公开(公告)号：US20210350210A1

公开(公告)日：2021-11-11

申请号：US17041336

申请日：2018-07-30

Applicant: INTEL CORPORATION

Inventor： Jiong GONG ,  Haihao SHEN ,  Xiao Dong LIN ,  Xiaoli LIU

IPC: G06N3/04 ,  G06F7/483 ,  G06F17/16 ,  G06N3/08 ,  G06N3/063

Abstract: A method and apparatus for keeping statistical inference accuracy with 8-bit winograd convolution. A calibration dataset and a pretrained CNN comprising 32-bit floating point weight values may be sampled to generate an input activation tensor and a weight tensor. A transformed input activation tensor may be generated by multiplying the input activation tensor and an input matrix to generate a transformed input activation tensor. A transformed weight tensor may be generated by multiplying the weight tensor and a weight matrix. A scale factor may be computed for each transformed tensor. An 8-bit CNN model including the scale factors may be generated.