| .. |
|
BatchNormalizationEngine.cpp
|
CNTK v2 library: Several bug fixes
|
2016-09-20 22:54:07 -07:00 |
|
BatchNormalizationEngine.h
|
minor change in definition of BatchNormEngine to avoid warnings during compilation with cuda8 compiler
|
2016-10-18 18:46:15 +02:00 |
|
BlockHandlerAVX.cpp
|
Implement code review feedback from clemensm
|
2016-06-24 08:40:48 -07:00 |
|
BlockHandlerAVX.h
|
Fix path differing only in case. Fix hard tabs.
|
2016-09-19 09:13:31 +02:00 |
|
BlockHandlerSSE.cpp
|
enabling cross-compile of the cntk sources for arm64 devices.
|
2016-10-14 10:09:40 +02:00 |
|
BlockHandlerSSE.h
|
Implement code review feedback from clemensm
|
2016-06-24 08:40:48 -07:00 |
|
BlockMultiplier.h
|
Windows: switch to Visual Studio 2015
|
2017-01-10 10:47:35 +01:00 |
|
BlockMultiplierMatrixUtil.h
|
Implement code review feedback from clemensm
|
2016-06-24 08:40:48 -07:00 |
|
BlockMultiplierPlatform.h
|
Implement comments from code review. Have not tried to compile.
|
2016-06-16 16:17:21 -07:00 |
|
CPUMatrix.cpp
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
CPUMatrix.h
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
CPURNGHandle.cpp
|
Windows: switch to Visual Studio 2015
|
2017-01-10 10:47:35 +01:00 |
|
CPURNGHandle.h
|
Align randomization across platforms
|
2016-12-21 17:31:22 +01:00 |
|
CPUSparseMatrix.cpp
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
CPUSparseMatrix.h
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
CUDAPageLockedMemAllocator.cpp
|
Fix path differing only in case. Fix hard tabs.
|
2016-09-19 09:13:31 +02:00 |
|
CUDAPageLockedMemAllocator.h
|
Re-format code using clang-format (plus some post-processing)
|
2016-01-18 09:36:14 +01:00 |
|
CntkBatchNormalization.cuh
|
Windows: switch to Visual Studio 2015
|
2017-01-10 10:47:35 +01:00 |
|
ColumnQuantizer.h
|
tuned >1-bit SGD: odd #quantization levels, range now 4 stddevs (before: 5)
|
2016-08-21 01:33:01 -07:00 |
|
CommonMatrix.h
|
added TempDeviceBuffer to CommonMatrix.h
|
2016-11-19 21:53:24 +01:00 |
|
Convolution.cuh
|
Addressed BS CR comments
|
2016-10-04 14:55:10 +02:00 |
|
ConvolutionEngine.cpp
|
CNTK core: Implemented an optimization to elide the initial zeroing and subsequent accumulation into gradients for nodes with just one parent/accestor node
|
2016-12-12 12:06:17 -08:00 |
|
ConvolutionEngine.h
|
CNTK core: Implemented an optimization to elide the initial zeroing and subsequent accumulation into gradients for nodes with just one parent/accestor node
|
2016-12-12 12:06:17 -08:00 |
|
ConvolveGeometry.h
|
Added test, addressed CR comments
|
2017-01-19 09:50:23 +01:00 |
|
CuDnnBatchNormalization.cu
|
[CNTK Core] Change BatchNormalization default to CuDnn in both BrainScript and V2 C++/Python, and add fallback to cntk engine when not supported by cudnn
|
2017-01-11 01:00:09 -08:00 |
|
CuDnnCommon.cpp
|
Changes Addressing Code Review for CUDNN RNNStack Node
|
2016-08-23 10:40:34 -07:00 |
|
CuDnnCommon.h
|
Address CR comments
|
2016-08-23 20:32:03 +02:00 |
|
CuDnnConvolutionEngine.cu
|
Commenting out conv reference engine
|
2017-01-09 15:33:36 +01:00 |
|
CuDnnFactories.h
|
Introducing global flags
|
2016-09-01 10:26:18 +02:00 |
|
CuDnnRNN.cpp
|
Changes Addressing Code Review for CUDNN RNNStack Node
|
2016-08-23 10:40:34 -07:00 |
|
CuDnnRNN.h
|
(made gcc happy)
|
2016-08-24 19:51:02 -07:00 |
|
DataTransferer.cpp
|
First version of GPU prefetch
|
2016-09-21 12:00:04 +02:00 |
|
DataTransferer.h
|
First version of GPU prefetch
|
2016-09-21 12:00:04 +02:00 |
|
GPUDataTransferer.cpp
|
Enable better logging in GPU transferer
|
2017-01-10 10:49:40 +01:00 |
|
GPUDataTransferer.h
|
Remove statics from prefetch transferer
|
2016-12-07 11:57:48 +01:00 |
|
GPUMatrix.cu
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
GPUMatrix.h
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
GPUMatrixCUDAKernels.cuh
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
GPURNGHandle.cu
|
Use 8 bytes for RNG(Handle) seed on both platforms.
|
2016-12-13 12:37:18 +01:00 |
|
GPURNGHandle.h
|
Use 8 bytes for RNG(Handle) seed on both platforms.
