I don't speak English so I'm translating at ChatGPT. Let me know if the English is weird.This is the SDXL version of <a target="_blank" rel="ugc" href="https://civarchive.com/models/86277?modelVersionId=188171">Manmaru Mix</a>. It is based on <a target="_blank" rel="ugc" href="https://civarchive.com/models/260267?modelVersionId=403131">Animagine XL V3.1</a>. We have conducted additional training using images produced by <a rel="ugc" href="https://civarchive.com/models/86277?modelVersionId=188171">Manmaru Mix</a> SD1.5. Honestly, we need to improve the quality of the training images further, but it seemed like SD3 might be released by the time we got everything perfect, so we are releasing this current version for now.<a rel="ugc" href="https://civarchive.com/models/86277?modelVersionId=188171">Manmaru Mix</a>のSDXL版です。ベースに<a rel="ugc" href="https://civarchive.com/models/260267?modelVersionId=403131">Animagine XL V3.1</a>を利用しています。 SD1.5の<a target="_blank" rel="ugc" href="https://civarchive.com/models/86277?modelVersionId=188171">Manmaru Mix</a>から出力した画像を使って追加学習を行いました。 正直なところもっと教師画像のクオリティを上げなければならないのですが、 のろのろとやってたらSD3がリリースされてしまいそうだったので 現時点でのバージョンをとりあえずリリースします。Training Image Breakdown / 教師画像内訳I am currently experimenting, and it is evident that the balance of image proportions is not ideal. My selection process has also been lenient, and there is a definite need for a quality improvement. I essentially need more time...以下は試行錯誤中で、画像の割合のバランスが悪いというのは現時点でわかっています。 画像選別も甘く、クオリティアップが必要と考えています。とにかく時間がいる・・・Outputs from SD1.5 Manmaru Mix / SD1.5のManmaru Mixで以下画像を出力<ul><li>Upper body images with various hairstyles and hair colors on a white background 背景白の様々な髪形、髪色の上半身画像 x 1500</li><li>Full body images with various poses on a white background 背景白の様々な姿勢の全身画像 x 500</li><li>Images with various poses on various backgrounds 様々な背景の様々な姿勢の画像 x 1000</li></ul>VAE<a rel="ugc" href="https://civarchive.com/models/152040?modelVersionId=264491">XL_VAE_C f1</a> has been merged. / <a rel="ugc" href="https://civarchive.com/models/152040?modelVersionId=264491">XL_VAE_C f1</a>をマージ済みです。License This model uses <a target="_blank" rel="ugc" href="https://civarchive.com/models/260267?modelVersionId=403131">Animagine XL V3.1</a> and therefore inherits the <a target="_blank" rel="ugc" href="https://freedevproject.org/faipl-1.0-sd/">Fair AI Public License 1.0-SD</a>. 本モデルは<a target="_blank" rel="ugc" href="https://civarchive.com/models/260267?modelVersionId=403131">Animagine XL V3.1</a>を利用しているため<a target="_blank" rel="ugc" href="https://freedevproject.org/faipl-1.0-sd/">Fair AI Public License 1.0-SD</a>を継承しています。

Files with SHA256: e686217ec8e8dc0fdf110eb7245b1a6d5516623f780543938ee5f2b85caa0756