Yorha commander from the 2023 anime series. Made on request but I'll share it here too in case anyone else needs it.There are two versions, pick whichever you want according to the following - 1.0-ALT is more flexible but possible less slightly less accurate (not that the anime style is especially detailed or anything). 1.0 is better at producing the default outfit but a bit more stubborn in displaying other types of clothing.How to use Use the "yorcom" tag and the 1girl, that's usually enough for most models. Default outfit tags: white coat, white gloves, white thigh boots, ascot, see-through cleavage Optional tags: hair ornament, finely detailed grey eyes (or whatever color you prefer)Fun fact: Using just "yorcom", and negatives "greyscale", "monochrome" on NAI produces screencaps from a futuristic YorHA high school anime with serious yuri undertones for some reason. Recommended strength is anywhere between 0.6 to 1.00 depending on prompt, model and tags.Notes Since the original outfit is all she wears in the anime (so far) it can sometimes bleed through and affect other outfits influencing their style. This is mainly the semi-transparent area in her default outfit since it didn't have a good tag to target it with. If this happens just modify the LoRA or outfit weights. Personally I don't bother since I feel the commander is tight laced enough to totally only wear YorHA logo clothing off duty as well. The logo thing is tagged as "emblem", it's possible to apply sometimes make it appear on it's own, see tattoo image sample.Misc 768 resolution trained on NAI without noise offset option for best LoRA compatibility. Dataset based on 17 screencaps from the TV-release first six episodes + two custom image from previous generation output. Tested on NAI + AOM2.Don't forget to review and upload your own images, it's always fun to see what you're creating! Like my LoRAs and want more? Please consider a small tip towards my electricity bill and further training &gt;.&lt;

YoRHa Commander

Added two genned images from previous version plus one new screencap. Retagged and retrain using up-to-date settings. This version might be a bit more flexible than the previous one.

yorcom-v2.safetensors

v1.0-ALT

v1.0

Yorha commander from the 2023 anime series. Made on request but I'll share it here too in case anyone else needs it.There are two versions, pick whichever you want according to the following - 1.0-ALT is more flexible but possible less slightly less accurate (not that the anime style is especially detailed or anything). 1.0 is better at producing the default outfit but a bit more stubborn in displaying other types of clothing.How to use Use the "yorcom" tag and the 1girl, that's usually enough for most models. Default outfit tags: white coat, white gloves, white thigh boots, ascot, see-through cleavage Optional tags: hair ornament, finely detailed grey eyes (or whatever color you prefer)Fun fact: Using just "yorcom", and negatives "greyscale", "monochrome" on NAI produces screencaps from a futuristic YorHA high school anime with serious yuri undertones for some reason. Recommended strength is anywhere between 0.6 to 1.00 depending on prompt, model and tags.Notes Since the original outfit is all she wears in the anime (so far) it can sometimes bleed through and affect other outfits influencing their style. This is mainly the semi-transparent area in her default outfit since it didn't have a good tag to target it with. If this happens just modify the LoRA or outfit weights. Personally I don't bother since I feel the commander is tight laced enough to totally only wear YorHA logo clothing off duty as well. The logo thing is tagged as "emblem", it's possible to apply sometimes make it appear on it's own, see tattoo image sample.Misc 768 resolution trained on NAI without noise offset option for best LoRA compatibility. Dataset based on 17 screencaps from the TV-release first six episodes + two custom image from previous generation output. Tested on NAI + AOM2.Don't forget to review and upload your own images, it's always fun to see what you're creating! Like my LoRAs and want more? Please consider a small tip towards my electricity bill and further training &gt;.&lt;

Files with SHA256: 6872d4c6d241f6e5f23c95faf0415509ac45f737b6e1758aab3708da84e3baa6