For partII, we implemented two models and include our source code in q2 folder. Simply extract the data to the root folder (same place as q2 and q3) and run "python train_and_test.py" in folder TextRnn/TextAttBiRNN to start training the model.
For partIII,
We have provided couple scripts to test different parameter setups for the TextRNN model.
For customized parameters:
python train_and_test.py epochs, learning_rate, batch_size, model_name layer_size
Example input:
python train_and_test.py 20 0.001 256 TextRNN 128