tf.nn.embedding_lookup和tf.gather实例

tf.nn.embedding_lookup的作用就是找到embedding data中对应行下的vector

tf.nn.embedding_lookup(params, ids, partition_strategy='mod', name=None, validate_indices=True, max_norm=None)#寻找params中索引为ids的vector

 

import pandas as pd
import numpy as np
from pandas import Series,DataFrame
import tensorflow as tf

p=tf.Variable(tf.random_normal([5,1]))#生成5*1的张量
b = tf.nn.embedding_lookup(p, [1, 3])#查找张量中的序号为1和3的
 
with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    print(sess.run(b))
    print('------')
    print(sess.run(p))

[[0.5533718]
 [0.6643605]]
------
[[ 0.16127495]
 [ 0.5533718 ]
 [-0.6191211 ]
 [ 0.6643605 ]
 [-1.9890579 ]]

tf.gather是从'params'中，按照axis坐标和indices标注的元素下标，把这些元素抽取出来组成新的tensor.

tf.gather(params, indices, validate_indices=None, name=None, axis=0)

import tensorflow as tf
temp = tf.range(0,10)*10 + tf.constant(1,shape=[10])
temp2 = tf.gather(temp,[2,5,8])

with tf.Session() as sess:
    print(sess.run(temp))
    print('------')
    print(sess.run(temp2))

[ 1 11 21 31 41 51 61 71 81 91]
------
[21 51 81]

扫描下方二维码关注领取程序员必备千套ppt模板，300本精选好书，丰富面经：

有酒有风