File size: 4,367 Bytes
5b2ab1c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
from typing import List, Optional, Tuple, Union
from PIL import Image
import numpy as np
import torch

from diffusers import (
    ControlNetModel,
    StableDiffusionControlNetInpaintPipeline,
    UniPCMultistepScheduler,
)

from .controlnet import StableDiffusionControlNet, MODEL_DICT


class StableDiffusionControlNetInpaint(StableDiffusionControlNet):
    """StableDiffusion with ControlNet model for inpainting images based on prompts.

    Args:
        control_model_name (str):
            Name of the controlnet processor.
        sd_model_name (str):
            Name of the StableDiffusion model.
    """

    def __init__(
        self,
        control_model_name: str,
        sd_model_name: Optional[str] = "runwayml/stable-diffusion-inpainting",
    ) -> None:
        super().__init__(
            control_model_name=control_model_name,
            sd_model_name=sd_model_name,
        )

    def create_pipe(
        self, sd_model_name: str, control_model_name: str
    ) -> StableDiffusionControlNetInpaintPipeline:
        """Create a StableDiffusionControlNetInpaintPipeline.

        Args:
            sd_model_name (str): StableDiffusion model name.
            control_model_name (str): Name of the ControlNet module.

        Returns:
            StableDiffusionControlNetInpaintPipeline
        """
        controlnet = ControlNetModel.from_pretrained(
            MODEL_DICT[control_model_name]["model"], torch_dtype=torch.float16
        )
        pipe = StableDiffusionControlNetInpaintPipeline.from_pretrained(
            sd_model_name, controlnet=controlnet, torch_dtype=torch.float16
        )

        pipe.scheduler = UniPCMultistepScheduler.from_config(pipe.scheduler.config)
        pipe.enable_model_cpu_offload()
        pipe.enable_xformers_memory_efficient_attention()

        return pipe

    def process(
        self,
        images: List[Image.Image],
        prompts: List[str],
        mask_images: List[Image.Image],
        control_images: Optional[List[Image.Image]] = None,
        negative_prompt: Optional[str] = None,
        n_outputs: Optional[int] = 1,
        num_inference_steps: Optional[int] = 30,
        strength: Optional[float] = 1.0,
        guidance_scale: Optional[float] = 7.5,
        eta: Optional[float] = 0.0,
    ) -> List[List[Image.Image]]:
        """Inpaint images based on `prompts` using `control_images` and `mask_images`.

        Args:
            images (List[Image.Image]): Input images.
            prompts (List[str]): List of prompts.
            mask_images (List[Image.Image]): List of mask images.
            control_images (Optional[List[Image.Image]], optional): List of control images. Defaults to None.
            negative_prompt (Optional[str], optional): Negative prompt. Defaults to None.
            n_outputs (Optional[int], optional): Number of generated outputs. Defaults to 1.
            num_inference_steps (Optional[int], optional): Number of inference iterations. Defaults to 30.

        Returns:
            List[List[Image.Image]]
        """

        if control_images is None:
            control_images = self.generate_control_images(images)

        assert len(prompts) == len(
            control_images
        ), "Number of prompts and input images must be equal."

        if n_outputs > 1:
            prompts = self._repeat(prompts, n=n_outputs)
            images = self._repeat(images, n=n_outputs)
            control_images = self._repeat(control_images, n=n_outputs)
            mask_images = self._repeat(mask_images, n=n_outputs)

        generator = [
            torch.Generator(device="cuda").manual_seed(int(i))
            for i in np.random.randint(max(len(prompts), 16), size=len(prompts))
        ]

        output = self.pipe(
            prompts,
            image=images,
            control_image=control_images,
            mask_image=mask_images,
            negative_prompt=[negative_prompt] * len(prompts),
            num_inference_steps=num_inference_steps,
            generator=generator,
        )

        output_images = [
            output.images[idx * n_outputs : (idx + 1) * n_outputs]
            for idx in range(len(images) // n_outputs)
        ]

        return {
            "output_images": output_images,
            "control_images": control_images,
            "mask_images": mask_images,
        }