Add word timestamps

[evamaxfield]

Follow the example: https://cloud.google.com/speech-to-text/docs/async-time-offsets

However, I think this means that the best solution for transcript storage ready for web will be a json list of dict like the following:

[
   {
      "0:00:00": "Hello and good morning everybody.",
      "0:00:05": "It is 3:30 pm on April 25, and this is the committee on testing.",
      "...": "..."
   }
]

Which means I need to also write a really simply function: create_raw_transcript for use when people want the raw format.