HuggingPapers· @HuggingPapers · X·· 2 天前AI 评分
腾讯发布STEPQuant量化框架
腾讯研究人员刚刚在 Hugging Face 上发布了 STEPQuant。 一个面向 Delta-rule 循环状态的时空后训练量化框架，用 6-bit 状态即可匹配 FP32 精度，同时将总服务内存削减最多 68.7%。
Tencent researchers just released STEPQuant on Hugging Face
A spatial-temporal post-training quantization framework for Delta-rule recurrent states, matching FP32 accuracy with 6-bit states while cutting total serving memory by up to 68.7%
来源：HuggingPapers · x.com