datawhalechina/tiny-universe
《大模型白盒子构建指南》：一个全手搓的Tiny-Universe
This project offers a comprehensive, 'from-scratch' guide to building large language model (LLM) systems. It starts with foundational principles and walks you through constructing a complete LLM ecosystem, including the core model, Retrieval-Augmented Generation (RAG) framework, Agent system, and evaluation tools. The target audience is deep learning practitioners who have some experience with LLMs and want to deepen their understanding by building systems from the ground up.
4,598 stars.
Use this if you are an AI/ML practitioner with basic LLM application knowledge and want to understand the underlying principles and build your own custom LLM components without relying on high-level frameworks.
Not ideal if you are looking for a quick way to integrate pre-built LLMs or frameworks into an application, as this project focuses on foundational understanding and manual implementation.
Stars
4,598
Forks
450
Language
Jupyter Notebook
License
—
Category
Last pushed
Feb 12, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/datawhalechina/tiny-universe"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
WangRongsheng/awesome-LLM-resources
🧑🚀 全世界最好的LLM资料总结（多模态生成、Agent、辅助编程、AI审稿、数据处理、模型训练、模型推理、o1 模型、MCP、小语言模型、视觉语言模型） | Summary of the...
SylphAI-Inc/AdalFlow
AdalFlow: The library to build & auto-optimize LLM applications.
LazyAGI/LazyLLM
Easiest and laziest way for building multi-agent LLMs applications.
luhengshiwo/LLMForEverybody
每个人都能看懂的大模型知识分享，LLMs春/秋招大模型面试前必看，让你和面试官侃侃而谈
katanaml/sparrow
Structured data extraction and instruction calling with ML, LLM and Vision LLM