serbocroatian_serbian_mfa v3.3.0

Serbocroatian (Serbian) MFA G2P model v3.3.0

Link to documentation on mfa-models

Jump to section:

• Model details
• Intended use
• Performance Factors
• Metrics
• Training data
• Evaluation data
• Ethical considerations
• Troubleshooting issues

Model details

• Maintainer: Montreal Forced Aligner
• Language: Serbo-Croatian
• Dialect: N/A
• Phone set: MFA
• Model type: G2P model
• Architecture: phonetisaurus
• Model version: v3.3.0
• Trained date: 2024-10-25
• Compatible MFA version: v3.3.0
• License: CC BY 4.0
• Citation:

@techreport{mfa_serbocroatian_serbian_mfa_g2p_2024,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Serbocroatian (Serbian) MFA G2P model v3.3.0},
	address={\url{https://mfa-models.readthedocs.io/G2P model/Serbocroatian/Serbocroatian (Serbian) MFA G2P model v3_3_0.html}},
	year={2024},
	month={Oct},
}

• If you have comments or questions about this model, you can check previous MFA model discussion posts or create a new one.

Installation

Install from the MFA command line:

mfa model download g2p serbocroatian_serbian_mfa

Or download from the release page.

Intended use

This model is intended for generating pronunciations of Serbo-Croatian transcripts.

This model uses the MFA phone set for Serbocroatian, and was trained from the pronunciation dictionaries above. Pronunciations generated with this G2P model can be appended and used when aligning or transcribing.

Performance Factors

The trained G2P models should be relatively quick and accurate, however the model may struggle when dealing with less common orthographic characters or word types outside of what it was trained on. If so, you may need to supplement the dictionary through generating, correcting, and re-training the G2P model as necessary.

Metrics

The model was trained on 90% of the dictionary and evaluated on a held-out 10% and evaluated with word error rate and phone error rate.

Training

This model was trained on the following data set:

• Words: 19,917
• Phones: 61
• Graphemes: 32

Evaluation

This model was evaluated on the following data set:

• Words: 2,212
• WER: 100.00%
• PER: 100.00%

Ethical considerations

Deploying any model involving language into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias

You should assume every machine learning model has demographic bias unless proven otherwise. For G2P models, the model will only process the types of tokens that it was trained on, and will not represent the full range of text or spoken words that native speakers will produce. If you are using this model in production, you should acknowledge this as a potential issue.

Surveillance

Speech-to-Text technologies may be misused to invade the privacy of others by recording and mining information from private conversations. This kind of individual privacy is protected by law in many countries. You should not assume consent to record and analyze private speech.

serbocroatian_croatian_mfa v3.3.0

Serbocroatian (Croatian) MFA G2P model v3.3.0

Link to documentation on mfa-models

Jump to section:

• Model details
• Intended use
• Performance Factors
• Metrics
• Training data
• Evaluation data
• Ethical considerations
• Troubleshooting issues

Model details

• Maintainer: Montreal Forced Aligner
• Language: Serbo-Croatian
• Dialect: N/A
• Phone set: MFA
• Model type: G2P model
• Architecture: phonetisaurus
• Model version: v3.3.0
• Trained date: 2024-10-25
• Compatible MFA version: v3.3.0
• License: CC BY 4.0
• Citation:

@techreport{mfa_serbocroatian_croatian_mfa_g2p_2024,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Serbocroatian (Croatian) MFA G2P model v3.3.0},
	address={\url{https://mfa-models.readthedocs.io/G2P model/Serbocroatian/Serbocroatian (Croatian) MFA G2P model v3_3_0.html}},
	year={2024},
	month={Oct},
}

• If you have comments or questions about this model, you can check previous MFA model discussion posts or create a new one.

Installation

Install from the MFA command line:

mfa model download g2p serbocroatian_croatian_mfa

Or download from the release page.

Intended use

This model is intended for generating pronunciations of Serbo-Croatian transcripts.

This model uses the MFA phone set for Serbocroatian, and was trained from the pronunciation dictionaries above. Pronunciations generated with this G2P model can be appended and used when aligning or transcribing.

