dranzerstar/SD-textual-inversion-embeddings-repo

SD-textual-inversion-embeddings/Lora repo

Lora Networks

Still Exploring on this training process

prompt: masterpiece, bestquality, clear details,1girl, cowboyshot, simple_background

with respective LoRa net [image] [image] [image] [image]

Lora characters and outfits

using char- and outfit- togeather

masterpiece, bestquality, clear details,1girl, reverseoutfit (pasties) (maebari) highheels \<lora:outfit-reverseoutfit:1\>, (fullbody), lookingat_viewer, floor , \<lora:char-seia:0.9\>,

[image]

Textual inversion embeddings

stable diffusion emeddings of characters and outfits

Check each image's PNG info inthe preview folder for excat gen params

Sample of shinymas/character embeddings

generated with the same prompt with interchanging character phrase (char-X)

prompt: masterpiece, best_quality, clear details, char-kogane ,shirt,1girl,upper body

Negative prompt: fakeanimalears, bad_prompt:0.8, (Cropped head), (Extra hands), (extra legs), (cropped), (missing legs), (duplicate), (morbid), cropped, (error), (bad anatomy), text, jpeg artifacts, (ugly), (morbid), (blurry), (low quality), (long leg), (poorly drawn), (bad proportions),

[image]

Sample of char-toru wearing various outfit embeddings

generated with the same prompt with interchanging outfit phrase (char-X)

prompt: masterpiece, best_quality, clear details, illustration of char-toru standing wearing outfit-null:1.1, ((full body)) , (smile),(solo), boots, floor,

Negative prompt: fakeanimalears, bad_prompt:0.8, (Cropped head), (Extra hands), (extra legs), (cropped), (missing legs), (duplicate), (morbid), cropped, (error), (bad anatomy), text, jpeg artifacts, (ugly), (morbid), (blurry), (low quality), (long leg), (poorly drawn), (bad proportions),

[image]

[image]

[image]

[image]

license: osl-3.0 ---