NerdyRodent

AniPortrait

Added 2024-03-30 22:28:01 +0000 UTC

AniPortrait is a fun new python app that can be used to... animate portraits :) A few commands are missing from the current documentation, so here they are in full!

== Setup / Install Command Summary ==

git clone https://github.com/Zejun-Yang/AniPortrait.git

cd AniPortrait

conda create --name aniportrait python=3.11

conda activate aniportrait

pip install -r requirements.txt

conda install ffmpeg -c conda-forge

== Download weights ==

You can use "huggingface-cli", "git clone" or the download links to get all the models you need from huggingface.

* Example downloads using huggingface-cli:

huggingface-cli download --resume-download ZJYang/AniPortrait --local-dir pretrained_model

huggingface-cli download --resume-download bdsqlsz/image_encoder --local-dir pretrained_model/image_encoder

huggingface-cli download --resume-download facebook/wav2vec2-base-960h --local-dir pretrained_model/wav2vec2-base-960h

huggingface-cli download --resume-download stabilityai/sd-vae-ft-mse --local-dir pretrained_model/sd-vae-ft-mse

huggingface-cli download --resume-download runwayml/stable-diffusion-v1-5 --revision fp16 --local-dir pretrained_model/stable-diffusion-v1-5

By default, the huggingface-cli command will download files to the cache directory defined by your HF_HOME environment variable value, but you can also specify a custom cache using --cache-dir if you like.

If pytorch is running in CPU mode, make sure you've installed a CUDA accelerated version. The table of install commands is available at https://pytorch.org/

For Microsoft Windows there is also a fork which has some basic changes -

https://github.com/sdbds/AniPortrait-for-windows

You can also apply their PR:

git fetch origin pull/13/head:testing

git checkout testing

ComfyUI verion - https://github.com/chaojie/ComfyUI-AniPortrait

RVC WebUI (Voice Cloning) - https://youtu.be/g4td-n8-QZk

Have fun! :)