LoRA of Gold City (umamusume)

MAYBE NOT COMPATIBLE WITH AOM3A2, AOM3A3 model.

Recommend options

• LoRA weight 0.7~1.0

• Trigger words

  • base

    • gold city \(umamusume\)

  • Racing competition clothes :

    • parted lips, black choker, gold chain, single glove, white gloves, long sleeves, bandeau, cleavage, strapless, tube top, navel, midriff, open jacket, open clothes, black jacket, black belt, denim shorts, cutoffs, mismatched footwear

  • Festival costume :

    • hair ornament, hair flower, gloves, japanese clothes, detached sleeves, wide sleeves, frilled sleeves, kimono, , official alternate costume, sash, nail polish, high heels, floral print, obi

    • (optional) folding fan

  • Casual clothes :

    • Because of lack of images, not included in here.

  • Tracen school uniform :

    • white thighhighs, tracen school uniform, purple shirt, pleated skirt, puffy short sleeves, white skirt, puffy sleeves, summer uniform, frilled skirt, sailor collar, sailor shirt, miniskirt, frills

Use animefull based model

• Settings

  • Stable Diffusion WebUI by Automatic1111

  • DPM++ Series(SDE Karras, 2M Karras, etc.)

  • about 20 steps, CFG scale 3.5~6.5

  • kl-f8-anime2.vae

  • CLIP skip = 1

  • Use highres. fix to get higher quality image

    • Upscaler : Latent series

    • Denoising strength 0.50~0.65

LoRA training info

• trained on sd-scripts by kohya_ss. Thank you a lot!

• based on Animefull-pruned

• 118 images of gold city

  • 4x low quality 41 images

  • 8x medium quality 32 images

  • 16x high quality 45 images

  • => 1 epoch = 1140 images

• replaced character feature tags with gold city \(umamusume\)

  • horse girl, horse tail, blonde hair, blue eyes, etc.

• resolution : 768x768

• rank=alpha=16

• batch size = 4, 4 epochs trained

• other settings included in training_data.zip

This LoRA may be compatible with animefull-based models (ex. AbyssOrangeMix Series...)

This LoRA can generate nsfw images of gold city.

All uploaded images are generated by AOM3A1