加载中...
加载中...
本页面提供当前主流大模型在代码能力上的评测结果，包括HumanEval和MBPP等基准数据集。
Top Model
Phi-3-mini 3.8B
Top Score
-
Model Count
17
Data version
-
Data source: 论文或GitHub评测结果
| Model | Parameters | HumanEval Pass@1 | MBPP Pass@1 | Organization | License |
|---|---|---|---|---|---|
| Phi-3-mini 3.8B | 38.0 | 58.50 | 70 | Microsoft Azure | / |
| Phi-1 | 13.0 | 50.60 | 55.50 | Microsoft Azure | / |
| MiniCPM-2B-DPO |
Data is for reference only. Official sources are authoritative. Click model names to view DataLearner model profiles.
| 24.0 |
| 50 |
| 47.31 |
| 面壁智能 |
| / |
| Phi-2 | 27.0 | 48.30 | 59.10 | Microsoft Azure | / |
| Qwen2.5-3B | 30.0 | 42.10 | 57.10 | 阿里巴巴 | / |
| Qwen2.5-1.5B | 15.0 | 37.20 | 60.20 | 阿里巴巴 | / |
| Stable LM Zephyr 3B | 30.0 | 35.37 | 31.85 | Stability AI | / |
| Phi-1.5 | 13.0 | 34.10 | 37.70 | Microsoft Azure | / |
| Qwen2-1.5B | 15.0 | 31.10 | 37.40 | 阿里巴巴 | / |
| Qwen2.5-0.5B | 5.0 | 30.50 | 39.30 | 阿里巴巴 | / |
| Gemma 2B | 20.0 | 22 | 29.20 | Google Research | / |
| Gemma 2B - It | 20.0 | 22 | 29.20 | Google Research | / |
| CodeGemma-2B | 20.0 | 22 | 29.20 | Google Research | / |
| Qwen2-0.5B | 4.0 | 22 | 22 | 阿里巴巴 | / |
| RecurrentGemma-2B | 27.0 | 21.30 | 28.80 | Google Research | / |
| Qwen-1.8B | 18.0 | 15.20 | / | 阿里巴巴 | / |
| TinyLlama | 11.0 | 6.71 | 19.91 | 新加坡科技与设计大学 | / |