A drawback in using SileroVAD for audio/video subtitle generation

[tentse]

I was building a application where i want to generate subtitle for audio currently. And since the maximum audio subtitle duration should be between 5-8 seconds since longer the audio longer the subtitle text will be. But for some audio SileroVAD is not able to detect gap in between the audio creating some exception case such as for some audio subtitle won't appropriately generated. Is there any solution for that where we can specify if x amount of gap is there then audio should be split. THANK YOU.

[tentse]

Example for this video (sorry since this is in my native language) Video Silero is generating a audio time stamp of 40 second like (1.1second to 41.2second) which when applying to speech to text, the subtitle won't look appropriate. This issue might have been already raised, if already solution is there kindly let me know and if it's not raised yet please give a solution for this issue.