AI Glossary
LoRA (Low-Rank Adaptation)
What is LoRA (Low-Rank Adaptation)?
A parameter-efficient fine-tuning technique that inserts small trainable matrices into a model's layers rather than updating all weights. LoRA dramatically reduces the memory and compute cost of fine-tuning large models. It has become the dominant method for producing custom domain-specific model variants, particularly in the open-source community.
Example in practice
A company that trains a custom Llama 3 adapter on their internal product documentation using LoRA on a single consumer GPU — at a fraction of full fine-tuning cost — is using the technique that made domain-specific AI accessible without enterprise-scale compute.
Learn more
See LoRA (Low-Rank Adaptation) applied in a professional context through this free course.
AI Strategy for Leaders →