mattzcarey / mini-quivr-demo

Mini Quivr Demo (with Bedrock)

Build a personal assistant with Langchain, Streamlit and AWS

Project requirements

Pyenv and Python 3.11.4 (Optional but highly recommended)

• Install pyenv to manage your Python versions and virtual environments:

  curl -sSL https://pyenv.run | bash

  • If you are on MacOS and experiencing errors on python install with pyenv, follow this comment
  • Add these lines to your ~/.bashrc or ~/.zshrc to be able to activate pyenv virtualenv:

    eval "$(pyenv init -)"
    eval "$(pyenv virtualenv-init -)"
    eval "$(pyenv init --path)"

  • Restart your shell

• Install the right version of Python with pyenv:

  pyenv install 3.11.4

Poetry

• Install Poetry to manage your dependencies and tooling configs:

  curl -sSL https://install.python-poetry.org | python - --version 1.5.1

  If you have not previously installed any Python version, you may need to set your global Python version before installing Poetry:

  pyenv global 3.11.4

Installation

Create a virtual environment

Create your virtual environment and link it to your project folder:

pyenv virtualenv 3.11.4 aws-comsum-gen-ai-workshop
pyenv local aws-comsum-gen-ai-workshop

Now, every time you are in your project directory your virtualenv will be activated thanks to pyenv!

Install Python dependencies through poetry

poetry install --no-root

Streamlit App

The project includes a Streamlit app in ./demo/main.py.

This is where you will build your personal assistant.

Next Steps

Please insure you have set up a virtual environment and installed the dependencies before moving on.

Continue the workshop by navigating to instructions/1-running-the-app.md