I am using Open AI whispher model for convert speech to text :
API : https://api.openai.com/v1/audio/translations
model : whisper-1
temprature : 0
->> In my audio file silence is present in last and API gives this response.
example : We have dynamic element and then we can handle by using the implicit weight and explicit weight. We can handle that. Okay, sir. We can use a contents method and start with methods. Yes, sir. Yes, sir. Yes, sir. Okay. Yes, sir. Yes, sir.
→ So, anyone have any solution to avoid this random response and what to set in parameters to get better and accurate result.