abbuibuibui/Aqua-Dataset

Aqua (KonoSuba) — Illustrious SDXL Character LoRA Training Dataset

中文说明 | English (current)

This is the actual training subset used to train abbuibuibui/Aqua-Lora. Files are a copy of the four image_dir folders named in dataset.toml under .../konosuba/_final_lora_dataset_20260912/Aqua/. Captions were not rewritten. Latent caches, WD14 logs, and review sheets 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)	60
Matching captions (images/**/*.txt)	60
Missing same-name captions	0
Orphan captions	0
num_repeats	3

Folder names are kept: Standard_Outfit, White_Dress, Travel_Cloak, Sleepwear. Captions start with identity and outfit:

aqua_konosuba, aqua_konosuba_standard_outfit, 1girl, solo, closed eyes, lying, anime screencap

Outfit token	Images
aqua_konosuba_standard_outfit	40
aqua_konosuba_white_dress	8
aqua_konosuba_travel_cloak	7
aqua_konosuba_sleepwear	5
Total	60

• Identity trigger: aqua_konosuba (long blue hair, blue eyes, hair bobbles, large hair bow)
• Training used shuffle_caption = true and keep_tokens = 2
• images/metadata.jsonl uses paths like Standard_Outfit/774.png relative to images/

Preview

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

Standard	Standard	Standard

White dress	White dress	Cloak

Cloak	Sleepwear

How it was trained

• Base: waiIllustriousSDXL v17, 1024 buckets
• 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
• 60 × repeats 3 = 180 optimizer steps per epoch; last resume round 1260 steps
• Recommended model: R2E6 (aqua_rE5S846-000006.safetensors) at strength 0.8

Related repositories

• Model: abbuibuibui/Aqua-Lora

License

• Character and source-work rights remain with © Natsume Akatsuki / Kurone Mishima / Kadokawa / 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.