Simple Whisper Base Template Based on model from https://huggingface.co/openai/whisper-base, made to run on Banana Serverless GPUs This template supports mp3 bytes in, returns the transcript, nothing crazy. See test.py for an example of how to convert mp3 to bytes