Code Generation on HumanEval (Normalized performance (%))
102.47Normalized Performance (%)ACE
Evaluation Results
| Method | Links | |
|---|---|---|
| ACEBackbone=LLaMA-32026.03 | 102.47 | |
| AverageBackbone=LLaMA-32026.03 | 93.82 | |
| Task ArithmeticBackbone=LLaMA-32026.03 | 93.82 |
Loading the SOTA2 catalog…