This LoRa is quite flexible, it works in both realism and anime, but if what you want is to easily obtain all the details of her clothing, I recommend using semi-realistic or anime models. If you use realistic models you need a little luck.

Models Tested

Example 1 : "XXMix_9realistic_v4.0"

Example 2 : "blue_pencil_realistic-v1"

Example 3 : "hassakuV1_3"

Example 4 : "majicMIX_sombrev2"

----------------------------------------------------------------

Trigger : astagc
Recommended Prompts : braid, one side up, blue eyes, pink hair.

*For clothes you don't need any prompt, just put the trigger and the magic will be done

Weight : 1

----------------------------------------------------------------

🧐Recommended settings (The examples were created with this configuration, if you prefer another, no problem.)

VAE: For realistic and semi realistic use "vae-ft-mse-840000-ema-pruned.ckpt"

If you will make anime use "kl-f8-anime2.ckpt" "YOZORA.vae" etc

Sampler: DPM++ SDE Karras 🟢 DPM++ 2M Karras

Steps: 20 🟢 30

CFG scale: 6

Clip skip: 2

Upscaler: The one you like the most (I use 4x-UltraSharp, but it's not relevant).

Hires steps: 20 🟢 30

Denoising strength: 0.4 🟢 0.5

-------------------------------------------------------------------------

Any questions, queries or comments you may have, please do so with confidence, I will read them and respond as soon as possible.

Mi idioma natal es el español, asi que tambien doy soporte en esta hermosa lengua.

My Twitter : https://twitter.com/GreenFkingCat

Asta (Honkai: Star Rail) / Flexible LoRa - v1