Automatic Speech Recognition on Fuzhou (Fzh.)
2.87WERKimi-Audio
Evaluation Results
| Method | Links | |
|---|---|---|
| Kimi-AudioSize=7B2025.09 | 2.87 | |
| VITA-AudioSize=7B2025.09 | 6.35 | |
| Pretrain+TtTSize=3B2025.09 | 18.99 | |
| TtTSize=3B2025.09 | 55.67 | |
| Qwen2.5-3B (NAR)Size=3B2025.09 | 68.94 | |
| Qwen2.5-3B (AR)Size=3B2025.09 | 90.32 | |
| SpeechGPTSize=7B2025.09 | 101.45 | |
| Mini-OmniSize=0.5B2025.09 | 182.73 |