This is another version of my latex diffusion loras with a reworked dataset and tagging.
The currently available versions are only trained on images with latex hoods (tagged as latex masks).
The reason for this is to separate my dataset into two parts: with hoods and without hoods
This should make each lora better for it's task instead of creating inaccurate masks or plastic looking faces.

The hood versions have two main triggers:
'[color] latex mask' - to trigger the correct hood (example: 'black latex mask')

'hairless / [color] [braided/] [pigtails/ponytail] - to trigger the hairstyle or none (examples: 'hairless', 'blonde braided pigtails', 'red ponytail')

Other tags:

'[color] latex catsuit' - catsuits of various colors (black, red, blue, green, pink, purple, white, bordeau)

'transparent latex' - some thin suits tagged with those (not colorless)

'[color] latex corset - various corsets

'laced catsuits' - tagged for back views

'[color] harness, o-rings'

'o-ring collar, spiked collar, [color] collar' - different collars

'chain leash'

'[color] platform high heels'

'[color] ball gag' - only red and black in the dataset

'[color] lipstick' - a few lip colors (pink, red, dark pink, purple, black)

'full lips, glossy lips, parted lips' - a few more lip tags
'latex hood' - only mentioned once but refers to a literal hood like from a cape/hoody/etc.
'latex nun, nun habit' - latex fetish nun outfits

'latex dress, tube dress, minidress' - various dresses

'cleavage, plunging neckline, unzipped catsuit, exposed breasts, boob window, laced cleavage' - types of cleavage

'bare hands, bare feet' - tagged for exposed hands and feet

'latex feet' - tagged for feet focused images

'black latex penis' - a few images of sucking a black rubber dildo or suit sheath