BloodAxe/pytorch-toolbelt

PyTorch extensions for fast R&D prototyping and Kaggle farming

56
/ 100
Established

This library offers pre-built components and utilities for researchers and machine learning practitioners focused on deep learning with PyTorch. It helps quickly assemble neural network models, particularly for image analysis tasks like segmentation and classification, by providing modular encoders, decoders, and specialized loss functions. It's designed for individuals developing and experimenting with computer vision models, especially those participating in machine learning competitions.

1,568 stars. No commits in the last 6 months. Available on PyPI.

Use this if you are a machine learning engineer or data scientist working with PyTorch and need to rapidly prototype computer vision models, utilize advanced loss functions, or efficiently perform inference on large images.

Not ideal if you are looking for a complete high-level deep learning framework like Keras or fast.ai, as this library is intended to complement existing PyTorch workflows rather than replace them.

computer-vision image-segmentation image-classification deep-learning-research machine-learning-competitions
Stale 6m
Maintenance 2 / 25
Adoption 10 / 25
Maturity 25 / 25
Community 19 / 25

How are scores calculated?

Stars

1,568

Forks

127

Language

Python

License

MIT

Last pushed

Oct 09, 2025

Commits (30d)

0

Dependencies

4

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/BloodAxe/pytorch-toolbelt"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.