In Tensorflow, get the names of all the Tensors in a graph

python tensorflow tensorboard skflow
P. Camilleri picture P. Camilleri · Apr 27, 2016 · Viewed 138.8k times · Source

I am creating neural nets with Tensorflow and skflow; for some reason I want to get the values of some inner tensors for a given input, so I am using myClassifier.get_layer_value(input, "tensorName"), myClassifier being a skflow.estimators.TensorFlowEstimator.

However, I find it difficult to find the correct syntax of the tensor name, even knowing its name (and I'm getting confused between operation and tensors), so I'm using tensorboard to plot the graph and look for the name.

Is there a way to enumerate all the tensors in a graph without using tensorboard?

Answer

Yaroslav Bulatov picture Yaroslav Bulatov · Apr 27, 2016

You can do

[n.name for n in tf.get_default_graph().as_graph_def().node]

Also, if you are prototyping in an IPython notebook, you can show the graph directly in notebook, see show_graph function in Alexander's Deep Dream notebook

Related questions

Tensorflow Tensorboard default port

Is there a way to change the default port (6006) on TensorBoard so we could open multiple TensorBoards? Maybe an option like --port="8008"?

Read More
can't open tensorboard 0.0.0.0:6006 or localhost:6006

when i use tensorboard in cmd with win10 D:\python document\tensorflow>tensorboard --logdir=D:\python document\tensorflow Starting TensorBoard b'47' at http://0.0.0.0:6006 But when i open the webpage，it shows "dial tcp 0.0.0.0:6006: connectex: The requested address is …

Read More
Show training and validation accuracy in TensorFlow using same graph

I have a TensorFlow model, and one part of this model evaluates the accuracy. The accuracy is just another node in the tensorflow graph, that takes in logits and labels. When I want to plot the training accuracy, this is …

Read More