[Meta] Llama4 EAGLE Support (#20591)

zhiweiz committed over 1 year ago

[CI] update typos config for CI pre-commit and fix some spells (#20919)

Peter Pan committed over 1 year ago

Avoid direct comparison of floating point numbers (#21002)

Maximilien de Bayser committed over 1 year ago

[Voxtral] Add more tests (#21010)

Patrick von Platen committed over 1 year ago

[Doc] Remove duplicate docstring (#21012)

Wentao Ye committed over 1 year ago

[Bugfix] Fix Mistral3 support on SM100/SM120 (#20998)

Michael Goin committed over 1 year ago

[CI][HPU] update for v0 deprecate by switching to VLLM_TARGET_DEVICE=empty (#21006)

Chendi.Xue committed over 1 year ago

Add Dockerfile argument for VLLM_USE_PRECOMPILED environment (#20943)

Doug Smith committed over 1 year ago

[Bugfix] Correct per_act_token in CompressedTensorsW8A8Fp8MoECutlassM… (#20937)

Ming Yang committed over 1 year ago

[Docs] Enhance Anyscale documentation, add quickstart links for vLLM (#21018)

Ricardo Decal committed over 1 year ago

[Misc] Refactor: Improve argument handling for `conda` command (#20481)

Reid committed over 1 year ago

[Chore] Remove outdated transformers check (#20989)

Brayden Zhong committed over 1 year ago

[Misc] bump xgrammar version to v0.1.21 (#20992)

Chauncey committed over 1 year ago

[Model] Add ModelConfig class for GraniteMoeHybrid to override default max_seq_len_to_capture (#20923)

Thomas Parnell committed over 1 year ago

[Frontend] Remove print left in FrontendArgs.add_cli_args (#21004)

Michael Goin committed over 1 year ago

[Frontend] OpenAI Responses API supports input image (#20975)

Chauncey committed over 1 year ago

[Nvidia] Integrate SM100 cudnn prefill API to MLA prefill (#20411)

Elfie Guo committed over 1 year ago

[Bug Fix] get_distributed_init_method should get the ip from get_ip i… (#20889)

Chen LI committed over 1 year ago

[Frontend] Support cache_salt in /v1/completions and /v1/responses (#20981)

Marko Rosenmueller committed over 1 year ago

[BugFix] fix 3 issues: (1) using metadata for causal-conv1d, (2) indexing overflow in v1 vLLM, and (3) init_states in v0 (#20838)

Tuan, Hoang-Trong committed over 1 year ago