CanonLora - Photorealistic Portrait LoRA

High-quality photorealistic LoRA trained for consistent face/identity generation. 

**Best Results With:**
- Juggernaut XL V9 (recommended)
- RealVisXL V4
- Any photorealistic SDXL checkpoint

**Trigger Words:** pairlora, ohwx

**Recommended Settings:**
- LoRA Strength: 0.7-0.85
- CFG Scale: 5-7
- Sampler: DPM++ 2M Karras
- Steps: 25-35

**Tips:**
- Place trigger words early in prompt for stronger effect
- Works best with RAW photo style prompts
- Stack with detail/enhancement LoRAs at lower weights
- Use ADetailer for face consistency at high resolutions