vllm-project/vllm

Pull Requests Commits

[torch.compile] caching of config fields should be opt-out by default (#26468)

vnadathur committed 22 days ago

Verified 1ffe934c

[CI] Reorganize compile tests so new tests are automatically included in CI (#28625)

gmagogsfm committed 22 days ago

Verified 2c8b9182

Relax Transformers modeling backend MoE experts check (#28952)

hmellor committed 22 days ago

Verified 4f5299f7

[Doc]: fix typos in various files (#29010)

didier-durand committed 22 days ago

Verified 09540cd9

[Feat][Perf] Enable deepep-low-latency with round-robin expert placement. (#28449)

cboss6 committed 22 days ago

Verified da2f6800

[config] Expose `get_total_num_hidden_layers()` in ModelConfig (#28961)

ptovam committed 22 days ago

Verified ba558c02

[Docs] Take env var definition out of folded admonition (#29005)

hmellor committed 22 days ago

Verified 97cfa99d

[CI/Build] Fix broken build on Apple M1 (#28999)

j20120307 committed 22 days ago

Verified bbc6c2f1

refactor(cpu_types_scalar.hpp): Unify scalar loop implementations using unroll_loop (#28847)

ihb2032 committed 22 days ago

Verified 81516095

[Docs] Clean up moe_kernel_features.md (#28530)

windsonsea committed 22 days ago

Verified fdf93486

add support for --fully-sharded-loras in fused_moe (#28761)

gnovack committed 22 days ago

Verified d69062c6

Add CPU support model (#28697)

louie-tsai committed 22 days ago

Verified ae4821a1

[Doc]: fix typos in various files (#28945)

didier-durand committed 22 days ago

Verified 7ed27f3c

Speed up macOS smoke test (#28954)

mgoin committed 22 days ago

Verified a4511e38

[Misc] Update embedding/cross encoder tests to use `mteb` v2 (#27329)

Samoed committed 22 days ago

Verified 71d0ae1c

[Model][QwenVL] Simplify cos/sin rotary embedding indexing (#28962)

lgeiger committed 22 days ago

Verified 3d4e7d34

[Docs] Update oneshot imports (#28188)

UranusSeven committed 22 days ago

Verified 6a25ea5f

[Bugfix] Fix typo in Qwen3 Next model executor (#28960)

Nepherpitou committed 22 days ago

Verified 73ff872d

[Bugfix] Fix FusedMoEModularKernel for triton backend (#28913)

xyang16 committed 22 days ago

Verified 468a8d72

[Attention] FlashAttention ViT support, make default backend (#28763)

MatthewBonanni committed 22 days ago

Verified 4c23690f

Enable bitsandbytes quantization on AMD GPUs that use warp size 32 (#27307)

sstamenk committed 22 days ago

Verified 814843e0

[CPU] Refactor CPU WNA16 (#28826)

bigPYJ1151 committed 22 days ago

Verified 20852c8f

[Core] Switch Flat logprob control from environment variable to SamplingParams (#28914)

Jialin committed 22 days ago

Verified 40b6b38f

Move online quantization to `model.load_weights` (#26327)

jerryzh168 committed 22 days ago

Verified da94c7c0

[Hybrid][torch.compile] Refactor mamba2 forward to avoid obscuring linear projections under custom op (#28587)

tomeras91 committed 22 days ago

Verified 1395461f

[Build] Add OpenAI triton_kernels (#28788)

varun-sundar-rabindranath committed 22 days ago

Verified 9912b8cc

[NVIDIA] Guard SM100 CUTLASS MoE macro to SM100 builds v2 (#28938)

johnnynunez committed 22 days ago

Verified 49ef847a

Supress verbose logs from model_hosting_container_standards (#28949)

mgoin committed 22 days ago

Verified 67745d18

Replace `torch.cuda.Event` with `torch.Event` for better hardware compatibility (#26985)

jikunshang committed 22 days ago

Verified 2a2d5d27

[NIXL] fix cpu PD after physical <> logical block_size PR (#28904)

xuechendi committed 22 days ago

Verified c3e29786

Newer Older