openvinotoolkit / openvino

OpenVINO™ is an open-source toolkit for optimizing and deploying AI inference
https://docs.openvino.ai
Apache License 2.0
7.12k stars 2.23k forks source link

[Bug] Unexpected Exception - model optimizer fails with Bug Report for onnx model #16660

Closed darshanraj-intc closed 8 months ago

darshanraj-intc commented 1 year ago
System information (version)
Detailed description

mo --version Version of Model Optimizer is: 2022.3.0-9052-9752fafe8eb-releases/2022/3

Command used

mo --input_model onnx_models/vision/object_detection_segmentation/mask-rcnn/model/MaskRCNN-12-int8.onnx

Output

mo --input_model onnx_models/vision/object_detection_segmentation/mask-rcnn/model/MaskRCNN-12-int8.onnx
[ ERROR ]  -------------------------------------------------
[ ERROR ]  ----------------- INTERNAL ERROR ----------------
[ ERROR ]  Unexpected exception happened.
[ ERROR ]  Please contact Model Optimizer developers and forward the following information:
[ ERROR ]  Check 'unknown_operators.empty()' failed at src/frontends/onnx/frontend/src/core/graph.cpp:213:
OpenVINO does not support the following ONNX operations: com.microsoft.QLinearAdd, com.microsoft.QLinearConcat, com.microsoft.QLinearSigmoid

[ ERROR ]  Traceback (most recent call last):
  File "/home/imt-adl-blue/openvino/openvino_env/lib/python3.10/site-packages/openvino/tools/mo/main.py", line 50, in main
    ngraph_function = convert_model(**argv)
  File "/home/imt-adl-blue/openvino/openvino_env/lib/python3.10/site-packages/openvino/tools/mo/convert.py", line 47, in convert_model
    return _convert(**args)
  File "/home/imt-adl-blue/openvino/openvino_env/lib/python3.10/site-packages/openvino/tools/mo/convert_impl.py", line 937, in _convert
    raise e.with_traceback(None)
RuntimeError: Check 'unknown_operators.empty()' failed at src/frontends/onnx/frontend/src/core/graph.cpp:213:
OpenVINO does not support the following ONNX operations: com.microsoft.QLinearAdd, com.microsoft.QLinearConcat, com.microsoft.QLinearSigmoid

[ ERROR ]  ---------------- END OF BUG REPORT --------------
[ ERROR ]  -------------------------------------------------
Steps to reproduce
  1. Use instruction from here to download mask-rcnn model.

  2. Download model from https://github.com/onnx/models/tree/main/vision/object_detection_segmentation/mask-rcnn/model

  3. Run command shown here to convert onnx model

Issue submission checklist
mitruska commented 1 year ago

Hello @darshanraj-intc, thank you for reporting this issue. Conversion of the operations listed in the error message is currently unsupported by OpenVINO ONNX Frontend, but we will consider them as a feature request to be planned for future releases.

> OpenVINO does not support the following ONNX operations: 
> com.microsoft.QLinearAdd, com.microsoft.QLinearConcat, com.microsoft.QLinearSigmoid

(ref. 107242)

darshanraj-intc commented 1 year ago

@mitruska : Thanks for the clarification and quick response time.

OpenVINO links are misleading -> https://docs.openvino.ai/latest/openvino_docs_MO_DG_prepare_model_convert_model_onnx_specific_Convert_Mask_RCNN.html and https://docs.openvino.ai/latest/openvino_docs_MO_DG_prepare_model_convert_model_Convert_Model_From_ONNX.html

They point to incorrect models and/or incorrect commands to run.

It would really benefit to update the documentation (at least mark those sections in document with a comment as not supported), to provide appropriate guidance to the user.

akladiev commented 1 year ago

This issue will be closed in 2 weeks in case of no activity.

mlukasze commented 1 year ago

@kblaszczak-intel could you check a last comment of Darshan and lead internally process of fixing it, please?

github-actions[bot] commented 8 months ago

This issue will be closed in a week because of 9 months of no activity.

github-actions[bot] commented 8 months ago

This issue was closed because it has been stalled for 9 months with no activity.