Hello to my 39 followers, we officially hit 1.2k downloads and 300+ likes on all my loras, which i thank everyone for :D. Now for the sad news, I have to reinstall both my debian install and my windows partition as well since I cant boot into my windows install and my debian one is buggy as hell basically not letting me install anything properly. Anyways this just meant I have to cut my testing short on this lora so I can reinstall everything and get back into the rhythm. I tried alot less tagging with this lora since for some reason not tagging the eyes in my previous lora made the eyes look much better, so I deducted that I can just do that with every other feature as well. I guess you, my testers will report back on if it works or not. Dont be afraid to tell me if its absolutely dog since I tried something new and I can just re-do it later if its really bad.For my suggested prompts (there isnt alot for her): vi0let, large breasts, nipples, robes/blue robes, fox tail. With her I could do a back view as well since she actually had good images, but for some reason her fox tail kinda doesnt appear all the time, so make sure to prompt it if you miss it. Trained on <a target="_blank" rel="ugc" href="https://civarchive.com/api/download/models/100675">MeinaHentai</a>, as base model. In conclusion: I will be uploading maybe Wednesday if I can solve all the issues, but dont hold your breath if I dont. Yeah do post your generated images cause I couldnt do much testing and I want to see if my new tagging method actually works, and if you are new please do drop a follow idk maybe ill do a 100 follower special. Thank you :)

Violet - Project QT

Less tagging = less flexibility i assume

Files with SHA256: 07080509de69c7be33b2de9b335918d4ca0b487198ae473be8696174fe3e419e