microsoft/Tutel · AI Repo StatsMImicrosoft/Tutel
Tutel MoE: Optimized Mixture-of-Experts Library, Support GLM/Kimi/Gpt/DeepSeek/Qwen using FP8/NVFP4/MXFP4
LLM & foundation modelsC MIT 111created 2021-08-06 #deepseek#llm#mixture-of-experts#moe#pytorch