vllm-project/vllm

Pull Requests Commits

LucasWilkinson committed 2 days ago

e5131146

[Model Runner V2] Remove unnecessary copies in PW CUDA graph capture (#34849)

WoosukKwon committed 2 days ago

Verified c878b43b

[CI][AMD][BugFix] Use torch.testing.assert_close instead of assert torch.allclose in test_rocm_skinny_gemms.py (#34181)

rasmith committed 2 days ago

Verified 2b84ac66

[Model Runner V2] support piecewise & mixed cudagraph (#32771)

izhuhaoran committed 2 days ago

Verified 11d3976b

[MoE Refactor] Convert mxfp4 marlin into modular kernel format (#34588)

zyongye committed 2 days ago

Verified 40da9625

Fix empty tool_call_id in Anthropic messages API tool result conversion (#34745)

sfeng33 committed 2 days ago

Verified 8d9babd4

[BUG] Fixing Weight Sync unit test (#34841)

hao-aaron committed 2 days ago

Verified e99ba957

[Docs] Clean up speculators docs (#34065)

kylesayrs committed 2 days ago

Verified 64ac1395

[Bugfix] Fix lora tests (#34834)

DarkLight1337 committed 2 days ago

Verified 61cf0876

[Bugfix][MoE Kernel] Fix incorrect routing selection for models without expert groups (e.g., MiniMax-M2.1) (#34673)

wwl2755 committed 2 days ago

Verified 847a57cd

[CI][AMD][BugFix] Skip tests in test_unquantized_backend_selection that should not run on ROCm (#34655)

rasmith committed 2 days ago

Verified fcd6ac97

[Model Runner V2] Minor simplification for DCP (#34786)

WoosukKwon committed 2 days ago

Verified 95be2a7f

[Bugfix] Remove assert causing hipErrorStreamCaptureUnsupported (#34455)

JadenMathias committed 2 days ago

Verified 0e60c925

[Misc] Add mooncake-transfer-engine to kv_connectors requirements (#34826)

stmatengss committed 2 days ago

Verified d7ff2220

[Bugfix] Redo Qwen3.5/Qwen3-Next GDN projector fusion (#34697)

Isotr0py committed 2 days ago

Verified c0bd8b13

[Bugfix] Fix NVFP4 TRTLLM MoE non-gated support; add gsm8k for Nemotron-3-Nano FP8+NVFP4 (#34725)

mgoin committed 2 days ago

Verified caeb887b

[CI][Bugfix] Fix multinode test script (#34820)

ilmarkov committed 2 days ago

Verified 6b3166a7

[CI] temporarily disable multi-node tests (#34825)

robertgshaw2-redhat committed 2 days ago

Verified 25e2e136

[Model Bash] DeepSeek R1 BF16 Min Latency QKV A GEMM (0.5% E2E Speedup) (#34758)

robertgshaw2-redhat committed 2 days ago

Verified 6874638b

Add unit tests for fp8 output fusion of triton_attn (#34228)

bringlein committed 2 days ago

Verified e24663c5

[Model Runner V2] Avoid prepare prefill kernel launch overhead (#34780)

njhill committed 2 days ago

Verified c50e105a

[Renderer] Deprecate code paths for old input processing (#34775)

DarkLight1337 committed 2 days ago

Verified a766b303

[Quantization] - Added uses_meta_device_weights to quant config (#34645)

Josephasafg committed 2 days ago

Verified 1faa8cb7

[Bugfix] fix activation in cpu_fused_moe_torch call (#34696)

michalowski-arm committed 2 days ago

Verified e89a91d9

[Bugfix] Fix prefix creation for Qwen3.5 (#34723)

mgoin committed 2 days ago

Verified 909b1471

[Bugfix] Fix quant RMS norm fusion for quantization with TMA-aligned scales (#33255)

ElizaWszola committed 2 days ago

Verified a88b3be7

[Model Runner V2] A bit more PP simplification (#34766)

njhill committed 2 days ago

Verified a49ea5a5

[CI/Build] Remove use of `skip_v1` (#34699)

DarkLight1337 committed 2 days ago

Verified 30ebe0dc

[ROCm][CI] Removed hard-coded attn backend requirement for Qwen VL (#34753)

AndreasKaratzas committed 2 days ago

Verified cef65f07

[Core] Fix SSRF bypass via backslash-@ URL parsing inconsistency (#34743)

russellb committed 2 days ago

Verified 6f3b2047

Older