VentusMix

CHECKPOINT
Original


Updated:

This is an illustrious anime model, focused on consistency and detailed backgrounds with a painting-like artstyle. It’s very easy to control and will yield good results without many quality tags. Works best for generation of the scenes requiring a high level of small details.

 V2 should be a decent improvement across the board, it has better prompt adherence, colors and nicer faces. It needs a bit lower CFG than V1.

Short usage guide

Use DPM++ 2M SDE, with CFG 2-3, 20-40 steps.

Hires fix is not required, but ADetailer is probably preferable.

 Prompts

 I use two sets of prompts with nearly the similar frequency.

  • First one allows to achieve the best lighting and detail, but not every image really needs it and excessive quality tag prompting can degrade the prompt adherence:

masterpiece, best quality, amazing quality, very aesthetic, absurdres,newest,detailed face,(volumetric lighting)

Negative

lowres, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, signature, watermark, old, oldest,simple background

  • Second one is simple, almost pure model output with two quality tags to fix small details:

masterpiece

Negative

worst quality

Also you can add this in negative as you see fit because I had problems with characters having longer than usual fangs or random sweating:

fangs,sweat,sharp teeth

Base models used in this merge:

NoobIEater by jyrrata https://civitai.com/models/879640?modelVersionId=1034051

NTR MIX by tkio https://civitai.com/models/926443?modelVersionId=1061268

Version Detail

Illustrious
Improved prompt adherence, better colors and faces. CFG 2-3

Project Permissions

    Use Permissions

  • Use in TENSOR Online

  • As a online training base model on TENSOR

  • Use without crediting me

  • Share merges of this model

  • Use different permissions on merges

    Commercial Use

  • Sell generated contents

  • Use on generation services

  • Sell this model or merges

Related Posts