fix(loss): use autograd-aware all_reduce in BarlowTwinsLoss (#1980)

* fix(loss): use autograd-aware all_reduce in BarlowTwinsLoss

The raw torch.distributed.all_reduce overwrote the cross-correlation
matrix in place and was not registered in the autograd graph. Under DDP
gradient averaging this scaled the backbone gradient down by
1/world_size relative to single-GPU Barlow Twins on the global batch.

Swap it for the autograd-aware torch.distributed.nn.all_reduce, whose
backward is itself an all_reduce, so the gradient is scaled back up by
world_size. The forward value is unchanged.

Fixes #1977. Same fix as SIGReg in #1923.

* test(loss): remove mock-based gather_distributed tests

Mocking all_reduce only tested the mock, not the real collective, and
would break on any change to the reduction. Real multi-rank coverage
is tracked in #1982.

---------

Co-authored-by: Saud Kamran <saud.kamran92@gmail.com>