vllm-project/vllm

Pull Requests Commits

zhuohan123 committed 1 year ago

f048f16b

zhuohan123 committed 1 year ago

180880dd

Revert "[Misc] Remove use of CUDA_VISIBLE_DEVICES for device selection (fix DP slow startup time &c) (#26709)"

zhuohan123 committed 1 year ago

3e0a770c

[KVConnector] Migrate the LMCache integration code to be vLLM native (#25542)

ApostaC committed 1 year ago

Verified 83f478bb

[Misc][DP] Guard mxfp4 implementation selection (#27484)

varun-sundar-rabindranath committed 1 year ago

Verified 269c4db0

[Log] Optimize Startup Log (#26740)

yewentao256 committed 1 year ago

Verified 52efc34e

[Distributed] Basic set of configuration for large EP deployment on GB200 (#27328)

wpc committed 1 year ago

Verified d95d0f4b

[Perf][Async Scheduling] Remove CPU->GPU sync in dummy_run (#27455)

lhtin committed 1 year ago

Verified 04024282

[Document] Add ms-swift library to rlhf.md (#27469)

hjh0119 committed 1 year ago

Verified 17af6aa0

[CI/Build] Fix test_torch_utils in AMD CI (#27317)

zhewenl committed 1 year ago

Verified fc168c33

[Bugfix] Fix interns1-vit qk norm code path (#27480)

Isotr0py committed 1 year ago

Verified acc78aeb

[Attention] Add MLA prefill backend: trtllm_ragged_attention_deepseek (#26397)

minosfuture committed 1 year ago

Verified 0f67d4d9

[Bugfix] Fix MultiConnector stats reconstruction across process boundaries (#27366)

kouroshHakha committed 1 year ago

Verified 7e1d697b

[NIXL][BUGFIX] delay done_recving queue cleanup to bottom of get_finished (#27297)

xuechendi committed 1 year ago

Verified 699d62e6

[compile] Turn standalone_compile back on (#27460)

zou3519 committed 1 year ago

Verified cd390b60

[cpu][fix] Fix onednn_mm crash on consecutive matmuls with same M,K,N and different dtype (#27472)

fadara01 committed 1 year ago

Verified 2080b050

[Doc] Fix minor issues in docs/design/metrics.md (#27436)

draftbk committed 1 year ago

Verified 6454afec

Fix test named tool use (#27458)

chaunceyjiang committed 1 year ago

Verified 41a62564

[MISC] `cudagraph_capture_sizes` related improvements (#26016)

fhl2000 committed 1 year ago

Verified 284cc922

Fix AArch64 CPU Docker pipeline (#27331)

ioghiban committed 1 year ago

Verified 435be10d

[Benchmark] Enable benchmark to run with `encoding_format="bytes"` (#27467)

DarkLight1337 committed 1 year ago

Verified b7030d96

[Refactor] move tool parsing logic from protocol.py to the tool parser (#27383)

chaunceyjiang committed 1 year ago

Verified 35678169

[BugFix] Fix torchrun DP with LLM class (#27395)

22quinn committed 1 year ago

Verified e0ef8a29

[MM][Bugfix] Replace `PatchEmbed`'s conv3d to linear layer (#27418)

Isotr0py committed 1 year ago

Verified 42efe609

[Docs] remove v1 column for embedding models (#27446)

piood committed 1 year ago

Verified 88d3141e

[Misc] Avoid "PyTorch non-writable tensors" warning in RayPPCommunicator (#27443)

ruisearch42 committed 1 year ago

Verified 09a6a49e

[Bugfix] Fix Pydantic union resolution for ResponseFunctionToolCall in Responses API (#26706)

strinczer committed 1 year ago

Verified 07447554

[Chore] remove structural tags logging lines (#27451)

aarnphm committed 1 year ago

Verified d4c574c3

Fix EventPublisherFactory logic for disabled KV cache events (#27419)

usberkeley committed 1 year ago

Verified c528b900

[Bugfix][CI] Move resolving cudagraph_mode before initializing attn_metadata_builder (#27427)

fhl2000 committed 1 year ago

Verified 85fee74b

Older