Question about length and resolution of the generated video

[Andressi]

Thanks for your great work!
I noticed that you added support for CogVideoX1.5 in infer.py, but did not update the relevant weights. If I want to generate videos of the same length and resolution as CogVideoX1.5, do I need to retrain a model based on CogVideoX1.5? Or do you have plans to open source the corresponding weights?

[SHYuanBest]

Thanks for interest! The current open-sourced weight only support 480x720x49, and we are working to open source the new model in the future.