I want to output the values from the input layer of my model, so I runned the following model optimizer command:
python mo_mxnet.py --input_model <MY_MODEL_TO_CONVERT_DIR>\model_file-0000.params --input_shape (1,3,128,128) --output_dir <MY_CONVERTED_IR_DIR> --output "/input_layer2"
So basically what we did was setting the /input_layer2 as the last layer of the model (output layer) in order to get the results from it.
The model convert runs successfully; a bin, mapping and xml is produced as usual; and the following output is given:
Model Optimizer arguments: Common parameters: - Path to the Input Model: C:\Program Files (x86)\IntelSWTools\openvino_2019.3.379\model_file-0000.params - Path for generated IR: C:\Program Files (x86)\IntelSWTools\openvino_2019.3.379\openvino_model/_input_layer2 - IR output name: model_file-0000 - Log level: ERROR - Batch: Not specified, inherited from the model - Input layers: Not specified, inherited from the model - Output layers: /input_layer2 - Input shapes: (1,3,128,128) - Mean values: Not specified - Scale values: Not specified - Scale factor: 1.0 - Precision of IR: FP32 - Enable fusing: True - Enable grouped convolutions fusing: True - Move mean values to preprocess section: False - Reverse input channels: False MXNet specific parameters: - Deploy-ready symbol file: None - Enable MXNet loader for models trained with MXNet version lower than 1.0.0: False - Prefix name for args.nd and argx.nd files: None - Pretrained model to be merged with the .nd files: None - Enable saving built parameters file from .nd files: False Model Optimizer version: 2019.3.0-408-gac8584cb7 [ SUCCESS ] Generated IR model. [ SUCCESS ] XML file: C:\Program Files (x86)\IntelSWTools\openvino_2019.3.379\openvino_model/_input_layer2\model_file-0000.xml [ SUCCESS ] BIN file: C:\Program Files (x86)\IntelSWTools\openvino_2019.3.379\openvino_model/_input_layer2\model_file-0000.bin [ SUCCESS ] Total execution time: 1.32 seconds.
The problem comes when I try to run the inference engine with this model, using the classification async script:
%MYPROFILE%\Documents\Intel\OpenVINO\inference_engine_samples_build\intel64\Release\classification_sample_async.exe -i <MY_TEST_IMAGES_DIR>\test.png -m <MY_CONVERTED_IR_DIR>\model_file-0000.xml -d CPU -l %USERPROFILE%\Documents\Intel\OpenVINO\inference_engine_samples_build\intel64\Release\cpu_extension.dll
Which gives the following error:
[ ERROR ] Sample supports topologies with 1 input only
I don't even understand the error message because I have only 1 input only, which is the input_layer.
Could you share your model with me to take a look? I can start a private message to share privately if needed.
Please tell me more about your model:
- What Topology is your model based on?
- Is it a pre-trained or custom trained model?
What Topology is your model based on?
Is based on Lenet, however I tested with a model with only one convolutional layer and one FC layer and the same error remains.
Is it a pre-trained or custom trained model?
Is a custom trained model, just one convolutional layer and a FC layer, I trained it for just one epoch because it is just a dummy model to test the input layer.
I can start a private message to share privately if needed.
Yes off course we can start a private chat, plz send me a PM. If we got it working, I will post here the solution and explanation.