Comment by euniceee3

Tensor chips are not open enough for an optimized mobile LLM to be ran on them.