generative-models
generative-models copied to clipboard
train SVD XT
Training SVD requires about 70G of GPU memory under the conditions of 576*1024 resolution and a batch size of 1. How much GPU memory is needed to train SVD XT?
How did you train SVD?
How did you train SVD?
How did you train SVD?
I write a SVD training script here: https://github.com/alibaba/animate-anything?tab=readme-ov-file#finetune-stable-video-diffusion