Fork自 PaddlePaddle / PaddleDetection
* Add a fake_quantize_op, which quantize an input tensor to a tensor with lower bits.
拖放文件到此处或点击上传