Simple Weather Agent created using Hugging Face's smolagents library

• Python v3.3 >
• Docker

This project uses Python's [venv][venv] to manage the Virtual Environment, you can create the .venv directory as follows:

python3 -m venv .venv

venv comes as part of Python's standard library since Python v3.3

.venv should not be commited to the Git Repository

Then activate the Virtual Environment by executing:

source .venv/bin/activate

Confirm the virtual environment is activated by checking Python binary path:

which python

While the virtual environment is active, the above command will output a filepath that includes the .venv directory, e.g. .venv/bin/python.

If you want to switch projects or leave your virtual environment, deactivate the environment:

deactivate

This project uses Dockerized Ollama as LLM, the included Justfile provides a convenient recipe to run the Ollama container.

just run-ollama

Then you can install the Model using:

just run-model

This project is using a memory cheap model: https://ollama.com/library/smollm:1.7b