test

run_whisper_finetuning.py

@@ -415,7 +415,6 @@ def main():
     # Make sure everything is in 16K
     train_dataset = train_dataset.cast_column(data_args.audio_column_name, Audio(sampling_rate=16000))
     eval_dataset = eval_dataset.cast_column(data_args.audio_column_name, Audio(sampling_rate=16000))
-    
    
     # Rename columns
     if data_args.audio_column_name != "audio":
@@ -430,7 +429,7 @@ def main():
             data_args.text_column_name, "sentence")
         eval_dataset = eval_dataset.rename_column(
             data_args.text_column_name, "sentence")
-
+        column_names.remove(data_args.text_column_name)
         
     # Prepare the dataset
     train_dataset = train_dataset.map(prepare_dataset, remove_columns=column_names)