Character model based on 38 images of Leander from Azur Lane, mostly fan art.I had been using <a target="_blank" rel="ugc" href="https://civarchive.com/models/312359/leander-azur-laen-xl?modelVersionId=350545">mara_123's Leander model</a> for the longest time (and a big thank you for that one!), but there were quite a number of issues with outfit/style/pose bleed, so eventually I decided to train my own. I've been using this model for a while, and figured that her launch day (or 'unberth' day as I like to call it 😄) was as good a time as any to post it. Happy 93rd, Leander o/<h3 id="usage-7u5y6w3jg">Usage</h3>BasicWorks well anywhere between a strength of <code>0.7</code> and <code>1.0</code>. Base trigger is<ul><li>Pony: <code>leander</code></li><li>NoobAI: <code>leandernai</code> Note: Tag was changed to avoid colliding with the built-in knowledge NAI has of Leander, but the LoRA does also work reasonably okay with <code>leander</code>. See XYZ comparison image for reference.</li></ul>Higher strengths may lead to outfit bleed (but shouldn't do if you specify clothing different to the official outfit), so you may wish to drop the strength to <code>0.8</code>.OutfitsOutfit components are tagged separately so you can choose what to include.<ul><li>Base skin: <code>white jacket, white shirt, red skirt, high-waist skirt, white gloves, elbow gloves, fingerless gloves, black thigh highs, garter straps</code></li></ul>Being that Leander's hairband is so integral to her identity, I decided not to tag it separately, but you could try adding <code>hairband</code> to the negative if you want to exclude it.There is technically a tag for the <code>anchor</code> fasteners on her thigh highs, but it is not at all consistent.<h3 id="extra-ystm12y89">Extra</h3>In the interest of openness and community-helping, I have uploaded the dataset along with the model, this includes: images, captions, and the training json (for kohya_ss).

Leander (Azur Lane) [Character] (Pony / NoobAI)

<ul><li>trained on 38 images</li></ul>

Files with SHA256: e5a590dfbc295013dc0ded249dbe39384f49776c58eb118d95e86ea74d32e15e