Merge branch 'nmt-main' into 'main'

inverse_square_root learning param schedule

See merge request ADLR/megatron-lm!466

megatron/arguments.py

@@ -649,7 +649,7 @@ def _add_learning_rate_args(parser):
                        'and initial warmup, the learing rate at each '
                        'iteration would be different.')
     group.add_argument('--lr-decay-style', type=str, default='linear',
-                       choices=['constant', 'linear', 'cosine'],
+                       choices=['constant', 'linear', 'cosine', 'inverse-square-root'],
                        help='Learning rate decay function.')
     group.add_argument('--lr-decay-iters', type=int, default=None,
                        help='number of iterations to decay learning rate over,'

megatron/optimizer_param_scheduler.py

@@ -90,8 +90,14 @@ class OptimizerParamScheduler(object):
         # For any steps larger than `self.lr_decay_steps`, use `self.min_lr`.
         if self.num_steps > self.lr_decay_steps:
             return self.min_lr
-        
+
         # If we are done with the warmup period, use the decay style.
+        if self.lr_decay_style == 'inverse-square-root':
+            warmup_steps = max(self.lr_warmup_steps, 1)
+            num_steps = max(self.num_steps, 1)
+            lr = self.max_lr * warmup_steps ** 0.5 / (num_steps ** 0.5)
+            return max(self.min_lr, lr)
+
         num_steps_ = self.num_steps - self.lr_warmup_steps
         decay_steps_ = self.lr_decay_steps - self.lr_warmup_steps
         decay_ratio = float(num_steps_) / float(decay_steps_)