This project looks at usage of poiter generators for generating random text based off news coverage about a particular subject The news coverage uses GDELT api to scrape the news articles that deal with some predetermined keywords with source country and language. The data is stored in a mongo database and then a LSTM based generator network would be trained on this text.
To install I recommend using virtualenv to sandbox the installation. To install virtualenv forllow this link
Once Virtual env is stalled, create an environment by
virtualenv .envActivate the environment by
source .env/bin/activateafter creating environment, install dependencies for this project
pip install -r requirements.txtThese steps should set you up for running the project.
depending on where the mongo is being run, you will need to point the connector to that IP and port. Currently this is hardcoded, but will be configurable soon.