[Docs] Improve `priority` parameter documentation (#29572)

maang-h committed 12 months ago

[Frontend][torch.compile] CompilationConfig Overhaul (#20283): Set up -O infrastructure (#26847)

Morrison Turnansky committed 12 months ago

[Bugfix] Update Ultravox compatibility (#29588)

Cyrus Leung committed 12 months ago

[Model Runner V2] Implement multi-step Eagle with CUDA graph (#29559)

Woosuk Kwon committed 12 months ago

[ROCm][CI] Fix test_cpu_offloading for ROCm (#29548)

Micah Williamson committed 12 months ago

Fix tpu-inference platform path (#29554)

Johnny Yang committed 12 months ago

[Bugfix] Fix getting device for MoE LoRA (#29475)

Jee Jee Li committed 12 months ago

[Model Runner V2] Refactor CudaGraphManager (#29583)

Woosuk Kwon committed 12 months ago

[cpu][fix] Fix Arm CI tests (#29552)

Fadi Arafeh committed 12 months ago

[DOC] Add vLLM Bangkok Meetup info (#29561)

TJian committed 12 months ago

[Model Runner V2] Minor cleanup for build_attn_metadata (#29576)

Woosuk Kwon committed 12 months ago

[Model Runner V2] Minor code cleanup (#29570)

Woosuk Kwon committed 12 months ago

add xpu supported model and model id for cpu (#29380)

Louie Tsai committed 12 months ago

[Bugfix] fix IMA issue in certain cases of the moe marlin kernel (#28619)

Jinzhen Lin committed 12 months ago

[Attention][Async] Eliminate `seq_lens_cpu` in FlashAttention metadata building with DCP > 1 (#29449)

Matthew Bonanni committed 12 months ago

[Bugfix] Make compressed-tensors MoEs respect ignored layers (#28878)

HDCharles committed 12 months ago

[TPU] add tpu_inference (#27277)

Johnny Yang committed 12 months ago

[Core] Refactor padding logic and pad for CUDA graphs before attention metadata building (#28579)

Lucas Wilkinson committed 12 months ago

[Attention] Remove imports from `vllm/attention/__init__.py` (#29342)

Matthew Bonanni committed 12 months ago

[CI/Build] allow user modify pplx and deepep ref by ENV or command line (#29131)

Alec committed 12 months ago