My LoRA Model
This LoRA was trained on custom images to capture a specific style/subject.
Model Details
- Base Model: Stable Diffusion 1.5
- Training: 500 steps with 20 images
- LoRA Rank: 16
- File Size: ~13MB
Usage
In ComfyUI:
- Download the
.safetensors file
- Place it in
ComfyUI/models/loras/
- Use the "Load LoRA" node
- Set strength between 0.7-1.0
In Automatic1111:
- Download the
.safetensors file
- Place it in
stable-diffusion-webui/models/Lora/
- Add to your prompt:
<lora:filename:0.8>
Example Prompt:
a photo of [your subject], high quality, detailed <lora:my_lora:0.8>Recommended Settings
- LoRA Strength: 0.7 - 1.0
- CFG Scale: 7-9
- Steps: 20-30
- Sampler: DPM++ 2M Karras or Euler a
Tips
- Start with strength 0.8 and adjust from there
- Lower strength (0.5-0.7) for subtle effects
- Higher strength (0.9-1.0) for stronger style adherence
- Works best with SD 1.5 based models
License
This model is licensed under the CreativeML Open RAIL-M license.
