Improving transcriptions

[chrome adder]

Hi! I'm wondering what other people do against bad transcriptions. We are streaming the transcription. With whisper we use the prompting feature which to give the model context of our domain/application which helps a lot.

Any similar things we can do with deepgram or any creative ways of improving transcripts?

[wheat basalt]

Thanks for asking your question. Please be sure to reply with as much detail as possible so we can assist you efficiently. Such as:

• Provide the request_id if you've a question about a transcription response.
• The options you used or the api.deepgram.com URL you sent your request to, including parameters.
• Any code snippets you can include.
• Any audio you can include, or if you can't share it here please email it to us at [email protected] and provide a link to this thread.

[grand aspen]

I assume you're using Whisper yourself and not Deepgram's Whisper

Whisper itself doesn't offer streaming transcription. Whereas Nova-2 is our most accurate model, and works fully on our streaming transcription service.

[wheat basalt]

Hey there @chrome adder~! Did you forget about this thread? Let us know if you still need help!

[proper siren]

hello,
please help to stream live audio in different format. streaming radio or live podcast? Thanks

[blazing geyser]

@proper siren please create your own thread and when you do, please include details like what you are trying to do because what you have doesnt provide any information, what audio encoding are you using, maybe even some sample code, etc. just saying "I need help stream live audio in different format" isnt helpful at all

[chrome adder]

Yes we're using Whisper via OpenAI API. We only use it to correct Deepgram streaming transcriptions in case sth is odd.