Skip to content
mlx.app

PEFT

Hugging Face's parameter-efficient fine-tuning library, the one everything wraps.

When to use it

Use this when you're already deep in the `transformers` stack and want LoRA as a library call rather than a standalone trainer.

Alternatives in Fine-tuning

MLX LM LoRA

The same mlx-lm package doubles as Apple's official fine-tuning path.

Unsloth

The fine-tuning library everyone quotes for faster training on less memory.

Axolotl

The YAML-config trainer that treats every method as a config file.