run_id = log_run_start('Train', hyper_dict, VERSION_DESCRIPTION)
print()
epoch_train_time_list = []
training_losses = []
validate_accuracies = []
for i in range(epochs):
t1 = time()
X_TRAIN, Y_TRAIN = shuffle(X_TRAIN, Y_TRAIN)
for offset in range(0, num_examples, batch_size):
end = offset + batch_size
batch_x, batch_y = X_TRAIN[offset:end], Y_TRAIN[offset:end]
_, loss = sess.run((training_operation, loss_operation), \
feed_dict={x: batch_x, y: batch_y, keep_prob: 0.5})
if TRAIN_RATIO < 1.0:
validation_accuracy = evaluate(X_VALID, Y_VALID, batch_size)
else:
validation_accuracy = None
print("EPOCH {} ...".format(i+1))
print("Training Loss = {:.3f}".format(loss))
if validation_accuracy is not None:
print("Validation Accuracy = {:.3f}".format(validation_accuracy))
print()
epoch_train_time_list.append(time() - t1)
training_losses.append(loss)
validate_accuracies.append(validation_accuracy)
saver.save(sess, save_file)
print("Model saved")
log_run_end(run_id, validation_accuracy, loss)
print()
import tensorflow as tf
from traffic_sign_model import evaluate
i = 1
fname = 'sign{:02d}.png'.format(i)
images = mpimg.imread(fname)
images = np.expand_dims(images, axis=0)
print("Image shape", images.shape)
for i in range(2,7):
fname = 'sign{:02d}.png'.format(i)
img = mpimg.imread(fname)
img = np.expand_dims(img, axis=0)
print("image %s shape" % (i), img.shape)
images = np.append(images, img, axis=0)
#printing out some stats and plotting
print('The extra data shape is:', images.shape)
y_extra = [14,28,13,27,17,26]
saver = tf.train.Saver()
save_file = 'traffic_signs.ckpt'
# Evaluate the model with test data
with tf.Session() as sess:
saver.restore(sess, save_file)
test_accuracy = evaluate(images, y_extra, 6)
print("Test Accuracy = {:.3f}".format(test_accuracy))