Fix/update audio to text dataset

nemo/collections/asr/data/audio_to_text_dataset.py

@@ -343,7 +343,8 @@ def get_tarred_dataset(
         logging.warning(f"dataset does not have explicitly defined labels")
 
     if 'max_utts' in config:
-        raise ValueError('"max_utts" parameter is not supported for tarred datasets')
+        logging.warning('"max_utts" parameter is not supported for tarred datasets')
+        # raise ValueError('"max_utts" parameter is not supported for tarred datasets')
 
     for dataset_idx, (tarred_audio_filepath, manifest_filepath) in enumerate(
         zip(tarred_audio_filepaths, manifest_filepaths)