Difference to model in MS-Image2Video?

#2
by Softology - opened

Hi,

This model code is
pipe = pipeline(task="video-to-video", model='damo/Video-to-Video')

MS-Image2Video second stage is
pipe2 = pipeline(task='video-to-video', model='damo/Video-to-Video', model_revision='v1.1.0', device='cuda:0')

Both reference the same model? Or does the new MS-Vid2Vid-XL model need a revision to make sure it gets the new model?

We haven't updated the MS-Vid2Vid-XL model yet, so you don't need to download the model again. We recommend using model_revision='v1.1.0' in the new version of the SDK=1.8.4. The current SDK has optimized VRAM usage and can support more GPU cards.

OK, thanks. I just wanted to clarify that it was the same model.

Sign up or log in to comment