Code Generation on MBPP EvalPlus
84.9Pass@1Llama 3.1 Instruct
Evaluation Results
| Method | Links | |
|---|---|---|
| Llama 3.1 InstructModel Scale=70B2025.04 | 84.9 | |
| Llama 3 InstructModel Scale=70B2025.04 | 82.8 | |
| ParamΔModel Scale=70B2025.04 | 80.7 | |
| ParamΔModel Scale=8B2025.04 | 73.3 | |
| Llama 3 InstructModel Scale=8B2025.04 | 70.6 | |
| Llama 3.1 InstructModel Scale=8B2025.04 | 70.4 | |
| Llama 3.1 BaseModel Scale=70B2025.04 | 70.4 | |
| Llama 3 BaseModel Scale=70B2025.04 | 66.9 | |
| Llama 3.1 BaseModel Scale=8B2025.04 | 59.3 | |
| Llama 3 BaseModel Scale=8B2025.04 | 59 |