Definition
Low-Rank Adaptation: an efficient fine-tuning technique that adapts a large model (image, video, language) to a specific style, persona, subject, or concept by training only a small additional low-rank layer. It makes it possible to personalize generative models with a few GBs of compute instead of retraining the full model. In image gen it is the standard for reproducing artistic styles or characters.
Relationships
Concepts linked to this one by a real relationship in the glossary. Hover over each one to see why it's related.