Fix: reattach @spaces.GPU to run_lora; helper defs moved above decorator (fixes SyntaxError and moves LoRA download out of the GPU window)

Prefetch + cache remapped style LoRA outside the GPU window; raise GPU duration to 120s

Fix custom LoRA loading: remap ai-toolkit diffusion_model.* keys to diffusers transformer.*

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Update app.py

Create app.py