Download as zip and unextract to use.
Lamma is a fake phone for talking to fake people.
This should run on any webbrowser.
This project requires LLamaCPP: https://github.com/ggml-org/llama.cpp/releases Download the zip for your platform and we need llama-server or llama-server.exe. Get a GGUF model from https://huggingface.co/ Gemma models seem to keep character well. Run llama-server like this on windows
llama-server.exe -m modelname.jjuf
and like this on linux/macos
./llama-server -m modelname.gguf
minimize it and open llamaphone.html
Once you're up and running, feel free to delete the (silly) built in chatacters. Hit the + and add a contact, describe your character and attach a 128x128 photo. I've included a handful off AI generated faces in faces.zip.
You can use the included charmaker.html to mass generate characters, but you have to add the images manually.