I run this on a Koboldcpp custom endpoint locally. When I chat with a model, it sends my first message fine and the model will start generating a response. Once the response finishes, that response is then looped right back to the LLM again instead of being sent to my 'chat window' for me to see. It's causing an endless loop of the AI 'talking to itself' in my powershell. I think you have a value in your code set incorrectly somewhere, I never get the response that is intended for me. It just keeps se...

After the last update, everything stopped working. The TTS can't generate audio, and the LLM can't answer questions; it just keeps showing 'please go on.' Everything is broken. And of course, after that, now I have to free up 30 GB of space to hear my language....

【如何调用工具或者让AI换装换动作】 软件左侧按钮点击第一个“AI模型”按钮，在右侧展开的AI模型管理面板中勾选 “允许AI使用工具和更换服装动作” 【为什么有时候AI无法正常调用工具】 1. 建议从创意工坊订阅或者下载7B或更高的AI模型，比默认的1.5B的模型更智能更精确调用工具。 2. 建议经常清理聊天记录，或者在软件右下角点击加号按钮新建会话，不然过多的聊天记录会干�...

老显示"Neither 'from' or 'files' was specified"

How much cpu/gpu/ram does this use I know it will be based on model but whats a decent llm to run with this and not eat up 100% resources so I can still play games and stream?