README under construction...
The file structure for the training data should look like this:
vision
│
└───data
│
└───train
│ │
│ └───label1
│ │ │
│ │ └───img1.jpeg
│ │ │
│ │ └───img2.jpeg
│ │ │
│ │ └───...
│ │
│ └───label2
│ │
│ └───...
│
└───test
│
└───label1
│
└───label2
│
└───...
Once that is set up all there is to do is to run train.py
and wait for it to finish.