batch processing
#30344 opened Dec 9, 2025 by anker-c2 Loading…
3 of 5 tasks
1
7
[CI] refine more logic when generating and using nightly wheels & indices ci/build
#30341 opened Dec 9, 2025 by Harry-Chen Loading…
3 of 5 tasks
10
Add Eagle and Eagle3 support to Transformers modeling backend
#30340 opened Dec 9, 2025 by hmellor • Draft
2 tasks
2
[CMake][Build]: Remove unused ACL CMake env variables ci/build
#30339 opened Dec 9, 2025 by Radu2k Loading…
5
Fix gigachat3 parser + update tests frontend tool-calling
#30338 opened Dec 9, 2025 by ajpqs Loading…
3 of 5 tasks
4
fix: enhance human_readable_int function
#30337 opened Dec 9, 2025 by andyxning Loading…
5 tasks
2
[Bugfix] Fix fp8 DeepGemm compilation issues bug Something isn't working ci-failure Issue about an unexpected test failure in CI deepseek Related to DeepSeek models ready ONLY add when PR is ready to merge/full CI is needed
#30336 opened Dec 9, 2025 by ElizaWszola Loading…
8
[Bugfix]: Streaming i/o of batch files. Resolves #30268 ci/build frontend
#30334 opened Dec 9, 2025 by umgefahren Loading…
3 of 5 tasks
7
[Bugfix] tpu_model_runner: set vllm config context when calling reset_dynamo_cache() ready ONLY add when PR is ready to merge/full CI is needed tpu Related to Google TPUs v1
#30331 opened Dec 9, 2025 by dtrifiro Loading…
7
[Bugfix] Fix cuda graph sizes when running with speculative decoding nvidia ready ONLY add when PR is ready to merge/full CI is needed
#30330 opened Dec 9, 2025 by PatrykSaffer Loading…
3
[Feature][CPU Backend]: Add PyTorch vectorized backend
#30329 opened Dec 9, 2025 by Radu2k Loading…
10
[BugFix] Fix hang issue in LMCache mp mode kv-connector v1
#30327 opened Dec 9, 2025 by wz1qqx Loading…
5 tasks
5
[Frontend] [Doc] Exclude log deltas feature frontend
#30322 opened Dec 9, 2025 by Catacomba Loading…
3 tasks done
2
[BugFix] Spec decode with VLLM_ENABLE_V1_MULTIPROCESSING=0 v1
#30319 opened Dec 9, 2025 by heheda12345 Loading…
5 tasks
3
[Frontend] Allow users to modify the scheduler configuration online in dev mode. frontend v1
#30316 opened Dec 9, 2025 by noooop • Draft
5 tasks
4
ProTip! Filter pull requests by the default branch with base:main.