1.认识下TensorBoard
TensorFlow不仅是一个软件库,而是一整套包括TensorFlow、TensorBoard、Tensor Serving在内的软件包。为了更大程度地利用TensorFlow,我们应该了解如何将它们串联起来应用。在和一部分,我们来探索下TensorBoard。
TensorBoard是一个图(graph)可视化软件,在(安装TensorFlow的时候会默认安装)。下面是谷歌的介绍:
The computations you'll use TensorFlow for - like training a massive deep neural network - can be complex and confusing. To make it easier to understand, debug, and optimize TensorFlow programs, we've included a suite of visualization tools called TensorBoard.
在运行一个包含一些运算的TensorFlow程序时,这些运算会导出成一个时间日志文件。TensorBoard 可以将这些日志文件可视化,以便更好观察程序的机构以及运行表现。TensorBoard和TensorFlow一并使用,会使工作更加有趣和更具生产力。
下面开始我们第一个TensorFlow程序,并使用TensorBoard可视化。
import tensorflow as tf
a = tf.constant(2)
b = tf.constant(3)
x = tf.add(a, b)
with tf.Session() as sess:
print(sess.run(x))
执行结果
为了将上面程序可视化,我们需要下面一行程序将日志写入文件:
writer = tf.summary.FileWriter([logdir], [graph])
[graph] 是运行程序所在的图,可以通过tf.get_default_graph()
返回程序默认图,也可以通过sess.graph返回当前会话中运行的图,后者需要你自己先创建一个session。无论哪种方式,都要你在定义graph之后创建一个writer,否则TensorBoard不能可视化程序。
[logdir]是存储日志文件的路径
import tensorflow as tf
a = tf.constant(2)
b = tf.constant(3)
x = tf.add(a, b)
writer = tf.summary.FileWriter('./graphs', tf.get_default_graph())
with tf.Session() as sess:
# writer = tf.summary.FileWriter('./graphs', sess.graph) # if you prefer creating your writer using session's graph
print(sess.run(x))
writer.close()
然后在cmd运行程序
$ python3 [my_program.py]
$ tensorboard --logdir="./graphs" --port 6006
在浏览器打开
可视化效果如下
“Const”和“Const_1”指的是a和b,节点“Add”指的是x,为了更好理解运算,我们可以给ops命名。
a = tf.constant(2, name="a")
b = tf.constant(2, name="b")
x = tf.add(a, b, name="add")
我们可以通过点击节点来查看它的值和类型:
2.常量op
op:图中的节点(operation 的缩写).
下面是创建constant的操作
tf.constant(value, dtype=None, shape=None, name='Const', verify_shape=False)
-
常数
num = tf.constant(2, name="num")
-
向量
# constant of 1d tensor (vector)
a = tf.constant([2, 2], name="vector")
-
矩阵
# constant of 2x2 tensor (matrix)
b = tf.constant([[0, 1], [2, 3]], name="matrix")
可以通过填充创建tensor,类似于numpy中的操作 -
tf.zeros(shape, dtype=tf.float32, name=None)
# create a tensor of shape and all elements are zeros
tf.zeros([2, 3], tf.int32) ==> [[0, 0, 0], [0, 0, 0]]
- tf.zeros_like(input_tensor, dtype=None, name=None, optimize=True)
# create a tensor of shape and type (unless type is specified) as the input_tensor but all elements are zeros.
# input_tensor [[0, 1], [2, 3], [4, 5]]
tf.zeros_like(input_tensor) ==> [[0, 0], [0, 0], [0, 0]]
- tf.ones(shape, dtype=tf.float32, name=None)
# create a tensor of shape and all elements are ones
tf.ones([2, 3], tf.int32) ==> [[1, 1, 1], [1, 1, 1]]
- tf.ones_like(input_tensor, dtype=None, name=None, optimize=True)
# create a tensor of shape and type (unless type is specified) as the input_tensor but all elements are ones.
# input_tensor is [[0, 1], [2, 3], [4, 5]]
tf.ones_like(input_tensor) ==> [[1, 1], [1, 1], [1, 1]]
- tf.fill(dims, value, name=None)
# create a tensor filled with a scalar value.
tf.fill([2, 3], 8) ==> [[8, 8, 8], [8, 8, 8]]
- tf.lin_space(start, stop, num, name=None)
tf.lin_space(start, stop, num, name=None)
# create a sequence of num evenly-spaced values are generated beginning at start. If num > 1, the values in the sequence increase by (stop - start) / (num - 1), so that the last one is exactly stop.