Performance Factors

The trained G2P models should be relatively quick and accurate, however the model may struggle when dealing with less common orthographic characters or word types outside of what it was trained on. If so, you may need to supplement the dictionary through generating, correcting, and re-training the G2P model as necessary.

Metrics

The model was trained on 90% of the dictionary and evaluated on a held-out 10% and evaluated with word error rate and phone error rate.

Training

This model was trained on the following data set:

• Words: 21,595
• Phones: 61
• Graphemes: 30

Evaluation

This model was evaluated on the following data set:

• Words: 1,769
• WER: 100.00%
• PER: 100.00%

Ethical considerations

Deploying any model involving language into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias

You should assume every machine learning model has demographic bias unless proven otherwise. For G2P models, the model will only process the types of tokens that it was trained on, and will not represent the full range of text or spoken words that native speakers will produce. If you are using this model in production, you should acknowledge this as a potential issue.

Surveillance

Speech-to-Text technologies may be misused to invade the privacy of others by recording and mining information from private conversations. This kind of individual privacy is protected by law in many countries. You should not assume consent to record and analyze private speech.

serbocroatian_mfa v3.3.0

Serbocroatian MFA dictionary v3.3.0

Link to documentation on mfa-models

Jump to section:

• Dictionary details
• Intended use
• Performance Factors
• Ethical considerations

Dictionary details

• Maintainer: Montreal Forced Aligner
• Language: Serbo-Croatian
• Dialect: N/A
• Phone set: MFA
• Number of words: 69,307
• Phones: a aː aː˦˨ aː˨˦ a˦˨ a˨˦ b d dʑ dʒ e eː eː˦˨ eː˨˦ e˦˨ e˨˦ f i iː iː˦˨ iː˨˦ i˦˨ i˨˦ j k l m n o oː oː˦˨ oː˨˦ o˦˨ o˨˦ p r rː r̩ r̩ː˦˨ r̩ː˨˦ r̩˦˨ r̩˨˦ s t ts tɕ tʃ u uː uː˦˨ uː˨˦ u˦˨ u˨˦ v x z ɡ ɲ ʃ ʎ ʒ
• License: CC BY 4.0
• Compatible MFA version: v3.3.0
• Citation:

@techreport{mfa_serbocroatian_mfa_dictionary_2025,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Serbocroatian MFA dictionary v3.3.0},
	address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/Serbocroatian/Serbocroatian MFA dictionary v3_3_0.html}},
	year={2025},
	month={Oct},
}

• If you have comments or questions about this dictionary or its phone set, you can check previous MFA model discussion posts or create a new one.
• The dictionary downloadable from this release has trained pronunciation and silence probabilities. The base dictionary is available here

##Installation

Install from the MFA command line:

mfa model download dictionary serbocroatian_mfa

Or download from the release page.

The dictionary available from the release page and command line installation has pronunciation and silence probabilities estimated as part acoustic model training (see Silence probability format and training pronunciation probabilities for more information. If you would like to use the version of this dictionary without probabilities, please see the [plain dictionary](https://raw.githubusercontent.com/MontrealCorpusTools/mfa-models/main/dictionary/serbocroatian/mfa/Serbocroatian MFA dictionary v3_3_0.dict).

Intended use

This dictionary is intended for forced alignment of Serbo-Croatian transcripts.

This dictionary uses the MFA phone set for Serbocroatian, and was used in training the Serbocroatian MFA acoustic model. Pronunciations can be added on top of the dictionary, as long as no additional phones are introduced.

Performance Factors

When trying to get better alignment accuracy, adding pronunciations is generally helpful, especially for different styles and dialects. The most impactful improvements will generally be seen when adding reduced variants that involve deleting segments/syllables common in spontaneous speech. Alignment must include all phones specified in the pronunciation of a word, and each phone has a minimum duration (by default 10ms). If a speaker pronounces a multisyllabic word with just a single syllable, it can be hard for MFA to fit all the segments in, so it will lead to alignment errors on adjacent words as well.

