我刚刚阅读了 Deep MNIST for Experts 教程并修改了 mnist_deep.py 代码以保存经过培训的模型
在创建会话之前saver = tf.train.Saver()
for循环训练模型后saver.save(sess, './mnist_deep_model', global_step=2000)。它似乎被正确保存，因为我在我的工作文件夹中得到了以下四个文件：
我还修改了 mnist_deep.py ，添加了以下两个函数，以便能够逐个测试各个测试图像上的模型：
def indexMax(list):
"""indexMax returns the index of the max element of the list."""
return list.index(max(list))
def identifyDigitInImage(sess, x, y_conv, keep_prob, image):
"""identifyDigitInImage apply the trained model to given image to identify the represented digit."""
result = sess.run(y_conv, {x:[image], keep_prob: 1.0})[0].tolist()
return indexMax(result)
我还在主函数的末尾添加了一个循环，在该循环中，我在测试集中随机选择一个测试图像，并尝试使用此函数将训练模型应用于每个测试图像。它似乎有效，因为我在这个测试循环中获得了相同的准确度：99.2％
然后我写了第二个程序： mnist_deep_restore_trained_model.py （也基于mnist_deep.py源代码）试图恢复以前保存的训练模型并将测试图像应用到它期望得到相同的精度。
当然，我删除了此程序创建，训练和测试模型所需的所有代码（deepnn()函数和所有相关函数，张量创建：x = tf.placeholder(tf.float32, [None, 784])，y_conv ，keep_prob = deepnn(x)，loss，optimizer和准确性内容......）我只是以这种方式恢复了已保存的模型:(会话开启后）
saver = tf.train.import_meta_graph('./mnist_deep_model-2000.meta')
saver.restore(sess, tf.train.latest_checkpoint('./'))
我还在会话开始时删除了全局变量初始化，因为应该从训练模型中恢复全局变量的值：
但是，为了能够应用模型以识别给定测试图像的数字（参考function identifyDigitInImage(sess, x, y_conv, keep_prob, image)），我仍然需要获得Tensor变量 x，y_conv和keep_prob 。所以我从磁盘恢复模型后添加了以下几行：
graph = tf.get_default_graph()
x = graph.get_tensor_by_name("x:0")
keep_prob = graph.get_tensor_by_name("keep_prob:0")
y_conv = graph.get_tensor_by_name("y_conv:0")
最后，我还在第二个程序的末尾添加了与mnist_deep.py相同的测试循环，期望从这个恢复的模型得到相同的结果......
不幸的是，第一次调用get_tensor_by_name（）时出现异常：
x = graph.get_tensor_by_name("x:0")
KeyError: "The name 'x:0' refers to a Tensor which does not exist. The operation 'x', does not exist in the graph."
其他get_tensor_by_name()调用也会引发同样的异常。
我做错了什么？为什么不能以这种方式获得这些张量？
以下是我的完整 mnist_deep.py 源代码：
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""A deep MNIST classifier using convolutional layers.
See extensive documentation at
https://www.tensorflow.org/get_started/mnist/pros
"""
# Disable linter warnings to maintain consistency with tutorial.
# pylint: disable=invalid-name
# pylint: disable=g-bad-import-order
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import sys
import tempfile
import random
from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf
FLAGS = None
def deepnn(x):
"""deepnn builds the graph for a deep net for classifying digits.
Args:
x: an input tensor with the dimensions (N_examples, 784), where 784 is the
number of pixels in a standard MNIST image.
Returns:
A tuple (y, keep_prob). y is a tensor of shape (N_examples, 10), with values
equal to the logits of classifying the digit into one of 10 classes (the
digits 0-9). keep_prob is a scalar placeholder for the probability of
dropout.
"""