# comparable to but slightly different from numpy.linspace
tf.lin_space(10.0, 13.0, 4, name="linspace") ==> [10.0 11.0 12.0 13.0]
- tf.range([start], limit=None, delta=1, dtype=None, name='range')
# create a sequence of numbers that begins at start and extends by increments of delta up to but not including limit
# slight different from range in Python
# 'start' is 3, 'limit' is 18, 'delta' is 3
tf.range(start, limit, delta) ==> [3, 6, 9, 12, 15]
# 'start' is 3, 'limit' is 1, 'delta' is -0.5
tf.range(start, limit, delta) ==> [3, 2.5, 2, 1.5]
# 'limit' is 5
tf.range(limit) ==> [0, 1, 2, 3, 4]
不像Numpy或者Python其他序列,TensorFlow序列不能迭代
for _ in np.linspace(0, 10, 4): # OK
for _ in tf.linspace(0.0, 10.0, 4): # TypeError: 'Tensor' object is not iterable.
for _ in range(4): # OK
for _ in tf.range(4): # TypeError: 'Tensor' object is not iterable.
也可以生成随机constant,具体请见API
tf.random_normal
tf.truncated_normal
tf.random_uniform
tf.random_shuffle
tf.random_crop
tf.multinomial
tf.random_gamma
tf.set_random_seed
3. 数学运算
tf.divide(a/b)才和Python的风格一样,a除以b
a = tf.constant([2, 2], name='a')
b = tf.constant([[0, 1], [2, 3]], name='b')
with tf.Session() as sess:
print(sess.run(tf.div(b, a))) ⇒ [[0 0] [1 1]]
print(sess.run(tf.divide(b, a))) ⇒ [[0. 0.5] [1. 1.5]]
print(sess.run(tf.truediv(b, a))) ⇒ [[0. 0.5] [1. 1.5]]
print(sess.run(tf.floordiv(b, a))) ⇒ [[0 0] [1 1]]
print(sess.run(tf.realdiv(b, a))) ⇒ # Error: only works for real values
print(sess.run(tf.truncatediv(b, a))) ⇒ [[0 0] [1 1]]
print(sess.run(tf.floor_div(b, a))) ⇒ [[0 0] [1 1]]
- tf.add_n
tf.add_n([a, b, b]) => equivalent to a + b + b
- 点积 Dot
a = tf.constant([10, 20], name='a')
b = tf.constant([2, 3], name='b')
with tf.Session() as sess:
print(sess.run(tf.multiply(a, b))) ⇒ [20 60] # element-wise multiplication
print(sess.run(tf.tensordot(a, b, 1))) ⇒ 80 # 按列相乘然后相加
下面是TensorFlow中运算表格,来自《Fundamentals of Deep Learning》
4 数据类型
- python基本数据类型
TensorFlow吸收了Python的一些基本类型,比如布尔值(boolean values),数值(int和float)以及字符串(string)。相应地,单值可以转化为0-d(0维)张量(tensor)(或者标量(scalars)),列表可以转化为1-d tensor (vector),元素为列表的列表可以转化为2-d tensor(matrix),等等。下面是是引自《TensorFlow for Machine Intelligence》的几个例子:
t_0 = 19 # Treated as a 0-d tensor, or "scalar"
tf.zeros_like(t_0) # ==> 0
tf.ones_like(t_0) # ==> 1
t_1 = [b"apple", b"peach", b"grape"] # treated as a 1-d tensor, or "vector"
tf.zeros_like(t_1) # ==> [b'' b'' b'']
tf.ones_like(t_1) # ==> TypeError
t_2 = [[True, False, False],
[False, False, True],
[False, True, False]] # treated as a 2-d tensor, or "matrix"
tf.zeros_like(t_2) # ==> 3x3 tensor, all elements are False
tf.ones_like(t_2) # ==> 3x3 tensor, all elements are True
-
TensorFlow基本数据类型
像NumPy一样,TensorFlow也有自己数据类型,例如:tf.int32,tf.float32.下面是TensorFlow所有数据类型截图:
