This LoRa achieves the best performance by selecting "WAI-illustrious-SDXL v15.0" as the checkpoint.
This LoRa model was specifically trained for the "Taimanin Asagi OVA" series, directed by Teruaki Murakami.
It was trained to generate anime-style images based on director Teruaki Murakami's work.
The background colors are white and red to clearly separate the generated results.
・White background evokes image of dignified Taimanin.
・Red background evokes image of ❤hardcore play❤.
This is a trial training model created using 700 materials, Network Dim 64, and 2000 steps.
This training model used "WAI-illustrious-SDXL v15.0".
Portait 832x1216 ............ ❤️Highly recommended❤️
Square 1024x1024 ......... ❤️Highly recommended❤️
Landscape 1216x832 .... ❤️Highly recommended❤️
The learning materials are available here. Please feel free to use them.
https://mega.nz/folder/5shlnJKa#PoR_T4A2HUN0JKpYt75eQQ
--SAMPLE PROMPT ------------------------------------------------------------------------------------
PROMPT:White BackGround (Taimanin image)
(Simple background), ((white background)),
<lora:MurakamiTeruaki_Taimanin_Anime_IL_V3:0.9>murakamiteruaki_taimanin, anime screencap, cel animation, slight film grain, retro artstyle, sharp focus,
with detailed ringed pupils , long eyelashes, striking angular eyes, glossy highlights on skin and hair, Intense and oppressive atmosphere, dramatic hard-edged shadows,
taimanin, shiny body, skin tight, taimanin costume, colorful costume, frills, shiny clothes, armor, lothing cutout, arm guards, fingerless gloves, elbow gloves,
covered nipples,
high leg cut, cameltoe,
fighting stance, spread legs, legs apart,
<***Character Prompt***>.
PROMPT:Red BackGround (❤Hardcore Play❤)
(Simple background), ((RED background)),
<lora:MurakamiTeruaki_Taimanin_Anime_IL_V3:0.9>murakamiteruaki_taimanin, anime screencap, cel animation, slight film grain, retro artstyle, sharp focus,
with detailed ringed pupils , long eyelashes, striking angular eyes, glossy highlights on skin and hair, Intense and oppressive atmosphere, dramatic hard-edged shadows,
taimanin, shiny body, skin tight, taimanin costume, colorful costume, frills, shiny clothes, armor, lothing cutout, arm guards, fingerless gloves, elbow gloves,
Captive, prisoner,
Torn clothes, meaningless clothes,open breasts, open pussy,
ringed eyes, wide-eyed, teeth, crazy eyes, empty eyes,@_@,
<***Character Prompt***>.
PROMPT:Red BackGround(❤Full SET Dynamic Prompts ❤)
(Simple background), ((RED background)),
<lora:MurakamiTeruaki_Taimanin_Anime_IL_V3:0.9>murakamiteruaki_taimanin, anime screencap, cel animation, slight film grain, retro artstyle, sharp focus,
with detailed ringed pupils , long eyelashes, striking angular eyes, glossy highlights on skin and hair, Intense and oppressive atmosphere, dramatic hard-edged shadows,
taimanin, shiny body, skin tight, taimanin costume, colorful costume, frills, shiny clothes, armor, lothing cutout, arm guards, fingerless gloves, elbow gloves,
Captive, prisoner,
Torn clothes, meaningless clothes,open breasts, open pussy,
ringed eyes, wide-eyed, teeth, crazy eyes, empty eyes,@_@,
{close-up face | cowboy shot | upper body},
((from { front | back | side} sex)), rough sex, epic sex,
3+boys, 3+penis, Multiple penises, dark-skinned male, interracial, Ugly male, fat male, naked male, faceless male, frame out male head, hetero,
excessive bukkake, face close-up, bukkake, facial, excessive cum, cum on hair,
{(Hyottoko blowjob), ( :>=), irrumatio, deepthroat, oral, | Tasting semen on tongue, semen pool in mouth), open mouth, }
gangbang, group sex, handjob, oral, fellatio, ((triple penetration)), ((double penetration)),
large insertion, large penis, thick penis, Grab body, grab legs, grab waist, grab head,
<***Character Prompt***>--Sorry Japanese Text ----------------------------------------------------------------------------------
ご好評をいただいている、むらかみてるあき監督作品の、対魔忍アサギOVAのエロアニメジェネレータの、大幅改善…というか、完全別物の作り直しの試作モデルです。
.
背景色の指定により、明確に生成イメージを分離しました。
・白背景は凛々しい対魔忍イメージ生成
・赤背景はスーパーハードコアプレイ生成
.
