Easy-to-use VGGNet

1. Prepare your Dataset.
2. Modify setting.yaml file. (Check out the # write here comments section.)
3. Run python main.py.

1. model.fit(x, y, validation_data, epochs=30, batch_size=4)
   : Training function is reconstructed like tensorflow-keras style.

2. model.evaluate(model, dataloader, valid_iter=1, batch_size=1, criterion=None)

3. model.predict(test_images, use_cpu=False)