Fix audio column name for task template

NPSC.py

@@ -132,8 +132,8 @@ class Npsc(datasets.GeneratorBasedBuilder):
             citation=_CITATION,
             task_templates=[
                 AutomaticSpeechRecognition(
-                    audio_file_path_column="path",
-                    transcription_column="sentence_text"
+                    audio_column="audio",
+                    transcription_column="sentence_text",
                 )
             ],
         )