Official Repository of the paper: Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference.

Project Page: https://latent-consistency-models.github.io

Try our 🤗 Hugging Face Demos: 🔥🔥🔥

• (🔥New) 2023/10/21 We support local gradio demo now. You can run the LCM model locally!! Please refer to the "Local gradio Demos".
• (🔥New) 2023/10/19 We provide a demo of LCM in 🤗 Hugging Face Space. Try it here.
• (🔥New) 2023/10/19 We provide the LCM model (Dreamshaper_v7) in 🤗 Hugging Face. Download here.
• (🔥New) 2023/10/19 LCM is integrated in 🧨 Diffusers library. Please refer to the "Usage".

To run the model locally, you can download the "local_gradio" folder:

1. Install Pytorch (CUDA). MacOS system can download the "MPS" version of Pytorch. Please refer to: https://pytorch.org
2. Install the main library:

pip install diffusers transformers accelerate gradio==3.48.0 

3. Launch the gradio:

python app.py

Ours Hugging Face Demo and Model are released ! Latent Consistency Models are supported in 🧨 diffusers.

Hugging Face Demo:

Replicate Demo:

LCM Model Download: LCM_Dreamshaper_v7

By distilling classifier-free guidance into the model's input, LCM can generate high-quality images in very short inference time. We compare the inference time at the setting of 768 x 768 resolution, CFG scale w=8, batchsize=4, using a A800 GPU.

You can try out Latency Consistency Models directly on:

To run the model yourself, you can leverage the 🧨 Diffusers library:

1. Install the library:

pip install diffusers transformers accelerate

2. Run the model:

from diffusers import DiffusionPipeline
import torch

pipe = DiffusionPipeline.from_pretrained("SimianLuo/LCM_Dreamshaper_v7", custom_pipeline="latent_consistency_txt2img", custom_revision="main")

# To save GPU memory, torch.float16 can be used, but it may compromise image quality.
pipe.to(torch_device="cuda", torch_dtype=torch.float32)

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

# Can be set to 1~50 steps. LCM support fast inference even <= 4 steps. Recommend: 1~8 steps.
num_inference_steps = 4 

images = pipe(prompt=prompt, num_inference_steps=num_inference_steps, guidance_scale=8.0, lcm_origin_steps=50, output_type="pil").images

@misc{luo2023latent,
      title={Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference}, 
      author={Simian Luo and Yiqin Tan and Longbo Huang and Jian Li and Hang Zhao},
      year={2023},
      eprint={2310.04378},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}