#1414th of 15 in ML frameworks and libraries
fastai
High-level deep learning library on PyTorch for vision, text, tabular and recommenders
Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 28.2k
- License
- Apache-2.0
- Last commit
- Sep 2026
- Last release
- Sep 2026
- Language
- Jupyter Notebook
27,769 on 24 Mar 202628,226 on 13 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
fastai is a Python deep learning library built on PyTorch. It offers high-level APIs that train image classifiers, segmentation, text sentiment, tabular and recommendation models in about 5 lines, on top of layered lower-level components for researchers. Core pieces include a data block API, a two-way callback system and an optimizer refactored into small composable parts.
Who it is for: Python practitioners and researchers training deep learning models with PyTorch
Strengths
- Layered API: 5-line high-level training down to low-level building blocks
- Two-way callbacks can modify data, model or optimizer at any point in training
- Docs are runnable notebooks, with a free course and a book
- Migration guides for plain PyTorch, Ignite, Lightning and Catalyst
Weaknesses
- Jupyter on Windows forces num_workers to 0, making vision training much slower
- Covers classic deep learning domains; README does not mention LLM or diffusion workflows
- README advises installing PyTorch separately before fastai
- GPU memory and RAM requirements are not stated
What it needs
- GPU optional
- Compose
- Needs PyTorch, fastcore
- Models: PyTorch
Also in ML frameworks and libraries
See all 15| Rank | Project | Score |
|---|---|---|
| 1 | transformersModel-definition library for text, vision, audio and multimodal models | 94 out of 100 |
| 2 | pytorchPython tensor library with GPU support and tape-based autograd for deep learning | 92 out of 100 |
| 3 | ultralyticsPython library and CLI for training and running YOLO vision models | 86 out of 100 |
| #4 | pytorch-image-modelsPyTorch library of image models, pretrained weights and reference training scripts | 82 out of 100 |
| #5 | tensorflowTensor and neural network framework with Python and C++ APIs | 80 out of 100 |