jingyaogong/minimind

🧠 Train a 64M-parameter LLM from scratch in just 2h!

3 best points · 1 comments · 1 Hacker News post · 64,963 GitHub stars · View on GitHub

Hacker News discussions

• MiniMind: End-to-end GPT-style LLM training pipeline in pure PyTorch (3 points, 1 comments, by dmonterocrespo, 2026-03-24)