概述
以前自己都利用别人搭好的工程，修改过来用，很少把模型搭建、导出模型、加载模型运行走一遍，搞了一遍才知道这个事情也不是那么简单的。
搭建模型和导出模型
参考《TensorFlow固化模型》，导出固化的模型有两种方式.
方式1：导出pb图结构和ckpt文件，然后用 freeze_graph 工具冻结生成一个pb（包含结构和参数）
在我的代码里测试了生成pb图结构和ckpt文件，但是没接着往下走，感觉有点麻烦。我用的是第二种方法。
注意我这里只在最后保存了一次ckpt，实际应该在训练中每隔一段时间就保存一次的。
saver = tf.train.Saver(max_to_keep=5)
#tf.train.write_graph(session.graph_def, FLAGS.model_dir, "nn_model.pbtxt", as_text=True)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
max_step = 2000
for i in range(max_step):
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})
print('test accuracy %g' % accuracy.eval(feed_dict={
x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0}))
# 保存pb和ckpt
print('save pb file and ckpt file')
tf.train.write_graph(sess.graph_def, graph_location, "graph.pb",as_text=False)
checkpoint_path = os.path.join(graph_location, "model.ckpt")
saver.save(sess, checkpoint_path, global_step=max_step)
方式2：convert_variables_to_constants
我实际使用的就是这种方法。
看名字也知道，就是把变量转化为常量保存，这样就可以愉快的加载使用了。
注意这里需要指明保存的输出节点，我的输出节点为'out/fc2'（我猜测会根据输出节点的依赖推断哪些部分是训练用到的，推理时用不到）。关于输出节点的名字是有规律的，其中out是一个name_scope名字，fc2是op节点的名字。
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
max_step = 2000
for i in range(max_step):
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})
print('test accuracy %g' % accuracy.eval(feed_dict={
x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0}))
print('save frozen file')
pb_path = os.path.join(graph_location, 'frozen_graph.pb')
print('pb_path:{}'.format(pb_path))
# 固化模型
output_graph_def = convert_variables_to_constants(sess, sess.graph_def, output_node_names=['out/fc2'])
with tf.gfile.FastGFile(pb_path, mode='wb') as f:
f.write(output_graph_def.SerializeToString())
上述代码会在训练后把训练好的计算图和参数保存到frozen_graph.pb文件。后续就可以用这个模型来测试图片了。
方式2的完整训练和保存模型代码
主要看main函数就行。另外注意deepnn函数最后节点的名字。
"""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 os
from tensorflow.examples.tutorials.mnist import input_data
from tensorflow.python.framework.graph_util import convert_variables_to_constants
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='ratio')
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('out'):
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='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 main(_):
# Import data
mnist = input_data.read_data_sets(FLAGS.data_dir)
# Create the model
with tf.name_scope('input'):
x = tf.placeholder(tf.float32, [None, 784], name='x')
# Define loss and optimizer
y_ = tf.placeholder(tf.int64, [None])
# Build the graph for the deep net
y_conv, keep_prob = deepnn(x)
with tf.name_scope('loss'):
cross_entropy = tf.losses.sparse_softmax_cross_entropy(
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), y_)
correct_prediction = tf.cast(correct_prediction, tf.float32)
accuracy = tf.reduce_mean(correct_prediction)
graph_location = './model'
print('Saving graph to: %s' % graph_location)
train_writer = tf.summary.FileWriter(graph_location)
train_writer.add_graph(tf.get_default_graph())
saver = tf.train.Saver(max_to_keep=5)
#tf.train.write_graph(session.graph_def, FLAGS.model_dir, "nn_model.pbtxt", as_text=True)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
max_step = 2000
for i in range(max_step):
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})
print('test accuracy %g' % accuracy.eval(feed_dict={
x: mnist.test.images, y_: mnist.test.labels, keep_prob: 1.0}))
# save pb file and ckpt file
#print('save pb file and ckpt file')
#tf.train.write_graph(sess.graph_def, graph_location, "graph.pb", as_text=False)
#checkpoint_path = os.path.join(graph_location, "model.ckpt")
#saver.save(sess, checkpoint_path, global_step=max_step)
print('save frozen file')
pb_path = os.path.join(graph_location, 'frozen_graph.pb')
print('pb_path:{}'.format(pb_path))
output_graph_def = convert_variables_to_constants(sess, sess.graph_def, output_node_names=['out/fc2'])
with tf.gfile.FastGFile(pb_path, mode='wb') as f:
f.write(output_graph_def.SerializeToString())
if __name__ == '__main__':
parser = argparse.ArgumentParser()
parser.add_argument('--data_dir', type=str,
default='./data',
help='Directory for storing input data')
FLAGS, unparsed = parser.parse_known_args()
tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
加载模型进行推理
上一节已经训练并导出了frozen_graph.pb。
这一节把它运行起来。
加载模型
