tensroflow：tf.sequence_mask

版权声明：本文为博主原创文章，未经博主允许不得转载。 https://blog.csdn.net/wxf2012301351/article/details/84401950

import tensorflow as tf

with tf.variable_scope('gg',reuse=tf.AUTO_REUSE):
    f=tf.placeholder(tf.int32,[None])
    f_f=tf.to_float(f > 0)
    user_emb_w = tf.get_variable("user_emb_w", [2, 5])
    g=tf.sequence_mask([1, 3, 2], 5,dtype=tf.float32)
    a=tf.expand_dims(g,-1)
    b=tf.tile(a,[1,1,5])
    t=tf.tile(g, [2,  5])
    print(t)
    num=tf.placeholder(tf.float32,[6,25])
    print(num)
    out=num*t
with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())

    f_g,g_g,t_g,out_g,a_g,b_g,f_fg=sess.run([f,g,t,out,a,b,f_f],feed_dict={f:[1,2.2,-2],num:np.random.rand(6,25)})
    print(g_g)
    print(sess.run(user_emb_w))
    print(f_g)
    print(t_g)
    print(out_g)
    print(a_g)
    print(b_g)
    print(f_fg)

Tensor("gg/Tile_1:0", shape=(6, 25), dtype=float32)
Tensor("gg/Placeholder_1:0", shape=(6, 25), dtype=float32)
[[1. 0. 0. 0. 0.]
 [1. 1. 1. 0. 0.]
 [1. 1. 0. 0. 0.]]
[[-0.29378003 -0.7962597  -0.40921748  0.32213974  0.8533299 ]
 [-0.09586155  0.8208542  -0.3765049  -0.20915747  0.09365308]]
[ 1  2 -2]
[[1. 0. 0. 0. 0. 1. 0. 0. 0. 0. 1. 0. 0. 0. 0. 1. 0. 0. 0. 0. 1. 0. 0. 0.
  0.]
 [1. 1. 1. 0. 0. 1. 1. 1. 0. 0. 1. 1. 1. 0. 0. 1. 1. 1. 0. 0. 1. 1. 1. 0.
  0.]
 [1. 1. 0. 0. 0. 1. 1. 0. 0. 0. 1. 1. 0. 0. 0. 1. 1. 0. 0. 0. 1. 1. 0. 0.
  0.]
 [1. 0. 0. 0. 0. 1. 0. 0. 0. 0. 1. 0. 0. 0. 0. 1. 0. 0. 0. 0. 1. 0. 0. 0.
  0.]
 [1. 1. 1. 0. 0. 1. 1. 1. 0. 0. 1. 1. 1. 0. 0. 1. 1. 1. 0. 0. 1. 1. 1. 0.
  0.]
 [1. 1. 0. 0. 0. 1. 1. 0. 0. 0. 1. 1. 0. 0. 0. 1. 1. 0. 0. 0. 1. 1. 0. 0.
  0.]]
[[2.70663410e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00
  0.00000000e+00 4.29145962e-01 0.00000000e+00 0.00000000e+00
  0.00000000e+00 0.00000000e+00 7.36451685e-01 0.00000000e+00
  0.00000000e+00 0.00000000e+00 0.00000000e+00 4.73007172e-01
  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00
  4.07942593e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00
  0.00000000e+00]
 [3.23744744e-01 2.89273918e-01 6.70506179e-01 0.00000000e+00
  0.00000000e+00 9.73211974e-02 3.00509661e-01 4.44553137e-01
  0.00000000e+00 0.00000000e+00 8.50156188e-01 5.30314028e-01
  2.36548066e-01 0.00000000e+00 0.00000000e+00 5.60386956e-01
  8.44906509e-01 6.66610241e-01 0.00000000e+00 0.00000000e+00
  3.69455725e-01 6.13740347e-02 9.97683644e-01 0.00000000e+00
  0.00000000e+00]
 [2.84169704e-01 2.20960617e-01 0.00000000e+00 0.00000000e+00
  0.00000000e+00 5.39233617e-04 1.87725261e-01 0.00000000e+00
  0.00000000e+00 0.00000000e+00 6.17116392e-01 5.16861856e-01
  0.00000000e+00 0.00000000e+00 0.00000000e+00 4.22688603e-01
  2.57003665e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00
  4.49555218e-01 5.30027866e-01 0.00000000e+00 0.00000000e+00
  0.00000000e+00]
 [1.13443434e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00
  0.00000000e+00 2.49865055e-01 0.00000000e+00 0.00000000e+00
  0.00000000e+00 0.00000000e+00 2.85478588e-02 0.00000000e+00
  0.00000000e+00 0.00000000e+00 0.00000000e+00 3.82318974e-01
  0.00000000e+00 0.00000000e+00 0.00000000e+00 0.00000000e+00
  2.68762916e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00
  0.00000000e+00]
 [7.23953605e-01 1.17065370e-01 4.79779989e-01 0.00000000e+00
  0.00000000e+00 7.77306557e-01 7.41112471e-01 4.85356927e-01
  0.00000000e+00 0.00000000e+00 7.84220934e-01 2.01409265e-01
  9.78608072e-01 0.00000000e+00 0.00000000e+00 6.99940920e-01
  3.87414396e-01 4.81573045e-01 0.00000000e+00 0.00000000e+00
  3.62429358e-02 1.97894201e-01 4.50255901e-01 0.00000000e+00
  0.00000000e+00]
 [9.00696814e-01 6.73199296e-01 0.00000000e+00 0.00000000e+00
  0.00000000e+00 2.99850136e-01 7.68026292e-01 0.00000000e+00
  0.00000000e+00 0.00000000e+00 9.06648815e-01 7.42762029e-01
  0.00000000e+00 0.00000000e+00 0.00000000e+00 9.68360782e-01
  2.91729569e-01 0.00000000e+00 0.00000000e+00 0.00000000e+00
  3.13812852e-01 4.07107353e-01 0.00000000e+00 0.00000000e+00
  0.00000000e+00]]
[[[1.]
  [0.]
  [0.]
  [0.]
  [0.]]

 [[1.]
  [1.]
  [1.]
  [0.]
  [0.]]

 [[1.]
  [1.]
  [0.]
  [0.]
  [0.]]]
[[[1. 1. 1. 1. 1.]
  [0. 0. 0. 0. 0.]
  [0. 0. 0. 0. 0.]
  [0. 0. 0. 0. 0.]
  [0. 0. 0. 0. 0.]]

 [[1. 1. 1. 1. 1.]
  [1. 1. 1. 1. 1.]
  [1. 1. 1. 1. 1.]
  [0. 0. 0. 0. 0.]
  [0. 0. 0. 0. 0.]]

 [[1. 1. 1. 1. 1.]
  [1. 1. 1. 1. 1.]
  [0. 0. 0. 0. 0.]
  [0. 0. 0. 0. 0.]
  [0. 0. 0. 0. 0.]]]
[1. 1. 0.]