Generates colourful squids and octos.
This model will help SDXL-based checkpoints draw:
in a style reminiscent of Splatoon concept and promotional 2D artwork.
concepts from Splatoon, including Inklings and Octolings.
Recommended checkpoints:
kohakuXL works well, and generally balances creativity against a neutral art style.
blue_pencilXL and CounterfeitXL work well, and introduce more noticeable idiosyncrasies.
Art style notes:
It may help to prefix a prompt with "copic marker style" to replicate Splatoon concept art style.
CLIP skip 2 can help to reduce chaos and detail from scenes.
Adding keywords like "highly detailed, masterpiece, highest quality, etc." as recommended by checkpoints may steer the network into drawing with more human-like body proportions.
Different samplers tend to produce different styled results, particularly with body proportions and scene ambience.
Negative prompt:
I have found "photorealistic, 3d model, bad, worse, worst, ugly, bad anatomy, blurry, close-up, disembodied limb" (taken from kohakuXL page) to be sufficient for most generations.
Adding negativeXL_D may help in some cases.
This model is a bit temperamental, however it should be fairly usable at this point. Have fun!
v2.0
Primary focus of changes were to improve generated Octoling accuracy, and general control over Inkling/Octoling features.
Cleaned up tagging.
Added more images to dataset.
Tuned training parameters.
v1.0
Initial release.
This model was trained against official Nintendo Splatoon artwork. Art was selected by hand to match a particular visual style (highly saturated flat-shaded art).
See the "About this version" card for information about the dataset and training parameters. Images, tags, and training configuration are provided in an optional zip file along the model.
This work was inspired by the following linked model by IWannaGoBack, which targets a similar niche for SD 1.5 based networks.
https://civitai.com/models/85425/splatoon-style
Thanks to IWannaGoBack. I enjoyed playing with your model, and it inspired me to dive into the deep rabbit hole of LoRA training.
The author of this generative AI model (model) and curator of its training database (database) releases rights and disclaims liabilities that pertain to the model/database themselves or usage thereof, under the mark of the CC0 1.0 legal code.
http://creativecommons.org/publicdomain/zero/1.0
This is a fan work that derives from intellectual property owned by Nintendo Co., Ltd. It is not endorsed by Nintendo.