120 lines
3.2 KiB
Markdown
120 lines
3.2 KiB
Markdown
# DrivelologyBinaryClassification
|
||
|
||
## 概述
|
||
|
||
Drivelology 二分类任务评估模型识别“drivelology”的能力——这是一种独特的语言现象，其特征是“有深度的胡言乱语”。这类话语在句法上连贯，但在语用层面具有悖论性、情感负载性或修辞颠覆性。
|
||
|
||
## 任务描述
|
||
|
||
- **任务类型**：二分类文本分类（是/否）
|
||
- **输入**：待分类的文本样本
|
||
- **输出**：若为 drivelology 则输出 "Yes"，否则输出 "No"
|
||
- **领域**：语言学分析、幽默检测、语用学
|
||
|
||
## 主要特点
|
||
|
||
- 测试对多层次语言含义的理解能力
|
||
- 能区分“有深度的胡言乱语”、纯粹的胡言乱语和正常文本
|
||
- 需要上下文理解与情感洞察力
|
||
- 涵盖幽默、反讽、讽刺等检测
|
||
- 提供多个难度级别
|
||
|
||
## 评估说明
|
||
|
||
- 默认配置使用 **0-shot** 评估
|
||
- 评估指标：准确率（Accuracy）、精确率（Precision）、召回率（Recall）、F1 分数（F1-Score）
|
||
- 子集：binary-english-easy、binary-english-hard、binary-chinese-easy、binary-chinese-hard
|
||
|
||
## 属性
|
||
|
||
| 属性 | 值 |
|
||
|----------|-------|
|
||
| **基准测试名称** | `drivel_binary` |
|
||
| **数据集ID** | [extraordinarylab/drivel-hub](https://modelscope.cn/datasets/extraordinarylab/drivel-hub/summary) |
|
||
| **论文** | N/A |
|
||
| **标签** | `Yes/No` |
|
||
| **指标** | `accuracy`, `precision`, `recall`, `f1_score`, `yes_ratio` |
|
||
| **默认示例数量** | 0-shot |
|
||
| **评估划分** | `test` |
|
||
| **聚合方式** | `f1` |
|
||
|
||
## 数据统计
|
||
|
||
| 指标 | 值 |
|
||
|--------|-------|
|
||
| 总样本数 | 1,200 |
|
||
| 提示词长度（平均） | 1056.08 字符 |
|
||
| 提示词长度（最小/最大） | 984 / 1449 字符 |
|
||
|
||
## 样例示例
|
||
|
||
**子集**: `binary-classification`
|
||
|
||
```json
|
||
{
|
||
"input": [
|
||
{
|
||
"id": "ddbda8da",
|
||
"content": [
|
||
{
|
||
"text": "#Instruction#:\nClassify whether the given text is a Drivelology sample or not.\n\n#Definition#:\n- Drivelology: Statements that appear logically coherent but contain deeper, often paradoxical meanings.\nThese challenge conventional interpretation ... [TRUNCATED] ... ology.\n\n#Output Format#:\nYou should try your best to answer \"Yes\" if the given input text is Drivelology, otherwise specify \"No\".\nThe answer you give MUST be \"Yes\" or \"No\"\".\n\n#Input Text#: A: Name? B: Henry. A: Age? B: E-N-R-Y.\n#Your Answer#:"
|
||
}
|
||
]
|
||
}
|
||
],
|
||
"target": "YES",
|
||
"id": 0,
|
||
"group_id": 0,
|
||
"metadata": {
|
||
"answer": "YES"
|
||
}
|
||
}
|
||
```
|
||
|
||
*注：部分内容因展示需要已被截断。*
|
||
|
||
## 提示模板
|
||
|
||
**提示模板：**
|
||
```text
|
||
{question}
|
||
```
|
||
|
||
<details>
|
||
<summary>少样本（Few-shot）模板</summary>
|
||
|
||
```text
|
||
{question}
|
||
```
|
||
|
||
</details>
|
||
|
||
## 使用方法
|
||
|
||
### 使用 CLI
|
||
|
||
```bash
|
||
evalscope eval \
|
||
--model YOUR_MODEL \
|
||
--api-url OPENAI_API_COMPAT_URL \
|
||
--api-key EMPTY_TOKEN \
|
||
--datasets drivel_binary \
|
||
--limit 10 # 正式评估时请删除此行
|
||
```
|
||
|
||
### 使用 Python
|
||
|
||
```python
|
||
from evalscope import run_task
|
||
from evalscope.config import TaskConfig
|
||
|
||
task_cfg = TaskConfig(
|
||
model='YOUR_MODEL',
|
||
api_url='OPENAI_API_COMPAT_URL',
|
||
api_key='EMPTY_TOKEN',
|
||
datasets=['drivel_binary'],
|
||
limit=10, # 正式评估时请删除此行
|
||
)
|
||
|
||
run_task(task_cfg=task_cfg)
|
||
``` |