Fantastic
So simple to use, no over-engineering like other models in this space. Can also work with local images and on MPS (Apple Silicon). The example below takes less than 20 seconds on M4 Max.
Example:
`import PIL
import requests
import torch
from diffusers import StableDiffusionInstructPix2PixPipeline, EulerAncestralDiscreteScheduler
model_id = "timbrooks/instruct-pix2pix"
pipe = StableDiffusionInstructPix2PixPipeline.from_pretrained(model_id, torch_dtype=torch.float16, safety_checker=None)
pipe.to("mps")
pipe.scheduler = EulerAncestralDiscreteScheduler.from_config(pipe.scheduler.config)
image = PIL.Image.open("images/pic.jpg")
image = image.resize((768, 768)) # Resize, can try varying sizes also
prompt = "make him older"
images = pipe(prompt, image=image, num_inference_steps=10, image_guidance_scale=1).images
images[0].show()
#images[0].save("images/pic_older.png")`