The Writing Fool

The Writing Fool is a writing style transfer creativity supporting tool.

How to Run it

Install the style27 anaconda environment using the environment.yml file provided in the root directory.
Add your Tensorflow model checkpoints of both predictive and style transfer text in a data folder under the root directory. For exact folder structure needed for this directory, please see the model related lines in server.py. Also add a preprocessing object of your preprocessed text, and the word embedding used by the model in the same data directory.
Choose which author's style you which to enable by commenting/uncommenting out the necessary model learning code in server.py. Current Writing Fool does not support showing multiple author's style at the same time.
With Flask installed and the style27 environment turned on, run python server.py command in terminal.
On the default setting you should be able to go to http://127.0.0.1:5000 to use/see the tool

First, train your model via the Language-RNN notebook in model, it'll save its models in the model/tmp directory
Then update rnn.py with the name of the model you want to use.
Install Flask, then run python rnn.py
On default setting, you need to send POST requests to http://127.0.0.1:8080 with a form body with the 'text' attribute set to the sentence you wish to complete and get a completed sentence in return.