v0.3.0

• User prompting
• Stream text output
• d-v1a native support
• Lots of refactoring

v0.2.0

• Detach the fork from the original repository
• Implement speaker diarization on top of new speaker segmentation
  • Change diarization_prefix parameter to speaker_delimiter
  • Warn users that models sizes lower than medium do not support it
  • Convert delimiter strings to functions that return delimiter strings
• Output raw transcript as opposed to ICU-tokenized words

v0.1.0

Initial release