shape=[None, IMAGE_PIXELS],
name='images')
labels_placeholder = tf.placeholder(tf.int64,
shape=[None],
name='image-labels')
logits = two_layer_fc.inference(images_placeholder,
IMAGE_PIXELS,
FLAGS.hidden1,
CLASSES,
reg_constant=FLAGS.reg_constant)
global_step = tf.Variable(0, name="global_step", trainable=False)
accuracy = two_layer_fc.evaluation(logits, labels_placeholder)
saver = tf.train.Saver()
with tf.Session() as sess:
ckpt = tf.train.get_checkpoint_state(FLAGS.train_dir)
if ckpt and ckpt.model_checkpoint_path:
print('Restoring variables from checkpoint')
saver.restore(sess, ckpt.model_checkpoint_path)
current_step = tf.train.global_step(sess, global_step)
print('Current step: {}'.format(current_step))
print('Test accuracy {:g}'.format(
accuracy.eval(
feed_dict={
images_placeholder: data_sets['images_test'],
beginTime = time.time()
data_sets = data_helpers.load_data()
images_placeholder = tf.placeholder(tf.float32, shape=[None, IMAGE_PIXELS],
name='images')
labels_placeholder = tf.placeholder(tf.int64, shape=[None], name='image-labels')
logits = two_layer_fc.inference(images_placeholder, IMAGE_PIXELS,
FLAGS.hidden1, CLASSES, reg_constant=FLAGS.reg_constant)
global_step = tf.Variable(0, name="global_step", trainable=False)
accuracy = two_layer_fc.evaluation(logits, labels_placeholder)
saver = tf.train.Saver()
with tf.Session() as sess:
ckpt = tf.train.get_checkpoint_state(FLAGS.train_dir)
if ckpt and ckpt.model_checkpoint_path:
print('Restoring variables from checkpoint')
saver.restore(sess, ckpt.model_checkpoint_path)
current_step = tf.train.global_step(sess, global_step)
print('Current step: {}'.format(current_step))
print('Test accuracy {:g}'.format(accuracy.eval(
feed_dict={ images_placeholder: data_sets['images_test'],
labels_placeholder: data_sets['labels_test']}
)))