Thanks to your team, great work!
In my test, I have tried the text-to-video and text-to-image. They are ok.
However, I have attempted several methods for the second_v_clip mode generation, but the generated videos are of very low quality, and unrelated to the given prompt video.
Could you provide a demo script to show how it works?