Basic AI Agent that runs using a local Ollama LLM instance (or uncomment the hf inference api if you have credit limits). Developed as part of the Huggingface AI Agents course.

Model

This agent is configured to use the llama 3.2B model, run locally via Ollama, and accessed with the LiteLLMModel.

Tools

The agent is facilitated with the following tools:

1. Fetch current time of a timezone
2. Fetch current weather of a city using OpenWeather APIs

Access Needed to Run

To allow this agent to work with all its tools, the following keys have to be provided in a secret.config file. Populate this file as <KEY_NAME>=<API_KEY> before running locally.

1. OPENWEATHER_API_KEY: Generate API_KEY from https://openweathermap.org/api and provide to the agent to be able to fetch real-time weather data.

Run Agent

The agent is run locally by executing the app.py script and can be interacted with via the gradio UI hosted on localhost.