Add Transcripts

[ephes]

• [x] Add a transcript model with vtt format for feed and json for the podlove player
• [x] Add the transcript tab to the podlove player
• [x] Include transcripts in the feed
• [ ] Validate transcript formats on upload (uploaded DOTe to podlove -> bad)
• [ ] Have a detail site containing the transcript
• [x] Add transcriptions to all episodes

Should we add a new transcript model or just reference the file in the json field of an audio model? A separate transcript model is probably cleaner but it's more complicated to avoid additional queries.