Wendong-Fan commited on
Commit
b0af2d3
·
1 Parent(s): 3b803b8

update readme

Browse files
Files changed (2) hide show
  1. README.md +41 -20
  2. README_zh.md +18 -4
README.md CHANGED
@@ -71,13 +71,14 @@ Our vision is to revolutionize how AI agents collaborate to solve real-world tas
71
  - [**Install Dependencies**](#install-dependencies)
72
  - [**Setup Environment Variables**](#setup-environment-variables)
73
  - [**Running with Docker**](#running-with-docker)
74
-
75
  - [🚀 Quick Start](#-quick-start)
 
76
  - [🌐 Web Interface](#-web-interface)
77
  - [🧪 Experiments](#-experiments)
78
  - [⏱️ Future Plans](#️-future-plans)
79
  - [📄 License](#-license)
80
  - [🖊️ Cite](#️-cite)
 
81
  - [🔥 Community](#-community)
82
  - [❓ FAQ](#-faq)
83
  - [⭐ Star History](#-star-history)
@@ -87,7 +88,8 @@ Our vision is to revolutionize how AI agents collaborate to solve real-world tas
87
 
88
  - **[2025.03.11]**: We added MCPToolkit, FileWriteToolkit, and TerminalToolkit to enhance OWL agents with MCP tool calling, file writing capabilities, and terminal command execution.
89
  - **[2025.03.09]**: We added a web-based user interface that makes it easier to interact with the system.
90
- - **[2025.03.07]**: We open-source the codebase of 🦉 OWL project.
 
91
 
92
  # 🎬 Demo Video
93
 
@@ -106,6 +108,8 @@ https://private-user-images.githubusercontent.com/55657767/420212194-e813fc05-13
106
 
107
  # 🛠️ Installation
108
 
 
 
109
  ## Option 1: Using uv (Recommended)
110
 
111
  ```bash
@@ -253,9 +257,7 @@ For more detailed Docker usage instructions, including cross-platform support, o
253
 
254
  # 🚀 Quick Start
255
 
256
-
257
-
258
- Run the following demo case:
259
 
260
  ```bash
261
  python owl/run.py
@@ -323,13 +325,17 @@ print(f"\033[94mAnswer: {answer}\033[0m")
323
  OWL will then automatically invoke document-related tools to process the file and extract the answer.
324
 
325
 
326
- Example tasks you can try:
 
 
 
327
  - "Find the latest stock price for Apple Inc."
328
  - "Analyze the sentiment of recent tweets about climate change"
329
  - "Help me debug this Python code: [your code here]"
330
  - "Summarize the main points from this research paper: [paper URL]"
 
331
 
332
- # 🧰 Configuring Toolkits
333
 
334
  > **Important**: Effective use of toolkits requires models with strong tool calling capabilities. For multimodal toolkits (Web, Image, Video), models must also have multimodal understanding abilities.
335
 
@@ -392,16 +398,19 @@ Selecting only necessary toolkits optimizes performance and reduces resource usa
392
 
393
  # 🌐 Web Interface
394
 
395
- OWL now includes a web-based user interface that makes it easier to interact with the system. To start the web interface, run:
 
 
396
 
397
  ```bash
 
398
  python run_app.py
399
 
400
- # For English Version
401
  python run_app_en.py
402
  ```
403
 
404
- The web interface provides the following features:
405
 
406
  - **Easy Model Selection**: Choose between different models (OpenAI, Qwen, DeepSeek, etc.)
407
  - **Environment Variable Management**: Configure your API keys and other settings directly from the UI
@@ -415,21 +424,25 @@ The web interface is built using Gradio and runs locally on your machine. No dat
415
  To reproduce OWL's GAIA benchmark score of 58.18:
416
 
417
  1. Switch to the `gaia58.18` branch:
418
- ```bash
419
- git checkout gaia58.18
420
- ```
421
 
422
- 1. Run the evaluation script:
423
- ```bash
424
- python run_gaia_roleplaying.py
425
- ```
 
