Fix 'eval_split_name' described as defaulting to 'train' (#15348)

The default is correct (`test`) but the description is not.

examples/pytorch/speech-recognition/run_speech_recognition_ctc.py

@@ -152,7 +152,7 @@ class DataTrainingArguments:
     eval_split_name: str = field(
         default="test",
         metadata={
-            "help": "The name of the training data set split to use (via the datasets library). Defaults to 'train'"
+            "help": "The name of the training data set split to use (via the datasets library). Defaults to 'test'"
         },
     )
     audio_column_name: str = field(