Lcm Lora Sdv1 5

🚀 潜在一致性模型 (LCM) LoRA: SDv1 - 5

潜在一致性模型 (LCM) LoRA 是一种蒸馏一致性适配器，可将推理步骤大幅减少至仅 2 - 8 步，显著提升了图像生成的效率。该模型由 Simian Luo、Yiqin Tan、Suraj Patil、Daniel Gu 等人在论文 LCM - LoRA: A universal Stable - Diffusion Acceleration Module 中提出。

🚀 快速开始

LCM - LoRA 自 🤗 Hugging Face Diffusers 库的 v0.23.0 版本起开始支持。若要运行该模型，需先安装最新版本的 Diffusers 库以及 peft、accelerate 和 transformers。可使用以下命令进行安装：

pip install --upgrade pip
pip install --upgrade diffusers transformers accelerate peft

⚠️ 重要提示

如需详细的使用示例，建议查看官方 [LCM - LoRA 文档](https://huggingface.co/docs/diffusers/main/en/using - diffusers/inference_with_lcm_lora)。

💻 使用示例

基础用法

文本到图像

适配器可与 SDv1 - 5 或其衍生模型一起加载。以下示例使用 [Lykon/dreamshaper - 7](https://huggingface.co/Lykon/dreamshaper - 7) 模型。需将调度器更改为 LCMScheduler，并将推理步骤减少至 2 到 8 步。请确保禁用 guidance_scale 或使用 1.0 到 2.0 之间的值。

import torch
from diffusers import LCMScheduler, AutoPipelineForText2Image

model_id = "Lykon/dreamshaper-7"
adapter_id = "latent-consistency/lcm-lora-sdv1-5"

pipe = AutoPipelineForText2Image.from_pretrained(model_id, torch_dtype=torch.float16, variant="fp16")
pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)
pipe.to("cuda")

# load and fuse lcm lora
pipe.load_lora_weights(adapter_id)
pipe.fuse_lora()


prompt = "Self-portrait oil painting, a beautiful cyborg with golden hair, 8k"

# disable guidance_scale by passing 0
image = pipe(prompt=prompt, num_inference_steps=4, guidance_scale=0).images[0]

图像到图像

LCM - LoRA 也可应用于图像到图像的任务。以下示例使用 [dreamshaper - 7](https://huggingface.co/Lykon/dreamshaper - 7) 模型和适用于 stable - diffusion - v1 - 5 的 LCM - LoRA。

import torch
from diffusers import AutoPipelineForImage2Image, LCMScheduler
from diffusers.utils import make_image_grid, load_image

pipe = AutoPipelineForImage2Image.from_pretrained(
    "Lykon/dreamshaper-7",
    torch_dtype=torch.float16,
    variant="fp16",
).to("cuda")

# set scheduler
pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)

# load LCM-LoRA
pipe.load_lora_weights("latent-consistency/lcm-lora-sdv1-5")
pipe.fuse_lora()

# prepare image
url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/img2img-init.png"
init_image = load_image(url)
prompt = "Astronauts in a jungle, cold color palette, muted colors, detailed, 8k"

# pass prompt and image to pipeline
generator = torch.manual_seed(0)
image = pipe(
    prompt,
    image=init_image,
    num_inference_steps=4,
    guidance_scale=1,
    strength=0.6,
    generator=generator
).images[0]
make_image_grid([init_image, image], rows=1, cols=2)

图像修复

LCM - LoRA 还可用于图像修复任务。

import torch
from diffusers import AutoPipelineForInpainting, LCMScheduler
from diffusers.utils import load_image, make_image_grid

pipe = AutoPipelineForInpainting.from_pretrained(
    "runwayml/stable-diffusion-inpainting",
    torch_dtype=torch.float16,
    variant="fp16",
).to("cuda")

# set scheduler
pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)

# load LCM-LoRA
pipe.load_lora_weights("latent-consistency/lcm-lora-sdv1-5")
pipe.fuse_lora()

# load base and mask image
init_image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/inpaint.png")
mask_image = load_image("https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/inpaint_mask.png")

# generator = torch.Generator("cuda").manual_seed(92)
prompt = "concept art digital painting of an elven castle, inspired by lord of the rings, highly detailed, 8k"
generator = torch.manual_seed(0)
image = pipe(
    prompt=prompt,
    image=init_image,
    mask_image=mask_image,
    generator=generator,
    num_inference_steps=4,
    guidance_scale=4, 
).images[0]
make_image_grid([init_image, mask_image, image], rows=1, cols=3)

ControlNet

以下示例使用 SD - v1 - 5 模型、适用于 SD - v1 - 5 的 LCM - LoRA 以及 Canny ControlNet。

import torch
import cv2
import numpy as np
from PIL import Image

from diffusers import StableDiffusionControlNetPipeline, ControlNetModel, LCMScheduler
from diffusers.utils import load_image

image = load_image(
    "https://hf.co/datasets/huggingface/documentation-images/resolve/main/diffusers/input_image_vermeer.png"
).resize((512, 512))

image = np.array(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)

controlnet = ControlNetModel.from_pretrained("lllyasviel/sd-controlnet-canny", torch_dtype=torch.float16)
pipe = StableDiffusionControlNetPipeline.from_pretrained(
    "runwayml/stable-diffusion-v1-5",
    controlnet=controlnet,
    torch_dtype=torch.float16,
    safety_checker=None,
    variant="fp16"
).to("cuda")

# set scheduler
pipe.scheduler = LCMScheduler.from_config(pipe.scheduler.config)

# load LCM-LoRA
pipe.load_lora_weights("latent-consistency/lcm-lora-sdv1-5")

generator = torch.manual_seed(0)
image = pipe(
    "the mona lisa",
    image=canny_image,
    num_inference_steps=4,
    guidance_scale=1.5,
    controlnet_conditioning_scale=0.8,
    cross_attention_kwargs={"scale": 1},
    generator=generator,
).images[0]
make_image_grid([canny_image, image], rows=1, cols=2)

🔧 技术细节

待补充。

📄 许可证

本项目采用 openrail++ 许可证。