RKNN-tools 1.4 转 pytorch模型 报错

以下是报错信息

--> config model
done
--> Loading model
/home/maple/anaconda3/lib/python3.6/site-packages/onnx_tf/common/__init__.py:87: UserWarning: FrontendHandler.get_outputs_names is deprecated. It will be removed in future release.. Use node.outputs instead.
  warnings.warn(message)
./resnet_car.pt ********************
WARNING: Token 'COMMENT' defined, but not used
WARNING: There is 1 unused token
Syntax error in input! LexToken(NAMED_IDENTIFIER,'str',7,340)
!!!!! Illegal character '"'
!!!!! Illegal character '-'
!!!!! Illegal character '"'
!!!!! Illegal character '"'
E Catch exception when loading pytorch model: ./resnet_car.pt!
E Traceback (most recent call last):
E   File "rknn/api/rknn_base.py", line 611, in rknn.api.rknn_base.RKNNBase.load_pytorch
E   File "rknn/base/RKNNlib/app/importer/import_pytorch.py", line 97, in rknn.base.RKNNlib.app.importer.import_pytorch.ImportPytorch.run
E   File "rknn/base/RKNNlib/converter/convert_pytorch.py", line 570, in rknn.base.RKNNlib.converter.convert_pytorch.convert_pytorch.__init__
E   File "rknn/base/RKNNlib/converter/convert_pytorch.py", line 641, in rknn.base.RKNNlib.converter.convert_pytorch.convert_pytorch.model_simplify
E   File "/home/maple/anaconda3/lib/python3.6/site-packages/torch/tensor.py", line 412, in __array__
E     return self.numpy()
E TypeError: can't convert CUDA tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.
Load pytorch model failed!
模型转换跟cuda有什么关系？

张量输入不能Tensor.to(device),创建ｔｅｎｓｏｒ后直接torch.jit.trace