koeri

This model is an aesthetic fine-tune of Illustrious XL, with bases in its v0.1, v1.0, and v2.0 versions, specialized for anime-style generation. It is designed for artist-style and tag-style prompting of characters, clothing, and scenes. The model uses v-prediction with ZTSNR and was trained on thousands of curated images using Uniform LogSNR sampling. It remains LoRA-compatible with its Illustrious bases.

---

Inference

• Tag-style: Model was finetuned with general/character tags, comma-separated
• Guidance: 4–6 is a good starting band
• Steps: 20–30 steps
• Resolution: Follow base SDXL/Illustrious defaults
• Styles: LSNet along with the Kaloscope model can query artist tags from any image
• Tools: Custom node pack

---

Training (v1)

• Base model: latent-space-dreams/LS_Pingo_XL
• Source: Illustrious XL merged variants (0.1 / 1.0 / 2.0)
• Sampling: Uniform LogSNR
• Schedule: v-prediction & ZTSNR
• Dataset: ~10,000 curated SFW/NSFW anime images
• Trainer: sdxl-trainer (Modified from Aozora)

Training was performed in three passes;

1. around four epochs.
2. one epoch at a lower learning rate.
3. short refinement pass on a curated subset of ~3500 images.

---

Acknowledgements

• Illustrious XL authors
• latent-space-dreams/LS_Pingo_XL as the base