Buckets:

hf-doc-build
/

doc

Files

xet

hf-doc-build/doc / diffusers /v0.43.0 /zh /tutorials /basic_training.md

mishig

2 days ago

preview code

download

raw

17 kB

# 训练扩散模型

无条件图像生成是扩散模型的一个热门应用，它能够生成与训练数据集中图像风格相似的图片。通常，在特定数据集上微调预训练模型可以获得最佳效果。你可以在 [Hub](https://huggingface.co/search/full-text?q=unconditional-image-generation&type=model) 上找到许多此类检查点，但如果你找不到满意的，也可以随时训练自己的模型！

本教程将教你如何从零开始，在 [Smithsonian Butterflies](https://huggingface.co/datasets/huggan/smithsonian_butterflies_subset) 数据集的一个子集上训练 [UNet2DModel](/docs/diffusers/v0.43.0/zh/api/models/unet2d#diffusers.UNet2DModel)，以生成你自己的 🦋 蝴蝶 🦋。

> [!TIP]
> 💡 本训练教程基于 [Training with 🧨 Diffusers](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/training_example.ipynb) 笔记本。如需了解扩散模型（例如其工作原理）的更多细节和背景信息，请查阅该笔记本！

在开始之前，请确保已安装 🤗 Datasets 以加载和预处理图像数据集，并安装 🤗 Accelerate 以简化在任意数量 GPU 上的训练。以下命令还会安装 [TensorBoard](https://www.tensorflow.org/tensorboard) 用于可视化训练指标（你也可以使用 [Weights & Biases](https://docs.wandb.ai/) 来跟踪训练过程）。

```py
# uncomment to install the necessary libraries in Colab
#!pip install diffusers[training]
```

我们鼓励您与社区分享您的模型。为此，您需要登录您的 Hugging Face 账户（如果您还没有账户，可以在此处[创建](https://hf.co/join)一个！）。您可以从 notebook 中登录，并在提示时输入您的 token。请确保您的 token 具有写入权限。

```py
>>> from huggingface_hub import notebook_login
>>> notebook_login()
```

或者从终端登录：

```bash
hf auth login
```

由于模型检查点文件较大，请安装 [Git-LFS](https://git-lfs.com/) 来对这些大文件进行版本管理：

```bash
!sudo apt -qq install git-lfs
!git config --global credential.helper store
```

## 训练配置

为了方便起见，创建一个包含训练超参数的 `TrainingConfig` 类（您可以自由调整这些参数）：

```py
>>> from dataclasses import dataclass
>>> @dataclass
... class TrainingConfig:
...     image_size = 128  # the generated image resolution
...     train_batch_size = 16
...     eval_batch_size = 16  # how many images to sample during evaluation
...     num_epochs = 50
...     gradient_accumulation_steps = 1
...     learning_rate = 1e-4
...     lr_warmup_steps = 500
...     save_image_epochs = 10
...     save_model_epochs = 30
...     mixed_precision = "fp16"  # `no` for float32, `fp16` for automatic mixed precision
...     output_dir = "ddpm-butterflies-128"  # the model name locally and on the HF Hub
...     push_to_hub = True  # whether to upload the saved model to the HF Hub
...     hub_model_id = "<your-username>/<my-awesome-model>"  # the name of the repository to create on the HF Hub
...     hub_private_repo = None
...     overwrite_output_dir = True  # overwrite the old model when re-running the notebook
...     seed = 0
>>> config = TrainingConfig()
```

## 加载数据集

您可以轻松使用 🤗 Datasets 库加载 [Smithsonian Butterflies](https://huggingface.co/datasets/huggan/smithsonian_butterflies_subset) 数据集：

```py
>>> from datasets import load_dataset
>>> config.dataset_name = "huggan/smithsonian_butterflies_subset"
>>> dataset = load_dataset(config.dataset_name, split="train")
```

