This repository contains the code for training and evaluating an intent classification model for the Saint Aveline project. This model is designed to take user commands and classify them into predefined intents to facilitate appropriate responses.
The model is trained in Python, but is consumed by C# and Unity.
- Python 3.11.x is being used for development
$> python -m venv venv
$> source venv/bin/activate # On Windows use `venv\Scripts\activate`This may be needed on some systems:
pip install --index-url https://download.pytorch.org/whl/cpu torch$> pip install -r requirements.txt$> python train.py$> python test.py$> python python train.py # outputs intent_model.pt
$> python export_model.py # converts intent_model.pt to intent_model.onnx
$> python export_vcab_json.py # exports the vocabulary to a JSON file