 
426
 
427
  # ⏱️ Future Plans
428
 
429
- - [ ] Write a technical blog post detailing our exploration and insights in multi-agent collaboration in real-world tasks.
430
- - [ ] Enhance the toolkit ecosystem with more specialized tools for domain-specific tasks.
431
- - [ ] Develop more sophisticated agent interaction patterns and communication protocols
432
 
 
 
 
 
433
 
434
  # 📄 License
435
 
@@ -479,6 +492,14 @@ Join us for further discussions!
479
 
480
  A: If OWL determines that a task can be completed using non-browser tools (such as search or code execution), the browser will not be launched. The browser window will only appear when OWL determines that browser-based interaction is necessary.
481
 
 
 
 
 
 
 
 
 
482
  # ⭐ Star History
483
 
484
  [![Star History Chart](https://api.star-history.com/svg?repos=camel-ai/owl&type=Date)](https://star-history.com/#camel-ai/owl&Date)
 
71
  - [**Install Dependencies**](#install-dependencies)
72
  - [**Setup Environment Variables**](#setup-environment-variables)
73
  - [**Running with Docker**](#running-with-docker)
 
74
  - [🚀 Quick Start](#-quick-start)
75
+ - [🧰 Toolkits and Capabilities](#-toolkits-and-capabilities)
76
  - [🌐 Web Interface](#-web-interface)
77
  - [🧪 Experiments](#-experiments)
78
  - [⏱️ Future Plans](#️-future-plans)
79
  - [📄 License](#-license)
80
  - [🖊️ Cite](#️-cite)
81
+ - [🤝 Contributing](#-contributing)
82
  - [🔥 Community](#-community)
83
  - [❓ FAQ](#-faq)
84
  - [⭐ Star History](#-star-history)
 
88
 
89
  - **[2025.03.11]**: We added MCPToolkit, FileWriteToolkit, and TerminalToolkit to enhance OWL agents with MCP tool calling, file writing capabilities, and terminal command execution.
90
  - **[2025.03.09]**: We added a web-based user interface that makes it easier to interact with the system.
91
+ - **[2025.03.07]**: We open-sourced the codebase of the 🦉 OWL project.
92
+ - **[2025.03.03]**: OWL achieved the #1 position among open-source frameworks on the GAIA benchmark with a score of 58.18.
93
 
94
  # 🎬 Demo Video
95
 
 
108
 
109
  # 🛠️ Installation
110
 
111
+ OWL supports multiple installation methods to fit your workflow preferences. Choose the option that works best for you.
112
+
113
  ## Option 1: Using uv (Recommended)
114
 
115
  ```bash
 
257
 
258
  # 🚀 Quick Start
259
 
260
+ After installation and setting up your environment variables, you can start using OWL right away:
 
 
261
 
262
  ```bash
263
  python owl/run.py
 
325
  OWL will then automatically invoke document-related tools to process the file and extract the answer.
326
 
327
 
328
+ ### Example Tasks
329
+
330
+ Here are some tasks you can try with OWL:
331
+
332
  - "Find the latest stock price for Apple Inc."
333
  - "Analyze the sentiment of recent tweets about climate change"
334
  - "Help me debug this Python code: [your code here]"
335
  - "Summarize the main points from this research paper: [paper URL]"
336
+ - "Create a data visualization for this dataset: [dataset path]"
337
 
338
+ # 🧰 Toolkits and Capabilities
339
 
340
  > **Important**: Effective use of toolkits requires models with strong tool calling capabilities. For multimodal toolkits (Web, Image, Video), models must also have multimodal understanding abilities.
341
 
 
398
 
399
  # 🌐 Web Interface
400
 
401
+ OWL includes an intuitive web-based user interface that makes it easier to interact with the system.
402
+
403
+ ## Starting the Web UI
404
 
405
  ```bash
406
+ # Start the Chinese version
407
  python run_app.py
408
 
409
+ # Start the English version
410
  python run_app_en.py
411
  ```
412
 
