Cue is a chatbot that helps neuro-divergent people or people with social anxiety to practice having conversations in a low-stakes environment that feels safe, supportive, and realistic.

Clone the repository

git clone https://github.com/IanRPage/unit4533-swamphacks
cd unit4533-swamphacks

1. Navigate to the frontend folder

   cd frontend

2. Install dependencies

   npm i

3. Start the frontend

   npm run dev

1. Create and activate virtual environment

   python3 -m venv venv
   source venv/bin/activate

2. Install dependencies

   pip install -r requirements.txt

3. Navigate to backend folder and start server

   cd backend
   flask run

We already provide the pretrained model (model/emotion_model.pth) in the repository. To test it, follow steps 1 and 2 from the backend setup. And then run

cd model
python test_model.py

If you want to train the model yourself, follow these steps:

IMPORTANT: I trained the model using PyTorch 2.10.0 for Linux, with CUDA 13.0 support. If you're on a different OS or have a different CUDA version, you may need to adjust the training process accordingly. Go to the PyTorch website to figure out how to install the right version of PyTorch for your system (you may need to remove torch and torchvision if you already have them installed). Generally though, you should be able to do the following:

1. Get the training data (from root)

   sh getFER-2013.sh

2. Install Dependencies (after creating virtual environment)

   pip install -r requirements.txt

3. Go to model directory

   cd model

4. Run data_preprocessing.py

   python data_preprocessing.py

5. Train the model

   python model/train_model.py

6. Test model

   python test_model.py