Skip to main content
Ctrl+K
Starting with Version 3.4, we are rolling out new model formats that integrate with Hugging Face! See here for the latest updates in MFA 3.4.

Montreal Forced Aligner

  • Getting started
  • User guide
  • API reference
  • Changelog
  • Pretrained MFA models
  • GitHub
  • PyPI
  • PyPI
  • Getting started
  • User guide
  • API reference
  • Changelog
  • Pretrained MFA models
  • GitHub
  • PyPI
  • PyPI
Ctrl+K

Section Navigation

  • Core functionality
    • Corpora
      • AcousticCorpus
      • TextCorpus
      • FileData
      • UtteranceData
      • AlignmentRemapper
    • Database
      • Dictionary
      • Dialect
      • Word
      • Pronunciation
      • Phone
      • Grapheme
      • File
      • TextFile
      • SoundFile
      • Speaker
      • Utterance
      • WordInterval
      • PhoneInterval
      • CorpusWorkflow
      • PhonologicalRule
      • RuleApplication
      • Job
      • M2MSymbol
      • M2M2Job
      • Word2Job
    • Pronunciation dictionaries
      • Main classes
        • MultispeakerDictionary
        • DictionaryRemapper
      • Helper classes and functions
        • DictionaryModel
        • DictionaryMixin
        • TemporaryDictionaryMixin
        • MultispeakerDictionaryMixin
      • Training pronunciation probabilities
        • DictionaryTrainer
    • Acoustic models
      • AcousticModel
      • Training acoustic models
        • TrainableAligner
        • MonophoneTrainer
        • TriphoneTrainer
        • LdaTrainer
        • SatTrainer
        • PronunciationProbabilityTrainer
      • Helper functionality
        • AcousticModelTrainingMixin
        • MonoAlignEqualFunction
        • TreeStatsFunction
        • ConvertAlignmentsFunction
        • LdaAccStatsFunction
        • CalcLdaMlltFunction
        • AccStatsTwoFeatsFunction
        • TransitionAccFunction
        • MonoAlignEqualArguments
        • TreeStatsArguments
        • ConvertAlignmentsArguments
        • LdaAccStatsArguments
        • CalcLdaMlltArguments
        • AccStatsTwoFeatsArguments
        • TransitionAccArguments
    • Grapheme-to-Phoneme (G2P) models
      • G2PModel
      • Training G2P models
        • PyniniTrainer
        • PyniniValidator
        • PhonetisaurusTrainer
        • PyniniTrainerMixin
        • PhonetisaurusTrainerMixin
      • Helper functionality
        • G2PMixin
        • G2PTopLevelMixin
        • G2PTrainer
        • RandomStartWorker
        • RandomStart
    • Language models
      • LanguageModel
      • Training language models
        • MfaLmCorpusTrainer
        • MfaLmDictionaryCorpusTrainer
        • MfaLmArpaTrainer
      • Helper functionality
        • LmTrainerMixin
        • LmCorpusTrainerMixin
        • LmDictionaryCorpusTrainerMixin
        • TrainSpeakerLmFunction
        • TrainSpeakerLmArguments
    • Ivector extraction
      • IvectorExtractorModel
      • Training ivector extractors
        • IvectorTrainer
        • DubmTrainer
        • TrainableIvectorExtractor
      • Training functionality
        • IvectorModelTrainingMixin
        • GmmGselectFunction
        • GmmGselectArguments
        • GaussToPostFunction
        • GaussToPostArguments
        • AccGlobalStatsFunction
        • AccGlobalStatsArguments
        • AccIvectorStatsFunction
        • AccIvectorStatsArguments
  • Workflows
    • Alignment
      • Alignment classes
        • CorpusAligner
        • AdaptingAligner
        • PretrainedAligner
      • Helper functionality
        • AlignMixin
        • AlignFunction
        • FineTuneFunction
        • CompileTrainGraphsFunction
        • AccStatsFunction
        • AlignmentExtractionFunction
        • ExportTextGridProcessWorker
        • PhoneConfidenceFunction
        • AlignArguments
        • AccStatsArguments
        • CompileTrainGraphsArguments
        • AlignmentExtractionArguments
        • ExportTextGridArguments
        • FineTuneArguments
        • PhoneConfidenceArguments
    • Validation
      • Validators
        • TrainingValidator
        • PretrainedValidator
        • DictionaryValidator
      • Helper functions
        • ValidationMixin
    • Generating dictionaries
      • Dictionary generation
        • PyniniCorpusGenerator
        • PyniniWordListGenerator
      • Helper functionality
        • PyniniGenerator
        • Rewriter
