MiniMind
Train a tiny language model from scratch on regular hardware
Build a ChatGPT-like LLM in PyTorch step by step
LLMs From Scratch is a hands-on book-style repository that teaches how to implement a ChatGPT-like large language model in PyTorch from the ground up, covering tokenization, attention, training and fine-tuning. It is for developers who want to understand and build LLMs rather than just call APIs.
Jupyter notebooks walk through each stage: preparing text data, implementing attention and the transformer block, pretraining, then fine-tuning for classification and instruction following. You run the notebooks sequentially, optionally scaling to GPU, and end with a small working LLM.
16 Cool GitHub Repos You WILL Use (no pressure)
Train a tiny language model from scratch on regular hardware
Free 12-week AI curriculum with runnable notebooks
Learn to build, deploy and iterate production ML apps
A pretend trading firm staffed entirely by AI agents
Multi-agent coding teams with roles, models and memory
LLM-powered multi-market stock analysis with dashboards and alerts