[Pony & IL] [B-LoRA] Chloe Welter クロエ・ヴェルター | Let This Grieving Soul Retire | Nageki no Bourei wa Intai shitai 嘆きの亡霊は引退したい

Chloe Welter from Let This Grieving Soul Retire (Nageki no Bourei wa Intai shitai 嘆きの亡霊は引退したい)

This model was trained on anime screencaps (up to Episode 11) only.

Changelog:

• Added a version trained on Illustrious XL v0.1.

Note:
This is a B-LoRA. Use the style LoRA to adjust the intensity of the anime style.

Prompts:-

• Trigger Words:

chloe_olis, 1girl, black hair, high ponytail, brown eyes, eyes visible through hair

• Outfit:

white bow, military uniform, blue jacket, white shirt, red bowtie, brown belt, blue skirt, black pantyhose

• Negative Prompts (optional):

blurry background
brick wall

Weights:

Character LoRA: 1
Style LoRA:     0     for a neutral style
                >=~8  for a more accurate anime style and outfit

The anime style works best with: (not recommended, just tested)
PonyXLv6 & AutismMix SDXL                 for PonyXL LoRA
Illustrious-XL v0.1 & NoobAI XL variants  for Illustrious-XL v0.1 LoRA

(Feel free to experiment with the weights.)

Eyes Inpainting: (for improving the quality of anime eyes, intended for use with the style LoRA)

• Use ADetailer with this model: https://civitai.com/models/330727/full-eyes-detection-adetailer
  OR manually mask her face (i.e. eyebrows, eyes, nose and mouth) and set Inpaint area: Only masked
  (The second method seems to work better due to how I prepared my dataset for portrait images.)

• Width/Height: 1024

• Padding Pixels: 0-32

• Steps: ~30

• Denoising Strength: no idea (I usually do a X/Y/Z plot with a value of 0.4-0.7(+0.1) and then choose the best one.)

• Positive Prompt: add both LoRA and brown eyes, eyes visible through hair, eye focus, close-up (and optional looking tags such as looking at viewer)

Known Issue:

• Minor details of the outfit are not accurate, since I didn't want to further overtrain the model. (I don't think many people care about this nowadays.)

Afterword:
This time, I decided to tag the outfit with the color variant instead of my usual approach, i.e. using white bow instead of hair bow. The reason is that, for a B-LoRA, the color is learnt in the style LoRA. This way could make the character more recognizable when using character LoRA alone.
I am also considering experimenting with IllustriousXL on this character next since I like her very much. However, I haven't decided which model to train on yet, perhaps NoobAI Eps v1. I just trained one on Illustrious XL v0.1. For my next models, I might stick to Illustrious XL or NoobAI and abandon PonyXLv6, since they handle some concepts better.
Life is going to get busy for me in 2025, as I really need to focus on getting a proper job. Sadly, that's just how life is.

Hopefully some of you will like this model.