MfaLmArpaTrainer#

class montreal_forced_aligner.language_modeling.trainer.MfaLmArpaTrainer(arpa_path, keep_case=False, **kwargs)[source]#

Bases: LmTrainerMixin, TopLevelMfaWorker, DatabaseMixin

Top-level worker to convert an existing ARPA-format language model to MFA format

See also

LmTrainerMixin: For language model training parsing parameters
TopLevelMfaWorker: For top-level parsing parameters

property data_directory#: Data directory

property data_source_identifier#: Data source identifier

property meta#: Metadata information for the trainer

setup()[source]#: Set up language model training

train()[source]#: Convert the arpa model to MFA format

property working_directory#: Current working directory