[traner] fix --lr_scheduler_type choices (huggingface#9800)

* fix --lr_scheduler_type choices * rewrite to fix for all enum-based cl args * cleanup * adjust test * style * Proposal that should work * Remove needless code * Fix test Co-authored-by: Sylvain Gugger <[email protected]>
Qbiwan · Jan 31, 2021 · b342ca2 · b342ca2
1 parent 2035586
commit b342ca2
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 8 deletions.
diff --git a/src/transformers/hf_argparser.py b/src/transformers/hf_argparser.py
@@ -94,8 +94,8 @@ def _add_dataclass_arguments(self, dtype: DataClassType):
                     field.type = prim_type
 
             if isinstance(field.type, type) and issubclass(field.type, Enum):
-                kwargs["choices"] = list(field.type)
-                kwargs["type"] = field.type
+                kwargs["choices"] = [x.value for x in field.type]
+                kwargs["type"] = type(kwargs["choices"][0])
                 if field.default is not dataclasses.MISSING:
                     kwargs["default"] = field.default
             elif field.type is bool or field.type is Optional[bool]:
@@ -198,7 +198,7 @@ def parse_json_file(self, json_file: str) -> Tuple[DataClass, ...]:
         data = json.loads(Path(json_file).read_text())
         outputs = []
         for dtype in self.dataclass_types:
-            keys = {f.name for f in dataclasses.fields(dtype)}
+            keys = {f.name for f in dataclasses.fields(dtype) if f.init}
             inputs = {k: v for k, v in data.items() if k in keys}
             obj = dtype(**inputs)
             outputs.append(obj)
@@ -211,7 +211,7 @@ def parse_dict(self, args: dict) -> Tuple[DataClass, ...]:
         """
         outputs = []
         for dtype in self.dataclass_types:
-            keys = {f.name for f in dataclasses.fields(dtype)}
+            keys = {f.name for f in dataclasses.fields(dtype) if f.init}
             inputs = {k: v for k, v in args.items() if k in keys}
             obj = dtype(**inputs)
             outputs.append(obj)

diff --git a/tests/test_hf_argparser.py b/tests/test_hf_argparser.py
@@ -55,7 +55,10 @@ class BasicEnum(Enum):
 
 @dataclass
 class EnumExample:
-    foo: BasicEnum = BasicEnum.toto
+    foo: BasicEnum = "toto"
+
+    def __post_init__(self):
+        self.foo = BasicEnum(self.foo)
 
 
 @dataclass
@@ -133,14 +136,18 @@ def test_with_enum(self):
         parser = HfArgumentParser(EnumExample)
 
         expected = argparse.ArgumentParser()
-        expected.add_argument("--foo", default=BasicEnum.toto, choices=list(BasicEnum), type=BasicEnum)
+        expected.add_argument("--foo", default="toto", choices=["titi", "toto"], type=str)
         self.argparsersEqual(parser, expected)
 
         args = parser.parse_args([])
-        self.assertEqual(args.foo, BasicEnum.toto)
+        self.assertEqual(args.foo, "toto")
+        enum_ex = parser.parse_args_into_dataclasses([])[0]
+        self.assertEqual(enum_ex.foo, BasicEnum.toto)
 
         args = parser.parse_args(["--foo", "titi"])
-        self.assertEqual(args.foo, BasicEnum.titi)
+        self.assertEqual(args.foo, "titi")
+        enum_ex = parser.parse_args_into_dataclasses(["--foo", "titi"])[0]
+        self.assertEqual(enum_ex.foo, BasicEnum.titi)
 
     def test_with_list(self):
         parser = HfArgumentParser(ListExample)