0.1.7

Full Changelog: 0.1.6...0.1.7

0.1.6

Full Changelog: 0.1.5...0.1.6

0.1.5

make sure contiguous

0.1.4

router z loss should be calculated on the unnoised gating logits

0.1.2

allow for noising of gates

0.1.1

researcher will want to log the unweighted auxiliary losses

0.1.0

rename loss_coef to balance_loss_coef, sum the balance and router z-l…

…oss and return the total auxiliary loss and add some comments in readme on what to do with it

0.0.30

handle variable sequence lengths if `allow_var_seq_len = True` on `Ex…

…perts`

0.0.29

any combinatino of number of experts and world size should not break

0.0.28

oops