This is very basic lora, infact this is my first one. Its only been trained on 10 images, but it came out pretty well, it maybe somewhat rigid , but you can use (Heavy Prompts) to make is more flexible.You can play with the weights , but i prefer you to use between 0.6 to 0.8, 0.7 worked best for me as you can see in the above samples.<h3 id="heading-905">Trigger words</h3>mariamizuseexp, reddish brown hair, blue eyes, twin tails, bangs, side bangs, hair clip, hair ornamentUse (((flat coloring))), (((anime style))), ((flat colors)) in negative prompt for better results or else the result have more anime type look with no volumeI didnt know she had two types of hairs and this is her disguise before training this lora, so you can generate her multi colored hair in this, ill try to make another one for it.

Files with SHA256: 5afddd6843d4dabd94d4bb1f7aea31e56aba6f503d1ccb9b94e017c936bfc1f4