# Reshape to use within a convolutional neural net.
# Last dimension is for "features" - there is only one here, since images are
# grayscale -- it would be 3 for an RGB image, 4 for RGBA, etc.
with tf.name_scope('reshape'):
x_image = tf.reshape(x, [-1, 28, 28, 1])
# First convolutional layer - maps one grayscale image to 32 feature maps.
with tf.name_scope('conv1'):
W_conv1 = weight_variable([5, 5, 1, 32])
b_conv1 = bias_variable([32])
h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)
# Pooling layer - downsamples by 2X.
with tf.name_scope('pool1'):
h_pool1 = max_pool_2x2(h_conv1)
# Second convolutional layer -- maps 32 feature maps to 64.
with tf.name_scope('conv2'):
W_conv2 = weight_variable([5, 5, 32, 64])
b_conv2 = bias_variable([64])
h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)
# Second pooling layer.
with tf.name_scope('pool2'):
h_pool2 = max_pool_2x2(h_conv2)
# Fully connected layer 1 -- after 2 round of downsampling, our 28x28 image
# is down to 7x7x64 feature maps -- maps this to 1024 features.
with tf.name_scope('fc1'):
W_fc1 = weight_variable([7 * 7 * 64, 1024])
b_fc1 = bias_variable([1024])
h_pool2_flat = tf.reshape(h_pool2, [-1, 7*7*64])
h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)
# Dropout - controls the complexity of the model, prevents co-adaptation of
# features.
with tf.name_scope('dropout'):
keep_prob = tf.placeholder(tf.float32)
h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)
# Map the 1024 features to 10 classes, one for each digit
with tf.name_scope('fc2'):
W_fc2 = weight_variable([1024, 10])
b_fc2 = bias_variable([10])
y_conv = tf.matmul(h_fc1_drop, W_fc2) + b_fc2
return y_conv, keep_prob
def conv2d(x, W):
"""conv2d returns a 2d convolution layer with full stride."""
return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')
def max_pool_2x2(x):
"""max_pool_2x2 downsamples a feature map by 2X."""
return tf.nn.max_pool(x, ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1], padding='SAME')
def weight_variable(shape):
"""weight_variable generates a weight variable of a given shape."""
initial = tf.truncated_normal(shape, stddev=0.1)
return tf.Variable(initial)
def bias_variable(shape):
"""bias_variable generates a bias variable of a given shape."""
initial = tf.constant(0.1, shape=shape)
return tf.Variable(initial)
def indexMax(list):
"""indexMax returns the index of the max element of the list."""
return list.index(max(list))
def identifyDigitInImage(sess, x, y_conv, keep_prob, image):
"""identifyDigitInImage apply the trained model to given image to identify the represented digit."""
result = sess.run(y_conv, {x:[image], keep_prob: 1.0})[0].tolist()
return indexMax(result)
def main(_):
# Import data
mnist = input_data.read_data_sets(FLAGS.data_dir, one_hot=True)
# Create the model
x = tf.placeholder(tf.float32, [None, 784])
# Define loss and optimizer
y_ = tf.placeholder(tf.float32, [None, 10])
# Build the graph for the deep net
y_conv, keep_prob = deepnn(x)
with tf.name_scope('loss'):
cross_entropy = tf.nn.softmax_cross_entropy_with_logits(labels=y_,
logits=y_conv)
cross_entropy = tf.reduce_mean(cross_entropy)
with tf.name_scope('adam_optimizer'):
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
with tf.name_scope('accuracy'):
correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))
correct_prediction = tf.cast(correct_prediction, tf.float32)
accuracy = tf.reduce_mean(correct_prediction)
#graph_location = tempfile.mkdtemp()
#print('Saving graph to: %s' % graph_location)
#train_writer = tf.summary.FileWriter(graph_location)
#train_writer.add_graph(tf.get_default_graph())
# Prepare a saver to save the trained model:
saver = tf.train.Saver()
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
# Save the untrained model:
saver.save(sess, './mnist_deep_model')
# Train the model:
for i in range(20000):
batch = mnist.train.next_batch(50)
if i % 100 == 0:
train_accuracy = accuracy.eval(feed_dict={
x: batch[0], y_: batch[1], keep_prob: 1.0})
print('step %d, training accuracy %g' % (i, train_accuracy))
train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.5})
# Save the trained model:
saver.save(sess, './mnist_deep_model', global_step=2000)
# Display the test accuracy:
print('test accuracy %g' % accuracy.eval(feed_dict={
x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0}))
# Now try to apply the model to randomly choosen test images, one by one:
stop = False
count = 0
ok_count = 0
while not stop:
# Choosing a test image index:
test_image_index = random.randint(0, len(mnist.test.images) - 1)
test_image = mnist.test.images[test_image_index]
# Applying the trained model to identify the digit from the test image:
identified_digit = identifyDigitInImage(sess, x, y_conv, keep_prob, test_image)
# Display the identified digit:
print("The written digit on the given image has been identified as a {}".format(identified_digit))
# Check the expected_digit from the test label of the choosen test image:
expected_digit = indexMax(mnist.test.labels[test_image_index].tolist())
# Display the expected digit:
print("Actually, the digit is a {}".format(expected_digit))
# Count the correctly identified digits:
if identified_digit == expected_digit:
ok_count += 1
# Stop the loop after 10000 iterations
count += 1
stop = count == 10000
# Display the measured accuracy during the test loop:
print("2nd Test accuracy = {}%".format(100 * (ok_count / count)))
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--data_dir', type=str,
default='/tmp/tensorflow/mnist/input_data',
help='Directory for storing input data')
FLAGS, unparsed = parser.parse_known_args()
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
这里是我的完整 mnist_deep_restore_trained_model.py 源代码：
# Copyright 2015 The TensorFlow Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# =============================================================================
"""A deep MNIST classifier using convolutional layers.
