Deploying Open AI’s whisper on Sagemaker for audio streaming

Hello Experts,

I have already been able to deploy OpenAI’s whisper on sagemaker EP for audio files. So here the EP is reading the files from bucket and then pushing back the transcription. Now, I need to deploy the Whisper for streaming audio and preferably through websocket. Is there any way I can achieve this?

Thanks
Amit

Hello @amitkayal,

SageMaker is currently only supporting POST requests to an endpoint. There is no support for streaming or WebSockets.
We tried to push for this in the past as well, but no success yet.

Can you give some details about the open ai deployment? I have trouble woth audios of more than 30 sec