/aws-bedrock

Developing generative AI applications using Amazon Bedrock

Primary LanguagePython

Prerequisites

  1. Make sure that you have set up AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY in .aws directory within credentials file
  2. Go to Amazon Bedrock Console's model access section to request an access for models
  3. After cloning the repository, run cd aws_bedrock
  4. Run export PYTHONPATH=$PYTHONPATH:$(pwd)/src
  5. Then, run chmod +x run.sh
  6. Using pyenv set up your python version above 3.11 e.g. pyenv local 3.11.5
  7. Run the following command to install project dependencies
poetry env use $(pyenv which python)
poetry install
  1. Afterwards, to activate virtual env run poetry shell

How to get demo applications up and running

Building Bonds

  1. Set env for Tavily Search and Nubela in .env file, check the naming convention from .env.example file
  2. Run ./run.sh building_bonds

Ingredient to Recipe

  1. Set HuggingFace API Token in .env
  2. Run ./run.sh ingredient_to_recipe

Resume Screening

  1. Spin up Postgres container with pgvector extension docker run --name pgvector-container -e POSTGRES_USER=russ -e POSTGRES_PASSWORD=russ -e POSTGRES_DB=resume_screener -p 6024:5432 -d pgvector/pgvector:pg16
  2. Run ./run.sh resume_screening

References

More