About model size and human pose

[CS123n]

Thanks for your wonderful job. I want to ask two questiones:

1. If I want a larger controlnet-xs, do I only need to change the contol_model_ratio configuration?
2. Does your pipeline work well for human poses?

[Sipirius]

1: With the current version, the only thing you need to change is indeed the control_model_ratio 2: We did not explicitly train on human poses, but there is no conceptual difference, so it should work just as well

cheers

[CS123n]

Thanks for your feedback! I have two more questiones:

1. Is learning embeddings important? It causes the initialization state to be incorrect.
2. What amount of data and number of steps are needed to train ControlNet-xs?