fudan-generative-vision / hallo

Hallo: Hierarchical Audio-Driven Visual Synthesis for Portrait Image Animation
https://fudan-generative-vision.github.io/hallo/
MIT License
8.89k stars 1.2k forks source link

Use Case Example - Great job on this devs! #57

Open Omarch47 opened 2 months ago

Omarch47 commented 2 months ago

I wanted to share a quick demonstration of a home use case for this that I made - though it is a bit silly: https://youtu.be/Q6PgKZfQeK8

I am running that on a 3090ti and Ubuntu 2022.04. As a side question, would there be a way to get this to utilize my second 3090ti as well? Great job on this project and thanks for sharing it openly!

xumingw commented 2 months ago

Currently we don’t support inference one case with multiple gpus. But you can set CUDA_VISIBLE_DEVICES to run multi different inference simultaneously try CUDA_VISIBLE_DEVICES=1 python -m scripts.inference .....

Omarch47 commented 2 months ago

Currently we don’t support inference one case with multiple gpus. But you can set CUDA_VISIBLE_DEVICES to run multi different inference simultaneously try CUDA_VISIBLE_DEVICES=1 python -m scripts.inference .....

Thanks very much for the tip! That would be awesome to utilize both cards for different generations!

zhjygit commented 2 months ago

一张显卡8G,带不动怎么搞呀?能否支持多卡的分布式。