"scaling_factor": 0.13025 video_latent = video_latent * scaling_factor when you scale the video latent, the video are too much noisy, that's have problem after training bad result from previous training... >>>[now train video] https://github.com/user-attachments/assets/11107fa6-8397-4f4e-82e1-4d30aff10ab8 >>> [z1 train video] 
"scaling_factor": 0.13025
video_latent = video_latent * scaling_factor
when you scale the video latent, the video are too much noisy, that's have problem after training bad result from previous training...
text_to_video_sample.mp4