RuntimeError: ‘lengths‘ argument should be a 1D CPU int64 tensor, but got 1D cuda:0 Long tensor

  File "/home/.../model/decoder.py", line 97, in forward
    packed_x = nn.utils.rnn.pack_padded_sequence(x_seq, lengths=sorted_lengths, batch_first=True)
  File "/home/.../.local/lib/python3.8/site-packages/torch/nn/utils/rnn.py", line 244, in pack_padded_sequence
    _VF._pack_padded_sequence(input, lengths, batch_first)
RuntimeError: 'lengths' argument should be a 1D CPU int64 tensor, but got 1D cuda:0 Long tensor

点击，打开rnn.py

  File "/home/.../.local/lib/python3.8/site-packages/torch/nn/utils/rnn.py"

把,244行左右

    data, batch_sizes = \
        _VF._pack_padded_sequence(input, lengths, batch_first)

改成，加cpu()

    data, batch_sizes = \
        _VF._pack_padded_sequence(input, lengths.cpu(), batch_first)

这是因为pytorch 1.5升级后，bilstm导致