Please Donate Buzz for FLUX Lora Training !

FLUX v1.0 :

Trained on FLUX.1 [dev] with 120 photos of American Model Gigi Hadid with detailed GPT-4 captions. 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:Gigi_Hadid_FLUX_v1-000037:1>

SDXL v2.0 :

Trained on Juggernaut X with 220 photos of Gigi Hadid with improved likeness due to consistency and an improved Lora model. Tested on Juggernaut X, Juggernaut v7, RealismEngine 2, RealVisXL3, AlbedoBaseV2! Use with keyword : "gixhdid" . Use around strength 1.0-1.1. CFG 5.0-7.0. Clipskip 1. Can be used for example as follows:

Positive : {Artstyle}, {Character and scene description}, gixhdid, <lora:gihadid_juggerX_xl_1_wocap_merger_61_113_162_03_05_02-gixhdid:1>

Negative : ugly, deformed, airbrushed, photoshop, rendered, (multiple people), child

SDXL v1.0:

Lora trained on 230 image of Gigi Hadid and tagged by GPT 4-Vision and COG VLM locally. The model is trained on the SDXL 1.0 Base with a small learning rate to ensure maximum control when choosing the best checkpoint. Lora strength should be between 0.7-1.1.