Suggestion : Pytorch compatibility get_last_lr

[L-Reichardt]

Hello,

first of all thank you very much for your implementation. I noticed a small "incompatibility" when using the scheduler. Since some update, Pytorch recommends using .get_last_lr() over .get_lr(). In their code they've added a warning.

def get_lr(self):
    if not self._get_lr_called_within_step:
        warnings.warn("To get the last learning rate computed by the scheduler, "
                      "please use `get_last_lr()`.", UserWarning)

When plugging your scheduler into existing (newer) code bases, this raises an error as some use get_last_lr(). A quick fix would be adding the following to your code:

    def get_last_lr(self):
        return self.get_lr()

Best Regards,
L