This model cannot provide accurate assessments of goodness of pronunciations or provide transcripts as it is trained to be accepting of variation in pronunciation to provide a reasonable alignment for Thai speech.

Bias, Risks, and Limitations

This model will perform best on the variety of speech that it was trained on. The speakers in the training data are all adult speakers, so child speech alignment may not be accurate.

Recommendations

When using this model on a variety that it was not trained on, better results can be attained by adapting the model to the data to be aligned first. See https://montreal-forced-aligner.readthedocs.io/en/latest/user_guide/workflows/adapt_acoustic_model.html and https://github.com/mmcauliffe/mfa-adaptation for example usage and scripts.

How to Get Started with the Model

Use the code below to get started with the model.

To get started, follow the instructions for installing MFA. To align files using this model, use the mfa align command.

Dictionary Details

Details for thai_mfa dictionary and G2P model

Source: wikipron
Orthography: Latin
Phone set: MFA
Words: 53,840

Phones: 45
Graphemes: 107

IPA chart

Consonants

Manner	Labial	Labiodental	Alveolar	Palatal	Velar	Glottal
Nasal	m		n		ŋ
Stop	p b		t d		k	ʔ
Aspirated	pʰ		tʰ		kʰ
Unreleased	p̚		t̚		k̚
Affricate				tɕ
Aspirated				tɕʰ
Sibilant			s
Fricative		f				h
Approximant	w			j
Trill			r
Lateral			l

Vowels

	Front	Central	Back
Close	i iː		ɯ ɯː u uː

Close-Mid	e eː		ɤ ɤː o oː

Open-Mid	ɛ ɛː		ɔ ɔː

Open		a aː

Diphthongs

Training Details

Training Data

Common Voice Thai

Source: https://voice.mozilla.org/en/datasets
License: CC-0
Dialects: N/A
Number of hours: 167.39
Number of utterances: 139,753
Number of speakers: 6,385
- Female speakers: 245
- Male speakers: 328
- Unknown speakers: 5,812

GlobalPhone Thai

Source: https://catalogue.elra.info/en-us/repository/browse/ELRA-S0321/
License: ELRA
Dialects: N/A
Number of hours: 28.17
Number of utterances: 14,039
Number of speakers: 98
- Female speakers: 65
- Male speakers: 27
- Unknown speakers: 6

Gowajee Corpus

Source: https://github.com/ekapolc/gowajee_corpus?tab=readme-ov-file
License: MIT
Dialects: N/A
Number of hours: 17.19
Number of utterances: 20,307
Number of speakers: 188
- Female speakers: 25
- Male speakers: 163
- Unknown speakers: 0

Lotus Corpus

Source: https://github.com/korakot/corpus/releases/tag/v1.0
License: CC BY-SA-NC 3.0
Dialects: N/A
Number of hours: 5.92
Number of utterances: 3,360
Number of speakers: 96
- Female speakers: 48
- Male speakers: 48
- Unknown speakers: 0

Thai Elderly Speech dataset by Data Wow and VISAI

Source: https://github.com/VISAI-DATAWOW/Thai-Elderly-Speech-dataset/releases/tag/v1.0.0
License: CC BY-SA 4.0
Dialects: N/A
Number of hours: 26.56
Number of utterances: 19,200
Number of speakers: 32
- Female speakers: 22
- Male speakers: 10
- Unknown speakers: 0

Training Procedure

Preprocessing

Preprocessing include fixes and orthographic standardization to various corpora.

Training Hyperparameters

Training regime: Training configuration

Evaluation

Testing Data, Factors & Metrics

Testing Data

N/A

Factors

N/A

Metrics

N/A

Results

N/A

Summary

Technical Specifications

Model Architecture and Objective

HMM-GMM model

Software

This model was trained via the Montreal Forced Aligner.

Citation

BibTeX:

@techreport{mfa_thai_mfa_acoustic_2026,
	author={McAuliffe, Michael and Sonderegger, Morgan},
	title={Thai MFA acoustic model v3.3.0},
	address={\url{https://huggingface.co/MontrealCorpusTools/thai_mfa}},
	year={2026},
	month={Jun},
}

APA:

McAuliffe, M. & Sonderegger, M. (2026). Thai MFA acoustic model v3.3.0. Available at https://huggingface.co/MontrealCorpusTools/thai_mfa.

Contact

For questions and issues, please file an issue either for this model at https://huggingface.co/MontrealCorpusTools/thai_mfa/discussions or for larger MFA issues at https://github.com/MontrealCorpusTools/Montreal-Forced-Aligner/issues.