Character Set from H anime "Record of Lodoss War ロードス島戦記 (OVA) ".
This model offers a mini-sized version of the character's original art style.
You can use the prompts below to create character:
All characters are adults
Deedlit ディードリット
mai_rolw,retro_artstyle,1990s_\(style\),deedlitln,elf,long_hair,blonde_hair,green_eyes,long_pointy_ears,circlet,earrings,shoulder_armor,breastplate,green_skirt,blue_cape,arm_warmers,leg_warmersPirotess ピロテース
mai_rolw,retro_artstyle,1990s_\(style\),pirotessln,mature_female,dark_elf,solo,long_hair,white_hair,pointy_ears,dark-skinned_female,grey_eyes,lips,circlet,white_dress,short_dress,center_opening,gloves,white_thigh_bootsKarla カーラ
mai_rolw,retro_artstyle,1990s_\(style\),karlarl,mature_female,ponytail,black_hair,blue_eyes,purple_lips,circlet,earrings,necklace,hood,black_cloak,high_collar,purple_dressShiris シーリス
mai_rolw,retro_artstyle,1990s_\(style\),shirisrl,mature_female,short_hair,red_hair,purple_eyes,armor,red_dress,fur_trim,short_dress,red_thigh_boots,capeRecommend preferred Strength: 0.7~1.
Using the prompt 'mai_rolw, retro_artstyle, 1990s_\(style\)' can obtain the style of the work.
A nostalgic masterpiece, the pioneer of Elf-themed anime. Since many of the OVA episodes were created by heavyweight animators, the style after learning with AI isn't really all that consistent. Art really can't be copied with just technique :)
Description
FAQ
Comments (11)
Oh, very nice. This takes me back.
\(★ﾟ∀ﾟ☆)/
I love the chara design do you will make Anima Lora ?
Making an Anima version could take about 4-6 hours to train based on the data size, and since the images in this dataset aren't very clear, it's not really suitable for training DiT-type models. So for now, I don't have any plans to make an Anima version model :(
@00x09901 ok thank no prob :)
Hello. What checkpoint or workflow did you use for your sample images? I can't recreate that quality at all. That angry Deedlit in white dress for example, always comes with different color dresses, even when using "white dress" ;-(
@Fluffia_Ai Hello, in the example images, my generation settings are all as shared. I usually generate example images directly using the WebUI. The image you pointed out is quite special; it’s a two-stage img2img generation image, and this type of image will have an img2img tag in the labels. If you want to generate a similar image in one go using txt2img, you can refer to this picture:
@00x09901 Oh! I see. Thank you! :-)
Σ(ﾟДﾟ) Wow!
@00x09901 hehe xD