> [!TIP]
> 💡 你可以从 [HugGan Community Event](https://huggingface.co/huggan) 获取额外的数据集，也可以通过创建本地 [`ImageFolder`](https://huggingface.co/docs/datasets/image_dataset#imagefolder) 来使用自己的数据集。如果数据集来自 HugGan Community Event，请将 `config.dataset_name` 设置为该数据集的仓库 id；如果你使用的是自己的图像，则将其设置为 `imagefolder`。
🤗 Datasets 使用 [Image](https://huggingface.co/docs/datasets/v5.2.0/en/package_reference/main_classes#datasets.Image) 功能自动解码图像数据，并将其加载为 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)，以便我们进行可视化：
```py
>>> import matplotlib.pyplot as plt
>>> fig, axs = plt.subplots(1, 4, figsize=(16, 4))
>>> for i, image in enumerate(dataset[:4]["image"]):
...     axs[i].imshow(image)
...     axs[i].set_axis_off()
>>> fig.show()
```
    
不过，这些图像的尺寸各不相同，因此您需要先对其进行预处理：
* `Resize` 会将图像尺寸更改为 `config.image_size` 中定义的尺寸。
* `RandomHorizontalFlip` 通过随机镜像图像来增强数据集。
* `Normalize` 很重要，它可以将像素值重新缩放到 [-1, 1] 范围内，这正是模型所期望的。

```py
>>> from torchvision import transforms
>>> preprocess = transforms.Compose(
...     [
...         transforms.Resize((config.image_size, config.image_size)),
...         transforms.RandomHorizontalFlip(),
...         transforms.ToTensor(),
...         transforms.Normalize([0.5], [0.5]),
...     ]
... )
```

使用 🤗 Datasets 的 [set_transform](https://huggingface.co/docs/datasets/v5.2.0/en/package_reference/main_classes#datasets.Dataset.set_transform) 方法，在训练过程中即时应用 `preprocess` 函数：

```py
>>> def transform(examples):
...     images = [preprocess(image.convert("RGB")) for image in examples["image"]]
...     return {"images": images}
>>> dataset.set_transform(transform)
```

您可以再次可视化图像，以确认它们已被调整大小。现在，您可以将数据集包装在 [DataLoader](https://pytorch.org/docs/stable/data#torch.utils.data.DataLoader) 中用于训练了！

```py
>>> import torch
>>> train_dataloader = torch.utils.data.DataLoader(dataset, batch_size=config.train_batch_size, shuffle=True)
```

## 创建 UNet2DModel

🧨 Diffusers 中的预训练模型可以轻松地通过其模型类以及所需的参数来创建。例如，要创建一个 [UNet2DModel](/docs/diffusers/v0.43.0/zh/api/models/unet2d#diffusers.UNet2DModel)：

```py
>>> from diffusers import UNet2DModel
>>> model = UNet2DModel(
...     sample_size=config.image_size,  # the target image resolution
...     in_channels=3,  # the number of input channels, 3 for RGB images
...     out_channels=3,  # the number of output channels
...     layers_per_block=2,  # how many ResNet layers to use per UNet block
...     block_out_channels=(128, 128, 256, 256, 512, 512),  # the number of output channels for each UNet block
...     down_block_types=(
...         "DownBlock2D",  # a regular ResNet downsampling block
...         "DownBlock2D",
...         "DownBlock2D",
...         "DownBlock2D",
...         "AttnDownBlock2D",  # a ResNet downsampling block with spatial self-attention
...         "DownBlock2D",
...     ),
...     up_block_types=(
...         "UpBlock2D",  # a regular ResNet upsampling block
...         "AttnUpBlock2D",  # a ResNet upsampling block with spatial self-attention
...         "UpBlock2D",
...         "UpBlock2D",
...         "UpBlock2D",
...         "UpBlock2D",
...     ),
... )
```

通常，快速检查样本图像的形状是否与模型输出的形状匹配是一个好主意：

```py
>>> sample_image = dataset[0]["images"].unsqueeze(0)
>>> print("Input shape:", sample_image.shape)
Input shape: torch.Size([1, 3, 128, 128])
>>> print("Output shape:", model(sample_image, timestep=0).sample.shape)
Output shape: torch.Size([1, 3, 128, 128])
```

很好！接下来，你需要一个调度器（scheduler）来为图像添加一些噪声。

## 创建调度器

调度器的行为取决于您是将模型用于训练还是推理。在推理阶段，调度器从噪声生成图像。在训练阶段，调度器会获取扩散过程中特定时间点的模型输出（或样本），并根据*噪声调度*和*更新规则*对图像施加噪声。

