LLM -- OpenAI GPT Applications

Sample Python Project To Show LLM Capabilities. For frontend, you can refer https://github.com/bignyap/llm-fe

Keycloak Authentication

• src/auth_service

Interactive Chat Application -- src/chat_service

• Company Name with simple prompt engineering
• Document Summarizer
• Video to Audio
• You Tube Video Summarizer

Non-Interactive Applications -- src/llm_service

• Interactive chat interface using LLMs

How to Run:

• Navigate to src/ folder and open terminal
• Create a .env file and fill in the config values mentioned in .env.sample
• Feel free to remove the mysql service from docker-compose file if you already have a mysql instance running. In that case, create the username and database appropriately. If you are using the mysql service, then you do not have worry about changing anything.
• Run docker-compose down; docker-compose up -d
• You should be able to open these 4 links in any browser:
  • https://localhost:8443/auth/
  • https://localhost/auth_service/docs
  • https://localhost/chat_service/docs
  • https://localhost/llm_service/docs

Requirements:

• Docker

Frameworks used

• Docker
• Envoy
• Python (fastapi, mysql.connector, openai, etc)
• SQL (MySQL)
• Keycloak

Concepts / Technologies

• LLM
• REST APIs
• RDBMS
• Proxy (Rate Limiting)
• Authentication (OIDC)
• Micro Services