bbaguette

eval.py

@@ -102,7 +102,7 @@ def main(args):
     # map function to decode audio
     def map_to_pred(batch):
         prediction = asr(
-            batch["audio"]["array"], chunk_length_s=args.chunk_length_s, stride_length_s=args.stride_length_s,,skip_special_tokens=True
+            batch["audio"]["array"], chunk_length_s=args.chunk_length_s, stride_length_s=args.stride_length_s
         )
 
         batch["prediction"] = prediction["text"]# "".join(prediction["text"].split("<s>"))

run_speech_recognition_ctc.py

@@ -643,7 +643,7 @@ def main():
 
         pred.label_ids[pred.label_ids == -100] = tokenizer.pad_token_id
 
-        pred_str = tokenizer.batch_decode(pred_ids, skip_special_tokens=True)
+        pred_str = tokenizer.batch_decode(pred_ids)
         # we do not want to group tokens when computing the metrics
         label_str = tokenizer.batch_decode(pred.label_ids, group_tokens=False)