Speech Recognition on CVC Kazakh (test)
19.1WER (or)Fine-tuning Community Model (KSC2)
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Fine-tuning Community Model (KSC2)Training Data=CVC + FLEURS2026.03 | 19.1 | 4.7 | 13.5 | 3.5 | |
| Fine-tuning Community Model (KSC2)Training Data=Songs + CVC2026.03 | 19.5 | 4.5 | 12.3 | 3.1 | |
| Fine-tuning Community Model (KSC2)Training Data=Songs + CVC + FLEURS2026.03 | 21 | 5.2 | 15.6 | 4.1 | |
| Fine-tuning Community Model (KSC2)Training Data=FLEURS only2026.03 | 26.4 | 6.1 | 14.4 | 3.7 | |
| Fine-tuning Community Model (KSC2)Training Data=Songs + FLEURS2026.03 | 26.8 | 7.9 | 17.1 | 6 | |
| Fine-tuning Whisper Large-V3 TurboTraining Data=Songs + CVC + FLEURS2026.03 | 32 | 7.4 | 27.6 | 6.5 | |
| Fine-tuning Whisper Large-V3 TurboTraining Data=CVC + FLEURS2026.03 | 33 | 7.7 | 28.1 | 6.6 | |
| Fine-tuning Whisper Large-V3 TurboTraining Data=Songs + CVC2026.03 | 33.9 | 8.7 | 29.6 | 7.8 | |
| Fine-tuning Whisper Large-V3 TurboTraining Data=Songs + FLEURS2026.03 | 40.9 | 11 | 34.1 | 9.5 | |
| Fine-tuning Community Model (KSC2)Training Data=Songs only2026.03 | 41.1 | 8.5 | 13.9 | 3.5 | |
| Fine-tuning Whisper Large-V3 TurboTraining Data=CVC only2026.03 | 42.3 | 10 | 39.1 | 9.2 | |
| Fine-tuning Whisper Large-V3 TurboTraining Data=Songs only2026.03 | 49.8 | 12.3 | 37.3 | 9.3 | |
| Fine-tuning Whisper Large-V3 TurboTraining Data=FLEURS only2026.03 | 51.6 | 14.8 | 43.6 | 12.8 | |
| Community fine-tuned (KSC2)Training=Baseline2026.03 | 56.2 | 10.9 | 12.5 | 3.1 | |
| Fine-tuning Community Model (KSC2)Training Data=CVC only2026.03 | 56.3 | 10.9 | 12.6 | 3.1 | |
| Whisper Large-V3Mode=zero-shot2026.03 | 61.2 | 22 | 56.5 | 20.7 | |
| Whisper Large-V3 TurboMode=zero-shot2026.03 | 70.6 | 28.6 | 47.7 | 23.8 |