mixture-of-experts 开源项目推荐（1,226 个）

deepspeedai /DeepSpeed

DeepSpeed 是一个深度学习优化库，使分布式训练和推理变得简单、高效且有效。

FareedKhan-dev /kimi-k3-in-c

2.78万亿参数Kimi K3在8.24 GB RAM中的单个中央处理器上运行推理。便携式C99：无BLAS、无框架、无图形处理器。

bojieli /ai-infra-book

《深入理解 AI Infra：量化分析与系统设计》（李博杰 著）开源书稿：从硬件约束和模型架构出发，量化推导 LLM 推理与训练系统设计。含全书正文、PDF、配套计算工具与实验

algorithmicsuperintelligence /optillm

优化 LLM 的推理代理

mw00 /project-maya

在你自己的 GPU 上运行 GLM-5.3-Flash（321B）、DeepSeek-V4.1-Flash（748B）和 MiniMax-M3.1-Flash（428B）——一个引擎、一个仪表盘，兼容 OpenAI 和 Anthropic 的 API。

learning-at-home /hivemind

Decentralized deep learning in PyTorch. Built to train models on thousands of volunteers across the world.

dvmazur /mixtral-offloading

Run Mixtral-8x7B models in Colab or consumer desktops

PKU-YuanGroup /MoE-LLaVA

【TMM 2025🔥】面向大型视觉语言模型的混合专家模型

davidmrau /mixture-of-experts

PyTorch Re-Implementation of "The Sparsely-Gated Mixture-of-Experts Layer" by Noam Shazeer et al. https://arxiv.org/abs/1701.06538

peonist-ai /halogen-flash-server

Local inference that behaves like a frontier API: Qwen3.8-Flash-Next on AMD Strix Halo plus small NPU models, over the OpenAI and Anthropic APIs on one port.

rhymes-ai /Aria

Codebase for Aria - an Open Multimodal Native MoE

microsoft /Tutel

Tutel MoE：优化专家混合库，支持 GptOss/DeepSeek/Kimi-K2/Qwen3，使用 FP8/NVFP4/MXFP4

pjlab-sys4nlp /llama-moe

⛷️ LLaMA-MoE: Building Mixture-of-Experts from LLaMA with Continual Pre-training (EMNLP 2024)

NVIDIA /cudnn-frontend

cuDNN 前端是 NVIDIA 现代开源的 cuDNN 库入口，以及一个不断增长的高性能开源内核集合。

SMTorg /smt

SMT：代理模型工具箱

simplexai-labs /PolyStrata

Run large AI models on your own PC: Qwen3.8-Flash-Next on Strata's engine, and GLM 5.3 Flash, DeepSeek V4 Flash, LFM2.5, Qwen3.6, Ornith 1.5, Gemma 4, Tiel Coder and MiniCPM-V on a second engine written for PolyStrata. One-click install for Windows and Linux, OpenAI/Anthropic API on localhost.

lucidrains /mixture-of-experts

A Pytorch implementation of Sparsely-Gated Mixture of Experts, for massively increasing the parameter count of language models

AviSoori1x /makeMoE

受Andrej Karpathy的makemore启发，从头开始实现稀疏混合专家语言模型：）

ARahim3 /kaggle-tpu-lab

Frontier-class open models on a free Kaggle TPU v5e-8: GLM-5.3-Flash 320B MoE (~165 tok/s, our own JAX engine) and Qwen3.8-27B bf16 (~130 tok/s), 262k context, prefix caching. Works with Claude Code, Codex, opencode and pi.

drawbridge /keras-mmoe

A TensorFlow Keras implementation of "Modeling Task Relationships in Multi-task Learning with Multi-gate Mixture-of-Experts" (KDD 2018)