Dialogue Evaluation on Alpaca
3.58Accepted Length (τ)DSpark
Evaluation Results
| Method | Links | |
|---|---|---|
| DSparktarget_model=Qwen3-8B2026.07 | 3.58 | |
| DSparktarget_model=Qwen3-14B2026.07 | 3.58 | |
| DSparktarget_model=Qwen3-4B2026.07 | 3.54 | |
| DSparktarget_model=Gemma4-12B2026.07 | 3.35 | |
| Eagle3target_model=Gemma4-12B2026.07 | 3.06 | |
| DFlashtarget_model=Qwen3-8B2026.07 | 2.98 | |
| DFlashtarget_model=Qwen3-4B2026.07 | 2.96 | |
| DFlashtarget_model=Qwen3-14B2026.07 | 2.94 | |
| DFlashtarget_model=Gemma4-12B2026.07 | 2.84 | |
| Eagle3target_model=Qwen3-8B2026.07 | 2.54 | |
| Eagle3target_model=Qwen3-14B2026.07 | 2.47 | |
| Eagle3target_model=Qwen3-4B2026.07 | 2.26 |