DeepSeek R1 Distill Qwen 14B
开源来自 DeepSeek · 发布于 2025-01-20
56.0
平均分
N/A
输入价格
N/A
输出价格
N/A
上下文窗口
text-generation
类型
Tested on 11 benchmarks with 56.0% average. Top scores: JCommonsenseQA (93.7%), JSQuAD (89.8%), JNLI (82.4%).
基准测试分数
| 基准测试 | 类别 | 分数 | Bar |
|---|---|---|---|
| JCommonsenseQA | language | 93.7 | |
| JSQuAD | language | 89.8 | |
| JNLI | language | 82.4 | |
| JMMLU | language | 63.4 | |
| MATH Level 5 | math | 57.0 | |
| LLM-JP — Overall | language | 56.8 | |
| IFEval | language | 43.8 | |
| MMLU-PRO | knowledge | 40.7 | |
| BBH (HuggingFace) | general | 40.7 | |
| MUSR | reasoning | 28.7 | |
| GPQA | knowledge | 18.3 |