TAMGAYAR LoRA v1.0 - User Guide

Bald Character & Stylized Portrait Model

🛠 Installation

Stable Diffusion Integration

• Place tamgayar.safetensors in:

  • Automatic1111: /models/Lora/

  • ComfyUI: /comfy/models/loras/

• Refresh model lists in your UI.

🎯 Optimal Usage

Base Model Compatibility

• Best Results: RealisticVision V6 or EpicRealism V5

• Avoid: Anime-focused models (e.g., AnythingV4).

Trigger Word

• Mandatory: Start prompts with TAMGAYAR (uppercase).

• Weight Tuning: Adjust with (TAMGAYAR:1.2) for intensity.

🧬 Key Parameters

ParameterRecommended RangeEffectDenoising0.5-0.7Balances LoRA influenceCFG Scale7-12Higher = stricter styleSteps25-35DPM++ 2M Karras samplerResolution1024x1024Avoids feature bleeding

🎨 Sample Prompts

python

Copy

# Mythic Warrior  
"TAMGAYAR, (molten obsidian armor:1.3), glowing broadsword, divine aura particles, shattered battlefield, artstation, 8k cinematic"  

# Sin City Noir  
"TAMGAYAR, (B&W contrast:1.4), crimson blood splatter, rain-drenched trenchcoat, Frank Miller halftones, speech bubble"  

# Renaissance TikToker  
"TAMGAYAR, holographic AR filters, Da Vinci sfumato, Vatican fresco backdrop, trending #RenaissanceCore hashtags"  

# Metal Concert  
"TAMGAYAR, (stage dive:1.2), laser grid, Marshall stack glow, Canon 5D Mark II stage photo, sweat droplet closeups"  

⚠️ Troubleshooting

IssueSolutionStyle BleedingLower LoRA weight (<1.1)Flat LightingAdd cinematic rim lightingOverfittingReduce steps (<25)Anatomy ErrorsUse (anatomical precision:1.3)

🔀 Advanced Blending

Combine with other LoRAs using bracketed syntax:

Copy

TAMGAYAR, (wearing cybernetic armor:1.1), (in Neo-Tokyo:0.9), <lora:cyberpunk_city:0.7>  

📜 Legal & Ethics

• Commercial Use: Allowed with attribution.

• Restrictions: No NSFW/misinformation outputs.

• Consent is Required if training data includes real persons.