Shouldn't loss for each example be weighed by the sqrt of the length?

[tsjain]

From the code it appears that the loss for an entire batch is scaled by either the crop length or the average of the sequence length for the batch.

Shouldn't each individual sample's loss be scaled by the sqrt of its length?

https://github.com/aqlaboratory/openfold/blob/main/openfold/utils/loss.py#L1776