Ethical considerations

Deploying any Speech-to-Text model into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias

You should assume every machine learning model has demographic bias unless proven otherwise. For pronunciation dictionaries, it is often the case that transcription accuracy and lexicon coverage for the prestige variety modeled in this dictionary compared to other variants. If you are using this dictionary in production, you should acknowledge this as a potential issue.

serbocroatian_mfa v3.3.0

acoustic-serbocroatian_mfa-v3.3.0

Remove hyphen from Serbo-Croatian for now

spanish_spain_mfa v3.3.0

Spanish (Spain) MFA G2P model v3.3.0

Link to documentation on mfa-models

Jump to section:

• Model details
• Intended use
• Performance Factors
• Metrics
• Training data
• Evaluation data
• Ethical considerations
• Troubleshooting issues

Model details

• Maintainer: Montreal Forced Aligner
• Language: Spanish
• Dialect: Peninsular Spanish
• Phone set: MFA
• Model type: G2P model
• Architecture: phonetisaurus
• Model version: v3.3.0
• Trained date: 2024-09-15
• Compatible MFA version: v3.3.0
• License: CC BY 4.0
• Citation:

@techreport{mfa_spanish_spain_mfa_g2p_2024,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Spanish (Spain) MFA G2P model v3.3.0},
	address={\url{https://mfa-models.readthedocs.io/G2P model/Spanish/Spanish (Spain) MFA G2P model v3_3_0.html}},
	year={2024},
	month={Sep},
}

• If you have comments or questions about this model, you can check previous MFA model discussion posts or create a new one.

Installation

Install from the MFA command line:

mfa model download g2p spanish_spain_mfa

Or download from the release page.

Intended use

This model is intended for generating pronunciations of Spanish transcripts.

This model uses the MFA phone set for Spanish, and was trained from the pronunciation dictionaries above. Pronunciations generated with this G2P model can be appended and used when aligning or transcribing.

Performance Factors

The trained G2P models should be relatively quick and accurate, however the model may struggle when dealing with less common orthographic characters or word types outside of what it was trained on. If so, you may need to supplement the dictionary through generating, correcting, and re-training the G2P model as necessary.

Metrics

The model was trained on 90% of the dictionary and evaluated on a held-out 10% and evaluated with word error rate and phone error rate.

Training

This model was trained on the following data set:

• Words: 139,661
• Phones: 43
• Graphemes: 38

Evaluation

This model was evaluated on the following data set:

• Words: 11,790
• WER: 100.00%
• PER: 100.00%

Ethical considerations

Deploying any model involving language into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias

You should assume every machine learning model has demographic bias unless proven otherwise. For G2P models, the model will only process the types of tokens that it was trained on, and will not represent the full range of text or spoken words that native speakers will produce. If you are using this model in production, you should acknowledge this as a potential issue.

Surveillance

Speech-to-Text technologies may be misused to invade the privacy of others by recording and mining information from private conversations. This kind of individual privacy is protected by law in many countries. You should not assume consent to record and analyze private speech.

spanish_latin_america_mfa v3.3.0

Spanish (Latin America) MFA G2P model v3.3.0

Link to documentation on mfa-models

Jump to section:

• Model details
• Intended use
• Performance Factors
• Metrics
• Training data
• Evaluation data
• Ethical considerations
• Troubleshooting issues

Model details

• Maintainer: Montreal Forced Aligner
• Language: Spanish
• Dialect: Spanish in the Americas
• Phone set: MFA
• Model type: G2P model
• Architecture: phonetisaurus
• Model version: v3.3.0
• Trained date: 2024-09-15
• Compatible MFA version: v3.3.0
• License: CC BY 4.0
• Citation:

@techreport{mfa_spanish_latin_america_mfa_g2p_2024,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Spanish (Latin America) MFA G2P model v3.3.0},
	address={\url{https://mfa-models.readthedocs.io/G2P model/Spanish/Spanish (Latin America) MFA G2P model v3_3_0.html}},
	year={2024},
	month={Sep},
}

• If you have comments or questions about this model, you can check previous MFA model discussion posts or create a new one.

Installation

Install from the MFA command line:

mfa model download g2p spanish_latin_america_mfa

Or download from the release page.

Intended use

This model is intended for generating pronunciations of Spanish transcripts.

