I noticed that LoRAs made for Sui ishida (TokyoGhoul author and artist) is either made for SD1.5 or just in color; colored ones capture his water-based style of drawing, not the ink-based style of his manga works.So; I gathered and edited 235 full stills of both TokyoGhoul mangas (without speech bubbles or manga sites logos) and made this LoRA against base SDXL 1.0. Then chose a checkpoint in which the general hatching and tones style is dominant but the face structure is unique each time; this was the one for me.Extra loras to be used with this (to achieve the showcase level of quality):1- <a target="_blank" rel="ugc" href="https://civarchive.com/models/122359/detail-tweaker-xl">https://civarchive.com/models/122359/detail-tweaker-xl</a>2- <a target="_blank" rel="ugc" href="https://civarchive.com/models/124347?modelVersionId=152309">https://civarchive.com/models/124347?modelVersionId=152309</a>Trigger word: sui ishida styleNote1: This LoRA works very well with base SDXL and not so much with other models; as they tend to product anime-style pictures. Sadly, base SDXL is the only one that gives accurate results so far in my testing.Note2: For better results, use it with 1.0 weight, 7.0 CFG scale. This LoRA works well with detail loras in general (more of that in the test pictures below - all made by civit's generator).

Files with SHA256: f77a1b803fa5c4e62936eab657302b878736d3df0bd9101f42480ac86b2fa3c4