Unit 0 — Welcome & Setup¶

Overview¶

This unit covers everything you need to get started:

Understanding what the course is about
Setting up your development environment
Getting access to the required APIs and tools

Environment setup¶

1. Clone this repository¶

git clone https://github.com/jcphysica5/hf-agents-systems-notes.git
cd hf-agents-systems-notes

2. Create the conda environment¶

conda env create -f environment.yml
conda activate hf-agents-env

This installs all packages at the exact versions used to develop the course notebooks (Python 3.11, smolagents 1.24, LangGraph 1.0, LlamaIndex 0.14, and more).

Colab / pip users: skip conda and run pip install -r requirements.txt instead.

3. Configure API keys¶

Create a .env file at the root of the project:

HF_TOKEN=hf_...          # Hugging Face token (Settings → Access Tokens)
OPENAI_API_KEY=sk-...    # Optional — only if you use OpenAI models

Warning

Never commit your .env file. It is already listed in .gitignore.

Key tools used in this course¶

Tool	Purpose
`smolagents`	Lightweight agent library by Hugging Face
`transformers`	Model loading & inference
`huggingface_hub`	Accessing HF Hub models and spaces
`langchain` / `langgraph`	Agent orchestration
`llama-index`	Data-centric agent framework