Sign In

[Pony] Raora Panthera

18
111
55
4
Verified:
SafeTensor
Type
LoRA
Stats
111
55
34
Reviews
Published
Jan 3, 2025
Base Model
Pony
Trigger Words
raora1st, panther girl, animal ears, animal ear fluff, blush, pink hair, streaked hair, large breasts, long hair, yellow eyes, goggles, black tube top, black hat, white thighhighs, cleavage, ascot, bare shoulders, thigh strap, white shorts, white vest, gloves, pink jacket, ankle boots, white boots, panther tail
raorapanthera, panther girl, panther tail, animal ears, animal ear fluff, blush, pink hair, streaked hair, large breasts, long hair
Hash
AutoV2
14EA8631C6
default creator card background decoration
vjumpkung's Avatar
vjumpkung

Raora Panthera

Disclaimer

  • Do not use my LoRA to produce AI image and tagging her fanart hashtag in Twitter/X or if you have an enough contribution in that image then fine.

trigger word

debut costume

raora1st, panther girl, animal ears, animal ear fluff, blush, pink hair, streaked hair, large breasts, long hair, yellow eyes, goggles, black tube top, black hat, white thighhighs, cleavage, ascot, bare shoulders, thigh strap, white shorts, white vest, gloves, pink jacket, ankle boots, white boots, panther tail

any costume

raorapanthera, panther girl, panther tail, animal ears, animal ear fluff, blush, pink hair, streaked hair, large breasts, long hair

Limitations

  • still cannot remove jacket in debut costume

  • clothing details in debut maybe inaccurate

  • newyear costume is not working (only twintails is working)

Training Details

LoRA rank

  • reduced dimension to 8 with sv_fro=0.9

dataset

  • 55 images debut costume

  • 20 images others

  • 7 images new year costume (not working)

parameters

  • resolution = 1024

  • batch size = 2

  • dim,alpha = 16,16 (for training)

  • mix/save precision = bf16/bf16

  • optmizer = AdEMAMix + weight_decay=0.025 betas=0.9,0.99,0.9999

  • UNet LR = 2e-4

  • TE LR = 1e-4

  • scheduler = cosine_with_min_lr (min_lr_ratio 0.33)

  • huber snr 0.85

steps

  • epochs = 10

  • total steps = 1975

  • repeat = 3,8,10

tools

  • kohya-ss GUI v24.3.0 FORKED by vjumpkung

  • torch 2.5.0 cu124

  • RTX 3060 12 GB + xformers + gradient_checkpointing

avg weight

  • UNet average weights : 0.013589977964081545

  • TE1 average weights : 0.011423828994641868

  • TE2 average weights : 0.009651259635575112

*This LoRA is for studying LoRA training with new technique so do not use for damaging the vtuber (also support her too).

*If you see my model reposted on other image generation service please report to me.