See extensive documentation at
https://www.tensorflow.org/get_started/mnist/pros
"""
# Disable linter warnings to maintain consistency with tutorial.
# pylint: disable=invalid-name
# pylint: disable=g-bad-import-order
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import sys
import random
from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf
FLAGS = None
def indexMax(list):
"""indexMax returns the index of the max element of the list."""
return list.index(max(list))
def identifyDigitInImage(sess, x, y_conv, keep_prob, image):
"""identifyDigitInImage apply the trained model to given image to identify the represented digit."""
result = sess.run(y_conv, {x:[image], keep_prob: 1.0})[0].tolist()
return indexMax(result)
def main(_):
# Import data
mnist = input_data.read_data_sets(FLAGS.data_dir, one_hot=True)
with tf.Session() as sess:
# Restoring the trained model previously saved:
saver = tf.train.import_meta_graph('./mnist_deep_model-2000.meta')
saver.restore(sess, tf.train.latest_checkpoint('./'))
# Trying to get back some required tensors variables from the restored graph:
graph = tf.get_default_graph()
x = graph.get_tensor_by_name("x:0")
# This call fails with the following exception:
# KeyError: "The name 'x:0' refers to a Tensor which does not exist. The operation 'x', does not exist in the graph."
keep_prob = graph.get_tensor_by_name("keep_prob:0")
y_conv = graph.get_tensor_by_name("y_conv:0")
# Now try to apply the model to randomly choosen test images, one by one:
stop = False
count = 0
ok_count = 0
while not stop:
# Choosing a test image index:
test_image_index = random.randint(0, len(mnist.test.images) - 1)
test_image = mnist.test.images[test_image_index]
# Applying the trained model to identify the digit from the test image:
identified_digit = identifyDigitInImage(sess, x, y_conv, keep_prob, test_image)
# Display the identified digit:
print("The written digit on the given image has been identified as a {}".format(identified_digit))
# Check the expected_digit from the test label of the choosen test image:
expected_digit = indexMax(mnist.test.labels[test_image_index].tolist())
# Display the expected digit:
print("Actually, the digit is a {}".format(expected_digit))
# Count the correctly identified digits:
if identified_digit == expected_digit:
ok_count += 1
# Stop the loop after 10000 iterations
count += 1
stop = count == 10000
# Display the measured accuracy during the test loop:
print("Test accuracy = {}%".format(100 * (ok_count / count)))
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--data_dir', type=str,
default='/tmp/tensorflow/mnist/input_data',
help='Directory for storing input data')
FLAGS, unparsed = parser.parse_known_args()
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
答案 0 :(得分：3)
您没有为占位符指定明确的名称：
# Create the model
x = tf.placeholder(tf.float32, [None, 784])
# Define loss and optimizer
y_ = tf.placeholder(tf.float32, [None, 10])
...因此，它们在已保存的图表中被命名为Placeholder和Placeholder_1，因此出现错误。将此代码更改为：
# Create the model
x = tf.placeholder(tf.float32, [None, 784], name='x')
# Define loss and optimizer
y_ = tf.placeholder(tf.float32, [None, 10], name='y')
...同样适用于keep_prob和y_conv（使用tf.add为+ op命名。顺便说一句，为所有变量和关键操作命名并使用scopes总是一个好主意。重新训练模型后，mnist_deep_restore_trained_model.py应该可以正常工作。
答案 1 :(得分：1)
感谢您的帮助Maxim。它现在正常工作。
这是我修复的mnist_deep.py代码：
# Disable linter warnings to maintain consistency with tutorial.
# pylint: disable=invalid-name
# pylint: disable=g-bad-import-order
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import sys
import tempfile
import random
from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf
FLAGS = None
def deepnn(x):
"""deepnn builds the graph for a deep net for classifying digits.
