Variable

tensorflow中有兩個(gè)關(guān)于variable的op，tf.Variable()與tf.get_variable()下面介紹這兩個(gè)的區(qū)別

tf.Variable與tf.get_variable()

tf.Variable(initial_value=None, trainable=True, collections=None, validate_shape=True, caching_device=None, name=None, variable_def=None, dtype=None, expected_shape=None, import_scope=None)

• 1

tf.get_variable(name, shape=None, dtype=None, initializer=None, regularizer=None, trainable=True, collections=None, caching_device=None, partitioner=None, validate_shape=True, custom_getter=None)

• 1

區(qū)別

使用tf.Variable時(shí)，如果檢測(cè)到命名沖突，系統(tǒng)會(huì)自己處理。使用tf.get_variable()時(shí)，系統(tǒng)不會(huì)處理沖突，而會(huì)報(bào)錯(cuò)

import tensorflow as tf w_1 = tf.Variable(3,name="w_1") w_2 = tf.Variable(1,name="w_1") print w_1.name print w_2.name #輸出 #w_1:0 #w_1_1:0

• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8

import tensorflow as tfw_1 = tf.get_variable(name="w_1",initializer=1) w_2 = tf.get_variable(name="w_1",initializer=2) #錯(cuò)誤信息 #ValueError: Variable w_1 already exists, disallowed. Did #you mean to set reuse=True in VarScope?

• 1
• 2
• 3
• 4
• 5
• 6
• 7

基于這兩個(gè)函數(shù)的特性，當(dāng)我們需要共享變量的時(shí)候，需要使用tf.get_variable()。在其他情況下，這兩個(gè)的用法是一樣的

get_variable()與Variable的實(shí)質(zhì)區(qū)別

來(lái)看下面一段代碼：

import tensorflow as tfwith tf.variable_scope("scope1"):w1 = tf.get_variable("w1", shape=[])w2 = tf.Variable(0.0, name="w2") with tf.variable_scope("scope1", reuse=True):w1_p = tf.get_variable("w1", shape=[])w2_p = tf.Variable(1.0, name="w2")print(w1 is w1_p, w2 is w2_p) #輸出 #True False

• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8
• 9
• 10
• 11
• 12

看到這，就可以明白官網(wǎng)上說(shuō)的參數(shù)復(fù)用的真面目了。由于tf.Variable()?每次都在創(chuàng)建新對(duì)象，所有reuse=True?和它并沒有什么關(guān)系。對(duì)于get_variable()，來(lái)說(shuō)，如果已經(jīng)創(chuàng)建的變量對(duì)象，就把那個(gè)對(duì)象返回，如果沒有創(chuàng)建變量對(duì)象的話，就創(chuàng)建一個(gè)新的。

random Tensor

可用于賦值給tf.Variable()的第一個(gè)參數(shù)

tf.random_normal(shape, mean=0.0, stddev=1.0, dtype=tf.float32, seed=None, name=None)tf.truncated_normal(shape, mean=0.0, stddev=1.0, dtype=tf.float32, seed=None, name=None)tf.random_uniform(shape, minval=0, maxval=None, dtype=tf.float32, seed=None, name=None)tf.random_shuffle(value, seed=None, name=None)tf.random_crop(value, size, seed=None, name=None)tf.multinomial(logits, num_samples, seed=None, name=None)tf.random_gamma(shape, alpha, beta=None, dtype=tf.float32, seed=None, name=None)tf.set_random_seed(seed)

• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8
• 9
• 10
• 11
• 12
• 13
• 14
• 15

constant value tensor

tf.zeros(shape, dtype=tf.float32, name=None)tf.zeros_like(tensor, dtype=None, name=None)tf.ones(shape, dtype=tf.float32, name=None)tf.ones_like(tensor, dtype=None, name=None)tf.fill(dims, value, name=None)tf.constant(value, dtype=None, shape=None, name='Const')

• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8
• 9
• 10
• 11

initializer

tf.constant_initializer(value=0, dtype=tf.float32) tf.random_normal_initializer(mean=0.0, stddev=1.0, seed=None, dtype=tf.float32) tf.truncated_normal_initializer(mean=0.0, stddev=1.0, seed=None, dtype=tf.float32) tf.random_uniform_initializer(minval=0, maxval=None, seed=None, dtype=tf.float32) tf.uniform_unit_scaling_initializer(factor=1.0, seed=None, dtype=tf.float32) tf.zeros_initializer(shape, dtype=tf.float32, partition_info=None) tf.ones_initializer(dtype=tf.float32, partition_info=None) tf.orthogonal_initializer(gain=1.0, dtype=tf.float32, seed=None)

• 1
• 2
• 3
• 4
• 5
• 6
• 7
• 8

參考資料?
https://www.tensorflow.org/api_docs/python/state_ops/variables#Variable?
https://www.tensorflow.org/api_docs/python/state_ops/sharing_variables#get_variable?
https://www.tensorflow.org/versions/r0.10/api_docs/python/constant_op/?
https://www.tensorflow.org/api_docs/python/state_ops/

轉(zhuǎn)自：http://blog.csdn.net/u012436149/article/details/53696970

總結(jié)

以上是生活随笔為你收集整理的variable与get_variable的全部?jī)?nèi)容，希望文章能夠幫你解決所遇到的問(wèn)題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。