アニメLORAですが、アニメモデルとの相性は良くありません、"kantanmix"以外のモデルで出力する場合は2.5Dモデルで試してみて下さい。アニメモデルの場合は、重量を下げ、プロンプトで補正したりする事が必要です。
Although it is an anime LORA, it is not compatible with anime models. If you want to output with a model other than "kantanmix", please try using a 2.5D model. For animated models, it may be necessary to lower the weight and make corrections with prompts.
*********************************************
好きな衣装や、背景、ポーズを組み合わせて下さい。
Please combine your favorite costumes, backgrounds, and poses.
*********************************************
weight 0.7、ADetailerを使用しface_yolov8nで顔修正を行っています。
高解像度補助:R-ESRGAN 4x+ Anime6B,倍率2倍
Weight 0.7 ADetailer is used and face_yolov8n is used for face correction.
Hires.fix:R-ESRGAN 4x+ Anime6B,2x magnification
*********************************************
"Detail Slider - LoRA"
https://civarchive.com/models/153562?modelVersionId=171989
"Detail Tweaker LoRA (细节调整LoRA)"
https://civarchive.com/models/58390?modelVersionId=62833
"flatBG"
https://civarchive.com/models/81298/flatbg
"noline"
https://civarchive.com/models/81309/noline
"boldline"
https://civarchive.com/models/81313/boldline
"Zoom Slider - LoRA"
https://civarchive.com/models/114460?modelVersionId=123732
"ADetailer"
(インストールは自己責任でお願いします)(Please install at your own risk.)
顔の詳細を修正する追加機能
Additional features to correct facial details
https://github.com/Bing-su/adetailer
Embeddings
"Fast Negative Embedding (+ FastNegativeV2)"
https://civarchive.com/models/71961?modelVersionId=94057
"EasyNegative"
https://civarchive.com/models/7808/easynegative
"negative_hand Negative Embedding"
https://civarchive.com/models/56519/negativehand-negative-embedding
Description
嫌顔の目が小さくならない
FAQ
Comments (7)
はやい
大好きな作品なので、かなり頑張って作りましたｗ
良ければ、LORAのテストや、投稿、評価をして頂けると嬉しいです。
Sadly, this one works well only with your kantanmix... if I use any other model, like eg. AnyLora or Anything v5, it's going to produce very "childish artstyle" results
You should train it on NAI (anime-full-pruned) model rather than your own.
It will give much better results and be more flexbile.
@LittleJelly
Thank you for your advice.
Where can I get the "NAI (anime-full-pruned) model"? , I've never tried it, so I'd like to try it.
The LORA I am creating is an anime-based LORA, but it is not suitable for outputting as an anime-based model. "kantanmix" is the best, but if you want to output as another model, use a 2.5D model. I think you need to output it.
I tried training with various models, but when I used strong learning to output like the official animation, the background of the anime model was lost and the output could not be done well, so I decided to use the current method. I'm taking it.
It is a model between an anime model and a 2.5D model, and it feels like it protects the background data while powerfully learning the character.
ご指摘ありがとうございます。
"NAI (anime-full-pruned) model"とは、どこで手に入る物でしょうか？、試した事が無いので、試してみたいです。
私が作成しているLORAは、アニメ系LORAですが、アニメ系モデルでの出力には向いていません、"kantanmix"が最適ですが、他のモデルで出力する場合は2.5D系のモデルで出力する必要が有ると思います。
色々なモデルで学習させてみましたが、公式アニメの様な出力をさせる為に、強力な学習をさせると、アニメ系モデルでは、背景が負けてしまい上手く出力が出来ず、今の手法を取っています。
アニメモデルと2.5Dモデルの中間のモデルで、キャラクターを強力に学習させつつ、背景データを保護している感じです。
@jibunsagasinotabi
1. download link -> https://huggingface.co/hollowstrawberry/stable-diffusion-guide/tree/main/models
2. the background of the anime:
a) i recommend separate LoRA for that, focused specifically on backgrounds
b) for image generation combine both, character & background lora (i've done that with my Yofukashi no Uta)
3. がんばって！新しいLORAを待っている
@LittleJelly
Thank you, I have downloaded it and am trying it out.
It was very helpful.
ありがとうございます、ダウンロードして色々試しています。
とても参考になりました。
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.