Tensorflow学习笔记
来源:互联网 发布:java同步锁的种类 编辑:程序博客网 时间:2024/06/01 10:25
- Install
- Basic Usage
- 构建图
- 交互式方法
- 变量
- MNIST Training
- MNIST Data
- Softmax
- CNN
- TensorFlow运作方式
- Inference
- Loss
- TensorBoard
Install
Ubuntu14.04:
#安装pipsudo apt-get install python-pip python-dev #安装tensorflowsudo pip install --upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.8.0-cp27-none-linux_x86_64.whl#安装python-numpy ,python-scipy,python-matplotlibsudo apt-get install python-numpysudo apt-get install python-scipysudo apt-get install python-matplotlib
官网上的tensorflow安装命令在虚拟机里面安装不上,可能是网络的问题。
测试是否安装成功:
>>> import tensorflow as tf>>> hello = tf.constant('Hello, TensorFlow!')>>> sess = tf.Session()>>> print(sess.run(hello))Hello, TensorFlow!>>> a = tf.constant(10)>>> b = tf.constant(32)>>> print(sess.run(a + b))42
Basic Usage
TensorFlow的基本内容:
- 用图(graph)来表示计算任务
- 在会话(Session)中执行图
- 用(tensor)表示数据
- 用变量(Variable)维护状态
- 用feed和fetch来赋值或取值
构建图
构建图,首先创建源op,比如(Constant)。
TensorFlow Python 库有一个默认图(default graph), op 构造器可以为其增加节点,通常这个默认图对许多程序来说已经够用了。
图构造完成后,需要启动图才能运行操作。启动图需要创建一个Session,参数缺省启动默认图。
以下是示例代码:
import tensorflow as tf# 创建一个常量 op, 产生一个 1x2 矩阵. 这个 op 被作为一个节点# 加到默认图中.## 构造器的返回值代表该常量 op 的返回值.matrix1 = tf.constant([[3., 3.]])# 创建另外一个常量 op, 产生一个 2x1 矩阵.matrix2 = tf.constant([[2.],[2.]])# 创建一个矩阵乘法 matmul op , 把 'matrix1' 和 'matrix2' 作为输入.# 返回值 'product' 代表矩阵乘法的结果.product = tf.matmul(matrix1, matrix2)# 启动默认图.sess = tf.Session()# 调用 sess 的 'run()' 方法来执行矩阵乘法 op, 传入 'product' 作为该方法的参数. # 上面提到, 'product' 代表了矩阵乘法 op 的输出, 传入它是向方法表明, 我们希望取回# 矩阵乘法 op 的输出.## 整个执行过程是自动化的, 会话负责传递 op 所需的全部输入. op 通常是并发执行的.# # 函数调用 'run(product)' 触发了图中三个 op (两个常量 op 和一个矩阵乘法 op) 的执行.## 返回值 'result' 是一个 numpy `ndarray` 对象.result = sess.run(product)print result# ==> [[ 12.]]# 任务完成, 关闭会话.sess.close()
Session对象使用完成后需要关闭释放资源。除了显示调用close以外,也可以用以下方式来代替原来方法:
with tf.Session() as sess: result = sess.run([product]) print result
在实现上,TensorFlow 将图形定义转换成分布式执行的操作,以充分利用可用的计算资源(如 CPU 或 GPU)。一般不需要显式指定使用 CPU 还是 GPU,TensorFlow 能自动检测。如果检测到 GPU,TensorFlow 会尽可能地利用找到的第一个 GPU 来执行操作。
如果机器上有超过一个可用的 GPU, 除第一个外的其它 GPU 默认是不参与计算的. 为了让 TensorFlow 使用这些 GPU, 你必须将 op 明确指派给它们执行. with…Device 语句用来指派特定的 CPU 或 GPU 执行操作:
with tf.Session() as sess: with tf.device("/gpu:1"): matrix1 = tf.constant([[3., 3.]]) matrix2 = tf.constant([[2.],[2.]]) product = tf.matmul(matrix1, matrix2)
设备用字符串进行标识,目前支持的设备包括:
"/cpu:0": 机器的 CPU."/gpu:0": 机器的第一个 GPU, 如果有的话."/gpu:1": 机器的第二个 GPU, 以此类推.
