diff --git a/README.md b/README.md index 57864c0..77b7920 100644 --- a/README.md +++ b/README.md @@ -1,11 +1,17 @@ + +> [!NOTE] +> 本文档由 WeHub 基于上游 README 翻译整理，属于社区翻译，非官方中文文档。 +> [English](./README.en.md) · [原始项目](https://github.com/microsoft/RD-Agent) · [上游 README](https://github.com/microsoft/RD-Agent/blob/HEAD/README.md) +> 原作者、版权与许可证归属以原始项目及本仓库 LICENSE 文件为准。 +

RA-Agent logo - 🖥️ Live Demo | - 🎥 Demo Video ▶️YouTube | - 📖 Documentation | - 📄 Tech Report | - 📃 Papers + 🖥️ 在线演示 | + 🎥 演示视频 ▶️YouTube | + 📖 文档 | + 📄 技术报告 | + 📃 论文

@@ -28,86 +34,86 @@ [![arXiv](https://img.shields.io/badge/arXiv-2505.14738-00ff00.svg)](https://arxiv.org/abs/2505.14738) -# 📰 News -| 🗞️ News | 📝 Description | +# 📰 新闻 +| 🗞️ 新闻 | 📝 说明 | | -- | ------ | -| ICML 2026 Acceptance | We are thrilled to announce that our paper [FT-Dojo: Towards Autonomous LLM Fine-Tuning with Language Agents](https://arxiv.org/abs/2603.01712) has been accepted to ICML 2026. The FT-Agent implementation is available in the [LLM fine-tuning guide](rdagent/app/finetune/llm/README.md). | -| ACL 2026 Findings Acceptance | We are thrilled to announce that our paper [Reasoning as Gradient](https://arxiv.org/abs/2603.01692) has been accepted to ACL 2026 Findings. Execution traces are available at [Gome GPT-5 Traces](https://huggingface.co/datasets/amstrongzyf/Gome-GPT5-Traces) | -| Web UI Release | We release a new frontend that can be built and served by `rdagent server_ui` for real-time interaction and trace viewing, currently excluding the `data_science` scenario. | -| NeurIPS 2025 Acceptance | We are thrilled to announce that our paper [R&D-Agent-Quant](https://arxiv.org/abs/2505.15155) has been accepted to NeurIPS 2025 | -| [Technical Report Release](#overall-technical-report) | Overall framework description and results on MLE-bench | -| [R&D-Agent-Quant Release](#deep-application-in-diverse-scenarios) | Apply R&D-Agent to quant trading | -| MLE-Bench Results Released | R&D-Agent currently leads as the [top-performing machine learning engineering agent](#-the-best-machine-learning-engineering-agent) on MLE-bench | -| Support LiteLLM Backend | We now fully support **[LiteLLM](https://github.com/BerriAI/litellm)** as our default backend for integration with multiple LLM providers. | -| General Data Science Agent | [Data Science Agent](https://rdagent.readthedocs.io/en/latest/scens/data_science.html) | -| Kaggle Scenario release | We release **[Kaggle Agent](https://rdagent.readthedocs.io/en/latest/scens/data_science.html)**, try the new features! | -| Official WeChat group release | We created a WeChat group, welcome to join! (🗪[QR Code](https://github.com/microsoft/RD-Agent/issues/880)) | -| Official Discord release | We launch our first chatting channel in Discord (🗪[![Chat](https://img.shields.io/badge/chat-discord-blue)](https://discord.gg/ybQ97B6Jjy)) | -| First release | **R&D-Agent** is released on GitHub | +| ICML 2026 录用 | 我们激动地宣布，论文 [FT-Dojo: Towards Autonomous LLM Fine-Tuning with Language Agents](https://arxiv.org/abs/2603.01712) 已被 ICML 2026 录用。FT-Agent 实现可在 [LLM 微调指南](rdagent/app/finetune/llm/README.md) 中获取。 | +| ACL 2026 Findings 录用 | 我们激动地宣布，论文 [Reasoning as Gradient](https://arxiv.org/abs/2603.01692) 已被 ACL 2026 Findings 录用。执行轨迹可在 [Gome GPT-5 Traces](https://huggingface.co/datasets/amstrongzyf/Gome-GPT5-Traces) 查看。 | +| Web UI 发布 | 我们发布了新的前端，可通过 `rdagent server_ui` 构建并提供服务，用于实时交互与轨迹查看，目前不包含 `data_science` 场景。 | +| NeurIPS 2025 录用 | 我们激动地宣布，论文 [R&D-Agent-Quant](https://arxiv.org/abs/2505.15155) 已被 NeurIPS 2025 录用 | +| [技术报告发布](#overall-technical-report) | MLE-bench 上的整体框架描述与结果 | +| [R&D-Agent-Quant 发布](#deep-application-in-diverse-scenarios) | 将 R&D-Agent 应用于量化交易 | +| MLE-Bench 结果发布 | R&D-Agent 目前在 MLE-bench 上领先，是[表现最佳的机器学习工程智能体](#-the-best-machine-learning-engineering-agent) | +| 支持 LiteLLM 后端 | 我们现已全面支持 **[LiteLLM](https://github.com/BerriAI/litellm)** 作为默认后端，以便与多家 LLM 提供商集成。 | +| 通用数据科学智能体 | [Data Science Agent](https://rdagent.readthedocs.io/en/latest/scens/data_science.html) | +| Kaggle 场景发布 | 我们发布了 **[Kaggle Agent](https://rdagent.readthedocs.io/en/latest/scens/data_science.html)**, 欢迎试用新功能！ | +| 官方微信群发布 | 我们创建了微信群，欢迎加入！（🗪[二维码](https://github.com/microsoft/RD-Agent/issues/880)) | +| 官方 Discord 发布 | 我们在 Discord 上开通了首个聊天频道（🗪[![Chat](https://img.shields.io/badge/chat-discord-blue)](https://discord.gg/ybQ97B6Jjy)) | +| 首次发布 | **R&D-Agent** 已在 GitHub 上发布 | -# 🏆 The Best Machine Learning Engineering Agent! +# 🏆 最佳机器学习工程智能体！ -[MLE-bench](https://github.com/openai/mle-bench) is a comprehensive benchmark evaluating the performance of AI agents on machine learning engineering tasks. Utilizing datasets from 75 Kaggle competitions, MLE-bench provides robust assessments of AI systems' capabilities in real-world ML engineering scenarios. +[MLE-bench](https://github.com/openai/mle-bench) 是一个综合基准，用于评估 AI 智能体在机器学习工程任务上的表现。MLE-bench 利用来自 75 场 Kaggle 竞赛的数据集，对 AI 系统在真实 ML 工程场景中的能力进行稳健评估。 -R&D-Agent currently leads as the top-performing machine learning engineering agent on MLE-bench: +R&D-Agent 目前在 MLE-bench 上领先，是表现最佳的机器学习工程智能体： -| Agent | Low == Lite (%) | Medium (%) | High (%) | All (%) | +| 智能体 | Low == Lite (%) | Medium (%) | High (%) | All (%) | |---------|--------|-----------|---------|----------| | R&D-Agent o3(R)+GPT-4.1(D) | 51.52 ± 6.9 | 19.3 ± 5.5 | 26.67 ± 0 | 30.22 ± 1.5 | | R&D-Agent o1-preview | 48.18 ± 2.49 | 8.95 ± 2.36 | 18.67 ± 2.98 | 22.4 ± 1.1 | | AIDE o1-preview | 34.3 ± 2.4 | 8.8 ± 1.1 | 10.0 ± 1.9 | 16.9 ± 1.1 | -**Notes:** -- **O3(R)+GPT-4.1(D)**: This version is designed to both reduce average time per loop and leverage a cost-effective combination of backend LLMs by seamlessly integrating Research Agent (o3) with Development Agent (GPT-4.1). -- **AIDE o1-preview**: Represents the previously best public result on MLE-bench as reported in the original MLE-bench paper. -- Average and standard deviation results for R&D-Agent o1-preview is based on a independent of 5 seeds and for R&D-Agent o3(R)+GPT-4.1(D) is based on 6 seeds. -- According to MLE-Bench, the 75 competitions are categorized into three levels of complexity: **Low==Lite** if we estimate that an experienced ML engineer can produce a sensible solution in under 2 hours, excluding the time taken to train any models; **Medium** if it takes between 2 and 10 hours; and **High** if it takes more than 10 hours. +**说明：** +- **O3(R)+GPT-4.1(D)**：该版本旨在缩短每轮循环的平均耗时，并通过将 Research Agent (o3) 与 Development Agent (GPT-4.1) 无缝集成，采用更具成本效益的后端 LLM 组合。 +- **AIDE o1-preview**：代表 MLE-bench 原始论文中报告的最佳公开结果。 +- R&D-Agent o1-preview 的平均值与标准差基于 5 个独立随机种子；R&D-Agent o3(R)+GPT-4.1(D) 基于 6 个种子。 +- 根据 MLE-Bench，75 场竞赛按复杂度分为三个等级：**Low==Lite** 表示我们估计有经验的 ML 工程师可在 2 小时内（不含模型训练时间）给出合理方案；**Medium** 表示需要 2 至 10 小时；**High** 表示需要超过 10 小时。 -You can inspect the detailed runs of the above results online. -- [R&D-Agent o1-preview detailed runs](https://aka.ms/RD-Agent_MLE-Bench_O1-preview) -- [R&D-Agent o3(R)+GPT-4.1(D) detailed runs](https://aka.ms/RD-Agent_MLE-Bench_O3_GPT41) +你可以在线查看上述结果的详细运行记录。 +- [R&D-Agent o1-preview 详细运行记录](https://aka.ms/RD-Agent_MLE-Bench_O1-preview) +- [R&D-Agent o3(R)+GPT-4.1(D) 详细运行记录](https://aka.ms/RD-Agent_MLE-Bench_O3_GPT41) -For running R&D-Agent on MLE-bench, refer to **[MLE-bench Guide: Running ML Engineering via MLE-bench](https://rdagent.readthedocs.io/en/latest/scens/data_science.html)** +要在 MLE-bench 上运行 R&D-Agent，请参阅 **[MLE-bench 指南：通过 MLE-bench 运行 ML 工程](https://rdagent.readthedocs.io/en/latest/scens/data_science.html)** -# 🥇 The First Data-Centric Quant Multi-Agent Framework! +# 🥇 首个以数据为中心的多智能体量化框架！ -R&D-Agent for Quantitative Finance, in short **RD-Agent(Q)**, is the first data-centric, multi-agent framework designed to automate the full-stack research and development of quantitative strategies via coordinated factor-model co-optimization. +面向量化金融的 R&D-Agent，简称 **RD-Agent(Q)**，是首个以数据为中心的多智能体框架，旨在通过因子与模型的协同优化，自动化量化策略的全栈研发。 ![image](https://github.com/user-attachments/assets/3198bc10-47ba-4ee0-8a8e-46d5ce44f45d) -Extensive experiments in real stock markets show that, at a cost under $10, RD-Agent(Q) achieves approximately 2× higher ARR than benchmark factor libraries while using over 70% fewer factors. It also surpasses state-of-the-art deep time-series models under smaller resource budgets. Its alternating factor–model optimization further delivers excellent trade-off between predictive accuracy and strategy robustness. +在真实股票市场上的大量实验表明，RD-Agent(Q) 在成本低于 10 美元的情况下，相较基准因子库可实现约 2 倍更高的 ARR（年化收益率），同时使用的因子数量减少超过 70%。在更小的资源预算下，它也超越了最先进的深度时间序列模型。其交替进行的因子—模型优化进一步在预测精度与策略稳健性之间实现了出色权衡。 -You can learn more details about **RD-Agent(Q)** through the [paper](https://arxiv.org/abs/2505.15155) and reproduce it through the [documentation](https://rdagent.readthedocs.io/en/latest/scens/quant_agent_fin.html). +你可以通过[论文](https://arxiv.org/abs/2505.15155)了解更多关于 **RD-Agent(Q)** 的细节，并通过[文档](https://rdagent.readthedocs.io/en/latest/scens/quant_agent_fin.html).进行复现 -# Data Science Agent Preview -Check out our demo video showcasing the current progress of our Data Science Agent under development: +# Data Science Agent 预览 +查看我们的演示视频，了解正在开发中的 Data Science Agent 当前进展： https://github.com/user-attachments/assets/3eccbecb-34a4-4c81-bce4-d3f8862f7305 -# 🌟 Introduction +# 🌟 简介
Our focused scenario