让我们来看看 [DDPMScheduler](/docs/diffusers/v0.43.0/zh/api/schedulers/ddpm#diffusers.DDPMScheduler)，并使用 `add_noise` 方法为之前的 `sample_image` 添加一些随机噪声：

```py
>>> import torch
>>> from PIL import Image
>>> from diffusers import DDPMScheduler
>>> noise_scheduler = DDPMScheduler(num_train_timesteps=1000)
>>> noise = torch.randn(sample_image.shape)
>>> timesteps = torch.LongTensor([50])
>>> noisy_image = noise_scheduler.add_noise(sample_image, noise, timesteps)
>>> Image.fromarray(((noisy_image.permute(0, 2, 3, 1) + 1.0) * 127.5).type(torch.uint8).numpy()[0])
```

    
该模型的训练目标是预测添加到图像中的噪声。此步骤的损失可通过以下公式计算：

```py
>>> import torch.nn.functional as F
>>> noise_pred = model(noisy_image, timesteps).sample
>>> loss = F.mse_loss(noise_pred, noise)
```

## 训练模型

至此，你已具备开始训练模型所需的大部分组件，剩下的就是将它们整合在一起。

首先，你需要一个优化器和学习率调度器：

```py
>>> from diffusers.optimization import get_cosine_schedule_with_warmup
>>> optimizer = torch.optim.AdamW(model.parameters(), lr=config.learning_rate)
>>> lr_scheduler = get_cosine_schedule_with_warmup(
...     optimizer=optimizer,
...     num_warmup_steps=config.lr_warmup_steps,
...     num_training_steps=(len(train_dataloader) * config.num_epochs),
... )
```

然后，你需要一种方式来评估模型。对于评估，你可以使用 [DDPMPipeline](/docs/diffusers/v0.43.0/zh/api/pipelines/ddpm#diffusers.DDPMPipeline) 生成一批样本图像并将其保存为网格：

```py
>>> from diffusers import DDPMPipeline
>>> from diffusers.utils import make_image_grid
>>> import os
>>> def evaluate(config, epoch, pipeline):
...     # Sample some images from random noise (this is the backward diffusion process).
...     # The default pipeline output type is `List[PIL.Image]`
...     images = pipeline(
...         batch_size=config.eval_batch_size,
...         generator=torch.Generator(device='cpu').manual_seed(config.seed), # Use a separate torch generator to avoid rewinding the random state of the main training loop
...     ).images
...     # Make a grid out of the images
...     image_grid = make_image_grid(images, rows=4, cols=4)
...     # Save the images
...     test_dir = os.path.join(config.output_dir, "samples")
...     os.makedirs(test_dir, exist_ok=True)
...     image_grid.save(f"{test_dir}/{epoch:04d}.png")
```

现在，你可以使用 🤗 Accelerate 将这些组件整合到一个训练循环中，以便轻松实现 TensorBoard 日志记录、梯度累积和混合精度训练。要将模型上传到 Hub，请编写一个函数以获取你的仓库名称和信息，然后将其推送到 Hub。

