ValueError: Trying to set a tensor of shape torch.Size([1280]) in "class_embedding" (which has shape torch.Size([768])), this looks incorrect.

Hi， when I run  cd diffusers/examples/world_model   python inference/img2video.py .
The error occurred: ValueError: Trying to set a tensor of shape torch.Size([1280]) in "class_embedding" (which has shape torch.Size([768])), this looks incorrect.
The pretrained model in the Hugging Face is incorrect?
 