Language Identification on FLORES-200 CLD3 label set 77 languages (test)
99.9F1 ScoreGlotLID-M
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GlotLID-MArchitecture=fastText, Training Corpus=GlotLID-C2026.02 | 99.9 | 0.0006 | |
| fastTextTraining Corpus=GlotLID-C2026.02 | 99.8 | 0.003 | |
| UniLIDTokenizer=UnigramLM, Training Corpus=GlotLID-C2026.02 | 99.7 | 0.0033 | |
| UniLID-Mistral-NemoTokenizer=Mistral-Nemo-Base-2407, Training Corpus=GlotLID-C2026.02 | 99.4 | 0.0081 | |
| CLD32026.02 | 97.8 | 0.0222 |