bugs when use pytorch

Traceback (most recent call last):
File "multilabel/runners/resnet_runner.py", line 26, in
main()
File "multilabel/runners/resnet_runner.py", line 22, in main
fire.Fire(resnet50)
File "/usr/local/lib/python3.6/dist-packages/fire/core.py", line 138, in Fire
component_trace = _Fire(component, args, parsed_flag_args, context, name)
File "/usr/local/lib/python3.6/dist-packages/fire/core.py", line 560, in _Fire
treatment='callable')
File "/usr/local/lib/python3.6/dist-packages/fire/core.py", line 675, in _CallAndUpdateTrace
component = fn(*varargs, **kwargs)
File "/usr/local/lib/python3.6/dist-packages/torch/nn/modules/module.py", line 489, in __call__
result = self.forward(*input, **kwargs)
TypeError: forward() missing 1 required positional argument: 'x'

When I rewrite pytorch nn.Module, I use x as params to override forward function. This bugs appear.
I have test. If I do not use Fire, it works well.
I suppose that because fire did not support this kind of override? Or someone can tell me how to fix it.
Thanks.