> [!TIP]
> 💡 下面的训练循环可能看起来有些吓人且冗长，但当你之后只需一行代码就能启动训练时，你会觉得这一切都是值得的！如果你等不及了，想立即开始生成图像，请随意复制并运行下面的代码。你随时可以回来仔细研究训练循环，比如在你等待模型完成训练的时候。🤗

```py
>>> from accelerate import Accelerator
>>> from huggingface_hub import create_repo, upload_folder
>>> from tqdm.auto import tqdm
>>> from pathlib import Path
>>> import os
>>> def train_loop(config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler):
...     # Initialize accelerator and tensorboard logging
...     accelerator = Accelerator(
...         mixed_precision=config.mixed_precision,
...         gradient_accumulation_steps=config.gradient_accumulation_steps,
...         log_with="tensorboard",
...         project_dir=os.path.join(config.output_dir, "logs"),
...     )
...     if accelerator.is_main_process:
...         if config.output_dir is not None:
...             os.makedirs(config.output_dir, exist_ok=True)
...         if config.push_to_hub:
...             repo_id = create_repo(
...                 repo_id=config.hub_model_id or Path(config.output_dir).name, exist_ok=True
...             ).repo_id
...         accelerator.init_trackers("train_example")
...     # Prepare everything
...     # There is no specific order to remember, you just need to unpack the
...     # objects in the same order you gave them to the prepare method.
...     model, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(
...         model, optimizer, train_dataloader, lr_scheduler
...     )
...     global_step = 0
...     # Now you train the model
...     for epoch in range(config.num_epochs):
...         progress_bar = tqdm(total=len(train_dataloader), disable=not accelerator.is_local_main_process)
...         progress_bar.set_description(f"Epoch {epoch}")
...         for step, batch in enumerate(train_dataloader):
...             clean_images = batch["images"]
...             # Sample noise to add to the images
...             noise = torch.randn(clean_images.shape, device=clean_images.device)
...             bs = clean_images.shape[0]
...             # Sample a random timestep for each image
...             timesteps = torch.randint(
...                 0, noise_scheduler.config.num_train_timesteps, (bs,), device=clean_images.device,
...                 dtype=torch.int64
...             )
...             # Add noise to the clean images according to the noise magnitude at each timestep
...             # (this is the forward diffusion process)
...             noisy_images = noise_scheduler.add_noise(clean_images, noise, timesteps)
...             with accelerator.accumulate(model):
...                 # Predict the noise residual
...                 noise_pred = model(noisy_images, timesteps, return_dict=False)[0]
...                 loss = F.mse_loss(noise_pred, noise)
...                 accelerator.backward(loss)
...                 if accelerator.sync_gradients:
...                     accelerator.clip_grad_norm_(model.parameters(), 1.0)
...                 optimizer.step()
...                 lr_scheduler.step()
...                 optimizer.zero_grad()
...             progress_bar.update(1)
...             logs = {"loss": loss.detach().item(), "lr": lr_scheduler.get_last_lr()[0], "step": global_step}
...             progress_bar.set_postfix(**logs)
...             accelerator.log(logs, step=global_step)
...             global_step += 1
...         # After each epoch you optionally sample some demo images with evaluate() and save the model
...         if accelerator.is_main_process:
...             pipeline = DDPMPipeline(unet=accelerator.unwrap_model(model), scheduler=noise_scheduler)
...             if (epoch + 1) % config.save_image_epochs == 0 or epoch == config.num_epochs - 1:
...                 evaluate(config, epoch, pipeline)
...             if (epoch + 1) % config.save_model_epochs == 0 or epoch == config.num_epochs - 1:
...                 if config.push_to_hub:
...                     upload_folder(
...                         repo_id=repo_id,
...                         folder_path=config.output_dir,
...                         commit_message=f"Epoch {epoch}",
...                         ignore_patterns=["step_*", "epoch_*"],
...                     )
...                 else:
...                     pipeline.save_pretrained(config.output_dir)
```

呼，代码量确实不少！不过现在你终于可以使用 🤗 Accelerate 的 [notebook_launcher](https://huggingface.co/docs/accelerate/v1.15.0/en/package_reference/launchers#accelerate.notebook_launcher) 函数来启动训练了。请将训练循环、所有训练参数以及用于训练的进程数（你可以将此值更改为你可用的 GPU 数量）作为参数传入该函数：

```py
>>> from accelerate import notebook_launcher
>>> args = (config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler)
>>> notebook_launcher(train_loop, args, num_processes=1)
```

训练完成后，看看你的扩散模型生成的最终 🦋 图像 🦋 吧！

```py
>>> import glob
>>> sample_images = sorted(glob.glob(f"{config.output_dir}/samples/*.png"))
>>> Image.open(sample_images[-1])
```

    
## 后续步骤

无条件图像生成是可以训练的任务之一。您可以访问 [🧨 Diffusers 训练示例](../training/overview) 页面来探索其他任务和训练技术。以下是一些您可以学习的内容示例：

* [Textual Inversion](../training/text_inversion)，一种用于让模型学习特定视觉概念并将其融入生成图像的算法。
* [DreamBooth](../training/dreambooth)，一种在给定主体若干输入图像的情况下，生成该主体个性化图像的技术。
* 微调 Stable Diffusion 模型的[指南](../training/text2image)，使用您自己的数据集。
* 使用 LoRA 的[指南](../training/lora)，这是一种内存高效的微调技术，可加速对超大模型的微调。

Xet Storage Details

Size:: 17 kB
Xet hash:: 23c4bd2498000301382e4b5d65cc281f9bc96bb24e7b46d65c8ae27af16facd0

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.

