Real-Time Latent Consistency Model

ControlNet

This demo showcases LCM Image to Image pipeline using Diffusers with a MJPEG stream server.

There are 0 user(s) sharing the same GPU, affecting real-time performance. Maximum queue size is 4. Duplicate and run it on your own GPU.

Prompt

Change the prompt to generate different images, accepts Compel syntax.

Advanced Options

Camera Options:

Inference Steps 4 LCM Inference Steps 50 Guidance Scale 8.0 Strength 0.5 ControlNet Condition Scale 0.8 ControlNet Guidance Start 0.0 ControlNet Guidance End 1.0 Canny Low Threshold 0.2 Canny High Threshold 0.8 Seed Image Dimensions

512x512

768x768

Debug Canny