RuntimeError: cur_offset == offset ASSERT FAILED

Hi all,

I try to utilize LSTM to classify the time series data. The program report this error.
When I try to load another dataset. This program runs well.
I don’t know what’s gonging wrong.

Environment

  • PyTorch Version: 1.0.0
  • OS (e.g., Linux): Windows 10
  • How you installed PyTorch: conda
  • Python version: 3.6.5
  • CUDA/cuDNN version: 8.0
  • GPU models and configuration: Nvidia GeForce GTX 1060

Here is the message:
Traceback (most recent call last):
File “test_version1.py”, line 236, in
loss.backward()
File “D:\ProgramFiles\Anaconda\lib\site-packages\torch\tensor.py”, line 102, in backward
torch.autograd.backward(self, gradient, retain_graph, create_graph)
File “D:\ProgramFiles\Anaconda\lib\site-packages\torch\autograd_init_.py”, line 90, in backward
allow_unreachable=True) # allow_unreachable flag
RuntimeError: cur_offset == offset ASSERT FAILED at …\aten\src\ATen\native\cudnn\RNN.cpp:471, please report a bug to PyTorch. cur_offset = 1860; offset = 930

            input = input.cuda(async = True)
            target = target.cuda(async = True)
            # Forward
            output = model(input)
            loss = F.cross_entropy(output, target)
            losses["train"] += loss.item()
            # Compute accuracy
            _,pred = output.data.max(1)
            correct = pred.eq(target.data).sum().item()
            accuracy = correct/input.data.size(0)
            accuracies[split] += accuracy
            # Backward and optimize
            optimizer.zero_grad()
            loss.backward()
            optimizer.step()

Thanks!