Tensorflow: Visualize freezed model file (.pb) in Tensorboard

To visualize freezed Tensorflow model file use Tensorboard:
– Download import_pb_to_tensorboard.py (https://github.com/tensorflow/tensorflow/blob/master/tensorflow/python/tools/import_pb_to_tensorboard.py)
– Create a directory (/path/to/tmp/folder) where intermediate files will be created to visualize model in Tensorboard.
– Execute
– python import_pb_to_tensorboard.py –model_dir="/path/to/folder/tensorflow_model_frozen.pb" –log_dir="/path/to/tmp/folder"
NOTE: Wrapped values to param # model_dir and param # log_dir in "". Otherwise it might not work.
– tensorboard –logdir="/path/to/tmp/folder"
– By default Tensorboard runs on 6006, so open http://localhost:6006 in your web-browser to see Tensorboard.
– JFI: To inspect model without loading in web browser:
– tensorboard –inspect –logdir="/path/to/tmp/folder"