	# 训练扩散模型

	无条件图像生成是扩散模型的一个热门应用，它能够生成与训练数据集中图像风格相似的图片。通常，在特定数据集上微调预训练模型可以获得最佳效果。你可以在 [Hub](https://huggingface.co/search/full-text?q=unconditional-image-generation&type=model) 上找到许多此类检查点，但如果你找不到满意的，也可以随时训练自己的模型！

	本教程将教你如何从零开始，在 [Smithsonian Butterflies](https://huggingface.co/datasets/huggan/smithsonian_butterflies_subset) 数据集的一个子集上训练 [UNet2DModel](/docs/diffusers/v0.43.0/zh/api/models/unet2d#diffusers.UNet2DModel)，以生成你自己的 🦋 蝴蝶 🦋。

	> [!TIP]
	> 💡 本训练教程基于 [Training with 🧨 Diffusers](https://colab.research.google.com/github/huggingface/notebooks/blob/main/diffusers/training_example.ipynb) 笔记本。如需了解扩散模型（例如其工作原理）的更多细节和背景信息，请查阅该笔记本！

	在开始之前，请确保已安装 🤗 Datasets 以加载和预处理图像数据集，并安装 🤗 Accelerate 以简化在任意数量 GPU 上的训练。以下命令还会安装 [TensorBoard](https://www.tensorflow.org/tensorboard) 用于可视化训练指标（你也可以使用 [Weights & Biases](https://docs.wandb.ai/) 来跟踪训练过程）。

	```py
	# uncomment to install the necessary libraries in Colab
	#!pip install diffusers[training]
	```

	我们鼓励您与社区分享您的模型。为此，您需要登录您的 Hugging Face 账户（如果您还没有账户，可以在此处[创建](https://hf.co/join)一个！）。您可以从 notebook 中登录，并在提示时输入您的 token。请确保您的 token 具有写入权限。

	```py
	>>> from huggingface_hub import notebook_login

	>>> notebook_login()
	```

	或者从终端登录：

	```bash
	hf auth login
	```

	由于模型检查点文件较大，请安装 [Git-LFS](https://git-lfs.com/) 来对这些大文件进行版本管理：

	```bash
	!sudo apt -qq install git-lfs
	!git config --global credential.helper store
	```

	## 训练配置

	为了方便起见，创建一个包含训练超参数的 `TrainingConfig` 类（您可以自由调整这些参数）：

	```py
	>>> from dataclasses import dataclass

	>>> @dataclass
	... class TrainingConfig:
	... image_size = 128 # the generated image resolution
	... train_batch_size = 16
	... eval_batch_size = 16 # how many images to sample during evaluation
	... num_epochs = 50
	... gradient_accumulation_steps = 1
	... learning_rate = 1e-4
	... lr_warmup_steps = 500
	... save_image_epochs = 10
	... save_model_epochs = 30
	... mixed_precision = "fp16" # `no` for float32, `fp16` for automatic mixed precision
	... output_dir = "ddpm-butterflies-128" # the model name locally and on the HF Hub

	... push_to_hub = True # whether to upload the saved model to the HF Hub
	... hub_model_id = "<your-username>/<my-awesome-model>" # the name of the repository to create on the HF Hub
	... hub_private_repo = None
	... overwrite_output_dir = True # overwrite the old model when re-running the notebook
	... seed = 0

	>>> config = TrainingConfig()
	```

	## 加载数据集

	您可以轻松使用 🤗 Datasets 库加载 [Smithsonian Butterflies](https://huggingface.co/datasets/huggan/smithsonian_butterflies_subset) 数据集：

	```py
	>>> from datasets import load_dataset

	>>> config.dataset_name = "huggan/smithsonian_butterflies_subset"
	>>> dataset = load_dataset(config.dataset_name, split="train")
	```

	> [!TIP]
	> 💡 你可以从 [HugGan Community Event](https://huggingface.co/huggan) 获取额外的数据集，也可以通过创建本地 [`ImageFolder`](https://huggingface.co/docs/datasets/image_dataset#imagefolder) 来使用自己的数据集。如果数据集来自 HugGan Community Event，请将 `config.dataset_name` 设置为该数据集的仓库 id；如果你使用的是自己的图像，则将其设置为 `imagefolder`。

