Update app.py

app.py

@@ -1,5 +1,5 @@
 # app.py
-# Version: 1.07 (08.24.24), ALPHA
+# Version: 1.07a (08.27.24), ALPHA
 #---------------------------------------------------------------------------------------------------------------------------------------------
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
@@ -61,27 +61,14 @@ SIDEBAR_INFO = f"""
 #                                                       ------------transcribe section------------
 
 
-
-# ============UPDATED============[convert m4a audio to wav]
-@spaces.GPU()
-def convert_to_wav(filepath):
-    _,file_ending = os.path.splitext(f'{filepath}')
-    audio_file = filepath.replace(file_ending, ".wav")
-    print("starting conversion to wav")
-    os.system(f'ffmpeg -i "{filepath}" -ar 16000 -ac 1 -c:a pcm_s16le "{audio_file}"')
-    return audio_file
-# ================================[------------------------]
-
-
 pipe = pipeline("automatic-speech-recognition", model="NbAiLab/nb-whisper-large", chunk_length_s=30, generate_kwargs={'task': 'transcribe', 'language': 'no'})
 
 @spaces.GPU()
-def transcribe_audio(audio_file, batch_size=16):
-    audio_file = convert_to_wav(audio_file)
+def transcribe_audio(audio, batch_size=16):
 
     with tempfile.NamedTemporaryFile(suffix=".wav") as temp_audio_file:
         # --copy contents of uploaded audio file to temporary file
-        temp_audio_file.write(open(audio_file, "rb").read())
+        temp_audio_file.write(open(audio, "rb").read())
         temp_audio_file.flush()
         # --use torchaudio to load it
         waveform, sample_rate = torchaudio.load(temp_audio_file.name)