WorkflowType#

class montreal_forced_aligner.data.WorkflowType(value, names=None, *, module=None, qualname=None, type=None, start=1, boundary=None)[source]#

Bases: Enum

Enum for workflows involving corpora

Parameters:
  • reference (int) – Load alignments from reference directory

  • alignment (int) – Align using corpus texts, acoustic model, and pronunciation dictionary

  • transcription (int) – Transcribe using acoustic model, pronunciation dictionary, and language model

  • phone_transcription (int) – Transcribe using acoustic model and phone-based language model

  • per_speaker_transcription (int) – Transcribe using acoustic model, pronunciation dictionary, and per-speaker language model generated by corpus texts

  • speaker_diarization (int) – Diarize speakers

  • online_alignment (int) – Online alignment

  • acoustic_training (int) – Acoustic model training

  • acoustic_model_adaptation (int) – Acoustic model adaptation

  • segmentation (int) – Segment based on speech activity