If you want to execute this notebook on your jupyter locally, navigate to the repository and enter pip3 install -r requirements.txt Then, simply open the imdb_bert.ipynb with your jupyter notebook.

Due to the long duration that the model takes to train, its weights are saved in the imdb_bert.h5 file. You won't have to wait one hour for the model to finish training ;)

Have fun

For more details, visit: https://github.com/amaiya/ktrain/blob/master/examples/text/IMDb-BERT.ipynb https://towardsdatascience.com/bert-text-classification-in-3-lines-of-code-using-keras-264db7e7a358