|
--- |
|
license: cc-by-nc-sa-4.0 |
|
--- |
|
|
|
This repository contains KenLM models for the Ukrainian language |
|
|
|
Metrics for the NEWS models (tested with an acoustic model of [wav2vec2-xls-r-300m model](https://huggingface.co/Yehor/wav2vec2-xls-r-300m-uk-with-small-lm)): |
|
|
|
| Model | CER | WER | |
|
|-|-|-| |
|
| no LM | 0.0412 | 0.2206 | |
|
| lm-3gram-50k | 0.0348 | 0.1826 | |
|
| lm-4gram-50k | 0.0347 | 0.1818 | |
|
| lm-5gram-50k | 0.0347 | 0.1821 | |
|
| lm-3gram-100k | 0.031 | 0.1588 | |
|
| lm-4gram-100k | 0.0308 | 0.1579 | |
|
| lm-5gram-100k | 0.0308 | 0.1579 | |
|
| lm-3gram-300k | 0.0261 | 0.1294 | |
|
| lm-4gram-300k | 0.0261 | 0.1293 | |
|
| lm-5gram-300k | 0.0261 | 0.1293 | |
|
| lm-3gram-500k | 0.0248 | 0.1209 | |
|
| lm-4gram-500k | 0.0247 | 0.1207 | |
|
| lm-5gram-500k | 0.0247 | 0.1209 | |
|
|
|
Files of the models are under the Files and versions section. |
|
|
|
Attribution to the NEWS models: |
|
- Chaplynskyi, D. et al. (2021) lang-uk Ukrainian Ubercorpus [Data set]. https://lang.org.ua/uk/corpora/#anchor4 |
|
|
