Setup

Follow these setup steps to prepare to run the Mistral fine tuning blog post. This should run on any *nix systems such as WSL2 w/ CUDA access. (Note: Created using Python 3.9.12. w/ Cuda installed)

This project uses virtual env to isolate a python environment and pulls all necessary requirments from a requirements-file.

Setup Instructions

# Create a virtual environment using venv (do this *inside* of the project directory)
python -m venv env

# Activate your environment
source ./env/bin/activate 

# Verify that you have a "(env)" next to your command prompt. You can deactivte the virtual environment later by running `deactivate`

# Install dependencies (NOTE: this loads more than is strictly required for this demo)

python -m pip install -r requirements.txt

#The `python -m` part ensures that you run your virtual environment's pip and not the global pip.

You can now run the finetune.ipynb. (If you are running on VS Code don't forget to connect it to your virtual environment)