CivArchive
    Wizard - v1.0
    NSFW
    Preview 1822328
    Preview 1822248
    Preview 1822251
    Preview 1822127
    Preview 1826046
    Preview 1822128
    Preview 1822124
    Preview 1822130
    Preview 1822297
    Preview 1822246
    Preview 1822249
    Preview 1822243
    Preview 1822125
    Preview 1822129
    Preview 1822244
    Preview 1822242
    Preview 1822250
    Preview 1822247
    Preview 1822362
    Preview 1822363

    Please put an objective assessment of my work , it will motivate me to make new models.
    add a ❤️ to receive future updates.

    Wizard:

    Size:
    square - 1024x1024
    rectangle - 1024x1365

    Full Negative prompt (recommend):
    layman work, worst quality, ugly, (deformed|distorted|disfigured:1.21), poorly drawn, bad anatomy, wrong anatomy, mutation, mutated, (mutated hands AND fingers:1.21), bad hands, bad fingers,loss of a limb, extra limb, missing limb, floating limbs, amputation, Yaeba



    This model has been trained on 1000 of the best images taken on the LehinaModel
    Additionally, several mergers with the base model + 4Guofeng4XL_v10RealBeta + dreamshaperXL10_alpha2Xl10 were performed, and the best option was selected.

    Use and share. Enjoy creating art for yourself and your loved ones.
    Best regards, Esher.

    Description

    FAQ

    Comments (6)

    Agent_SmthAug 1, 2023
    CivitAI

    what is this? a finetune? a mix?
    no infos

    PolygonAug 1, 2023
    CivitAI

    A lot of the faces, and particularly the hair, look a little too blurry. Is that a feature of the model, or the prompt?

    Are you using the refiner (it helps with that)?
    Also, what did you merge to get to this?

    Its because the current trained xl models dont have an appopriate trained refiner checkpoint. I have tested all variations XL with/without refiner, SD 1.5 as refiner. XL base, XL refiner with 1.5 checkpoint for upscale. My summary: Best results XL 0.9 and 0.9 refiner and a high Res step works best for my skullart. XL base without refiner or 1.5 model as refiner is definitely behind the quality of a good trained "stand alone" 1.5 model.

    Daedalus_7Aug 1, 2023

    @sikasolutionsworldwide709 It's because you are using Euler a. It behaves differently than in SD1.5, and in SDXL it creates blurry images. Try DPM++ 2SA Karras at 70 steps and a CFG of 7+.

    Also, try to use High Rez fix to being the images to 1280x1280 (for example) and use Adetailer to improve the face. I'll post an example later :)

    PolygonAug 1, 2023

    @Daedalus_7  I have used refiner on non base models with decent results.

    Daedalus_7Aug 1, 2023

    @Polygon If you use the refiner on an image that has been through the High Res fix it gives the image less details, though :(

    Checkpoint
    SDXL 1.0

    Details

    Downloads
    1,624
    Platform
    CivitAI
    Platform Status
    Available
    Created
    8/1/2023
    Updated
    5/11/2026
    Deleted
    -

    Files

    wizard_v10.safetensors

    Mirrors

    HuggingFace (1 mirrors)
    CivitAI (1 mirrors)
    TensorArt (1 mirrors)