[Comfyui] Any Model and any Lora 4 Step Highspeed Image Generation based on LCM (SD and SDXL compatible)
121
2.5k
38
Generates images based on the new LCM Loras von Huggingface. It requires only 4 steps and is therefore very fast.
The workflow includes an optional Upscaling in 2 Steps.
You can use it with any additional Lora or Controlnet. Workflow can be used with SD or SDXL Model.
Make sure to update Comfy UI and all modules to the latest Version to be able to use lcm as sampler.
Enable 1. Uscale and 2. Upscale if needed
SD LCM Lora: https://huggingface.co/latent-consistency/lcm-lora-sdv1-5/resolve/main/pytorch_lora_weights.safetensors?download=true
SDXL LCM Lora:
CFG should be 1-2
Steps should be 4-5
For more Information:
Workflow by cblinde