交互式方法
文档中的 Python 示例使用一个会话 Session 来 启动图,并调用 Session.run() 方法执行操作。
为了便于使用诸如 IPython 之类的 Python 交互环境, 可以使用 InteractiveSession 代替 Session 类,使用 Tensor.eval() 和 Operation.run() 方法代替 Session.run()。这样可以避免使用一个变量来持有会话。
以下是示例代码:
# 进入一个交互式 TensorFlow 会话.import tensorflow as tfsess = tf.InteractiveSession()x = tf.Variable([1.0, 2.0])a = tf.constant([3.0, 3.0])# 使用初始化器 initializer op 的 run() 方法初始化 'x' x.initializer.run()# 增加一个减法 sub op, 从 'x' 减去 'a'. 运行减法 op, 输出结果 sub = tf.sub(x, a)print sub.eval()# ==> [-2. -1.]
变量
变量维护图执行过程中的状态信息,下面的例子演示了如何使用变量实现一个简单的计数器:
# 创建一个变量, 初始化为标量 0.state = tf.Variable(0, name="counter")# 创建一个 op, 其作用是使 state 增加 1one = tf.constant(1)new_value = tf.add(state, one)update = tf.assign(state, new_value)# 启动图后, 变量必须先经过`初始化` (init) op 初始化,# 首先必须增加一个`初始化` op 到图中.init_op = tf.initialize_all_variables()# 启动图, 运行 opwith tf.Session() as sess: # 运行 'init' op sess.run(init_op) # 打印 'state' 的初始值 print sess.run(state) # 运行 op, 更新 'state', 并打印 'state' for _ in range(3): sess.run(update) print sess.run(state)# 输出:# 0# 1# 2# 3
在调用 run() 执行表达式之前,图所描绘的表达式(assign(), add() ….)并不会真正执行赋值操作。
用[ ]可以在run的时候取回多个tensor:
input1 = tf.constant(3.0)input2 = tf.constant(2.0)input3 = tf.constant(5.0)intermed = tf.add(input2, input3)mul = tf.mul(input1, intermed)with tf.Session(): result = sess.run([mul, intermed]) print result# 输出:# [array([ 21.], dtype=float32), array([ 7.], dtype=float32)]
还可以用一下方式来待定参数,在run的时候再设定输入:
input1 = tf.placeholder(tf.float32)input2 = tf.placeholder(tf.float32)output = tf.mul(input1, input2)with tf.Session() as sess: print sess.run([output], feed_dict={input1:[7.], input2:[2.]})# 输出:# [array([ 14.], dtype=float32)]
MNIST Training
MNIST是在机器学习领域中的一个经典问题。该问题解决的是把28x28像素的灰度手写数字图片识别为相应的数字,其中数字的范围从0到9。
MNIST是ML界的’hello world’,这个比喻还挺有意思的。
MNIST Data
MNIST数据集有四个文件:
- train-images-idx3-ubyte.gz:训练集图片 - 55000 张 训练图片,5000 张 验证图片。
- train-labels-idx1-ubyte.gz:训练集图片对应的数字标签。
- t10k-images-idx3-ubyte.gz:测试集图片 - 10000 张 图片。
- t10k-labels-idx1-ubyte.gz:测试集图片对应的数字标签。
这些文件本身并没有使用标准的图片格式存储。在下面代码中extract_images()和extract_labels()函数可以手动解压他们。
图片数据将被解压成2维的tensor:[image index, pixel index] 其中每一项表示某一图片中特定像素的强度值。”image index”代表数据集中图片的编号,从0到数据集的上限值。”pixel index”代表该图片中像素点的个数, 从0到图片的像素上限值。
以train-*开头的文件中包括60000个样本,其中分割出55000个样本作为训练集,其余的5000个样本作为验证集。因为所有数据集中28x28像素的灰度图片的尺寸为784,所以训练集输出的tensor格式为[55000, 784]。
数字标签数据被解压成1维的tensor:[image index],它定义了每个样本数值的类别分类。对于训练集的标签来说,这个数据规模就是:[55000]。
解压重构图片和标签数据之后,会得到如下数据集对象:
- data_sets.train:55000 组 图片和标签,用于训练。
- data_sets.validation:5000 组 图片和标签,用于迭代验证训练的准确性。
- data_sets.test:10000 组 图片和标签, 用于最终测试训练的准确性。
调用以下代码中的read_data_sets()函数,将会返回一个DataSet实例,其中包含了以上三个数据集。
函数DataSet.next_batch()是用于获取以batch_size为大小的一个元组,其中包含了一组图片和标签,该元组会被用于当前的TensorFlow运算会话中。
下载数据的代码:
