[Perf] Use FlashInfer RoPE for RotaryEmbedding.forward_cuda when available (#21126)

Michael Goin committed about 1 year ago

Specify platform in `pip-compile` `pre-commit` hook so it runs on MacOS (#25273)

Harry Mellor committed about 1 year ago

[Bugfix] Fix chunked a2_scales in modular kernels (#25264)

bnellnm committed about 1 year ago

[Bugfix] GPT OSS Attritbute error on H100 (#25228)

Varun Sundar Rabindranath committed about 1 year ago

Optimize triton unified attention performance for sliding window attention (#24390)

qizixi committed about 1 year ago

[KV offload][4/N] Offloading KV connector (#22595)

Or Ozeri committed about 1 year ago

[bugfix] fix structured outputs key missing issue from #24929 (#25195)

Lucia Fang committed about 1 year ago

[Docs] add __init__.py to vllm/model_executor/layers/quantization/compressed_tensors/transform (#24974)

samzong committed about 1 year ago

[Core][Prefix Hash] Fix prefix hash metrics sliding window maintainance (#24990)

Jialin Ouyang committed about 1 year ago

[CI/Build] fix test function_calling (#25072)

Chauncey committed about 1 year ago

Encoder model support for the Transformers backend (#25174)

Harry Mellor committed about 1 year ago

Remove Redundant Assignment in Qwen3_VisionPatchMerger (#25224)

LJH-LBJ committed about 1 year ago

[Core] Modify the initialization parameters of the lora manager (#25249)

Jee Jee Li committed about 1 year ago

[Frontend] Pass API server count to each process (#23717)

Cyrus Leung committed about 1 year ago

[KV offload][3/N] Add worker-side CPU support (#21448)

Or Ozeri committed about 1 year ago

refactor(benchmarks): add type annotations to wait_for_endpoint parameters (#25218)

samzong committed about 1 year ago

Move `ModelConfig` from `config/__init__.py` to `config/model.py` (#25252)

Harry Mellor committed about 1 year ago

Update CODEOWNERS (#25269)

Harry Mellor committed about 1 year ago

[Qwen] Remove cuda hard-code in qwen3 next (#25243)

Icey committed about 1 year ago

[V0 Deprecation] Remove V0 logic from `get_input_embeddings` interface (#25242)

Cyrus Leung committed about 1 year ago