tf.slice()解释

#tf.slice(input, begin, size)

>>> sess.run(n)
array([[[ 1,  2,  3,  4,  5,  6],
        [ 7,  8,  9, 10, 11, 12]]])

>>> n1 = tf.slice(n, [0,0,0], [-1,-1,3])

>>> sess.run(n1)
array([[[1, 2, 3],
        [7, 8, 9]]])

>>> n2 = tf.slice(n, [0,0,3], [-1,-1,3])

>>> sess.run(n2)
array([[[ 4,  5,  6],
        [10, 11, 12]]])
>>> 


>>> n3 = tf.slice(n, [0,0,3], [1,1,3])
>>> sess.run(n3)
array([[[4, 5, 6]]])

你可能感兴趣的:(tf.slice()解释)