Spaces:

BSC-LT
/

asr-inference

Running on Zero

ssolito commited on Dec 4, 2024

Commit

5b25bd7

verified ·

1 Parent(s): fe1504e

Update whisper.py

Files changed (1) hide show

whisper.py CHANGED Viewed

@@ -225,6 +225,6 @@ def generate(audio_path, use_v5):
         task = "transcribe"
         output = transcribe_pipeline(format_audio(audio_path), task)
-    clean_output = post_process_transcription(output, max_repeats=1)
     return clean_output

         task = "transcribe"
         output = transcribe_pipeline(format_audio(audio_path), task)
+    clean_output = post_process_transcription(output)
     return clean_output