allennlp test run

test.out

----------------------------- Captured stderr call -----------------------------
2017-11-09 22:23:15,113 - INFO - allennlp.common.checks - Pytorch version: 0.3.0b0+1f694e9
_________________________ EntropyTest.test_masked_case _________________________

self = <tests.training.metrics.entropy_test.EntropyTest testMethod=test_masked_case>

    def test_masked_case(self):
        metric = Entropy()
        # This would have non-zero entropy without the mask.
        logits = torch.Tensor([[1, 1, 1, 1],
                               [10000, -10000, -10000, -1000]])
        mask = torch.Tensor([0, 1])
>       metric(logits, mask)

tests/training/metrics/entropy_test.py:39:
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
allennlp/training/metrics/entropy.py:34: in __call__
    log_probs = torch.nn.functional.log_softmax(logits).data
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

input =
     1      1      1      1
 10000 -10000 -10000  -1000
[torch.FloatTensor of size 2x4]

dim = 1, _stacklevel = 3

    def log_softmax(input, dim=None, _stacklevel=3):
        r"""Applies a softmax followed by a logarithm.

        While mathematically equivalent to log(softmax(x)), doing these two
        operations separately is slower, and numerically unstable. This function
        uses an alternative formulation to compute the output and gradient correctly.

        See :class:`~torch.nn.LogSoftmax` for more details.

        Arguments:
            input (Variable): input
            dim (int): A dimension along which log_softmax will be computed.
        """
        if dim is None:
            dim = _get_softmax_dim('log_softmax', input.dim(), _stacklevel)
>       return torch._C._nn.log_softmax(input, dim)
E       RuntimeError: log_softmax(): argument 'input' (position 1) must be Variable, not torch.Fl
oatTensor

/opt/conda/envs/pytorch-py3.6/lib/python3.6/site-packages/torch/nn/functional.py:786: RuntimeErro
r