Openvino2023 toolkit

[Mazr]

I want a guide on where to find mo.py on openvino2023, for openvino2021 it was easy to get because it's located in ==> deployment_tools/model_optimizer.

or if there is another way i can convert yolov3.weight to get IR format 

mostly whith openvino2021 we used to use this command "python mo.py --input_model yolo3.weights --data_type FP16
"

which involves mo.py but i can't find it in openvino2023

thanks...

[Megat_Intel]

Hi Mazr,

Thank you for reaching out to us.

Starting from 2022 Releases, tools such as Model Optimizer, Benchmark Tool, Accuracy Checker, Annotation Converter, Post-Training Optimization Tool, Model Downloader, and other Open Model Zoo Tools are not included in OpenVINO™ Runtime and are only available in OpenVINO™ Development Tools. OpenVINO™ Development Tools can be installed via PyPi.

On the other hand, OpenVINO™ 2023.1.0 has introduced a new OpenVINO converter tool (OVC). This tool offers functionality similar to Model Optimizer and is designed to be its lightweight version. You can read more in the Release Notes under the "OpenVINO™ Development Tools" subsection under the "New and Changed in 2023.1" section

OpenVINO™ 2023.1.0 now offers two possible approaches for model conversion: the Python API functions (openvino.convert_model and openvino.save_model) and the OVC CLI method. Check out the Model Preparation page for more information. You can also refer to the Transition from Legacy Conversion API page for the differences between mo and ovc.

For your information, I installed OpenVINO™ 2023.1.0, converted the yolo-v3-tf using ovc, and ran the benchmark app. I share my result below:

 

Regards,

Megat

[Megat_Intel]

Hi Mazr,

This thread will no longer be monitored since we have provided suggestions. If you need any additional information from Intel, please submit a new question.

Regards,

Megat