t01nstyle LoRA for Qwen-Image-Edit-2509

Style LoRA trained in T2I mode (--model_version original) on Qwen-Image-Edit-2509 weights. Loads on the full Qwen-Image-Edit-2509 pipeline at inference.

Trigger: t01nstyle

Recommended inference setup:

  • Pipeline: Qwen-Image-Edit-2509 (multi-reference + ControlNet pose stack)
  • LoRA scale: 0.7–0.9 (not 1.0 β€” preserves identity from reference images)
  • Sampler: FlowMatch / Euler, 25–40 steps
  • True CFG: 4.0
  • Resolution: 1024Β² or native 1328Β²

Training:

  • Tuner: musubi-tuner (kohya-ss), --model_version original
  • Dataset: 256 art targets in t01nstyle, captions only (no control images)
  • Network rank/alpha: 32 / 32
  • Learning rate: 1e-4 (adamw8bit, constant_with_warmup, 200 warmup steps)
  • Timestep sampling: shift, discrete_flow_shift 2.2, num_timestep_buckets 4
  • 30 epochs, ~4,140 steps
  • Hardware: H100 SXM 80GB

Why T2I mode on an Edit model: training in T2I mode (without control images) gives the LoRA a clean style-only signal. At inference on the Edit pipeline, ControlNet handles pose, multi-reference handles identity, and this LoRA only handles style β€” no conflicts between components.

Checkpoints provided: epochs 12–28 (every 2 epochs) + final (epoch 30). Best checkpoint typically around epochs 18–25.

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. πŸ™‹ Ask for provider support

Model tree for KassandraToon/t01nstyle-qie2509-lora

Adapter
(87)
this model