fastai and fastbook
The library provides the implementation of deep learning techniques while the book provides jupyter notebook tutorials and explanations that teach how to use that library, making them complements designed to be used together.
About fastai
fastai/fastai
The fastai deep learning library
Built on PyTorch, fastai implements a layered architecture with high-level APIs for rapid prototyping across vision, NLP, and tabular domains, while exposing low-level composable abstractions for research. Key innovations include a type dispatch system for Python, a GPU-optimized computer vision library, a refactored optimizer framework, and a 2-way callback system enabling fine-grained control over training dynamics. It integrates seamlessly with existing PyTorch workflows and provides migration paths from libraries like Lightning and Ignite.
About fastbook
fastai/fastbook
The fastai book, published as Jupyter Notebooks
Scores updated daily from GitHub, PyPI, and npm data. How scores work