パラメータ調整の試行錯誤の結果、モデルサイズは445MBと肥大化しました。
既存モデルと内部構成が完全に別物なので、先日投稿した姉モデルとは別モデル扱いとして投稿しました。
本モデルは、よりOVAアニメを意識した生成を行います。
楽しんでいただければ幸いです。
.
学習素材はこちらです。ご自由にお使いください。
https://mega.nz/folder/5shlnJKa#PoR_T4A2HUN0JKpYt75eQQ
サンプルスクリプトの <***Character Prompt***> 部分にお好きなキャラクターを指定してご使用ください。
ダイナミックプロンプトがご利用になれる場合は、「PROMPT:Red BackGround(❤Full SET Dynamic Prompts ❤)」からダイナミックなパターンを生成できます。
このモデルは、今までの生成方法とは異なるパラメータで作成されたモデルです。以前の公開版との比較等、忌憚のない意見を聞かせてもらえると改善が捗ります。
私は本LORAの利用に、制限・権利・責任を一切主張しません。
Description
FAQ
Comments (13)
This model intentionally generates thicker lines with different parameter values than previous LoRa models.
Please adjust the LoRa strength to find your preferred value.
We look forward to your feedback.
LoRAの味付け的には、かなり濃厚な味付けのLoRaになりました。久留米ご当地の激臭超濃厚とんこつラーメンくらいの好きと嫌いが分かれる、人を選ぶ味付けだと思います
まぁ、皆さん意外かと思うでしょうが…私はね、このクッソどぎっついド変態臭の濃厚な味がね、大好きなんですよ。
画風LoRAは刺さるかどうか(Yes/No)なので、クセが強い方がいいと思います!
変わったかどうかわからない程度じゃ、意味が無いですし。
Another masterpiece of deviance!
💕👍🏻
hi link is dead : https://mega.nz/folder/5shlnJKa#PoR_T4A2HUN0JKpYt75eQQ
@XT_404
Thank you for the report!
I had access. I was able to access it as a guest with other accounts as well.
If anyone else cannot access it, please let us know.
I will talk to MEGA.
@suteakasuteakasuteka434 : thx for feedback :)
So then... Are yous guys doing these animations all in Automatic1111 img2vid? Or do you generate the image there and move to something like ComfyUI?
I just recently got a handle on doing animations and wanted to try this but am a bit confused.
This model generates images using Illustrious checkpoints.
The generated images are specified as the START-END of ComfyUI's WAN2.2 I2V workflow to generate a video.
@suteakasuteakasuteka434 Ah okay. Cool. Thanks for the tip!
Edit: ̶W̶a̶i̶t̶ ̶a̶c̶t̶u̶a̶l̶l̶y̶ ̶I̶ ̶d̶o̶ ̶h̶a̶v̶e̶ ̶a̶ ̶q̶u̶e̶s̶t̶i̶o̶n̶ ̶t̶h̶e̶n̶.̶ ̶W̶h̶e̶n̶ ̶y̶o̶u̶ ̶s̶a̶y̶ ̶t̶h̶a̶t̶ ̶y̶o̶u̶ ̶a̶r̶e̶ ̶u̶s̶i̶n̶g̶ ̶S̶t̶a̶r̶t̶-̶E̶n̶d̶,̶ ̶d̶o̶ ̶y̶o̶u̶ ̶m̶e̶a̶n̶ ̶l̶i̶k̶e̶ ̶t̶h̶e̶y̶ ̶a̶r̶e̶ ̶t̶h̶e̶ ̶s̶a̶m̶e̶ ̶t̶w̶o̶ ̶i̶m̶a̶g̶e̶s̶ ̶b̶u̶t̶ ̶o̶n̶e̶ ̶i̶s̶ ̶z̶o̶o̶m̶e̶d̶ ̶i̶n̶?̶ ̶B̶e̶c̶a̶u̶s̶e̶ ̶f̶o̶r̶ ̶m̶o̶s̶t̶ ̶o̶f̶ ̶t̶h̶e̶m̶ ̶i̶t̶ ̶d̶o̶e̶s̶n̶'̶t̶ ̶s̶e̶e̶m̶ ̶l̶i̶k̶e̶ ̶a̶ ̶r̶a̶d̶i̶c̶a̶l̶ ̶c̶h̶a̶n̶g̶e̶ ̶i̶n̶ ̶p̶o̶s̶e̶.̶
Nevermind I figured it out lol. Your guides are really helpful, thanks again!
@calphamark1181
I'm so glad that this problem seems to have been solved.
By the way, I failed miserably at creating the WAN2.2 I2V animation model for Taimanin Asagi, directed by Teruaki Murakami. 🤪
I plan to try again.
@suteakasuteakasuteka434 Ah! Cool man, I wish you luck then! I really like the animation direction. It would be cool if it could be easily applied to other art styles. c:
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.















