Agents

Overview

Architecture

How to run

Local

Before running codes, you need to install the dependencies.

$ poetry install

To see some components are properly working, you can see the available commands as below.

$ poetry run python main.py --help

To run the Streamlit app, you can run the following commands.

$ poetry run streamlit run main.py streamlit-app

Docker

# see the available commands
$ docker run --rm \
    -v ${PWD}/.env:/app/.env \
    ks6088ts/workshop-llm-agents:latest \
    python main.py --help

# mount the .env file to the container and expose the port 8501
$ docker run --rm \
    -v ${PWD}/.env:/app/.env \
    -p 8501:8501 \
    ks6088ts/workshop-llm-agents:latest

Use cases

Chatbot with Tools

Graph:

Screenshot:

References

• 🚀 LangGraph Quick Start > Part 2: 🛠️ Enhancing the Chatbot with Tools

Research Agent with Human-in-the-loop

Graph:

Screenshot:

YouTube: Research Agent using Azure

References

• Research Agent with Human-in-the-loop

Documentation Agent

Graph:

Screenshot:

Tools Agent

Graph:

Screenshot:

Summarize Agent

Graph:

$ poetry run python main.py agents-summarize-run \
    --png ./docs/images/summarize_agent.png

References

• How to summarize text through parallelization