Make iter persistent for AdagradW (pytorch#4147)

Summary:
Pull Request resolved: pytorch#4147

X-link: facebookresearch/FBGEMM#1228

Make iter persistent for AdagradW optimizer state saving.
This is to avoid potential loss of the iter information when training is restarted.

Reviewed By: q10

Differential Revision: D74717848

Author: minhua-chen

fbgemm_gpu/fbgemm_gpu/split_table_batched_embeddings_ops_training.py

@@ -1122,6 +1122,7 @@ def __init__(  # noqa C901
         self._max_counter_update_freq: int = -1
         # Extract parameters from CounterBasedRegularizationDefinition or CowClipDefinition
         # which are passed as entries for OptimizerArgs
+        self._used_rowwise_adagrad_with_adagradw: bool = False
         if self._used_rowwise_adagrad_with_counter:
             if self.weight_decay_mode == WeightDecayMode.COUNTER:
                 self._max_counter_update_freq = (
@@ -1131,6 +1132,9 @@ def __init__(  # noqa C901
                     counter_based_regularization.counter_weight_decay_mode
                 )
                 counter_halflife = counter_based_regularization.counter_halflife
+                self._used_rowwise_adagrad_with_adagradw = (
+                    opt_arg_weight_decay_mode == CounterWeightDecayMode.ADAGRADW
+                )
             else:
                 opt_arg_weight_decay_mode = (
                     cowclip_regularization.counter_weight_decay_mode
@@ -1359,6 +1363,7 @@ def __init__(  # noqa C901
                     OptimType.EMAINPLACE_ROWWISE_ADAGRAD,
                 )
                 or self._used_rowwise_adagrad_with_global_weight_decay
+                or self._used_rowwise_adagrad_with_adagradw
             ):
                 self.register_buffer(
                     "iter",
@@ -2766,10 +2771,7 @@ def get_optimizer_state(self) -> List[Dict[str, torch.Tensor]]:
                             "row_counter": states[2],
                             "iter": self.iter,
                         }
-                        if self.optimizer_args.regularization_mode
-                        == WeightDecayMode.COUNTER.value
-                        and self.optimizer_args.weight_decay_mode
-                        == CounterWeightDecayMode.ADAGRADW.value
+                        if self._used_rowwise_adagrad_with_adagradw
                         else {
                             "sum": states[0],
                             "prev_iter": states[1],