Good morning, I need a reference for the formula of the relative error L2 used for the validation on the desktop and validation on target of the DL model with the generated C model on the host. Thank you very much. Best regards Luigi Bibbo´

[LBibb.1]

..

[Amel NASRI]

Hi @Luigi Bibbo´​ ,

I'm not sure to understand the context of your question. Is it really related to an STM32 product? If yes, which one?

-Amek

[LBibb.1]

It's STM32L475VG

[LBibb.1]

This is link of the board B-L475E-IOT01A - STM32L4 Discovery kit IoT node, low-power wireless, BLE, NFC, SubGHz, Wi-Fi - STMicroelectronics

DISCO-L475VG-IOT01A (B-L475E-IOT01A) | Mbed

[Amel NASRI]

OK, thanks for precising used board.

Now coming back to your question: "I need a reference for the formula of the relative error L2 used for the validation on the desktop and validation on target of the DL model with the generated C model on the host."

It remains not clear for me about which C model are you speaking? Is there any particular document you are referring to?

-Amel

[LBibb.1]

This is the document I am referring to: UM2526 User Manual – Getting started with X-CUBE-AI Expansion Package for Artificial Intelligence (AI), January 2022.