	🤗 Datasets 使用 [Image](https://huggingface.co/docs/datasets/v5.2.0/en/package_reference/main_classes#datasets.Image) 功能自动解码图像数据，并将其加载为 [`PIL.Image`](https://pillow.readthedocs.io/en/stable/reference/Image.html)，以便我们进行可视化：

	```py
	>>> import matplotlib.pyplot as plt

	>>> fig, axs = plt.subplots(1, 4, figsize=(16, 4))
	>>> for i, image in enumerate(dataset[:4]["image"]):
	... axs[i].imshow(image)
	... axs[i].set_axis_off()
	>>> fig.show()
	```



	不过，这些图像的尺寸各不相同，因此您需要先对其进行预处理：

	* `Resize` 会将图像尺寸更改为 `config.image_size` 中定义的尺寸。
	* `RandomHorizontalFlip` 通过随机镜像图像来增强数据集。
	* `Normalize` 很重要，它可以将像素值重新缩放到 [-1, 1] 范围内，这正是模型所期望的。

	```py
	>>> from torchvision import transforms

	>>> preprocess = transforms.Compose(
	... [
	... transforms.Resize((config.image_size, config.image_size)),
	... transforms.RandomHorizontalFlip(),
	... transforms.ToTensor(),
	... transforms.Normalize([0.5], [0.5]),
	... ]
	... )
	```

	使用 🤗 Datasets 的 [set_transform](https://huggingface.co/docs/datasets/v5.2.0/en/package_reference/main_classes#datasets.Dataset.set_transform) 方法，在训练过程中即时应用 `preprocess` 函数：

	```py
	>>> def transform(examples):
	... images = [preprocess(image.convert("RGB")) for image in examples["image"]]
	... return {"images": images}

	>>> dataset.set_transform(transform)
	```

	您可以再次可视化图像，以确认它们已被调整大小。现在，您可以将数据集包装在 [DataLoader](https://pytorch.org/docs/stable/data#torch.utils.data.DataLoader) 中用于训练了！

	```py
	>>> import torch

	>>> train_dataloader = torch.utils.data.DataLoader(dataset, batch_size=config.train_batch_size, shuffle=True)
	```

	## 创建 UNet2DModel

	🧨 Diffusers 中的预训练模型可以轻松地通过其模型类以及所需的参数来创建。例如，要创建一个 [UNet2DModel](/docs/diffusers/v0.43.0/zh/api/models/unet2d#diffusers.UNet2DModel)：

	```py
	>>> from diffusers import UNet2DModel

	>>> model = UNet2DModel(
	... sample_size=config.image_size, # the target image resolution
	... in_channels=3, # the number of input channels, 3 for RGB images
	... out_channels=3, # the number of output channels
	... layers_per_block=2, # how many ResNet layers to use per UNet block
	... block_out_channels=(128, 128, 256, 256, 512, 512), # the number of output channels for each UNet block
	... down_block_types=(
	... "DownBlock2D", # a regular ResNet downsampling block
	... "DownBlock2D",
	... "DownBlock2D",
	... "DownBlock2D",
	... "AttnDownBlock2D", # a ResNet downsampling block with spatial self-attention
	... "DownBlock2D",
	... ),
	... up_block_types=(
	... "UpBlock2D", # a regular ResNet upsampling block
	... "AttnUpBlock2D", # a ResNet upsampling block with spatial self-attention
	... "UpBlock2D",
	... "UpBlock2D",
	... "UpBlock2D",
	... "UpBlock2D",
	... ),
	... )
	```

	通常，快速检查样本图像的形状是否与模型输出的形状匹配是一个好主意：

	```py
	>>> sample_image = dataset[0]["images"].unsqueeze(0)
	>>> print("Input shape:", sample_image.shape)
	Input shape: torch.Size([1, 3, 128, 128])

	>>> print("Output shape:", model(sample_image, timestep=0).sample.shape)
	Output shape: torch.Size([1, 3, 128, 128])
	```

	很好！接下来，你需要一个调度器（scheduler）来为图像添加一些噪声。

	## 创建调度器

	调度器的行为取决于您是将模型用于训练还是推理。在推理阶段，调度器从噪声生成图像。在训练阶段，调度器会获取扩散过程中特定时间点的模型输出（或样本），并根据噪声调度和更新规则对图像施加噪声。

