Gryvain - half Dragons

Originally inspired by Trials in Tainted Space - Gryvains - a Race of voluptuous half Dragons with plenty of curves and stunning colors.

you like Dragons?

you want Women??

plenty of curves???

Scales?????

then this Lora is for you!!!

Unlike the Gryvains from TiTs these ones tend to be more half Dragon ish with only one pair of Wings, elfin Ears and colors for hair, horns, lips and eyes linked to their scales. (saucy bits included)

-- usage --

by default the lora will produce nsfw/nude ish images

for sfw images add your clothing of choice

add background of choice

wings and tail are optional and can go into the negative to remove them

scales can be multicolored when only the scales tag is added

scale colors can blend when 2 tags for scale color are added

the lora loves 3d and realistic tags

-- known issues --

• in rare cases the skin can be off color and blend with the scale color when no skin color like tanned skin is given - this is an issue with the base model falling back to the dragon tag

• the lora is quite hefty and will produce a dominant styling and character design even on lower weights as low as 0.2 but is stable up to a wight of 1.6. (which is bonkers tbh)

• the lora will fight with the base model on default weight if the base model has flat/shaded colors or cartoony style - lower lora weight to 0.5 or less or add style lora on hight weight

-- source images for the dataset --

all source images were generated using perchance furry ai and natural language prompts, then ran through comfyui locally with xavier base model for upscale, fix and 3d + semi-realistic style tags. source images are available with metadata.

 

-- training --

trained on regular illustrious model

epochs: 10

num repeats: 4

batch size: 4

steps: 700

resolution: 1024

bucket: on

schuffle tags: false

keep tokens: 0

clipskip: 2

unet learning rate: 0.0008

text encoder learning rate: 0.00007

shedulder: cosine_with_restarts

shedulder cycles: 3

min snr gamma: 5

network dim: 32

network alpha: 16

noise offset: 0.1

the Lora made consistent images similar to source images from generation 4 onward during training with it being well balanced on generation 10.

72 source images ranging from full body to closeups from diffrent angles and poses withouth word spam preventing useless clogging of lora memory and unwanted concepts. (every tag gets baked into the lora as a trigger word!! - duh)