MultiAgent demo

Overview

This repository showcases a collaborative framework using CrewAI to efficiently manage tasks among multiple agents. The project provides an organized structure to leverage CrewAI's capabilities, agents, tasks, and configurations. The modular architecture enables easy scalability and facilitates the addition of new functionalities or agents in the future.

Features

1. CrewAI Implementation

Crew Management: Demonstrates the setup and execution of a Crew using the CrewAI framework.
Agent Allocation: Defines various agents, their roles, goals, and tools for task execution.
Task Assignment: Provides examples of tasks delegated to different agents within the CrewAI framework.

2. Agent and OpenAIWrapper Integration

Agent Creation: Illustrates the creation of agents with specific roles and objectives.
OpenAI Integration: Shows integration with the OpenAIWrapper for specific agent tasks.

3. Local Model with LM Studio Server

Model Setup: Incorporates a local machine learning model, specifically leveraging the Zephyr Beta 7b version for testing purposes. While testing, the usage of a smaller model was considered; however, most of the time it would stop due to iteration limit or time limit.
Server Configuration: Configured to operate on port 3008.

Running the Code

Navigate to the project directory.
Execute the main file or specific modules:
```
python main.py
```

Configuration

Agents

Researcher Agent (researcher):
- Role: Senior Research Analyst
- Objective: Utilizes DuckDuckGo search for in-depth research on user queries.
Writer Agent (writer):
- Role: Tech Content Strategist
- Objective: Generates blog posts using data collected by the Researcher Agent.

Tasks

Tasks are defined in the tasks.py file, outlining specific objectives for agents.

ViralOne/multiagent-crewai