openVINO benchmark_app : how to run precision INT8 ?

[Bhuvaneshwara]

openVINO benchmark_app : this tool gives 2 option to specifiy precision 

--infer_precision Optional. Specifies the inference precision. Example #1: '-infer_precision bf16'. Example #2: '-infer_precision CPU:bf16,GPU:f32'

other one is Preprocessing options:
-op <value> Optional. Specifies precision for all output layers of the model.

which is appropriate to run int8 precision?

[Zulkifli_Intel]

Hi Bhuvaneshwara,

Thank you for reaching out.

You can run benchmark_app without the model precision option. You can try run the benchmark_app using this command:

benchmark_app -m model_name.xml

The Benchmark Tool demonstrates how to use the benchmark_app to estimate deep learning inference performance on supported devices.

Regards,

Zul

[Zulkifli_Intel]

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