Buckets:

hf-doc-build
/

doc

Files

xet

hf-doc-build/doc / diffusers /v0.43.0 /zh /using-diffusers /legacy_adapters.md

mishig

2 days ago

preview code

download

raw

6.69 kB

# 旧版适配器

这些方法仍然受到支持，尤其是在 Stable Diffusion 系列的检查点上。对于新的工作，建议优先使用 [LoRA](../tutorials/using_peft_for_inference)、[IP-Adapter](./ip_adapter) 和 [ControlNet](./controlnet)。

## T2I-Adapter

[T2I-Adapter](https://huggingface.co/papers/2302.08453) 是一种用于可控生成的适配器。它学习从控制信号（例如深度图或 Canny 边缘）到预训练模型内部特征的映射，并在生成过程中加入该引导。其作用类似于 [ControlNet](./controlnet)，但它是一个更轻量级的即插即用适配器。

加载一个受控适配器，然后使用 [from_pretrained()](/docs/diffusers/v0.43.0/zh/api/pipelines/overview#diffusers.DiffusionPipeline.from_pretrained) 加载流水线，并通过 `adapter=` 传入该适配器。

```py
import torch
from diffusers import T2IAdapter, StableDiffusionXLAdapterPipeline, AutoencoderKL
t2i_adapter = T2IAdapter.from_pretrained(
    "TencentARC/t2i-adapter-canny-sdxl-1.0",
    dtype=torch.float16,
)
```

使用 [opencv-python](https://github.com/opencv/opencv-python) 生成 canny 图像。

```py
import cv2
import numpy as np
from PIL import Image
from diffusers.utils import load_image
original_image = load_image(
    "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/non-enhanced-prompt.png"
)
image = np.array(original_image)
low_threshold = 100
high_threshold = 200
image = cv2.Canny(image, low_threshold, high_threshold)
image = image[:, :, None]
image = np.concatenate([image, image, image], axis=2)
canny_image = Image.fromarray(image)
```

将 canny 图像传入 pipeline 以生成图像。

```py
vae = AutoencoderKL.from_pretrained("madebyollin/sdxl-vae-fp16-fix", dtype=torch.float16)
pipeline = StableDiffusionXLAdapterPipeline.from_pretrained(
    "stabilityai/stable-diffusion-xl-base-1.0",
    adapter=t2i_adapter,
    vae=vae,
    dtype=torch.float16,
).to("cuda")  # or "mps", "xpu", "cpu"
prompt = """
A photorealistic overhead image of a cat reclining sideways in a flamingo pool floatie holding a margarita. 
The cat is floating leisurely in the pool and completely relaxed and happy.
"""
pipeline(
    prompt, 
    image=canny_image,
    num_inference_steps=100, 
    guidance_scale=10,
).images[0]
```

  
    
    原始图像
  
  
    
    canny 图像
  
  
    
    生成的图像
  
另请参阅 [T2I-Adapter 训练](../training/t2i_adapters)。

### MultiAdapter

使用 `MultiAdapter` 组合多个控制（例如 canny 和 depth）。将适配器列表传入 `MultiAdapter`，然后将对应的控制图像以列表形式传入 pipeline。

```py
import torch
from diffusers.utils import load_image
from diffusers import StableDiffusionXLAdapterPipeline, AutoencoderKL, MultiAdapter, T2IAdapter
canny_image = load_image(
    "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/canny-cat.png"
)
depth_image = load_image(
    "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/sdxl_depth_image.png"
)
controls = [canny_image, depth_image]
prompt = ["""
a relaxed rabbit sitting on a striped towel next to a pool with a tropical drink nearby, 
bright sunny day, vacation scene, 35mm photograph, film, professional, 4k, highly detailed
"""]
adapters = MultiAdapter(
    [
        T2IAdapter.from_pretrained("TencentARC/t2i-adapter-canny-sdxl-1.0", dtype=torch.float16),
        T2IAdapter.from_pretrained("TencentARC/t2i-adapter-depth-midas-sdxl-1.0", dtype=torch.float16),
    ]
)
```

将适配器、提示词和控制图像传入 [StableDiffusionXLAdapterPipeline](/docs/diffusers/v0.43.0/zh/api/pipelines/stable_diffusion/adapter#diffusers.StableDiffusionXLAdapterPipeline)。使用 `adapter_conditioning_scale` 参数来确定为每个控制分配多少权重。

```py
vae = AutoencoderKL.from_pretrained("madebyollin/sdxl-vae-fp16-fix", dtype=torch.float16)
pipeline = StableDiffusionXLAdapterPipeline.from_pretrained(
    "stabilityai/stable-diffusion-xl-base-1.0",
    dtype=torch.float16,
    vae=vae,
    adapter=adapters,
).to("cuda")  # or "mps", "xpu", "cpu"
pipeline(
    prompt,
    image=controls,
    height=1024,
    width=1024,
    adapter_conditioning_scale=[0.7, 0.7]
).images[0]
```

  
    
    canny 图像
  
  
    
    深度图
  
   
    
