Проблемы с запуском на графическом процессоре в MXNet.

Question

Я воспроизвожу код по адресу https://github.com/cartus/dcgcn в Colab. Но возникает ошибка при использовании графического процессора. Между тем, если я использую процессор, эта ошибка не появляется. Есть ли какое-нибудь решение?

Примечание. Я использую mxnet_cu116 и версию cuda: 12.0.

Спасибо

      [ERROR:__main__] Uncaught exception
Traceback (most recent call last):
  File "/usr/local/lib/python3.8/dist-packages/mxnet/symbol/symbol.py", line 1903, in simple_bind
    check_call(_LIB.MXExecutorSimpleBindEx(self.handle,
  File "/usr/local/lib/python3.8/dist-packages/mxnet/base.py", line 246, in check_call
    raise get_last_ffi_error()
mxnet.base.MXNetError: _Map_base: :at

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/usr/lib/python3.8/runpy.py", line 194, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "/usr/lib/python3.8/runpy.py", line 87, in _run_code
    exec(code, run_globals)
  File "/content/drive/.shortcut-targets-by-id/1NKrSJUW7eWh--JL4gjWIO0WHT-FHe2mS/Kuliah/Magister/Semester 4/thesis/code/DCGCN/sockeye/train.py", line 955, in <module>
    main()
  File "/content/drive/.shortcut-targets-by-id/1NKrSJUW7eWh--JL4gjWIO0WHT-FHe2mS/Kuliah/Magister/Semester 4/thesis/code/DCGCN/sockeye/train.py", line 829, in main
    train(args)
  File "/content/drive/.shortcut-targets-by-id/1NKrSJUW7eWh--JL4gjWIO0WHT-FHe2mS/Kuliah/Magister/Semester 4/thesis/code/DCGCN/sockeye/train.py", line 900, in train
    training_model = create_training_model(config=model_config,
  File "/content/drive/.shortcut-targets-by-id/1NKrSJUW7eWh--JL4gjWIO0WHT-FHe2mS/Kuliah/Magister/Semester 4/thesis/code/DCGCN/sockeye/train.py", line 736, in create_training_model
    training_model = training.TrainingModel(config=config,
  File "/content/drive/.shortcut-targets-by-id/1NKrSJUW7eWh--JL4gjWIO0WHT-FHe2mS/Kuliah/Magister/Semester 4/thesis/code/DCGCN/sockeye/training.py", line 78, in __init__
    self._initialize(provide_data, provide_label, default_bucket_key)
  File "/content/drive/.shortcut-targets-by-id/1NKrSJUW7eWh--JL4gjWIO0WHT-FHe2mS/Kuliah/Magister/Semester 4/thesis/code/DCGCN/sockeye/training.py", line 183, in _initialize
    self.module.bind(data_shapes=provide_data,
  File "/usr/local/lib/python3.8/dist-packages/mxnet/module/bucketing_module.py", line 352, in bind
    module.bind(data_shapes, label_shapes, for_training, inputs_need_grad,
  File "/usr/local/lib/python3.8/dist-packages/mxnet/module/module.py", line 422, in bind
    self._exec_group = DataParallelExecutorGroup(self._symbol, self._context,
  File "/usr/local/lib/python3.8/dist-packages/mxnet/module/executor_group.py", line 280, in __init__
    self.bind_exec(data_shapes, label_shapes, shared_group)
  File "/usr/local/lib/python3.8/dist-packages/mxnet/module/executor_group.py", line 383, in bind_exec
    self.execs.append(self._bind_ith_exec(i, data_shapes_i, label_shapes_i,
  File "/usr/local/lib/python3.8/dist-packages/mxnet/module/executor_group.py", line 675, in _bind_ith_exec
    executor = self.symbol.simple_bind(ctx=context, grad_req=self.grad_req,
  File "/usr/local/lib/python3.8/dist-packages/mxnet/symbol/symbol.py", line 1944, in simple_bind
    raise RuntimeError(error_msg)
RuntimeError: simple_bind error. Arguments:
source: (4, 200, 1)
target: (4, 200)
source_graphs: (4, 200, 200)
source_positions: (4, 200)
target_label: (4, 200)
_Map_base: :at

Может работать на графическом процессоре в Google Colab.