I want to refine/retrain Stable Diffusion Inpainting
I have some specific medical images: (1) VAE I can refine separately (2) Instead of CLIP can use MEDCLIP (3) also succeeded to refine denoiser in unconditional image generation with diffusion
But I don鈥檛 see how to refine the model presented in runwayml/stable-diffusion-inpainting 路 Hugging Face. In general, I want to run their training code, but with my images, text and masks; but I don鈥檛 find their training code. Can someone guide me how to use Trainer object in this case?