5/15/2024 Add keep_alive support
0 means releasing the video memory immediately after loading the model, while 60m represents that the model will not be released from the video memory until 60m has passed since it was loaded.
This is just an integrated project. I refer to the ComfyUI-Prompt-MZ project prompt word template setup and comfyui-ollama project to make the prompt word generation more consistent with the requirements of stable diffusion
The projects which I referenced are
https://github.com/stavsap/comfyui-ollama
https://github.com/MinusZoneAI/ComfyUI-Prompt-MZ
I would like to express my special thanks to the authors of these two projects
1、Install ComfyUI
2、git clone in the custom_nodes folder inside your ComfyUI installation or download as zip and unzip the contents to custom_nodes/ComfyUi-Ollama-YN. The git git command is git clone https://github.com/wujm424606/ComfyUi-Ollama-YN.git
3、Start/restart ComfyUI
You can download and install ollama from this website
1、You can choose models by https://ollama.com/library/llama3:8b-instruct-q4_K_M website
2、Input such as ollama run llama3:8b-instruct-q4_K_M in CMD in any folders
First function:
This function is to deduce the prompt words according to the picture
The default modle is llava model. If you have a better model, you can change the model by changing the model name
Second function:
This function is a simple question and answer function.
The default modle is llama3 model. If you have a better model, you can change the model by changing the model name
Third function:
This function allows the model to be embellished according to your prompt words
Four function:
This function is to answer questions by contacting above
Five function：
This function is to generate prompt words that more closely follow the stable diffusion pattern
If you get an error that says there is no model, install the model as described above
This project involves installing model commands in CMD
1、ollama run llama3:8b-instruct-q4_K_M
2、ollama run llama3
3、ollama run phi3
4、ollama run phi3:3.8b-mini-instruct-4k-q4_K_M
5、ollama run phi3:3.8b-mini-instruct-4k-fp16
6、ollama run llava
5/15/2024 加入keep_alive支持
0为加载模型后即释放显存，60m代表模型在显存内存在60m后才会释放
这是一个整合的项目，我在comfyui-ollama 项目的基础上参考了 ComfyUI-Prompt-MZ 项目的提示词和正则代码使生成的提示词更符合stable diffusion的格式
参考的两个项目是
https://github.com/stavsap/comfyui-ollama
https://github.com/MinusZoneAI/ComfyUI-Prompt-MZ
在这里我向这两个项目的作者表示由衷的感谢
1、安装ComfyUI
2、在custom_nodes 文件夹下在cmd页面输入git clone https://github.com/wujm424606/ComfyUi-Ollama-YN.git 完成项目的安装 或者 下载zip文件将文件解压到custom_nodes/ComfyUi-Ollama-YN目录
3、重启ComfyUI
点开上面的网址，然后点击下载，然后进行安装即可
0、模型安装的默认位置是C盘的用户文件夹下的.ollama文件夹里，这个下载目录可以通过设置全局变量进行修改，修改方法参考 https://blog.csdn.net/Yurixu/article/details/136443395
1、安装完ollama后在后台中需要将ollama保持启动状态
2、点开 https://ollama.com/library 网站，搜索然后选择好要安装的模型
例如
3、然后在任意文件夹里面调出cmd页面，输入ollama run llama3:8b-instruct-q4_K_M
4、安装完成后关闭cmd页面即可
功能一:
这个功能是推导图片的提示词
这个默认模型是 llava模型。如果你有更好的模型，可以在安装好新模型后，在原有模型名字文字修改成新模型的名字，即可以调用
功能二:
这个功能是简单问答功能
这个默认模型是 llama3模型。如果你有更好的模型，可以在安装好新模型后，在原有模型名字文字修改成新模型的名字，即可以调用
功能三:
这个功能是根据文本进行提示词润色
功能四:
这个功能是通过联系上文进行问答功能
功能五：
这个功能是根据输入生成更符合stable diffusion格式的提示词
如果你的后台出现没有模型的错误，请参考上面选择对应的模型进行安装
这是这个项目中涉及的安装的模型的运行命令
1、ollama run llama3:8b-instruct-q4_K_M
2、ollama run llama3
3、ollama run phi3
4、ollama run phi3:3.8b-mini-instruct-4k-q4_K_M
5、ollama run phi3:3.8b-mini-instruct-4k-fp16
6、ollama run llava