run.py

import sys
import numpy as np

from model import ConvolutionalNet
from data_manager import load_data


def main(args):
  """
      Description: Create, train and test the conv net
  """

  # Get the data
  train_data, train_labels, test_data, test_labels = load_data(args[1])

  print(np.shape(train_data))

  # Build our ConvNet
  conv_net = ConvolutionalNet(
      int(args[5]),
      int(args[6]),
      np.shape(train_data)[1],
      np.shape(test_data[0])[1]
  )

  # Train the ConvNet
  conv_net.train(train_data, train_labels, int(args[2]), int(args[3]), float(args[4]))

  # Test the ConvNet
  conv_net.test(test_data, test_labels)


if __name__ == '__main__':
  # Usage: run.py [data_path] [batch_size] [iteration_count] [alpha] [kernel_count] [kernel_size]
  main(sys.argv)