3 ms·Using 8-bit quantization to speed up inference on mobile with TensorFlow Lite1 points by austin_kodra 8y ago