When I generate image with huggingface stable diffusion pipeline, it sometimes generate with weird green, yellow and red lines as shown in the picture(https://i.stack.imgur.com/L9RpJ.jpg).
Is anyone who knows the reason of this, and how I can remove this lines?
To reproduce :
model_id = "../stable-diffusion-2" # I download stable-diffusion-2's weight and code
prompt = "best quality, extremely detailed"
scheduler = LMSDiscreteScheduler.from_pretrained(model_id, subfolder="scheduler")
pipe = StableDiffusionPipeline.from_pretrained(model_id, scheduler=scheduler, torch_dtype=torch.float16)
image = pipe(prompt).images[0]