Skip to content

setup_gemma.md

Latest commit

 

History

History
68 lines (52 loc) · 1.33 KB

setup_gemma.md

File metadata and controls

68 lines (52 loc) · 1.33 KB

🚀 Gemma Guard Setup Guide

Step 1: Install Ollama

  1. Download Ollama from https://ollama.ai
  2. Install and start Ollama service

Step 2: Download Gemma Model

# For Gemma 3n (latest version - recommended)
ollama pull gemma:latest

# Or for specific versions
ollama pull gemma:7b
ollama pull gemma:2b

Step 3: Verify Ollama Installation

# Check if Ollama is running
ollama list

# Test Gemma
ollama run gemma:latest "Hello, how are you?"

Step 4: Setup Python Environment

# Create virtual environment
python -m venv venv

# Activate (Windows)
venv\Scripts\activate

# Activate (macOS/Linux)
source venv/bin/activate

# Install dependencies
pip install -r requirements.txt

Step 5: Configure Environment

# Copy environment template
copy env.example .env

# Edit .env file and set:
ENABLE_OLLAMA_INTEGRATION=true
OLLAMA_MODEL=gemma:latest

Step 6: Run the Application

streamlit run app/main.py

🔧 Troubleshooting

If Ollama connection fails:

  • Check if Ollama service is running
  • Verify model is downloaded: ollama list
  • Check Ollama logs for errors

If Gemma model not found:

  • Download the model: ollama pull gemma:latest
  • Update OLLAMA_MODEL in .env file

Fallback Mode:

Set ENABLE_OLLAMA_INTEGRATION=false in .env to use simulation mode while testing.