Please Donate Buzz for FLUX Lora Training !
FLUX v2.0 :
Trained on FLUX.1 [dev] with 180 photos of Josephine Skriver with GPT-4o captions and a larger model. Tested on FLUX 1.D (full) and FLUX fp8 and FLUX nf4 ! Use around strength 0.8-1.2. No keywords needed! Distilled CFG around 1-4 and CFG 1.0 (without negative prompt). Clipskip 1. Can be used for example as follows:
Positive : {Artstyle, Character and scene description in usual FLUX fashion}, <lora:Josephine_Skriver_FLUX_larger_dataset_v2_merger_37_58_07_03:1.0>
FLUX v1.0 :
Trained on FLUX.1 [dev] with 163 photos of Josephine Skriver. Tested on FLUX 1.D (full) and FLUX fp8 and FLUX nf4 ! Use around strength 0.8-1.2. No keywords needed! Distilled CFG around 1-4 and CFG 1.0 (without negative prompt). Clipskip 1. Can be used for example as follows:
Positive : {Artstyle, Character and scene description in usual FLUX fashion}, <lora:Josephine_Skriver_FLUX_v1-000034:1>
SDXL v2.0 :
Trained on Juggernaut X with 220 photos of Josephine Skriver with improved likeness due to consistency and an improved Lora model. Tested on Juggernaut X, Juggernaut v7, RealismEngine 2, RealVisXL3, AlbedoBaseV2! Use with keyword : "joxskriver" . Use around strength 1.0. CFG 5.0-7.0. Clipskip 1. Can be used for example as follows:
Positive : {Artstyle}, {Character and scene description}, joxskriver, <lora:joskriver_juggerX_xl_1_woc-joxskriver-000093:1>
Negative : ugly, deformed, airbrushed, photoshop, rendered, (multiple people), child
SDXL v1.0:
Lora trained on 150 images of Model Josephine Skriver and captioned by GPT4-V. The strength should be kept between 0.9-1.2. The model is biased towards realism and is better for closeups.
