Zero-shot Speech Evaluation on sWUGGY
84.1sWUGGY In-Vocab ScoreTWIST-13B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| TWIST-13BInitialization=Warm (TWIST), Model Scale=13B, Tokenizer Frequency=25Hz, Number of Tokens=5002023.05 | 84.1 | 74.5 | |
| Borsos et al.2023.05 | 83.7 | 71.5 | |
| TWIST-7BInitialization=Warm (TWIST), Model Scale=7B, Tokenizer Frequency=25Hz, Number of Tokens=5002023.05 | 83.6 | 73.9 | |
| TWIST-1.3BInitialization=Warm (TWIST), Model Scale=1.3B, Tokenizer Frequency=25Hz, Number of Tokens=5002023.05 | 82.5 | 72.7 | |
| COLD-INIT-1.3BInitialization=Cold, Model Scale=1.3B, Tokenizer Frequency=25Hz, Number of Tokens=5002023.05 | 81.9 | 72.2 | |
| van Niekerk et al.2023.05 | 72.3 | 64.3 | |
| Lakhotia et al.2023.05 | 68.7 | — |