下方的代码用来加载模型。推理时计算图里共两个placeholder需要填充数据，一个是图片（这不废话吗），一个是drouout_ratio，drouout_ratio用一个常量作为输入，后续就只需要输入图片了。
graph_location = './model'
pb_path = os.path.join(graph_location, 'frozen_graph.pb')
print('pb_path:{}'.format(pb_path))
newInput_X = tf.placeholder(tf.float32, [None, 784], name="X")
drouout_ratio = tf.constant(1., name="drouout")
with open(pb_path, 'rb') as f:
graph_def = tf.GraphDef()
graph_def.ParseFromString(f.read())
output = tf.import_graph_def(graph_def,
input_map={'input/x:0': newInput_X, 'dropout/ratio:0':drouout_ratio},
return_elements=['out/fc2:0'])
input_map参数并不是必须的。如果不用input_map，可以在run之前用tf.get_default_graph().get_tensor_by_name获取tensor的句柄。但是我觉得这种方法不是很友好，我这里没用这种方法。
注意input_map里的tensor名字是和搭计算图时的name_scope和op名字有关的，而且后面要补一个‘：0'(这点我还没细究)。
同时要注意，newInput_X的形状是[None, 784]，第一维是batch大小，推理时和训练要一致。
（我用的是mnist图片，训练时每个bacth的形状是[batchsize, 784],每个图片是28x28）
运行模型
我是一张张图片单独测试的，运行模型之前先把图片变为[1, 784]，以符合newInput_X的维数。
with tf.Session( ) as sess:
file_list = os.listdir(test_image_dir)
# 遍历文件
for file in file_list:
full_path = os.path.join(test_image_dir, file)
print('full_path:{}'.format(full_path))
# 只要黑白的，大小控制在(28,28)
img = cv2.imread(full_path, cv2.IMREAD_GRAYSCALE )
res_img = cv2.resize(img,(28,28),interpolation=cv2.INTER_CUBIC)
# 变成长784的一维数据
new_img = res_img.reshape((784))
# 增加一个维度，变为 [1, 784]
image_np_expanded = np.expand_dims(new_img, axis=0)
image_np_expanded.astype('float32') # 类型也要满足要求
print('image_np_expanded shape:{}'.format(image_np_expanded.shape))
# 注意注意，我要调用模型了
result = sess.run(output, feed_dict={newInput_X: image_np_expanded})
# 出来的结果去掉没用的维度
result = np.squeeze(result)
print('result:{}'.format(result))
#print('result:{}'.format(sess.run(output, feed_dict={newInput_X: image_np_expanded})))
# 输出结果是长度为10（对应0-9）的一维数据，最大值的下标就是预测的数字
print('result:{}'.format( (np.where(result==np.max(result)))[0][0] ))
注意模型的输出是一个长度为10的一维数组，也就是计算图里全连接的输出。这里没有softmax，只要取最大值的下标即可得到结果。
输出结果：
full_path:./test_images/97_7.jpg image_np_expanded shape:(1, 784) result:[-1340.37145996 -283.72436523 1305.03320312 437.6053772 -413.69961548 -1218.08166504 -1004.83807373 1953.33984375 42.00457001 -504.43829346] result:7 full_path:./test_images/98_6.jpg image_np_expanded shape:(1, 784) result:[ 567.4041748 -550.20904541 623.83496094 -1152.56884766 -217.92695618 1033.45239258 2496.44750977 -1139.23620605 -5.64091825 -615.28491211] result:6 full_path:./test_images/99_9.jpg image_np_expanded shape:(1, 784) result:[ -532.26409912 -1429.47277832 -368.58096313 505.82876587 358.42163086 -317.48199463 -1108.6829834 1198.08752441 289.12286377 3083.52539062] result:9
加载模型进行推理的完整代码
import sys
import os
import cv2
import numpy as np
import tensorflow as tf
test_image_dir = './test_images/'
graph_location = './model'
pb_path = os.path.join(graph_location, 'frozen_graph.pb')
print('pb_path:{}'.format(pb_path))
newInput_X = tf.placeholder(tf.float32, [None, 784], name="X")
drouout_ratio = tf.constant(1., name="drouout")
with open(pb_path, 'rb') as f:
graph_def = tf.GraphDef()
graph_def.ParseFromString(f.read())
#output = tf.import_graph_def(graph_def)
output = tf.import_graph_def(graph_def,
input_map={'input/x:0': newInput_X, 'dropout/ratio:0':drouout_ratio},
return_elements=['out/fc2:0'])
with tf.Session( ) as sess:
file_list = os.listdir(test_image_dir)
# 遍历文件
for file in file_list:
full_path = os.path.join(test_image_dir, file)
print('full_path:{}'.format(full_path))
# 只要黑白的，大小控制在(28,28)
img = cv2.imread(full_path, cv2.IMREAD_GRAYSCALE )
res_img = cv2.resize(img,(28,28),interpolation=cv2.INTER_CUBIC)
# 变成长784的一维数据
new_img = res_img.reshape((784))
# 增加一个维度，变为 [1, 784]
image_np_expanded = np.expand_dims(new_img, axis=0)
image_np_expanded.astype('float32') # 类型也要满足要求
print('image_np_expanded shape:{}'.format(image_np_expanded.shape))
# 注意注意，我要调用模型了
result = sess.run(output, feed_dict={newInput_X: image_np_expanded})
# 出来的结果去掉没用的维度
result = np.squeeze(result)
print('result:{}'.format(result))
#print('result:{}'.format(sess.run(output, feed_dict={newInput_X: image_np_expanded})))
# 输出结果是长度为10（对应0-9）的一维数据，最大值的下标就是预测的数字
print('result:{}'.format( (np.where(result==np.max(result)))[0][0] ))
以上这篇tensorflow 20:搭网络,导出模型,运行模型的实例就是小编分享给大家的全部内容了，希望能给大家一个参考，也希望大家多多支持呐喊教程。
声明：本文内容来源于网络，版权归原作者所有，内容由互联网用户自发贡献自行上传，本网站不拥有所有权，未作人工编辑处理，也不承担相关法律责任。如果您发现有涉嫌版权的内容，欢迎发送邮件至：notice#nhooo.com（发邮件时，请将#更换为@）进行举报，并提供相关证据，一经查实，本站将立刻删除涉嫌侵权内容。