MyStyle v2 is a stylized anime portrait LoRA trained for Stable Diffusion 1.5.

It focuses on clean facial features, expressive eyes, smooth shading, and a polished illustration look.

This LoRA works best for close-up portraits and half-body images and is intended for stylized, anime-inspired artwork rather than photorealism.

Recommended use cases:

- Anime-style female portraits

- Clean illustration art

- Character-focused compositions

Not recommended for:

- Photorealistic images

- Background-heavy scenes

- Realistic anatomy studies

The model was trained on original and curated datasets and does not depict any real person.

Usage Tips:

- Base Model: Stable Diffusion 1.5

- LoRA Strength: 0.7 – 1.0 (recommended: 0.8)

- Sampler: Euler a / DPM++

- Steps: 20–35

- CFG: 6–8

Trigger words:

- mystyle