Automatic Speech Recognition on CV-yue
7.44CERZipformer
Evaluation Results
| Method | Links | |
|---|---|---|
| ZipformerTraining Data=ASR+ST, Training Steps=320K, Decoding Strategy=Greedy2026.01 | 7.44 | |
| ZipformerTraining Data=ASR-only, Training Steps=312K, Decoding Strategy=Greedy2026.01 | 7.48 | |
| Qwen3-ASR-1.7BParameters=1.7B2026.01 | 7.57 | |
| Qwen3-ASR-0.6BParameters=0.6B2026.01 | 9.5 | |
| GPT-4o-Transcribe2026.01 | 11.36 | |
| Doubao-ASR2026.01 | 13.2 | |
| Whisper-large-v32026.01 | 16.23 | |
| Gemini-2.5-Pro2026.01 | 18.76 | |
| FireRed-AED2026.01 | 45.79 | |
| Paraformer2026.01 | 73.23 |