AlekseyCalvin/Eszter_W_style_art_Ver3_Flux_LoRA_BySilverAgePoets

Eszter Winkelmayer Style Psychedelicate Art VERSION 3

HISTORIC 𝒜𝓡𝒯𝚒𝕤𝕥 STYLE ℒ𝓸ℜ𝐴𝘴

||| By SilverAgePoets.com |||

Herein lives the 3rd variant of our Eszter Winkelmayer Art Style Low-Rank Adapter (LoRA) for FLUX-family models...
Like the second variant of the Eszter LoRA, this version draws from an expanded (from Version 1) manually-recaptioned dataset...
This Rank 32 (dim and alpha) variant was fine-tuned over the new Ablated Modification (v2) of FluxDev by aoxo and the modified LongCLIP by zer0int...
At following hyperparameters: 2.0e-04 Transformer Learning Rate, 1e-12 Text Encoder Learning Rate, batch size 1, regular guidance, 786x bucketing...

Alternately,
The first variant of this Eszter LoRA, over vanilla FluxDev, is found HERE!
The second variant, over GENOVA_APEX Flux model by DNA_1_618, is found HERE!

Fine-tuned on around 50 unforgettably self-singular ink illustrations –
...each like a potent alchemical extract of a phenomenon or a concept, of an apparition or a dimension, carried thru under the easy law of a masterful hand...
Set to glue behind our eyes every cosmic topology and unreserved encapsulation spawned amid the limitless vision of their artist:
East Bay's own legendary reality-stylist (and up-and-coming attorney):
ESZTER WINKELMAYER!

Prompt with:

Eszter style psychedelicate illustration art, figuration distinctively stylized by Eszter Winkelmeyer, geometric patterns, ink, intricate linework, irreverent cosmic punk, mandala-like, hyper-structural neo-Symbolism, etc...

A few of the original artworks by Eszter Winkelmayer used in our training data:

Use it with the 🧨 diffusers library

from diffusers import AutoPipelineForText2Image
import torch

pipeline = AutoPipelineForText2Image.from_pretrained('black-forest-labs/FLUX.1-dev', torch_dtype=torch.float16).to('cuda')
pipeline.load_lora_weights('AlekseyCalvin/Eszter_W_style_art_Ver3_Flux_LoRA_BySilverAgePoets')
image = pipeline('your prompt').images[0]