Args:
x: an input tensor with the dimensions (N_examples, 784), where 784 is the
number of pixels in a standard MNIST image.
Returns:
A tuple (y, keep_prob). y is a tensor of shape (N_examples, 10), with values
equal to the logits of classifying the digit into one of 10 classes (the
digits 0-9). keep_prob is a scalar placeholder for the probability of
dropout.
"""
# Reshape to use within a convolutional neural net.
# Last dimension is for "features" - there is only one here, since images are
# grayscale -- it would be 3 for an RGB image, 4 for RGBA, etc.
with tf.name_scope('reshape'):
x_image = tf.reshape(x, [-1, 28, 28, 1])
# First convolutional layer - maps one grayscale image to 32 feature maps.
with tf.name_scope('conv1'):
W_conv1 = weight_variable([5, 5, 1, 32])
b_conv1 = bias_variable([32])
h_conv1 = tf.nn.relu(conv2d(x_image, W_conv1) + b_conv1)
# Pooling layer - downsamples by 2X.
with tf.name_scope('pool1'):
h_pool1 = max_pool_2x2(h_conv1)
# Second convolutional layer -- maps 32 feature maps to 64.
with tf.name_scope('conv2'):
W_conv2 = weight_variable([5, 5, 32, 64])
b_conv2 = bias_variable([64])
h_conv2 = tf.nn.relu(conv2d(h_pool1, W_conv2) + b_conv2)
# Second pooling layer.
with tf.name_scope('pool2'):
h_pool2 = max_pool_2x2(h_conv2)
# Fully connected layer 1 -- after 2 round of downsampling, our 28x28 image
# is down to 7x7x64 feature maps -- maps this to 1024 features.
with tf.name_scope('fc1'):
W_fc1 = weight_variable([7 * 7 * 64, 1024])
b_fc1 = bias_variable([1024])
h_pool2_flat = tf.reshape(h_pool2, [-1, 7*7*64])
h_fc1 = tf.nn.relu(tf.matmul(h_pool2_flat, W_fc1) + b_fc1)
# Dropout - controls the complexity of the model, prevents co-adaptation of
# features.
with tf.name_scope('dropout'):
keep_prob = tf.placeholder(tf.float32, name='keep_prob')
h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)
# Map the 1024 features to 10 classes, one for each digit
with tf.name_scope('fc2'):
W_fc2 = weight_variable([1024, 10])
b_fc2 = bias_variable([10])
y_conv = tf.add(tf.matmul(h_fc1_drop, W_fc2), b_fc2, name='y_conv')
return y_conv, keep_prob
def conv2d(x, W):
"""conv2d returns a 2d convolution layer with full stride."""
return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')
def max_pool_2x2(x):
"""max_pool_2x2 downsamples a feature map by 2X."""
return tf.nn.max_pool(x, ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1], padding='SAME')
def weight_variable(shape):
"""weight_variable generates a weight variable of a given shape."""
initial = tf.truncated_normal(shape, stddev=0.1)
return tf.Variable(initial)
def bias_variable(shape):
"""bias_variable generates a bias variable of a given shape."""
initial = tf.constant(0.1, shape=shape)
return tf.Variable(initial)
def indexMax(list):
"""indexMax returns the index of the max element of the list."""
return list.index(max(list))
def identifyDigitInImage(sess, x, y_conv, keep_prob, image):
"""identifyDigitInImage apply the trained model to given image to identify the represented digit."""
