PJ0 QwenImageArt stage1

A small FFT of the QwenImageArt model.
This is the beginning of work on this model, focused on improving realism.
I welcome constructive feedback and suggestions.


Performance

  • Recommended samplers: Euler/Beta57, Re_multistep/Simple
  • Generating one 1024?1024 image with 20 steps on RTX 3090 takes ~1m 15s

Examples

пример 1
пример 2
пример 3
пример 4
пример 5
пример 6
пример 7
пример 8


License

Apache-2.0

Downloads last month
93
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support