Run on Google Colab
Each file has a different method of prediction
First, run the code which will process your data into the format needed for an LSTM
Then run the sequence-to-sequence model(s) using the eye and hand fixation sequences as the input/output