Hi SDK - LLM Development Kit for Raspberry Pi

A Python SDK for interacting with LLM models on Raspberry Pi.

Installation Options

Option 1: Pre-built Raspberry Pi Image

Download our pre-configured Raspberry Pi image with all dependencies: Download Hi SDK Image

This is a pre-built image with all dependencies installed on a Raspberry Pi OS 6.6.31+rpt-rpi-v8.

Option 2: Quick Setup

Prerequisites:

• Python 3.11.2 or higher
• Raspberry Pi OS

# Clone and install
git clone https://github.com/svenplb/hi-sdk.git
cd hi-sdk
pip install -e .

# Run setup (this will install Ollama, download models, and start services)
python3 -m hi setup --dev

After setup completes, you can start using the SDK:

hi chat  # Start chatting
hi models  # List available models

Option 3: Manual Setup (Without Image)

If you prefer to set up dependencies manually on your existing system:

1. Install system dependencies:

sudo apt-get update
sudo apt-get install -y python3-pip git

2. Install Ollama (LLM backend):

curl https://ollama.ai/install.sh | sh
ollama pull gemma:2b

3. Clone and set up the SDK:

git clone https://github.com/svenplb/hi-sdk.git
cd hi-sdk
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt
pip install -e .

4. Run Ollama on new terminal:

ollama serve

5. Start the API server on new terminal:

python3 -m uvicorn main:app --reload

Now you can use the SDK as shown in the Quick Start section.

Quick Start

from sdk.client import HiClient

client = HiClient()
client.load_model("qwen:1.8b", temperature=0.7)
response = client.chat("Tell me a joke")
print(response)

Features

• Multiple model support (gemma2:2b, qwen:1.8b, qwen:0.5b)
• Streaming responses
• System prompts
• Performance metrics
• Conversation tracking
• CLI interface