Some questions about the number of channels of Unet and its' initialization

[Delicious-Bitter-Melon]

Thanks for your excellent work.

Do you expand all convolutional layers of UNet to 13 channels initialized with zero weights or only expand the first convolutional layer of UNet to 13 channels? Do you use the pre-trained SDXL inpainting models to initialize the denoiser inpainting Unet?

[yisol]

Hello, we initialized additional channels with zero weights and used sdxl inpainting model for initialization of TryOnNet.