David Chiang

Doesn't work with PyTorch 2?

2

The `pyproject.toml` (if I understand correctly) requires PyTorch >= 1.1 and < 2.0. Is that necessary?

Allow repeated and unmatched output indices?

2

Although the original einsum doesn't allow it, it would be convenient if: - `einsum('i->ii', torch.tensor([1., 2.]))` returns `torch.tensor([[1., 0.], [0., 2.]])` (that is, it's like `torch.diag_embed`) - `einsum('->i', torch.tensor(1.)` returns...

Broadcasting

3

torch.einsum now supports broadcasting: https://github.com/pytorch/pytorch/issues/48420

Boolean semiring

3

How hard would it be to add the Boolean semiring?

A common error I get is in Equation.validate_sizes(): IndexError: Dimension out of range (expected to be in range of [-1, 0], but got 1) which is not as clear as...

out= option

2

Can an `out=` option be added to overwite an existing tensor instead of creating a new one?

More than 26 indices?

3

torch.einsum limits you to 26 indices but can semiring_einsum provide a non-string interface that allows more than 26?

Derived functions

1

This implements several PyTorch functions, like tensordot and matmul, in terms of einsum.

Fix bug (?) in length penalty computation

12

I think there was an off-by-one error in the length computation. Assuming that the length penalty is computed for real words only (not BOS or EOS), the length of a...

Extensions to beam_decode for random sampling

1

It looks to me like it's working, but @tnq177 if you can take a look to make sure that (1) it looks right and (2) you're okay with adding some...