Metadata-Version: 2.4
Name: yaunet
Version: 0.1.2
Summary: A simple, modular U-Net implementation in PyTorch.
Author-email: "Simon J. Larsen" <simonhffh@gmail.com>
Maintainer-email: "Simon J. Larsen" <simonhffh@gmail.com>
License-Expression: MIT
Project-URL: Documentation, https://yaunet.readthedocs.io
Project-URL: Repository, https://github.com/SimonLarsen/yaunet.git
Project-URL: Issues, https://github.com/SimonLarsen/yaunet/issues
Keywords: pytorch,u-net
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: numpy>=2.2.6
Requires-Dist: einops>=0.8.2
Requires-Dist: torch>=2.0.0
Dynamic: license-file

Yet Another U-Net
=================

yaunet (*ya' you net*) is a simple, modular U-Net implementation in PyTorch.

Features:
* Simple implementation only dependent on [PyTorch](https://pytorch.org) and [einops](https://github.com/arogozhnikov/einops).
* Aims to be small, modular and highly configurable.
* Provides a decent number of commonly used building blocks.
* All blocks support AdaLN-Zero-like conditioning for training conditional models e.g. class- and/or time-conditioned diffusion models.

## Installation

```sh
pip install yaunet  # or `uv add yaunet`
```

## Documentation

Read the API documentation at [yaunet.readthedocs.io](https://yaunet.readthedocs.io).
