add a private function to find adam opt pass

python/paddle/fluid/distribute_transpiler.py

@@ -401,10 +401,7 @@ class DistributeTranspiler:
         # HACK: optimization global ops only used to scale beta1 and beta2
         # replace it with dependency engine.
         for op in self.optimize_ops:
-            if op.type == "scale":
-                for in_name in op.input_arg_names:
-                    if in_name.startswith("beta1_pow_acc") or \
-                            in_name.startswith("beta2_pow_acc"):
+            if self._is_adam_connected_op(op):
                 global_ops.append(op)
 
         def __append_optimize_op__(op, block, grad_to_block_id):
@@ -1152,13 +1149,20 @@ class DistributeTranspiler:
                     op.input("Param")[0]),
                                      self.origin_program.global_block().var(
                                          op.input("Grad")[0])))
-            elif op.type == "scale":
-                # for adam optimize op
-                for in_name in op.input_arg_names:
-                    if in_name.startswith("beta1_pow_acc") or \
-                            in_name.startswith("beta2_pow_acc"):
+            elif self._is_adam_connected_op(op):
                 opt_ops.append(op)
-                        break
             else:
                 pass
         return opt_ops, params_grads
+
+    def _is_adam_connected_op(self, op):
+        """
+        A hack function to determinate whether the input operator
+        is connected to optimize operator.
+        """
+        if op.type == "scale":
+            for in_name in op.input_arg_names:
+                if in_name.startswith("beta1_pow_acc") or \
+                        in_name.startswith("beta2_pow_acc"):
+                    return True
+        return False