Transcript Alignment on Common Voice English 8 (test)
58.9Character GLELWA
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| LWABackbone Model=Whisper (v3, 1.6B)2025.09 | 58.9 | 54.2 | |
| LWABackbone Model=Parakeet CTC (0.6B)2025.09 | 59.7 | 55.4 | |
| LWABackbone Model=Phi-4-multimodal (5.6B)2025.09 | 59.8 | 55.2 | |
| LWABackbone Model=Parakeet TDT (v2, 0.6B)2025.09 | 60.3 | 55.5 | |
| OWABackbone Model=Whisper (v3, 1.6B)2025.09 | 65.8 | 60.6 | |
| OWABackbone Model=Parakeet CTC (0.6B)2025.09 | 65.9 | 61.3 | |
| OWABackbone Model=Phi-4-multimodal (5.6B)2025.09 | 66 | 61.2 | |
| OWABackbone Model=Parakeet TDT (v2, 0.6B)2025.09 | 66.2 | 61.1 | |
| PWRBackbone Model=Parakeet CTC (0.6B)2025.09 | 75.2 | 71.5 | |
| PWRBackbone Model=Phi-4-multimodal (5.6B)2025.09 | 76.8 | 73.2 | |
| PWRBackbone Model=Whisper (v3, 1.6B)2025.09 | 77 | 73.1 | |
| error-alignBackbone Model=Parakeet CTC (0.6B), Beam size=1002025.09 | 77 | 72.6 | |
| PWRBackbone Model=Parakeet TDT (v2, 0.6B)2025.09 | 77.9 | 73.8 | |
| error-alignBackbone Model=Phi-4-multimodal (5.6B), Beam size=1002025.09 | 78.6 | 74.2 | |
| error-alignBackbone Model=Whisper (v3, 1.6B), Beam size=1002025.09 | 78.8 | 74.2 | |
| error-alignBackbone Model=Parakeet TDT (v2, 0.6B), Beam size=1002025.09 | 79.5 | 74.7 |