7 ### CPU Backend Operation
8 - CPU backend supports per-channel hybrid quantization of int8 type weight and float activation. (TFLite's dynamic range quantization)
10 ### On-device Quantization
11 - _onert_ supports new experimental API for on-device quantization.
12 - As the 1st step, _onert_ supports per-channel hybrid quantization of int8/int16 type weight and float activation.
13 - API requires file path to export quantized model.
16 - _onert_` support minmax recording of each layer as experimental feature. It is not supported by API yet.
17 - Output file format is HDF5. (File format may change later).