Skip to content

Pull requests: vllm-project/vllm

New pull request New
507 Open 5,719 Closed
507 Open 5,719 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
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[ci] Add more source file dependencies for some tests ci/build
#13123 opened Feb 12, 2025 by khluu Loading…
1
[Bugfix] Allow fallback to AWQ from AWQMarlin at per-layer granularity quantization ready ONLY add when PR is ready to merge/full CI is needed
#13119 opened Feb 11, 2025 by mgoin Loading…
1
[ci] Consolidate Qwen models ready ONLY add when PR is ready to merge/full CI is needed structured-output
#13118 opened Feb 11, 2025 by khluu Loading…
1
[Misc] Log time consumption of sleep and wake-up
#13115 opened Feb 11, 2025 by waltforme Loading…
1
[Frontend] Pass pre-created socket to uvicorn frontend
#13113 opened Feb 11, 2025 by russellb Loading…
2
[Attention] Update to lastest FA3 code that supports different K and V head dims ci/build perf-benchmarks
#13111 opened Feb 11, 2025 by LucasWilkinson Loading…
2
[Kernel] LoRA - Refactor sgmv kernels
#13110 opened Feb 11, 2025 by varun-sundar-rabindranath Loading…
4
Support AWQMarlin with MLA
#13109 opened Feb 11, 2025 by mgoin • Draft
3
[Bug] [V1] Try fetching stop_reason from EngineOutput before checking the request ready ONLY add when PR is ready to merge/full CI is needed v1
#13108 opened Feb 11, 2025 by bnellnm Loading…
@njhill
5
Further reduce the HTTP calls to huggingface.co ready ONLY add when PR is ready to merge/full CI is needed
#13107 opened Feb 11, 2025 by maxdebayser Loading…
2
[BUG] Addreses #3935 and #3683, by making intial_incremental_detokenization_offset configurable frontend v1
#13106 opened Feb 11, 2025 by sinamoeini Loading…
1
1
[Quant] Add SupportsQuant to phi3 and clip
#13104 opened Feb 11, 2025 by kylesayrs Loading…
4
[Bugfix] Do not crash V0 engine on input errors
#13101 opened Feb 11, 2025 by joerunde Loading…
2
[Bugfix] deepseek_r1_reasoning_parser put reason content in wrong field in certain edge case frontend ready ONLY add when PR is ready to merge/full CI is needed
#13097 opened Feb 11, 2025 by LikeSundayLikeRain Loading…
4
[V1] LoRA - Add triton kernels for V1 v1
#13096 opened Feb 11, 2025 by varun-sundar-rabindranath • Draft
1
[V1][Kernel] Refactor the prefix_prefill kernel so that the caller no longer has to pass in the context lengths
#13095 opened Feb 11, 2025 by SageMoore Loading…
1
Consolidate Llama model usage in tests ready ONLY add when PR is ready to merge/full CI is needed speculative-decoding v1
#13094 opened Feb 11, 2025 by hmellor Loading…
12
[V1][Bugfix] DeepSeek-V3 v1 attn_backend miss q_lora_rank
#13092 opened Feb 11, 2025 by AoyuQC Loading…
1
[V1][Minor] Restore V1 compatibility with LLMEngine class
#13090 opened Feb 11, 2025 by Ryp Loading…
1
[core][dist] init device with current_platform.device_type
#13086 opened Feb 11, 2025 by MengqingCao Loading…
1
[Misc] Add model list API in disagg proxy
#13083 opened Feb 11, 2025 by ggaaooppeenngg Loading…
1
[V1][Pixtral-HF] Add custom slice_encoder_output for Pixtral v1
#13080 opened Feb 11, 2025 by lk-chen Loading…
@ywang96
1
Support logit_bias in v1 Sampler v1
#13079 opened Feb 11, 2025 by houseroad • Draft
7
[V1] Allow sliding window + prefix caching
#13069 opened Feb 11, 2025 by WoosukKwon Loading…
2
Run v1 latency benchmark and integrate with PyTorch OSS benchmark database ci/build
#13068 opened Feb 11, 2025 by huydhn Loading…
@ywang96
12
ProTip! Type g i on any issue or pull request to go back to the issue listing page.