google

/

gemma-2-2b-it

like 553

Hi @omarabb315
Currently, the Gemma Cookbook contains three tutorials on fine-tuning the Gemma model:

• Finetune_with_XTuner.ipynb
• Finetune_with_Axolotl.ipynb
• Finetune_with_LLaMA_Factory.ipynb

Thanks this is great, can you please tell me the difference between those methods?

Basically, the notebooks utilize different datasets and toolkits—XTuner, Axolotl, and LLaMA Factory—to fine-tune Gemma models, each with its own unique features and capabilities.

1. Finetune_with_XTuner.ipynb

• Toolkit: XTuner - An efficient, flexible, and full-featured toolkit specifically designed for fine-tuning LLMs.
• Model: Gemma-2b-it
• Dataset: Alpaca dataset

2. Finetune_with_Axolotl.ipynb

• Toolkit: Axolotl - A tool designed to streamline the fine-tuning of various AI models, offering support for multiple configurations and architectures.  
• Model: Gemma-2b
• Dataset: mhenrichsen/alpaca_2k_test

3. Finetune_with_LLaMA_Factory.ipynb

• Toolkit: LLaMA Factory - A WebUI for Efficient Fine-Tuning of 100+ LLMs.
• Model: Gemma-2b
• Dataset: Demo Alpaca dataset