Axolotl

Axolotl wraps training tools into a YAML-driven workflow for fine-tuning language models. You specify the base model, dataset format and method, such as LoRA, and run one command.

It supports multi-GPU training and many model families, making it a common choice for community fine-tunes. Documentation and examples help newcomers.

It is best for people who want repeatable experiments without writing training code from scratch.