Tensorflow implementation of some text classification models.

• Python 3.6
• Tensorflow == 1.14.0

• The datasets (imdb, yelp-13, yelp-14, yelp15) are collected and processed by Duyu Tang, and the resources can be downloaded here.
• The glove word embedding.

To preprocess the dataset (make sure all the data are in correct directories)

python data_helpers.py

To train the model (please check all the hyper-parameters)

python train.py

To evaluate the trained model

python eval.py

fastText, CNN, RNN w/o attention, RCNN, GatedNN, HAN, DMN