-R&D-Agent aims to automate the most critical and valuable aspects of the industrial R&D process, and we begin with focusing on the data-driven scenarios to streamline the development of models and data. -Methodologically, we have identified a framework with two key components: 'R' for proposing new ideas and 'D' for implementing them. -We believe that the automatic evolution of R&D will lead to solutions of significant industrial value. +R&D-Agent 旨在自动化工业研发（R&D）流程中最关键、最有价值的环节，我们从聚焦数据驱动场景入手，以简化模型与数据的开发流程。 +在方法论上，我们确定了一个包含两个关键组件的框架：「R」用于提出新想法，「D」用于实现这些想法。 +我们相信研发的自动演化将带来具有重大工业价值的解决方案。 -R&D is a very general scenario. The advent of R&D-Agent can be your -- 💰 **Automatic Quant Factory** ([🎥Demo Video](https://rdagent.azurewebsites.net/factor_loop)|[▶️YouTube](https://www.youtube.com/watch?v=X4DK2QZKaKY&t=6s)) -- 🤖 **Data Mining Agent:** Iteratively proposing data & models ([🎥Demo Video 1](https://rdagent.azurewebsites.net/model_loop)|[▶️YouTube](https://www.youtube.com/watch?v=dm0dWL49Bc0&t=104s)) ([🎥Demo Video 2](https://rdagent.azurewebsites.net/dmm)|[▶️YouTube](https://www.youtube.com/watch?v=VIaSTZuoZg4)) and implementing them by gaining knowledge from data. -- 🦾 **Research Copilot:** Auto read research papers ([🎥Demo Video](https://rdagent.azurewebsites.net/report_model)|[▶️YouTube](https://www.youtube.com/watch?v=BiA2SfdKQ7o)) / financial reports ([🎥Demo Video](https://rdagent.azurewebsites.net/report_factor)|[▶️YouTube](https://www.youtube.com/watch?v=ECLTXVcSx-c)) and implement model structures or building datasets. -- 🤖 **Kaggle Agent:** Auto Model Tuning and Feature Engineering([🎥Demo Video Coming Soon...]()) and implementing them to achieve more in competitions. -- 🧪 **FT-Agent:** Autonomous LLM fine-tuning for benchmark-driven domain adaptation. See the [LLM fine-tuning guide](rdagent/app/finetune/llm/README.md). +研发是一个非常通用的场景。R&D-Agent 的出现可以助你实现： +- 💰 **Automatic Quant Factory（自动量化工厂）** ([🎥Demo Video](https://rdagent.azurewebsites.net/factor_loop)|[▶️YouTube](https://www.youtube.com/watch?v=X4DK2QZKaKY&t=6s)) +- 🤖 **Data Mining Agent（数据挖掘智能体）：** 迭代式提出数据与模型方案 ([🎥Demo Video 1](https://rdagent.azurewebsites.net/model_loop)|[▶️YouTube](https://www.youtube.com/watch?v=dm0dWL49Bc0&t=104s)) ([🎥Demo Video 2](https://rdagent.azurewebsites.net/dmm)|[▶️YouTube](https://www.youtube.com/watch?v=VIaSTZuoZg4))，并通过从数据中获取知识来实现它们。 +- 🦾 **Research Copilot（研究副驾驶）：** 自动阅读研究论文 ([🎥Demo Video](https://rdagent.azurewebsites.net/report_model)|[▶️YouTube](https://www.youtube.com/watch?v=BiA2SfdKQ7o)) / 财务报告 ([🎥Demo Video](https://rdagent.azurewebsites.net/report_factor)|[▶️YouTube](https://www.youtube.com/watch?v=ECLTXVcSx-c))，并实现模型结构或构建数据集。 +- 🤖 **Kaggle Agent：** 自动模型调优与特征工程 ([🎥Demo Video Coming Soon...]())，并在竞赛中实现它们以取得更好成绩。 +- 🧪 **FT-Agent：** 面向基准测试驱动的领域适配的自主 LLM 微调。参见 [LLM fine-tuning guide](rdagent/app/finetune/llm/README.md)。 - ... -You can click the links above to view the demo. We're continuously adding more methods and scenarios to the project to enhance your R&D processes and boost productivity. +你可以点击上方链接观看演示。我们持续为项目添加更多方法与场景，以增强你的研发流程并提升生产力。 -Additionally, you can take a closer look at the examples in our **[🖥️ Live Demo](https://rdagent.azurewebsites.net/)**. +此外，你还可以更仔细地查看我们 **[🖥️ Live Demo](https://rdagent.azurewebsites.net/)**. 中的示例
@@ -116,68 +122,68 @@ Additionally, you can take a closer look at the examples in our **[🖥️ Live
-# ⚡ Quick start +# ⚡ 快速开始 -### RD-Agent currently only supports Linux. +### RD-Agent 目前仅支持 Linux。 -You can try above demos by running the following command: +你可以通过运行以下命令来试用上述演示： -### 🐳 Docker installation. -Users must ensure Docker is installed before attempting most scenarios. Please refer to the [official 🐳Docker page](https://docs.docker.com/engine/install/) for installation instructions. -Ensure the current user can run Docker commands **without using sudo**. You can verify this by executing `docker run hello-world`. +### 🐳 Docker 安装 +用户在尝试大多数场景之前，必须确保已安装 Docker。安装说明请参阅[官方 🐳Docker 页面](https://docs.docker.com/engine/install/)。 +确保当前用户可以在**不使用 sudo** 的情况下运行 Docker 命令。你可以通过执行 `docker run hello-world` 来验证这一点。 -### 🐍 Create a Conda Environment -- Create a new conda environment with Python (3.10 and 3.11 are well-tested in our CI): +### 🐍 创建 Conda 环境 +- 使用 Python 创建新的 conda 环境（我们的 CI 中已充分测试 3.10 和 3.11 版本）： ```sh conda create -n rdagent python=3.10 ``` -- Activate the environment: +- 激活环境： ```sh conda activate rdagent ``` -### 🛠️ Install the R&D-Agent +### 🛠️ 安装 R&D-Agent -#### For Users -- You can directly install the R&D-Agent package from PyPI: +#### 面向用户 +- 你可以直接从 PyPI 安装 R&D-Agent 包： ```sh pip install rdagent ``` -#### For Developers -- If you want to try the latest version or contribute to RD-Agent, you can install it from the source and follow the development setup: +#### 面向开发者 +- 如果你想试用最新版本或为 RD-Agent 做贡献，可以从源码安装并按照开发环境配置进行设置： ```sh git clone https://github.com/microsoft/RD-Agent cd RD-Agent make dev ``` -More details can be found in the [development setup](https://rdagent.readthedocs.io/en/latest/development.html). +更多细节请参阅[开发环境配置](https://rdagent.readthedocs.io/en/latest/development.html). -### 💊 Health check -- rdagent provides a health check that currently checks two things. - - whether the docker installation was successful. - - whether the default port used by the [rdagent ui](https://github.com/microsoft/RD-Agent?tab=readme-ov-file#%EF%B8%8F-monitor-the-application-results) is occupied. +### 💊 健康检查 +- rdagent 提供健康检查，目前会检查两项内容。 + - docker 安装是否成功。 + - [rdagent ui](https://github.com/microsoft/RD-Agent?tab=readme-ov-file#%EF%B8%8F-monitor-the-application-results) 使用的默认端口是否被占用。 ```sh rdagent health_check --no-check-env ``` -### ⚙️ Configuration -- The demos requires following ability: +### ⚙️ 配置 +- 演示需要以下能力： - ChatCompletion - json_mode - embedding query - You can set your Chat Model and Embedding Model in the following ways: + 你可以通过以下方式设置 Chat Model 和 Embedding Model： - > **🔥 Attention**: We now provide experimental support for **DeepSeek** models! You can use DeepSeek's official API for cost-effective and high-performance inference. See the configuration example below for DeepSeek setup. + > **🔥 注意**：我们现已为 **DeepSeek** 模型提供实验性支持！你可以使用 DeepSeek 官方 API 进行高性价比、高性能的推理。DeepSeek 配置示例见下文。 -- **Using LiteLLM (Default)**: We now support LiteLLM as a backend for integration with multiple LLM providers. You can configure in multiple ways: +- **使用 LiteLLM（默认）**：我们现已支持 LiteLLM 作为后端，以便与多个 LLM 提供商集成。你可以通过多种方式进行配置： - **Option 1: Unified API base for both models** + **选项 1：为两个模型使用统一的 API base** - *Configuration Example: `OpenAI` Setup :* + *配置示例：`OpenAI` Setup :* ```bash cat << EOF > .env @@ -189,9 +195,9 @@ More details can be found in the [development setup](https://rdagent.readthedocs OPENAI_API_KEY= ``` - *Configuration Example: `Azure OpenAI` Setup :* + *配置示例：`Azure OpenAI` Setup :* - > Before using this configuration, please confirm in advance that your `Azure OpenAI API key` supports `embedded models`. + > 在使用此配置之前，请事先确认你的 `Azure OpenAI API key` 支持 `embedded models`。 ```bash cat << EOF > .env @@ -202,7 +208,7 @@ More details can be found in the [development setup](https://rdagent.readthedocs AZURE_API_VERSION= ``` - **Option 2: Separate API bases for Chat and Embedding models** + **选项 2：为 Chat 和 Embedding 模型使用独立的 API base** ```bash cat << EOF > .env # Set to any model supported by LiteLLM. @@ -221,9 +227,9 @@ More details can be found in the [development setup](https://rdagent.readthedocs LITELLM_PROXY_API_BASE=https://api.siliconflow.cn/v1 ``` - *Configuration Example: `DeepSeek` Setup :* + *配置示例：`DeepSeek` Setup :* - >Since many users encounter configuration errors when setting up DeepSeek. Here's a complete working example for DeepSeek Setup: + > 由于许多用户在配置 DeepSeek 时会遇到配置错误，以下是一个完整可用的 DeepSeek 配置示例： ```bash cat << EOF > .env # CHAT MODEL: Using DeepSeek Official API @@ -237,98 +243,98 @@ More details can be found in the [development setup](https://rdagent.readthedocs LITELLM_PROXY_API_BASE=https://api.siliconflow.cn/v1 ``` - Notice: If you are using reasoning models that include thought processes in their responses (such as \ tags), you need to set the following environment variable: + 注意：如果你使用的是在响应中包含思考过程（例如 \ 标签）的推理模型，需要设置以下环境变量： ```bash REASONING_THINK_RM=True ``` - You can also use a deprecated backend if you only use `OpenAI API` or `Azure OpenAI` directly. For this deprecated setting and more configuration information, please refer to the [documentation](https://rdagent.readthedocs.io/en/latest/installation_and_configuration.html). + 如果你仅直接使用 `OpenAI API` 或 `Azure OpenAI`，也可以使用已弃用的后端。有关此弃用设置及更多配置信息，请参阅[文档](https://rdagent.readthedocs.io/en/latest/installation_and_configuration.html). -- If your environment configuration is complete, please execute the following commands to check if your configuration is valid. This step is necessary. +- 如果环境配置已完成，请执行以下命令检查配置是否有效。此步骤是必要的。 ```bash rdagent health_check ``` -### 🚀 Run the Application +### 🚀 运行应用 -The **[🖥️ Live Demo](https://rdagent.azurewebsites.net/)** is implemented by the following commands(each item represents one demo, you can select the one you prefer): +**[🖥️ Live Demo](https://rdagent.azurewebsites.net/)** 由以下命令实现（每一项代表一个演示，你可以选择你偏好的那一个）： -- Run the **Automated Quantitative Trading & Iterative Factors Model Joint Evolution**: [Qlib](http://github.com/microsoft/qlib) self-loop factor & model proposal and implementation application +- 运行 **Automated Quantitative Trading & Iterative Factors Model Joint Evolution（自动化量化交易与迭代因子模型联合演化）**：[Qlib](http://github.com/microsoft/qlib) 自循环因子与模型提出及实现应用 ```sh rdagent fin_quant ``` -- Run the **Automated Quantitative Trading & Iterative Factors Evolution**: [Qlib](http://github.com/microsoft/qlib) self-loop factor proposal and implementation application +- 运行 **自动化量化交易与迭代因子演化**：[Qlib](http://github.com/microsoft/qlib) 自循环因子提案与实现应用 ```sh rdagent fin_factor ``` -- Run the **Automated Quantitative Trading & Iterative Model Evolution**: [Qlib](http://github.com/microsoft/qlib) self-loop model proposal and implementation application +- 运行 **自动化量化交易与迭代模型演化**：[Qlib](http://github.com/microsoft/qlib) 自循环模型提案与实现应用 ```sh rdagent fin_model ``` -- Run the **Automated Quantitative Trading & Factors Extraction from Financial Reports**: Run the [Qlib](http://github.com/microsoft/qlib) factor extraction and implementation application based on financial reports +- 运行 **自动化量化交易与基于财务报告的因子提取**：运行基于财务报告的 [Qlib](http://github.com/microsoft/qlib) 因子提取与实现应用 ```sh - # 1. Generally, you can run this scenario using the following command: + # 1. 通常，你可以使用以下命令运行该场景： rdagent fin_factor_report --report-folder= - # 2. Specifically, you need to prepare some financial reports first. You can follow this concrete example: + # 2. 