abbuibuibui/ChikaFujiwara-Dataset

Chika Fujiwara (Kaguya-sama) — Illustrious SDXL Character LoRA Training Dataset

中文说明 | English (current)

This is the actual training subset used to train abbuibuibui/ChikaFujiwara-Lora, an unofficial Chika Fujiwara character LoRA on waiIllustriousSDXL v17. The files here are a copy of the directory named in dataset.toml (image_dir = .../_扩充终选_待打标_20260908/藤原千花). Captions were not rewritten for this release. Latent caches, WD14 logs, and review sheets that sat in the same folder are not included.

This is a fan-made derivative, not an official product. Rights to the character and source work remain with their respective owners.

Contents

Item	Count
Training images (images/*.png)	42
Matching captions (images/*.txt)	42
Missing same-name captions	0
Orphan captions	0
num_repeats	4

Filenames are unchanged. Captions are comma-separated Danbooru-style tags with the identity and outfit triggers fixed first:

chika_fujiwara, chika_school_uniform, 1girl, solo, open mouth, closed eyes

Outfit token	Images
chika_hoodie	17
chika_school_uniform	12
chika_casual	12
chika_dress	1
Total	42

• Identity trigger: chika_fujiwara (pink hair, blue eyes, side ponytail, black hair bow)
• Outfit tokens: chika_hoodie / chika_school_uniform / chika_casual / chika_dress
• Training used shuffle_caption = true and keep_tokens = 2, so the first two tokens stay pinned.
• Captions in this set are sparse (few hair/eye tags). The LoRA card adds those identity clues at eval time; they were not inserted into these .txt files.
• images/metadata.jsonl maps all 42 filenames to those captions for the Dataset Viewer.

Preview

These eight images are copies of real training stills, not model outputs.

Hoodie	Hoodie	Hoodie

School	School	School

Casual	Casual

How it was trained

• Base: waiIllustriousSDXL v17, 1024 buckets, bucket_no_upscale
• Network: LoRA rank 32 / alpha 16, UNet plus both text encoders
• Optimizer: AdamW8bit; UNet LR 1e-4, text-encoder LR 5e-5; cosine, 7 epochs per round
• 42 × repeats 4 = 168 optimizer steps per epoch; last resume round 1176 steps
• Recommended model from that run: R4E6 (chika_fujiwara_rE6S951-000006.safetensors) at strength 0.8

Related repositories

• Model: abbuibuibui/ChikaFujiwara-Lora — all unique epoch weights, bilingual card, and the public epoch × strength evaluation.

License

• Character and source-work rights remain with © Aka Akasaka / Shueisha / the relevant production committees.
• Dataset files are released under CreativeML OpenRAIL-M.
• Intended for non-commercial fan LoRA training and research.
• Do not use this dataset for unlawful, deceptive, harassing, or exploitative purposes.