Intel® Distribution of OpenVINO™ Toolkit
Community support and discussions about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all things computer vision-related on Intel® platforms.
5762 Discussions

Raspbian Toolkit version 2020.1.023 ngraph RuntimeError

Williams__Rhys
Beginner
216 Views

Hey there! First of all I just wanted to say that I'm having a great time with the Neural Compute Stick 2, it really is incredible!

I've just recently updated to the 2020.1.023 version of OpenVINO toolkit on Ubuntu 16.04, likewise I've updated the Raspbian version accordingly. However after freezing and optimizing my Keras model for inference I'm getting the dreaded ngraph issue I've seen pop in these forums:

  File "ie_api.pyx", line 980, in openvino.inference_engine.ie_api.IENetwork.__cinit__
RuntimeError: Check 'axis < static_cast<size_t>(input_rank)' failed at /teamcity/work/scoring_engine_build/releases_2020_1/ngraph/src/ngraph/op/gather.cpp:140:
While validating node 'Gather[Gather_189](patternLabel_185: float{10,20,30}, patternLabel_186: int64_t{5}, patternLabel_188: int64_t{1}) -> (??)':
The axis must => 0 and <= input_rank (axis: 4294967295).

I took the advice I've seen you guys give and tried an IR optimised the last toolkit version and it runs just fine. I guess my question is what's the latest update as to this issue? I'd love to continue using Tensorflow 1.15. For reference I'm using the optimization tool as follows:

 

python3 /opt/intel/openvino_2020.1.023/deployment_tools/model_optimizer/mo_tf.py  \
--input_model avoidance.pb \
--input_shape [1,224,224,3] --scale 255 \
--data_type FP16 --output activation_1/Sigmoid \
--output_dir {output_dir} \
--reverse_input_channels

Cheers for your help in advance!

 

0 Kudos
1 Reply
RandallMan_B_Intel
216 Views

Hi Williams,

Thank you for reaching out.

You can check a similar issue in the following link: https://software.intel.com/en-us/forums/intel-distribution-of-openvino-toolkit/topic/848966

Hope this helps.

Regards,

Randall B.

Reply