Please Donate Buzz for FLUX Lora Training ! This training cost almost 4000 Buzz and it is not something I can sustain!

FLUX v1.0 (2011-2014):

Trained on FLUX dev with 90 photos of Alexandra Daddario (2011-2014) with detailed GPT-4 captions. Tested on FLUX 1.dev (full) and FLUX fp8 and FLUX nf4 ! Use around strength 0.8-1.0. 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:Alexandra_Daddario_FLUX_v1_merger_40_58_03_07:1>

SDXL v2.0 :

Trained on Juggernaut X with 220 photos of Alexandra Daddario. Tested on Juggernaut X, Juggernaut v7, RealismEngine 2, RealVisXL3, AlbedoBaseV2! Use with keyword : "alxddrio" . Use around strength 0.9-1.1. CFG 5.0-7.0. Clipskip 1. Can be used for example as follows:

Positive : {Artstyle}, {Character and scene description}, alxddrio, <lora:aldaddario_juggerX_xl_1_wocap_merger_79_134_07_03-alxddrio:1>

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

SDXL v1.0 :

Lora trained on 298 images of Alexandra Daddario on the SDXL 1.0 base model. The recommended Lora strength for most use cases is between 0.8-1.2. This model is biased towards realism so if you want highly stylized imagry I would add something like "photo, photorealism, realistic, " etc. in the negative prompt to make generation easier.