Automatic Speech Recognition on MDCC
6.28CERZipformer
Evaluation Results
| Method | Links | |
|---|---|---|
| ZipformerTraining Data=ASR-only, Training Steps=312K, Decoding Strategy=Greedy2026.01 | 6.28 | |
| ZipformerTraining Data=ASR+ST, Training Steps=320K, Decoding Strategy=Greedy2026.01 | 6.37 | |
| FireRed-AED2026.01 | 36.18 | |
| Paraformer2026.01 | 47.72 |