AI ART Discover · Learn · Create
AI Art > Glossary > LoRA (Low-Rank Adaptation)
Browse Full AI Glossary
L AI Glossary Term • Term 122 of 224
Plain English Definition

What does “LoRA (Low-Rank Adaptation)” mean?

Core Definition

A parameter-efficient fine-tuning method that adds low-rank adapter matrices.

Operational Mechanism

How LoRA (Low-Rank Adaptation) Works

1

Frozen pre-trained weights, unchanged

2

Small trainable adapter matrices, only these learn

3

Combined at run time: adapted model.

Real-World Scenario

Practical Example

Instead of updating all billions of weights in a language model, a team trains small adapter matrices for a customer-service style. The adapter file is small, and the same base model can swap different adapters for different clients.

Core Significance

Why It Matters

LoRA makes fine-tuning possible on modest hardware and makes adapters easy to store and share.

Expert Q&A

Common Questions About LoRA (Low-Rank Adaptation)

Q: What does low-rank mean?

The weight changes are represented by two thin matrices multiplied together, which needs far fewer numbers than a full update.

Q: Does LoRA match full fine-tuning?

Often close on many tasks, though full fine-tuning may do better on some.

Q: Can adapters be combined?

Sometimes, with care. They can be merged into the base weights or swapped depending on the task.