TensorFlow基础笔记(13) Mobilenet训练测试mnist数据

主要是四个文件

mnist_train.py

#coding: utf-8
import os
 
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data
 
import mnist_inference
 
BATCH_SIZE = 100
LEARNING_RATE_BASE = 0.8
LEARNING_RATE_DECAY = 0.99
REGULARAZTION_RATE = 0.0001
TRAINING_STEPS =10000
MOVING_AVERAGE_DECAY = 0.99
MODEL_SAVE_PATH = "./mobilenet_v1_model/"
MODEL_NAME = "model.ckpt"
channels = 1
 
def train_MLP(mnist):
    x = tf.placeholder(tf.float32, [None, mnist_inference.INPUT_NODE], name='x-input')
    y_ = tf.placeholder(tf.float32, [None, mnist_inference.OUTPUT_NODE], name='y-input')
    regularizer = tf.contrib.layers.l2_regularizer(REGULARAZTION_RATE)
 
    y = mnist_inference.inference_MLP(x, regularizer)
 
    global_step = tf.Variable(0, trainable=False)
 
    variable_averages = tf.train.ExponentialMovingAverage(MOVING_AVERAGE_DECAY, global_step)
    variable_averages_op = variable_averages.apply(tf.trainable_variables())
    cross_entropy = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=y, labels=tf.argmax(y_, 1))
    cross_entropy_mean = tf.reduce_mean(cross_entropy)
    loss = cross_entropy_mean + tf.add_n(tf.get_collection('losses'))
    learning_rate = tf.train.exponential_decay(LEARNING_RATE_BASE, global_step, mnist.train.num_examples / BATCH_SIZE, LEARNING_RATE_DECAY)
    train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss, global_step=global_step)
 
    with tf.control_dependencies([train_step, variable_averages_op]):
        train_op = tf.no_op(name='train')
 
    saver = tf.train.Saver()
    with tf.Session() as sess:
        tf.initialize_all_variables().run()
 
        for i in range(TRAINING_STEPS):
            xs, ys = mnist.train.next_batch(BATCH_SIZE)
            _, loss_value, step = sess.run([train_op, loss, global_step], feed_dict={x: xs, y_: ys})
 
            if i % 1000 == 0:
                print("After %d training step(s), loss on training batch is %g." % (step, loss_value))
                # print os.path.join(MODEL_SAVE_PATH, MODEL_NAME)
                saver.save(sess, os.path.join(MODEL_SAVE_PATH, MODEL_NAME), global_step=global_step)
 
def train_mobilenet(mnist):
    x = tf.placeholder(tf.float32, [None, mnist_inference.INPUT_NODE], name='x-input')
    y_ = tf.placeholder(tf.float32, [None, mnist_inference.OUTPUT_NODE], name='y-input')
    regularizer = tf.contrib.layers.l2_regularizer(REGULARAZTION_RATE)
 
    #mobilenet 把输入数据变成与w矩阵同纬度的
    x_image = tf.reshape(x, [-1,28,28,1])
    x_image = tf.image.resize_image_with_crop_or_pad(x_image, 28*4,28*4)
    y = mnist_inference.inference_mobilenet(x_image, regularizer)
 
    global_step = tf.Variable(0, trainable=False)
 
    variable_averages = tf.train.ExponentialMovingAverage(MOVING_AVERAGE_DECAY, global_step)
    variable_averages_op = variable_averages.apply(tf.trainable_variables())
    cross_entropy = tf.nn.sparse_softmax_cross_entropy_with_logits(logits=y, labels=tf.argmax(y_, 1))
    cross_entropy_mean = tf.reduce_mean(cross_entropy)
    loss = cross_entropy_mean #+ tf.add_n(tf.get_collection('losses'))
    learning_rate = tf.train.exponential_decay(LEARNING_RATE_BASE, global_step, mnist.train.num_examples / BATCH_SIZE, LEARNING_RATE_DECAY)
    train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss, global_step=global_step)
 
    with tf.control_dependencies([train_step, variable_averages_op]):
        train_op = tf.no_op(name='train')
 
    saver = tf.train.Saver()
    with tf.Session() as sess:
        tf.initialize_all_variables().run()
 
        for i in range(TRAINING_STEPS):
            xs, ys = mnist.train.next_batch(BATCH_SIZE)
            _, loss_value, step = sess.run([train_op, loss, global_step], feed_dict={x: xs, y_: ys})
 
            if i % 1000 == 0:
                print("After %d training step(s), loss on training batch is %g." % (step, loss_value))
                # print os.path.join(MODEL_SAVE_PATH, MODEL_NAME)
                saver.save(sess, os.path.join(MODEL_SAVE_PATH, MODEL_NAME), global_step=global_step)
            else:
                print("After %d training step(s), loss on training batch is %g." % (step, loss_value))
 
def main(argv=None):
    mnist = input_data.read_data_sets("../MNIST_data", one_hot=True)
    train_mobilenet(mnist)
 
if __name__ == '__main__':
    tf.app.run()

mnist_eval.py

#coding: utf-8
import time
import tensorflow as tf
from tensorflow.examples.tutorials.mnist import input_data
 
import mnist_inference
import mnist_train
 
#every 10 sec load the newest model
EVAL_INTERVAL_SECS = 10
 
def evaluate_MLP(mnist):
    with tf.Graph().as_default() as g:
        x = tf.placeholder(tf.float32, [None, mnist_inference.INPUT_NODE], name='x-input')
        y_ = tf.placeholder(tf.float32, [None, mnist_inference.OUTPUT_NODE], name='y-input')
        validate_feed = {x: mnist.validation.images, y_: mnist.validation.labels}
 
        y = mnist_inference.inference(x, None)
 
        correcgt_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))
        accuracy = tf.reduce_mean(tf.cast(correcgt_prediction, tf.float32))
 
        variable_averages = tf.train.ExponentialMovingAverage(mnist_train.MOVING_AVERAGE_DECAY)
        variable_to_restore = variable_averages.variables_to_restore()
        saver = tf.train.Saver(variable_to_restore)
 
