Explore live open-source projects and AI models.

Build a modern LLM from scratch. Every line commented. Explained like we are five.

Demystify RAG by building it from scratch. Local LLMs, no black boxes - real understanding of embeddings, vector search, retrieval, and context-augmented generation.

From-scratch PyTorch implementation of Google's TurboQuant (ICLR 2026) for LLM KV cache compression. 5x compression at 3-bit with 99.5% attention fidelity.

No description available from source.

[ICLR 2025] Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing. Your efficient and high-quality synthetic data generation pipeline!

An LLM-powered advanced RAG pipeline built from scratch

Train a 1B LLM with 1T tokens from scratch by personal

Single File, Single GPU, From Scratch, Efficient, Full Parameter Tuning library for "RL for LLMs"

Building LLM-Enabled Multi Agent Applications from Scratch

从0到1构建一个MiniLLM (pretrain+sft+dpo实践中)