Unifying training argument type annotations (#17934)

* doc: Unify training arg type annotations

* wip: extracting enum type from Union

* blackening

src/transformers/hf_argparser.py

@@ -92,7 +92,11 @@ def _parse_dataclass_field(parser: ArgumentParser, field: dataclasses.Field):
                     " the argument parser only supports one type per argument."
                     f" Problem encountered in field '{field.name}'."
                 )
-            if bool not in field.type.__args__:
+            if type(None) not in field.type.__args__:
+                # filter `str` in Union
+                field.type = field.type.__args__[0] if field.type.__args__[1] == str else field.type.__args__[1]
+                origin_type = getattr(field.type, "__origin__", field.type)
+            elif bool not in field.type.__args__:
                 # filter `NoneType` in Union (except for `Union[bool, NoneType]`)
                 field.type = (
                     field.type.__args__[0] if isinstance(None, field.type.__args__[1]) else field.type.__args__[1]

src/transformers/training_args.py

@@ -20,7 +20,7 @@
 from dataclasses import asdict, dataclass, field
 from enum import Enum
 from pathlib import Path
-from typing import Any, Dict, List, Optional
+from typing import Any, Dict, List, Optional, Union
 
 from .debug_utils import DebugOption
 from .trainer_utils import (
@@ -493,7 +493,7 @@ class TrainingArguments:
     do_train: bool = field(default=False, metadata={"help": "Whether to run training."})
     do_eval: bool = field(default=False, metadata={"help": "Whether to run eval on the dev set."})
     do_predict: bool = field(default=False, metadata={"help": "Whether to run predictions on the test set."})
-    evaluation_strategy: IntervalStrategy = field(
+    evaluation_strategy: Union[IntervalStrategy, str] = field(
         default="no",
         metadata={"help": "The evaluation strategy to use."},
     )
@@ -559,7 +559,7 @@ class TrainingArguments:
         default=-1,
         metadata={"help": "If > 0: set total number of training steps to perform. Override num_train_epochs."},
     )
-    lr_scheduler_type: SchedulerType = field(
+    lr_scheduler_type: Union[SchedulerType, str] = field(
         default="linear",
         metadata={"help": "The scheduler type to use."},
     )
@@ -596,14 +596,14 @@ class TrainingArguments:
         },
     )
     logging_dir: Optional[str] = field(default=None, metadata={"help": "Tensorboard log dir."})
-    logging_strategy: IntervalStrategy = field(
+    logging_strategy: Union[IntervalStrategy, str] = field(
         default="steps",
         metadata={"help": "The logging strategy to use."},
     )
     logging_first_step: bool = field(default=False, metadata={"help": "Log the first global_step"})
     logging_steps: int = field(default=500, metadata={"help": "Log every X updates steps."})
     logging_nan_inf_filter: bool = field(default=True, metadata={"help": "Filter nan and inf losses for logging."})
-    save_strategy: IntervalStrategy = field(
+    save_strategy: Union[IntervalStrategy, str] = field(
         default="steps",
         metadata={"help": "The checkpoint save strategy to use."},
     )
@@ -815,7 +815,7 @@ class TrainingArguments:
     label_smoothing_factor: float = field(
         default=0.0, metadata={"help": "The label smoothing epsilon to apply (zero means no label smoothing)."}
     )
-    optim: OptimizerNames = field(
+    optim: Union[OptimizerNames, str] = field(
         default="adamw_hf",
         metadata={"help": "The optimizer to use."},
     )
@@ -868,7 +868,7 @@ class TrainingArguments:
     hub_model_id: Optional[str] = field(
         default=None, metadata={"help": "The name of the repository to keep in sync with the local `output_dir`."}
     )
-    hub_strategy: HubStrategy = field(
+    hub_strategy: Union[HubStrategy, str] = field(
         default="every_save",
         metadata={"help": "The hub strategy to use when `--push_to_hub` is activated."},
     )