[ENHANCEMENT]

[githwd2016]

In the method _get_param_groups from megatron/core/optimizer/__init__.py , when set wd_mult and lr_mult, the code is:

if not no_wd and not scale_lr:
    wd_mult, lr_mult = 1.0, 1.0
elif not no_wd and scale_lr:
    wd_mult, lr_mult = 1.0, lr_mult
elif no_wd and not scale_lr:
    wd_mult, lr_mult = 0.0, 1.0
else:
    wd_mult, lr_mult = 0.0, lr_mult

I think the following code is more concise and clearer. Could I make a PR?

wd_mult = 1.0 if not no_wd else 0.0
lr_mult = 1.0 if not scale_lr else lr_mult