CivArchive
    Illustrious Rabbit - v0.7
    ```I am excited to announce the release of my newest model, Harlequin Rabbit: the first model in my new Leporid model series``` **_Check it out here:_** https://yodayo.com/models/a101487c-756f-4828-b981-ccda296e54d4?index=0&modelversion=4b58417c-bb79-4ea5-bf6f-4cb0cc086650 ```I'm please to finally announce that Illustrious Rabbit v1.0 is finally here! The beta versions 0.3 and 0.7 will remain available for use. The only place you will find this model is right here on Moescape/Yodayo This is the first model I have uploaded. It's a detailed merge of several of my favorite models, plus a couple of models that had a really nice style I rather enjoyed. The base style produced without LoRAs is a subtle look with soft shadows and smooth colors. Certain elements like highlights, blushing, and eyes have an almost hand drawn look to them. The simple aesthetic looks great by itself, and it's very flexible with additional style tags and has a wide range of LoRA compatibility. I strongly encourage experimentation with quality, genre/them, era, and artist style tags. For more information on prompting and changes in each version, I strongly suggest reading the version information. With each new iteration, I sought to expand on the capabilities and versatility while preserving the base aesthetic. The end result is a model made with about 10 other models including Illustrious-XL. What's next? I have begun work on a NoobAI-XL variant, which I hope to have up in the next month or two. Please feel free to reach out to me on Discord or tag me in comments if you have any feedback.```

    Description

    I am very pleased to finally release the next public version of Illustrious Rabbit, v0.7! Please take note, this is not a full release just yet, I am still refining it and perfecting certain areas. =Ideal Settings= Sampler: Euler A, DPM++ 2M or SDE Karras Steps: 20 to 35 Scale: 5 to 6.5 Upscaler: 4x-AnimeSharp, 4x-UltraSharp, 4x_BooruGan_650K Hires Steps: I use between 10 and 15 most of the time Denoising Strength: 0.15 - 0.25 NOTE: If you are getting distorted elements like a second belly button, lower either your total sampling steps or your denoising strength =Details= For this version I focused on generating a wider range of species for furry and monster girls, including various dog and cat breeds as well as improving on Natural Language Prompting. This means that you can mix and match Danbooru keywords/tags with a more natural way of speaking and still achieve fantastic results. For furry or monster girls: Add the correct keywords for what you desire, such as 'furry, anthro' or 'monster girl, anthro' to the main prompt, if needed put 'human' into the negative prompt. I've done my best to improve on these features without sacrificing the core style of the model. As I said, these are still early versions and I welcome any feedback.

    FAQ

    checkpoint
    Illustrious

    Details

    Downloads
    9,185,149
    Platform
    Yodayo
    Platform Status
    Available
    Created
    3/31/2025
    Updated
    5/19/2025
    Deleted
    -

    Files

    1c44cb93-342c-41f8-8ae6-228f8d068251.safetensors

    Size:
    6.46 GB
    SHA256:

    Mirrors