vlms-zero-to-hero

Why we included this project

Most courses on vision-language models expect you to already know the field. This series instead builds up from the ground, pairing each landmark paper, from Word2Vec and Attention Is All You Need to BERT, CLIP, and LLaVA, with a PyTorch notebook you can open straight in Colab. Reading the original research and then following code that implements it makes the ideas concrete rather than abstract. The author's open-source computer vision background shows in the teaching, and because every concept ships as a runnable notebook, you can experiment instead of just reading. For developers and students who want genuine intuition about how image and text representations combine, this is a coherent path through the field's milestones, not a scattered collection of tutorials.