mavericb / ftune

README

This repository uses the LLaMA-Efficient-Tuning repository to perform fine-tuning and prediction on HPC resources.

Usage

1. Download the notebook by clicking on "Code" and selecting "Download ZIP"
2. Open JupyterLab and import the downloaded notebook
3. Follow the instructions in the notebook to perform training and prediction. You will need to provide a train.json and a test.json file according to the characteristics indicated in the notebook, and update the parameters according to the instructions provided

The list of currently supported models can be found at the LLaMA-Efficient-Tuning repository.