	让我们来看看 [DDPMScheduler](/docs/diffusers/v0.43.0/zh/api/schedulers/ddpm#diffusers.DDPMScheduler)，并使用 `add_noise` 方法为之前的 `sample_image` 添加一些随机噪声：

	```py
	>>> import torch
	>>> from PIL import Image
	>>> from diffusers import DDPMScheduler

	>>> noise_scheduler = DDPMScheduler(num_train_timesteps=1000)
	>>> noise = torch.randn(sample_image.shape)
	>>> timesteps = torch.LongTensor([50])
	>>> noisy_image = noise_scheduler.add_noise(sample_image, noise, timesteps)

	>>> Image.fromarray(((noisy_image.permute(0, 2, 3, 1) + 1.0) * 127.5).type(torch.uint8).numpy()[0])
	```



	该模型的训练目标是预测添加到图像中的噪声。此步骤的损失可通过以下公式计算：

	```py
	>>> import torch.nn.functional as F

	>>> noise_pred = model(noisy_image, timesteps).sample
	>>> loss = F.mse_loss(noise_pred, noise)
	```

	## 训练模型

	至此，你已具备开始训练模型所需的大部分组件，剩下的就是将它们整合在一起。

	首先，你需要一个优化器和学习率调度器：

	```py
	>>> from diffusers.optimization import get_cosine_schedule_with_warmup

	>>> optimizer = torch.optim.AdamW(model.parameters(), lr=config.learning_rate)
	>>> lr_scheduler = get_cosine_schedule_with_warmup(
	... optimizer=optimizer,
	... num_warmup_steps=config.lr_warmup_steps,
	... num_training_steps=(len(train_dataloader) * config.num_epochs),
	... )
	```

	然后，你需要一种方式来评估模型。对于评估，你可以使用 [DDPMPipeline](/docs/diffusers/v0.43.0/zh/api/pipelines/ddpm#diffusers.DDPMPipeline) 生成一批样本图像并将其保存为网格：

	```py
	>>> from diffusers import DDPMPipeline
	>>> from diffusers.utils import make_image_grid
	>>> import os

	>>> def evaluate(config, epoch, pipeline):
	... # Sample some images from random noise (this is the backward diffusion process).
	... # The default pipeline output type is `List[PIL.Image]`
	... images = pipeline(
	... batch_size=config.eval_batch_size,
	... generator=torch.Generator(device='cpu').manual_seed(config.seed), # Use a separate torch generator to avoid rewinding the random state of the main training loop
	... ).images

	... # Make a grid out of the images
	... image_grid = make_image_grid(images, rows=4, cols=4)

	... # Save the images
	... test_dir = os.path.join(config.output_dir, "samples")
	... os.makedirs(test_dir, exist_ok=True)
	... image_grid.save(f"{test_dir}/{epoch:04d}.png")
	```

	现在，你可以使用 🤗 Accelerate 将这些组件整合到一个训练循环中，以便轻松实现 TensorBoard 日志记录、梯度累积和混合精度训练。要将模型上传到 Hub，请编写一个函数以获取你的仓库名称和信息，然后将其推送到 Hub。

