CogVideo
CogVideo copied to clipboard
How to finetune with multi resolution?
Is there any plan release the code of multi resolution finetune with sat? I found that the current sft.yaml file is not adjust with the v1.5 ckpt
In the fine-tuning section, we will provide a fine-tuning version of diffusers. The model version of diffusers will be released this week, and it is expected to be fine-tuned on cogvideox-factory
@zRzRzRzRzRzRzR Could you add some comments to this code? What is the role of OFS embedding? https://github.com/THUDM/CogVideo/blob/main/sat/dit_video_concat.py#L721
ofs is a constant, mainly adding a constant to the embed of the model