iChristGit

> I think you want > > ``` > ### System:\n{{preprompt}}\n\n### User:{{#each messages}}{{#ifUser}}{{content}}\n\n### Assistant:{{/ifUser}}{{#ifAssistant}}{{content}}\n\n### User:{{/ifAssistant}}{{/each}} > ``` > > Although I haven't tried it. Make sure you add the correct...

If you try llamacpp it will work fine, its something with Text-gen-webui specifically, hopefully devs will find out why.

@nsarrazin I think he mean just pasting an image url into the chat to have it shown as beauty, not using ViT or Clip to interrogate it.

Should this mean we could also add Text files as well as HTML files? this would be amazing

> Hi, any update here? I have successfully enabled the path: llama.cpp-./server-llama 3 70B.gguf->chat-ui, from localhost:8080, my Mac can see the webpage. However, my phone from the same LAN can't...

Confirmed as well with my iPhone 11 Pro, after latest update it crashes with the "native video player" option, without the option is runs fine!

Also for me with a 3090 Inference took too long I just use a 20GB command-r verison, normal ollama chat works superfast

> You can add a similar logic to the STscript to skip messages on certain conditions without triggering the search. > > Alternatively, if you disable the search trigger on...

@Cohee1207 Any idea about it? If that is something that can be done via a script

> Web search is dumb, it doesn't have a context of your conversions. Queries for follow-up questions like these need to be done with LLM calls, no doubt. So it...