docstring in DecoupledAdaLRLion is not cohere with the code

[QinLuo]

The docstring of DecoupledAdaLRLion say the LR is scaled down by min(`lr_penalty` ** N, `min_scale`), but the code implemented adjust_lr here is lr * max(min_scale, lr_penalty**num_times).

So which is the right one, max or min?

[Jeremy D]

you're correct. the doctoring is wrong, thanks for catching the typo! fixing with #563