        • RewriterWorker
    • Transcription
      • Transcriber
        • Transcriber
      • Helper functions
        • TranscriberMixin
        • CreateHclgFunction
        • CreateHclgArguments
        • DecodeFunction
        • DecodeArguments
        • LmRescoreFunction
        • LmRescoreArguments
        • CarpaLmRescoreFunction
        • CarpaLmRescoreArguments
        • InitialFmllrFunction
        • InitialFmllrArguments
        • FmllrRescoreFunction
        • FmllrRescoreArguments
        • FinalFmllrFunction
        • FinalFmllrArguments
    • Segmentation
      • Segmenter
        • VadSegmenter
        • TranscriptionSegmenter
      • Helper functions
        • SegmentTranscriptFunction
        • SegmentTranscriptArguments
        • SegmentVadFunction
        • SegmentVadArguments
        • get_initial_segmentation
        • merge_segments
        • segment_utterance_transcript
        • segment_utterance_vad
    • Speaker diarization
      • Speaker Diarization
        • SpeakerDiarizer
      • Helper functions
        • PldaClassificationFunction
        • PldaClassificationArguments
        • ComputeEerFunction
        • ComputeEerArguments
        • SpeechbrainEmbeddingFunction
        • SpeechbrainClassificationFunction
        • SpeechbrainArguments
        • cluster_matrix
    • Tokenizers
      • TokenizerModel
      • Training tokenizer models
        • TokenizerTrainer
      • Corpus tokenizer
        • CorpusTokenizer
        • TokenizerValidator
      • Helper functionality
        • TokenizerRewriter
        • TokenizerArguments
        • TokenizerFunction
        • SanitizeFunction
        • SplitWordsFunction
  • Managing MFA servers
    • configure_pg
    • initialize_server
    • check_databases
    • start_server
    • stop_server
    • delete_server
  • Helper
    • Abstract Base Classes
      • KaldiFunction
      • MfaModel
      • MfaWorker
      • TopLevelMfaWorker
      • TrainerMixin
      • TemporaryDirectoryMixin
      • DatabaseMixin
      • AdapterMixin
      • ExporterMixin
      • FileExporterMixin
      • ModelExporterMixin
    • Archive
    • MFA configuration
      • MfaConfiguration
      • MfaProfile
      • get_temporary_directory
      • generate_config_path
      • generate_command_history_path
      • load_command_history
      • update_command_history
    • Data classes
      • MfaArguments
      • TextFileType
      • SoundFileType
      • SoundFileInformation
      • PhoneSetType
      • WordData
      • WordType
      • PhoneType
      • WorkflowType
      • DatabaseImportData
      • PronunciationProbabilityCounter
    • Exception classes
      • MFAError
      • SoxError
      • G2PError
      • ConfigError
      • LMError
      • LanguageModelNotFoundError
      • ModelExtensionError
      • ThirdpartyError
      • TrainerError
      • ModelError
      • CorpusError
      • ModelLoadError
      • CorpusReadError
      • ArgumentError
      • AlignmentExportError
      • NoSuccessfulAlignments
      • KaldiProcessingError
      • TextParseError
      • TextGridParseError
      • DictionaryError
      • NoDefaultSpeakerDictionaryError
      • DictionaryPathError
      • DictionaryFileError
      • FileArgumentNotFoundError
      • PretrainedModelNotFoundError
      • MultipleModelTypesFoundError
      • ModelTypeNotSupportedError
      • PronunciationAcousticMismatchError
      • RootDirectoryError
    • Helper functions
      • comma_join
      • make_safe
      • make_scp_safe
      • load_scp
      • load_scp_safe
      • score_wer
      • edit_distance
      • output_mapping
    • Textgrid utilities
      • process_ctm_line
      • export_textgrid
      • construct_output_path
      • output_textgrid_writing_errors
    • Utility functions
      • Counter
      • run_kaldi_function
      • thirdparty_binary
      • log_kaldi_errors
      • parse_logs
  • MFA API
  • Core functionality
  • Pronunciation dictionaries
  • Main classes

Main classes#

MultispeakerDictionary([dictionary, ...])

Class for processing multi- and single-speaker pronunciation dictionaries

DictionaryRemapper([acoustic_model, ...])

previous

Pronunciation dictionaries

next

MultispeakerDictionary
Show Source

© Copyright 2018-2026, Montreal Corpus Tools.

Created using Sphinx 9.1.0.

Built with the PyData Sphinx Theme 0.19.0.