Error optimizing tensorflow2 model

[AR92]

Hi,

i have trained this  http://download.tensorflow.org/models/object_detection/tf2/20200711/ssd_mobilenet_v2_320x320_coco17_tpu-8.tar.gz model  with tensorflow2 , now i am trying to optimize it with mo_tf.py file with below command 

 python mo_tf.py --saved_model_dir E:\tensorflow_models\my_mobilenetv2\saved_model\ --tensorflow_object_detection_api_pipeline_config E:\tensorflow_models\my_mobilenetv2\pipeline.config --tensorflow_use_custom_operations_config ssd_v2_support.json --data_type FP32 --reverse_input_channels

but i got below error:

Model Optimizer arguments:
Common parameters:
        - Path to the Input Model:      None
        - Path for generated IR:        C:\Program Files (x86)\IntelSWTools\openvino_2020.3.194\deployment_tools\model_optimizer\.
        - IR output name:       saved_model
        - Log level:    ERROR
        - Batch:        Not specified, inherited from the model
        - Input layers:         Not specified, inherited from the model
        - Output layers:        Not specified, inherited from the model
        - Input shapes:         Not specified, inherited from the model
        - Mean values:  Not specified
        - Scale values:         Not specified
        - Scale factor:         Not specified
        - Precision of IR:      FP32
        - Enable fusing:        True
        - Enable grouped convolutions fusing:   True
        - Move mean values to preprocess section:       False
        - Reverse input channels:       True
TensorFlow specific parameters:
        - Input model in text protobuf format:  True
        - Path to model dump for TensorBoard:   None
        - List of shared libraries with TensorFlow custom layers implementation:        None
        - Update the configuration file with input/output node names:   None
        - Use configuration file used to generate the model with Object Detection API:  E:\tensorflow_models\my_mobilenetv2\pipeline.config
        - Use the config file:  C:\Program Files (x86)\IntelSWTools\openvino_2020.3.194\deployment_tools\model_optimizer\ssd_v2_support.json
Model Optimizer version:
2020-09-04 18:57:47.481179: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'cudart64_100.dll'; dlerror: cudart64_100.dll not found
2020-09-04 18:57:47.481332: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
2020-09-04 18:57:51.581332: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'nvcuda.dll'; dlerror: nvcuda.dll not found
2020-09-04 18:57:51.581465: E tensorflow/stream_executor/cuda/cuda_driver.cc:318] failed call to cuInit: UNKNOWN ERROR (303)
2020-09-04 18:57:51.588243: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:169] retrieving CUDA diagnostic information for host: UECLAB1
2020-09-04 18:57:51.588455: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:176] hostname: UECLAB1
2020-09-04 18:57:51.589145: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2
[ ERROR ]  Exception occurred during running replacer "REPLACEMENT_ID" (<class 'extensions.load.tf.loader.TFLoader'>): Unexpected exception happened during extracting attributes for node ssd_mobile_net_v2keras_feature_extractor/FeatureMaps/layer_19_2_Conv2d_5_3x3_s2_128_batchnorm/moving_variance/Read/ReadVariableOp.
Original exception message: 'ascii' codec can't decode byte 0x93 in position 183: ordinal not in range(128)

with --tensorflow_use_custom_operations_config ssd_v2_support.json i also tried  ssd_support_api_v1.15.

previously i have trained a ssd inception v2 module with tensorflow1.15 and optimization with that worked.

so is tensorflow 2 is supported with openvino

Thanks & Regards

Amit Rawat 

[Iffa_Intel]

Greetings,

Openvino does support TF2. You can refer here for more details: https://docs.openvinotoolkit.org/latest/openvino_docs_MO_DG_prepare_model_convert_model_Convert_Model_From_TensorFlow.html#Convert_From_TF2X

the error 'ascii' codec can't decode byte 0x93 in position 183: ordinal not in range(128) indicates some type of Unicode Decode error. Perhaps you can try these workaround:

https://github.com/facebookresearch/mmf/issues/101

https://github.com/OfflineIMAP/offlineimap/issues/388

Sincerely,

Iffa

[Iffa_Intel]

Greetings,

Intel will no longer monitor this thread since we have provided a solution. If you need any additional information from Intel, please submit a new question. 

Sincerely,

Iffa