从 MiniMind 源码读起,再延伸到现代大模型技术体系的中文学习笔记。主线逐行精读预训练 / SFT / DPO / PPO / GRPO 与训练机制;附录 17 篇进阶卷覆盖量化、投机解码、RLHF 全景、模型代际史等 MiniMind 没涉及、但进阶绕不开的主题。

🚀 [从零构建 LLM] 极简大模型训练原理与实践指南。包含 Transformer, Pretraining, SFT 核心代码与对照实验。 | A minimal, principle-first guide to understanding and building LLMs from scratch.

The Cowork Agent for Everything — trainable advertising AI + 14 platform MCP servers + agent skills. Based on minimind (42k stars). Train from zero in 2 hours.

A beginner-friendly VLA project starter: CPU smoke, ACT + PushT-style imitation learning, rollout eval, and resume-ready docs.

练习时长两年半的 AI 大模型 (实际 26M params，2.5B = 两年半) | ikun meme-culture chatbot 🐔🏀

知识蒸馏 | Knowledge Distillation from teacher model 🎓

从零预训练语言模型 | Pretrain LLM from scratch 📚

混合专家模型 | Mixture of Experts architecture 🧩

偏好对齐训练 | Direct Preference Optimization 👍👎

推理模型训练 | Reasoning with chain-of-thought 🧠

从零复现 26M 中文小语言模型 MiniMind｜预训练→SFT→LoRA→DPO→蒸馏→RLHF 全流程学习复现（基于 jingyaogong/minimind）

模型部署指南 | Deploy to API/Web/ollama/vllm 🚀

从零训练分词器 | BPE tokenizer from scratch 🔤

强化学习对齐 | Group Relative Policy Optimization 🎮

从0到1手写的中文小型大语言模型 — minimind 风格教学项目：RMSNorm·RoPE·GQA·SwiGLU，完整 tokenizer→pretrain→SFT 管线，单张 RTX 3060 可训。A from-scratch Chinese LLM for learning.

📝 Enhance your understanding of large model training with detailed, clear annotations for the MiniMind project, tailored for Chinese learners.

多模态视觉语言模型 | Vision-Language Model 👁️