Alexis

This LoRA was trained on a custom dataset to create a consistent character called Alexis, a young woman with a slim body, toned figure, and natural beauty.
It works well for:

• Semi-realistic & realistic portraits

• Full body shots (bikini, casual outfits, lingerie, etc.)

• Social media style photos (Instagram, Fanvue, OnlyFans aesthetics)

• Creative scenes with controlled style and stable face consistency

Strengths:

• Keeps face consistency across different themes.

• Adapts well to SDXL and semi-realistic checkpoints.

• Handles both close-up portraits and wide full body compositions.

Limitations:

• Not designed for anime/cartoony checkpoints.

• Works best with SDXL Standard (other bases may reduce accuracy).

• Avoid over-training prompts with too many style tags (can distort face).

Recommended use:

• Resolution 1024x1024 or higher

• LoRA weight between 0.6 – 0.9 for most prompts

• Add simple quality tags like (masterpiece, best quality) for best results