ML-Notebooks

Why we included this project

This is a solid teaching collection for anyone who wants to see how machine learning really works. The notebooks start with computational graphs and a PyTorch hello world, then move through linear regression built from scratch with gradient descent, and on to transformers, GNNs, GANs, and LoRA/QLoRA fine-tuning of models like TinyLlama and Mistral. Each notebook is deliberately minimal and self-contained, so you can open it in Colab or a Codespace, run it end to end, and tweak it for your own experiments without wading through a large codebase. If you need a quick refresher or a way to ramp up on both classic algorithms and modern fine-tuning techniques, having clear runnable examples side by side is hard to beat.