PulseLoop brand style (LoRA)

A style LoRA for Stable Diffusion 1.5: soft pastel 3D, glossy clay shapes and gradient backgrounds in the PulseLoop palette. Trigger word: plsloop style.

  • Data: 23 captioned images, varied subjects so it learns the style, not objects
  • Training: LoRA rank 16 on the UNet attention layers, 1,500 steps, 27 minutes on a 6 GB laptop GPU (diffusers + PEFT)
  • Files: plsloop_style_kohya.safetensors (ComfyUI / A1111 / CivitAI) and plsloop_style.safetensors (diffusers)

Base model vs. with this LoRA, same prompts and seeds, subjects not in the training set

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for niyeahhtea/plsloop-style-lora

Adapter
(672)
this model