# Copyright 2015 Google Inc. 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.# =============================================================================="""Functions for downloading and reading MNIST data."""from __future__ import absolute_importfrom __future__ import divisionfrom __future__ import print_functionimport gzipimport osimport tensorflow.python.platformimport numpyfrom six.moves import urllibfrom six.moves import xrange # pylint: disable=redefined-builtinimport tensorflow as tfSOURCE_URL = 'http://yann.lecun.com/exdb/mnist/'def maybe_download(filename, work_directory): """Download the data from Yann's website, unless it's already here.""" if not os.path.exists(work_directory): os.mkdir(work_directory) filepath = os.path.join(work_directory, filename) if not os.path.exists(filepath): filepath, _ = urllib.request.urlretrieve(SOURCE_URL + filename, filepath) statinfo = os.stat(filepath) print('Successfully downloaded', filename, statinfo.st_size, 'bytes.') return filepathdef _read32(bytestream): dt = numpy.dtype(numpy.uint32).newbyteorder('>') return numpy.frombuffer(bytestream.read(4), dtype=dt)[0]def extract_images(filename): """Extract the images into a 4D uint8 numpy array [index, y, x, depth].""" print('Extracting', filename) with gzip.open(filename) as bytestream: magic = _read32(bytestream) if magic != 2051: raise ValueError( 'Invalid magic number %d in MNIST image file: %s' % (magic, filename)) num_images = _read32(bytestream) rows = _read32(bytestream) cols = _read32(bytestream) buf = bytestream.read(rows * cols * num_images) data = numpy.frombuffer(buf, dtype=numpy.uint8) data = data.reshape(num_images, rows, cols, 1) return datadef dense_to_one_hot(labels_dense, num_classes=10): """Convert class labels from scalars to one-hot vectors.""" num_labels = labels_dense.shape[0] index_offset = numpy.arange(num_labels) * num_classes labels_one_hot = numpy.zeros((num_labels, num_classes)) labels_one_hot.flat[index_offset + labels_dense.ravel()] = 1 return labels_one_hotdef extract_labels(filename, one_hot=False): """Extract the labels into a 1D uint8 numpy array [index].""" print('Extracting', filename) with gzip.open(filename) as bytestream: magic = _read32(bytestream) if magic != 2049: raise ValueError( 'Invalid magic number %d in MNIST label file: %s' % (magic, filename)) num_items = _read32(bytestream) buf = bytestream.read(num_items) labels = numpy.frombuffer(buf, dtype=numpy.uint8) if one_hot: return dense_to_one_hot(labels) return labelsclass DataSet(object): def __init__(self, images, labels, fake_data=False, one_hot=False, dtype=tf.float32): """Construct a DataSet. one_hot arg is used only if fake_data is true. `dtype` can be either `uint8` to leave the input as `[0, 255]`, or `float32` to rescale into `[0, 1]`. """ dtype = tf.as_dtype(dtype).base_dtype if dtype not in (tf.uint8, tf.float32): raise TypeError('Invalid image dtype %r, expected uint8 or float32' % dtype) if fake_data: self._num_examples = 10000 self.one_hot = one_hot else: assert images.shape[0] == labels.shape[0], ( 'images.shape: %s labels.shape: %s' % (images.shape, labels.shape)) self._num_examples = images.shape[0] # Convert shape from [num examples, rows, columns, depth] # to [num examples, rows*columns] (assuming depth == 1) assert images.shape[3] == 1 images = images.reshape(images.shape[0], images.shape[1] * images.shape[2]) if dtype == tf.float32: # Convert from [0, 255] -> [0.0, 1.0]. images = images.astype(numpy.float32) images = numpy.multiply(images, 1.0 / 255.0) self._images = images self._labels = labels self._epochs_completed = 0 