🚀 Z-Image Turbo Image-to-Image is Live — Unlimited Generations for Premium Users

Try Now
HotArt - HotArt3-SPO thumbnail

ModelsLab/HotArt - HotArt3-SPO

hotart-hotart3-spo
Text to Image Community ModelFree for Premium UsersLLMs.txt

This version adds Step-aware Preference Optimization (SPO) LORA, [a novel post-training approach that independently evaluates and adjusts the denoising performance at each step, using a step-aware preference model and a step-wise resampler to ensure accurate step-aware supervision... blah, blah, blah = makes the output way artfully beautiful, a.k.a. - NICER !!!]

P.S.: The SPO approach is better than Direct Preference Optimization (DPO) that has extended its success from aligning large language models (LLMs) to aligning text-to-image diffusion models with human preferences. Unlike most existing DPO methods, which assume that all diffusion steps share a consistent preference order with the final generated images, this assumption neglects step-specific denoising performance and that preference labels should be tailored to each step's contribution.

API PlaygroundAPI Documentation

Input

Customize your API request

Select models...

Per image generation will cost 0.0047$
For premium plan image generation will cost 0.00$ i.e Free.

Output

Generation complete

Idle

Unknown content type

Related Models

Discover similar models you might be interested in

Skoi Street Scooter [SDXL] - V1

Skoi Street Scooter [SDXL] - V1

Lila - V1

Lila - V1

La Principessa's Jinx from League of Legends - XL PONY - v2.0

La Principessa's Jinx from League of Legends - XL PONY - v2.0

The Von Carstein Couple - Classic Warhammer Fantasy - SDXL v1.0

The Von Carstein Couple - Classic Warhammer Fantasy - SDXL v1.0

BLGrl1 - V1

BLGrl1 - V1