AcousticModel

class montreal_forced_aligner.models.AcousticModel(source, root_directory=None)

Bases: Archive

Class for storing acoustic models in MFA, exported as zip files containing the necessary Kaldi files to be reused

add_meta_file(trainer)

Add metadata file from a model trainer

Parameters:

trainer (ModelExporterMixin) – Trainer to supply metadata information about the acoustic model

add_model(source)

Add file into archive

Parameters:

source (str) – File to add

add_pronunciation_models(source, dictionary_base_names)

Add file into archive

Parameters:

• source (str) – File to add

• dictionary_base_names (list[str]) – Base names of dictionaries to add pronunciation models

property alignment_model_path

Alignment model path

export_model(destination)

Extract the model files to a new directory

Parameters:

destination (str) – Destination directory to extract files to

property lda_mat_path

Current acoustic model path

property lda_options

Parameters to use in computing MFCC features.

log_details()

Log metadata information to a logger

property meta

Metadata information for the acoustic model

property mfcc_options

Parameters to use in computing MFCC features.

property model_path

Current acoustic model path

property parameters

Parameters to pass to top-level workers

property phone_symbol_path

Path to phone symbol table

property pitch_options

Parameters to use in computing MFCC features.

pretty_print()

Prints the metadata information to the terminal

property rules_path

Path to phone symbol table

property tree_path

Current acoustic model path

validate(dictionary)

Validate this acoustic model against a pronunciation dictionary to ensure their phone sets are compatible

Parameters:

dictionary (DictionaryMixin) – DictionaryMixin to compare phone sets with

Raises:

PronunciationAcousticMismatchError – If there are phones missing from the acoustic model