具体来说，你需要先准备一些财务报告。你可以参考以下具体示例： wget https://github.com/SunsetWolf/rdagent_resource/releases/download/reports/all_reports.zip unzip all_reports.zip -d git_ignore_folder/reports rdagent fin_factor_report --report-folder=git_ignore_folder/reports ``` -- Run the **Automated Model Research & Development Copilot**: model extraction and implementation application +- 运行 **自动化模型研发 Copilot**：模型提取与实现应用 ```sh - # 1. Generally, you can run your own papers/reports with the following command: + # 1. 通常，你可以使用以下命令运行自己的论文/报告： rdagent general_model - # 2. Specifically, you can do it like this. For more details and additional paper examples, use `rdagent general_model -h`: + # 2. 具体来说，你可以这样做。更多详情和其他论文示例，请参阅 `rdagent general_model -h`： rdagent general_model "https://arxiv.org/pdf/2210.09789" ``` -- Run the **Automated Medical Prediction Model Evolution**: Medical self-loop model proposal and implementation application +- 运行 **自动化医学预测模型演化**：医学自循环模型提案与实现应用 ```bash - # Generally, you can run the data science program with the following command: + # 通常，你可以使用以下命令运行数据科学程序： rdagent data_science --competition - # Specifically, you need to create a folder for storing competition files (e.g., competition description file, competition datasets, etc.), and configure the path to the folder in your environment. In addition, you need to use chromedriver when you download the competition descriptors, which you can follow for this specific example: + # 具体来说，你需要创建一个文件夹来存放竞赛文件（例如竞赛说明文件、竞赛数据集等），并在环境中配置该文件夹路径。此外，下载竞赛描述信息时需要使用 chromedriver，你可以参考以下具体示例： - # 1. Download the dataset, extract it to the target folder. + # 1. 下载数据集，解压到目标文件夹。 wget https://github.com/SunsetWolf/rdagent_resource/releases/download/ds_data/arf-12-hours-prediction-task.zip unzip arf-12-hours-prediction-task.zip -d ./git_ignore_folder/ds_data/ - # 2. Configure environment variables in the `.env` file + # 2. 在 `.env` 文件中配置环境变量 dotenv set DS_LOCAL_DATA_PATH "$(pwd)/git_ignore_folder/ds_data" dotenv set DS_CODER_ON_WHOLE_PIPELINE True dotenv set DS_IF_USING_MLE_DATA False dotenv set DS_SAMPLE_DATA_BY_LLM False dotenv set DS_SCEN rdagent.scenarios.data_science.scen.DataScienceScen - # 3. run the application + # 3. 运行应用 rdagent data_science --competition arf-12-hours-prediction-task ``` - **NOTE:** For more information about the dataset, please refer to the [documentation](https://rdagent.readthedocs.io/en/latest/scens/data_science.html). + **注意：** 有关数据集的更多信息，请参阅 [文档](https://rdagent.readthedocs.io/en/latest/scens/data_science.html). -- Run the **Automated Kaggle Model Tuning & Feature Engineering**: self-loop model proposal and feature engineering implementation application
- > Using **tabular-playground-series-dec-2021** as an example.
- > 1. Register and login on the [Kaggle](https://www.kaggle.com/) website.
- > 2. Configuring the Kaggle API.
- > (1) Click on the avatar (usually in the top right corner of the page) -> `Settings` -> `Create New Token`, A file called `kaggle.json` will be downloaded.
- > (2) Move `kaggle.json` to `~/.config/kaggle/`
- > (3) Modify the permissions of the kaggle.json file. Reference command: `chmod 600 ~/.config/kaggle/kaggle.json`
- > 3. Join the competition: Click `Join the competition` -> `I Understand and Accept` at the bottom of the [competition details page](https://www.kaggle.com/competitions/tabular-playground-series-dec-2021/data). +- 运行 **自动化 Kaggle 模型调优与特征工程**：自循环模型提案与特征工程实现应用
+ > 以 **tabular-playground-series-dec-2021** 为例。
+ > 1. 在 [Kaggle](https://www.kaggle.com/) 网站上注册并登录。
+ > 2. 配置 Kaggle API。
+ > (1) 点击头像（通常在页面右上角）-> `Settings` -> `Create New Token`，将下载名为 `kaggle.json` 的文件。
+ > (2) 将 `kaggle.json` 移动到 `~/.config/kaggle/`
+ > (3) 修改 kaggle.json 文件的权限。参考命令：`chmod 600 ~/.config/kaggle/kaggle.json`
+ > 3. 加入竞赛：在 [竞赛详情页](https://www.kaggle.com/competitions/tabular-playground-series-dec-2021/data). 底部点击 `Join the competition` -> `I Understand and Accept` ```bash - # Generally, you can run the Kaggle competition program with the following command: + # 通常，你可以使用以下命令运行 Kaggle 竞赛程序： rdagent data_science --competition - # 1. Configure environment variables in the `.env` file + # 1. 在 `.env` 文件中配置环境变量 mkdir -p ./git_ignore_folder/ds_data dotenv set DS_LOCAL_DATA_PATH "$(pwd)/git_ignore_folder/ds_data" dotenv set DS_CODER_ON_WHOLE_PIPELINE True @@ -336,127 +342,127 @@ The **[🖥️ Live Demo](https://rdagent.azurewebsites.net/)** is implemented b dotenv set DS_SAMPLE_DATA_BY_LLM True dotenv set DS_SCEN rdagent.scenarios.data_science.scen.KaggleScen - # 2. run the application + # 2. 运行应用 rdagent data_science --competition tabular-playground-series-dec-2021 ``` -- Run **FT-Agent for Autonomous LLM Fine-Tuning**: an ICML 2026 LLM fine-tuning scenario for benchmark-driven data processing, training, evaluation, and feedback-guided refinement. +- 运行 **FT-Agent 自主 LLM 微调**：一个 ICML 2026 LLM 微调场景，支持基于基准测试的数据处理、训练、评估以及反馈引导的优化。 ```bash - # See the full setup, benchmark descriptions, dataset notes, and examples: + # 查看完整设置、基准测试描述、数据集说明和示例： # rdagent/app/finetune/llm/README.md - # Configure FT_TARGET_BENCHMARK and FT_BENCHMARK_DESCRIPTION before running. + # 运行前请配置 FT_TARGET_BENCHMARK 和 FT_BENCHMARK_DESCRIPTION。 rdagent llm_finetune --base-model Qwen/Qwen2.5-7B-Instruct ``` -### 🖥️ Monitor the Application Results +### 🖥️ 监控应用结果 #### Streamlit UI -Use the Streamlit UI to view run logs, especially for the `data_science` scenario. +使用 Streamlit UI 查看运行日志，尤其是 `data_science` 场景。 ```sh rdagent ui --port 19899 --log-dir --data-science ``` -About the `data_science` parameter: If you want to see the logs of the data science scenario, set the `data_science` parameter to `True`; otherwise set it to `False`. +关于 `data_science` 参数：若要查看数据科学场景的运行日志，请将 `data_science` 参数设置为 `True`；否则请设置为 `False`。 #### Web UI -We also provide a separate web frontend in `web/` for the Flask backend started by `server_ui`. +我们还为 `server_ui` 启动的 Flask 后端在 `web/` 中提供了独立的前端。 -**NOTE:** This web UI is different from `rdagent ui`. The current web UI does not support the `data_science` scenario yet. For the `data_science` scenario, please continue to use `rdagent ui --data-science`. +**注意：** 此 Web UI 与 `rdagent ui` 不同。当前的 Web UI 尚不支持 `data_science` 场景。对于 `data_science` 场景，请继续使用 `rdagent ui --data-science`。 ```sh cd web npm install ``` -To build the frontend for the Flask backend, generate the static assets into the default directory used by `server_ui`: +要为 Flask 后端构建前端，请将静态资源生成到 `server_ui` 使用的默认目录： ```sh cd web npm run build:flask ``` -By default, `server_ui` serves static files from `./git_ignore_folder/static`. If you need a different location, set the `UI_STATIC_PATH` environment variable before starting the backend. +默认情况下，`server_ui` 从 `./git_ignore_folder/static` 提供静态文件。若需要使用其他位置，请在启动后端前设置 `UI_STATIC_PATH` 环境变量。 -Start the Flask backend and serve the built frontend together with the real-time APIs: +启动 Flask 后端，同时提供已构建的前端和实时 API： ```sh rdagent server_ui --port 19899 ``` -After that, open `http://127.0.0.1:19899` in your browser. +然后，在浏览器中打开 `http://127.0.0.1:19899`。 -#### Common Notes +#### 通用说明 -Port `19899` is used in the examples above. Before starting either UI, check whether this port is already occupied. If it is, please change it to another available port. +上述示例使用了端口 `19899`。在启动任一 UI 之前，请检查该端口是否已被占用。若已占用，请更换为其他可用端口。 -You can check whether the port is occupied by running: +你可以通过运行以下命令检查端口是否被占用： ```sh rdagent health_check --no-check-env --no-check-docker ``` -# 🏭 Scenarios +# 🏭 场景 -We have applied R&D-Agent to multiple valuable data-driven industrial scenarios. +我们已将 R&D-Agent 应用于多个有价值的数据驱动工业场景。 -## 🎯 Goal: Agent for Data-driven R&D +## 🎯 目标：面向数据驱动研发的 Agent -In this project, we are aiming to build an Agent to automate Data-Driven R\&D that can -+ 📄 Read real-world material (reports, papers, etc.) and **extract** key formulas, descriptions of interested **features** and **models**, which are the key components of data-driven R&D . -+ 🛠️ **Implement** the extracted formulas (e.g., features, factors, and models) in runnable codes. - + Due to the limited ability of LLM in implementing at once, build an evolving process for the agent to improve performance by learning from feedback and knowledge. -+ 💡 Propose **new ideas** based on current knowledge and observations. +在本项目中，我们的目标是构建一个能够自动化数据驱动研发（Data-Driven R\&D）的 Agent，它能够 ++ 📄 阅读真实世界的材料（报告、论文等），并**提取**关键公式、感兴趣的**特征**和**模型**描述，这些都是数据驱动研发的关键组成部分。 ++ 🛠️ 将提取的公式（例如特征、因子和模型）**实现**为可运行的代码。 + + 由于 LLM 一次性实现的能力有限，构建一个演化流程，让 Agent 通过从反馈和知识中学习来提升性能。 ++ 💡 基于当前知识和观察**提出新想法**。 -## 📈 Scenarios/Demos +## 📈 场景/演示 -In the two key areas of data-driven scenarios, model implementation and data building, our system aims to serve two main roles: 🦾Copilot and 🤖Agent. -- The 🦾Copilot follows human instructions to automate repetitive tasks. -- The 🤖Agent, being more autonomous, actively proposes ideas for better results in the future. +在数据驱动场景的两个关键领域——模型实现和数据构建中，我们的系统旨在承担两种主要角色：🦾Copilot 和 🤖Agent。 +- 🦾Copilot 遵循人类指令，自动化重复性任务。 +- 🤖Agent 更具自主性，会主动提出想法以在未来获得更好的结果。 -The supported scenarios are listed below: +支持的场景如下： -| Scenario/Target | Model Implementation | Data Building | +| 场景/目标 | 模型实现 | 数据构建 | | -- | -- | -- | -| **💹 Finance** | 🤖 [Iteratively Proposing Ideas & Evolving](https://rdagent.azurewebsites.net/model_loop)[▶️YouTube](https://www.youtube.com/watch?v=dm0dWL49Bc0&t=104s) | 🤖 [Iteratively Proposing Ideas & Evolving](https://rdagent.azurewebsites.net/factor_loop) [▶️YouTube](https://www.youtube.com/watch?v=X4DK2QZKaKY&t=6s)
🦾 [Auto reports reading & implementation](https://rdagent.azurewebsites.net/report_factor)[▶️YouTube](https://www.youtube.com/watch?v=ECLTXVcSx-c) | -| **🩺 Medical** | 🤖 [Iteratively Proposing Ideas & Evolving](https://rdagent.azurewebsites.net/dmm)[▶️YouTube](https://www.youtube.com/watch?v=VIaSTZuoZg4) | - | -| **🏭 General** | 🦾 [Auto paper reading & implementation](https://rdagent.azurewebsites.net/report_model)[▶️YouTube](https://www.youtube.com/watch?v=BiA2SfdKQ7o)
