Phoneme Recognition on MyST (test)
0.0139PERK-WFST
Evaluation results
| Method | Links | ||
|---|---|---|---|
| K-WFSTModel=Kids-FT (fine-tuned on MyST), K-Selection=K=12025.07 | 0.0139 | 0.0983 | |
| K-WFSTModel=Base (pre-trained Wav2Vec2.0), K-Selection=K=12025.07 | 0.0372 | 0.279 | |
| K-WFSTModel=Base (pre-trained Wav2Vec2.0), K-Selection=K=22025.07 | 0.0691 | 0.29 | |
| K-WFSTModel=Kids-FT (fine-tuned on MyST), K-Selection=K=22025.07 | 0.0831 | 0.1467 | |
| GreedyModel=Kids-FT (fine-tuned on MyST)2025.07 | 0.1186 | 0.6589 | |
| GreedyModel=Base (pre-trained Wav2Vec2.0)2025.07 | 0.4026 | 0.6692 |