AI Glossary

Parameter-efficient fine-tuning

Also called PEFT

What does “Parameter-efficient fine-tuning” mean?

Parameter-efficient fine-tuning: Adapting a model by training a small set of parameters while most original weights remain fixed.

Parameter-efficient fine-tuning visual guide: Adapting a model by training a small set of parameters while most original weights remain fixed.
Visual guide Open full size (new tab)

What is an example of “Parameter-efficient fine-tuning”?

A team adds small LoRA adapters to specialize an existing model.

Keep learning

See also

Fine-tuningTermInstruction tuningTermKnowledge distillationTermContinual learningTermCatastrophic forgettingTerm

Same topic

Adapter layerTermSoft prompt tuningTermPolicy in reinforcement learningTermOversamplingTermPooling layerTerm