Spaces:
Running
Running
File size: 5,071 Bytes
372531f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 |
# LangGraph x GPT Researcher
[LangGraph](https://python.langchain.com/docs/langgraph) is a library for building stateful, multi-actor applications with LLMs.
This example uses Langgraph to automate the process of an in depth research on any given topic.
## Use case
By using Langgraph, the research process can be significantly improved in depth and quality by leveraging multiple agents with specialized skills.
Inspired by the recent [STORM](https://arxiv.org/abs/2402.14207) paper, this example showcases how a team of AI agents can work together to conduct research on a given topic, from planning to publication.
An average run generates a 5-6 page research report in multiple formats such as PDF, Docx and Markdown.
Please note: This example uses the OpenAI API only for optimized performance.
## The Multi Agent Team
The research team is made up of 8 agents:
- **Human** - The human in the loop that oversees the process and provides feedback to the agents.
- **Chief Editor** - Oversees the research process and manages the team. This is the "master" agent that coordinates the other agents using Langgraph.
- **Researcher** (gpt-researcher) - A specialized autonomous agent that conducts in depth research on a given topic.
- **Editor** - Responsible for planning the research outline and structure.
- **Reviewer** - Validates the correctness of the research results given a set of criteria.
- **Revisor** - Revises the research results based on the feedback from the reviewer.
- **Writer** - Responsible for compiling and writing the final report.
- **Publisher** - Responsible for publishing the final report in various formats.
## How it works
Generally, the process is based on the following stages:
1. Planning stage
2. Data collection and analysis
3. Review and revision
4. Writing and submission
5. Publication
### Architecture
<div align="center">
<img align="center" height="600" src="https://github.com/user-attachments/assets/ef561295-05f4-40a8-a57d-8178be687b18">
</div>
<br clear="all"/>
### Steps
More specifically (as seen in the architecture diagram) the process is as follows:
- Browser (gpt-researcher) - Browses the internet for initial research based on the given research task.
- Editor - Plans the report outline and structure based on the initial research.
- For each outline topic (in parallel):
- Researcher (gpt-researcher) - Runs an in depth research on the subtopics and writes a draft.
- Reviewer - Validates the correctness of the draft given a set of criteria and provides feedback.
- Revisor - Revises the draft until it is satisfactory based on the reviewer feedback.
- Writer - Compiles and writes the final report including an introduction, conclusion and references section from the given research findings.
- Publisher - Publishes the final report to multi formats such as PDF, Docx, Markdown, etc.
## How to run
1. Install required packages:
```bash
pip install -r requirements.txt
```
3. Update env variables
```bash
export OPENAI_API_KEY={Your OpenAI API Key here}
export TAVILY_API_KEY={Your Tavily API Key here}
```
2. Run the application:
```bash
python main.py
```
## Usage
To change the research query and customize the report, edit the `task.json` file in the main directory.
#### Task.json contains the following fields:
- `query` - The research query or task.
- `model` - The OpenAI LLM to use for the agents.
- `max_sections` - The maximum number of sections in the report. Each section is a subtopic of the research query.
- `include_human_feedback` - If true, the user can provide feedback to the agents. If false, the agents will work autonomously.
- `publish_formats` - The formats to publish the report in. The reports will be written in the `output` directory.
- `source` - The location from which to conduct the research. Options: `web` or `local`. For local, please add `DOC_PATH` env var.
- `follow_guidelines` - If true, the research report will follow the guidelines below. It will take longer to complete. If false, the report will be generated faster but may not follow the guidelines.
- `guidelines` - A list of guidelines that the report must follow.
- `verbose` - If true, the application will print detailed logs to the console.
#### For example:
```json
{
"query": "Is AI in a hype cycle?",
"model": "gpt-4o",
"max_sections": 3,
"publish_formats": {
"markdown": true,
"pdf": true,
"docx": true
},
"include_human_feedback": false,
"source": "web",
"follow_guidelines": true,
"guidelines": [
"The report MUST fully answer the original question",
"The report MUST be written in apa format",
"The report MUST be written in english"
],
"verbose": true
}
```
## To Deploy
```shell
pip install langgraph-cli
langgraph up
```
From there, see documentation [here](https://github.com/langchain-ai/langgraph-example) on how to use the streaming and async endpoints, as well as the playground.
|