Hugging Face Agents Course · Final project

Local Research Agent

A Qwen 3.5 agent that researches sources, uses tools, and keeps a trace of its work. Run it locally, review its answers here, then submit to the official course evaluator.

Inspect the agent code ↗

1. Run your agent

Install Ollama and the Python requirements from the source repository, then run:

ollama pull qwen3.5:9b
python -m pip install -r requirements.txt
python agent.py --output evaluation

The agent saves tool traces and evaluation/answers.json. Inaccessible attachments are recorded as unavailable. Model inference stays on your computer.

2. Review and submit

Choose the results file to review its answers.

Submission sends your username, these answers, and this public project’s code link to the official course scoring service.