Fix dtype getter (#17668)

* Fix dtype getters

* Proper fix for dtype getter

* Style and commant

* Always use last for consistency

* Quality

src/transformers/modeling_utils.py

@@ -139,7 +139,7 @@ def get_first_parameter_dtype(parameter: Union[nn.Module, GenerationMixin, "Modu
     try:
         return next(parameter.parameters()).dtype
     except StopIteration:
-        # For nn.DataParallel compatibility in PyTorch 1.5
+        # For nn.DataParallel compatibility in PyTorch > 1.5
 
         def find_tensor_attributes(module: nn.Module) -> List[Tuple[str, Tensor]]:
             tuples = [(k, v) for k, v in module.__dict__.items() if torch.is_tensor(v)]
@@ -152,31 +152,33 @@ def find_tensor_attributes(module: nn.Module) -> List[Tuple[str, Tensor]]:
 
 def get_parameter_dtype(parameter: Union[nn.Module, GenerationMixin, "ModuleUtilsMixin"]):
     """
-    Returns the first found floating dtype in parameters if there is one, otherwise returns the first dtype it found.
+    Returns the first found floating dtype in parameters if there is one, otherwise returns the last dtype it found.
     """
-    try:
-        for t in parameter.parameters():
-            if t.is_floating_point():
-                return t.dtype
-        # if no floating dtype was found return whatever the first dtype is
-        else:
-            return next(parameter.parameters()).dtype
+    last_dtype = None
+    for t in parameter.parameters():
+        last_dtype = t.dtype
+        if t.is_floating_point():
+            return t.dtype
 
-    except StopIteration:
-        # For nn.DataParallel compatibility in PyTorch 1.5
+    if last_dtype is not None:
+        # if no floating dtype was found return whatever the first dtype is
+        return last_dtype
 
+    else:
+        # For nn.DataParallel compatibility in PyTorch > 1.5
         def find_tensor_attributes(module: nn.Module) -> List[Tuple[str, Tensor]]:
             tuples = [(k, v) for k, v in module.__dict__.items() if torch.is_tensor(v)]
             return tuples
 
         gen = parameter._named_members(get_members_fn=find_tensor_attributes)
+        last_tuple = None
         for tuple in gen:
+            last_tuple = tuple
             if tuple[1].is_floating_point():
                 return tuple[1].dtype
-        # fallback to any dtype the model has even if not floating
-        else:
-            first_tuple = next(gen)
-            return first_tuple[1].dtype
+
+        # fallback to the last dtype
+        return last_tuple[1].dtype
 
 
 def get_state_dict_float_dtype(state_dict):