Error when converting YoloV4 model using the converter

Trying to convert YoloV4 model to IR using converter.py:

python3 /home/open_model_zoo/tools/downloader/converter.py --name yolo-v4-tf --mo /opt/intel/openvino/deployment_tools/model_optimizer/mo.py

But get the error:

tensorflow.python.framework.errors_impl.FailedPreconditionError: /home/open_model_zoo/public/yolo-v4-tf/yolo-v4.savedmodel is not a directory
Traceback (most recent call last):
  File "/home/open_model_zoo/models/public/yolo-v4-tf/pre-convert.py", line 40, in <module>
    main()
  File "/home/open_model_zoo/models/public/yolo-v4-tf/pre-convert.py", line 36, in main
    ], check=True)
  File "/usr/lib/python3.6/subprocess.py", line 438, in run
    output=stdout, stderr=stderr)
subprocess.CalledProcessError: Command '['/usr/bin/python3', '--', '/home/open_model_zoo/public/yolo-v4-tf/keras-YOLOv3-model-set/tools/model_converter/convert.py', '/home/open_model_zoo/public/yolo-v4-tf/keras-YOLOv3-model-set/cfg/yolov4.cfg', '/home/open_model_zoo/public/yolo-v4-tf/yolov4.weights', '/home/open_model_zoo/public/yolo-v4-tf/yolo-v4.savedmodel', '--yolo4_reorder']' returned non-zero exit status 1.

FAILED:
yolo-v4-tf

[eaidova]

@vladimir-dudnik we internally discussed issue. The issue connected with running downloader and converter from different release versions. in 2021.4 we migrate on usage saved model instead keras h5 format to simplify model conversion. It made downloader postprocessing from 2021.3 is not compatible with pre-convert script in 2021.4. Now, it is solved by usage scripts from the same release.

[vladimir-dudnik]

@likholat You’ve closed the issue, so seems you solved the problem? If someone will look for yolo-v4 related issues it might be not clear if you close the issue because simple decide to stop work with yolo-v4 model or you have find and solved problem.