Welcome new user! You can search existing questions and answers without registering, but please register to post new questions and receive answers. Note that due to large amounts of spam attempts, your first three posts will be manually moderated, so please be patient.
Because of un-manageable amounts of spam despite our use of CAPTCHAs, email authorization, and other tools, we have discontinued this forum (see the 700k+ registered users with validated email addresses at right?). Please email us any questions or post bug reports and feature requests on GitHub at https://github.com/jevois -- The content below remains available for future reference.
Welcome to JeVois Tech Zone, where you can ask questions and receive answers from other members of the community.

error in Training custom TensorFlow networks for JeVois tutorial

0 votes

While doing the "Training custom TensorFlow networks for JeVois" tutorial at http://jevois.org/tutorials/UserTensorFlowTraining.html , I got error when executing the command:

toco \

--input_file=tf_files/retrained_graph.pb \

--output_file=tf_files/optimized_graph.pb \

--input_format=TENSORFLOW_GRAPHDEF \

--output_format=TENSORFLOW_GRAPHDEF \

--input_shape=1,${IMAGE_SIZE},${IMAGE_SIZE},3 \

--input_array=input \


The error message is: 

toco: error: argument --output_format: invalid choice: 'TENSORFLOW_GRAPHDEF' (choose from 'TFLITE', 'GRAPHVIZ_DOT')

asked Oct 23, 2018 in User questions by YigalB (250 points)

1 Answer

0 votes
yes, you are probably using a newer version of toco.

I just edited that tutorial yesterday about that but the new version is not yet live as I want to try it first.

Basically, you should locate the original tensorflow for poets: tflite android, and, in there, look at how they invoke toco. Instead of 2 calls to toco as was the case when we wrote the tutorial, now there is only one call to toco.
answered Oct 23, 2018 by JeVois (46,580 points)