Skip to content

Latest commit

 

History

History
71 lines (52 loc) · 3.05 KB

README.md

File metadata and controls

71 lines (52 loc) · 3.05 KB

Code Explorer

This app allows you to ask questions and get answers regarding your code provided the folder location of your code. It is a RAG LLM chain. The app was based on the GenAI stack from https://github.com/docker/genai-stack.

App

Diagram

Configure

Create a .env file from the environment template file .example.env

Available variables:

Variable Name Default value Description
OLLAMA_BASE_URL http://host.docker.internal:11434 REQUIRED - URL to Ollama LLM API
NEO4J_URI neo4j://database:7687 REQUIRED - URL to Neo4j database
NEO4J_USERNAME neo4j REQUIRED - Username for Neo4j database
NEO4J_PASSWORD password REQUIRED - Password for Neo4j database
LLM codellama:7b-instruct REQUIRED - Can be any Ollama model tag, or gpt-4 or gpt-3.5 or claudev2
EMBEDDING_MODEL ollama REQUIRED - Can be sentence_transformer, openai, aws or ollama
AWS_ACCESS_KEY_ID REQUIRED - Only if LLM=claudev2 or embedding_model=aws
AWS_SECRET_ACCESS_KEY REQUIRED - Only if LLM=claudev2 or embedding_model=aws
AWS_DEFAULT_REGION REQUIRED - Only if LLM=claudev2 or embedding_model=aws
OPENAI_API_KEY REQUIRED - Only if LLM=gpt-4 or LLM=gpt-3.5 or embedding_model=openai
LANGCHAIN_ENDPOINT "https://api.smith.langchain.com" OPTIONAL - URL to Langchain Smith API
LANGCHAIN_TRACING_V2 false OPTIONAL - Enable Langchain tracing v2
LANGCHAIN_PROJECT OPTIONAL - Langchain project name
LANGCHAIN_API_KEY OPTIONAL - Langchain API key

NOTE: If using EMBEDDING_MODEL=sentence_transformer, uncomment code in requirements.txt & chains.py. It was commented out to reduce code size.

NOTE: Make sure to set the OLLAMA_BASE_URL=http://llm:11434 in the .env file when using Ollama docker container.

Docker (Linux only)

Build only

docker compose --profile linux build

To start everything (Linux)

docker compose --profile linux up --build

To enter watch mode (auto rebuild on file changes). First start everything, then in new terminal:

docker compose --profile linux alpha watch

Shutdown If health check fails or containers don't start up as expected, shutdown completely to start up again.

docker compose --profile linux down

Application

Access the app at http://localhost:8501. In the sidebar, enter the path to your code folder and click "Process files". Then start asking questions about your code in the main chat. The detailed mode toggle switches between asking the QA agent only (detailedMode=true) and asking an agent which uses the QA agent as a tool (detailedMode=false). In testing, the agent appears to summarize rather than giving a technical response as opposed to the QA agent only.

The Neo4j vector database can be explored at http://localhost:7474.