Error while running tutorial custom_detector pytorch_yolov5 on zedbox

Hi there,
I want to run example Custom Detector pytorch_yolov5 python3 on my ZED BOX - Xavier NX. I followed the instructions in the readme and encountered the following error:

freeglut (  ERROR:  Internal error <FBConfig with necessary capabilities not found> in function fgOpenWindow
Exception in thread Thread-1:
Traceback (most recent call last):
  File "/usr/lib/python3.6/", line 916, in _bootstrap_inner
  File "/usr/lib/python3.6/", line 864, in run
    self._target(*self._args, **self._kwargs)
  File "", line 104, in torch_thread
    model = attempt_load(weights, map_location=device)  # load FP32
  File "/home/user/catkin_ws/src/yolov5_pytorch_ros/src/models/", line 96, in attempt_load
    ckpt = torch.load(attempt_download(w), map_location=map_location)  # load
  File "/home/user/.local/lib/python3.6/site-packages/torch/", line 592, in load
    return _load(opened_zipfile, map_location, pickle_module, **pickle_load_args)
  File "/home/user/.local/lib/python3.6/site-packages/torch/", line 851, in _load
    result = unpickler.load()
  File "/home/user/.local/lib/python3.6/site-packages/torch/", line 843, in persistent_load
    load_tensor(data_type, size, key, _maybe_decode_ascii(location))
  File "/home/user/.local/lib/python3.6/site-packages/torch/", line 832, in load_tensor
    loaded_storages[key] = restore_location(storage, location)
  File "/home/user/.local/lib/python3.6/site-packages/torch/", line 812, in restore_location
    return default_restore_location(storage, str(map_location))
  File "/home/user/.local/lib/python3.6/site-packages/torch/", line 175, in default_restore_location
    result = fn(storage, location)
  File "/home/user/.local/lib/python3.6/site-packages/torch/", line 157, in _cuda_deserialize
    return obj.cuda(device)
  File "/home/user/.local/lib/python3.6/site-packages/torch/", line 80, in _cuda
    return new_type(self.size()).copy_(self, non_blocking)
  File "/home/user/.local/lib/python3.6/site-packages/torch/cuda/", line 484, in _lazy_new
    return super(_CudaBase, cls).__new__(cls, *args, **kwargs)
RuntimeError: CUDA error: driver shutting down

pytorch version 1.8.0
cuda version 10.2

did you solve it, please?

Yes, sorry.
If I remember correctly I had some unfitting weights.