wwjCMP

> > 目前暗黑模式下，加粗字体是这样的现实。保证了文字颜色的一致性。 对比度不足，需要和其他不加粗文字一起看才能看出来

> 新版本 v1.3.6 已经可以发送图片给自定义模型了 使用ollama 部署的 llava 模型接受不到图片

I think it is more valuable that the new MOC is on the same canvas.

> ok,now i run a customized embedding model [Dmeta-embedding-zh](https://huggingface.co/DMetaSoul/Dmeta-embedding-zh) successfully with [ollama](https://github.com/ollama/ollama), since [ollama](https://github.com/ollama/ollama) using [llama.cpp](https://github.com/ggerganov/llama.cpp) as inference, and it support BERT which is architecture of most of embedding model,...

> ok,now i run a customized embedding model [Dmeta-embedding-zh](https://huggingface.co/DMetaSoul/Dmeta-embedding-zh) successfully with [ollama](https://github.com/ollama/ollama), since [ollama](https://github.com/ollama/ollama) using [llama.cpp](https://github.com/ggerganov/llama.cpp) as inference, and it support BERT which is architecture of most of embedding model,...

> > > 好的，现在我用[ollama](https://github.com/ollama/ollama)成功运行了一个定制的嵌入模型[Dmeta-embedding-zh](https://huggingface.co/DMetaSoul/Dmeta-embedding-zh)，因为[ollama](https://github.com/ollama/ollama)使用[llama.cpp](https://github.com/ggerganov/llama.cpp)作为推理，并且它支持 BERT，这是大多数嵌入模型的架构，以下是步骤：[](https://github.com/ollama/ollama)[](https://github.com/ollama/ollama)[](https://github.com/ggerganov/llama.cpp) > > > > > > * 将拥抱脸模型转换为GGUF文件 > > > > > > ``` > > > 1. git clone https://github.com/ggerganov/llama.cpp.git > > >...

I think I got it

> 还是没有反应，已经升级到2.2.4，pdf用chrome浏览器打开

> @wwjCMP 是框选后右键使用吗, pdf不支持浮动工具 我在另一台电脑发现好用，应该是我这边的问题

> Exactly! > […](#) > On Sun, Mar 17, 2024 at 21:20:51, Wills < ***@***.*** > wrote: > > > @ StanGirard ( https://github.com/StanGirard ) Thanks for looking into >...