SemEval Task 2: Predicting Valence-Arousal Score for Longitudinal Texts

Environment set up

1. Create a new python virtual environment on your machine and activate

2. Install PyTorch using the official instruction: https://pytorch.org/

3. Run pip install -r requirements.txt to install the rest of the libraries

Running the inference

Subtask 1

python BiLSTM_predict.py --valence_model models/subtask1_val.pt --arousal_model models/subtask1_aro.pt --input .\data\test_subtask1.csv --output pred_subtask1.csv

Subtask 2a

python tft_subtask2a.py

Subtask 2b

python lm_rnn_subtask2b.py