Model.cuda can't work

1

I don’t know how to figure the follow problem.

Traceback (most recent call last):
  File "pos.py", line 177, in <module>
    model = model.cuda()
  File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 147, in cuda
    return self._apply(lambda t: t.cuda(device_id))
  File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/module.py", line 118, in _apply
    module._apply(fn)
  File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/rnn.py", line 116, in _apply
    self.flatten_parameters()
  File "/usr/local/lib/python3.5/dist-packages/torch/nn/modules/rnn.py", line 78, in flatten_parameters
    handle = cudnn.get_handle()
  File "/usr/local/lib/python3.5/dist-packages/torch/backends/cudnn/__init__.py", line 268, in get_handle
    handle = CuDNNHandle()
  File "/usr/local/lib/python3.5/dist-packages/torch/backends/cudnn/__init__.py", line 87, in __init__
    check_error(lib.cudnnCreate(ctypes.byref(ptr)))
  File "/usr/local/lib/python3.5/dist-packages/torch/backends/cudnn/__init__.py", line 255, in check_error
    raise CuDNNError(status)
torch.backends.cudnn.CuDNNError: 4: b'CUDNN_STATUS_INTERNAL_ERROR'
Exception ignored in: <bound method CuDNNHandle.__del__ of <torch.backends.cudnn.CuDNNHandle object at 0x7fae98efdcf8>>
Traceback (most recent call last):
  File "/usr/local/lib/python3.5/dist-packages/torch/backends/cudnn/__init__.py", line 91, in __del__
    check_error(lib.cudnnDestroy(self))
ctypes.ArgumentError: argument 1: <class 'TypeError'>: Don't know how to convert parameter 1
2

Yikes. Could you post some more details (what code you’re running, version of pytorch, your system cuda/cudnn setup?)

3

I am running a simple lstm model.
Here is the main code.https://paste.ubuntu.com/26301189/
The pytorch version is 0.2.0_3.
The cuda version is 8.0.61.

In fact, I was able to run this code successfully last week. So I have no idea where the bug come from.

4

I’m not sure why your model would work last week but not now, if you haven’t changed anything. Does upgrading pytorch to 0.3 help?