[Bugfix][Qwen3-Next] fixes the varlen issue in qwen3-next's MTP implementation. (#24957)

Tao He committed about 1 year ago

[fix] lora benchmarks pass no_lora_flag_cpu (#23774)

dolpm committed about 1 year ago

Apply fixes for CUDA 13 (#24599)

Aidyn-A committed about 1 year ago

[Bugfix] Fix Stream usage in CPU model runner and OneDNN kernel check (#25046)

Li, Jiang committed about 1 year ago

Add RADIO Vision Encoder Support to vLLM (#24595)

danielafrimi committed about 1 year ago

[Misc] Avoid use of deprecated `AutoModelForVision2Seq` (#25065)

Cyrus Leung committed about 1 year ago

[BugFix] enable DOTALL to match multi-line tool_call parameters in extract_tool_call_required_streaming (#24668)

Shijun Yin committed about 1 year ago

[Frontend] Support returning all prompt logprobs (#24956)

Chauncey committed about 1 year ago

[DP] Create placement groups by ray_device_key (#25026)

Xinyu Chen committed about 1 year ago

[Docs] Fix griffe warning in base_static_graph.py (#25018)

Michael Yao committed about 1 year ago

[Core] Remove tokenizer group in vLLM (#24078)

Zhuohan Li committed about 1 year ago

[Model] Apply SharedFusedMoE to glm4_moe. (#24849)

whx committed about 1 year ago

[Model] Pass param prefix to LLMHead (#24862)

whx committed about 1 year ago

[Core][MultiModalHasher] Hash images without converting image mode (#24969)

Lukas Geiger committed about 1 year ago

[EPLB] Support EPLB for Mixtral Model (#22842)

rouchenzi committed about 1 year ago

[XPU] Fix xpu model runner call torch.cuda APIs (#25011)

Kunshang Ji committed about 1 year ago

Add more documentation and improve usability of lognormal dist (benchmark_serving_multi_turn) (#23255)

Daniel Serebrenik committed about 1 year ago

[Rocm] [quantization] Fix quark ptpc moe and add test case (#24649)

haoyangli-amd committed about 1 year ago

[Model] Support Qwen3-VL Model Series (#24727)

Roger Wang committed about 1 year ago

[V0 Deprecation] Remove MQLLMEngine (#25019)

Woosuk Kwon committed about 1 year ago