CivArchive
    MetersXD Style by LoRA Factory - SwinV2 WD14 Tags
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined
    Preview undefined

    MetersXD Style

    This style LoRA is trained to generate furry characters with bold shapes and soft lighting with a melancholic tone.

    Use danbooru tags and e621 tags in tandem with English caption for best result.

    Recommend Strength: 0.6 - 1.0

    Base model: ACIDFUR (no VAE) VAE: OrangeMix VAE

    Best used with: badhandv4, adetailer CLIP Skip: 2

    Hi-Res Fix Steps: 7 - 14

    Latent (nearest exact) or R-ESRGAN 4x+ Anime6B

    Denoising Strength: 0.3 - 0.4+

    Upscale: 1.25- 1.5+

    adetailer model: face_yolo8n.pt (put in "stable-difussion-webui/models/adetailer")

    adetailer prompt: eyes

    LoRA Factory 2023

    Description

    Training data auto-captioned using SmilingWolf/wd-v1-4-swinv2-tagger-v2

    FAQ

    LORA
    SD 1.5

    Details

    Downloads
    136
    Platform
    CivitAI
    Platform Status
    Deleted
    Created
    6/10/2023
    Updated
    7/19/2026
    Deleted
    8/7/2025

    Files

    LRFactory-MetersXD-swinv2.safetensors