self._index_in_epoch = 0 @property def images(self): return self._images @property def labels(self): return self._labels @property def num_examples(self): return self._num_examples @property def epochs_completed(self): return self._epochs_completed def next_batch(self, batch_size, fake_data=False): """Return the next `batch_size` examples from this data set.""" if fake_data: fake_image = [1] * 784 if self.one_hot: fake_label = [1] + [0] * 9 else: fake_label = 0 return [fake_image for _ in xrange(batch_size)], [ fake_label for _ in xrange(batch_size)] start = self._index_in_epoch self._index_in_epoch += batch_size if self._index_in_epoch > self._num_examples: # Finished epoch self._epochs_completed += 1 # Shuffle the data perm = numpy.arange(self._num_examples) numpy.random.shuffle(perm) self._images = self._images[perm] self._labels = self._labels[perm] # Start next epoch start = 0 self._index_in_epoch = batch_size assert batch_size <= self._num_examples end = self._index_in_epoch return self._images[start:end], self._labels[start:end]def read_data_sets(train_dir, fake_data=False, one_hot=False, dtype=tf.float32): class DataSets(object): pass data_sets = DataSets() if fake_data: def fake(): return DataSet([], [], fake_data=True, one_hot=one_hot, dtype=dtype) data_sets.train = fake() data_sets.validation = fake() data_sets.test = fake() return data_sets TRAIN_IMAGES = 'train-images-idx3-ubyte.gz' TRAIN_LABELS = 'train-labels-idx1-ubyte.gz' TEST_IMAGES = 't10k-images-idx3-ubyte.gz' TEST_LABELS = 't10k-labels-idx1-ubyte.gz' VALIDATION_SIZE = 5000 local_file = maybe_download(TRAIN_IMAGES, train_dir) train_images = extract_images(local_file) local_file = maybe_download(TRAIN_LABELS, train_dir) train_labels = extract_labels(local_file, one_hot=one_hot) local_file = maybe_download(TEST_IMAGES, train_dir) test_images = extract_images(local_file) local_file = maybe_download(TEST_LABELS, train_dir) test_labels = extract_labels(local_file, one_hot=one_hot) validation_images = train_images[:VALIDATION_SIZE] validation_labels = train_labels[:VALIDATION_SIZE] train_images = train_images[VALIDATION_SIZE:] train_labels = train_labels[VALIDATION_SIZE:] data_sets.train = DataSet(train_images, train_labels, dtype=dtype) data_sets.validation = DataSet(validation_images, validation_labels, dtype=dtype) data_sets.test = DataSet(test_images, test_labels, dtype=dtype) return data_sets
每一个MNIST数据单元有两部分组成:一张包含手写数字的图片和一个对应的标签。我们把这些图片设为“xs”,把这些标签设为“ys”。训练数据集和测试数据集都包含xs和ys,比如训练数据集的图片是 mnist.train.images ,训练数据集的标签是 mnist.train.labels。
其中,mnist.train.images 是一个形状为 [60000, 784] 的张量,第一个维度数字用来索引图片,第二个维度数字用来索引每张图片中的像素点。在此张量里的每一个元素,都表示某张图片里的某个像素的强度值,值介于0和1之间。
数字n将表示成一个只有在第n维度(从0开始)数字为1的10维向量。比如,标签0将表示成([1,0,0,0,0,0,0,0,0,0,0])。因此, mnist.train.labels 是一个 [60000, 10] 的数字矩阵。
Softmax
softmax模型可以用来给不同的对象分配概率。
简单版本Softmax模型实现代码:
import tensorflow as tfimport input_data#read Datamnist = input_data.read_data_sets("Mnist_data/", one_hot = True)#parametex = tf.placeholder(tf.float32, [None, 784])W = tf.Variable(tf.zeros([784, 10]))b = tf.Variable(tf.zeros([10]))#calc yy = tf.nn.softmax(tf.matmul(x, W) + b)#true yy_true = tf.placeholder(tf.float32, [None, 10])#cost functioncross_entropy = -tf.reduce_sum(y_true * tf.log(y))#gradientDescenttrain_step = tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy)#initinit = tf.initialize_all_variables()sess = tf.Session()sess.run(init)#trainfor i in range(1000): batch_xs, batch_ys = mnist.train.next_batch(100) sess.run(train_step, feed_dict = {x : batch_xs, y_true : batch_ys}) print "loop " + str(i) + " done."correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_true, 1))accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))print sess.run(accuracy, feed_dict = {x : mnist.test.images, y_true : mnist.test.labels})
代码中有几个注意点:
- tf.placeholder()相当于c/c++里面的定义的还未输入(赋值)的变量,浮点型表示用”float”或者tf.float32都可以。x = tf.placeholder(“float”, [None, 784]),第二个参数表示这个张量的形状是[None, 784],表示第一维不定,第二维固定。
- tf.Variable()相当于定义了带初值的变量。
- tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy),表示使用GD来训练,其Learning rate为0.01,cost function为cross_entropy。
- mnist.train.next_batch(100)会每次随机读取数据中的100个数据点来批处理。
- tf.argmax(a, b)可以返回矩阵a中,第b维中最大的数的index。
- tf.cast()可以进行类型转换。
最终训练结果正确率大概91%。
使用tf.InteractiveSession()实现的版本代码(别人家的代码- -,为何都如此好看):
import input_dataimport tensorflow as tfmnist = input_data.read_data_sets("Mnist_data/", one_hot=True)sess = tf.InteractiveSession()# Create the modelx = tf.placeholder(tf.float32, [None, 784])W = tf.Variable(tf.zeros([784, 10]))b = tf.Variable(tf.zeros([10]))y = tf.nn.softmax(tf.matmul(x, W) + b)# Define loss and optimizery_ = tf.placeholder(tf.float32, [None, 10])cross_entropy = -tf.reduce_sum(y_ * tf.log(y))train_step = tf.train.GradientDescentOptimizer(0.01).minimize(cross_entropy)# Traintf.initialize_all_variables().run()for i in range(1000): batch_xs, batch_ys = mnist.train.next_batch(100) train_step.run({x: batch_xs, y_: batch_ys})# Test trained modelcorrect_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1))accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32))print(accuracy.eval({x: mnist.test.images, y_: mnist.test.labels}))
CNN
用CNN来训练MNIST,代码:
# load MNIST dataimport input_datamnist = input_data.read_data_sets("Mnist_data/", one_hot=True)# start tensorflow interactiveSessionimport tensorflow as tfsess = tf.InteractiveSession()# weight initializationdef weight_variable(shape): initial = tf.truncated_normal(shape, stddev=0.1) return tf.Variable(initial)def bias_variable(shape): initial = tf.constant(0.1, shape = shape) return tf.Variable(initial)# convolutiondef conv2d(x, W): return tf.nn.conv2d(x, W, strides=[1, 1, 1, 1], padding='SAME')# poolingdef max_pool_2x2(x): return tf.nn.max_pool(x, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='SAME')# Create the model# placeholderx = tf.placeholder("float", [None, 784])y_ = tf.placeholder("float", [None, 10])# variablesW = tf.Variable(tf.zeros([784,10]))b = tf.Variable(tf.zeros([10]))y = tf.nn.softmax(tf.matmul(x,W) + b)# first convolutinal layerw_conv1 = weight_variable([5, 5, 1, 32])b_conv1 = bias_variable([32])x_image = tf.reshape(x, [-1, 28, 28, 1])h_conv1 = tf.nn.relu(conv2d(x_image, w_conv1) + b_conv1)h_pool1 = max_pool_2x2(h_conv1)# second convolutional layerw_conv2 = weight_variable([5, 5, 32, 64])b_conv2 = bias_variable([64])h_conv2 = tf.nn.relu(conv2d(h_pool1, w_conv2) + b_conv2)h_pool2 = max_pool_2x2(h_conv2)# densely connected layerw_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)# dropoutkeep_prob = tf.placeholder("float")h_fc1_drop = tf.nn.dropout(h_fc1, keep_prob)# readout layerw_fc2 = weight_variable([1024, 10])b_fc2 = bias_variable([10])y_conv = tf.nn.softmax(tf.matmul(h_fc1_drop, w_fc2) + b_fc2)# train and evaluate the modelcross_entropy = -tf.reduce_sum(y_*tf.log(y_conv))train_step = tf.train.GradientDescentOptimizer(1e-3).minimize(cross_entropy)#train_step = tf.train.AdagradOptimizer(1e-5).minimize(cross_entropy)correct_prediction = tf.equal(tf.argmax(y_conv, 1), tf.argmax(y_, 1))accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))sess.run(tf.initialize_all_variables())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, train 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})
几个注意点:
- tf.truncated_normal(shape, stddev=0.1)创建一个标准差为0.1,大小为shape的随机截断正态分布矩阵。
- tf.constant(a, shape)创建一个值为a,大小为shape的矩阵。
- tf.nn.conv2d中strides一定为strides[0] = strides[3] = 1,其余两维表示每次卷积框移动的step。
TensorFlow运作方式
Inference
其函数定义如下:
def inference(images, hidden1_units, hidden2_units): """Build the MNIST model up to where it may be used for inference. Args: images: Images placeholder, from inputs(). hidden1_units: Size of the first hidden layer. hidden2_units: Size of the second hidden layer. Returns: softmax_linear: Output tensor with the computed logits. """ # Hidden 1 with tf.name_scope('hidden1'): weights = tf.Variable( tf.truncated_normal([IMAGE_PIXELS, hidden1_units], stddev=1.0 / math.sqrt(float(IMAGE_PIXELS))), name='weights') biases = tf.Variable(tf.zeros([hidden1_units]), name='biases') hidden1 = tf.nn.relu(tf.matmul(images, weights) + biases) # Hidden 2 with tf.name_scope('hidden2'): weights = tf.Variable( tf.truncated_normal([hidden1_units, hidden2_units], stddev=1.0 / math.sqrt(float(hidden1_units))), name='weights') biases = tf.Variable(tf.zeros([hidden2_units]), name='biases') hidden2 = tf.nn.relu(tf.matmul(hidden1, weights) + biases) # Linear with tf.name_scope('softmax_linear'): weights = tf.Variable( tf.truncated_normal([hidden2_units, NUM_CLASSES], stddev=1.0 / math.sqrt(float(hidden2_units))), name='weights') biases = tf.Variable(tf.zeros([NUM_CLASSES]), name='biases') logits = tf.matmul(hidden2, weights) + biases return logits
这个函数主要作用说白了就是做前向传播,其中的一些库的用法也在之前也接触过了。
Loss
这个函数的定义好像新版本和旧版本的不一样。
新版本的简单多了:
def loss(logits, labels): """Calculates the loss from the logits and the labels. Args: logits: Logits tensor, float - [batch_size, NUM_CLASSES]. labels: Labels tensor, int32 - [batch_size]. Returns: loss: Loss tensor of type float. """ labels = tf.to_int64(labels) cross_entropy = tf.nn.sparse_softmax_cross_entropy_with_logits( logits, labels, name='xentropy') loss = tf.reduce_mean(cross_entropy, name='xentropy_mean') return loss
旧版本,这个版本的代码有几个函数没有理解透彻,花了半天时间才搞懂:
def loss(logits, labels): """Calculates the loss from the logits and the labels. Args: logits: Logits tensor, float - [batch_size, NUM_CLASSES]. labels: Labels tensor, int32 - [batch_size]. Returns: loss: Loss tensor of type float. """ # Convert from sparse integer labels in the range [0, NUM_CLASSES) # to 1-hot dense float vectors (that is we will have batch_size vectors, # each with NUM_CLASSES values, all of which are 0.0 except there will # be a 1.0 in the entry corresponding to the label). batch_size = tf.size(labels) labels = tf.expand_dims(labels, 1) indices = tf.expand_dims(tf.range(0, batch_size), 1) concated = tf.concat(1, [indices, labels]) onehot_labels = tf.sparse_to_dense( concated, tf.pack([batch_size, NUM_CLASSES]), 1.0, 0.0) cross_entropy = tf.nn.softmax_cross_entropy_with_logits(logits, onehot_labels, name='xentropy') loss = tf.reduce_mean(cross_entropy, name='xentropy_mean') return loss
代码里面需要注意的几个点:
一是tf.expand_dims(Tensor, dim)函数,作用是为张量Tensor加一维:
sess = tf.InteractiveSession()labels = [1,2,3]x = tf.expand_dims(labels, 0)print(sess.run(x))x = tf.expand_dims(labels, 1)print(sess.run(x))#>>>[[1 2 3]]#>>>[[1]# [2]# [3]]
二是tf.pack(values, axis=0, name=”pack”)函数,将一个R维张量列表沿着axis轴组合成一个R+1维的张量:
# 'x' is [1, 4] # 'y' is [2, 5] # 'z' is [3, 6] pack([x, y, z]) => [[1, 4], [2, 5], [3, 6]] # Pack along first dim. pack([x, y, z], axis=1) => [[1, 2, 3], [4, 5, 6]]
三是tf.concat(concat_dim, values, name=”concat”) 函数,将张量沿着指定维数拼接起来:
t1 = [[1, 2, 3], [4, 5, 6]]t2 = [[7, 8, 9], [10, 11, 12]]tf.concat(0, [t1, t2]) #==> [[1, 2, 3], [4, 5, 6], [7, 8, 9], [10, 11, 12]]tf.concat(1, [t1, t2]) #==> [[1, 2, 3, 7, 8, 9], [4, 5, 6, 10, 11, 12]]
这里要注意的是,如果是两个向量,是无法调用第二个维度的:
# t1, t2为向量,只有一个维度tf.concat(1, [t1, t2])
因此,要连接他们,只能调用tf.expand_dims来扩维:
t1=tf.constant([1,2,3]) t2=tf.constant([4,5,6]) #concated = tf.concat(1, [t1,t2]) 错误 t1=tf.expand_dims(tf.constant([1,2,3]),1) t2=tf.expand_dims(tf.constant([4,5,6]),1) concated = tf.concat(1, [t1,t2]) #正确
四是tf.sparse_to_dense,将系数矩阵转成密集矩阵,例子如下:
import tensorflow as tf import numpy BATCHSIZE=6 label=tf.expand_dims(tf.constant([0,2,3,6,7,9]),1) index=tf.expand_dims(tf.range(0, BATCHSIZE),1) #use a matrix concated = tf.concat(1, [index, label]) onehot_labels = tf.sparse_to_dense(concated, tf.pack([BATCHSIZE,10]), 1.0, 0.0) #use a vector concated2=tf.constant([1,3,4]) #onehot_labels2 = tf.sparse_to_dense(concated2, tf.pack([BATCHSIZE,10]), 1.0, 0.0)#cant use ,because output_shape is not a vector onehot_labels2 = tf.sparse_to_dense(concated2, tf.pack([10]), 1.0, 0.0)#can use #use a scalar concated3=tf.constant(5) onehot_labels3 = tf.sparse_to_dense(concated3, tf.pack([10]), 1.0, 0.0) with tf.Session() as sess: result1=sess.run(onehot_labels) result2 = sess.run(onehot_labels2) result3 = sess.run(onehot_labels3) print ("This is result1:") print (result1) print ("This is result2:") print (result2) print ("This is result3:") print (result3)
结果:
This is result1: [[ 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.] [ 0. 0. 1. 0. 0. 0. 0. 0. 0. 0.] [ 0. 0. 0. 1. 0. 0. 0. 0. 0. 0.] [ 0. 0. 0. 0. 0. 0. 1. 0. 0. 0.] [ 0. 0. 0. 0. 0. 0. 0. 1. 0. 0.] [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 1.]] This is result2: [ 0. 1. 0. 1. 1. 0. 0. 0. 0. 0.] This is result3: [ 0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]
TensorBoard
尝试使用了一下TensorBoard来可视化。
代码如下:
import input_dataimport tensorflow as tfflags = tf.app.flagsFLAGS = flags.FLAGSflags.DEFINE_boolean('fake_data', False, 'If true, uses fake data ' 'for unit testing.')flags.DEFINE_integer('max_steps', 1000, 'Number of steps to run trainer.')flags.DEFINE_float('learning_rate', 0.01, 'Initial learning rate.')def main(_): # Import data mnist = input_data.read_data_sets('Mnist_data/', one_hot=True, fake_data=FLAGS.fake_data) sess = tf.InteractiveSession() # Create the model x = tf.placeholder(tf.float32, [None, 784], name='x-input') W = tf.Variable(tf.zeros([784, 10]), name='weights') b = tf.Variable(tf.zeros([10], name='bias')) # Use a name scope to organize nodes in the graph visualizer with tf.name_scope('Wx_b'): y = tf.nn.softmax(tf.matmul(x, W) + b) # Add summary ops to collect data _ = tf.histogram_summary('weights', W) _ = tf.histogram_summary('biases', b) _ = tf.histogram_summary('y', y) # Define loss and optimizer y_ = tf.placeholder(tf.float32, [None, 10], name='y-input') # More name scopes will clean up the graph representation with tf.name_scope('xent'): cross_entropy = -tf.reduce_sum(y_ * tf.log(y)) _ = tf.scalar_summary('cross entropy', cross_entropy) with tf.name_scope('train'): train_step = tf.train.GradientDescentOptimizer( FLAGS.learning_rate).minimize(cross_entropy) with tf.name_scope('test'): correct_prediction = tf.equal(tf.argmax(y, 1), tf.argmax(y_, 1)) accuracy = tf.reduce_mean(tf.cast(correct_prediction, tf.float32)) _ = tf.scalar_summary('accuracy', accuracy) # Merge all the summaries and write them out to /tmp/mnist_logs merged = tf.merge_all_summaries() writer = tf.train.SummaryWriter('/tmp/mnist_logs', sess.graph_def) tf.initialize_all_variables().run() # Train the model, and feed in test data and record summaries every 10 steps for i in range(FLAGS.max_steps): if i % 10 == 0: # Record summary data and the accuracy if FLAGS.fake_data: batch_xs, batch_ys = mnist.train.next_batch( 100, fake_data=FLAGS.fake_data) feed = {x: batch_xs, y_: batch_ys} else: feed = {x: mnist.test.images, y_: mnist.test.labels} result = sess.run([merged, accuracy], feed_dict=feed) summary_str = result[0] acc = result[1] writer.add_summary(summary_str, i) print('Accuracy at step %s: %s' % (i, acc)) else: batch_xs, batch_ys = mnist.train.next_batch( 100, fake_data=FLAGS.fake_data) feed = {x: batch_xs, y_: batch_ys} sess.run(train_step, feed_dict=feed)if __name__ == '__main__': tf.app.run()
在运行完之后,在terminal里面启动tensorboard:
tensorboard --logdir=/tmp/mnist_logs
必须与writer = tf.train.SummaryWriter(‘/tmp/mnist_logs’, sess.graph_def)中的路径相一致。
接着用浏览器打开:http://localhost:6006 就可以看到效果了。
- TensorFlow学习笔记-1
- TensorFlow学习笔记
- TensorFlow 深度学习笔记
- TensorFlow学习笔记1
- tensorflow-Alexnet学习笔记
- TensorFlow学习笔记
- Tensorflow学习笔记
- Tensorflow学习笔记(1)
- Tensorflow学习笔记(2)
- tensorflow学习笔记
- Tensorflow学习笔记
- tensorflow基础学习笔记
- TensorFlow--学习笔记
- tensorflow学习笔记
- TensorFlow学习笔记----TensorBoard_1
- TensorFlow学习笔记----TensorBoard_2
- Tensorflow学习笔记
- TensorFlow学习笔记
- 堆
- Android动画----Scale
- 如何检测棋盘格
- Angular2 表单form
- js案例 移动端左右滑动touch+transform
- Tensorflow学习笔记
- 大话数据结构8之栈和队列2
- (十六)备忘录模式详解(都市异能版)
- Servlet 监听
- 栅栏染色-动态规划
- 那些年在WebView上踩过的坑
- 如何确定CPU是大端字节序还是小端字节序?
- (十七)迭代器模式详解(foreach的精髓)
- C语言——使用scanf函数时需要注意的问题