result = sess.run(y_conv, {x:[image], keep_prob: 1.0})[0].tolist()
return indexMax(result)
def main(_):
# Import data
mnist = input_data.read_data_sets(FLAGS.data_dir, one_hot=True)
# Create the model
x = tf.placeholder(tf.float32, [None, 784], name = 'x')
# Define loss and optimizer
y_ = tf.placeholder(tf.float32, [None, 10], name = 'y_')
# Build the graph for the deep net
y_conv, keep_prob = deepnn(x)
with tf.name_scope('loss'):
cross_entropy = tf.nn.softmax_cross_entropy_with_logits(labels=y_,
logits=y_conv)
cross_entropy = tf.reduce_mean(cross_entropy)
with tf.name_scope('adam_optimizer'):
train_step = tf.train.AdamOptimizer(1e-4).minimize(cross_entropy)
with tf.name_scope('accuracy'):
correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))
correct_prediction = tf.cast(correct_prediction, tf.float32)
accuracy = tf.reduce_mean(correct_prediction)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
# Train the model:
for i in range(20000):
batch = mnist.train.next_batch(50)
if i % 100 == 0:
train_accuracy = accuracy.eval(feed_dict={
x: batch[0], y_: batch[1], keep_prob: 1.0})
print('step %d, training accuracy %g' % (i, train_accuracy))
train_step.run(feed_dict={x: batch[0], y_: batch[1], keep_prob: 0.5})
# Save the trained model:
saver = tf.train.Saver()
saver.save(sess, './mnist_deep_model', global_step=2000)
# Display the test accuracy:
print('test accuracy %g' % accuracy.eval(feed_dict={
x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0}))
# Now try to apply the model to randomly choosen test images, one by one:
count = 0
ok_count = 0
while count < 10000:
# Choosing a test image index:
test_image_index = random.randint(0, len(mnist.test.images) - 1)
test_image = mnist.test.images[test_image_index]
# Applying the trained model to identify the digit from the test image:
identified_digit = identifyDigitInImage(sess, x, y_conv, keep_prob, test_image)
# Display the identified digit:
print("The written digit on the given image has been identified as a {}".format(identified_digit))
# Check the expected_digit from the test label of the choosen test image:
expected_digit = indexMax(mnist.test.labels[test_image_index].tolist())
# Display the expected digit:
print("Actually, the digit is a {}".format(expected_digit))
# Count the correctly identified digits:
if identified_digit == expected_digit:
ok_count += 1
# Stop the loop after 10000 iterations
count += 1
# Display the measured accuracy during the test loop:
print("2nd Test accuracy = {}%".format(100 * (ok_count / count)))
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--data_dir', type=str,
default='/tmp/tensorflow/mnist/input_data',
help='Directory for storing input data')
FLAGS, unparsed = parser.parse_known_args()
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
和相应的固定mnist_deep_restore_train_model.py代码：
# Disable linter warnings to maintain consistency with tutorial.
# pylint: disable=invalid-name
# pylint: disable=g-bad-import-order
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import argparse
import sys
import random
from tensorflow.examples.tutorials.mnist import input_data
import tensorflow as tf
FLAGS = None
def indexMax(list):
"""indexMax returns the index of the max element of the list."""
return list.index(max(list))
def identifyDigitInImage(sess, x, y_conv, keep_prob, image):
"""identifyDigitInImage apply the trained model to given image to identify the represented digit."""
result = sess.run(y_conv, {x:[image], keep_prob: 1.0})[0].tolist()
return indexMax(result)
def main(_):
# Import data
mnist = input_data.read_data_sets(FLAGS.data_dir, one_hot=True)
with tf.Session() as sess:
# Restoring the trained model previously saved:
saver = tf.train.import_meta_graph('./mnist_deep_model-2000.meta')
saver.restore(sess, tf.train.latest_checkpoint('./'))
# Trying to get back some required tensors variables from the restored graph:
graph = tf.get_default_graph()
x = graph.get_tensor_by_name("x:0")
keep_prob = graph.get_tensor_by_name("dropout/keep_prob:0")
y_conv = graph.get_tensor_by_name("fc2/y_conv:0")
# Now try to apply the model to randomly choosen test images, one by one:
count = 0
ok_count = 0
while count < 10000:
# Choosing a test image index:
test_image_index = random.randint(0, len(mnist.test.images) - 1)
test_image = mnist.test.images[test_image_index]
# Applying the trained model to identify the digit from the test image:
identified_digit = identifyDigitInImage(sess, x, y_conv, keep_prob, test_image)
# Display the identified digit:
print("The written digit on the given image has been identified as a {}".format(identified_digit))
# Check the expected_digit from the test label of the choosen test image:
expected_digit = indexMax(mnist.test.labels[test_image_index].tolist())
# Display the expected digit:
print("Actually, the digit is a {}".format(expected_digit))
# Count the correctly identified digits:
if identified_digit == expected_digit:
ok_count += 1
# Stop the loop after 10000 iterations
count += 1
# Display the measured accuracy during the test loop:
print("Test accuracy = {}%".format(100 * (ok_count / count)))
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--data_dir', type=str,
default='/tmp/tensorflow/mnist/input_data',
help='Directory for storing input data')
FLAGS, unparsed = parser.parse_known_args()
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)