- NumPy数据类型
截止到现在,你会发现TensorFlow与NumPy有很多相似之处,TensorFlow被设计地与这个数据科学领“通用语言”NumPy无缝衔接。
TensorFlow的数据类型基于NumPy,实际上,np.int32==tf.int32
,你可以将NumPy数据类型传递给TensorFlow算子。
tf.ones([2, 2], np.float32) ==> [[1.0 1.0], [1.0 1.0]]
5.变量
constan与variable的区别:
- 常量也是常数,经常我们需要在训练模型的时候更新权重与偏置矩阵
- 一个常量存储于图中,并且当图加载时需要重新复制;一个变量独立地存储,并且存在与参数服务中。
以上两点解释了,当权重很大时,constant消耗代价很大,并且减慢了加载图的速度。我们看下图存储的内容:
import tensorflow as tf
my_const = tf.constant([1.0, 2.0], name="my_const")
print(tf.get_default_graph().as_graph_def())
node {
name: "my_const"
op: "Const"
attr {
key: "dtype"
value {
type: DT_FLOAT
}
}
attr {
key: "value"
value {
tensor {
dtype: DT_FLOAT
tensor_shape {
dim {
size: 2
}
}
tensor_content: "\000\000\200?\000\000\000@"
}
}
}
}
versions {
producer: 21
}
[Finished in 3.0s]
- 创建变量
为了声明一个变量,你需要实例化一个tf.Variable,注意V
大写。
x = tf.Variable(...)
x.initializer # init
x.value() # read op
x.assign(...) # write op
x.assign_add(...)
# and more
创建变量的旧方式是调用tf.Variable(<initial-value>, name=<optional-name>)
s = tf.Variable(2, name="scalar")
m = tf.Variable([[0, 1], [2, 3]], name="matrix")
W = tf.Variable(tf.zeros([784,10]))
然而这种方式被TensorFlow摒弃了,推荐我们使用tf.get_variable来创建,因为这样可以更好地实现变量共享。通过tf.get_variable
,我们可以internal name,shape,type和initializer提供给初始值。注意当我们通过tf.constant作为initializer时,不需要提供shape
tf.get_variable(
name,
shape=None,
dtype=None,
initializer=None,
regularizer=None,
trainable=True,
collections=None,
caching_device=None,
partitioner=None,
validate_shape=True,
use_resource=None,
custom_getter=None,
constraint=None
)
s = tf.get_variable("scalar", initializer=tf.constant(2))
m = tf.get_variable("matrix", initializer=tf.constant([[0, 1], [2, 3]]))
W = tf.get_variable("big_matrix", shape=(784, 10), initializer=tf.zeros_initializer())
- 初始化变量
在使用一个变量之前,需要先初始化,否则将会报错:FailedPreconditionError: Attempting to use uninitialized value
.可以通过以下语句来打印出没有初始化的变量:
print(session.run(tf.report_uninitialized_variables()))
(1)初始化所有的变量:
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
(2)初始化部分变量tf.variables_initializer()
with tf.Session() as sess:
sess.run(tf.variables_initializer([a, b]))
(3) 初始化单个变量tf.Variable.initializer
with tf.Session() as sess:
sess.run(W.initializer)
- 查看变量的值
从session取出值
# V is a 784 x 10 variable of random values
V = tf.get_variable("normal_matrix", shape=(784, 10),
initializer=tf.truncated_normal_initializer())
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
print(sess.run(V))
通过tf.Variable.eval()取出值
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
print(V.eval())
- 变量赋值
通过tf.Variable.assign()
W = tf.Variable(10)
W.assign(100)
with tf.Session() as sess:
sess.run(W.initializer)
print(W.eval()) # >> 10
为什么输出的是10而不是100呢?W.assign(100)
