edgchen1 6c7da5e9d3 Optimize CUDA Sum op kernel and refactor CUDA elementwise variadic input op kernels (#4418) ... For the special case where all variadic inputs of a kernel are the same shape (i.e. no broadcasting is required) and there are few enough of them, we perform the entire computation in a single kernel. The general implementation (which was previously used for this special case) handles broadcasting by repeatedly invoking a binary kernel on successive inputs.		2020-07-10 10:20:23 -07:00
..
alloc_kind.h
allocator.h	Cleanup SessionState. Move allocator lookup to SessionState. (#4194)	2020-06-28 14:55:42 +10:00
customregistry.h
data_types.h	address PR comments (#3312)	2020-03-25 19:35:12 -07:00
data_types_internal.h	Fix some warnings on Windows (#2560)	2020-01-22 15:59:11 -08:00
endian.h
execution_provider.h	Cleanup SessionState. Move allocator lookup to SessionState. (#4194)	2020-06-28 14:55:42 +10:00
fence.h
framework_common.h
func_api.h	Ryanunderhill/mkldnn dll (#3314)	2020-05-06 00:57:09 -07:00
kernel_def_builder.h	Introduce training changes.	2020-03-11 14:39:03 -07:00
kernel_registry.h	Parallel all the activations ops (#3722)	2020-05-05 01:18:17 -07:00
ml_value.h	Introduce container type runtime checks and other improvements (#2522)	2019-12-04 16:04:17 -08:00
op_kernel.h	Optimize CUDA Sum op kernel and refactor CUDA elementwise variadic input op kernels (#4418)	2020-07-10 10:20:23 -07:00
op_kernel_info.h
op_node_proto_helper.h
run_options.h	Support training_mode flag in eval (#4324)	2020-07-08 10:38:54 -07:00
sparse_tensor.h
tensor.h	View Op - new unit tests and add support for tensor memcpy by offset/size (#3439)	2020-04-07 13:07:11 -07:00
tensor_shape.h