FULX text2img with IPA Text-to-image generation based on style transfer

Style transfer using text-to-image has better performance than image-to-image（Render Y image in X style）Has higher universality，比如“用X图风格画A panda/An astronaut/A lady”，On the one hand, you don't need to find the Y image you want（A panda/An astronaut/A lady），On the one hand, it also naturally avoids the influence of the style of Y image on the X style you want to transfer。Based on the latest FLUX components,The computational cost is about 28（But it fluctuates with the server's energy consumption calculation rules and scale）