Automatic Speech Recognition on In-house ASR Nutrition domain (evaluation set)
6.63EERQwen3-Omni-30B-A3B-Instruct
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-Omni-30B-A3B-InstructLLM=Qwen3, Size=30B2026.04 | 6.63 | |
| CoT-ASRLLM=Phi4, Size=3.8B2026.04 | 6.66 | |
| Whisper Large-v3LLM=Whisper, Size=1.5B2026.04 | 7.13 | |
| Phi4MM-style BaselineLLM=Phi4, Size=3.8B2026.04 | 7.36 | |
| Voxtral-Mini-3B-2507LLM=Voxtral, Size=3B2026.04 | 7.36 | |
| Gemma 3n E4B ITLLM=Gemma 3n, Size=8B2026.04 | 11.77 | |
| Qwen2.5-Omni-7BLLM=Qwen2.5, Size=7B2026.04 | 14.72 |