Axolotl

Axolotl is an open-source framework that streamlines fine-tuning of open LLMs, supporting LoRA/QLoRA, full fine-tunes, many architectures, and YAML-configured training pipelines on single or multi-GPU setups.