Add Cogvideo i2v support.

[foreverpiano]

🔥🔥 News: 2024/9/25: CogVideoX web demo is available on Replicate. Try the text-to-video model CogVideoX-5B here Replicate and image-to-video model CogVideoX-5B-I2V here Replicate.

[xibosun]

CogVideoX-5B-I2V uses the same transformer model as CogVideoX-5B. They only vary on the structure of the input latent: CogVideoX-5B uses the video latent solely, whereas CogVideoX-5B-I2V combines the video latent and the image latent to pass to the transformer. We plan to add CogVideoX-5B-I2V to xDiT. It would also be wonderful if you could help us on it.