fastai
fast.aiHigh-level deep-learning library on PyTorch from fast.ai, built for training strong models in a few lines. Its Learner API covers image classification and segmentation, text, tabular data and recommendation, with transfer learning from pretrained torchvision and timm backbones, a learning-rate finder, one-cycle training and callbacks, and it underpins fast.ai's free deep-learning course.
Openness
5 high confidence- license
- Apache-2.0(OSI)
- source
- public(github.com/fastai/fastai)
- core features withheld
- no — published by fast.ai, a research institute
Apache-2.0 and built from the public repository. It is published by fast.ai, a research institute that sells no edition of it.
- https://raw.githubusercontent.com/fastai/fastai/HEAD/LICENSE recorded 2026-09-27
LICENSE is the Apache License, Version 2.0.
- https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/fastai%2Ffastai recorded 2026-09-27
Repository record for fastai/fastai, not archived and not a fork, license apache-2.0.
- https://ungh.cc/repos/fastai/fastai/files/main recorded 2026-09-27
Repository tree of 257 paths with no ee, enterprise, commercial or pro directory.
- https://www.fast.ai/about.html recorded 2026-09-27
About page: fast.ai is "a research institute dedicated to making deep learning more accessible"; no paid tier or commercial edition of the library.
Adoption
3 high confidenceMeasured on monthly PyPI downloads of the fastai package.
- https://pypistats.org/api/packages/fastai/recent recorded 2026-09-27
last_month 326,098 downloads of fastai
Capability
3 high confidencefastai makes it quick to fine-tune a pretrained network on the user's own labeled data across several domains, and ships the training recipes to do it well. The user still picks the architecture and supplies the training run, as with timm, so it sits with the toolkits.
- https://docs.fast.ai/vision.learner.html recorded 2026-09-27
vision.learner: "vision_learner and unet_learner ... will help you define a Learner using a pretrained model."
- https://raw.githubusercontent.com/fastai/fastai/HEAD/README.md recorded 2026-09-27
README: "fastai is a deep learning library which provides practitioners with high-level components that can quickly and easily provide state-of-the-art results in standard deep learning domains"; Quick Start builds "an image classifier, an image segmentation model, a text sentiment model, a recommendation system, and a tabular model".
Verified 2026-09-27