v2.0：

与v1.0一样，我希望能尽量减少工作流的操作量与保持界面简洁。

添加了二次采样放大，添加了扩图功能、颜色匹配与使用模型放大，为文生图模式添加了controlnet控制，为单图编辑添加了角度控制。这些功能单独拆出来都很好做，但全部综合在一起又要减少操作量，确实稍微麻烦了一些。

具体一些的使用方法写在工作流内。

（碎碎念：我知道我的工作流排版有点乱，因为我老是想让节点之间少一点空隙......您可以随意调整节点组的位置，它们之间都是独立的，组与组之间没有直接的连线，不用担心变得更乱。）

---

v2.0:

As with v1.0, the goal remains to minimize the number of operations in the workflow and maintain a clean interface.

New additions include hi-res, an image expansion (outpainting) feature, color matching, and upscaling. ControlNet support has been added for text-to-image generation, and angle control has been added for single-image editing. Implementing each of these features individually would be straightforward, but integrating them all together while still reducing operational complexity presented a bit of a challenge.

More detailed usage instructions are provided in the workflow.

(PS: I know the layout of my workflow looks a bit messy — that’s because I always try to leave as little empty space between nodes as possible. Feel free to rearrange the node groups however you like. They’re independent from each other, with no direct connections line between groups, so don’t worry about it getting even messier.)

v1.0：

自用的，没有什么额外的功能，也没有用什么额外的插件节点，只是将文生图、单图编辑与双图编辑整合在一个工作流里：简洁、方便。

用法：默认为文生图， bypass节点为单图编辑与双图编辑开关，需要用什么就打开什么，打开后文生图自动失效。

---

For personal use. It has no extra features and doesn't require custom nodes—it simply combines text-to-image, single-image editing, and dual-image editing into one workflow: clean and convenient.

how to use: default: text-to-image,

single-image and dual-image edit switches are in the bypass node; turn on whichever you need—once enabled, text-to-image is automatically disabled.