You can support me at https://ko-fi.com/elderagent
This is just a test as my first lora, on lower strengths than 1 it struggles with the eye/glasses aspect, don't expect much from it since it has only uses 13 images for training, in general it works consistently
Since the show style has flat colors it helped to get decent reference images, from the low amount of training images it will most likely not deviate much from the design but it might give repetitive results. Add 'short' or 'chibi' to your prompt to get a more show-accurate result.
Prompt
lisaloud, 1girl, solo, green sweater, pink pants, brown footwear, glasses, standing in a bedroom, looking at viewer, brown hair, shortSince I'm out of buzz I don't think I'll update another version anytime soon, feel free to improve upon it.
Description
You can support me at https://ko-fi.com/elderagent
This is just a test as my first lora, on lower strengths than 1 it struggles with the eye/glasses aspect, don't expect much from it since it has only uses 13 images for training, in general it works consistently.
FAQ
Comments (6)
They're Wonderful
Whenever you have the time would love an illustrious version, your lora's are amazing fam keep up the amazing work
I was about to but I lost the data set
@elderagent Ah damn all good brodie
hey, what setting you used in Lora_Trainer_XL colab ?
Oof I don't even remember
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.


