# Pirarara AI — Unraid Setup Guide

## Prerequisites
- Unraid 6.12+
- Nvidia Driver plugin installed (for GPU support)
- Community Apps plugin installed

## Step 1 — Install Ollama
1. Go to Apps → search "Ollama" → install ich777's template
2. Set container name: `pirarara-ollama`
3. Port: `11434`
4. Data path: `/mnt/user/appdata/pirarara/ollama`
5. Extra Parameters: `--runtime=nvidia`
6. Add variable: `NVIDIA_VISIBLE_DEVICES` = `all`

## Step 2 — Pull a model
Open Unraid Terminal and run:
```bash
docker exec -it pirarara-ollama ollama pull qwen2.5-coder:7b-instruct-q4_K_M
```

## Recommended models by VRAM
| VRAM   | Model                              |
|--------|------------------------------------|
| 6GB    | qwen2.5-coder:7b-instruct-q4_K_M  |
| 8GB    | qwen2.5-coder:14b-instruct-q4_K_M |
| 16GB+  | qwen2.5-coder:32b-instruct-q4_K_M |