Scale and customize LLMs across different AI models using our *free* open-source solutions
Join the Network: Learn from the growing Axolotl community – with over 170 contributors and 500+ active Discord members, we have answers to every kind of question
No Hardware Needed: Start training models right away with pre-existing configurations (in YAML) in the OSS repo without having to tweak any settings
Ask the Experts: Axolotl users integrate with Runpod, Latitude, Modal, and Jarvislabs with recent product updates allowing GRPO support, direct to Modal pushes and more...
From FSDP+qLoRA fixes, LoRA+, Multipack, PEFT/LoRA, you can rely on Axolotl for quick integration of high performance fine-tuning techniques
Leverage the latest technologies to quickly and efficiently fine-tune models – sometimes 3-5x faster than elsewhere with new product updates every few weeks
We support a wide range of model architectures by Huggingface transformers as well as the ability to compose techniques quickly with new model structures
You can run Axolotl anywhere – in your own cloud, with your own Docker and k8s setup, whatever you need. Enterprises use Axolotl on their own terms with massive results
BYOD: Bring your own data – Your data doesn’t have to be uploaded to the internet ex. OpenAI for robust compliance and data governance. Axolotl allows full control
Trust: Axolotl is used by research companies, model builders, Gen AI platforms, and practitioners like Teknium/Nous Research, Modal, Replicate, OpenPipe
Axolotl is an Open Source tool making fine-tuning AI models faster without sacrificing functionality or scale.
Get started with our pre made recipes and notebooks
