-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy patheager_basic.py
More file actions
124 lines (94 loc) · 2.94 KB
/
eager_basic.py
File metadata and controls
124 lines (94 loc) · 2.94 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
#!/usr/bin/env python
# _*_ coding: utf-8 _*_
"""
@author : lightnine
@site : https://ligntnine.github.io
@version : 1.0
@file : eager_basic.py
@software : PyCharm
@time : 2019/1/21 19:26
展示eager execution的一些基本用法
https://www.tensorflow.org/tutorials/eager/eager_basics?hl=zh-cn
"""
import os
# 取消警告信息的显示
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
import timeit
import tempfile
import numpy as np
import tensorflow as tf
# 启用eager模式
tf.enable_eager_execution()
def tensors_ops():
print(tf.add(1, 2))
print(tf.add([1, 2], [3, 4]))
print(tf.square(5))
print(tf.reduce_sum([1, 2, 3]))
print(tf.encode_base64("hello world"))
# Operator overloading is also supported
print(tf.square(2) + tf.square(3))
x = tf.matmul([[1]], [[2, 3]])
print(x.shape)
print(x.dtype)
def numpy_vs_tensor():
ndarray = np.ones([3, 3])
print("tensorflow 操作自动将numpy类型的数组转为tensor")
tensor = tf.multiply(ndarray, 42)
print(tensor)
print("numpy操作自动将tensor转为numpy类型")
print(np.add(tensor, 1))
print("numpy()方法明确将tensor转为numpy array")
print(tensor.numpy())
def gpu_acceleration():
x = tf.random_uniform([3, 3])
print("GPU是否可用:")
print(tf.test.is_gpu_available())
print("tensor 的device属性:", x.device)
print("tensor 是否在 GPU#0上:")
print(x.device.endswith("GPU:0"))
def time_matmul(x):
t = timeit.timeit(lambda: tf.matmul(x, x), number=1)
print("时间消耗:", t)
# %timeit tf.matmul(x, x)
def device_placement():
print("CPU上执行:")
with tf.device("CPU:0"):
x = tf.random_uniform([1000, 1000])
assert x.device.endswith("CPU:0")
time_matmul(x)
# 如果GPU可以用，则在GPU上执行
if tf.test.is_gpu_available():
print("在GPU上执行")
with tf.device("GPU:0"):
x = tf.random_uniform([1000, 1000])
assert x.device.endswith("GPU:0")
time_matmul(x)
def dataset_ops():
ds_tensors = tf.data.Dataset.from_tensor_slices([1, 2, 3, 4, 5, 6])
# 创建临时文件
_, filename = tempfile.mkstemp()
print(_)
print("文件名:", filename)
with open(filename, 'w') as f:
f.write("""Line 1
Line 2
Line 3
""")
ds_file = tf.data.TextLineDataset(filename)
print("ds_file type：", type(ds_file))
print("ds_file:", ds_file)
ds_tensors = ds_tensors.map(tf.square).shuffle(2).batch(2)
ds_file = ds_file.batch(2)
print("ds_file:", ds_file)
print('ds_tensors中的元素:')
for x in ds_tensors:
print(x)
print('ds_file中的元素:')
for x in ds_file:
print(x)
if __name__ == '__main__':
tensors_ops()
numpy_vs_tensor()
gpu_acceleration()
device_placement()
dataset_ops()