154 lines
4.4 KiB
Markdown
154 lines
4.4 KiB
Markdown
# MMStar
|
||
|
||
## 概述
|
||
|
||
MMStar 是一个精英级的视觉不可或缺型多模态基准测试，旨在确保评估过程中真正依赖视觉信息。每个样本都经过精心筛选，必须依靠实际的视觉理解才能作答，从而最大限度地减少数据泄露，并测试高级多模态能力。
|
||
|
||
## 任务描述
|
||
|
||
- **任务类型**：视觉依赖型多项选择问答（Vision-Dependent Multiple-Choice QA）
|
||
- **输入**：图像 + 需要视觉理解的多项选择题
|
||
- **输出**：单个答案字母（A/B/C/D）
|
||
- **领域**：感知、推理、数学、科学技术
|
||
|
||
## 核心特性
|
||
|
||
- 确保视觉依赖性 —— 无图像则无法回答问题
|
||
- 最小化训练语料中的数据泄露
|
||
- 测试高级多模态推理能力
|
||
- 六大类别：粗粒度感知、细粒度感知、实例推理、逻辑推理、数学、科学技术
|
||
- 高质量人工筛选样本，经验证确需视觉信息
|
||
|
||
## 评估说明
|
||
|
||
- 默认使用 **val** 划分进行评估
|
||
- 主要指标：多项选择题的 **准确率（Accuracy）**
|
||
- 使用思维链（Chain-of-Thought, CoT）提示，格式为 "ANSWER: [LETTER]"
|
||
- 结果按类别和整体分别报告
|
||
|
||
## 属性
|
||
|
||
| 属性 | 值 |
|
||
|----------|-------|
|
||
| **基准测试名称** | `mm_star` |
|
||
| **数据集ID** | [evalscope/MMStar](https://modelscope.cn/datasets/evalscope/MMStar/summary) |
|
||
| **论文** | N/A |
|
||
| **标签** | `Knowledge`, `MCQ`, `MultiModal` |
|
||
| **指标** | `acc` |
|
||
| **默认示例数（Shots）** | 0-shot |
|
||
| **评估划分** | `val` |
|
||
|
||
## 数据统计
|
||
|
||
| 指标 | 值 |
|
||
|--------|-------|
|
||
| 总样本数 | 1,500 |
|
||
| 提示词长度（平均） | 390.23 字符 |
|
||
| 提示词长度（最小/最大） | 272 / 2023 字符 |
|
||
|
||
**各子集统计数据：**
|
||
|
||
| 子集 | 样本数 | 提示词平均长度 | 提示词最小长度 | 提示词最大长度 |
|
||
|--------|---------|-------------|------------|------------|
|
||
| `coarse perception` | 250 | 350.8 | 282 | 784 |
|
||
| `fine-grained perception` | 250 | 334.98 | 277 | 608 |
|
||
| `instance reasoning` | 250 | 379.02 | 273 | 684 |
|
||
| `logical reasoning` | 250 | 427.22 | 284 | 2023 |
|
||
| `math` | 250 | 467.36 | 292 | 891 |
|
||
| `science & technology` | 250 | 381.98 | 272 | 1173 |
|
||
|
||
**图像统计数据：**
|
||
|
||
| 指标 | 值 |
|
||
|--------|-------|
|
||
| 总图像数 | 1,500 |
|
||
| 每样本图像数 | 最小: 1, 最大: 1, 平均: 1 |
|
||
| 分辨率范围 | 114x66 - 3160x2136 |
|
||
| 格式 | jpeg |
|
||
|
||
## 样例示例
|
||
|
||
**子集**: `coarse perception`
|
||
|
||
```json
|
||
{
|
||
"input": [
|
||
{
|
||
"id": "57e256b8",
|
||
"content": [
|
||
{
|
||
"text": "Answer the following multiple choice question.\nThe last line of your response should be of the following format:\n'ANSWER: [LETTER]' (without quotes)\nwhere [LETTER] is one of A,B,C,D. Think step by step before answering.\n\nWhich option describe the object relationship in the image correctly?\nOptions: A: The suitcase is on the book., B: The suitcase is beneath the cat., C: The suitcase is beneath the bed., D: The suitcase is beneath the book."
|
||
},
|
||
{
|
||
"image": "[BASE64_IMAGE: jpeg, ~37.2KB]"
|
||
}
|
||
]
|
||
}
|
||
],
|
||
"choices": [
|
||
"A",
|
||
"B",
|
||
"C",
|
||
"D"
|
||
],
|
||
"target": "A",
|
||
"id": 0,
|
||
"group_id": 0,
|
||
"subset_key": "coarse perception",
|
||
"metadata": {
|
||
"index": 0,
|
||
"category": "coarse perception",
|
||
"l2_category": "image scene and topic",
|
||
"source": "MMBench",
|
||
"split": "val",
|
||
"image_path": "images/0.jpg"
|
||
}
|
||
}
|
||
```
|
||
|
||
## 提示模板
|
||
|
||
**提示模板：**
|
||
```text
|
||
Answer the following multiple choice question.
|
||
The last line of your response should be of the following format:
|
||
'ANSWER: [LETTER]' (without quotes)
|
||
where [LETTER] is one of A,B,C,D. Think step by step before answering.
|
||
|
||
{question}
|
||
```
|
||
|
||
## 使用方法
|
||
|
||
### 使用命令行（CLI）
|
||
|
||
```bash
|
||
evalscope eval \
|
||
--model YOUR_MODEL \
|
||
--api-url OPENAI_API_COMPAT_URL \
|
||
--api-key EMPTY_TOKEN \
|
||
--datasets mm_star \
|
||
--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=['mm_star'],
|
||
dataset_args={
|
||
'mm_star': {
|
||
# subset_list: ['coarse perception', 'fine-grained perception', 'instance reasoning'] # 可选，用于评估特定子集
|
||
}
|
||
},
|
||
limit=10, # 正式评估时请删除此行
|
||
)
|
||
|
||
run_task(task_cfg=task_cfg)
|
||
``` |