Skip to content

Latest commit

 

History

History
54 lines (36 loc) · 2.04 KB

README.md

File metadata and controls

54 lines (36 loc) · 2.04 KB
Raw

Meta-Unlearning on Diffusion Models: Preventing Relearning Unlearned Concepts

This repository is the official implementation for the paper: Meta-Unlearning on Diffusion Models:Preventing Relearning Unlearned Concepts

Installation

We follow the Diffuser to install the required dependencies, please run the following commands:

conda create -n meta python=3.10
conda activate meta
pip install --upgrade diffusers[torch]

All experiments are conducted on NVIDIA A100 GPUs with 80GB of memory.

Generate images datasets

First you can use the shell scripts to generate hrm dataset, irt dataset and target dataset. In our code, the hrm dataset contains the unlearned concept, while both the irt and target dataset contain retain concepts. irt dataset means the concept unrelated to unlearned concept, and the target dataset means the concept related to unlearned concept.

Here is the exmaple generate scipt, you can change the text prompt according to your goal.

bash scripts/gen_image.sh

Baseline: unlearning

bash scripts/esd_baseline.sh
bash scripts/sdd_baseline.sh

UCE and RECE can be trained by the original code in their paper.

Our paradigm: meta-unlearning

For ESD and SDD based meta-unlearning:

bash scripts/esd_meta.sh
bash scripts/sdd_meta.sh

Note that we only give the example hyperparameter in code and you should change the hyperparameter refer to our paper (We use fix_timesteps = False in our paper, and the fixed_timesteps = True is just used for accelerating training) .

For UCE and RECE based meta-unlearning:

bash scripts/meta.sh

Note that you should change the model path to your unlearned model.

Acknowledgments

This repository is based on the codebase of the SDD. Thanks for their impressive works!