Installation

Requirements

• Python 3.10+
• 4GB+ RAM (8GB+ for local models)

Install

pip install spade_llm

Verify Installation

import spade_llm
from spade_llm import LLMAgent, LLMProvider

print(f"SPADE_LLM version: {spade_llm.__version__}")

LLM Provider Setup

Choose one provider:

OpenAI

export OPENAI_API_KEY="your-api-key"

Ollama (Local)

# Install Ollama
curl -fsSL https://ollama.ai/install.sh | sh

# Download a model
ollama pull llama3.1:8b
ollama serve

LM Studio (Local)

1. Download LM Studio
2. Download a model through the GUI
3. Start the local server

XMPP Server

For development, use a public XMPP server like jabber.at or set up Prosody locally.

Development Install

git clone https://github.com/sosanzma/spade_llm.git
cd spade_llm
pip install -e ".[dev]"

Troubleshooting

Import errors: Ensure you're in the correct Python environment

python -m pip install spade_llm

SSL errors: For development only, disable SSL verification:

agent = LLMAgent(..., verify_security=False)

Ollama connection: Check if Ollama is running:

curl http://localhost:11434/v1/models