feat(api) Implementtranscriptions Endpoint (or validate that it still works)

[barronstone]

User Story

As an AI application developer
I want to upload audio/video files via the transcriptions API endpoint
So that I can get a speech-to-text transcription of the recording

Acceptance Criteria

Given an audio or video file containing speech in a Whisper supported language
When that file is uploaded to the transcriptions endpoint
Then it returns a text transcription in the input language

Additional context

The transcriptions endpoint exists today; however, it has not been touched in several months and likely needs some refactoring. (It looks like there is some unused/dead code.) Additionally, we need to validate that it is compliant with the OpenAI API specification.

An issue exists to increase testing for the transcriptions endpoint: #577

References:

• Transcriptions Quickstart Guide
• Transcriptions API Reference