PEFT (Parameter-Efficient Fine-Tuning)

Umbrella term for fine-tuning techniques that update only a fraction of model parameters — LoRA is the most common.

Definition

PEFT is the family of techniques (LoRA, QLoRA, prefix tuning, prompt tuning) that fine-tune large models cheaply by only adjusting a small parameter set. Standard practice in 2026.

Examples

  • LoRA, QLoRA (quantized LoRA), AdaLoRA, prefix tuning

Related terms

Looking for actual gigs that use PEFT (Parameter-Efficient Fine-Tuning)? Browse all live listings or filter by $100+/hr.