Now, let's train our network:
for i in range(1, 30001): sess.run(train_op1, feed_dict={feature_data: train_features.values, actual_classes: train_labels.values.reshape(len(train_labels.values), 2)}) if i % 5000 == 0: print(i, sess.run(accuracy, feed_dict={feature_data: train_features.values, actual_classes: train_labels.values.reshape(len(train_labels.values), 2)}))