I have trained a custom yolov7t Pytorch model. I have converted it to “.nb” via the tool. I am running it using sd-card, the example successfully runs, as it loads the model from sd-card and continue to start detecting objects. The issue here is that it does random detections from the default objects list, and doesn’t detect the custom object which the model was trained for. I believe it uses default parameters/classes that are used for coco dataset. How can this be solved so that it just detect what the model is trained for?