vllm-project/vllm

Pull Requests Commits

Move`VllmConfig` from `config/__init__.py` to `config/vllm.py` (#25271)

hmellor committed 93 days ago

Verified 61aedb5f

[Benchmark] Support benchmark throughput for external launcher DP (#25913)

zhuohan123 committed 94 days ago

Verified d3bd1711

[Bug] Fix Weight Loading for Block FP8 Cutlass SM90 (#25909)

yewentao256 committed 94 days ago

Verified 89e4050a

Test Prompt Embeds/LoRA compatibility and Enable LoRA Support for OPT Models (#25717)

qthequartermasterman committed 94 days ago

Verified 78a47f87

[V0 Deprecation] Remove `vllm.worker` and update according imports (#25901)

aarnphm committed 94 days ago

Verified 6a113d9a

[NIXL] Increase default KV block eviction timeout on P (#25897)

NickLucche committed 94 days ago

Verified 2e4fe48c

[Doc] Polish example for torchrun dp (#25899)

zhuohan123 committed 94 days ago

Verified 8eb0a1d9

[Kernel] Chunk-aligned mamba2 (#24683)

tdoublep committed 94 days ago

Verified fea3e476

[Bugfix][ROCm] Fixing trying to import non-existent symbols from libnccl.so (#25605)

gshtras committed 94 days ago

Verified 61a34316

[Doc] Add documentation for vLLM continuous benchmarking and profiling (#25819)

namanlalitnyu committed 94 days ago

Verified 9bedac96

[BugFix][torch.compile] KV scale calculation issues with FP8 quantization (#25513)

adabeyta committed 94 days ago

Verified c42ff4f4

[Bugfix] Use correct key "ignore" for config.json non-quantized layers (#25706)

leejnau committed 94 days ago

Verified d5ab2851

[Model] Remove MotifForCausalLM (#25866)

jeejeelee committed 94 days ago

Verified e61eb5e0

[CI/Build] Include Transformers backend test in nightly transformers test (#25885)

Isotr0py committed 94 days ago

Verified 0899ba5b

[Bugfix][Speculative Decoding] Fix Eagle3 quantization config issue (#25883)

rahul-tuli committed 94 days ago

Verified 145ac733

[Nixl][P/D] Add cuda2cpu support (HD->DH transfer) (#24690)

chenxi-yang committed 94 days ago

Verified d0d138bc

[torch.compile] serialize cudagraph_mode as its enum name instead of value (#25868)

ZJY0516 committed 94 days ago

Verified 43227236

[Model][Bugfix] Fix issues in MiDashengLM implementation for quantized models (#25854)

zhoukezi committed 94 days ago

Verified 8616300a

[Bugfix] Fix requirements paths in install instructions (#25827)

yingjun-mou committed 94 days ago

Verified edbaadd9

update to latest deepgemm for dsv3.2 (#25871)

youkaichao committed 94 days ago

Verified 9360d34f

[Misc] Remove more `get_input_embeddings_v0` (#25857)

DarkLight1337 committed 94 days ago

Verified 1b67b046

[V0 Deprecation][Models] Remove all V0 condition for mm embeddings merge (#25331)

Isotr0py committed 94 days ago

Verified bd51f78e

[Bugfix] Fallback ViT attn backend to SDPA for blackwell (#25851)

ywang96 committed 94 days ago

Verified 65ecb4f1

[XPU]Fix xpu spec decoding UTs, avoid using cuda graph (#25847)

jikunshang committed 94 days ago

Verified 143844fa

Add Phi4FlashForCausalLM to _PREVIOUSLY_SUPPORTED_MODELS (#25832)

tdoublep committed 94 days ago

Verified 219cfbe7

[P/D] NIXL Updates (#25844)

robertgshaw2-redhat committed 94 days ago

Verified 9b44a7d9

[Misc] fix tests failure by using current_platform (#25825)

kingsmad committed 94 days ago

Verified a3ae45a3

Remove redundant cudagraph dispatcher warning (#25841)

mgoin committed 95 days ago

Verified 0307428d

[Bugfix] fix Qwen3VLMoe load when pp > 1 (#25838)

JJJYmmm committed 95 days ago

Verified 471997ad

Update GLM-4.5 Doc transformers version (#25830)

zRzRzRzRzRzRzR committed 95 days ago

Verified b1ded114

Newer Older