vllm-project/vllm

Pull Requests Commits

khluu committed 55 days ago

43e970bc

Add support for ModelOpt MXFP8 dense models (#33786)

danisereb committed 67 days ago

Verified 084aa19f

glm 4.6 fused tuned inference config for B200 (#32958)

navmarri14 committed 67 days ago

Verified 1ecfabe5

[torch.compile] Add an option to force-enable the MOE cold start optimization (#33735)

zou3519 committed 67 days ago

Verified 4df841fe

[BugFix] Change support no act and mul for marlin (#34088)

TomerBN-Nvidia committed 67 days ago

Verified a263aa61

[Revert] Fix performance regression for GLM-4.7-GPTQ decode and MTP acceptance rate (#33771)

aabbccddwasd committed 67 days ago

Verified 179ae7da

Add embedding input functionality for disabled modalities [remake] (#32493)

reaganjlee committed 67 days ago

Verified c4df59ad

[ROCm] [CI] Reduce Resource of two test groups (#34059)

tjtanaa committed 68 days ago

Verified 785cf28f

[Perf] Simplify DeepseekV32 tokenizer, ensure fast detokenization used (#33855)

njhill committed 68 days ago

Verified a96197f5

[ROCm][Bugfix] fix act_quant_fusion module import error (#34069)

AndreasKaratzas committed 68 days ago

Verified ab10d798

[CI/Build] Skip GCS test (#34057)

DarkLight1337 committed 68 days ago

Verified 7fcb705b

[Doc] Fix run_batch docs (#34056)

DarkLight1337 committed 68 days ago

Verified b956cdf8

Perf tuning and expansion of cases covered for wvSplitKrc (#33493)

amd-hhashemi committed 68 days ago

Verified ed17f54c

Make directory exist ok for ray spinning up multiple replicas on a single instance (#33604)

jiangwu300 committed 68 days ago

Verified 860981d8

Update DeepGEMM version pin in Dockerfile to match #32479 (#33935)

zifeitong committed 68 days ago

Verified 52181baa

move checks out of `unified_kv_cache_update` custom op (#33943)

Rohan138 committed 68 days ago

Verified de3869bb

[PluggableLayer][3/N] Apply PluggableLayer to mamba layers. (#33660)

whx-sjtu committed 68 days ago

Verified ce9b3cd3

[Model] Enable Step3p5ForCausalLM testing (#33755)

jeejeelee committed 68 days ago

Verified db4ede97

[Frontend]Add support for transcriptions and translations to run_batch (#33934)

pooyadavoodi committed 68 days ago

Verified 2cb2340f

Enable Eagle3 speculative decoding for Mistral3ForConditionalGeneration to support eagle3 (#33939)

TundeAtSN committed 68 days ago

Verified 4df44c16

[torch.compile] Stop compiling identical artifacts (#34003)

zou3519 committed 68 days ago

Verified 81fe69ca

[Kernel] Add KernelConfig flag to enable/disable FlashInfer autotune (#34006)

mmangkad committed 68 days ago

Verified dd6a6e11

[Renderer] Define `render_cmpl` and `render_chat` (#34039)

DarkLight1337 committed 68 days ago

Verified edb359cc

[CI][Build] Pin grpcio-tools==1.78.0 (#34048)

noooop committed 68 days ago

Verified 6ed5eda3

[Misc] Simplify `get_max_tokens` (#34036)

DarkLight1337 committed 69 days ago

Verified 11a4c9d3

Fix spelling errors (#33978)

sleepcoo committed 69 days ago

Verified 15a0b9e5

[ROCm][CI] Pinning lm-eval version to resolve multi-modal small eval bug (#34038)

AndreasKaratzas committed 69 days ago

Verified c490d8cc

[Misc] Make `PlaceholderRange.get_num_embeds` a method (#34035)

DarkLight1337 committed 69 days ago

Verified 48312e57

[Kernel] Add enable_sm120_or_later for SM121 (DGX Spark) CUTLASS support (#33517)

Code4me2 committed 69 days ago

Verified bc32444b

[Revert] Add util `handle_deprecated` back (#33998)

yewentao256 committed 69 days ago

Verified 18e85452

Older