Zhen Li
Zhen Li
You can try sending multiple images from different faces See Fig.2 in our arxiv paper for details.
We're checking this PR. Thank you for your contribution!
Now, you could update the environment building through: ```bash conda create --name photomaker python=3.10 pip install -U pip # Install requirements pip install -r requirements.txt # Install photomaker pip install...
https://civitai.com/models/139562?modelVersionId=268861
Download this model and edit `PhotoMakerStableDiffusionXLPipeline.from_pretrained` to `PhotoMakerStableDiffusionXLPipeline.from_single_file`
https://huggingface.co/docs/diffusers/v0.26.3/en/api/loaders/single_file#diffusers.loaders.FromSingleFileMixin.from_single_file.example
Yes, of course
You can download models from https://civitai.com/models. Then, you can use `from_single_file` function to load them.
Hi, for the model reported in the paper, we used filtered 110K images to train. No in-depth exploration of changes in the scale of training datasets.
@cckuailong @EdwardOwusuAdjei Hi, any suggestions?