CivArchive
    Preview 6673179
    Preview 6673176
    Preview 6673178
    Preview 6673181
    Preview 6673198
    Preview 6705114
    Preview 6705113
    Preview 6705117
    Preview 6705115
    Preview 6705116
    Preview 6706405
    Preview 6706404
    Preview 6706399
    Preview 6706398

    -1.5 Update-

    Sampling Step 10 ~ 20, CFG scale 1.5 ~ 3, Euler a, SGM Uniform

    Trigger words for persona style: persona

    At the positive prompt: masterpiece, best quality, amazing quality, very aesthetic, absurdres

    At the negative prompt: lowres, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, old, oldest, censored, bar_censor, copyright_name, (dialogue), text, speech bubble, Dialogue box, error, fewer, extra, missing, worst quality, low quality, watermark, signature, extra digits, username, scan, abstract, multiple views, (censored),worst quality, low quality, logo, bad hands, mutated hands,

    Nothing special, just a little bit of light source modification.

    It might have made it easier to use other artist prompts and LORA by reducing the personality of the basic drawing style. Or not.

    By the way, I wanted to use a Persona-style image as the cover, but it didn't pass, so I'm using a Pokémon-style image as the cover.

    -ilxl1.0 + NOOB1.1 Update-

    Sampling Step 10 ~ 20, CFG scale 1.5 ~ 3, Euler a, SGM Uniform

    Trigger words for persona style: persona

    At the positive prompt: masterpiece, very awa, best quality, amazing quality, very aesthetic, absurdres, newest, intricate details

    At the negative prompt: lowres, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, old, oldest, censored, bar_censor, copyright_name, (dialogue), text, speech bubble, Dialogue box, error, fewer, extra, missing, worst quality, low quality, watermark, signature, extra digits, username, scan, abstract, multiple views, (censored),worst quality, low quality, logo, bad hands, mutated hands,

    This model is more suitable for high-resolution images by merging ilxl1.0.

    The usage method is the same as the previous model, but the trigger word has been simplified to 'Perosna'.

    -NOOB v1 Update-

    Sampling Step 10 ~ 20, CFG scale 1.5 ~ 3, Euler a, SGM Uniform

    Trigger words for persona style: persona, soejima shigenori, anime coloring

    At the positive prompt: masterpiece, very awa, best quality, amazing quality, very aesthetic, absurdres, newest, intricate details

    At the negative prompt: lowres, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, old, oldest, censored, bar_censor, copyright_name, (dialogue), text, speech bubble, Dialogue box, error, fewer, extra, missing, worst quality, low quality, watermark, signature, extra digits, username, scan, abstract, multiple views, (censored),worst quality, low quality, logo, bad hands, mutated hands,

    The Persona style models I've made so far were too characteristic to be used universally.

    However, the new models made using Noob 1.1 can be used more universally.

    Now, the characteristic drawing style of Persona doesn't appear without a trigger word.

    You can also merge DMD to quickly generate drawings. The unmerged version of DMD didn't work, so even if you don't like it, you can quickly generate drawings.

    -PONY v3 Update-

    This model is more easier to use.

    Sampling Step 20, CFG scale 4, Euler a SGM Uniform

    -PONY v2 Update-

    This model is more easier to use.

    At the positive prompt, score_9, score_8_up, score_7_up,

    At the negative prompt score_4, score_3, score_2, score_1

    Sampling Step 20, CFG scale 4, Euler a SGM Uniform

    -PONY Update-

    This is a checkpoint created based on the PONY model.

    I'm aiming for something that just takes the style of the Persona series... So you still need LORA to draw a Persona character.

    -V2 Update-

    I made this model because I was dozing off and suddenly thought of a revolutionary way to make my fingers prettier.

    I eventually realised that the idea I had while dozing off was useless, so I wasted my time! But I did make something, so here it is.

    -Hard Update-

    A hard model is one that pursues art style at the expense of most things except art style.

    It's not a good model to use on its own, but sometimes it's what you need.

    -Lite Update-

    It's a model that aims to improve prompt recognition, backgrounds, light sources, and overall things like finger representation at the expense of some art style.

    It's not a huge improvement, but hey, at least they tried, right?

    Description

    Checkpoint
    SD 1.5

    Details

    Downloads
    4,738
    Platform
    CivitAI
    Platform Status
    Available
    Created
    2/15/2024
    Updated
    3/28/2026
    Deleted
    -