Metadata-Version: 2.4
Name: fourtune
Version: 0.0.1
Summary: FourTune: fully 4-bit LoRA training and inference for diffusion models (placeholder, release coming soon)
Author-email: Bowen Xue <kbrask@ustc.edu>
License-Expression: Apache-2.0
Project-URL: Paper, https://arxiv.org/abs/2607.05711
Project-URL: Models, https://huggingface.co/FourTune
Classifier: Development Status :: 1 - Planning
Classifier: Programming Language :: Python :: 3
Requires-Python: >=3.10
Description-Content-Type: text/markdown

# FourTune

Fully 4-bit LoRA training and inference for diffusion models: weights, activations and gradients in 4 bits.

This is a placeholder release that reserves the package name. It contains no code yet. The first release is coming soon.

- Paper: [FourTune: Towards Fully 4-Bit Efficient Post-Training for Diffusion Models](https://arxiv.org/abs/2607.05711)
- Models: [huggingface.co/FourTune](https://huggingface.co/FourTune)
