Model Card for Model ID
Model Details
Model Description
Luxembourgish MFA acoustic model (lb_acoustic_lux) with pronunciation dictionary and G2P for forced alignment.
- Developed by: ['Peter Gilles', 'University of Luxembourg']
- Funded by: University of Luxembourg
- Model type: Montreal Forced Aligner model
- Language(s) (NLP): Luxembourgish
- License: cc-by-4.0
Uses
Direct Use
Forced alignment of Luxembourgish speech using MFA align_hf.
Out-of-Scope Use
This model cannot provide accurate assessments of goodness of pronunciations or provide transcripts.
Bias, Risks, and Limitations
Best on Luxembourgish varieties similar to the training data.
Recommendations
Adapt the model when aligning substantially different varieties.
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:
mfa align_hf CORPUS pgilles/luxembourgish_mfa OUTPUT \
--dialect luxembourgish --use_g2p
Dictionary Details
Details for luxembourgish_run6 dictionary and G2P model
- Source: wikipron
- Orthography: [More Information Needed]
- Phone set: [More Information Needed]
- Words: 58,731
- Phones: 70
- Graphemes: 59
IPA chart
Consonants
| Manner |
|---|
Vowels
| Front | Near-Front | Central | Near-Back | Back | |
|---|---|---|---|---|---|
| Close | |||||
| Close-Mid | |||||
| Open-Mid | |||||
| Open |
Other
- ð
- ɲ
- œː
- p
- uː
- y
- h
- u
- n
- ɔ
- ɕ
- ɥ
- æːɪ
- o
- yː
- ɑɪ
- ɹ
- s
- mː
- tʃ
- ɜɪ
- ŋ
- f
- iː
- ʒ
- æl
- l
- b
- iə
- ɛː
- uə
- œ
- ən
- t
- dʒ
- øː
- əʊ
- k
- aː
- j
- ɛ
- χ
- ɑː
- i
- e
- ɔː
- w
- oː
- a
- oɪ
- eː
- v
- ts
- ʁ
- ɔɪ
- ɡ
- d̥ː
- ʑ
- m
- θ
- ø
- ɑ
- z
- æː
- ɑʊ
- ʀ
- dz
- d
- ʃ
- æːʊ
Training Details
Training Data
lb_acoustic_lux
- Source: [More Information Needed]
- License: cc-by-4.0
- Dialects: N/A
- Number of hours: 67.53
- Number of utterances: 88,183
- Number of speakers: 7
- Female speakers: 0
- Male speakers: 0
- Unknown speakers: 0
Training Procedure
Preprocessing
Standard MFA corpus preparation for Luxembourgish speech.
Training Hyperparameters
- Training regime: Training configuration
Evaluation
Testing Data, Factors & Metrics
Testing Data
[More Information Needed]
Factors
[More Information Needed]
Metrics
[More Information Needed]
Results
[More Information Needed]
Summary
Technical Specifications
Model Architecture and Objective
HMM-GMM model
Software
Montreal Forced Aligner 3.4
Citation
BibTeX:
[More Information Needed]
APA:
[More Information Needed]
Contact
[More Information Needed]