build-on-aws / llm-rag-vectordb-python Goto Github PK

Explore sample applications and tutorials demonstrating the prowess of Amazon Bedrock with Python. Learn to integrate Bedrock with databases, use RAG techniques, and showcase experiments with langchain and streamlit.

License: MIT No Attribution

Python 19.27% HTML 0.46% CSS 0.18% JavaScript 0.76% Jupyter Notebook 63.68% Dockerfile 0.11% Shell 0.13% TypeScript 15.41%

amazon-bedrock langchain-python llms node-js python3 rag

llm-rag-vectordb-python's Introduction

☁️🐍 Getting started with Amazon Bedrock, RAG, and Vector database in Python

🔍 Introduction

In this repository, you'll find sample applications and tutorials that showcase the power of Amazon Bedrock with Python. These resources are designed to help Python developers understand how to harness Amazon Bedrock in building generative AI-enabled applications. You'll also discover how to integrate Bedrock with vector databases using RAG (Retrieval-augmented generation), and services like Amazon Aurora, RDS, and OpenSearch. Additionally, get insights into using langchain and streamlit to create applications that demonstrate your experiments effectively.

📑 Table of Contents

📚🦜 Unified AI Q&A: Harnessing `pgvector`, Amazon Aurora & Amazon Bedrock

Craft sophisticated Q&A bots for specialized tasks, and experience the union of pgvector with Amazon Aurora PostgreSQL and the prowess of Titan LLMs under the RAG paradigm.

📖 Guide & Setup
🌠 Key Features:
- Seamless integration with Streamlit.
- Efficient backend with Amazon Bedrock and Aurora.

🚀 Integrated Fullstack Showcase

Harness the power of Stable Diffusion AI using Amazon Bedrock.

🖥 Live Demo
📖 Guide & Setup
🌠 Key Features:
- Seamless integration: Lambda, API Gateway, Bedrock, Amplify
- Deployment via Serverless stack.

📄 Resume Screening App

Streamline resume screening based on specific job descriptions.

📖 Guide & Setup
🌠 Key Features:
- Seamless integration with Streamlit.
- Efficient backend with Amazon Bedrock and Aurora.

🤝 Building Bonds

Revolutionize introductions by fetching LinkedIn profiles and generating engaging summaries.

📖 Guide & Setup
🌠 Key Features:
- Instant LinkedIn profile retrieval.
- Automated summaries & ice-breakers via Amazon Bedrock and LangChain.

📊 Data Analysis Tool

Analyze CSV data with a streamlined Streamlit application.

📖 Guide & Setup
🌠 Key Features:
- Smooth UI with Streamlit.
- Advanced functions via Langchain.

🥘 Instant Recipe Generator

Build a streamlined Streamlit application to generate recipes given an image of all the ingredients.

📖 Guide & Setup
🌠 Key Features:
- Smooth web application interface via Streamlit.
- Advanced functionalities through Langchain.
- Integration with Hugging Face.
- Generative AI applications with Amazon Bedrock.

💼 Getting Started

📥 Clone this repository.
🗂 Navigate to the desired project directory:
- Resume Screening App: README guide.
- Building Bonds: README guide.
- Stable Diffusion AI App: README guide.
- Instant Recipe Generator: README guide.
- Data Analysis Tool: README guide.
🔧 Set up a virtual environment, .env files, and install dependencies as outlined in each README.
🚀 Launch the desired Streamlit app and delve in!

🔒 Security

See more on security.

📜 License

Licensed under the MIT-0 License. View License.

llm-rag-vectordb-python's People

Contributors

Stargazers

Watchers

Forkers

debnsuma koushal2018 thubz09 anandmandilwar zhenghongpeng stophobia shashipal95 keshiriotinto kukrev dhirajsir mtrl1514 guyvani jas0n1iu dutch-arrow

llm-rag-vectordb-python's Issues

searching-from-my-data : ValueError: invalid literal for int() with base 10: 'None'

Traceback:


File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/data-search-env/lib/python3.10/site-packages/streamlit/runtime/scriptrunner/script_runner.py", line 556, in _run_script
    exec(code, module.__dict__)
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/app.py", line 284, in <module>
    main()
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/app.py", line 267, in main
    st.session_state.conversation = get_conversation_chain(get_vectorstore(None), selected_llm)
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/app.py", line 75, in get_vectorstore
    raise e  # Raise the exception to see the actual error
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/app.py", line 69, in get_vectorstore
    return PGVector(
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/data-search-env/lib/python3.10/site-packages/langchain/vectorstores/pgvector.py", line 126, in __init__
    self.__post_init__()
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/data-search-env/lib/python3.10/site-packages/langchain/vectorstores/pgvector.py", line 134, in __post_init__
    self._conn = self.connect()
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/data-search-env/lib/python3.10/site-packages/langchain/vectorstores/pgvector.py", line 151, in connect
    engine = sqlalchemy.create_engine(self.connection_string)
File "<string>", line 2, in create_engine
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/data-search-env/lib/python3.10/site-packages/sqlalchemy/util/deprecations.py", line 281, in warned
    return fn(*args, **kwargs)  # type: ignore[no-any-return]
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/data-search-env/lib/python3.10/site-packages/sqlalchemy/engine/create.py", line 548, in create_engine
    u = _url.make_url(url)
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/data-search-env/lib/python3.10/site-packages/sqlalchemy/engine/url.py", line 838, in make_url
    return _parse_url(name_or_url)
File "/var/www/python/llm-rag-vectordb-python/searching-from-my-data/data-search-env/lib/python3.10/site-packages/sqlalchemy/engine/url.py", line 899, in _parse_url
    components["port"] = int(components["port"])

env

PGVECTOR_DRIVER='psycopg2'
PGVECTOR_USER='postgres'
PGVECTOR_PASSWORD='12345'
PGVECTOR_HOST='postdemo124.cluster-ch7t2d1ht3ag.us-east-1.rds.amazonaws.com'
PGVECTOR_PORT=5432
PGVECTOR_DATABASE='table1'

app.py

CONNECTION_STRING = PGVector.connection_string_from_db_params(
    driver=os.environ.get("PGVECTOR_DRIVER"),
    user=os.environ.get("PGVECTOR_USER"),
    password=os.environ.get("PGVECTOR_PASSWORD"),
    host=os.environ.get("PGVECTOR_HOST"),
    port=os.environ.get("PGVECTOR_PORT"), 
    database=os.environ.get("PGVECTOR_DATABASE")
)

Currently using SQLAlchemy 2.0.23 , psycopg 3.1.10

tried to cast the port to int
port=int(os.environ.get("PGVECTOR_PORT", 5432)),
It throws Can't load plugin: sqlalchemy.dialects:postgresql.None