<h2 id="heading-448">Trigger words: onbashira</h2><h3 id="heading-449">optional words: shimenawa, shide</h3><p>Scenery words: scenery, landscape mountain, nature, wideshot</p><h3 id="heading-450">*Note: Due to a bit of concept bleeding from Kanako's Shimenawa from Touhou (she's in ~40% of the dataset), you randomly get a round floating shimenawa (~10% chance) if you prompt "Shimenawa." I will try to fix this in a newer version.</h3><p>Trained on 108 images.<br /><br />Version 1: Similar to V2, but less concept bleeding from Kanako, but lower success rate with shimenawa (round rope around the pillar) and shide (connected square papers on shimenawa).<br /></p><p>I'll appreciate it if you support me on <a target="_blank" rel="ugc" href="https://www.patreon.com/user?u=11876296">Patreon</a>. I plan on sharing early access to my models and releasing more based on my milestones on Patreon.</p><p>I have a discord server (<a target="_blank" rel="ugc" href="https://discord.gg/AmdETXDfHs">https://discord.gg/AmdETXDfHs</a>) where you can have AI and off-topic discussions, ask for help, and having fun (gaming, playing with AI, etc).</p><p>Best place to DM me is via Twitter: <a target="_blank" rel="ugc" href="http://twitter.com/CatNatsuhi928">twitter.com/CatNatsuhi928</a></p>

onbashira (Japanese ceremonial pillars)

<p>More images compared to version 2, but more Kanako's Shimenwa concept bleeding than V1.  Will be fixed in V3.</p>

Files with SHA256: 8096954bf6dfeff99f09e2df5d333da33e35f14f96ce81342e36ef7f37d5d678