This model uses the MFA phone set for Spanish, and was trained from the pronunciation dictionaries above. Pronunciations generated with this G2P model can be appended and used when aligning or transcribing.

Performance Factors

The trained G2P models should be relatively quick and accurate, however the model may struggle when dealing with less common orthographic characters or word types outside of what it was trained on. If so, you may need to supplement the dictionary through generating, correcting, and re-training the G2P model as necessary.

Metrics

The model was trained on 90% of the dictionary and evaluated on a held-out 10% and evaluated with word error rate and phone error rate.

Training

This model was trained on the following data set:

• Words: 139,675
• Phones: 42
• Graphemes: 38

Evaluation

This model was evaluated on the following data set:

• Words: 11,776
• WER: 100.00%
• PER: 100.00%

Ethical considerations

Deploying any model involving language into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias

You should assume every machine learning model has demographic bias unless proven otherwise. For G2P models, the model will only process the types of tokens that it was trained on, and will not represent the full range of text or spoken words that native speakers will produce. If you are using this model in production, you should acknowledge this as a potential issue.

Surveillance

Speech-to-Text technologies may be misused to invade the privacy of others by recording and mining information from private conversations. This kind of individual privacy is protected by law in many countries. You should not assume consent to record and analyze private speech.

czech_mfa v3.3.0

Czech MFA G2P model v3.3.0

Link to documentation on mfa-models

Jump to section:

• Model details
• Intended use
• Performance Factors
• Metrics
• Training data
• Evaluation data
• Ethical considerations
• Troubleshooting issues

Model details

• Maintainer: Montreal Forced Aligner
• Language: Czech
• Dialect: N/A
• Phone set: MFA
• Model type: G2P model
• Architecture: phonetisaurus
• Model version: v3.3.0
• Trained date: 2024-10-31
• Compatible MFA version: v3.3.0
• License: CC BY 4.0
• Citation:

@techreport{mfa_czech_mfa_g2p_2024,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Czech MFA G2P model v3.3.0},
	address={\url{https://mfa-models.readthedocs.io/G2P model/Czech/Czech MFA G2P model v3_3_0.html}},
	year={2024},
	month={Oct},
}

• If you have comments or questions about this model, you can check previous MFA model discussion posts or create a new one.

Installation

Install from the MFA command line:

mfa model download g2p czech_mfa

Or download from the release page.

Intended use

This model is intended for generating pronunciations of Czech transcripts.

This model uses the MFA phone set for Czech, and was trained from the pronunciation dictionaries above. Pronunciations generated with this G2P model can be appended and used when aligning or transcribing.

Performance Factors

The trained G2P models should be relatively quick and accurate, however the model may struggle when dealing with less common orthographic characters or word types outside of what it was trained on. If so, you may need to supplement the dictionary through generating, correcting, and re-training the G2P model as necessary.

Metrics

The model was trained on 90% of the dictionary and evaluated on a held-out 10% and evaluated with word error rate and phone error rate.

Training

This model was trained on the following data set:

• Words: 41,230
• Phones: 47
• Graphemes: 46

Evaluation

This model was evaluated on the following data set:

• Words: 1,617
• WER: 100.00%
• PER: 100.00%

Ethical considerations

Deploying any model involving language into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias

You should assume every machine learning model has demographic bias unless proven otherwise. For G2P models, the model will only process the types of tokens that it was trained on, and will not represent the full range of text or spoken words that native speakers will produce. If you are using this model in production, you should acknowledge this as a potential issue.

