Note: This version still sucks. I don't really know what I'm doing wrong.
Training changes from the previously posted run (v5):
Training Steps: Maximum of
6000>> 10000Base Model Precision:
int8-quanto>> fp8-quantoLearning Rate: 1.0
Optimizer: Prodigy
Batch Size:
1>> 2Gradient Accumulation Steps: 1
Rank:
16>> 8Gradient clipping:
default (probably 2)>> 0.1Caption Strategy:
Triple>> Instance prompt
Input
Per image generation will cost 0.0047$
For premium plan image generation will cost 0.00$ i.e Free.
Output
Unknown content type
Related Models
Discover similar models you might be interested in
High Braid Pony Tail - v1.0
Dewy skin - v1.0
Waldemar von Kozak - V1
Ferrari SF90 XX Stradale - 2024 - v1.0 FLUX
Dreamy and Colorful - v1.0
Flux.2 Dev Image To Image (Image Editing)
anime-art-stylelora
Ringbrothers Octavia - 2025 - v1.0 FLUX
Milo Manara - V1
Underflowered - V1
Artificial Caucasian Girl for Flux - v1.0
Sausage Dishes - White Sausage-Dev v1.0