diffusers icon indicating copy to clipboard operation
diffusers copied to clipboard
Published 1 month ago • verified publisher icon huggingface

Add VidTok AutoEncoders

Open annitang1997 opened this issue 9 months ago • 5 comments

We add VidTok, a versatile and state-of-the-art video tokenizer, as an autoencoder model to diffusers.

Paper: https://arxiv.org/pdf/2412.13061 Code: https://github.com/microsoft/VidTok Model: https://huggingface.co/microsoft/VidTok

annitang1997 avatar Apr 09 '25 17:04 annitang1997

Thank you for the PR @annitang1997! I will review this in depth soon. cc @yiyixuxu too

a-r-r-o-w avatar Apr 10 '25 06:04 a-r-r-o-w

Is there any updates on the review process? 👀 Looking forward to use VidTok with diffusers.

deeptimhe avatar Apr 20 '25 09:04 deeptimhe

Hello, I have improved the code based on your feedback. Please check it. 🤗

annitang1997 avatar May 09 '25 16:05 annitang1997

Any updates in this thread? :)

deeptimhe avatar May 23 '25 10:05 deeptimhe

@deeptimhe Sorry for the delay, I'm on leave at the moment, and so is @yiyixuxu. I'll try to test the PR and give it a look next week when I'm back

a-r-r-o-w avatar May 23 '25 19:05 a-r-r-o-w

Hello, I have cleaned the code by removing small methods/functions based on your feedback. Please check it. 🤗

annitang1997 avatar Jul 02 '25 07:07 annitang1997

Any updates in this thread? :)

annitang1997 avatar Jul 28 '25 14:07 annitang1997