RuntimeError: bool value of Tensor with more than one value is ambiguous

[deepali-c]

Steps to reproduce

Execute steps to test Custom Tensorizer - https://pytext.readthedocs.io/en/master/tensorizer.html

Observed Results

    assert words_tensors == tensor([[2, 3, 4, 5], [6, 7, 8, 1]])
RuntimeError: bool value of Tensor with more than one value is ambiguous

Expected Results

The test should pass.

[deepali-c]

Updating code locally to correctly compare the tensors at the end helps here.
The example in the document needs to be updated likewise:

assert torch.all(torch.eq(words_tensors, tensor([[2, 3, 4, 5], [6, 7, 8, 1]])))
assert torch.all(torch.eq(seq_len_tensors, tensor([4, 3])))

[deepali-c]

Thanks @hudeven :-) .

[deepali-c]

One more thing, there is an additional bracket at the following line -
words_tensors, seq_len_tensors = tensorizer.tensorize(numberized_rows))

[mwu1993]

#971