mirror of
https://github.com/open-compass/opencompass.git
synced 2025-05-30 16:03:24 +08:00
314 lines
6.0 KiB
Markdown
314 lines
6.0 KiB
Markdown
<div align="center">
|
||
<img src="docs/zh_cn/_static/image/logo.svg" width="500px"/>
|
||
<br />
|
||
<br />
|
||
|
||
[](https://opencompass.readthedocs.io/en/dev-1.x/?badge=dev-1.x)
|
||
[](https://github.com/InternLM/opencompass/blob/main/LICENSE)
|
||
[](https://pypi.org/project/opencompass/)
|
||
|
||
[🌐Website](https://opencompass.org.cn/) |
|
||
[📘Documentation](https://opencompass.readthedocs.io/en/latest/) |
|
||
[🛠️Installation](https://opencompass.readthedocs.io/en/latest/get_started/install.html) |
|
||
[🤔Reporting Issues](https://github.com/InternLM/opencompass/issues/new/choose)
|
||
|
||
[English](/README.md) | 简体中文
|
||
|
||
</div>
|
||
|
||
欢迎来到OpenCompass！
|
||
|
||
就像指南针在我们的旅程中为我们导航一样，我们希望OpenCompass能够帮助你穿越评估大型语言模型的重重迷雾。OpenCompass提供丰富的算法和功能支持，期待OpenCompass能够帮助社区更便捷地对NLP模型的性能进行公平全面的评估。
|
||
|
||
## 介绍
|
||
|
||
OpenCompass 是面向大模型评测的一站式平台。其主要特点如下：
|
||
|
||
- **开源可复现**：提供公平、公开、可复现的大模型评测方案
|
||
|
||
- **全面的能力维度**：五大维度设计，提供 50+ 个数据集约 30 万题的的模型评测方案，全面评估模型能力
|
||
|
||
- **丰富的模型支持**：已支持 20+ HuggingFace 及 API 模型
|
||
|
||
- **分布式高效评测**：一行命令实现任务分割和分布式评测，数小时即可完成千亿模型全量评测
|
||
|
||
- **多样化评测范式**：支持零样本、小样本及思维链评测，结合标准型或对话型提示词模板，轻松激发各种模型最大性能
|
||
|
||
- **灵活化拓展**：想增加新模型或数据集？想要自定义更高级的任务分割策略，甚至接入新的集群管理系统？OpenCompass 的一切均可轻松扩展！
|
||
|
||
## 性能榜单
|
||
|
||
我们将陆续提供开源模型和API模型的具体性能榜单，请见 [OpenCompass Leaderbaord](https://opencompass.org.cn/rank) 。如需加入评测，请提供模型仓库地址或标准的 API 接口至邮箱 `opencompass@pjlab.org.cn`.
|
||
|
||
|
||
|
||
## 数据集支持
|
||
|
||
<table align="center">
|
||
<tbody>
|
||
<tr align="center" valign="bottom">
|
||
<td>
|
||
<b>语言</b>
|
||
</td>
|
||
<td>
|
||
<b>知识</b>
|
||
</td>
|
||
<td>
|
||
<b>推理</b>
|
||
</td>
|
||
<td>
|
||
<b>学科</b>
|
||
</td>
|
||
<td>
|
||
<b>理解</b>
|
||
</td>
|
||
</tr>
|
||
<tr valign="top">
|
||
<td>
|
||
<details open>
|
||
<summary><b>字词释义</b></summary>
|
||
|
||
- WiC
|
||
- SummEdits
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>成语习语</b></summary>
|
||
|
||
- CHID
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>语义相似度</b></summary>
|
||
|
||
- AFQMC
|
||
- BUSTM
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>指代消解</b></summary>
|
||
|
||
- CLUEWSC
|
||
- WSC
|
||
- WinoGrande
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>翻译</b></summary>
|
||
|
||
- Flores
|
||
|
||
</details>
|
||
</td>
|
||
<td>
|
||
<details open>
|
||
<summary><b>知识问答</b></summary>
|
||
|
||
- BoolQ
|
||
- CommonSenseQA
|
||
- NaturalQuestion
|
||
- TrivialQA
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>多语种问答</b></summary>
|
||
|
||
- TyDi-QA
|
||
|
||
</details>
|
||
</td>
|
||
<td>
|
||
<details open>
|
||
<summary><b>文本蕴含</b></summary>
|
||
|
||
- CMNLI
|
||
- OCNLI
|
||
- OCNLI_FC
|
||
- AX-b
|
||
- AX-g
|
||
- CB
|
||
- RTE
