caffe2::Int8FCDNNLowPPackedWeightBlob Struct Reference

Packed weight matrix for DNNLOWP Int8FC operator. More...

#include <fbgemm_pack_blob.h>

Inheritance diagram for caffe2::Int8FCDNNLowPPackedWeightBlob:

Data Fields
std::vector< dnnlowp::TensorQuantizationParams >	qparams
std::shared_ptr< std::vector< std::int32_t > >	column_offsets
Tensor	original_tensor {CPU}
std::shared_ptr< std::vector< std::int32_t > >	bias
std::shared_ptr< fbgemm::PackBMatrix< std::int8_t > >	W
std::shared_ptr< fbgemm::PackBMatrix< std::int8_t, std::int16_t > >	W_acc16
std::shared_ptr< fbgemm::CompressedSparseColumn >	W_outlier
int	nbits_in_non_outlier

Detailed Description

Packed weight matrix for DNNLOWP Int8FC operator.

Definition at line 15 of file fbgemm_pack_blob.h.

---

The documentation for this struct was generated from the following file:

• caffe2/quantization/server/fbgemm_pack_blob.h