413
+ ## Features
414
 
415
  - **Easy Model Selection**: Choose between different models (OpenAI, Qwen, DeepSeek, etc.)
416
  - **Environment Variable Management**: Configure your API keys and other settings directly from the UI
 
424
  To reproduce OWL's GAIA benchmark score of 58.18:
425
 
426
  1. Switch to the `gaia58.18` branch:
427
+ ```bash
428
+ git checkout gaia58.18
429
+ ```
430
 
431
+ 2. Run the evaluation script:
432
+ ```bash
433
+ python run_gaia_roleplaying.py
434
+ ```
435
+
436
+ This will execute the same configuration that achieved our top-ranking performance on the GAIA benchmark.
437
 
438
  # ⏱️ Future Plans
439
 
440
+ We're continuously working to improve OWL. Here's what's on our roadmap:
 
 
441
 
442
+ - [ ] Write a technical blog post detailing our exploration and insights in multi-agent collaboration in real-world tasks
443
+ - [ ] Enhance the toolkit ecosystem with more specialized tools for domain-specific tasks
444
+ - [ ] Develop more sophisticated agent interaction patterns and communication protocols
445
+ - [ ] Improve performance on complex multi-step reasoning tasks
446
 
447
  # 📄 License
448
 
 
492
 
493
  A: If OWL determines that a task can be completed using non-browser tools (such as search or code execution), the browser will not be launched. The browser window will only appear when OWL determines that browser-based interaction is necessary.
494
 
495
+ **Q: Which Python version should I use?**
496
+
497
+ A: OWL supports Python 3.10, 3.11, and 3.12.
498
+
499
+ **Q: How can I contribute to the project?**
500
+
501
+ A: See our [Contributing](#-contributing) section for details on how to get involved. We welcome contributions of all kinds, from code improvements to documentation updates.
502
+
503
  # ⭐ Star History
504
 
505
  [![Star History Chart](https://api.star-history.com/svg?repos=camel-ai/owl&type=Date)](https://star-history.com/#camel-ai/owl&Date)
README_zh.md CHANGED
@@ -73,13 +73,16 @@
73
  - [**设置环境变量**](#设置环境变量)
74
  - [**使用Docker运行**](#使用docker运行)
75
  - [🚀 快速开始](#-快速开始)
 
76
  - [🌐 网页界面](#-网页界面)
77
  - [🧪 实验](#-实验)
78
  - [⏱️ 未来计划](#️-未来计划)
79
  - [📄 许可证](#-许可证)
80
  - [🖊️ 引用](#️-引用)
 
81
  - [🔥 社区](#-社区)
82
  - [❓ 常见问题](#-常见问题)
 
83
 
84
 
85
  # 🔥 新闻
@@ -87,6 +90,7 @@
87
  - **[2025.03.11]**: 我们添加了 MCPToolkit、FileWriteToolkit 和 TerminalToolkit,增强 OWL Agent的工具调用、文件写入能力和终端命令执行功能。
88
  - **[2025.03.09]**: 我们添加了基于网页的用户界面,使系统交互变得更加简便。
89
  - **[2025.03.07]**: 我们开源了 🦉 OWL 项目的代码库。
 
90
 
91
  # 🎬 演示视频
92
 
@@ -325,7 +329,7 @@ OWL 将自动调用与文档相关的工具来处理文件并提取答案。
325
  - "帮我调试这段 Python 代码:[在此粘贴你的代码]"
326
  - "总结这篇研究论文的主要观点:[论文URL]"
327
 
328
- # 🧰 配置工具包
329
 
330
  > **重要提示**:有效使用工具包需要具备强大工具调用能力的模型。对于多模态工具包(Web、图像、视频),模型还必须具备多模态理解能力。
331
 
@@ -423,10 +427,12 @@ python run_gaia_roleplaying.py
423
 
424
  # ⏱️ 未来计划
425
 
426
- - [ ] 撰写一篇技术博客,详细介绍我们在现实任务中多智能体协作方面的探索与见解。
427
- - [ ] 通过引入更多针对特定领域任务的专业工具,进一步完善工具生态系统。
428
- - [ ] 开发更复杂的智能体交互模式和通信协议
429
 
