File size: 5,377 Bytes
1d24639
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
127
128
129
130
131
# Copyright 2023 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.

import inspect
from typing import List, Optional, Tuple, Union

import torch

from ...models import UNet2DModel, VQModel
from ...schedulers import DDIMScheduler
from ...utils.torch_utils import randn_tensor
from ..pipeline_utils import DiffusionPipeline, ImagePipelineOutput


class LDMPipeline(DiffusionPipeline):
    r"""
    Pipeline for unconditional image generation using latent diffusion.

    This model inherits from [`DiffusionPipeline`]. Check the superclass documentation for the generic methods
    implemented for all pipelines (downloading, saving, running on a particular device, etc.).

    Parameters:
        vqvae ([`VQModel`]):
            Vector-quantized (VQ) model to encode and decode images to and from latent representations.
        unet ([`UNet2DModel`]):
            A `UNet2DModel` to denoise the encoded image latents.
        scheduler ([`SchedulerMixin`]):
            [`DDIMScheduler`] is used in combination with `unet` to denoise the encoded image latents.
    """

    def __init__(self, vqvae: VQModel, unet: UNet2DModel, scheduler: DDIMScheduler):
        super().__init__()
        self.register_modules(vqvae=vqvae, unet=unet, scheduler=scheduler)

    @torch.no_grad()
    def __call__(
        self,
        batch_size: int = 1,
        generator: Optional[Union[torch.Generator, List[torch.Generator]]] = None,
        eta: float = 0.0,
        num_inference_steps: int = 50,
        output_type: Optional[str] = "pil",
        return_dict: bool = True,
        **kwargs,
    ) -> Union[Tuple, ImagePipelineOutput]:
        r"""
        The call function to the pipeline for generation.

        Args:
            batch_size (`int`, *optional*, defaults to 1):
                Number of images to generate.
            generator (`torch.Generator`, *optional*):
                A [`torch.Generator`](https://pytorch.org/docs/stable/generated/torch.Generator.html) to make
                generation deterministic.
            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.
            output_type (`str`, *optional*, defaults to `"pil"`):
                The output format of the generated image. Choose between `PIL.Image` or `np.array`.
            return_dict (`bool`, *optional*, defaults to `True`):
                Whether or not to return a [`~pipelines.ImagePipelineOutput`] instead of a plain tuple.

        Example:

        ```py
        >>> from diffusers import LDMPipeline

        >>> # load model and scheduler
        >>> pipe = LDMPipeline.from_pretrained("CompVis/ldm-celebahq-256")

        >>> # run pipeline in inference (sample random noise and denoise)
        >>> image = pipe().images[0]
        ```

        Returns:
            [`~pipelines.ImagePipelineOutput`] or `tuple`:
                If `return_dict` is `True`, [`~pipelines.ImagePipelineOutput`] is returned, otherwise a `tuple` is
                returned where the first element is a list with the generated images
        """

        latents = randn_tensor(
            (batch_size, self.unet.config.in_channels, self.unet.config.sample_size, self.unet.config.sample_size),
            generator=generator,
        )
        latents = latents.to(self.device)

        # scale the initial noise by the standard deviation required by the scheduler
        latents = latents * self.scheduler.init_noise_sigma

        self.scheduler.set_timesteps(num_inference_steps)

        # prepare extra kwargs for the scheduler step, since not all schedulers have the same signature
        accepts_eta = "eta" in set(inspect.signature(self.scheduler.step).parameters.keys())

        extra_kwargs = {}
        if accepts_eta:
            extra_kwargs["eta"] = eta

        for t in self.progress_bar(self.scheduler.timesteps):
            latent_model_input = self.scheduler.scale_model_input(latents, t)
            # predict the noise residual
            noise_prediction = self.unet(latent_model_input, t).sample
            # compute the previous noisy sample x_t -> x_t-1
            latents = self.scheduler.step(noise_prediction, t, latents, **extra_kwargs).prev_sample

        # adjust latents with inverse of vae scale
        latents = latents / self.vqvae.config.scaling_factor
        # decode the image latents with the VAE
        image = self.vqvae.decode(latents).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)