Llamaindex S3 Index Storage
Using Llamaindex to query an vector index stored in S3. The new LLM Stack. Yay!
How to install a virtual environment using venv
pip install virtualenv
To use venv in your project, in your terminal, create a new project folder, cd to the project folder in your terminal, and run the following command:
mkdir llamaindex-s3-index-storage
cd llamaindex-s3-index-storage
python3 -m venv env
How to activate the virtual environment
You can then activate your new python virtual environment running the command below.
source ./venv/bin/activate
Install python packages
Install the packages the application requires
pip install -r requirements.txt
Environment variables
You will need a few secrets for this to work. You will have to rename .env.dev
to .env
for them to load into the application
OPENAI_API_KEY=
AWS_KEY=
AWS_SECRET=
BUCKET=
S3 bucket
You will need to have a public S3 bucket to use as storage for the index/vectors. Note: I used a pubic bucket i have not tested it with a private one.
llamaindex-demo/storage
Running the app
This command will run the steamlit app and open it in a new tab.
streamlit run app.py
Running the app via Docker
Build the image
docker build -t mystreamlitapp .
Run the new build image
docker run -p 8501:8501 mystreamlitapp