Chen Fu cd0af7ad44 Symmetric quantized convolution kernel ARM64 (#9772) ... Adding a symmetric quantized convolution kernel for ARM64 Note: Indirect conv performs worse for shallow convs (input channels are small). This is much more so for low end pre-dot CPUs, where only 128 or deeper conv is faster with indirect conv. With DOT-CPUs, 32 deep conv is already faster Co-authored-by: Chen Fu <fuchen@microsoft.com>		2021-12-13 21:14:45 -08:00
..
coreml	[CoreML EP]Make coreml ep build on non-macOS platform (#8677)	2021-08-18 09:35:32 -07:00
cpu	Symmetric quantized convolution kernel ARM64 (#9772)	2021-12-13 21:14:45 -08:00
cuda	Change TensorShape to typically not allocate heap memory (#9542)	2021-11-08 10:29:54 -08:00
dnnl	Add Transpose, Reshape, Pow and LeakyRelu ops to DNNL execution provider (#9180)	2021-10-13 10:20:07 -07:00
internal_testing	Change some const to constexpr in unit tests code(#10002)	2021-12-10 12:26:59 -08:00
nnapi	[NNAPI EP] Fix MaxPool error using uint8 (#9129)	2021-09-21 00:44:43 -07:00
rknpu	Change onnxruntime::make_unique to std::make_unique (#7502)	2021-04-29 17:04:53 -07:00
tensorrt	Ryanunderhill/cuda shared (#7626)	2021-05-20 07:53:47 -07:00
compare_provider_test_utils.cc	Update C++ Standard from 14 to 17 (#8041)	2021-06-25 14:08:01 -07:00
compare_provider_test_utils.h	Add SoftmaxCrossEntropyLossInternal to Support Dynamic ignore_index Input (#7899)	2021-06-09 10:29:46 +08:00
get_execution_providers_test.cc
kernel_def_hash_test.cc	add fallback of s8s8 support on x64 (#9995)	2021-12-10 11:33:19 -08:00
memcpy_test.cc	Reduce templatization of C API and refactor for InitOrtValue (#8700)	2021-08-12 16:51:18 -07:00
op_kernel_type_control_test.cc
provider_test_utils.cc	Change TensorShape to typically not allocate heap memory (#9542)	2021-11-08 10:29:54 -08:00
provider_test_utils.h	Support disabling support for the optional type in ORT builds (#9745)	2021-11-17 19:13:28 -08:00