[Audio] createTranscription Added `timestamp_granularities[]` request parameter and updated response object

[mihirshahxenlabs]

Fixes #156

As per the docs: https://platform.openai.com/docs/api-reference/audio/createTranscription Added timestamp_granularities[] in request parameter

As per the docs: https://platform.openai.com/docs/api-reference/audio/verbose-json-object Added task, language, duration, words, segment in response object. This is used when responseType is verbose_json

---

Note: Adding both values words and segments to the parameter timestamp_granularities, doesn't work.

[anasfik]

Very cool, thank you!

[anasfik]

I did make use of it in the example/create_audio_transcription.dart.