about 11 . inception

[turkeyB]

Very interesting tutorial indeed. Thanks a lot.

i tried the code in 11.inception and i found that i can run a code like this

import tensorflow as tf hello = tf.constant('hello,tensor!') sess = tf.Session() print(sess.run(hello)) hello,tensor!
but when i insert your .py file like retrain.py and predict.py into /tensorflow folder, and run the code you listed above like this
# python retrain.py \ --bottleneck_dir=./workspace/bottlenecks \ --model_dir=./workspace/inception \ --output_graph=./workspace/flowers_graph.pb \ --output_labels=./workspace/flowers_labels.txt \ --image_dir ./workspace/flower_photos \ --how_many_training_steps 1000
it suggests that
Traceback (most recent call last): File "retrain.py", line 81, in <module> from tensorflow.python.framework import graph_util ImportError: cannot import name graph_util
and i dont know how to do next...
Any comment on what I do wrong.
many thanks
Eric

[Jin Kim]

Hi, What's your TensorFlow version?

[turkeyB]

it is 1.3.0
After I asked this question, I rebuilt tensorflow by scource and finally solved the problem. your tutorial is fabulous , thx anyway