Another small ai model from google running locally on a Samsung Tablet

To run these ai model locally the stack is ternux a unix like terminal. Download llama.cpp and compile it for android. Then simply run llama-run and provide the url of the ai model gguf file on huggingface.co.

The url

https://huggingface.co/unsloth/gemma-3-270m-it-GGUF

The model size is around 500mb, the smallest i have seen .