README.md

llama-2-finetune: Finetune Llama 2 7b Model

Step-by-step Guide

1. Prepare Your Dataset

   Make sure your dataset has ONLY two columns: 'Human' and 'Assistant'.

2. Upload the Dataset to Google Drive

   • Upload the dataset to your Google Drive.
   • Click on the "Share" button for the uploaded file and change the access to "edit" for "Anyone with the link."
   • Copy the link to the file.

   

3. Clone the Repository

   Run the following command to clone the llama-2-finetune repository:

   !git clone https://github.com/hrushik98/llama-2-finetune.git 

4. Install Dependencies

Change the current directory to llama-2-finetune and install the required dependencies with the following commands:

    %cd llama-2-finetune/
    !pip install -r requirements.txt    

5. Start Finetuning

Run the finetune.py script with your dataset link as an argument:

!python3 finetune.py "<dataset_link>"  

Infererence of the finetuned model

1. !python3 infer.py "<prompt>"

Time to Download the weights

1. Download the weights from the "outputs" folder