Tensorflow doc changes on loss output size

[janjitse]

What does this PR do?

Fixes #9771, by changing the documentation to correctly state the size of the output loss function.

I did not change the documentation for TFSeq2SeqQuestionAnsweringModelOutput and TFSeq2SeqSequenceClassifierOutput, as I could not find any code using this, so I was unsure what the correct output size would be.

I also fixed a few instances where I found the documentation referring to torch.LongTensor when tf.tensor should be used.

Before submitting

• This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
• Did you read the contributor guideline,
  Pull Request section?
• Was this discussed/approved via a Github issue or the forum? Please add a link
  to it if that's the case.
• Did you make sure to update the documentation with your changes? Here are the
  documentation guidelines, and
  here are tips on formatting docstrings.
• Did you write any new necessary tests?

Who can review?

@jplu

[jplu]

Thanks a lot for taking care of this!!