vllm-project/vllm

Pull Requests Commits

Merge branch 'main' into wentao-small-refactor

yewentao256 committed 12 hours ago

Verified db006e8d

[Feature]Add EVS (Efficient Video Sampling) Support for Qwen3-VL (#29752)

skyloevil committed 14 hours ago

Verified ae88aada

[Bugfix] Improve error messages in ModelConfig validation (#30213)

yifant-code committed 14 hours ago

Verified 5ccf0efa

[Bugfix] Fix fusion for VL models (#30244)

ElizaWszola committed 14 hours ago

Verified 994acec0

[Bugfix] Revert Qwen2-VL part of change in #28271 (#30542)

zifeitong committed 14 hours ago

Verified 48b8456f

[Bugfix] Update get_processor_data to use get_all method (#30583)

dbotwinick committed 14 hours ago

Verified 5b64ac21

[Misc] Add a script to benchmark compilation time (#29919)

desertfire committed 14 hours ago

Verified a8ec4865

[Bugfix] fix _get_quant_method of FusedMoE for deepseekV3.2 on non-NV… (#30057)

tom-zju committed 17 hours ago

Verified 6ecc1e41

Nvidia ModelOpt workaround for issue 28072 (#30164)

shengliangxu committed 17 hours ago

Verified 0bb0bae4

fix: Update json features supported by xGrammar (#30390)

johannesflommersfeld committed 17 hours ago

Verified 06089365

[Bugfix] awq_gemm: fix argument order swap (#30364)

mgehre-amd committed 17 hours ago

Verified e9add129

[torch.compile] Add encoder tag for compilation (#30489)

ilmarkov committed 17 hours ago

Verified 3224ea99

Add AudioFlamingo3 model support (#30539)

lashahub committed 17 hours ago

Verified 3a20450d

[Doc]: fixing typos in various files (#30540)

didier-durand committed 17 hours ago

Verified 1a55cfaf

[main][BugFix] Fixed an accuracy bug of Qwen3-next-MTP when batched inferring (#30632)

drslark committed 17 hours ago

Verified add1b9d3

[Chore] Remove redundant `RequestPrompt` (#30612)

DarkLight1337 committed 18 hours ago

Verified dcb31196

enable unbacked with aot_compile (#30462)

laithsakka committed 19 hours ago

Verified f569c654

[ROCm][CI] Add "Qwen3-Next-80B-A3B-Instruct MTP Async EPLB Accuracy Test" Back Into AMD CI (#30590)

micah-wil committed 20 hours ago

Verified 97f2f160

Improve parse_raw_prompt test cases for invalid input .v2 (#30512)

mivehk committed 1 day ago

Verified 29f7d977

[Bug][KVConnector][Metrics] Remove a vacuous assertion breaking external-launcher (#30577)

QierLi committed 1 day ago

Verified dc7fb5be

[Doc] Add instructions for building docker image on GB300 with CUDA13 (#30414)

soodoshll committed 1 day ago

Verified 24429d59

[Logs] Optimize startup logs 4 (#29903)

yewentao256 committed 1 day ago

Verified 6e78ed6b

[Doc] Add documents for multi-node distributed serving with MP backend (#30509)

Isotr0py committed 1 day ago

Verified 7c16f3fb

[Docs] Clarify Expert Parallel behavior for attention and MoE layers (#30615)

majiayu000 committed 1 day ago

Verified ddbfbe52

Merge branch 'main' into wentao-small-refactor

robertgshaw2-redhat committed 1 day ago

Verified 968934e7

set assume_32bit_indexing and pass unbacked hints (#30459)

laithsakka committed 1 day ago

Verified 763963aa

[Refactor] `TokenizerRegistry` only uses lazy imports (#30609)

DarkLight1337 committed 1 day ago

Verified 39cefbdf

[Bugfix] Qwen3-next with --hf-overrides \{\"num_hidden_layers\":8\} (#30433)

heheda12345 committed 1 day ago

Verified ace34e37

[CI/Build] Fix broken mm processor test Mistral-3-large (#30597)

Isotr0py committed 1 day ago

Verified e5db3e27

[Chore] Adjust tokenizer import to avoid circular imports (#30601)

DarkLight1337 committed 1 day ago

Verified 64251f48

Older