vllm-project/vllm

Pull Requests Commits

remove cuda graph copy

yewentao256 committed 117 days ago

a409cf42

yewentao256 committed 117 days ago

10cf1e48

Merge branch 'main' into wentao-fix-torch-compile-issue

yewentao256 committed 117 days ago

d9cba20e

yewentao256 committed 117 days ago

a3da9190

[Doc] Add vllm-metal to hardware plugin documentation (#31174)

mgoin committed 117 days ago

Verified 95863540

[SM100] Enable fp8 compute for prefill MLA (#30746)

pavanimajety committed 117 days ago

Verified b10f41c8

[MoE Refactor][9/N] Use modular kernel for unquantized Triton MoE (#31052)

zyongye committed 117 days ago

Verified 7b926e89

[ROCm][CI/Build] Fix triton version to one that has triton_kernels required for gpt-oss to run (#31159)

gshtras committed 117 days ago

Verified ab3a85fd

[UX] improve profiler error message (#31125)

BoyuanFeng committed 117 days ago

Verified 8dd0db68

[ROCm] [Critical]: Remove unused variable (#31156)

tjtanaa committed 117 days ago

Verified 022f3cea

[AMD][CI] Add "V1 Test e2e + engine" to mi325_8 Agent Pool (#31040)

micah-wil committed 117 days ago

Verified a5bc77c2

[CI][Bugfix] Fix `entrypoints/openai/test_audio.py` (#31151)

NickLucche committed 117 days ago

Verified b1c3f96a

[BugFix] skip language model in Encoder (#30242)

Bounty-hunter committed 117 days ago

Verified 8f8f469b

[CI] add polling for precompiled wheel in python_only_compile.sh, fix index generation for releases (#30781)

Harry-Chen committed 117 days ago

Verified 2cf91c2e

[gpt-oss] Fix harmony parser in streaming responses (#30205)

AlonKejzman committed 117 days ago

Verified bd6d5a74

[Model] Fix bagel failed to run (#31132)

Potabk committed 117 days ago

Verified 256a33ec

Update MiniMax-M2 ToolCall and add MiniMax-M2.1 in Docs (#31083)

rogeryoungh committed 117 days ago

Verified c02a2705

[Bugfix][ROCm] Fix typo: is_linear_fp8_enaled -> is_linear_fp8_enabled (#31109)

c0de128 committed 117 days ago

Verified cf8eed7b

[Misc] Fix typo: 'occured' -> 'occurred' (#31120)

c0de128 committed 117 days ago

Verified 44ae85f7

[Misc] Fix spelling typos in model comments (#31117)

c0de128 committed 117 days ago

Verified 14c3e6ad

[Misc] Fix grammar errors in comments and messages (#31115)

c0de128 committed 117 days ago

Verified 42b42824

[Misc] Fix quantization-related typos (#31116)

c0de128 committed 117 days ago

Verified ec58c10c

[Misc] Fix spelling typos in comments (#31114)

c0de128 committed 117 days ago

Verified 8c084de5

[Feature]: Support NVIDIA ModelOpt HF FP8 variants FP8_PER_CHANNEL_PER_TOKEN and FP8_PB_WO in vLLM (#30957)

CedricHwong committed 117 days ago

Verified 19cc9468

[Kernel] Enable fused_qknorm_rope_kernel supports partial rope (#30821)

jeejeelee committed 117 days ago

Verified 097978a1

[CI] Fix "2 Node Tests (4 GPUs in total)" (#31090)

LucasWilkinson committed 117 days ago

Verified 7e065eba

[Doc] Clarify FP8 KV cache computation workflow (#31071)

westers committed 117 days ago

Verified 9d701e90

[NVFP4][Perf] Tune NVFP4 input quant kernel for small batch size (#30897)

mgoin committed 118 days ago

Verified 06d49028

[MoE Refactor][4/N] Marlin Fp8 Mk (#31036)

robertgshaw2-redhat committed 118 days ago

Verified b471092d

ci: add nvidia-smi warmup before Prime-RL integration test (#31093)

AmeenP committed 118 days ago

Verified 93cabc41

Older