File size: 4,049 Bytes
522606a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
# Copyright 2022 The HuggingFace Team. All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.


from typing import Optional, Tuple, Union

import torch

from ...models import UNet2DModel
from ...pipeline_utils import DiffusionPipeline, ImagePipelineOutput
from ...schedulers import PNDMScheduler


class PNDMPipeline(DiffusionPipeline):
    r"""
    This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for the generic methods the
    library implements for all the pipelines (such as downloading or saving, running on a particular device, etc.)

    Parameters:
        unet (`UNet2DModel`): U-Net architecture to denoise the encoded image latents.
        scheduler ([`SchedulerMixin`]):
            The `PNDMScheduler` to be used in combination with `unet` to denoise the encoded image.
    """

    unet: UNet2DModel
    scheduler: PNDMScheduler

    def __init__(self, unet: UNet2DModel, scheduler: PNDMScheduler):
        super().__init__()
        self.register_modules(unet=unet, scheduler=scheduler)

    @torch.no_grad()
    def __call__(
        self,
        batch_size: int = 1,
        num_inference_steps: int = 50,
        generator: Optional[torch.Generator] = None,
        output_type: Optional[str] = "pil",
        return_dict: bool = True,
        **kwargs,
    ) -> Union[ImagePipelineOutput, Tuple]:
        r"""
        Args:
            batch_size (`int`, `optional`, defaults to 1): The number of images to generate.
            num_inference_steps (`int`, `optional`, defaults to 50):
                The number of denoising steps. More denoising steps usually lead to a higher quality image at the
                expense of slower inference.
            generator (`torch.Generator`, `optional`): A [torch
                generator](https://pytorch.org/docs/stable/generated/torch.Generator.html) to make generation
                deterministic.
            output_type (`str`, `optional`, defaults to `"pil"`): The output format of the generate image. Choose
                between [PIL](https://pillow.readthedocs.io/en/stable/): `PIL.Image.Image` or `np.array`.
            return_dict (`bool`, `optional`, defaults to `True`): Whether or not to return a
                [`~pipeline_utils.ImagePipelineOutput`] instead of a plain tuple.

        Returns:
            [`~pipeline_utils.ImagePipelineOutput`] or `tuple`: [`~pipelines.utils.ImagePipelineOutput`] if
            `return_dict` is True, otherwise a `tuple. When returning a tuple, the first element is a list with the
            generated images.
        """
        # For more information on the sampling method you can take a look at Algorithm 2 of
        # the official paper: https://arxiv.org/pdf/2202.09778.pdf

        # Sample gaussian noise to begin loop
        image = torch.randn(
            (batch_size, self.unet.in_channels, self.unet.sample_size, self.unet.sample_size),
            generator=generator,
        )
        image = image.to(self.device)

        self.scheduler.set_timesteps(num_inference_steps)
        for t in self.progress_bar(self.scheduler.timesteps):
            model_output = self.unet(image, t).sample

            image = self.scheduler.step(model_output, t, image).prev_sample

        image = (image / 2 + 0.5).clamp(0, 1)
        image = image.cpu().permute(0, 2, 3, 1).numpy()
        if output_type == "pil":
            image = self.numpy_to_pil(image)

        if not return_dict:
            return (image,)

        return ImagePipelineOutput(images=image)