john-adeojo / custom_websearch_agent

Custom Websearch Agent Built with Local Models, vLLM, and OpenAI
MIT License
124 stars 55 forks source link

Custom Agent

A custom websearch agent useable with Ollama, OpenAI and vLLM.

Agent Schema:

Agent Schema

Prerequisites

Environment Setup

  1. Install Anaconda:
    Download Anaconda from https://www.anaconda.com/.

  2. Create a Virtual Environment:

    conda create -n agent_env python=3.10 pip
  3. Activate the Virtual Environment:

    conda activate agent_env

    Setup Ollama Server

  4. Download Ollama: Download https://ollama.com/download

  5. Download an Ollama Model:

    curl http://localhost:11434/api/pull -d "{\"name\": \"llama3\"}"

    Ollama[https://github.com/ollama/ollama/blob/main/docs/api.md#list-local-models](API documentionation)

Clone and Navigate to the Repository

  1. Clone the Repo:

    git clone https://github.com/john-adeojo/custom_agent_tutorial.git
  2. Navigate to the Repo:

    cd /path/to/your-repo/custom_agent_tutorial
  3. Install Requirements:

    pip install -r requirements.txt

Configure API Keys

  1. Open the config.yaml:

    nano config.yaml
  2. Enter API Keys:

Run Your Query

python agent.py run

Then enter your query.