Official Implementation of understanding the latent space of diffusion models through the lens of riemannian geometry (NeurIPS 2023)
- examples
weekly-supervised editing (text-conditioned)
conda create -n pullback python=3.10
pip install -r requirements.txt
- unconditional diffusion model
Currently, it's implemented only for CelebA-HQ.
In the paper, we used the mode from SDEdit, but since it's deprecated, here, we use a Hugging Face model here.
cd src
bash scripts/main_celeba_hf_local_encoder_pullback.sh
- stable diffusion
If you want to experiment with a new image, you can do the following: 1) Insert the desired image into "datasets/examples", and 2) adjust the "sample_idx" accordingly.
cd src
bash scripts/main_various_local_encoder_pullback_without_edit_prompt.sh # w/o text condition
bash scripts/main_various_local_encoder_pullback_with_edit_prompt.sh # w/ text condition