Closed XieJunchen closed 2 months ago
The project has been updated, the ksampler needs two input (plus or minus), maybe it needs to run the clip model?
Don't use the default clip text encoder, there is a CogVideo Text Encoder node.
look this: https://github.com/kijai/ComfyUI-CogVideoXWrapper/issues/19#issuecomment-2313915458
The project has been updated, the ksampler needs two input (plus or minus), maybe it needs to run the clip model?