EtienneAb3d / WhisperHallu

Experimental code: sound file preprocessing to optimize Whisper transcriptions without hallucinated texts
275 stars 22 forks source link

New Google Colab Error #28

Closed mad1m closed 9 months ago

mad1m commented 9 months ago

Google Colab "Standard Whisper"


TypeError Traceback (most recent call last) in <cell line: 21>() 19 20 loadModel("0") ---> 21 result = transcribePrompt(path=path, lng=lng, prompt=prompt)

/content/WhisperHallu/transcribeHallu.py in transcribePrompt(path, lng, prompt, lngInput, isMusic, addSRT, truncDuration, maxDuration) 212 print("PROMPT="+prompt,flush=True) 213 opts = dict(language=lng,initial_prompt=prompt) --> 214 return transcribeOpts(path, opts,lngInput,isMusic=isMusic,addSRT=addSRT,truncDuration=truncDuration,maxDuration=maxDuration) 215 216 def transcribeOpts(path: str,opts: dict

TypeError: transcribeOpts() got an unexpected keyword argument 'truncDuration

EtienneAb3d commented 9 months ago

@mad1m Thanks for the bug report. Fixed. ;-)

mad1m commented 9 months ago

@mad1m Thanks for the bug report. Fixed. ;-)

Thank you, it got fixed. However, audio files longer than 600 seconds aren't supported. That wasn't the case in the old version, i was able to to use the audio files that are longer than 40 mins!. is there a way to bypass the limit? DURATION=838 max 600 result = [Too long (838s)]

EtienneAb3d commented 9 months ago

@mad1m It's just a simple config in the code. You can change it.

mad1m commented 9 months ago

Noted Thanks for replying