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 8, 22:45 UTC
main
Michael Goin committed 9 months ago
[Voxtral Realtime] Change name (#33716)
Patrick von Platen committed 9 months ago
[MISC] Fix Tensor Parallelism for Quantized Mamba Models with n_groups=1 (#33257)
Vadim Gimpelson committed 9 months ago
Turn `@config` into a `dataclass_transform` (#31541)
Harry Mellor committed 9 months ago
[torch.compile] Significantly speed up cold start times (#33641)
Richard Zou committed 9 months ago
[Attention][FA3] Update FA3 to include new swizzle optimization (#23465)
Lucas Wilkinson committed 9 months ago
[P/D] rework mooncake connector and introduce its bootstrap server (#31034)
dtc committed 9 months ago
[Bugfix] Fix startup hang for Granite Speech (#33699)
Cyrus Leung committed 9 months ago
[Voxtral models] Skip warm-up to skip confusing error message in warm-up (#33576)
Patrick von Platen committed 9 months ago
[MM] Pass `prefix` parameter to MMEncoderAttention (#33674)
Shanshan Shen committed 9 months ago
wang.yuqi committed 9 months ago
Feat/add nemotron nano v3 tests (#33345)
shaharmor98 committed 9 months ago
Kuntai Du committed 9 months ago
Document NixlConnector backend selection via kv_connector_extra_config (#33552)
Krish Gupta committed 9 months ago
Fix Gemma3n audio encoder for Transformers v5 (#33673)
Harry Mellor committed 9 months ago
[Models] Intern-S1-Pro (#33636)
zxy committed 9 months ago
Fix Gemma3 GGUF for Transformers v5 (#33683)
Harry Mellor committed 9 months ago
Fix offline test for Transformers v5 (#33682)
Harry Mellor committed 9 months ago
[Bugfix] fix qwen3-asr response error (#33644)
Song Zhixin committed 9 months ago
[Misc] Update default image format of `encode_base64` (#33656)
Cyrus Leung committed 9 months ago