Each tutorial remains in its own GitHub repository. This lets you clone only the tutorial you want and follow setup instructions tailored to it.
Run in Google Colab¶
For supported tutorials, use the Open in Colab badge in the notebook. Colab runs in your browser and does not require a local Python installation.
You may still need API credentials (instructions provided within the notebooks). Store them using Colab Secrets when the tutorial directs you to do so. Never type credentials directly into a notebook that you might save or share.
Run a tutorial locally¶
Open the tutorial’s GitHub repository and follow its SETUP.md. The general workflow is:
git clone REPOSITORY_URL
cd REPOSITORY_DIRECTORYThen create the environment and install dependencies exactly as described in that repository. Tutorial environments are intentionally separate because their Python packages and supporting services differ.
Tutorial repositories¶
| Tutorial | Local setup | Cloud option |
|---|---|---|
| Reflection with OpenRouter | Repository | Open in Colab |
| RAG with OpenRouter and Supabase | Repository | Open in Colab |
| Local RAG with Ollama and PostgreSQL | Repository | Local only |
| ReAct with Hugging Face | Repository | Open in Colab |
| ReAct with LangChain | Repository | Open in Colab |
| Multi-Agent Collaboration with LangGraph | Repository | Open in Colab |
Local RAG requirements¶
The local RAG tutorial cannot run in a standard Colab environment. It requires locally available Docker, PostgreSQL, and Ollama services. Clone its repository and follow its SETUP.md from beginning to end before opening the notebook.
API keys¶
Depending on the tutorial, you may need credentials for OpenRouter, Supabase, SerpAPI, or LangSmith. Use environment variables, a local .env file excluded from Git, or Colab Secrets as directed by the tutorial.