- hand categorise some images, and put this data into training_data_redacted.py in the format shown
- create the model and do some initial training
python train_model.py - you can now ask the model to categorise new images using the check_image() function
- make a database, and a table with the following columns: hash (str), url (str), username (str), date (str), version (int), tag (str), strength (float)
- put all the urls of images that weren't in the training data (ie, the ones you want to categorise) in a csv
- choose a version number, eg 2
- Run this to categorise all images in the csv and put them in the DB
python database_putter.py 'non_training_images.csv' 2