Chao Liu a65ef90308 device_implicit_gemm_convolution_1_chwn_csrk_khwn: use tensor copy (instead of pointwise) for writing output, 3x3 increased from 78% to 84%, 5x5 from 80% to 84%

2019-02-19 11:47:46 -06:00

..

CMakeLists.txt

update build

2019-02-15 02:06:34 -06:00

device_direct_convolution_1.hpp

change file extension to hip.hpp and hip.cpp

2019-02-15 02:13:21 -06:00

device_direct_convolution_2.hpp

change file extension to hip.hpp and hip.cpp

2019-02-15 02:13:21 -06:00

device_implicit_gemm_convolution_1_chwn_csrk_khwn_padded.hpp

change file extension to hip.hpp and hip.cpp

2019-02-15 02:13:21 -06:00

device_implicit_gemm_convolution_1_chwn_csrk_khwn.hpp

device_implicit_gemm_convolution_1_chwn_csrk_khwn: use tensor copy (instead of pointwise) for writing output, 3x3 increased from 78% to 84%, 5x5 from 80% to 84%

2019-02-19 11:47:46 -06:00

device_implicit_gemm_convolution_1_nchw_kcsr_nkhw.hpp

change file extension to hip.hpp and hip.cpp

2019-02-15 02:13:21 -06:00

device_implicit_gemm_convolution_1_nchw_srck_nkhw.hpp

change file extension to hip.hpp and hip.cpp

2019-02-15 02:13:21 -06:00

device_implicit_gemm_convolution_2_chwn_csrk_khwn.hpp

2-type implicit gemm using chwn

2019-02-15 22:51:51 -06:00

device_implicit_gemm_convolution_2_cnhw_csrk_knhw.hpp

change file extension to hip.hpp and hip.cpp

2019-02-15 02:13:21 -06:00

driver.cu

update build

2019-02-15 02:06:34 -06:00

driver.hip.cpp

device_implicit_gemm_convolution_1_chwn_csrk_khwn: use tensor copy (instead of pointwise) for writing output, 3x3 increased from 78% to 84%, 5x5 from 80% to 84%

2019-02-19 11:47:46 -06:00