Code Generation on HumanEval (SpeedUp and tau)
5.71Inference SpeedUpPARD-2
Evaluation results
| Method | Links | ||
|---|---|---|---|
| PARD-2Target Model=Qwen3-8B, Inference Draft Length=16, Temperature=0, Evaluation Mode=Target-independent2026.05 | 5.71 | 7.38 | |
| PARD-2Target Model=Qwen3-32B, Inference Draft Length=16, Temperature=0, Evaluation Mode=Target-independent2026.05 | 5.71 | 7.12 | |
| PARD-2Target Model=Qwen3-14B, Inference Draft Length=16, Temperature=0, Evaluation Mode=Target-independent2026.05 | 5.63 | 7.16 | |
| PARDTarget Model=Qwen3-32B, Inference Draft Length=16, Temperature=0, Evaluation Mode=Target-independent2026.05 | 5.18 | 6.46 | |
| PARDTarget Model=Qwen3-14B, Inference Draft Length=16, Temperature=0, Evaluation Mode=Target-independent2026.05 | 5.16 | 5.25 | |
| PARDTarget Model=Qwen3-8B, Inference Draft Length=16, Temperature=0, Evaluation Mode=Target-independent2026.05 | 5.02 | 6.33 | |
| ARTarget Model=Qwen3-8B, Temperature=0, Evaluation Mode=Target-independent2026.05 | 1 | 1 | |
| ARTarget Model=Qwen3-14B, Temperature=0, Evaluation Mode=Target-independent2026.05 | 1 | 1 | |
| ARTarget Model=Qwen3-32B, Temperature=0, Evaluation Mode=Target-independent2026.05 | 1 | 1 |