PronunciationProbabilityCounter#
- class montreal_forced_aligner.data.PronunciationProbabilityCounter(ngram_counts=<dataclassy.dataclass.Factory object>, word_pronunciation_counts=<dataclassy.dataclass.Factory object>, silence_following_counts=<dataclassy.dataclass.Factory object>, non_silence_following_counts=<dataclassy.dataclass.Factory object>, silence_before_counts=<dataclassy.dataclass.Factory object>, non_silence_before_counts=<dataclassy.dataclass.Factory object>, *args, **kwargs)[source]#
Bases:
object
Data class for count information used in pronunciation probability modeling
- Parameters:
ngram_counts (collections.defaultdict) – Counts of ngrams
word_pronunciation_counts (collections.defaultdict) – Counts of word pronunciations
silence_following_counts (collections.Counter) – Counts of silence following pronunciation
non_silence_following_counts (collections.Counter) – Counts of non-silence following pronunciation
silence_before_counts (collections.Counter) – Counts of silence before pronunciation
non_silence_before_counts (collections.Counter) – Counts of non-silence before pronunciation
- add_counts(other_counter)[source]#
Combine counts of two
PronunciationProbabilityCounter
- Parameters:
other_counter (
PronunciationProbabilityCounter
) – Other object with pronunciation probability counts