[ICCV 2023] Tune-A-Video: One-Shot Tuning of Image Diffusion Models for Text-to-Video Generation
4.15k
stars
377
forks
source link
Is it possible to train only `ddim_inv_latent` and reuse Unet3d model from AnimateDiff? #84
Open
actionless opened 10 months ago
AnimateDiff's unet 3d model for reference: https://huggingface.co/spaces/guoyww/AnimateDiff/tree/main/models/Motion_Module