Documentation for Kamiwaza 0.6.0

This is documentation for Kamiwaza 0.6.0, which is no longer actively maintained. For the current GA release, see 1.0.1.

Prerequisites

Before you begin, make sure you have:

• Kamiwaza installed and configured (see our Installation Guide)
• At least 16GB of available RAM
• A stable internet connection for downloading models

Step 1: Start Kamiwaza

First, let's get the Kamiwaza platform running on your system.

For Community Edition (Ubuntu .deb package)

If you installed via the .deb package, Kamiwaza should start automatically as a system service. You can check the status with:

kamiwaza status

If it's not running, start it with:

kamiwaza start

For Manual Installations

Navigate to your Kamiwaza installation directory and start the platform:

cd /path/to/kamiwaza

bash startup/kamiwazad.sh start

Verify Kamiwaza is Running

Open your web browser and navigate to:

• Frontend: https://localhost
• API Documentation: http://localhost/api/docs

You should see the Kamiwaza interface load successfully. Use the following credentials in the Sign In screen:

• Username: admin
• Password: kamiwaza

Step 2: Deploy Your First Model

Now let's deploy a small, fast language model that's perfect for getting started.

Access the Models Section

1. In the Kamiwaza frontend, navigate to the Models section

2. Enter unsloth/Qwen3-0.6B-GGUF into the search field, click Exact match, then click Search.

   • Click the "Download" button, unselect all the files, just select the box next to Qwen3-0.6B-Q6_K.gguf - 495.11 MB
   • Click "Download Selected Files".
   • You should now see the file downloading in a new pop-up modal.

3. Deploy the model.

   • Scroll down to the "Your Models" section, and click the name of the model (Qwen3-0.6B-GGUF)
   • On the next screen, look for the "Deploy" button in the upper right (under Model Configurations) and click it.
   • After a few moments, you will see the model appear under "Model Deployments."

Review

For the first deployment, we deployed a modern large language model:

• Model: Qwen3 0.6B (very light weight, suitable for basic conversation)
  • Note that this is a GGUF-quantized version of the model suitable for CPU-based inference
• Engine: llamacpp (CPU-friendly)

Step 3: Launch an AI Chatbot from the App Garden

Now let's use your deployed model in a real application from the App Garden.

Access the App Garden

1. Navigate to the App Garden section in the Kamiwaza interface
2. Browse the available applications
3. Look for the "AI Chatbot" app

[Screenshot placeholder: App Garden interface showing available apps]

Configure and Launch the Chatbot

1. Click on the AI Chatbot app's "Deploy" button
2. In the configuration screen, keep the default configuration, and click "Deploy"
3. Click "Launch App"

Access Your Running Chatbot

• Once launched, click the "Open App" button in the deployed app listing,
• This will open the app in a new browser tab.
• It will automatically use your deployed model.

Try Your Chatbot

You now have a fully functional AI-powered chat application!

Next Steps

Congratulations! You've successfully:

✅ Started Kamiwaza

✅ Deployed your first AI model

✅ Launched a real application from the App Garden

Explore More

Now that you have the basics down, here are some next steps to explore:

• Try Different Models: Deploy larger, more capable models
• Explore More Apps: Check out other applications in the App Garden
• Learn the Architecture: Understand how Kamiwaza works under the hood
• Use the SDK: Build custom applications using the Kamiwaza Python SDK
• Set Up Your Data: Connect your own data sources for RAG applications

Need Help?

If you run into any issues:

• Check the troubleshooting section for common problems
• Join our Discord community for real-time help
• Contact our support team

What You've Learned

In this quickstart, you've experienced the core Kamiwaza workflow:

1. Model Management: How to deploy and test AI models
2. App Garden: How to launch pre-built applications
3. Integration: How models and apps work together seamlessly

This same pattern scales from simple chatbots to complex enterprise AI applications. Welcome to Kamiwaza!