A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 05:30 UTC
main
[NIXL] Add support for MLA caches with different latent dim (#25902)
Nicolò Lucchesi committed about 1 year ago
[perf] Use CPU tensor to reduce GPU->CPU sync (#25884)
Lehua Ding committed about 1 year ago
[Model] Move `vision_feature_select_strategy` into `resolve_visual_encoder_outputs` (#25938)
Cyrus Leung committed about 1 year ago
[Bugfix][Model]fix ernie45 moe gate&bias dtype to float32 (#25936)
CSWYF3634076 committed about 1 year ago
Updated TRL integration docs (#25684)
Sergio Paniego Blanco committed about 1 year ago
[Doc] Add Cambricon MLU support (#25942)
a120092009 committed about 1 year ago
[New Model] DeepSeek-V3.2 (Rebased to Main) (#25896)
Yongye Zhu committed about 1 year ago
[Bugfix]: Clean up chunked prefill logging when using whisper (#25075)
Simon Danielsson committed about 1 year ago
[Model][Bugfix] Fix MiDashengLM audio encoder mask by removing incorrect `logical_not` (#25925)
Zhou Jiahao committed about 1 year ago
[BugFix] Pass config_format via try_get_generation_config (#25912)
acisseJZhong committed about 1 year ago
[BugFix] Fix DP/EP hang (#25906)
Lucas Wilkinson committed about 1 year ago
Move`VllmConfig` from `config/__init__.py` to `config/vllm.py` (#25271)
Harry Mellor committed about 1 year ago
[Benchmark] Support benchmark throughput for external launcher DP (#25913)
Zhuohan Li committed about 1 year ago
[Bug] Fix Weight Loading for Block FP8 Cutlass SM90 (#25909)
Wentao Ye committed about 1 year ago
Test Prompt Embeds/LoRA compatibility and Enable LoRA Support for OPT Models (#25717)
Andrew Sansom committed about 1 year ago
[V0 Deprecation] Remove `vllm.worker` and update according imports (#25901)
Aaron Pham committed about 1 year ago
[NIXL] Increase default KV block eviction timeout on P (#25897)
Nicolò Lucchesi committed about 1 year ago
[Doc] Polish example for torchrun dp (#25899)
Zhuohan Li committed about 1 year ago
[Kernel] Chunk-aligned mamba2 (#24683)
Thomas Parnell committed about 1 year ago
[Bugfix][ROCm] Fixing trying to import non-existent symbols from libnccl.so (#25605)
Gregory Shtrasberg committed about 1 year ago