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>常识推理</b></summary>
|
||
|
||
- StoryCloze
|
||
- StoryCloze-CN（即将上线）
|
||
- COPA
|
||
- ReCoRD
|
||
- HellaSwag
|
||
- PIQA
|
||
- SIQA
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>数学推理</b></summary>
|
||
|
||
- MATH
|
||
- GSM8K
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>定理应用</b></summary>
|
||
|
||
- TheoremQA
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>代码</b></summary>
|
||
|
||
- HumanEval
|
||
- MBPP
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>综合推理</b></summary>
|
||
|
||
- BBH
|
||
|
||
</details>
|
||
</td>
|
||
<td>
|
||
<details open>
|
||
<summary><b>初中/高中/大学/职业考试</b></summary>
|
||
|
||
- GAOKAO-2023
|
||
- CEval
|
||
- AGIEval
|
||
- MMLU
|
||
- GAOKAO-Bench
|
||
- MMLU-CN (即将上线)
|
||
- ARC
|
||
|
||
</details>
|
||
</td>
|
||
<td>
|
||
<details open>
|
||
<summary><b>阅读理解</b></summary>
|
||
|
||
- C3
|
||
- CMRC
|
||
- DRCD
|
||
- MultiRC
|
||
- RACE
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>内容总结</b></summary>
|
||
|
||
- CSL
|
||
- LCSTS
|
||
- XSum
|
||
|
||
</details>
|
||
|
||
<details open>
|
||
<summary><b>内容分析</b></summary>
|
||
|
||
- EPRSTMT
|
||
- LAMBADA
|
||
- TNEWS
|
||
|
||
</details>
|
||
</td>
|
||
</tr>
|
||
</td>
|
||
</tr>
|
||
</tbody>
|
||
</table>
|
||
|
||
## 模型支持
|
||
|
||
<table align="center">
|
||
<tbody>
|
||
<tr align="center" valign="bottom">
|
||
<td>
|
||
<b>开源模型</b>
|
||
</td>
|
||
<td>
|
||
<b>API 模型</b>
|
||
</td>
|
||
<!-- <td>
|
||
<b>自定义模型</b>
|
||
</td> -->
|
||
</tr>
|
||
<tr valign="top">
|
||
<td>
|
||
|
||
- LLaMA
|
||
- Vicuna
|
||
- Alpaca
|
||
- Baichuan
|
||
- WizardLM
|
||
- ChatGLM-6B
|
||
- ChatGLM2-6B
|
||
- MPT
|
||
- Falcon
|
||
- TigerBot
|
||
- MOSS
|
||
- ……
|
||
|
||
</td>
|
||
<td>
|
||
|
||
- OpenAI (即将推出)
|
||
- Claude (即将推出)
|
||
- PaLM (即将推出)
|
||
- ……
|
||
|
||
</td>
|
||
<!-- <td>
|
||
|
||
- GLM
|
||
- ……
|
||
|
||
</td> -->
|
||
</tr>
|
||
</tbody>
|
||
</table>
|
||
|
||
# 安装
|
||
|
||
下面展示了快速安装的步骤。有部分第三方功能可能需要额外步骤才能正常运行，详细步骤请参考[安装指南](https://opencompass.readthedocs.io/zh_cn/latest/get_started.html)。
|
||
|
||
```Python
|
||
conda create --name opencompass python=3.8 pytorch torchvision -c pytorch -y
|
||
conda activate opencompass
|
||
git clone https://github.com/InternLM/opencompass opencompass
|
||
cd opencompass
|
||
pip install -e .
|
||
# 下载数据集到 data/ 处
|
||
# TODO: ....
|
||
```
|
||
|
||
## 评测
|
||
|
||
请阅读[快速上手](https://opencompass.readthedocs.io/zh_cn/latest/get_started.html)了解如何运行一个评测任务。
|
||
|
||
## 致谢
|
||
|
||
该项目部分的代码引用并修改自 [OpenICL](https://github.com/Shark-NLP/OpenICL)。
|
||
|
||
## 引用
|
||
|
||
```bibtex
|
||
@misc{2023opencompass,
|
||
title={OpenCompass: A Universal Evaluation Platform for Foundation Models},
|
||
author={OpenCompass Contributors},
|
||
howpublished = {\url{https://github.com/InternLM/OpenCompass}},
|
||
year={2023}
|
||
}
|
||
```
|