wwjCMP
wwjCMP
> > hermes-3-llama-3.1:8b-q8_0 has an infinite output issue . > > Works fine on my end, what prompt did you use? Did you set any parameters or changed anything or...
> > hermes-3-llama-3.1:8b-q8_0 has an infinite output issue . > > Works fine on my end, what prompt did you use? Did you set any parameters or changed anything or...
> This is partially because the RAG prompts are in English. And as you mentioned it also depends on the model you use. > > Will need to experiment with...
I understand, and I hope at least to achieve multi-conversation saving, so it's convenient to continue a conversation at any time.
> Closing this as conversation history loading is supported now.  Please cancel automatic saving of conversations and use manual saving.
After loading chat, the location of chat box will change.
> @fabiensc0ville have you checked the doc and set the context window explicitly in Ollama? https://github.com/logancyang/obsidian-copilot/blob/master/local_copilot.md#ollama > > Also you didn't mention which model you were using, how long is...
通常是在打开 edit card in the library的时候出现
有可能是新打开的（新创建的）文件直接放到了card library的编辑界面里面导致的错误？
好像是和readitlater插件相冲突。两个插件同时打开的时候用它新建的文件会直接替换掉canvas中的文件的内容。