Pull Requests vllm-project/vllm

Upstream fp8 with static scales gpt oss needs-rebase gpt-oss

#30357 opened 2025-12-09 19:49 by maleksan85

[CI][DeepSeek] Add nightly DeepSeek R1 `lm_eval` tests on H200 ready ci/build deepseek

#30356 opened 2025-12-09 18:05 by MatthewBonanni

[Fix] Handle multiple tool calls in Qwen3-MTP tool parser frontend tool-calling qwen

#30353 opened 2025-12-09 17:48 by ArkVex

[Bugfix] Cache added_vocab to avoid per-token overhead

#30351 opened 2025-12-09 16:57 by scratch-ml

Remove virtual engine handling tpu needs-rebase v1 codex qwen kv-connector

#30350 opened 2025-12-09 16:34 by WoosukKwon

[BugFix] Fix minimax m2 model rope_parameters

#30349 opened 2025-12-09 15:42 by esmeetu

[Docs]: adds a new metric vllm:request_prefill_kv_computed_tokens in docs documentation

#30348 opened 2025-12-09 15:24 by googs1025

[cpu][ci] Add CPU Attention Tests for Neon Backend

#30347 opened 2025-12-09 15:14 by fadara01

[Core] Major fix catch backend grammar exceptions (xgrammar, outlines, etc) in scheduler v1

#30346 opened 2025-12-09 14:58 by blancsw

Fix typos in comments across multiple files documentation ready v1

#30345 opened 2025-12-09 14:56 by wilsonwu

[Bugfix] Fix HunyuanOCR cross-image contamination in batch processing

#30344 opened 2025-12-09 14:49 by anker-c2

[CI] refine more logic when generating and using nightly wheels & indices ci/build

#30341 opened 2025-12-09 14:17 by Harry-Chen

Add Eagle and Eagle3 support to Transformers modeling backend

#30340 opened 2025-12-09 14:09 by hmellor

[CMake][Build]: Remove unused ACL CMake env variables ci/build

#30339 opened 2025-12-09 14:09 by Radu2k

Fix gigachat3 parser + update tests frontend tool-calling

#30338 opened 2025-12-09 13:37 by ajpqs

fix: enhance human_readable_int function

#30337 opened 2025-12-09 13:28 by andyxning

[Bugfix] Fix fp8 DeepGemm compilation issues bug ready ci-failure deepseek

#30336 opened 2025-12-09 12:41 by ElizaWszola

[Bugfix]: Streaming i/o of batch files. Resolves #30268 frontend ci/build

#30334 opened 2025-12-09 11:32 by umgefahren

[Bugfix] tpu_model_runner: set vllm config context when calling reset_dynamo_cache() tpu ready v1

#30331 opened 2025-12-09 11:08 by dtrifiro

[Bugfix] Fix cuda graph sizes when running with speculative decoding nvidia

#30330 opened 2025-12-09 10:45 by PatrykSaffer

[Feature][CPU Backend]: Add PyTorch vectorized backend

#30329 opened 2025-12-09 10:18 by Radu2k

[BugFix] Fix hang issue in LMCache mp mode v1 kv-connector

#30327 opened 2025-12-09 09:57 by wz1qqx

[Frontend] [Doc] Exclude log deltas feature frontend

#30322 opened 2025-12-09 09:13 by Catacomba

[BugFix] Spec decode with VLLM_ENABLE_V1_MULTIPROCESSING=0 v1

#30319 opened 2025-12-09 08:26 by heheda12345

[Frontend] Allow users to modify the scheduler configuration online in dev mode. frontend v1

#30316 opened 2025-12-09 08:06 by noooop

Generalize pooling model support with multi-task, multi-layer, multi-label classification that can be pooled from both hidden states and LM head's logits.

#30315 opened 2025-12-09 07:56 by kflu

[fix] fix SM check for Flashinfer TRTLLM MOE nvidia

#30314 opened 2025-12-09 07:12 by jiahanc

[Misc][Quantization] Clarify the intent of GGUF `FusedMoE` weight materialization ready

#30310 opened 2025-12-09 06:40 by a4lg

[bugfix][quantization] fix quark qwen3 kv_cache quantization ready qwen

#30308 opened 2025-12-09 05:41 by haoyangli-amd

Fix incomplete response generation for tool call outputs frontend deepseek fb-exported meta-exported

#30304 opened 2025-12-09 05:01 by qandrew