Skip to content
Trending storyDeveloping

PrunaAI releases P-Q-Image-2.1 LoRA adapters for six-step Qwen-Image-2.1 generation

1 article1 sourcesince Oct 28Last article 1h ago ·

Overview

AISummary of 1 article

PrunaAI releases P-Q-Image-2.1, a set of v0.2 LoRA adapters that run Qwen-Image-2.1 in six steps without classifier-free guidance (CFG), with generation up to 6.3× faster, according to a ModelScope post on X. The adapters cover text-to-image generation and single- or multi-image editing with up to three reference images. A variant named 6-steps-upsampled folds prompt upscaling into the adapter.

P-Q-Image-2.1 is a LoRA adapter, not a standalone model, so it requires the Qwen-Image-2.1 base weights. It is released under the Qwen Research License Agreement.

Written by AI from the articles below · updated Oct 28, 8:06 AM ET

Check the sources:

Latest developmentOct 28, 8:00 AM ET
PrunaAI releases P-Q-Image-2.1, a six-step LoRA for Qwen-Image-2.1

Article timeline

The articles in this story. Times are ET.

Oct 28
  1. ModelScopeOfficial
    PrunaAI releases P-Q-Image-2.1, a six-step LoRA for Qwen-Image-2.1

    AIPrunaAI releases P-Q-Image-2.1, a set of v0.2 LoRA adapters that run on Qwen-Image-2.1 in six steps without CFG, with generation up to 6.3× faster. The adapters handle text-to-image generation and single- or multi-image editing with up to three reference images, and the 6-steps-upsampled variant folds prompt-upscaling into the adapter. The adapter requires the Qwen-Image-2.1 base weights and is released under the Qwen Research License Agreement.

    Image from @ModelScope2022's post

Heat Trend

Not enough continuous observations to show a trend yet.