如何在Tensorflow中显示隐藏层输出 [英] How to show hidden layer outputs in Tensorflow

问题描述

将模型与其存储的protobuf版本进行比较时，我的输出存在差异(通过此转换脚本).为了调试，我分别比较了这两层.对于权重和测试序列期间的实际图层输出，我会收到相同的输出，因此我不确定如何访问隐藏的图层.

I'm having differences of the outputs when comparing a model with its stored protobuf version (via this conversion script). For debugging I'm comparing both layers respectively. For the weights and the actual layer output during a test sequence I receive the identical outputs, thus I'm not sure how to access the hidden layers.

这是我加载图层的方式

    input  = graph.get_tensor_by_name("lstm_1_input_1:0")
    layer1 = graph.get_tensor_by_name("lstm_1_1/kernel:0")
    layer2 = graph.get_tensor_by_name("lstm_1_1/recurrent_kernel:0")
    layer3 = graph.get_tensor_by_name("time_distributed_1_1/kernel:0")
    output = graph.get_tensor_by_name("activation_1_1/div:0")

这就是我想展示各个元素的方式.

Here is the way what I thought to show the respective elements.

显示权重:

with tf.Session(graph=graph) as sess:
       print sess.run(layer1)
       print sess.run(layer2)
       print sess.run(layer3)

显示输出:

with tf.Session(graph=graph) as sess:
    y_out, l1_out, l2_out, l3_out = sess.run([output, layer1, layer2, layer3], feed_dict={input: X_test})

使用此代码sess.run(layer1) == sess.run(layer1,feed_dict={input:X_test})，该代码不正确.

With this code sess.run(layer1) == sess.run(layer1,feed_dict={input:X_test}) which shouldn't be true.

有人可以帮我吗?

推荐答案

运行sess.run(layer1)时，您要告诉tensorflow计算layer1张量的值，即...

When you run sess.run(layer1), you're telling tensorflow to compute the value of layer1 tensor, which is ...

layer1 = graph.get_tensor_by_name("lstm_1_1/kernel:0")

...根据您的定义.请注意，LSTM内核是weights变量.它不取决于input，这就是为什么您使用sess.run(layer1, feed_dict={input:X_test})得到相同结果的原因.如果提供了输入，这并不像tensorflow正在计算输出-它正在计算指定的张量，在这种情况下为layer1.

... according to your definition. Note that LSTM kernel is the weights variable. It does not depend on the input, that's why you get the same result with sess.run(layer1, feed_dict={input:X_test}). It's not like tensorflow is computing the output if the input is provided -- it's computing the specified tensor(s), in this case layer1.

什么时候input重要呢?有依赖关系时.例如:

When does input matter then? When there is a dependency on it. For example:

• sess.run(output).如果没有input或任何允许计算input的张量，它将根本无法工作.
• 优化操作，例如tf.train.AdapOptimizer(...).minimize(loss).运行此操作将更改layer1，但也需要input来完成.

• sess.run(output). It simply won't work without an input, or any tensor that will allow to compute the input.
• The optimization op, such as tf.train.AdapOptimizer(...).minimize(loss). Running this op will change layer1, but it also needs the input to do so.

这篇关于如何在Tensorflow中显示隐藏层输出的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！