LoRA of Kana Yuuki 結城夏那 (For Pony)

Description

This LoRA trained using kohya_ss on 60 images 1024×1024 up to 6360 steps without captions and with regularisation images.
Source model: ‘Source model: PonyDiffusionV6XL,’ so this LoRA should works well with most of Pony family checkpoints.
learning rate: 0.00003.
Precision: fp16.
LR Scheduler: constant_with_warmup.
LR warmup: 30%.

Preview image generation data:
Positive Prompt: score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, masterpiece, best quality, perfect anatomy, very aesthetic, a beautiful AIDA_PXLR_ykana2 in a kimono dress posing against black background, teen, asian, beautiful eyes, skin pores, cute, long falling hair, black background, portrait, beauty, makeup, hair, fashion, cute, innocent, portrait photography, studio photo, black backdrop, 1girl, solo, long_hair, brown_hair, black_hair, long_hair, hair_ornament, looking_at_viewer, lips, eyelashes, portrait, black_background, realistic, nose, 14 years old, isolated on black, diffused light, parted lips, glossy pink lipstick, pale
Negative prompt: score_6, score_5, score_4, blurry, tie, backlight, hair clips, hair_clips, hair accessory, hair_accessory, flowers, headband, bow, jewelry, text, full body, source_anime, hands, fingers, arms, source_furry, too long legs, chubby, fat, huge breasts, source_pony, source_cartoon, lowres, bad_anatomy, bad_hands, deformed anatomy, worst_quality, low_quality, normal_quality, jpeg_artifacts, signature, watermark, monochrome, very displeasing, colorful, saturated
Steps: 32,
Sampler: Euler a,
Schedule type: Uniform,
CFG scale: 5,
Seed: 534009099,
Size: 1200×1200,
Model hash: 653cf3eadd, Model: cyberrealisticPony_v63,
VAE hash: 235745af8d, VAE: sdxl_vae.safetensors,
Clip skip: 2,
Lora hashes: “AIDA_PXLR_ykana2: d98df80be3e7”.

Download link to AIDA_PXLR_ykana2.safetensors will be available for you after purchase. For your comfort I packed AIDA_PXLR_ykana2.safetensors into .zip archive with AIDA_PXLR_ykana2.png and AIDA_PXLR_ykana2.txt, preview and description are compatible with AUTOMATIC1111 extra networks menu.

Purchasing this asset you will achieve rights for selling images generated with AIDA_PXLR_ykana2.safetensors LoRA. If you want say special thanks to me, you can add additional amount to your payment.

Please contact me if you need some extended rights for sharing or selling AIDA_PXLR_ykana2, merging AIDA_PXLR_ykana2 with other models, using AIDA_PXLR_ykana2 in generation services and so on.