def eval()

in content/advanced/420_kubeflow/kubeflow.files/mnist-tensorflow-docker.py [0:0]

• 3 lines of code
• 1 McCabe index (conditional complexity)

def eval(model, test_images, test_labels):
  test_loss, test_acc = model.evaluate(test_images, test_labels)
  print('\nTest accuracy: {}'.format(test_acc))