![[IL] Sheila the Thief (Dungeons and Dragons cartoon) / シイラ – 盗人 (ダンジョンズ&ドラゴンズ アニメ)](https://comfypilot.com/api/v1/img?u=https%3A%2F%2Fimage.civitai.com%2FxG1nkqKTMzGDvpLrqFT7WA%2F9471d570-73f3-4efd-a650-86d663be86e6%2Foriginal%3Dtrue%2F149449600.jpeg%3Fwidth%3D450)
触发词
文件清单 线路:镜像(hf-mirror.com) · 切换
下载后放进对应的 ComfyUI 目录。models/ 指 ComfyUI 的模型根目录。
说明
[IL] Sheila the Thief (Dungeons and Dragons cartoon) / シイラ – 盗人 (ダンジョンズ&ドラゴンズ アニメ)
作者:NanashiAnon
The party’s nominal thief. I say nominal because as far as I can tell she finds a trap once (at least it's a magic trap?), successfully pickpockets once, picks zero locks (fewer than the cavalier!), and can’t sneak anywhere without a Cloak of Invisibility. Tags: D_D_S_T, orange hair, medium hair, freckles, pink lips Outfit: hood down, hooded cloak, short dress, sash, fold-over boots, thigh boots ("hood up" is also trained, as is her without cloak) Style: Trained with "official art" (for production art) and "1980s (style)" (for screencaps), plus three "figure (medium)" pictures. Made with 12 production art, 35 screencaps, and 4 pics of toys (this last one would prove of mistake: see below). Saw this was on the official D&D channel in decent quality (in addition to a bunch of horrible Critical Role wannabes that now dominates the channel.), and I managed to find some concept art in good quality so I decided to give it a go and see if I could improve on the existing models. Threw in some pics of toys as an experiment to see if it would give her features a bit more “depth”: While the LoRA works, I wouldn't recommend doing that as it seems to add artifacts to her boots and give her skin a gloss when used without the trained style tags. I think I'll be back for a V2 that excludes those relatively soon.