Surveillance

Speech-to-Text technologies may be misused to invade the privacy of others by recording and mining information from private conversations. This kind of individual privacy is protected by law in many countries. You should not assume consent to record and analyze private speech.

spanish_spain_mfa v3.3.0

Spanish (Spain) MFA dictionary v3.3.0

Link to documentation on mfa-models

Jump to section:

• Dictionary details
• Intended use
• Performance Factors
• Ethical considerations

Dictionary details

• Maintainer: Montreal Forced Aligner
• Language: Spanish
• Dialect: Peninsular Spanish
• Phone set: MFA
• Number of words: 87,777
• Phones: a ã b c d̪ e ẽ f h i ĩ j k l m n o õ p r s tʃ t̪ u ũ v w x z ç ð ŋ ɟ ɟʝ ɡ ɣ ɱ ɲ ɾ ʃ ʎ ʝ β θ
• License: CC BY 4.0
• Compatible MFA version: v3.3.0
• Citation:

@techreport{mfa_spanish_spain_mfa_dictionary_2025,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Spanish (Spain) MFA dictionary v3.3.0},
	address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/Spanish/Spanish (Spain) MFA dictionary v3_3_0.html}},
	year={2025},
	month={Sep},
}

• If you have comments or questions about this dictionary or its phone set, you can check previous MFA model discussion posts or create a new one.
• The dictionary downloadable from this release has trained pronunciation and silence probabilities. The base dictionary is available here

##Installation

Install from the MFA command line:

mfa model download dictionary spanish_spain_mfa

Or download from the release page.

The dictionary available from the release page and command line installation has pronunciation and silence probabilities estimated as part acoustic model training (see Silence probability format and training pronunciation probabilities for more information. If you would like to use the version of this dictionary without probabilities, please see the [plain dictionary](https://raw.githubusercontent.com/MontrealCorpusTools/mfa-models/main/dictionary/spanish/mfa/Spanish (Spain) MFA dictionary v3_3_0.dict).

Intended use

This dictionary is intended for forced alignment of Spanish transcripts.

This dictionary uses the MFA phone set for Spanish, and was used in training the Spanish MFA acoustic model. Pronunciations can be added on top of the dictionary, as long as no additional phones are introduced.

Performance Factors

When trying to get better alignment accuracy, adding pronunciations is generally helpful, especially for different styles and dialects. The most impactful improvements will generally be seen when adding reduced variants that involve deleting segments/syllables common in spontaneous speech. Alignment must include all phones specified in the pronunciation of a word, and each phone has a minimum duration (by default 10ms). If a speaker pronounces a multisyllabic word with just a single syllable, it can be hard for MFA to fit all the segments in, so it will lead to alignment errors on adjacent words as well.

Ethical considerations

Deploying any Speech-to-Text model into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias

You should assume every machine learning model has demographic bias unless proven otherwise. For pronunciation dictionaries, it is often the case that transcription accuracy and lexicon coverage for the prestige variety modeled in this dictionary compared to other variants. If you are using this dictionary in production, you should acknowledge this as a potential issue.

spanish_latin_america_mfa v3.3.0

Spanish (Latin America) MFA dictionary v3.3.0

Link to documentation on mfa-models

Jump to section:

• Dictionary details
• Intended use
• Performance Factors
• Ethical considerations

Dictionary details

• Maintainer: Montreal Forced Aligner
• Language: Spanish
• Dialect: Spanish in the Americas
• Phone set: MFA
• Number of words: 51,946
• Phones: a ã b c d̪ e ẽ f h i ĩ j k l m n o õ p r s tʃ t̪ u ũ v w x z ç ð ŋ ɟ ɟʝ ɡ ɣ ɱ ɲ ɾ ʃ ʎ ʝ β
• License: CC BY 4.0
• Compatible MFA version: v3.3.0
• Citation:

@techreport{mfa_spanish_latin_america_mfa_dictionary_2025,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Spanish (Latin America) MFA dictionary v3.3.0},
	address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/Spanish/Spanish (Latin America) MFA dictionary v3_3_0.html}},
	year={2025},
	month={Sep},
}

• If you have comments or questions about this dictionary or its phone set, you can check previous MFA model discussion posts or create a new one.
• The dictionary downloadable from this release has trained pronunciation and silence probabilities. The base dictionary is available here

##Installation

Install from the MFA command line:

mfa model download dictionary spanish_latin_america_mfa

Or download from the release page.

The dictionary available from the release page and command line installation has pronunciation and silence probabilities estimated as part acoustic model training (see Silence probability format and training pronunciation probabilities for more information. If you would like to use the version of this dictionary without probabilities, please see the [plain dictionary](https://raw.githubusercontent.com/MontrealCorpusTools/mfa-models/main/dictionary/spanish/mfa/Spanish (Latin America) MFA dictionary v3_3_0.dict).

