Limit Audio File Whisper Large v2

I have deployed the whisper largev2 model through JumpStart with SageMaker. It has deployed an endpoint which works correctly. However, in the endpoint's response, it only transcribes a part of the audio, which corresponds to 30 seconds of the audio. When I have tested the model locally, it works with audios of any length. Is it possible to process longer audios in a single call to the endpoint in some way?

Start by reproducing the SageMaker JumpStart Whisper Large v2 endpoint behavior with audio longer than 30 seconds and compare it with the local model run. Check whether the endpoint request, response, or documented inference behavior defines this limit; done means a documented, reproducible explanation or an explicitly scoped change request.

Written by the indexing model from the issue text.