Automatic Speech Recognition on Fleurs en (test)
0.336WERUnited-MedASR
Evaluation results
| Method | Links | |
|---|---|---|
| United-MedASRlanguage=English2024.11 | 0.336 | |
| Step-Audio-2-miniType=Audio, Size=8B2026.02 | 4.51 | |
| Qwen3-Omni-InstructType=Omni, Size=30B-A3B2026.02 | 5.04 | |
| Eureka-Audio-InstructType=Ours, Size=1.7B2026.02 | 5.39 | |
| Qwen2.5-Omni-7BType=Omni, Size=7B2026.02 | 5.49 | |
| Qwen2.5-Omni-3BType=Omni, Size=3B2026.02 | 5.65 | |
| Eureka-Audio-BaseType=Ours, Size=1.7B2026.02 | 5.73 | |
| Ming-Lite-Omni-1.5Type=Omni, Size=19B-A2.8B2026.02 | 5.82 | |
| Kimi-Audio-7B-InstructType=Audio, Size=7B2026.02 | 6.11 | |
| MiniCPM-oType=Omni, Size=9B2026.02 | 6.18 | |
| Audio Flamingo 3Type=Audio, Size=8B2026.02 | 6.3 | |
| Qwen2-AudioType=Audio, Size=7B2026.02 | 6.92 | |
| Baichuan-omniParameters=7B2024.10 | 14.7 | |
| Qwen2-Audio-InstructParameters=7B2024.10 | 15.7 | |
| Whisper-large-v3Parameters=1.55B2024.10 | 17.2 | |
| SeamlessM4T Mediummodel_size=Medium2024.11 | 21.9 | |
| SeamlessM4T Largemodel_size=Large2024.11 | 23.1 |