Hi,
I have a Single Shot Detector (SSD) PyTorch model that I trained using mmdetection and then converted to openvino format. I am trying to covert it to blob for use on an OAK-D Lite device but I'm getting the following error. I used the blobconverter pip package.
python3 -m blobconverter --openvino-xml wd/end2end.xml --openvino-bin wd/end2end.bin --shaves 2
{
"exit_code": 1,
"message": "Command failed with exit code 1, command: /opt/intel/openvino/deployment_tools/inference_engine/lib/intel64/myriad_compile -m /tmp/blobconverter/74659247ee464cc2b458b385d14372ff/end2end/FP16/end2end.xml -o /tmp/blobconverter/74659247ee464cc2b458b385d14372ff/end2end/FP16/end2end.blob -c /tmp/blobconverter/74659247ee464cc2b458b385d14372ff/myriad_compile_config.txt -ip U8",
"stderr": "[ GENERAL_ERROR ] \n/home/jenkins/agent/workspace/private-ci/ie/build-linux-ubuntu20/b/repos/openvino/inference-engine/src/vpu/common/src/ngraph/transformations/dynamic_to_static_shape.cpp:197 DynamicToStaticShape transformation encountered dynamic node Tile_1405 of type Tile ver. 0, but only [Ceiling ver. 0, Clamp ver. 0, Concat ver. 0, Convert ver. 0, Exp ver. 0, ExpGatherElements ver. 0, Floor ver. 0, Log ver. 0, MatMul ver. 0, Relu ver. 0, ...] types are supported for dynamic nodes\n",
"stdout": "Inference Engine: \n\tIE version ......... 2021.4.0\n\tBuild ........... 2021.4.0-3839-cd81789d294-releases/2021/4\n\u001b[1;33m[Warning][VPU][Config] Deprecated option was used : VPU_MYRIAD_PLATFORM\u001b[0m\n"
}
I also have the model in ONNX format and got a different error when I try to convert it from ONNX to blob
python3 -m blobconverter --onnx-model wd/end2end.onnx --shaves 2
{
"exit_code": 1,
"message": "Command failed with exit code 1, command: /app/venvs/venv2021_4/bin/python /app/model_compiler/openvino_2021.4/converter.py --precisions FP16 --output_dir /tmp/blobconverter/043440a652594c5f8623dd8e00035983 --download_dir /tmp/blobconverter/043440a652594c5f8623dd8e00035983 --name end2end --model_root /tmp/blobconverter/043440a652594c5f8623dd8e00035983",
"stderr": "[ ERROR ] Cannot infer shapes or values for node \"Conv_6\".\n[ ERROR ] Data after padding has dimension less than window size. Possible reason of error is incorrectly specified model input shape(s).\n[ ERROR ] \n[ ERROR ] It can happen due to bug in custom shape infer function <function Convolution.infer at 0x7fcfd6af2e50>.\n[ ERROR ] Or because the node inputs have incorrect values/shapes.\n[ ERROR ] Or because input shapes are incorrect (embedded to the model or passed via --input_shape).\n[ ERROR ] Run Model Optimizer with --log_level=DEBUG for more information.\n[ ERROR ] Exception occurred during running replacer \"REPLACEMENT_ID\" (<class 'extensions.middle.PartialInfer.PartialInfer'>): Stopped shape/value propagation at \"Conv_6\" node. \n For more information please refer to Model Optimizer FAQ, question #38. (https://docs.openvinotoolkit.org/latest/openvino_docs_MO_DG_prepare_model_Model_Optimizer_FAQ.html?question=38#question-38)\n",
"stdout": "========== Converting end2end to IR (FP16)\nConversion command: /app/venvs/venv2021_4/bin/python -m mo --framework=onnx --data_type=FP16 --output_dir=/tmp/blobconverter/043440a652594c5f8623dd8e00035983/end2end/FP16 --model_name=end2end --data_type=FP16 --input_model=/tmp/blobconverter/043440a652594c5f8623dd8e00035983/end2end/FP16/end2end.onnx\n\nModel Optimizer arguments:\nCommon parameters:\n\t- Path to the Input Model: \t/tmp/blobconverter/043440a652594c5f8623dd8e00035983/end2end/FP16/end2end.onnx\n\t- Path for generated IR: \t/tmp/blobconverter/043440a652594c5f8623dd8e00035983/end2end/FP16\n\t- IR output name: \tend2end\n\t- Log level: \tERROR\n\t- Batch: \tNot specified, inherited from the model\n\t- Input layers: \tNot specified, inherited from the model\n\t- Output layers: \tNot specified, inherited from the model\n\t- Input shapes: \tNot specified, inherited from the model\n\t- Mean values: \tNot specified\n\t- Scale values: \tNot specified\n\t- Scale factor: \tNot specified\n\t- Precision of IR: \tFP16\n\t- Enable fusing: \tTrue\n\t- Enable grouped convolutions fusing: \tTrue\n\t- Move mean values to preprocess section: \tNone\n\t- Reverse input channels: \tFalse\nONNX specific parameters:\n[ WARNING ] Failed to import Inference Engine Python API in: PYTHONPATH\n[ WARNING ] libpython3.6m.so.1.0: cannot open shared object file: No such file or directory\n[ WARNING ] Failed to import Inference Engine Python API in: /opt/intel/openvino_2021.4.582/python/python3.8\n[ WARNING ] libpython3.6m.so.1.0: cannot open shared object file: No such file or directory\n[ WARNING ] Could not find the Inference Engine Python API. At this moment, the Inference Engine dependency is not required, but will be required in future releases.\n[ WARNING ] Consider building the Inference Engine Python API from sources or try to install OpenVINO (TM) Toolkit using \"install_prerequisites.sh\"\nModel Optimizer version: \t2021.4.0-3839-cd81789d294-releases/2021/4\nFAILED:\nend2end\n"
}