小能豆

BasicRNNCell 中的内部变量

py

我有以下示例代码要测试BasicRNNCell。我想获取其内部矩阵,以便我可以使用自己的代码计算 的值,output_resnewstate_res确保我可以重现 的值output_resnewstate_res

在 tensorflow 源代码中,它显示output = new_state = act(W * input + U * state + B)。有人知道如何获取W和 吗U?(我试图访问cell._kernel,但不可用。)

$ cat ./main.py
#!/usr/bin/env python
# vim: set noexpandtab tabstop=2 shiftwidth=2 softtabstop=-1 fileencoding=utf-8:

import tensorflow as tf
import numpy as np

batch_size = 4
vector_size = 3

inputs = tf.placeholder(
        tf.float32
        , [batch_size, vector_size]
        )

num_units = 2
state = tf.zeros([batch_size, num_units], tf.float32)

cell = tf.contrib.rnn.BasicRNNCell(num_units=num_units)
output, newstate = cell(inputs = inputs, state = state)

X = np.zeros([batch_size, vector_size])
#X = np.ones([batch_size, vector_size])
with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())

    output_res, newstate_res = sess.run([output, newstate], feed_dict = {inputs: X})
    print(output_res)
    print(newstate_res)
sess.close()

$ ./main.py
[[ 0.  0.]
 [ 0.  0.]
 [ 0.  0.]
 [ 0.  0.]]
[[ 0.  0.]
 [ 0.  0.]
 [ 0.  0.]
 [ 0.  0.]]

阅读 23

收藏
2024-12-24

共1个答案

小能豆

简短回答:您意识到自己在追求cell._kernel。以下是使用属性获取内核(和偏差)的一些代码variables,该属性在大多数 TensorFlow RNN 中都有:

import tensorflow as tf
import numpy as np

batch_size = 4
vector_size = 3
inputs = tf.placeholder(tf.float32, [batch_size, vector_size])

num_units = 2
state = tf.zeros([batch_size, num_units], tf.float32)

cell = tf.contrib.rnn.BasicRNNCell(num_units=num_units)
output, newstate = cell(inputs=inputs, state=state)

print("Output of cell.variables is a list of Tensors:")
print(cell.variables)
kernel, bias = cell.variables

X = np.zeros([batch_size, vector_size])
with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())

    output_, newstate_, k_, b_ = sess.run(
        [output, newstate, kernel, bias], feed_dict = {inputs: X})
    print("Output:")
    print(output_)
    print("New State == Output:")
    print(newstate_)
    print("\nKernel:")
    print(k_)
    print("\nBias:")
    print(b_)

输出

Output of cell.variables is a list of Tensors:
[<tf.Variable 'basic_rnn_cell/kernel:0' shape=(5, 2) dtype=float32_ref>, 
<tf.Variable 'basic_rnn_cell/bias:0' shape=(2,) dtype=float32_ref>]
Output:
[[ 0.  0.]
 [ 0.  0.]
 [ 0.  0.]
 [ 0.  0.]]
New State == Output:
[[ 0.  0.]
 [ 0.  0.]
 [ 0.  0.]
 [ 0.  0.]]

Kernel:
[[ 0.41417515 -0.64997244]
 [-0.40868729 -0.90995187]
 [ 0.62134564 -0.88962835]
 [-0.35878009 -0.25680023]
 [ 0.35606658 -0.83596271]]

Bias:
[ 0.  0.]

长答案:您还询问如何获得 W 和 U。让我复制实现call并讨论 W 和 U 在哪里。

def call(self, inputs, state):
     """Most basic RNN: output = new_state = act(W * input + U * state + B)."""

    gate_inputs = math_ops.matmul(
        array_ops.concat([inputs, state], 1), self._kernel)
    gate_inputs = nn_ops.bias_add(gate_inputs, self._bias)
    output = self._activation(gate_inputs)
    return output, output

看起来不像是W 和 U,但它们确实存在。本质上,vector_size内核的第一行是 W,num_units内核的下一行是 U。也许看看 LaTeX 中的元素数学会有所帮助:

1.png

我使用m作为通用批次索引,v作为vector_sizen作为num_unitsb作为batch_size。此外,[ ; ]表示连接。由于 TensorFlow 是批次为主的,因此实现通常使用右乘矩阵。

由于这是一个非常基本的 RNN,output == new_state下一次迭代的“历史”只是当前迭代的输出。

2024-12-24