    生成的图像
  
## 文本反转

[Textual Inversion](https://huggingface.co/papers/2208.01618) 通过微调绑定到唯一标记（`<sks>`）的词嵌入，使模型从 3-5 张图片中学习特定概念。之后，你可以在提示词中使用该标记来生成该概念（例如，像素艺术）。

Textual Inversion 权重通常只有几 KB，因为它们仅包含词嵌入。请在调用 [from_pretrained()](/docs/diffusers/v0.43.0/zh/api/pipelines/overview#diffusers.DiffusionPipeline.from_pretrained) 之后，使用 [load_textual_inversion()](/docs/diffusers/v0.43.0/zh/api/loaders/textual_inversion#diffusers.loaders.TextualInversionLoaderMixin.load_textual_inversion) 加载这些权重，并在提示词中包含该唯一标记以触发生成。

```py
import torch
from diffusers import AutoPipelineForText2Image
pipeline = AutoPipelineForText2Image.from_pretrained(
    "stable-diffusion-v1-5/stable-diffusion-v1-5",
    dtype=torch.float16
).to("cuda")  # or "mps", "xpu", "cpu"
```

```py
pipeline.load_textual_inversion("sd-concepts-library/gta5-artwork")
prompt = "A cute brown bear eating a slice of pizza, stunning color scheme, masterpiece, illustration, <gta5-artwork> style"
pipeline(prompt).images[0]
```

    
要训练嵌入，请参阅 [训练文本反转](../training/text_inversion)。

Textual Inversion 还可以训练以学习*负向嵌入*，从而引导生成过程避开诸如“模糊”或“丑陋”等不良特征，有助于提升图像质量。

EasyNegative 是一种广泛使用的负向嵌入，其中包含多个学习到的负向概念。加载负向嵌入，并指定与负向嵌入关联的文件名和标记。将该标记传入管道中的 `negative_prompt` 以启用它。

```py
import torch
from diffusers import AutoPipelineForText2Image
pipeline = AutoPipelineForText2Image.from_pretrained(
    "stable-diffusion-v1-5/stable-diffusion-v1-5",
    dtype=torch.float16
).to("cuda")  # or "mps", "xpu", "cpu"
pipeline.load_textual_inversion(
    "EvilEngine/easynegative",
    weight_name="easynegative.safetensors",
    token="easynegative"
)
prompt = "A cute brown bear eating a slice of pizza, stunning color scheme, masterpiece, illustration"
negative_prompt = "easynegative"
pipeline(prompt, negative_prompt=negative_prompt).images[0]
```

    

Xet Storage Details

Size:: 6.69 kB
Xet hash:: cde6d6a34a35ea9bca7bcbc85c6b987823ec2351e2f4f1f469bba56915179336

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

	# 旧版适配器

	这些方法仍然受到支持，尤其是在 Stable Diffusion 系列的检查点上。对于新的工作，建议优先使用 [LoRA](../tutorials/using_peft_for_inference)、[IP-Adapter](./ip_adapter) 和 [ControlNet](./controlnet)。

	## T2I-Adapter

	[T2I-Adapter](https://huggingface.co/papers/2302.08453) 是一种用于可控生成的适配器。它学习从控制信号（例如深度图或 Canny 边缘）到预训练模型内部特征的映射，并在生成过程中加入该引导。其作用类似于 [ControlNet](./controlnet)，但它是一个更轻量级的即插即用适配器。

	加载一个受控适配器，然后使用 [from_pretrained()](/docs/diffusers/v0.43.0/zh/api/pipelines/overview#diffusers.DiffusionPipeline.from_pretrained) 加载流水线，并通过 `adapter=` 传入该适配器。

	```py
	import torch
	from diffusers import T2IAdapter, StableDiffusionXLAdapterPipeline, AutoencoderKL

	t2i_adapter = T2IAdapter.from_pretrained(
	"TencentARC/t2i-adapter-canny-sdxl-1.0",
	dtype=torch.float16,
	)
	```

	使用 [opencv-python](https://github.com/opencv/opencv-python) 生成 canny 图像。

	```py
	import cv2
	import numpy as np
	from PIL import Image
	from diffusers.utils import load_image

	original_image = load_image(
	"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/non-enhanced-prompt.png"
	)

	image = np.array(original_image)

	low_threshold = 100
	high_threshold = 200

	image = cv2.Canny(image, low_threshold, high_threshold)
	image = image[:, :, None]
	image = np.concatenate([image, image, image], axis=2)
	canny_image = Image.fromarray(image)
	```

	将 canny 图像传入 pipeline 以生成图像。

	```py
	vae = AutoencoderKL.from_pretrained("madebyollin/sdxl-vae-fp16-fix", dtype=torch.float16)
	pipeline = StableDiffusionXLAdapterPipeline.from_pretrained(
	"stabilityai/stable-diffusion-xl-base-1.0",
	adapter=t2i_adapter,
	vae=vae,
	dtype=torch.float16,
	).to("cuda") # or "mps", "xpu", "cpu"

	prompt = """
	A photorealistic overhead image of a cat reclining sideways in a flamingo pool floatie holding a margarita.
	The cat is floating leisurely in the pool and completely relaxed and happy.
	"""

	pipeline(
	prompt,
	image=canny_image,
	num_inference_steps=100,
	guidance_scale=10,
	).images[0]
	```



