trained on 40 Film images with unique borders.
If you want to render text in the borders go for 50+ steps. To simply emulate film strip design, 32 steps is enough.
This lora was trained on Flux 1 dev Model and works well with the base model. Krea Dev might have some issues with the border effect and place texts inside image rather than on the border.

Left, Flux.1 Dev; right, Flux.1 Krea Dev