Convert _VALID_DICT_FIELDS to class attribute for shared dict parsing in subclasses

src/transformers/training_args.py

@@ -188,19 +188,6 @@ class OptimizerNames(ExplicitEnum):
     APOLLO_ADAMW_LAYERWISE = "apollo_adamw_layerwise"
 
 
-# Sometimes users will pass in a `str` repr of a dict in the CLI
-# We need to track what fields those can be. Each time a new arg
-# has a dict type, it must be added to this list.
-# Important: These should be typed with Optional[Union[dict,str,...]]
-_VALID_DICT_FIELDS = [
-    "accelerator_config",
-    "fsdp_config",
-    "deepspeed",
-    "gradient_checkpointing_kwargs",
-    "lr_scheduler_kwargs",
-]
-
-
 def _convert_str_dict(passed_value: dict):
     "Safely checks that a passed value is a dictionary and converts any string values to their appropriate types."
     for key, value in passed_value.items():
@@ -814,6 +801,18 @@ class TrainingArguments:
             https://github.com/huggingface/transformers/issues/34242
     """
 
+    # Sometimes users will pass in a `str` repr of a dict in the CLI
+    # We need to track what fields those can be. Each time a new arg
+    # has a dict type, it must be added to this list.
+    # Important: These should be typed with Optional[Union[dict,str,...]]
+    _VALID_DICT_FIELDS = [
+        "accelerator_config",
+        "fsdp_config",
+        "deepspeed",
+        "gradient_checkpointing_kwargs",
+        "lr_scheduler_kwargs",
+    ]
+
     framework = "pt"
     output_dir: Optional[str] = field(
         default=None,
@@ -1561,7 +1560,7 @@ def __post_init__(self):
             )
 
         # Parse in args that could be `dict` sent in from the CLI as a string
-        for field in _VALID_DICT_FIELDS:
+        for field in self._VALID_DICT_FIELDS:
             passed_value = getattr(self, field)
             # We only want to do this if the str starts with a bracket to indicate a `dict`
             # else its likely a filename if supported

tests/utils/test_hf_argparser.py

@@ -29,7 +29,6 @@
 from transformers import HfArgumentParser, TrainingArguments
 from transformers.hf_argparser import make_choice_type_function, string_to_bool
 from transformers.testing_utils import require_torch
-from transformers.training_args import _VALID_DICT_FIELDS
 
 
 # Since Python 3.10, we can use the builtin `|` operator for Union types
@@ -412,7 +411,8 @@ def test_parse_yaml(self):
         args = BasicExample(**args_dict_for_yaml)
         self.assertEqual(parsed_args, args)
 
-    def test_integration_training_args(self):
+    def test_z_integration_training_args(self):
+        # make sure that this test executes last in the test suite
         parser = HfArgumentParser(TrainingArguments)
         self.assertIsNotNone(parser)
 
@@ -424,7 +424,7 @@ def test_valid_dict_annotation(self):
         If this fails, a type annotation change is
         needed on a new input
         """
-        base_list = _VALID_DICT_FIELDS.copy()
+        base_list = TrainingArguments._VALID_DICT_FIELDS.copy()
         args = TrainingArguments
 
         # First find any annotations that contain `dict`
@@ -468,7 +468,7 @@ def test_valid_dict_annotation(self):
             self.assertIn(
                 field.name,
                 base_list,
-                f"Optional dict field `{field.name}` is not in the base list of valid fields. Please add it to `training_args._VALID_DICT_FIELDS`",
+                f"Optional dict field `{field.name}` is not in the base list of valid fields. Please add it to `TrainingArguments._VALID_DICT_FIELDS`",
             )
 
     @require_torch