	原始图像



	canny 图像



	生成的图像


	另请参阅 [T2I-Adapter 训练](../training/t2i_adapters)。

	### MultiAdapter

	使用 `MultiAdapter` 组合多个控制（例如 canny 和 depth）。将适配器列表传入 `MultiAdapter`，然后将对应的控制图像以列表形式传入 pipeline。

	```py
	import torch
	from diffusers.utils import load_image
	from diffusers import StableDiffusionXLAdapterPipeline, AutoencoderKL, MultiAdapter, T2IAdapter

	canny_image = load_image(
	"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/canny-cat.png"
	)
	depth_image = load_image(
	"https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/sdxl_depth_image.png"
	)
	controls = [canny_image, depth_image]
	prompt = ["""
	a relaxed rabbit sitting on a striped towel next to a pool with a tropical drink nearby,
	bright sunny day, vacation scene, 35mm photograph, film, professional, 4k, highly detailed
	"""]

	adapters = MultiAdapter(
	[
	T2IAdapter.from_pretrained("TencentARC/t2i-adapter-canny-sdxl-1.0", dtype=torch.float16),
	T2IAdapter.from_pretrained("TencentARC/t2i-adapter-depth-midas-sdxl-1.0", dtype=torch.float16),
	]
	)
	```

	将适配器、提示词和控制图像传入 [StableDiffusionXLAdapterPipeline](/docs/diffusers/v0.43.0/zh/api/pipelines/stable_diffusion/adapter#diffusers.StableDiffusionXLAdapterPipeline)。使用 `adapter_conditioning_scale` 参数来确定为每个控制分配多少权重。

	```py
	vae = AutoencoderKL.from_pretrained("madebyollin/sdxl-vae-fp16-fix", dtype=torch.float16)
	pipeline = StableDiffusionXLAdapterPipeline.from_pretrained(
	"stabilityai/stable-diffusion-xl-base-1.0",
	dtype=torch.float16,
	vae=vae,
	adapter=adapters,
	).to("cuda") # or "mps", "xpu", "cpu"

	pipeline(
	prompt,
	image=controls,
	height=1024,
	width=1024,
	adapter_conditioning_scale=[0.7, 0.7]
	).images[0]
	```



	canny 图像



	深度图



	生成的图像


	## 文本反转

	[Textual Inversion](https://huggingface.co/papers/2208.01618) 通过微调绑定到唯一标记（`<sks>`）的词嵌入，使模型从 3-5 张图片中学习特定概念。之后，你可以在提示词中使用该标记来生成该概念（例如，像素艺术）。

	Textual Inversion 权重通常只有几 KB，因为它们仅包含词嵌入。请在调用 [from_pretrained()](/docs/diffusers/v0.43.0/zh/api/pipelines/overview#diffusers.DiffusionPipeline.from_pretrained) 之后，使用 [load_textual_inversion()](/docs/diffusers/v0.43.0/zh/api/loaders/textual_inversion#diffusers.loaders.TextualInversionLoaderMixin.load_textual_inversion) 加载这些权重，并在提示词中包含该唯一标记以触发生成。

	```py
	import torch
	from diffusers import AutoPipelineForText2Image

	pipeline = AutoPipelineForText2Image.from_pretrained(
	"stable-diffusion-v1-5/stable-diffusion-v1-5",
	dtype=torch.float16
	).to("cuda") # or "mps", "xpu", "cpu"
	```

	```py
	pipeline.load_textual_inversion("sd-concepts-library/gta5-artwork")
	prompt = "A cute brown bear eating a slice of pizza, stunning color scheme, masterpiece, illustration, <gta5-artwork> style"
	pipeline(prompt).images[0]
	```



	要训练嵌入，请参阅 [训练文本反转](../training/text_inversion)。

	Textual Inversion 还可以训练以学习负向嵌入，从而引导生成过程避开诸如“模糊”或“丑陋”等不良特征，有助于提升图像质量。

	EasyNegative 是一种广泛使用的负向嵌入，其中包含多个学习到的负向概念。加载负向嵌入，并指定与负向嵌入关联的文件名和标记。将该标记传入管道中的 `negative_prompt` 以启用它。

	```py
	import torch
	from diffusers import AutoPipelineForText2Image

	pipeline = AutoPipelineForText2Image.from_pretrained(
	"stable-diffusion-v1-5/stable-diffusion-v1-5",
	dtype=torch.float16
	).to("cuda") # or "mps", "xpu", "cpu"
	pipeline.load_textual_inversion(
	"EvilEngine/easynegative",
	weight_name="easynegative.safetensors",
	token="easynegative"
	)
	prompt = "A cute brown bear eating a slice of pizza, stunning color scheme, masterpiece, illustration"
	negative_prompt = "easynegative"
	pipeline(prompt, negative_prompt=negative_prompt).images[0]
	```