vllm-project/vllm

Pull Requests Commits

[V1] port xformers backend to v1 (#21342)

TheEpicDolphin committed 127 days ago

Verified 469b3ffa

[Refactor] Remove Unused Environment Variable `VLLM_NO_DEPRECATION_WARNING` (#22199)

yewentao256 committed 127 days ago

Verified ae87ddd0

[CI/Build] Update flashinfer to 0.2.9 (#22233)

mgoin committed 127 days ago

Verified a7cb6101

Use UV_LINK_MODE=copy in Dockerfile to avoid hardlink fail (#22128)

mgoin committed 127 days ago

Verified c494f96f

[V0 Deprecation][TPU] Remove V1 flag check from tests (#22248)

NickLucche committed 127 days ago

Verified 0c275ad5

[Misc] correct static type check for GroupCoordinator (#21946)

andyxning committed 128 days ago

Verified 74333ae2

[NVIDIA] Support Flashinfer TRT-LLM Prefill Attention Kernel (#22095)

elvischenv committed 128 days ago

Verified 83156c7b

[Feature] Non-contiguous Support for FP8 Quantization (#21961)

yewentao256 committed 128 days ago

Verified 4771df7b

Migrate KimiVLImagePixelInputs to TensorSchema (#21769)

bbeckca committed 128 days ago

Verified 05fae021

[Docs][TPU] Highlight TPU Software version selection (#22242)

NickLucche committed 128 days ago

Verified d1bf1b97

[Model] Pooling model activation supports per request control by PoolingParams (#20538)

noooop committed 128 days ago

Verified 586f2867

[Core] Factor out common logic for MM budget calculation (#22228)

DarkLight1337 committed 128 days ago

Verified 811ac13d

[UX] Fail if an invalid attention backend is specified (#22217)

mgoin committed 128 days ago

Verified e79a12fc

[Bugfix] Misaligned params in TreeAttentionImpl (#22226)

DarkLight1337 committed 128 days ago

Verified cdfd6871

Optimize configuration access with LRU cache in custom ops (#22204)

skyloevil committed 128 days ago

Verified 4b3e4474

[Misc] log more detailed message for ensure_model_parallel_initialized (#22144)

andyxning committed 128 days ago

Verified bd3db7f4

[Doc] add backend to doc string of initialize_model_parallel (#22142)

andyxning committed 128 days ago

Verified 29b97c09

[Misc] Remove pass_config from CompilationConfig dump_json excluded (#21911)

elvischenv committed 128 days ago

Verified 7b455cf1

fix: kimi_k2 return empty tool call list (#22149)

tlipoca9 committed 128 days ago

Verified 8a6e108e

[Log] DeepGEMM Update Log for Unaligned Problem Size (#22208)

yewentao256 committed 128 days ago

Verified d7b28f34

self.gate dtype update for GLM-4.5 (#22203)

zRzRzRzRzRzRzR committed 128 days ago

Verified 6fa41e0c

[ROCm][Bugfix] Compilation passes fix (#22202)

gshtras committed 128 days ago

Verified 031ca762

[FEAT] Refactor ROPE into module (#22192)

tjtanaa committed 128 days ago

Verified 6ad6b8e1

[V0 deprecation][P/D] Deprecate v0 `KVConnectorBase` code (1/2) (#21785)

lk-chen committed 128 days ago

Verified f4f4e7ef

[V1] reduce block size for tree attention correctness test to fix 'ou… (#22207)

TheEpicDolphin committed 128 days ago

Verified 5ea71ff4

Revert "[Bugfix] V1 Fix the cursor leakage issue during request scheduling." (#22223)

WoosukKwon committed 128 days ago

Verified 71758176

[Bugfix] V1 Fix the cursor leakage issue during request scheduling. (#21173)

CLFutureX committed 128 days ago

Verified 2dffac46

[NVIDIA] Auto detect modelopt quant and fix DSR1-FP4 weight loading (#22073)

nvpohanh committed 128 days ago

Verified bdcb42e4

[Bugfix][V1][P/D]Fix the uneven polling issue in the toy proxy for P2pNcclConnector (#21819)

Abatom committed 128 days ago

Verified c09efff9

[Bug] Update auto_tune.sh to separate benchmarking and profiling. (#21629)

ericehanley committed 128 days ago

Verified 309c1bb8

Newer Older