vllm-project/vllm

Pull Requests Commits

[MRV2] Support Breakable CUDA graph

WoosukKwon committed 12 hours ago

8d8016e4

[Refactor] Remove dead current_tool_name_sent assignments from tool parsers (#43997)

sfeng33 committed 14 hours ago

Verified 1a096d82

offload prompt_embeds decode in render_prompts_async to avoid blocking (#43792)

gagandhakrey committed 14 hours ago

Verified 1e2ce5d1

[PERF]MiniMax-M2 gate kernel (#38445)

jeejeelee committed 15 hours ago

Verified 559d6710

Revert "[MoE Refactor] Migrate MoeWNA16Method quantization to MK orac… (#44033)

bnellnm committed 16 hours ago

Verified 187457a9

[CI] Fix smoke test step key to bypass block gate (#43974)

khluu committed 17 hours ago

Verified 8fad2665

[CI] Remove duplicate Harmony test coverage (#44023)

sfeng33 committed 17 hours ago

Verified 8c6daf6e

[MoE Refactor] Remove supports_expert_map (#43108)

bnellnm committed 19 hours ago

Verified 7b98f498

[MoE Refactor] Migrate MoeWNA16Method quantization to MK oracle (#42647)

bnellnm committed 19 hours ago

Verified 106aa92f

[Fronten] Clean up stop_token_ids override for Harmony (#44009)

yzong-rh committed 20 hours ago

Verified 46409fd2

[Metrics] Exclude KV transfer tokens from iteration_tokens_total (#43346)

tlrmchlsmth committed 20 hours ago

Verified 38b864d8

[Feature] SSL support for dp supervisor (#43688)

yewentao256 committed 21 hours ago

Verified 5dbf1605

Add @khluu to CODEOWNERS (#44019)

khluu committed 21 hours ago

Verified acbc2033

[CI] Remove redundant test_chat_with_tool_reasoning.py (#44011)

sfeng33 committed 21 hours ago

Verified 6de08e8b

[CI] Make Model Executor test hangs fail fast with a traceback (#43971)

khluu committed 21 hours ago

Verified 6aabe221

[Bug] Fix torch device issue for MOE permute (#44005)

yewentao256 committed 21 hours ago

Verified 739096a0

[Bugfix] Fix Ray placement group allocation with grouped nodes (#43998)

czhu-cohere committed 21 hours ago

Verified 8b9deeec

[Bugfix] Use storage_block_size in KV cache reshape for compressed specs (DeepSeek V4) (#43988)

zixi-qi committed 22 hours ago

Verified d07ad069

[EPLB] Make async EPLB default (#43219)

ilmarkov committed 22 hours ago

Verified 4aaba00f

[MoE Refactor] WNA16 MoE backend selection into oracle module (#42553)

bnellnm committed 23 hours ago

Verified 84b2a8a7

[Bugfix] Disable allreduce_rms_fusion when pipeline_parallel_size > 1 (#43616)

zixi-qi committed 1 day ago

Verified 4ff865c3

[Misc] added unit tests for the core pooling methods (#43818)

taneem-ibrahim committed 1 day ago

Verified 5502c3b5

docs: clarify ITL acronym in optimization docs (#43922)

chunyang-wen committed 1 day ago

Verified f191d563

Add vLLM library info to Hugging Face Hub requests (#43857)

Wauplin committed 1 day ago

Verified 11dfa316

[Bugfix][CPU] Remove invalid extra deps (#43977)

bigPYJ1151 committed 1 day ago

Verified 3f6f508e

Skip docs build if PR doesn't affect docs (#43972)

hmellor committed 1 day ago

Verified 0585b5ba

[Bugfix] Corrupted MLA + linear attention (#43961)

gau-nernst committed 1 day ago

Verified d2889722

[ROCm][Perf] DSv3.2 MI355X TP4 decode-step orchestration cleanup (3 micro-opts) (#42982)

frida-andersson committed 1 day ago

Verified 0b56815a

[Bugfix] [ROCm] [DSV4] Fix AITER MXFP4 MoE weight loading and shuffle… (#42595)

MHYangAMD committed 1 day ago

Verified ab12aab1

[Kernel][ROCm] Native W4A16 kernel for AMD RDNA3 (gfx1100) — fp16 + bf16 (#41394)

JartX committed 1 day ago

Verified 0cff0741

Older