🤖 Auto Kaggle Model Tuning | 🤖Auto Kaggle feature Engineering | +| **💹 金融** | 🤖 [迭代提出想法与演化](https://rdagent.azurewebsites.net/model_loop)[▶️YouTube](https://www.youtube.com/watch?v=dm0dWL49Bc0&t=104s) | 🤖 [迭代提出想法与演化](https://rdagent.azurewebsites.net/factor_loop) [▶️YouTube](https://www.youtube.com/watch?v=X4DK2QZKaKY&t=6s)
🦾 [自动读取报告与实现](https://rdagent.azurewebsites.net/report_factor)[▶️YouTube](https://www.youtube.com/watch?v=ECLTXVcSx-c) | +| **🩺 医疗** | 🤖 [迭代提出想法与演化](https://rdagent.azurewebsites.net/dmm)[▶️YouTube](https://www.youtube.com/watch?v=VIaSTZuoZg4) | - | +| **🏭 通用** | 🦾 [自动阅读论文与实现](https://rdagent.azurewebsites.net/report_model)[▶️YouTube](https://www.youtube.com/watch?v=BiA2SfdKQ7o)
🤖 自动 Kaggle 模型调优 | 🤖 自动 Kaggle 特征工程 | -- **[RoadMap](https://rdagent.readthedocs.io/en/latest/scens/data_science.html#roadmap)**: Currently, we are working hard to add new features to the Kaggle scenario. +- **[RoadMap](https://rdagent.readthedocs.io/en/latest/scens/data_science.html#roadmap)**: 目前，我们正在努力为 Kaggle 场景添加新功能。 -Different scenarios vary in entrance and configuration. Please check the detailed setup tutorial in the scenarios documents. +不同场景的入口和配置各不相同。请参阅场景文档中的详细设置教程。 -Here is a gallery of [successful explorations](https://github.com/SunsetWolf/rdagent_resource/releases/download/demo_traces/demo_traces.zip) (5 traces showed in **[🖥️ Live Demo](https://rdagent.azurewebsites.net/)**). You can download and view the execution trace using [this command](https://github.com/microsoft/RD-Agent?tab=readme-ov-file#%EF%B8%8F-monitor-the-application-results) from the documentation. +以下是 [成功案例集锦](https://github.com/SunsetWolf/rdagent_resource/releases/download/demo_traces/demo_traces.zip)（展示了 5 条轨迹，详见 **[🖥️ Live Demo](https://rdagent.azurewebsites.net/)**). 你可以使用文档中的 [这条命令](https://github.com/microsoft/RD-Agent?tab=readme-ov-file#%EF%B8%8F-monitor-the-application-results) 下载并查看执行轨迹。 -Please refer to **[📖readthedocs_scen](https://rdagent.readthedocs.io/en/latest/scens/catalog.html)** for more details of the scenarios. +有关场景的更多详情，请参阅 **[📖readthedocs_scen](https://rdagent.readthedocs.io/en/latest/scens/catalog.html)**。 -# ⚙️ Framework +# ⚙️ 框架
Framework-RDAgent
-Automating the R&D process in data science is a highly valuable yet underexplored area in industry. We propose a framework to push the boundaries of this important research field. +在工业界，自动化数据科学领域的研发（R&D）流程是一个极具价值但尚未被充分探索的方向。我们提出了一个框架，以推动这一重要研究领域的边界。 -The research questions within this framework can be divided into three main categories: -| Research Area | Paper/Work List | +该框架内的研究问题可分为三大类： +| 研究领域 | 论文/工作列表 | |--------------------|-----------------| -| **Benchmark the R&D abilities** | [Benchmark](#benchmark) | -| **Idea proposal:** Explore new ideas or refine existing ones | [Research](#research) | -| **Ability to realize ideas:** Implement and execute ideas | [Development](#development) | +| **研发能力基准测试（Benchmark）** | [Benchmark](#benchmark) | +| **创意提案（Idea proposal）：** 探索新想法或完善现有想法 | [Research](#research) | +| **实现创意的能力：** 将想法落地并执行 | [Development](#development) | -We believe that the key to delivering high-quality solutions lies in the ability to evolve R&D capabilities. Agents should learn like human experts, continuously improving their R&D skills. +我们相信，交付高质量解决方案的关键在于研发能力的持续演进。智能体（Agent）应像人类专家一样学习，不断提升自身的研发技能。 -More documents can be found in the **[📖 readthedocs](https://rdagent.readthedocs.io/)**. +更多文档请参阅 **[📖 readthedocs](https://rdagent.readthedocs.io/)**. -# 📃 Paper/Work list +# 📃 论文/工作列表 -## Overall Technical Report +## 整体技术报告 - [R&D-Agent: An LLM-Agent Framework Towards Autonomous Data Science](https://arxiv.org/abs/2505.14738) ```BibTeX @misc{yang2025rdagentllmagentframeworkautonomous, @@ -471,7 +477,7 @@ More documents can be found in the **[📖 readthedocs](https://rdagent.readthed ``` ![image](https://github.com/user-attachments/assets/28b0488d-a546-4fef-8dc5-563ed64a9b4d) -## 📊 Benchmark +## 📊 基准测试（Benchmark） - [Towards Data-Centric Automatic R&D](https://arxiv.org/abs/2404.11276) ```BibTeX @misc{chen2024datacentric, @@ -485,15 +491,15 @@ More documents can be found in the **[📖 readthedocs](https://rdagent.readthed ``` ![image](https://github.com/user-attachments/assets/494f55d3-de9e-4e73-ba3d-a787e8f9e841) -## 🔍 Research +## 🔍 研究（Research） -In a data mining expert's daily research and development process, they propose a hypothesis (e.g., a model structure like RNN can capture patterns in time-series data), design experiments (e.g., finance data contains time-series and we can verify the hypothesis in this scenario), implement the experiment as code (e.g., Pytorch model structure), and then execute the code to get feedback (e.g., metrics, loss curve, etc.). The experts learn from the feedback and improve in the next iteration. +在数据挖掘专家的日常研发过程中，他们会提出假设（例如，像 RNN 这样的模型结构可以捕捉时间序列数据中的模式），设计实验（例如，金融数据包含时间序列，我们可以在此场景中验证该假设），将实验实现为代码（例如 PyTorch 模型结构），然后执行代码以获得反馈（例如指标、损失曲线等）。专家从反馈中学习，并在下一次迭代中改进。 -Based on the principles above, we have established a basic method framework that continuously proposes hypotheses, verifies them, and gets feedback from the real-world practice. This is the first scientific research automation framework that supports linking with real-world verification. +基于上述原则，我们建立了一个基本方法框架，可持续提出假设、验证假设，并从真实世界的实践中获得反馈。这是首个支持与现实世界验证相衔接的科学研究自动化框架。 -For more detail, please refer to our **[🖥️ Live Demo page](https://rdagent.azurewebsites.net)**. +更多详情，请参阅我们的 **[🖥️ Live Demo page](https://rdagent.azurewebsites.net)**. -## 🛠️ Development +## 🛠️ 开发（Development） - [Collaborative Evolving Strategy for Automatic Data-Centric Development](https://arxiv.org/abs/2407.18690) ```BibTeX @@ -508,7 +514,7 @@ For more detail, please refer to our **[🖥️ Live Demo page](https://rdagent. ``` ![image](https://github.com/user-attachments/assets/75d9769b-0edd-4caf-9d45-57d1e577054b) -## Deep Application in Diverse Scenarios +## 多样化场景中的深度应用 - [FT-Dojo: Towards Autonomous LLM Fine-Tuning with Language Agents](https://arxiv.org/abs/2603.01712) @@ -524,7 +530,7 @@ For more detail, please refer to our **[🖥️ Live Demo page](https://rdagent. } ``` -FT-Agent, the autonomous LLM fine-tuning scenario from this paper, is available through the [LLM fine-tuning guide](rdagent/app/finetune/llm/README.md). +本文中的自主 LLM 微调场景 FT-Agent，可通过 [LLM 微调指南](rdagent/app/finetune/llm/README.md) 使用。 - [R&D-Agent-Quant: A Multi-Agent Framework for Data-Centric Factors and Model Joint Optimization](https://arxiv.org/abs/2505.15155) ```BibTeX @@ -551,19 +557,19 @@ FT-Agent, the autonomous LLM fine-tuning scenario from this paper, is available } ``` -You can check the detailed execution traces online at [Gome GPT-5 Traces](https://huggingface.co/datasets/amstrongzyf/Gome-GPT5-Traces). +你可以在线查看详细执行轨迹：[Gome GPT-5 Traces](https://huggingface.co/datasets/amstrongzyf/Gome-GPT5-Traces). -# 🤝 Contributing +# 🤝 贡献（Contributing） -We welcome contributions and suggestions to improve R&D-Agent. Please refer to the [Contributing Guide](CONTRIBUTING.md) for more details on how to contribute. +我们欢迎贡献和建议，以改进 R&D-Agent。有关如何贡献的更多详情，请参阅 [贡献指南](CONTRIBUTING.md)。 -Before submitting a pull request, ensure that your code passes the automatic CI checks. +在提交拉取请求（pull request）之前，请确保你的代码通过了自动 CI 检查。 -## 📝 Guidelines -This project welcomes contributions and suggestions. -Contributing to this project is straightforward and rewarding. Whether it's solving an issue, addressing a bug, enhancing documentation, or even correcting a typo, every contribution is valuable and helps improve R&D-Agent. +## 📝 指南（Guidelines） +本项目欢迎贡献和建议。 +为本项目做贡献既简单又富有成就感。无论是解决问题、修复 bug、改进文档，还是修正错别字，每一份贡献都很有价值，有助于改进 R&D-Agent。 -To get started, you can explore the issues list, or search for `TODO:` comments in the codebase by running the command `grep -r "TODO:"`. +要入门，你可以浏览 issue 列表，或通过运行命令 `grep -r "TODO:"` 在代码库中搜索 `TODO:` 注释。 @@ -571,7 +577,7 @@ To get started, you can explore the issues list, or search for `TODO:` comments -Before we released R&D-Agent as an open-source project on GitHub, it was an internal project within our group. Unfortunately, the internal commit history was not preserved when we removed some confidential code. As a result, some contributions from our group members, including Haotian Chen, Wenjun Feng, Haoxue Wang, Zeqi Ye, Xinjie Shen, and Jinhui Li, were not included in the public commits. +在我们将 R&D-Agent 作为开源项目在 GitHub 上发布之前，它曾是我们组内的内部项目。遗憾的是，在移除部分保密代码时，内部提交历史未能保留。因此，部分组成员的贡献——包括 Haotian Chen、Wenjun Feng、Haoxue Wang、Zeqi Ye、Xinjie Shen 和 Jinhui Li——未能纳入公开提交记录。 -# ⚖️ Legal disclaimer -

The RD-agent is provided “as is”, without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose and noninfringement. The RD-agent is aimed to facilitate research and development process in the financial industry and not ready-to-use for any financial investment or advice. Users shall independently assess and test the risks of the RD-agent in a specific use scenario, ensure the responsible use of AI technology, including but not limited to developing and integrating risk mitigation measures, and comply with all applicable laws and regulations in all applicable jurisdictions. The RD-agent does not provide financial opinions or reflect the opinions of Microsoft, nor is it designed to replace the role of qualified financial professionals in formulating, assessing, and approving finance products. The inputs and outputs of the RD-agent belong to the users and users shall assume all liability under any theory of liability, whether in contract, torts, regulatory, negligence, products liability, or otherwise, associated with use of the RD-agent and any inputs and outputs thereof.

+# ⚖️ 法律免责声明 +

RD-agent 按“原样”提供，不作任何明示或暗示的保证，包括但不限于适销性、特定用途适用性和不侵权的保证。RD-agent 旨在促进金融行业的研发流程，并非可直接用于任何金融投资或建议。用户应针对具体使用场景独立评估和测试 RD-agent 的风险，确保负责任地使用 AI 技术，包括但不限于开发和整合风险缓解措施，并遵守所有适用司法管辖区的所有适用法律和法规。RD-agent 不提供金融意见，也不反映 Microsoft 的意见，且无意取代合格金融专业人员在制定、评估和批准金融产品方面的角色。RD-agent 的输入和输出归用户所有，用户应对与使用 RD-agent 及其任何输入和输出相关的一切责任理论下的全部责任负责，无论是合同、侵权、监管、过失、产品责任还是其他。