Automatic Speech Recognition on 23K-word (test)
61.1Frame Accuracy10xEnsemble
Evaluation results
| Method | Links | ||
|---|---|---|---|
| 10xEnsembleEnsemble size=102015.03 | 61.1 | 10.7 | |
| Distilled Single modelSystem=Distilled, Number of models in teacher ensemble=102015.03 | 60.8 | 10.7 | |
| BaselineSystem=Baseline2015.03 | 58.9 | 10.9 |