Yolox_x in super-gradients

[yinmayoo185]

In current version, I cannot try yolox_x. Some well pretrained yolo model for detection used yolox_x and I cannot try with yolox_x in current repo.

[mikel-brostrom]

Should be fairly easy. Notice that the SuperGradients repo is not super user friendly... Pun intended. A lot of things have to be set manually in order to adapt to you custom models. We cannot do much about this in this repo.

Substitute this line of code:

https://github.com/mikel-brostrom/yolo_tracking/blob/057435d17d6837b2e510e27503d925f99a0d1694/examples/detectors/yolonas.py#L39-L42

By:

yolo_nas = models.get(
    'yolox_x',
    checkpoint_path=<path/to/your/yolox-x/weights.pt>
).to(device) ```