Disclaimer: this issue will be a little bit tricky as the code is kind of a mess.
The folder with the training code is ./Docker/train/
The code to train the neural network is contained within a jupyter notebook CNN_train_pred.ipynb.
There are several other .py files but they can likely be discarded.
The docker file for this service should start up a jupyter notebook and be started from a docker-compose that can connect to the mongodb service to retrieve training data.
Users will be able to train models from the notebook.
Status
Assigning to @ Please use this branch https://github.com/N2ITN/are-you-fake-news/tree/develop-dockerize This issue depends on https://github.com/Data4Democracy/are-you-fake-news/issues/19 being completed.
Issue
Disclaimer: this issue will be a little bit tricky as the code is kind of a mess. The folder with the training code is
./Docker/train/The code to train the neural network is contained within a jupyter notebook CNN_train_pred.ipynb. There are several other .py files but they can likely be discarded. The docker file for this service should start up a jupyter notebook and be started from a docker-compose that can connect to the mongodb service to retrieve training data. Users will be able to train models from the notebook.Tasks
Reminder: this issue cannot proceed until this issue is completed: https://github.com/Data4Democracy/are-you-fake-news/issues/19
pyfiles that are not needed by the notebook.