本文采用LeNet-5卷积网络来完成对MNIST数据集的分类任务，以下将详细介绍具体思考模式和流程。
1. 创建神经网络模型
def model(data, train=False):
"""模型定义"""
# 2D卷积, with 'SAME' padding (输出的feature map与输入大小一致)
# strides向量分别对应[image索引, y, x, depth]
conv = tf.nn.conv2d(data,
conv1_weights,
strides=[1, 1, 1, 1],
padding='SAME')
# conv1_biases为卷积层常数项对应的变量，也作为被训练参数之一
relu = tf.nn.relu(tf.nn.bias_add(conv, conv1_biases))
# Max pooling. 池化大小为2*2,且stripe为2*2
pool = tf.nn.max_pool(relu,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME')
#第二层卷积
conv = tf.nn.conv2d(pool,
conv2_weights,
strides=[1, 1, 1, 1],
padding='SAME')
relu = tf.nn.relu(tf.nn.bias_add(conv, conv2_biases))
pool = tf.nn.max_pool(relu,
ksize=[1, 2, 2, 1],
strides=[1, 2, 2, 1],
padding='SAME')
# 将feature map转化为2d矩阵，目的是适配全连接层
pool_shape = pool.get_shape().as_list()
reshape = tf.reshape(
pool,
[pool_shape[0], pool_shape[1] * pool_shape[2] * pool_shape[3]])
# 全连接层
hidden = tf.nn.relu(tf.matmul(reshape, fc1_weights) + fc1_biases)
# 仅仅在训练阶段添加50%的dropout,作为对参数的正则化
if train:
hidden = tf.nn.dropout(hidden, 0.5, seed=SEED)
return tf.matmul(hidden, fc2_weights) + fc2_biases
# 构建计算图
train_data_node = tf.placeholder(
tf.float32,
shape=(BATCH_SIZE, IMAGE_SIZE, IMAGE_SIZE, NUM_CHANNELS))
train_labels_node = tf.placeholder(tf.int64, shape=(BATCH_SIZE,))
eval_data = tf.placeholder(
tf.float32,
shape=(EVAL_BATCH_SIZE, IMAGE_SIZE, IMAGE_SIZE, NUM_CHANNELS))
# 以下为可训练的所有权重参数.
# 初始化命令为：{tf.global_variables_initializer().run()}
conv1_weights = tf.Variable(
# 5x5 filter, depth 32.
tf.truncated_normal([5, 5, NUM_CHANNELS, 32],
stddev=0.1,
seed=SEED, dtype=tf.float32))
conv1_biases = tf.Variable(tf.zeros([32], dtype=tf.float32))
conv2_weights = tf.Variable(tf.truncated_normal(
[5, 5, 32, 64], stddev=0.1,
seed=SEED, dtype=tf.float32))
conv2_biases = tf.Variable(tf.constant(0.1, shape=[64],
dtype=tf.float32))
fc1_weights = tf.Variable( # fully connected, depth 512.
tf.truncated_normal([IMAGE_SIZE // 4 * IMAGE_SIZE // 4 * 64, 512],
stddev=0.1,
seed=SEED,
dtype=tf.float32))
fc1_biases = tf.Variable(tf.constant(0.1, shape=[512],
dtype=tf.float32))
fc2_weights = tf.Variable(tf.truncated_normal([512, NUM_LABELS],
stddev=0.1,
seed=SEED,
dtype=tf.float32))
fc2_biases = tf.Variable(tf.constant(
0.1, shape=[NUM_LABELS], dtype=tf.float32))
#调用以上的model函数，得到模型输出，并计算softmax交叉熵损失函数
logits = model(train_data_node, True)
loss = tf.reduce_mean(tf.nn.sparse_softmax_cross_entropy_with_logits(
labels=train_labels_node, logits=logits))
# 全连接层参数的L2正则化
regularizers = (tf.nn.l2_loss(fc1_weights)
+ tf.nn.l2_loss(fc1_biases)
+ tf.nn.l2_loss(fc2_weights)
+ tf.nn.l2_loss(fc2_biases))
loss += 5e-4 * regularizers
# 每批次增加1，用于控制decay rate
batch = tf.Variable(0, dtype=tf.float32)
# Decay once per epoch, using an exponential schedule starting at 0.01.
learning_rate = tf.train.exponential_decay(
0.01, # Base learning rate.
batch * BATCH_SIZE, # Current index into the dataset.
train_size, # Decay step.
0.95, # Decay rate.
staircase=True)
# 优化器
optimizer = tf.train.MomentumOptimizer(learning_rate,
0.9).minimize(loss,
global_step=batch)
# 得到当前批次得到的预测分类值
train_prediction = tf.nn.softmax(logits)
# 对测试集和验证数据集进行预测
eval_prediction = tf.nn.softmax(model(eval_data))
2. 数据的准备
#下载文件
def download(filename):
"""从网站下载"""
if not os.path.exists(WORK_DIRECTORY):
os.makedirs(WORK_DIRECTORY)
filepath = os.path.join(WORK_DIRECTORY, filename)
if not os.path.exists(filepath):
filepath, _ = urllib.request.urlretrieve(SOURCE_URL + filename,
filepath)
size = os.stat(filepath).st_size
return filepath
# 从文件中提取出图片数据
def extract_data(filename, num_images):
"""将图片生成4D张量数据[image index, y, x, channels].
