3 ms·
Also the 3588 chips can run LLMs on the NPU. Not quite llama.cpp level easy but definitely doable. For 7B class models the speed is usable
by Havoc 2y ago
Also the 3588 chips can run LLMs on the NPU.
Not quite llama.cpp level easy but definitely doable.
For 7B class models the speed is usable