Learn about Low-Rank Adaptation (LoRA) in Stable Diffusion, a lightweight training technique that allows for fine-tuning large language and diffusion models without full model training. LoRA models are smaller, more efficient, and only update a subset of a checkpoint model's parameters. Examples of LoRA models include character LoRA, style LoRA, and clothing LoRA.
Table of contents
OverviewWhat Is Low-Rank AdaptationCheckpoint or LoRA?Examples of LoRA modelsFurther ReadingsSummary5 Impressions