How to restore Tensorflow model from .pb file in python?

Question

I have an tensorflow .pb file which I would like to load into python DNN, restore the graph and get the predictions. I am doing this to test out whether the .pb file created can make the predictions similar to the normal Saver.save() model.

My basic problem is am getting a very different value of predictions when I make them on Android using the above mentioned .pb file

My .pb file creation code:

frozen_graph = tf.graph_util.convert_variables_to_constants(
        session,
        session.graph_def,
        ['outputLayer/Softmax']
    )
with open('frozen_model.pb', 'wb') as f:
  f.write(frozen_graph.SerializeToString())

So I have two major concerns:

How can I load the above mentioned .pb file to python Tensorflow model ?
Why am I getting completely different values of prediction in python and android ?

MaxPowers · Accepted Answer · 2018-06-12 04:07:57Z

33

The following code will read the model and print out the names of the nodes in the graph.

import tensorflow as tf
from tensorflow.python.platform import gfile
GRAPH_PB_PATH = './frozen_model.pb'
with tf.Session() as sess:
   print("load graph")
   with gfile.FastGFile(GRAPH_PB_PATH,'rb') as f:
       graph_def = tf.GraphDef()
   graph_def.ParseFromString(f.read())
   sess.graph.as_default()
   tf.import_graph_def(graph_def, name='')
   graph_nodes=[n for n in graph_def.node]
   names = []
   for t in graph_nodes:
      names.append(t.name)
   print(names)

You are freezing the graph properly that is why you are getting different results basically weights are not getting stored in your model. You can use the freeze_graph.py (link) for getting a correctly stored graph.

edited Jun 12, 2018 at 4:07

MaxPowers

5,5292 gold badges50 silver badges70 bronze badges

answered Jun 2, 2018 at 0:20

Pranjal Sahu

1,4694 gold badges18 silver badges27 bronze badges

Sign up to request clarification or add additional context in comments.

6 Comments

mrgloom Over a year ago

What is sess.graph.as_default() doing?

volperossa Over a year ago

i get on graph_def.ParseFromString(f.read()) DecodeError: Error parsing message

tigertang Over a year ago

Use tf.gfile.GFile instead of gfile.FastGFile in 2019

user11758292 Over a year ago

When I try to display node names with your program and facenet model , it is giving me this error

ValueError: Input 0 of node InceptionResnetV1/Conv2d_1a_3x3/BatchNorm/cond/Switch was passed float from phase_train:0 incompatible with expected bool.

Do you have any idea why is it happening ? Thanks

Pranjal Sahu Over a year ago

@Sneha You are probably passing wrong datatype. It expects bool but it is getting float.

|

caylus · Accepted Answer · 2020-05-07 11:48:53Z

9

Here is the updated code for tensorflow 2.

import tensorflow as tf

GRAPH_PB_PATH = './frozen_model.pb'
with tf.compat.v1.Session() as sess:
   print("load graph")
   with tf.io.gfile.GFile(GRAPH_PB_PATH,'rb') as f:
       graph_def = tf.compat.v1.GraphDef()
   graph_def.ParseFromString(f.read())
   sess.graph.as_default()
   tf.import_graph_def(graph_def, name='')
   graph_nodes=[n for n in graph_def.node]
   names = []
   for t in graph_nodes:
      names.append(t.name)
   print(names)

edited May 7, 2020 at 11:48

answered May 7, 2020 at 11:43

caylus

4804 silver badges12 bronze badges

1 Comment

JoyfulPanda Over a year ago

I got DecodeError: Error parsing message with type 'tensorflow.GraphDef'.

Collectives™ on Stack Overflow

How to restore Tensorflow model from .pb file in python?

2 Answers 2

6 Comments

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

6 Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related