Yufeng Li ceeb1a65d6 Add quantization support of GEMM directly with QGemm (#8447) ... QGemm takes in quantized A, B, C, and quantization parameters of output Y, in which C and quantization parameters of Y are optional. Its output can be quantized or full precision, which depends on whether quantization parameters of Y exists or not. If quant params of Y are provided, the output will be requantized or is full precision. Comparing with QLinearMatMul and MatMulInteger, QGemm supports transpose, apha and beta attribute. The formula for quantized GEMM is: Y = alpha * scale_a * scale_b * ((A_int8 - zp_a) * (B_int8 - zp_b) + C_int32), in which, C_int32 is quantized with formula: C_int32 = (beta * C) / (alpha * scale_a * scale_b)		2021-07-27 21:21:49 -07:00
..
logging	Fix logs getting skipped in single-line conditionals. (#7589)	2021-05-07 15:40:47 -07:00
cuda_op_test_utils.h
denormal_test.cc	Update googletest to latest commit to fix build issues with GCC11 (#7984)	2021-06-08 16:06:53 -07:00
path_test.cc
quantization_test_utils.h	Add quantization support of GEMM directly with QGemm (#8447)	2021-07-27 21:21:49 -07:00
string_utils_test.cc	[NNAPI EP] Make partitioning stop ops configurable. (#8444)	2021-07-22 09:21:42 -07:00
tensor_op_test_utils.cc	Loosen tolerance of CudaKernelTest.ReduceSum_MidTensor, allow test random seed to be regenerated within a test run. (#5675)	2020-11-03 10:37:00 -08:00
tensor_op_test_utils.h	QDQ implementation (#7033)	2021-03-25 09:17:23 -07:00
utf8_util_test.cc	Replace functions with secured version for OSX compliance (#7586)	2021-07-08 11:02:36 -07:00