Lora + Image-to-Image + FLUX

This is a workflow that uses images to reverse-engineer prompts through natural language, with automatic caption and prompt reverse-engineering. The resulting prompts are better aligned with FLUX's understanding logic, and the image generation or tagging results are more suitable for FLUX compared to typical reverse-engineering tools. Personally, I find the results nearly perfect (though some adjustments may be needed depending on the Lora used).
Have fun, and feel free to leave any questions in the comments section!

---------------------------------------------------------------------------------------------------------------------

lora+图生图+FLux

这是一个利用图片通过自然语言反推提示词的工作流，支持自动打标与提示词反推。生成的提示词更加贴合FLUX的理解逻辑，图像生成效果或打标效果相比一般的反推工具，显得更加契合FLUX。个人认为效果接近完美（不过不同的Lora可能会有所差异，还需要根据实际情况进行调整）。
希望大家玩得开心，遇到问题可以在评论区留言哦！