Ollama-扩充提示词 / Add details to prompt words

3.11 / DeepSpeek을 사용하세요

---

嗨嗨嗨，懒癌发作了，又折腾了个小玩意

想跑涩涩但不知道该选什么场景？答：让AI来搞定

最原始的想法是，我只要输入：1girl,sex,indoors，它就能自动帮我生成场景、动作和角色

但Ollama的大模型不支持涩涩内容，那我们退而求其次，先生成场景和角色，之后再自己加进去吧（

我懒，所以想了个办法帮自己生成图片

最初的构想是，只要输入“1girl,sex,indoors”，它就能自动生成图片

但因为不支持涩涩内容，所以我们改用它来生成角色和场景

示例如下：/ Here's an example:

最上面那一排是Ollama的节点，左边那个类似GPT，能做很多事，我还没仔细研究，只是简单拿来当翻译用；右边那个框里的是我们的扩展节点

The top row of nodes is Ollama's nodes, with the left one being similar to GPT (it can do many things) - I haven't studied it in detail yet, I'm just using it as a simple translator. The right frame contains our custom node.

关于step5，我觉得还有很多可以优化改进的地方，目前我暂时够用就不想动了（懒），大家集思广益一下，有什么优化建议欢迎评论！

Regarding step 5, I think there are still many places that can be optimized and improved. For now, I don't have any ideas to update (lazy). Feel free to comment if you have any suggestions for optimization.

参考/Related article：https://www.bilibili.com/video/BV1Mr42177Y3/?spm_id_from=333.1391.0.0&vd_source=91b0ae62c126dc5c3bd019f5fec437b4

https://openart.ai/workflows/datou/prompt-polisher-ollama/m1Q4k4YnpKmtAmb0yH7Y

https://civitai.com/images/45045266

step5 的参考tag ↑ 一开始也是发现他这个自然语言跑的图才起了兴趣折腾的这个Ollama →https://civitai.com/images/45045477

Step 5 reference tag ↑. I started to be interested in generating images using natural language processing with Ollama because of the pictures from this author.→https://civitai.com/images/45045477