	> [!TIP]
	> 💡 下面的训练循环可能看起来有些吓人且冗长，但当你之后只需一行代码就能启动训练时，你会觉得这一切都是值得的！如果你等不及了，想立即开始生成图像，请随意复制并运行下面的代码。你随时可以回来仔细研究训练循环，比如在你等待模型完成训练的时候。🤗

	```py
	>>> from accelerate import Accelerator
	>>> from huggingface_hub import create_repo, upload_folder
	>>> from tqdm.auto import tqdm
	>>> from pathlib import Path
	>>> import os

	>>> def train_loop(config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler):
	... # Initialize accelerator and tensorboard logging
	... accelerator = Accelerator(
	... mixed_precision=config.mixed_precision,
	... gradient_accumulation_steps=config.gradient_accumulation_steps,
	... log_with="tensorboard",
	... project_dir=os.path.join(config.output_dir, "logs"),
	... )
	... if accelerator.is_main_process:
	... if config.output_dir is not None:
	... os.makedirs(config.output_dir, exist_ok=True)
	... if config.push_to_hub:
	... repo_id = create_repo(
	... repo_id=config.hub_model_id or Path(config.output_dir).name, exist_ok=True
	... ).repo_id
	... accelerator.init_trackers("train_example")

	... # Prepare everything
	... # There is no specific order to remember, you just need to unpack the
	... # objects in the same order you gave them to the prepare method.
	... model, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(
	... model, optimizer, train_dataloader, lr_scheduler
	... )

	... global_step = 0

	... # Now you train the model
	... for epoch in range(config.num_epochs):
	... progress_bar = tqdm(total=len(train_dataloader), disable=not accelerator.is_local_main_process)
	... progress_bar.set_description(f"Epoch {epoch}")

	... for step, batch in enumerate(train_dataloader):
	... clean_images = batch["images"]
	... # Sample noise to add to the images
	... noise = torch.randn(clean_images.shape, device=clean_images.device)
	... bs = clean_images.shape[0]

	... # Sample a random timestep for each image
	... timesteps = torch.randint(
	... 0, noise_scheduler.config.num_train_timesteps, (bs,), device=clean_images.device,
	... dtype=torch.int64
	... )

	... # Add noise to the clean images according to the noise magnitude at each timestep
	... # (this is the forward diffusion process)
	... noisy_images = noise_scheduler.add_noise(clean_images, noise, timesteps)

	... with accelerator.accumulate(model):
	... # Predict the noise residual
	... noise_pred = model(noisy_images, timesteps, return_dict=False)[0]
	... loss = F.mse_loss(noise_pred, noise)
	... accelerator.backward(loss)

	... if accelerator.sync_gradients:
	... accelerator.clip_grad_norm_(model.parameters(), 1.0)
	... optimizer.step()
	... lr_scheduler.step()
	... optimizer.zero_grad()

	... progress_bar.update(1)
	... logs = {"loss": loss.detach().item(), "lr": lr_scheduler.get_last_lr()[0], "step": global_step}
	... progress_bar.set_postfix(**logs)
	... accelerator.log(logs, step=global_step)
	... global_step += 1

	... # After each epoch you optionally sample some demo images with evaluate() and save the model
	... if accelerator.is_main_process:
	... pipeline = DDPMPipeline(unet=accelerator.unwrap_model(model), scheduler=noise_scheduler)

	... if (epoch + 1) % config.save_image_epochs == 0 or epoch == config.num_epochs - 1:
	... evaluate(config, epoch, pipeline)

	... if (epoch + 1) % config.save_model_epochs == 0 or epoch == config.num_epochs - 1:
	... if config.push_to_hub:
	... upload_folder(
	... repo_id=repo_id,
	... folder_path=config.output_dir,
	... commit_message=f"Epoch {epoch}",
	... ignore_patterns=["step_", "epoch_"],
	... )
	... else:
	... pipeline.save_pretrained(config.output_dir)
	```

	呼，代码量确实不少！不过现在你终于可以使用 🤗 Accelerate 的 [notebook_launcher](https://huggingface.co/docs/accelerate/v1.15.0/en/package_reference/launchers#accelerate.notebook_launcher) 函数来启动训练了。请将训练循环、所有训练参数以及用于训练的进程数（你可以将此值更改为你可用的 GPU 数量）作为参数传入该函数：

	```py
	>>> from accelerate import notebook_launcher

	>>> args = (config, model, noise_scheduler, optimizer, train_dataloader, lr_scheduler)

	>>> notebook_launcher(train_loop, args, num_processes=1)
	```

	训练完成后，看看你的扩散模型生成的最终 🦋 图像 🦋 吧！

	```py
	>>> import glob

	>>> sample_images = sorted(glob.glob(f"{config.output_dir}/samples/*.png"))
	>>> Image.open(sample_images[-1])
	```



	## 后续步骤

	无条件图像生成是可以训练的任务之一。您可以访问 [🧨 Diffusers 训练示例](../training/overview) 页面来探索其他任务和训练技术。以下是一些您可以学习的内容示例：

	* [Textual Inversion](../training/text_inversion)，一种用于让模型学习特定视觉概念并将其融入生成图像的算法。
	* [DreamBooth](../training/dreambooth)，一种在给定主体若干输入图像的情况下，生成该主体个性化图像的技术。
	* 微调 Stable Diffusion 模型的[指南](../training/text2image)，使用您自己的数据集。
	* 使用 LoRA 的[指南](../training/lora)，这是一种内存高效的微调技术，可加速对超大模型的微调。