 
 
 
 
430
 
431
  # 📄 许可证
432
 
@@ -476,6 +482,14 @@ python run_gaia_roleplaying.py
476
 
477
  A: 当OWL判断某个任务可以使用非浏览器工具(如搜索、代码分析等)完成时,浏览器就不会启动。只有在判断需要使用浏览器工具的时候,本地才会弹出浏览器窗口,并进行浏览器模拟交互。
478
 
 
 
 
 
 
 
 
 
479
  # ⭐ Star History
480
 
481
  [![Star History Chart](https://api.star-history.com/svg?repos=camel-ai/owl&type=Date)](https://star-history.com/#camel-ai/owl&Date)
 
73
  - [**设置环境变量**](#设置环境变量)
74
  - [**使用Docker运行**](#使用docker运行)
75
  - [🚀 快速开始](#-快速开始)
76
+ - [🧰 工具包与功能](#-工具包与功能)
77
  - [🌐 网页界面](#-网页界面)
78
  - [🧪 实验](#-实验)
79
  - [⏱️ 未来计划](#️-未来计划)
80
  - [📄 许可证](#-许可证)
81
  - [🖊️ 引用](#️-引用)
82
+ - [🤝 贡献](#-贡献)
83
  - [🔥 社区](#-社区)
84
  - [❓ 常见问题](#-常见问题)
85
+ - [⭐ Star History](#-star-history)
86
 
87
 
88
  # 🔥 新闻
 
90
  - **[2025.03.11]**: 我们添加了 MCPToolkit、FileWriteToolkit 和 TerminalToolkit,增强 OWL Agent的工具调用、文件写入能力和终端命令执行功能。
91
  - **[2025.03.09]**: 我们添加了基于网页的用户界面,使系统交互变得更加简便。
92
  - **[2025.03.07]**: 我们开源了 🦉 OWL 项目的代码库。
93
+ - **[2025.03.03]**: OWL 在 GAIA 基准测试中取得 58.18 平均分,在开源框架中排名第一!
94
 
95
  # 🎬 演示视频
96
 
 
329
  - "帮我调试这段 Python 代码:[在此粘贴你的代码]"
330
  - "总结这篇研究论文的主要观点:[论文URL]"
331
 
332
+ # 🧰 工具包与功能
333
 
334
  > **重要提示**:有效使用工具包需要具备强大工具调用能力的模型。对于多模态工具包(Web、图像、视频),模型还必须具备多模态理解能力。
335
 
 
427
 
428
  # ⏱️ 未来计划
429
 
430
+ 我们正在不断努力改进 OWL。以下是我们的路线图:
 
 
431
 
432
+ - [ ] 撰写技术博客,详细介绍我们在现实任务中多智能体协作方面的探索与见解
433
+ - [ ] 通过引入更多针对特定领域任务的专业工具,进一步完善工具生态系统
434
+ - [ ] 开发更复杂的智能体交互模式和通信协议
435
+ - [ ] 提高复杂多步推理任务的性能
436
 
437
  # 📄 许可证
438
 
 
482
 
483
  A: 当OWL判断某个任务可以使用非浏览器工具(如搜索、代码分析等)完成时,浏览器就不会启动。只有在判断需要使用浏览器工具的时候,本地才会弹出浏览器窗口,并进行浏览器模拟交互。
484
 
485
+ **Q: 我应该使用哪个Python版本?**
486
+
487
+ A: OWL支持Python 3.10、3.11和3.12。为了与所有依赖项获得最佳兼容性,我们推荐使用Python 3.10。
488
+
489
+ **Q: 我如何为项目做贡献?**
490
+
491
+ A: 请参阅我们的[贡献](#-贡献)部分,了解如何参与的详细信息。我们欢迎各种形式的贡献,从代码改进到文档更新。
492
+
493
  # ⭐ Star History
494
 
495
  [![Star History Chart](https://api.star-history.com/svg?repos=camel-ai/owl&type=Date)](https://star-history.com/#camel-ai/owl&Date)