kth8/Qwen3.6-27B-insurance-benchmark
收藏Hugging Face2026-04-23 更新2026-04-26 收录
下载链接：
https://hf-mirror.com/datasets/kth8/Qwen3.6-27B-insurance-benchmark
下载链接
链接失效反馈官方服务：
资源简介：
---
license: apache-2.0
language:
- en
base_model: Qwen/Qwen3.6-27B
datasets:
- kth8/insurance
---
Benchmark of [Qwen/Qwen3.6-27B](https://huggingface.co/Qwen/Qwen3.6-27B) against [kth8/insurance](https://huggingface.co/datasets/kth8/insurance) dataset.
Accuracy: 92.0%.
| Metric | Value |
|----------------------|---------------|
| **Correct** | 46 |
| **Incorrect** | 4 |
| **Errors** | 0 |
| **Total samples** | 50 |
| **Total completion tokens** | 57,899 |
Raw stats:
```json
{
"accuracy": 0.92,
"correct": 46,
"incorrect": 4,
"error": 0,
"total": 50,
"python_tool_calls": 0,
"completion_tokens": 57899
}
```
---
许可证：Apache-2.0
语言：
- 英语
基础模型：Qwen/Qwen3.6-27B
关联数据集：
- kth8/insurance
---
本文件为针对[Qwen/Qwen3.6-27B](https://huggingface.co/Qwen/Qwen3.6-27B)与[kth8/insurance](https://huggingface.co/datasets/kth8/insurance)数据集的基准测试报告。
本次测试准确率为92.0%，详细指标如下表所示：
| 指标(Metric) | 数值(Value) |
|----------------------------|---------------------|
| **正确样本数(Correct)** | 46 |
| **错误样本数(Incorrect)** | 4 |
| **异常数(Errors)** | 0 |
| **总样本数(Total samples)** | 50 |
| **总补全Token(Token)数** | 57,899 |
原始统计数据如下：
json
{
"accuracy": 0.92,
"correct": 46,
"incorrect": 4,
"error": 0,
"total": 50,
"python_tool_calls": 0,
"completion_tokens": 57899
}
提供机构：
kth8
