I decided to start with something simple - to run chess figures detection model in yolov5s (with TIM-VX). But I got the error:
Idk what’s wrong. I’ve quantized the weights, I did everything, here’s my model output in netron:
(I have 13 classes, that’s why 18=13+5 as in usual yolo)
And I cann’t see any differences between your model and mine. This is your model:
Do you have any suggestions what to do?