File size: 1,005 Bytes
d58ac58
ac94249
 
bce3703
ac94249
4b1eeea
4666901
2950d3f
 
ac94249
 
2d19487
d58ac58
ac94249
 
4211211
 
ac94249
44fbbac
 
 
 
84bc0a5
ac94249
4211211
ce05317
56f32b6
 
 
 
 
ac94249
 
83b5c4e
56f32b6
ac94249
ce05317
ac94249
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
import gradio as gr
from diffusers import StableDiffusionInpaintPipeline
import torch

pipeline = StableDiffusionInpaintPipeline.from_pretrained(
    "runwayml/stable-diffusion-inpainting",
    torch_dtype=torch.float16,
    use_safetensors=True,
    variant="fp16"
)
 
pipeline = pipeline.to("cuda")

def predict(mask_img):
    prompt = "a green frog, highly detailed, natural lighting"
    # get size of the input image
    size = mask_img["image"].size
    image = pipeline(prompt=prompt,
                     num_inference_steps=35,
                     image=mask_img["image"].convert("RGB"), 
                     mask_image=mask_img["mask"].convert("RGB"),
                     guidance_scale=9
                    ).images[0]
     
    return image.resize(size)

# def mirror(img):
#     print(img['mask'])
#     print(img['image'])
#     return img['image'].rotate(45)

demo = gr.Interface(
    fn=predict,
    inputs=gr.Image(tool = 'sketch', type='pil'),
    outputs=gr.Image()
)

demo.launch()