Instuctions for running GP solution to hackathon

Intro :

Below is a codebase for our team's solution to the 2021 Cochlear Implant Hackathon

This solution relies hevily on the DEAP package. Reading the documentation there for genetic programming could be useful before understanding what's at play here.

Getting started:

Navigate to the gp_refactor folder and run the following actions.

conda create -n deep_vibe python=3.7
brew install portaudio
python -m pip install -r requirements.txt 

You've got the key ingredients now you can run the evolution to find a solution.

This is not fleshed out as a CLI tool, but rather just for scripting. As such, you're going to use run_evolution.py for all of your needs.

Basically the key variables here are

• wavefile_path this is a given wav file you want to optimze for.
• pop_size This is the size of the populations of solution you're going to run, the bigger, the longer it takes (3 is a good starting place if you want to watch it finish, rule of thumb - one minute per individual on a single core)
• end_gen How many iterations do you want to run the simulation for? More iterations takes more time.