Innovate Futures @ Benji
patreon
CogVideoX 5B AI Video Model Updated With Img2Vid In ComfyUI
🕑 Added 2024-09-19 13:00:11 +0000 UTC
Video : https://youtu.be/sy6XHF2LEfI
In this video, we dive into a comprehensive review of CogVideoX, specifically focusing on the latest update for image-to-video models. We explore the advancements in image encoding and how it enhances the generation of motions based on reference images and various settings. I'll walk you through the parameters, runtime usage, and the process of setting up the CogVideoX 5B image-to-video models, providing insights on how to optimize your workflow efficiently.
As we delve deeper into the functionalities of CogVideoX, I showcase a step-by-step guide on how to download and integrate the latest image-to-video models seamlessly into your ComfyUI setup. From understanding the model's requirements to creating subfolders for efficient organization, this video serves as a comprehensive tutorial for both beginners and advanced users looking to leverage AI video technologies. Additionally, I touch upon the limitations of the current model in terms of video duration and quality, offering insights on potential enhancements for future iterations.
ComfyUI-CogVideoXWrapper
https://github.com/kijai/ComfyUI-CogVideoXWrapper
CogVideoX-5b-I2V
https://huggingface.co/THUDM/CogVideoX-5b-I2V/tree/main
Attached workflow file use in this video. A basic workflow for CogVideoX and AnimateDiff v2v.
The V2V is not a perfect one for refining Cogvideox generated video yet.
It might need to change in the future.
Comments
Neon Square
Hello! I keep getting this error Given groups=1, weight of size [3072, 16, 2, 2], expected input[26, 32, 60, 90] to have 16 channels, but got 32 channels instead
