Sahaj.blogs

Day page

10 Nov 2023

Paper summary

Paper summary: LCM-LoRA: A Universal Stable-Diffusion Acceleration Module

LCM-LoRA packages latent-consistency distillation as a LoRA that plugs into Stable Diffusion models and their fine-tunes to generate images in very few steps without extra training.

  • The LoRA weights learned during consistency distillation work as a reusable "accelerator" for other SD fine-tunes.
  • The authors frame it as a plug-in neural solver for the diffusion probability-flow ODE.

Read the post →