localGPT Backend

Simple Python backend that connects your frontend to Ollama for local LLM chat.

Prerequisites

1. Install Ollama (if not already installed):

   bashCopy

   # Visit https://ollama.ai or run:
   curl -fsSL https://ollama.ai/install.sh | sh
   

2. Start Ollama:

   bashCopy

   ollama serve
   

3. Pull a model (optional, server will suggest if needed):

   bashCopy

   ollama pull llama3.2
   

Setup

1. Install Python dependencies:

   bashCopy

   pip install -r requirements.txt
   

2. Test Ollama connection:

   bashCopy

   python ollama_client.py
   

3. Start the backend server:

   bashCopy

   python server.py
   

Server will run on http://localhost:8000

API Endpoints

Health Check

GET /health

Returns server status and available models.

Chat

POST /chat
Content-Type: application/json

{
  "message": "Hello!",
  "model": "llama3.2:latest",
  "conversation_history": []
}

Returns:

{
  "response": "Hello! How can I help you?",
  "model": "llama3.2:latest",
  "message_count": 1
}

Testing

Test the chat endpoint:

curl -X POST http://localhost:8000/chat \
  -H "Content-Type: application/json" \
  -d '{"message": "Hello!", "model": "llama3.2:latest"}'

Frontend Integration

Your React frontend should connect to:

• Backend: http://localhost:8000
• Chat endpoint: http://localhost:8000/chat

What's Next

This simple backend is ready for:

• ✅ Real-time chat with local LLMs
• 🔜 Document upload for RAG
• 🔜 Vector database integration
• 🔜 Streaming responses
• 🔜 Chat history persistence