https://github.com/astrodevil/opendeepseek-researcher
https://github.com/astrodevil/opendeepseek-researcher
Last synced: 4 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/astrodevil/opendeepseek-researcher
- Owner: Astrodevil
- License: mit
- Created: 2025-02-12T10:30:57.000Z (11 months ago)
- Default Branch: main
- Last Pushed: 2025-02-12T11:06:33.000Z (11 months ago)
- Last Synced: 2025-02-12T11:46:54.581Z (11 months ago)
- Language: Jupyter Notebook
- Size: 18.6 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# OpenDeepSeek-Researcher
This project is an AI-powered deep researcher that automates online research using **DeepSeek-V3** via [Nebius Ai Studio](https://dub.sh/AIStudio) and **[Exa](https://exa.ai/)** search APIs. It continuously generates search queries, extracts relevant content, evaluates information, and produces a comprehensive research report based on all relevant information collected.
## Features
* **Automated Research**: Uses AI to generate search queries and evaluate results.
* **Iterative Search**: Runs multiple iterations to refine and improve the quality of results.
* **Content Extraction & Evaluation**: Identifies and extracts relevant information from search results.
* **Final Report Generation**: Summarizes findings into a well-structured report.
* **Gradio UI**: Provides a user-friendly web interface for easy interaction.
## Installation
To run this project, install the required dependencies:
```sh
!pip install nest_asyncio gradio aiohttp openai exa_py
```
## Configuration
Before running the researcher, set up your API keys in the notebook:
```python
NEBIUS_API_KEY = "your_nebius_api_key" # Replace with your Nebius API key
EXA_API_KEY = "your_exa_api_key" # Replace with your EXA API key
```
## Usage
You can run the research assistant either via a function call or using [Google Colab](https://colab.research.google.com/drive/1MUKaQocLT4kP82u1PlPxUIcxPHD8cfwm?usp=sharing)
## Parameters
* **User Query**: The research topic or question you want to investigate.
* **Iteration Limit**: The maximum number of iterations the AI should perform while refining search queries.
## Example Queries
Try out the following example queries:
* "What are the latest advancements in quantum computing?"
* "How does intermittent fasting impact metabolism?"
* "Best practices for deploying large-scale AI models."
* "Comparison of cloud AI providers: AWS vs. GCP vs. Azure."
## Project Structure
* `async_research()`: Handles the asynchronous execution of research.
* `call_nebius_async()`: Calls Nebius AI for generating responses.
* `perform_search_async()`: Uses Exa API to fetch search results.
* `is_page_useful_async()`: Evaluates the relevance of a webpage.
* `extract_relevant_context_async()`: Extracts meaningful content from pages.
* `generate_final_report_async()`: Compiles all information into a final research report.
* `gradio_run()`: Wraps everything into a Gradio UI.
## Contributing
Contributions are welcome! Please feel free to open issues or submit pull requests.
## License
This project is licensed under the [MIT License](LICENSE).
### Acknowledgments
- This researcher is a extended fork of [OpenDeepResearcher](https://github.com/mshumer/OpenDeepResearcher)
- DeepSeek Model is used via [Nebius AI Studio](https://dub.sh/AIStudio)