Codename "KW" is an artificial character that was created in research lab. The dataset was heavily edited in multiple software and not intended to imitate any person in real.Lora was created for research and evaluation purposes.===========================================================Beta110 was trained without Text Encoder quantization, and optimizer EPS changed to 1e-8===========================================================Both Beta63 and Beta71 increase the precision to fp32, but Beta63 is dedicated for NSFW===========================================================Beta30 increases the dimensions to 256.===========================================================Beta29 is the rework on 128 dimensions due to overtrain found, according to the latest test list.===========================================================Beta26 increases the dimensions to 128.===========================================================Beta22 increases the dimensions to 64.===========================================================Beta15 increases the dimensions to 32, in order to hold more details.===========================================================Beta2 was found over trained after many pictures were drawn. Therefore the Beta17 is released with the steps reduced.===========================================================kwFemale FLUX Beta2 is the prototype version of LoRA for the flux.1-dev modelkwFemale series are created in lab for research purposes. The objectives are looking for optimal dataset and training setting in order to produce the best result.After the introducing of new model FLUX, the research of lora is also expanded to this model. Please feel free to visit other loras for more details and create pictures.

kwFemale FLUX/KREA

The AI-Toolkit settings are as follows:Optimizer: AdamW8bitMix Precision: bf16Save Precision: float16Learning rate: 1e-4Unet LR: 1e-4Linear: 16Linear_alpha: 16Resolution: [ 1536 ]Batch Size: 1

Files with SHA256: 3832defa8f0205372714302cf16f8c2cc6eb4eaa8828266c8257f8394dc9520e