Automatic Speech Recognition on AISHELL-2 i (iOS)
2.6WERQwen2.5-Omni
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Qwen2.5-OmniModel Size=7B2025.08 | 2.6 | 2.6 | |
| Kimi-Audio-InstructModel Size=7B2025.08 | 2.6 | 2.6 | |
| OSUM2026.03 | 2.75 | 2.75 | |
| MiDashengLMModel Size=7B2025.08 | 2.9 | 2.9 | |
| OSUM-Pangu2026.03 | 2.98 | 2.98 | |
| Qwen2-Audio2026.03 | 3 | 3 |