ml-stable-diffusion
ml-stable-diffusion copied to clipboard
Adding multiple ControlNets has no effect
After doing some tests passing more than one ControlNet in the array when loading the model and more than one controlNetInputs when generating has no effect on the generated image. Only the first ControlNet is used
The pipeline would have to pass the end result of one ControlNet to the next ControlNet as the starting image, and it seems that it does not do this presently. A clever coder could probably make it do so, as an option. I hope someone will undertake this coding task.