This repository contains code to train a rhetorical figure detector.
All data is extracted from here: rhetoricon.uwaterloo.ca/admin/export
Store the 3 json files in the input directory.
Create a virtual environment (python3 -m venv .venv). Activate it by running source .venv/bin/activate.
Then, run pip install -r requirements.txt.
Then, run extract.py, english.py, and phonetize.py in order to generate the CSV file usable for training.
Then, run all cells in train.ipynb.