Trained on PonyV6 with 20 images in LoKR.<hr />The base trigger word is <code>DoraemonStare, doraemon, alternate costume,</code>Recommended strength: <code>0.9 ~ 1.3</code>Recommended size: <code>1152 x 1152</code> (1:1)Please refer to the example images for the working prompt.It's a little difficult to use with other character LoRA. Check this post if you are willing to try it: <a target="_blank" rel="ugc" href="https://civarchive.com/posts/4940653">https://civarchive.com/posts/4940653</a><hr />This is a meme called 「哆啦A夢的凝視」 that was popular in the Chinese community in 2022 and has recently been brought up again.Here is the explanation of this meme (in Chinese.): <a target="_blank" rel="ugc" href="https://chinesedora.com/news/40372.htm">https://chinesedora.com/news/40372.htm</a> <a target="_blank" rel="ugc" href="https://www.thepaper.cn/newsDetail_forward_19261803">https://www.thepaper.cn/newsDetail_forward_19261803</a>↓This is the training dataset↓<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/28e13aca-9b8a-4206-8ae1-d0d2a7ea24f3/width=525/28e13aca-9b8a-4206-8ae1-d0d2a7ea24f3.jpeg" />↑This is the training dataset, not the AI-generated images.↑

DoraemonStare XL | meme

This LoRA is an attempt to train on Civitai on-site trainer. It seems somewhat flatter.I can't say it's better than vPony.1, but it's not bad either.

Files with SHA256: d5fd82a849470e1b2f8603771ae0db371bfc32825d2a11e8f0828caa23e552e1