<h2 id="heading-354">LCM-LoRA - Acceleration Module!</h2> Tested with ComfyUI, although I hear it's working with Auto1111 now! Step 1) Download LoRAStep 2) Add LoRA alongside any SDXL Model (or 1.5 model, if using the SD 1.5 version)Step 3) Set CFG to ~1.5 and Steps to 3Step 4) Generate images in ~&lt;1 second (instantaneously on a 4090)Basic LCM Comfy workflow attached as "Training Images" in zip format.Read more about the concept here; <a target="_blank" rel="ugc" href="https://github.com/luosiallen/latent-consistency-model/blob/main/LCM-LoRA%20Technical%20Report/LCM-LoRA-Technical-Report.pdf">https://github.com/luosiallen/latent-consistency-model/blob/main/LCM-LoRA%20Technical%20Report/LCM-LoRA-Technical-Report.pdf</a>Original HF link - SDXL; <a target="_blank" rel="ugc" href="https://huggingface.co/latent-consistency/lcm-lora-sdxl/tree/main">https://huggingface.co/latent-consistency/lcm-lora-sdxl/tree/main</a>Original HF link - SD 1.5; <a target="_blank" rel="ugc" href="https://huggingface.co/latent-consistency/lcm-lora-sdv1-5">https://huggingface.co/latent-consistency/lcm-lora-sdv1-5</a>

LCM-LoRA Weights - Stable Diffusion Acceleration Module

<a target="_blank" rel="ugc" href="https://huggingface.co/latent-consistency/lcm-lora-sdxl/tree/main">https://huggingface.co/latent-consistency/lcm-lora-sdxl/tree/main</a>

Files with SHA256: a764e6859b6e04047cd761c08ff0cee96413a8e004c9f07707530cd776b19141