⚡️ Qwen-Image-Edit Lightning

✨ 4-step inferencing with lightx2v's LoRA.

📝 Local Prompt Enhancement, Batched Multi-image Generation, 🎨 Preset Batches

Preset Batch Generation

Setting this to true will pass the basic prompt(s) generated via the static preset template to a secondary LLM tasked with improving the overall cohesiveness and details of the final generation prompt.

0 2147483647
1 10
1 16
1 4
Hint: depending on the original image, prompt quality, and complexity, you can often get away with 3 steps, even 2 steps without much loss in quality.