Honestly surprised with how this LoRA came out given the low quality of the dataset I had to work with.
-------------------------------------------------------------------------------------------------------------

Syouta Kaneda (金田正太) from the series Himitsu Gattai Dailaon/秘密合体ダイラオン

Trigger words: syouta kaneda, brown hair, spiked hair, brown eyes
Syouta comes with two outfits:

• Main Outfit tags: HGD_syouta_uniform, cropped jacket, two-tone jacket, red jacket, white jacket, turtleneck, two-tone shorts, white shorts, red shorts, blue socks, kneehighs, boots

• Alt Summer Outfit tags: HGD_syouta_summer_outfit, t-shirt, red shirt, raglan sleeves, blue sleeves, number print, white shorts, black belt, sandals

-------------------------------------------------------------------------------------------------------------

• Recommended parameters: 20-25 steps, Euler a/Automatic (or DPM++ 2M/Karras)

• Upscaled with 4x-Ultrasharp; Adetailer also recommended!

• The dataset I used to train this LoRA already has a 2000s anime aesthetic baked in, but you can still use tags like "anime screenshot" or "anime coloring" to reduce that sheen effect in some generations

📢🚨 THIS MODEL IS NOT FOR COMMERCIAL PURPOSES ❗🙅‍♂️