Follow the 2017 fast.ai course notebooks to learn practical deep learning from scratch as a working programmer.
Run Jupyter notebooks to experiment with image classification, NLP, and other deep learning tasks hands-on.
Use the fast.ai library examples to understand how to train neural networks with less boilerplate than raw PyTorch.
| fastai/courses | jeffheaton/t81_558_deep_learning | chaoningzhang/mobilesam | |
|---|---|---|---|
| Stars | 5,735 | 5,744 | 5,752 |
| Language | Jupyter Notebook | Jupyter Notebook | Jupyter Notebook |
| Setup difficulty | moderate | easy | moderate |
| Complexity | 2/5 | 3/5 | 3/5 |
| Audience | data | researcher | researcher |
Figures from each repo's GitHub metadata at analysis time.
Training exercises are impractical without a GPU, use a cloud GPU service like Kaggle or Paperspace.
This repository contains the lecture materials from the 2017 edition of Practical Deep Learning for Coders, a course offered by fast.ai. The course is aimed at working programmers who want to learn how to build and train machine learning models, with an emphasis on getting hands-on results quickly rather than starting from mathematical theory. The materials are stored as Jupyter notebooks, which are documents that combine written explanations with executable code in the same file. Learners can open them, read through the content, and run the code examples directly to see results. This format makes it easier to follow along and experiment with changes without setting up a separate project. The README is brief and mostly directs students to the course website, the community forums, and the wiki for support. The forums and wiki are described as the main resources for getting help when something goes wrong, with the advice that most common questions have already been answered there. GitHub Issues on this repository are not the right place for debugging questions, according to the README. Fast.ai has since released updated course versions, so this repository represents an older iteration of their curriculum. The content covers practical deep learning techniques using the fast.ai library, which wraps around PyTorch to make training models more accessible. Students interested in the current course materials would typically look at fast.ai's more recent repositories rather than this one.
The 2017 fast.ai Practical Deep Learning for Coders course, Jupyter notebooks teaching working programmers to build and train deep learning models using the fast.ai library on top of PyTorch, with a hands-on-first approach.
Mainly Jupyter Notebook. The stack also includes Python, Jupyter Notebook, PyTorch.
Setup difficulty is rated moderate, with roughly 1h+ to a first successful run.
Mainly data.
This repo across BitVibe Labs
Verify against the repo before relying on details.