并没有将100赋值给 W,而是创建了一个assign op.为了使这个op起到效果,我们需要在session运行op
W = tf.Variable(10)
assign_op = W.assign(100)
with tf.Session() as sess:
sess.run(assign_op)
print(W.eval()) # >> 100
注意此时我们不必去初始化W,因为assign()已经为我们实现。
# in the [source code](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/python/ops/variables.py)
self._initializer_op = state_ops.assign(self._variable, self._initial_value, validate_shape=validate_shape).op**
有趣的例子:
# create a variable whose original value is 2
a = tf.get_variable('scalar', initializer=tf.constant(2))
a_times_two = a.assign(a * 2)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
sess.run(a_times_two) # >> 4
sess.run(a_times_two) # >> 8
sess.run(a_times_two) # >> 16
tf.Variable.assign_add()
和 tf.Variable.assign_sub()
,这两个操作需要初始化
W = tf.Variable(10)
with tf.Session() as sess:
sess.run(W.initializer)
print(sess.run(W.assign_add(10))) # >> 20
print(sess.run(W.assign_sub(2))) # >> 18
session之间的变量相互独立
W = tf.Variable(10)
sess1 = tf.Session()
sess2 = tf.Session()
sess1.run(W.initializer)
sess2.run(W.initializer)
print(sess1.run(W.assign_add(10))) # >> 20
print(sess2.run(W.assign_sub(2))) # >> 8
print(sess1.run(W.assign_add(100))) # >> 120
print(sess2.run(W.assign_sub(50))) # >> -42
sess1.close()
sess2.close()
如果一个变量依赖于另一个变量,需要初始化
# W is a random 700 x 10 tensor
W = tf.Variable(tf.truncated_normal([700, 10]))
U = tf.Variable(W * 2)
U = tf.Variable(W.initialized_value() * 2)
6 交互的会话
InteractiveSession是我们在shell或者IPython操作很方便
sess = tf.InteractiveSession()
a = tf.constant(5.0)
b = tf.constant(6.0)
c = a * b
print(c.eval()) # we can use 'c.eval()' without explicitly stating a session
sess.close()
7. 控制依赖
有时候,graph有多个运算时op,并且我们想指定它们的执行顺序时,可以通过tf.Graph.control_dependencies([control_inputs])实现。
# your graph g have 5 ops: a, b, c, d, e
with g.control_dependencies([a, b, c]):
# `d` and `e` will only run after `a`, `b`, and `c` have executed.
d = ...
e = …
8. 导入数据
8.1 placeholder和feed_dict(旧方法)
我们在教程1提到过,TensorFlow程序执行通常包括两个阶段
阶段1:声明一个graph
阶段2:使用一个session来执行计算,来评估图中的变量
我们声明graphs的时候,不许要知道计算所需变量的值,这如同实名一个关于x和y的函数:f(x,y)=2x+y
,其中x和y是实际值的占位符(placeholder)
graph创建之后,我们需要稍后再提供值的时候来定义一个placeholder:
tf.placeholder(dtype, shape=None, name=None)
需要注意的是dtype
和shape
需要自己声明的;当shape=none
的时候, 表名可以接受任意shape的张量tensors。
下面是一个实例:
a = tf.placeholder(tf.float32, shape=[3]) # a is placeholder for a vector of 3 elements
b = tf.constant([5, 5, 5], tf.float32)
c = a + b # use the placeholder as you would any tensor
with tf.Session() as sess:
print(sess.run(c))
我们知道,此时执行上面程序会报错,因为还没有提供给a值,下面使用feed_dict把值传给a
with tf.Session() as sess:
# compute the value of c given the value of a is [1, 2, 3]
print(sess.run(c, feed_dict={a: [1, 2, 3]})) # [6. 7. 8.]
8.2 tf.data(新方法)
此部分结合教程3讲解
网友评论