Chatbot tutorial

This repository contains some jupyter-notebooks to serve as a tutorial on how to create a RAG agent with LangChain.

Pre-requesits

Please run these steps before the workshop. There are three ways to be ready for the workshop

1. Google Colab: Requires the least amount of set-up but requires a google account
2. GitHub Codespace: A browser version of VSCode requires 2 minutes of preparation (recommended for those familiar with VSCode).
3. Local: You can run all of this locally, but you would need to install a couple of things first

Google Colab (strongly recommended)

Please sign-up with you Google account and go to: https://colab.research.google.com/

If prompted click on "Authorize with Github" from settings and refresh

You can select GitHub and put this repository URL: https://github.com/ericksantillan-planday/chatbot-tutorial.git You should look somethig like this:

On the page select Google_Colab.ipynb

Run the first two cells to be sure that you are ready for the workshop! And that's it :)

GitHub Codespace

Alternatively you can work directly on the GitHub Codespace:
To do this

1. Go to the repository website and press shift + .
   this will open a new tab of VScode with the current project.
2. Open the terminal with ctrl + ~
3. You will be asked where you want the new environment. Choose the GitHub Codespace, and from the top command bar, choose the desired resources (take the lowest one).
4. A new terminal will open at the bottom. Click on bash on the bottom right panel.
5. You may check out to your own branch git checkout -b my_branch
6. Go to the file 0-Basics_python.ipynb and a pop-up will suggest you install the Python extension. Do it :)
7. Alternatively, go to extensions on the left options bar, search, and install the Python package.
8. The first time you try to run a Jupyter cell, you will be asked, on the top command bar, to install/enable the Python + Jupiter extension. Do it :)
9. Then, on the same top command bar, you will need to choose kernel, choose' Python Environment...`, and then choose the only Python bin path that you see.
10. and Walla! You are ready to go as if you were on your local machine!

Locally

You can of course run this locally, but you may encounter different issues depending of your OS, if you already have a python installation, your environment etc..

You have to make sure that you have:

• Python installed
• Git installed
• Jupyter-lab or Jupyter-notebooks installed

Install Python

For python you can use Self-service: Search for python and click installed. Once it's been installed open a terminal and run the following to make sure you have python

python3 --version

For more info please see: https://www.python.org/downloads/

Now, clone the repository:

git clone https://github.com/ericksantillan-planday/chatbot-tutorial.git
cd chatbot-tutorial

Install jupyter lab

You can follow the steps here It mainly consists on running:

pip3 install jupyterlab

However, we found the more consistent way in mac is to use brew:

brew install jupyterlab

You may need to run

jupyter-lab build

Create virtual environment

Open a new terminal (this should refresh your PATH) and run:

# Create virtual environment
python3 -m venv venv
# Activate it
source venv/bin/activate
# Install requirements
pip3 install -r requirements.txt
# Install python interactive kernel for this environment
pip3 install ipykernel
# Make kernel available to jupyer (you may need sudo to run this, depending on your jupyter installation)
python3 -m ipykernel install --name chatbot-tutorial

Run Jupyter-lab

On a new terminal run

jupyter-lab

If prompted select the chatbot-tutorial kernel and that should be it :)

If something is not running feel free to ping us!