OpenVINO ERROR whenConverting Object Detection Models

[Petrica]

I have a problem running when trying to use the convertion script with the following parameters: 

mo_tf.py --saved_model_dir ./test/saved_model --transformations_config /opt/intel/openvino_2021/deployment_tools/model_optimizer/extensions/front/tf/ssd_support_api_v2.4.json --tensorflow_object_detection_api_pipeline_config ./test/pipeline.config --data_type FP16

 it shows error :

Exception occurred during running replacer "ObjectDetectionAPIOutputReplacement" (<class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIOutputReplacement'>): Graph contains 1 node after executing <class 'extensions.front.tf.ObjectDetectionAPI.ObjectDetectionAPIOutputReplacement'>. It considered as error because resulting IR will be empty which is not usual

[Wan_Intel]

Hi Petrica,

Thanks for reaching out to us.

We noticed that you posted another question in the OpenVINO™ community, the commands that you used for both questions are similar.

May I know that are you using the same model for both questions?

Regards,

Wan

[Petrica]

Yes, I am using slightly the same models, one is optimized during training with adam optimizer, and the other with momentum optimizer. I want to add that I was able to convert an older version again, and it looks like is not working any more.

[Wan_Intel · Accepted solution]

Hi Petrica,

Thanks for your information.

Could you please share your model that trained with momentum optimizer here?

Or you can share it on the following case, so we can update you in one case.
https://community.intel.com/t5/Intel-Distribution-of-OpenVINO/Unable-to-run-conversion-for-a-tensorflow-quot-saved-model-quot/m-p/1377847

Thanks, and regards,

Wan

[Wan_Intel]

Hi Petrica,

Thank you for your question.

If you need any additional information from Intel, please submit a new question as this thread is no longer being monitored.

Regards,

Wan