daiaji
daiaji
Can this feature be supported? Basically, when I deploy in pod, I can directly use the localhost:3000 reverse proxy for NextWeb, but [one-api](https://github.com/songquanpeng/one-api) also uses port number 3000, which causes...
> this one you need to changed it > > https://github.com/ChatGPTNextWeb/ChatGPT-Next-Web/blob/main/Dockerfile#L43C1-L44C1 This appears to be just docker's port mapping.
用V2转发就没事了😆. 挺难泵的。
> 这种功能需要建立向量索引，在纯前端比较难搞，你看到的开源项目都是把文件传到服务器做处理。 > > 不过我在研究怎么在浏览器里跑 embedding 模型，之后会加此功能。 用js跑embedding模型？纯前端听起来很酷。
> 此功能将于 v2.9 版本加入。 结合现有的历史摘要功能，是否可以实现把每一个生成的历史摘要向量化到向量数据库里，然后实现GPT对于整个事件的长期记忆，而不是只局限于上下文和近期的历史摘要？
> > 此功能将于 v2.9 版本加入。 > > 结合现有的历史摘要功能，是否可以实现把每一个生成的历史摘要向量化到向量数据库里，然后实现GPT对于整个事件的长期记忆，而不是只局限于上下文和近期的历史摘要？ 从这个实践的最后示例来看，似乎是可行的。 https://github.com/TomLBZ/koishi-plugin-openai
@duncanthrax While the process of building the kernel is quite painful But I try to build receivers and report after I've built the kernel and rootfs.
@duncanthrax I don't know what to do. ``` aarch64-linux-gnu-gcc -Wall -o scream-pulse scream-pulse.c -lpulse-simple -lpulse /usr/lib/gcc-cross/aarch64-linux-gnu/8/../../../../aarch64-linux-gnu/bin/ld: cannot find -lpulse-simple /usr/lib/gcc-cross/aarch64-linux-gnu/8/../../../../aarch64-linux-gnu/bin/ld: cannot find -lpulse collect2: error: ld returned 1 exit status...
@duncanthrax In fact, I gave up cross compilation. It doesn't seem so slow to compile Receivers on ARM devices. After installing pulseaudio, scream-pulse seems to be able to run. But...