Intended use

This dictionary is intended for forced alignment of Spanish transcripts.

This dictionary uses the MFA phone set for Spanish, and was used in training the Spanish MFA acoustic model. Pronunciations can be added on top of the dictionary, as long as no additional phones are introduced.

Performance Factors

When trying to get better alignment accuracy, adding pronunciations is generally helpful, especially for different styles and dialects. The most impactful improvements will generally be seen when adding reduced variants that involve deleting segments/syllables common in spontaneous speech. Alignment must include all phones specified in the pronunciation of a word, and each phone has a minimum duration (by default 10ms). If a speaker pronounces a multisyllabic word with just a single syllable, it can be hard for MFA to fit all the segments in, so it will lead to alignment errors on adjacent words as well.

Ethical considerations

Deploying any Speech-to-Text model into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias

You should assume every machine learning model has demographic bias unless proven otherwise. For pronunciation dictionaries, it is often the case that transcription accuracy and lexicon coverage for the prestige variety modeled in this dictionary compared to other variants. If you are using this dictionary in production, you should acknowledge this as a potential issue.

czech_mfa v3.3.0

Czech MFA dictionary v3.3.0

Link to documentation on mfa-models

Jump to section:

• Dictionary details
• Intended use
• Performance Factors
• Ethical considerations

Dictionary details

• Maintainer: Montreal Forced Aligner
• Language: Czech
• Dialect: N/A
• Phone set: MFA
• Number of words: 67,316
• Phones: a aw aː b c d dz dʒ ew f iː j k l l̩ m m̩ n n̩ o ow oː p r r̝ r̝̊ r̩ s t ts tʃ u uː v x z ŋ ə ɛ ɛː ɟ ɡ ɦ ɪ ɲ ʃ ʒ ʔ
• License: CC BY 4.0
• Compatible MFA version: v3.3.0
• Citation:

@techreport{mfa_czech_mfa_dictionary_2025,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Czech MFA dictionary v3.3.0},
	address={\url{https://mfa-models.readthedocs.io/pronunciation dictionary/Czech/Czech MFA dictionary v3_3_0.html}},
	year={2025},
	month={Sep},
}

• If you have comments or questions about this dictionary or its phone set, you can check previous MFA model discussion posts or create a new one.
• The dictionary downloadable from this release has trained pronunciation and silence probabilities. The base dictionary is available here

##Installation

Install from the MFA command line:

mfa model download dictionary czech_mfa

Or download from the release page.

The dictionary available from the release page and command line installation has pronunciation and silence probabilities estimated as part acoustic model training (see Silence probability format and training pronunciation probabilities for more information. If you would like to use the version of this dictionary without probabilities, please see the [plain dictionary](https://raw.githubusercontent.com/MontrealCorpusTools/mfa-models/main/dictionary/czech/mfa/Czech MFA dictionary v3_3_0.dict).

Intended use

This dictionary is intended for forced alignment of Czech transcripts.

This dictionary uses the MFA phone set for Czech, and was used in training the Czech MFA acoustic model. Pronunciations can be added on top of the dictionary, as long as no additional phones are introduced.

Performance Factors

When trying to get better alignment accuracy, adding pronunciations is generally helpful, especially for different styles and dialects. The most impactful improvements will generally be seen when adding reduced variants that involve deleting segments/syllables common in spontaneous speech. Alignment must include all phones specified in the pronunciation of a word, and each phone has a minimum duration (by default 10ms). If a speaker pronounces a multisyllabic word with just a single syllable, it can be hard for MFA to fit all the segments in, so it will lead to alignment errors on adjacent words as well.

Ethical considerations

Deploying any Speech-to-Text model into any production setting has ethical implications. You should consider these implications before use.

Demographic Bias

You should assume every machine learning model has demographic bias unless proven otherwise. For pronunciation dictionaries, it is often the case that transcription accuracy and lexicon coverage for the prestige variety modeled in this dictionary compared to other variants. If you are using this dictionary in production, you should acknowledge this as a potential issue.