Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[Bug] Cannot convert models with MMCVModulatedReformConv2d from onnx to openvino #2835

Open
3 tasks done
wy787 opened this issue Nov 4, 2024 · 0 comments
Open
3 tasks done

Comments

@wy787
Copy link

wy787 commented Nov 4, 2024

Checklist

  • I have searched related issues but cannot get the expected help.
  • 2. I have read the FAQ documentation but cannot get the expected help.
  • 3. The bug has not been fixed in the latest version.

Describe the bug

When I tried to convert the mmocr-dbnetpp-dcnv2 model to openvino through the script tools/deploy, the following problem occurred:
I was able to successfully export the model as onnx, but failed when converting onnx to openvino.

When I switched to the mo command, it showed that openvino does not support mmdeploy.MMCVModulatedReformConv2d.

Reproduction

python tools/deploy.py configs/mmocr/text-detection/text-detection_openvino_dynamic-640x640.py ../mmocr-main/configs/textdet/dbnetpp/dbnetpp_resnet50-dcnv2_fpnc_1200e_icdar2015.py ../mmocr-main/work_dirs/dbnetpp_resnet50-dcnv2_fpnc_1200e_icdar2015/epoch_130.pth ../dataset/0.bmp --test-img ../dataset/0.bmp --work-dir output_mmocr/output_openvino/ --device cpu --show

mo --input_model="output_mmocr/output_openvino/end2end.onnx" --output_dir="output_mmocr/output_openvino/" --output="output" --input="input" --input_shape="[1, 3, 640, 640]"

Environment

11/04 08:20:24 - mmengine - INFO - sys.platform: linux
11/04 08:20:24 - mmengine - INFO - Python: 3.8.20 (default, Oct  3 2024, 15:24:27) [GCC 11.2.0]
11/04 08:20:24 - mmengine - INFO - CUDA available: True
11/04 08:20:24 - mmengine - INFO - GCC: gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0
11/04 08:20:24 - mmengine - INFO - PyTorch: 1.8.0+cu111

11/04 08:20:24 - mmengine - INFO - TorchVision: 0.9.0+cu111
11/04 08:20:24 - mmengine - INFO - OpenCV: 4.10.0
11/04 08:20:24 - mmengine - INFO - MMEngine: 0.10.5
11/04 08:20:24 - mmengine - INFO - MMCV: 2.0.0rc4
11/04 08:20:24 - mmengine - INFO - MMCV Compiler: GCC 7.3
11/04 08:20:24 - mmengine - INFO - MMCV CUDA Compiler: 11.1
11/04 08:20:24 - mmengine - INFO - MMDeploy: 1.3.1+

11/04 08:20:24 - mmengine - INFO - ONNXRuntime: 1.8.1
11/04 08:20:24 - mmengine - INFO - ONNXRuntime-gpu:     None
11/04 08:20:24 - mmengine - INFO - ONNXRuntime custom ops:      Available
11/04 08:20:24 - mmengine - INFO - openvino:    2023.2.0
11/04 08:20:24 - mmengine - INFO - mmocr:       1.0.1

Error traceback

File "/root/miniconda3/envs/mmlab/lib/python3.8/subprocess.py", line 516, in run
    raise CalledProcessError(retcode, process.args,
subprocess.CalledProcessError: Command 'mo --input_model="output_mmocr/output_openvino/end2end.onnx" --output_dir="output_mmocr/output_openvino/" --output="output"

RuntimeError: Check 'error_message.empty()' failed at src/frontends/onnx/frontend/src/frontend.cpp:124:
OpenVINO does not support the following ONNX operations: mmdeploy.MMCVModulatedDeformConv2d
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant