v2I lowered the LR by increasing the resolution of the images in the dataset to improve the artifacts.This LORA is still under-trained, in my opinion, but this method is too time consuming, so training was discontinued.<hr />832 x 1216, 30 images.No caption.sd-scripts : fine tuningoptimizer : adafactorscheduler : constantLR : 2e-6batch size : 6epochs : 350repeats : 1<hr />v1Slightly artifacted.Enabled by prompts such as anime, manga, etc.<hr />Trained on 416 x 608, 27 images.No caption.Parameterssd-scripts : fine tuningoptimizer : adafactorscheduler : cosineLR : 7.5e-6batch size : 3epochs : 210repeats : 1

【Style】 Wasabi - FLUX

FLUX_Style1863322422_v2_Dim16.safetensors

LORA Dim16 v2

Unet v2

LORA Dim16

Unet

<div>Slightly artifacted.Enabled by prompts such as anime, manga, etc.<hr />Trained on 416 x 608, 27 images.No caption.Parameterssd-scripts : fine tuningoptimizer : adafactorscheduler : cosineLR : 7.5e-6batch size : 3epochs : 210repeats : 1

Files with SHA256: 6523918e9d15daf3d911769832d04fd4061c2e7ab246a281bfa4b9be0120a01a