Overview

LUMIERE is the spiritual successor to Stelra, which was based on SD 1.5 and combined various models. Built on SDXL 1.0 and fine-tuned from the ground up, LUMIERE is one of the best models for generating high-quality anime-style images.

Following the structure of popular and widely used community models, LUMIERE is easy to use even for first-time users. In addition, it has been trained on a vast amount of open-source anime datasets as well as high-quality proprietary datasets, making it safe and reliable for anyone to create stunning anime-style images with confidence.

Disclaimer

• The LUMIERE model and any derivatives thereof may be used freely by users for their own creative purposes. However, all copyrights, legal responsibilities, and social or ethical consequences arising from any images, videos, or other outputs generated using this model remain solely with the user.

• The creation and use of both SFW (Safe For Work) and NSFW (Not Safe For Work, including but not limited to sexually explicit, violent, gory, or otherwise disturbing content) outputs are entirely at the discretion of the user. The developer and distributor of this model bear no responsibility or obligation regarding such content.

• By using this model, the user agrees to take full and exclusive responsibility for any legal disputes, social issues, infringement of third-party rights, defamation, or unlawful acts that may arise. The developer, distributor, and maintainer of this model accept no civil or criminal liability in any such cases.

• The user must comply with all applicable laws and ethical standards in their jurisdiction when using this model. Any consequences resulting from violations of such laws or standards shall be borne solely by the user.

• Under no circumstances shall the developer or distributor of this model be liable for any direct, indirect, incidental, special, or consequential damages arising out of the use or inability to use this model.

Model Details

• Model Name : LUMIERE

• Developer : BACKGWA

• Fine-tuned : Stable Diffusion XL 1.0

• License : CreativeML Open RAIL++-M

Prompt Guide

LUMIERE uses prompting techniques and methods commonly employed with the most widely used SDXL models. Please refer to the prompting techniques below.

• Prompt Structure

  (GENDER), CHARACTER NAME, SITUATION OR OTHER TAGS, QUALITY TAGS

• Quality Enhancement Prompt

  masterpiece, high score, great score, absurdres

• Negative Prompt

  lowres, bad anatomy, bad hands, text, error, missing finger, extra digits, fewer digits, cropped, worst quality, low quality, low score, bad score, average score, signature, watermark, username, blurry

Recommended Setting

Sampling method : Euler a
Sampling steps : 20 ~ 40
CFG : 5 - 7

Recommended Resolutions

- 768 x 1344 : Vertical (9:16)
- 915 x 1144 : Portrait (4:5)
- 1024 x 1024 : Square (1:1)
- 1182 x 886 : Photo (4:3)
- 1254 x 836 : Landscape (3:2)
- 1365 x 768 : Widescreen (16:9)
- 1564 x 670 : Cinematic (21:9)

Prompt / Setting Example

(1girl), march 7th \(honkai: star rail\), solo, long hair, open mouth, sleeping, dim light, masterpiece, best quality, highres, absurdres, newest

lowres, bad anatomy, bad hands, text, error, missing finger, extra digits, fewer digits, cropped, worst quality, low quality, low score, bad score, average score, signature, watermark, username, blurry

Resolution : 1200*504
Sampling method : Euler a
Sampling steps : 40
CFG : 5