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 11, 20:00 UTC
main
online fp8 quant with streaming weight post-processing (#29196)
Vasiliy Kuznetsov committed 11 months ago
[TPU] Bump tpu-inference to 0.12.0 (#30221)
Johnny Yang committed 11 months ago
Add latent MoE support (#30203)
shaharmor98 committed 11 months ago
[MP executor] fix get device count for multi node of mp executor feature (#30042)
weiguihua2 committed 11 months ago
[Misc] Split the LoRA code (#30253)
Jee Jee Li committed 11 months ago
Add evaluate_guards option to DynamicShapesConfig (#27432)
Laith Sakka committed 11 months ago
[DeepSeek v3.2] Make top-k work for any logit values. (#27568)
Daniel Cámpora committed 11 months ago
[ROCm] Guard group quant RMS norm fusion patterns (#30239)
Ye (Charlotte) Qi committed 11 months ago
[LoRA] Reduce the loading time of MoE LoRA (#30243)
Jee Jee Li committed 11 months ago
Add tip for `mypy` and `markdownlint` to the pre-commit comment (#30259)
Harry Mellor committed 11 months ago
[docs] governance documents (#24801)
Simon Mo committed 11 months ago
wang.yuqi committed 11 months ago
[CI] Prevents triggering of an inactive issue/PR check for forked repository. (#29654)
Shiming Zhang committed 11 months ago
wang.yuqi committed 11 months ago
[Perf] Remove sync point in vit torch sdpa attn backend (#30232)
Dazhi Jiang committed 11 months ago
[Misc] Rename TensorRT Model Optimizer to Model Optimizer (#30091)
Zhiyu committed 11 months ago
[Bugfix] fix fuse_allreduce_rms when tp =1 (#30178)
Jiangyun Zhu committed 11 months ago
[ROCm] [Fused Moe EP] Use binary expert mask for aiter fused moe kernel (#29773)
Zhiwei committed 11 months ago
[BugFix] Unblock use of LoRA with data parallel mode (#30220)
Nick Hill committed 11 months ago
Address comment to mergify.yml in #30117 (#30219)
Zhijian Jiang committed 11 months ago