configuration.md

⚙️ Configuration Guide

🐳 Docker Installation

Prerequisites

• Complete the standard installation and setup first
• Configure your LoRAs using the Lora_editor tool
• Docker and Docker desktop installed on your system

🤖 Discord Bot Setup

1️⃣ Creating Your Discord Application

1. Visit the Discord Developer Portal
2. Create New Application:
   • Click "New Application"
   • Name your bot
   • Navigate to "Bot" tab
   • Click "Add Bot"

2️⃣ Bot Configuration

1. Essential Bot Settings

   • Copy Bot Token (keep secure!)
   • Enable Required Intents: ✓ Presence Intent ✓ Server Members Intent ✓ Message Content Intent

2. Bot Permissions Setup

   • Go to "OAuth2" tab
   • Select "bot" under Scopes
   • Required Permissions:

     ✓ Send Messages
     ✓ Manage Messages
     ✓ Embed Links
     ✓ Attach Files
     ✓ Read Message History
     ✓ Use Slash Commands
     

3. Server Integration

   • Copy OAuth2 URL
   • Open in new tab
   • Select target server

🛠️ Bot Configuration

Initial Setup

1. Create .env file in root directory:

   COMMAND_PREFIX=/

2. Run Setup Tool:

   python setup.py
   # or use Setup.exe

📝 Configuration Details

1. Path Configuration

   • Browse to ComfyUI models folder
   • Example: C:/Comfyui_windows_portable/ComfyUI/models

2. API Tokens

   • Civitai Token:

     1. Create account on Civitai
     2. Profile → Settings → Add API Key

   • Huggingface Token:

     1. Create account on Huggingface
     2. Profile → Access Tokens
     3. Create token with "read access to public gated repos"

3. Server Configuration

   • Bot Server Address: Usually 127.0.0.1 (try 0.0.0.0 or localhost if needed)
   • ComfyUI Server: Same as bot (use remote IP if on different machine)

4. Discord Settings

   • Server IDs: Right-click server → Copy ID (e.g., 123456789,987654321,654321987)
   • Channel IDs: Right-click channel → Copy ID (e.g., 123456789,987654321,654321987)
   • Multiple IDs: Use comma separation (e.g., 123456789,987654321)
   • BOT_MANAGER_ROLE_ID (optional): For bot administration
     • This role is not required for the bot to work. It's an optional configuration that provides an additional way to grant certain users bot management permissions without giving them full Discord admin rights.
       If you don't set it, the bot will still function normally - only Discord administrators will be able to use the administrative commands.

5. Model Selection

   • Choose checkpoint based on GPU VRAM:
     • Support for 6GB to 24GB cards
     • Select version matching your VRAM capacity

🔧 Advanced Setup

Creating Executables

If you prefer building your own executables:

# Install PyInstaller
pip install pyinstaller

# Build Setup Tool
pyinstaller setup.spec --clean --noconfirm

# Build LoRA Editor
pyinstaller LoraEditor.spec --clean --noconfirm
# requires UPX for compression or executeable will be extremely large

⚠️ Important Notes

• Required files will be automatically moved to their respective folders
• Checkpoint changes: Simply rerun setup tool
• Model changes require bot restart
• Antivirus may flag executables
  • Build your own using provided specs
  • OS-dependent compilation

Docker Setup Steps

1. Initial Build and Start

   docker-compose up --build

2. Subsequent Starts After the initial build, you can start the container with:

   docker-compose up

Troubleshooting Connection

If the bot doesn't connect to your ComfyUI instance, you may need to adjust the server address in the Dockerfile:

1. Locate Line 33 in the Dockerfile:

   ENV server_address=0.0.0.0

2. Try one of these alternatives:

   • Local connection: 127.0.0.1
   • Your network IP: LOCAL NETWORK IP ADDRESS
   • ComfyUI host IP (if running on different machine)

🎨 LoRA Models Configuration

🌟 Introduction to LoRA Models

LoRA (Low-Rank Adaptation) models are the secret sauce that brings magic to your image generations! These models can:

• Enhance image quality
• Apply specific artistic styles
• Add unique characteristics to generations
• Transform your prompts into stunning visuals

📥 Getting LoRA Models

1. Visit Civitai.com/models
2. Apply filters:
   • Model Type: LoRA
   • Base Model: Flux.1s and Flux.1 D
3. Download your chosen models
4. Place files in: comfyui/models/Lora folder

🛠️ LoRA Editor Tool

python lora_editor.py
# or build your own exe using the .spec file and pyinstaller

✨ Latest Updates

• 🔗 URL links support for /lorainfo command
• 🚀 Support for 600+ LoRAs
• 💾 SQLite3 database for persistent storage
• 🎯 Persistent trigger words
• 🤗 Huggingface support
• 🤗 Civitai support
• 🔄 Reset All Loras
• ⚡ Instant LoRA availability
• 📋 Lora arrangement abilities

🚀 Getting Started

1. Launch the editor:

   python lora_editor.py
   # or build your own exe using the .spec file and pyinstaller

🎯 Key Features

• 🔄 Auto-download LoRAs from URLs
• 🎯 Automatic trigger word population
• ⚖️ Pre-configured weights (default: 1.0) -if lora weight is less than 1, keep it do not change
• 📊 Easy weight customization
• 📋 Flexible list organization
• 🔗 Activate and deactivate loras

📝 Manual Configuration

If you prefer manual setup, edit datasets/lora.json:

{
  "name": "Your LoRA Name",
  "add_prompt": "required_trigger_word",
  "file": "MyLoRA.safetensors",
  "weight": 1.0,
  "url": "https://civitai.com/api/download/models/123456789/MyLoRA.safetensors"
}

💡 Best Practices

1. Weight Management

   • Default: 1.0
   • Adjust lower for subtle effects
   • Test different values for optimal results

2. Organization

   • Use descriptive names
   • Editing the trigger words may result in loras not working
   • URL is auto populated.

🔍 Troubleshooting

• Verify file extensions (.safetensors)
• Confirm trigger words
• Test API connectivity

🔧 Bot Configuration

Environment Variables

All environment variables are auto created when setup is ran

Server Configuration

1. Enable --listen on ComfyUI server
2. Configure port settings (default: 8188)
3. Set up SSL if needed (recommended for production)