        #while True:
        if 1:
            with tf.Session() as sess:
                ckpt = tf.train.get_checkpoint_state(mnist_train.MODEL_SAVE_PATH)
                if ckpt and ckpt.model_checkpoint_path:
                    #load the model
                    saver.restore(sess, ckpt.model_checkpoint_path)
                    global_step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]
                    accuracy_score = sess.run(accuracy, feed_dict=validate_feed)
                    print("After %s training steps, validation accuracy = %g" % (global_step, accuracy_score))
 
                else:
                    print('No checkpoint file found')
                    return
            #time.sleep(EVAL_INTERVAL_SECS)
 
def evaluate_mobilenet(mnist):
    with tf.Graph().as_default() as g:
        x = tf.placeholder(tf.float32, [None, mnist_inference.INPUT_NODE], name='x-input')
        y_ = tf.placeholder(tf.float32, [None, mnist_inference.OUTPUT_NODE], name='y-input')
 
        #mobilenet 把输入数据变成与w矩阵同纬度的
        x_image = tf.reshape(x, [-1,28,28,1])
        x_image = tf.image.resize_image_with_crop_or_pad(x_image, 28*4,28*4)
        y = mnist_inference.inference_mobilenet(x_image, None)
 
        correcgt_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))
        accuracy = tf.reduce_mean(tf.cast(correcgt_prediction, tf.float32))
 
        variable_averages = tf.train.ExponentialMovingAverage(mnist_train.MOVING_AVERAGE_DECAY)
        variable_to_restore = variable_averages.variables_to_restore()
        saver = tf.train.Saver(variable_to_restore)
 
        input  = mnist.validation.images
        label = mnist.validation.labels
        batch_size = 100
        TEST_STEPS = input.shape[0] / batch_size
        sum_accury = 0.0
        #while True:
        if 1:
            with tf.Session() as sess:
                ckpt = tf.train.get_checkpoint_state(mnist_train.MODEL_SAVE_PATH)
                if ckpt and ckpt.model_checkpoint_path:
                    #load the model
                    saver.restore(sess, ckpt.model_checkpoint_path)
                    global_step = ckpt.model_checkpoint_path.split('/')[-1].split('-')[-1]
                    for i in range(int(TEST_STEPS)):
                        input_batch = input[i*batch_size : (i + 1)*batch_size, :]
                        label_batch = label[i*batch_size : (i + 1)*batch_size, :]
                        validate_feed = {x: input_batch, y_: label_batch}
                        # 取出部分数据测试
                        accuracy_score = sess.run(accuracy, feed_dict=validate_feed)
                        sum_accury += accuracy_score
                        print("test %s batch steps, validation accuracy = %g" % (i, accuracy_score))
 
                else:
                    print('No checkpoint file found')
                    return
            #time.sleep(EVAL_INTERVAL_SECS)
        print("After %s training steps, all validation accuracy = %g" % (global_step, sum_accury / TEST_STEPS))
 
def main(argv=None):
    mnist = input_data.read_data_sets("../MNIST_data", one_hot=True)
    evaluate_mobilenet(mnist)
 
if __name__ == '__main__':
    tf.app.run()

mnist_inference.py

#coding: utf-8
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
 
import numpy as np
import tensorflow as tf
 
import mobilenet_v1
 
slim = tf.contrib.slim
 
#define the variables of nerual network
INPUT_NODE = 784
OUTPUT_NODE = 10
LAYER1_NODE = 500
 
def get_weight_variable(shape, regularizer):
    weights = tf.get_variable("weights", shape, initializer=tf.truncated_normal_initializer(stddev=0.1))
 
    if regularizer != None:
        tf.add_to_collection('losses', regularizer(weights))
 
    return weights
 
#define the forward network with MLPnet
def inference_MLP(input_tensor, regularizer):
    with tf.variable_scope('layer1'):
        weights = get_weight_variable([INPUT_NODE, LAYER1_NODE], regularizer)
        biases = tf.get_variable("biases", [LAYER1_NODE], initializer=tf.constant_initializer(0.0))
        layer1 = tf.nn.relu(tf.matmul(input_tensor, weights) + biases)
 
    with tf.variable_scope('layer2'):
        weights = get_weight_variable([LAYER1_NODE, OUTPUT_NODE], regularizer)
        biases = tf.get_variable("biases", [OUTPUT_NODE], initializer=tf.constant_initializer(0.0))
        layer2 = tf.matmul(layer1, weights) + biases
 
    return layer2
 
#define the forward network with mobilenet_v1
def inference_mobilenet(input_tensor, regularizer):
    #inputs = tf.random_uniform((batch_size, height, width, 3))
    with slim.arg_scope([slim.conv2d, slim.separable_conv2d],
                      normalizer_fn=slim.batch_norm):
        logits, end_points = mobilenet_v1.mobilenet_v1(
            input_tensor,
            num_classes=OUTPUT_NODE,
            dropout_keep_prob=0.8,
            is_training=True,
            min_depth=8,
            depth_multiplier=1.0,
            conv_defs=None,
            prediction_fn=tf.contrib.layers.softmax,
            spatial_squeeze=True,
            reuse=None,
            scope='MobilenetV1',
            global_pool=False
        )
 
    return logits

mobilenet_v1.py

从此处下载

https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet_v1.py