PEFT

Parameter-Efficient Fine-Tuning (PEFT) methods for large pretrained models

Machine LearningLLM ToolsOpen sourceOpen source
Visit tool

github.com

About PEFT

PEFT is a library that enables efficient adaptation of large pretrained models to various downstream applications by only fine-tuning a small number of model parameters. It is integrated with Transformers, Diffusers, and Accelerate for easy model training and inference.

Description summarised by AI from the sources listed below.

Key features

  • Parameter-Efficient Fine-Tuning (PEFT) methods
  • parameter-efficient fine-tuning
  • Integration with Transformers
  • reduced computational costs
  • reduced computational and storage costs
  • integration with Transformers, Diffusers, and Accelerate
  • reduced storage costs
  • Diffusers for conveniently managing different adapters
  • Accelerate for distributed training and inference

Use cases

  • downstream applications
  • large pretrained models

Pricing

Pricing model: Open source. Detailed plans are not recorded; check the official website for current prices.

Pricing summarised by AI from the sources listed below.