trigger words:<pre><code>@swd3e2style</code></pre>period:<pre><code>newest</code></pre><pre><code>recent</code></pre><pre><code>mid</code></pre>style:<pre><code>sketch</code></pre><hr />The training set has been cropped to remove the signature watermark.<hr />Recommended settings: LoRA Weight ≈ 0.8, Resolution ≈ 1024×1600, 20–30 Steps, CFG 4.5. For enhanced details, a 1.2× Hires/Detailer redraw can be applied after generation.<hr />Comparison of style outputs using different trigger combinations, including Newest, Recent, Mid, and Sketch variants. <img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/1ad2c17f-a33c-46ff-b75e-948712a60f16/original=true/1ad2c17f-a33c-46ff-b75e-948712a60f16.jpeg" /><hr />The training set contains a portion of illustrations with prominent backgrounds and smaller character presence. Therefore, this LoRA can be used not only for character artwork, but also for generating stylized backgrounds and environmental scenes.

Files with SHA256: 6bafeb406486f8a0c5ee6b846689dfb60a296513ad8a9de3d9f38ae480d9edcf