Limiting the model result to a specific shape

[barnacl (Rekil Prashanth)]

Looks interesting, are you trying to do something similar to inpainting?

Pipei2i error: Input type (torch.cuda.FloatTensor) and weight type (torch.cuda.HalfTensor)

There seems to be similar issue with the inpainting pipeline. Have to use autocast currently to get around it If anyone wants to play around with it, here is the code from diffusers import StableDiffusionInpaintPipeline from PIL import ImageDraw pipe_inpaint = StableDiffusionInpaintPipeline.from_pretrained( "CompVis/stable-diffusion-v1-4", revision="fp16", torch_dtype=torch.float16, ).to("cuda") p = FastDownload().download("https://i.redd.it/hdxi9adiquw21.jpg")#cre…