vllm-project/vllm

Pull Requests Commits

Initial stubs for P/D scheduling changes

tlrmchlsmth committed 274 days ago

161010c3

[New Model]: Snowflake Arctic Embed (Family) (#16649)

noooop committed 274 days ago

Verified 3d3ab368

Fix `nullable_kvs` fallback (#16837)

hmellor committed 274 days ago

Verified 686623c5

[Misc] Clean up Kimi-VL (#16833)

DarkLight1337 committed 274 days ago

Verified aadb6565

[Model] use AutoWeightsLoader for BigCode, GPT-J (#16823)

jonghyunchoe committed 274 days ago

Verified 87e067de

[Docs] Fix a link and grammar issue in production-stack.md (#16809)

windsonsea committed 274 days ago

Verified 26507f89

[Doc] add podman setup instructions for official image (#16796)

nathan-weinberg committed 274 days ago

Verified 9c1d5b45

[Bugfix] fix pp for llama4 (#16746)

luccafong committed 274 days ago

Verified e31045f9

[ROCm] [Attention] Cleanup ROCm output passing (#16431)

ProExpertProg committed 274 days ago

Verified aaec845f

[Misc] add collect_env to cli and docker image (#16759)

lengrongfu committed 274 days ago

Verified 7bdfd29a

Improve-mm-and-pooler-and-decoding-configs (#16789)

hmellor committed 274 days ago

Verified e78587a6

[BugFix] Accuracy fix for llama4 int4 - improperly casted scales (#16801)

LucasWilkinson committed 274 days ago

Verified 7eb42556

Add hardware print to TPU V1 test (#16792)

mgoin committed 274 days ago

Verified 6a0f5475

[V1][Structured Output] Minor modification to `_validate_structured_output()` (#16748)

shen-shanshan committed 274 days ago

Verified 30ed81b7

[Misc] Update outdated note: LMCache now supports chunked prefill (#16697)

chaunceyjiang committed 274 days ago

Verified 7a4a5de7

[Doc] Improve help examples for `--compilation-config` (#16729)

DarkLight1337 committed 274 days ago

Verified c16fb5da

Add property-based testing for vLLM endpoints using an API defined by an OpenAPI 3.1 schema (#16721)

tarukumar committed 274 days ago

Verified e37073ef

[Attention] Update to lastest FA3 code (#13111)

LucasWilkinson committed 275 days ago

Verified 183dad7a

[P/D][V1] KV Connector API V1 (#15960)

ApostaC committed 275 days ago

Verified 3408e471

[MLA] Simplification to batch P/D reordering (#16673)

njhill committed 275 days ago

Verified 0377b831

[V1][Metrics] Fix http metrics middleware (#15894)

markmc committed 275 days ago

Verified e4755f7f

[ROCM] enable aiter fused moe kernel for llama4 bf16 checkpoints (#16674)

sijiac committed 275 days ago

Verified 92edf358

[V1][TPU] Enable Top K (#15489)

NickLucche committed 275 days ago

Verified eb5819b2

[TPU][V1] Fix padding recompilation when `max-num-batched-tokens` is not even (#16726)

NickLucche committed 275 days ago

Verified 5989f468

[Model] use AutoWeightsLoader for olmoe,opt,orion,persimmon,phi3_small (#16548)

lengrongfu committed 275 days ago

Verified 5125d72f

[Kernel] Add fp8_w8a8 fused MoE kernel tuning configs for DeepSeek V3/R1 on NVIDIA H20 (#16753)

Ximingwang-09 committed 275 days ago

Verified a018e555

[Bugfix]Fix index out of range error in api server log (#16787)

WangErXiao committed 275 days ago

Verified 6211b922

[V1][Perf] Faster incremental detokenization (#15137)

njhill committed 275 days ago

Verified 05fcd1b4

[Doc] Changed explanation of generation_tokens_total and prompt_tokens_total counter type metrics to avoid confusion (#16784)

insukim1994 committed 275 days ago

Verified 7c02d6a1

[Doc] Document Matryoshka Representation Learning support (#16770)

noooop committed 275 days ago

Verified 11c3b984

Older