Automatic Speech Recognition (Ko) on Fleurs Common Voice 15
3.82CERFastSLM
Evaluation Results
| Method | Links | |
|---|---|---|
| FastSLMModel Size=4.8B, #speech tokens/30 Sec.=502026.01 | 3.82 | |
| Gemini-2.5-Flash#speech tokens/30 Sec.=9602026.01 | 4.55 | |
| WhisperModel Size=1.5B, #speech tokens/30 Sec.=15002026.01 | 7.92 |