CivArchive
    Not-Midjourney-V3-0.3 - Not-Midjourney-V3-0.3
    Preview 7465456Preview 7465459Preview 7465457Preview 7465458

    This model is the same as https://civarchive.com/models/333491 only with 4 times the ranks and adapter size which I think is better in most cases to use. While this version can be a substitute for that one (V0.2) the V0.1 is still a distinct style itself https://civarchive.com/models/333102/not-midjourney-v3

    Recommended settings:

    CFG: 2

    Model: Dreamshapers 8

    It's
    strongly recommended that you experiment with a variety of models and settings! Change everything, sampler, CFG... and post comments! I'll be actively improving this project with that feedback!

    Use Midjourney V3 prompting for best results. Start with one word prompts like “unicorn” or “bird” then go longer.

    Tricks like unreal engine should work.

    Description

    Trained on 1000 images just like 0.2 but with 4 times dims and rank (64, 128)

    LORA
    SD 1.5

    Details

    Downloads
    172
    Platform
    CivitAI
    Platform Status
    Available
    Created
    3/4/2024
    Updated
    9/28/2025
    Deleted
    -

    Files

    373845_training_data.zip

    Mirrors

    CivitAI (1 mirrors)

    Not-Midjourney-V3-0.3.safetensors

    Mirrors