并且将值从 [0, 255] 转化为 [-0.5, 0.5]
"""
print('Extracting', filename)
with gzip.open(filename) as bytestream:
bytestream.read(16)
buf = bytestream.read(
IMAGE_SIZE * IMAGE_SIZE * num_images * NUM_CHANNELS)
data = numpy.frombuffer(buf, dtype=numpy.uint8).astype(
numpy.float32)
data = (data - (255/2.0))/255
data = data.reshape(num_images, IMAGE_SIZE, IMAGE_SIZE,
NUM_CHANNELS)
return data
#提取标签
def extract_labels(filename, num_images):
"""转化为一个类型为int64的向量"""
print('Extracting', filename)
with gzip.open(filename) as bytestream:
# Discard header.
bytestream.read(8)
# Read bytes for labels.
buf = bytestream.read(num_images)
labels = numpy.frombuffer(buf, dtype=numpy.uint8).astype(
numpy.int64)
return labels
SOURCE_URL = 'http://yann.lecun.com/exdb/mnist/'
WORK_DIRECTORY = 'data'
IMAGE_SIZE = 28
NUM_CHANNELS = 1
PIXEL_DEPTH = 255
NUM_LABELS = 10
VALIDATION_SIZE = 5000 # Size of the validation set.
SEED = 66478 # Set to None for random seed.
BATCH_SIZE = 64
NUM_EPOCHS = 10
EVAL_BATCH_SIZE = 64
EVAL_FREQUENCY = 100 # Number of steps between evaluations.
train_data_filename = download('train-images-idx3-ubyte.gz')
train_labels_filename = download('train-labels-idx1-ubyte.gz')
test_data_filename = download('t10k-images-idx3-ubyte.gz')
test_labels_filename = download('t10k-labels-idx1-ubyte.gz')
train_data = extract_data(train_data_filename, 60000)
train_labels = extract_labels(train_labels_filename, 60000)
test_data = extract_data(test_data_filename, 10000)
test_labels = extract_labels(test_labels_filename, 10000)
validation_data = train_data[:VALIDATION_SIZE, ...]
validation_labels = train_labels[:VALIDATION_SIZE]
train_data = train_data[VALIDATION_SIZE:, ...]
train_labels = train_labels[VALIDATION_SIZE:]
num_epochs = NUM_EPOCHS
train_size = train_labels.shape[0]
3. 模型训练
def eval_in_batches(data, sess):
"""批量模型推理预测"""
size = data.shape[0]
if size < EVAL_BATCH_SIZE:
raise ValueError("batch size for evals larger than dataset: %d"
% size)
predictions = numpy.ndarray(shape=(size, NUM_LABELS),
dtype=numpy.float32)
for begin in xrange(0, size, EVAL_BATCH_SIZE):
end = begin + EVAL_BATCH_SIZE
if end <= size:
predictions[begin:end, :] = sess.run(
eval_prediction,
feed_dict={eval_data: data[begin:end, ...]})
else:
batch_predictions = sess.run(
eval_prediction,
feed_dict={eval_data: data[-EVAL_BATCH_SIZE:, ...]})
predictions[begin:, :] = batch_predictions[begin - size:, :]
return predictions
def error_rate(predictions, labels):
"""计算分类的错误率"""
return 100.0 - (
100.0 *
numpy.sum(numpy.argmax(predictions, 1) == labels) /
predictions.shape[0])
start_time = time.time()
with tf.Session() as sess:
tf.global_variables_initializer().run()
# 训练步长
for step in xrange(int(num_epochs * train_size) // BATCH_SIZE):
# 得到每一步长对应批量数据
offset = (step * BATCH_SIZE) % (train_size - BATCH_SIZE)
batch_data = train_data[offset:(offset + BATCH_SIZE), ...]
batch_labels = train_labels[offset:(offset + BATCH_SIZE)]
feed_dict = {train_data_node: batch_data,
train_labels_node: batch_labels}
# 更新权重等参数
sess.run(optimizer, feed_dict=feed_dict)
# 一定阶段后输出对应的变量值，作为评估
if step % EVAL_FREQUENCY == 0:
# fetch some extra nodes' data
l, lr, predictions = sess.run([loss, learning_rate,
train_prediction],
feed_dict=feed_dict)
elapsed_time = time.time() - start_time
start_time = time.time()
print('Step %d (epoch %.2f), %.1f ms' %
(step, float(step) * BATCH_SIZE / train_size,
1000 * elapsed_time / EVAL_FREQUENCY))
print('Minibatch loss: %.3f, learning rate: %.6f' % (l, lr))
print('Minibatch error: %.1f%%'
% error_rate(predictions, batch_labels))
print('Validation error: %.1f%%' % error_rate(
eval_in_batches(validation_data, sess), validation_labels))
sys.stdout.flush()
# 最终结果
test_error = error_rate(eval_in_batches(test_data, sess),
test_labels)
print('Test error: %.1f%%' % test_error)
网友评论