|
2016-12-13 12:37:18 +01:00 |
|
GPUSparseMatrix.cu
|
Fix broken sparse CSC in InnerProductOfMatrices
|
2017-01-19 11:54:13 -08:00 |
|
GPUSparseMatrix.h
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
GPUTensor.cu
|
update to fix window build error
|
2017-01-10 10:15:55 -08:00 |
|
GPUTensor.h
|
merged from master
|
2016-07-26 16:51:19 -07:00 |
|
GPUWatcher.cu
|
normalized inconsistent commenting style from "//text" to "// text"
|
2016-01-22 13:58:47 -08:00 |
|
GPUWatcher.h
|
License change
|
2016-01-18 09:36:17 +01:00 |
|
Helpers.h
|
Inlining checks
|
2016-06-21 17:38:54 +02:00 |
|
Math.vcxproj
|
Windows: switch to Visual Studio 2015
|
2017-01-10 10:47:35 +01:00 |
|
Math.vcxproj.filters
|
Quantized multiplier v1
|
2016-11-27 21:25:26 -08:00 |
|
MathCUDA.vcxproj
|
Windows: switch to Visual Studio 2015
|
2017-01-10 10:47:35 +01:00 |
|
MathCUDA.vcxproj.filters
|
merge
|
2016-10-12 13:25:17 +02:00 |
|
Matrix.cpp
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
Matrix.h
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
MatrixQuantizerCPU.cpp
|
tuned >1-bit SGD: odd #quantization levels, range now 4 stddevs (before: 5)
|
2016-08-21 01:33:01 -07:00 |
|
MatrixQuantizerCPU.h
|
White-space (remove CRs that crept in)
|
2016-01-25 10:24:10 +01:00 |
|
MatrixQuantizerGPU.cu
|
First version of distributed aggregator for v2
|
2016-10-19 13:49:10 +02:00 |
|
MatrixQuantizerGPU.h
|
normalized inconsistent commenting style from "//text" to "// text"
|
2016-01-22 13:58:47 -08:00 |
|
MatrixQuantizerImpl.cpp
|
Re-format code using clang-format (plus some post-processing)
|
2016-01-18 09:36:14 +01:00 |
|
MatrixQuantizerImpl.h
|
First version of all reduce aggregator
|
2016-10-24 19:25:36 +02:00 |
|
MatrixQuantizer_kernel.cu
|
normalized inconsistent commenting style from "//text" to "// text"
|
2016-01-22 13:58:47 -08:00 |
|
MemAllocator.h
|
Re-format code using clang-format (plus some post-processing)
|
2016-01-18 09:36:14 +01:00 |
|
NcclComm.cpp
|
Source/Math/NcclComm.cpp: tune logging
|
2016-11-09 14:00:53 +00:00 |
|
NcclComm.h
|
Added NCCL support to SimpleDistGradAggregator
|
2016-11-08 18:15:01 -08:00 |
|
NoGPU.cpp
|
Add unit-gain momentum flag
|
2017-01-16 09:39:39 +01:00 |
|
QuantizedMatrix.cpp
|
addressing CR comments.
|
2016-11-23 23:12:32 +01:00 |
|
QuantizedMatrix.h
|
First version of all reduce aggregator
|
2016-10-24 19:25:36 +02:00 |
|
QuantizedOperations.h
|
Quantized multiplier v1
|
2016-11-27 21:25:26 -08:00 |
|
Quantizers.h
|
Quantized multiplier v1
|
2016-11-27 21:25:26 -08:00 |
|
RNGHandle.cpp
|
Use 8 bytes for RNG(Handle) seed on both platforms.
|
2016-12-13 12:37:18 +01:00 |
|
RNGHandle.h
|
Use 8 bytes for RNG(Handle) seed on both platforms.
|
2016-12-13 12:37:18 +01:00 |
|
RNNCommon.h
|
Incorporated CR feedback
|
2016-11-17 17:14:34 -08:00 |
|
TensorOps.h
|
ReduceLogSum: beautifications
|
2016-08-05 14:08:54 +02:00 |
|
TensorView.cpp
|
Quantized multiplier v1
|
2016-11-27 21:25:26 -08:00 |
|
TensorView.h
|
Quantized multiplier v1
|
2016-11-27 21:25:26 -08:00 |
|
ValueQuantizer.h
|
fix more warning descriptions
|
2016-11-23 12:01:31 +01:00 |
|
cudabasetypes.h
|
Re-format code using clang-format (plus some post-processing)
|
2016-01-18 09:36:14 +01:00 |
|
cudadevice.h
|
Removing unused dependencies
|
2016-06-21 17:38:54 +02:00 |
|
cudalattice.cpp
|
cuda related vs2015 changes
|
2016-11-29 11:57:01 +01:00 |
|
cudalattice.h
|
Re-format code using clang-format (plus some post-processing)
|
2016-01-18 09:36:14 +01:00 |
|
cudalatticeops.cu
|
Re-format code using clang-format (plus some post-processing)
|
2016-01-18 09:36:14 +01:00 |
|
cudalatticeops.cu.h
|
cuda related vs2015 changes
|
2016-11-29 11:57:01 +01:00 |
|
cudalatticeops.h
|
make parameter explicit by adding namespace
|
2016-11-23 10:39:35 +01:00 |
|
cudalib.cpp
|
normalized inconsistent commenting style from "//text" to "// text"
|
2016-01-22 13:58:47 -08:00 |
|
cudalib.h
|
Re-format code using clang-format (plus some post-processing)
|
2016-01-18 09:36:14 +01:00 |
|
dllmain.cpp
|
License change
|
2016-01-18 09:36:17 +01:00 |
|
fast_divmod.h
|
Initial divmod implementation
|
2017-01-04 12:23:08 -08:00 |
|
latticefunctionskernels.h
|
reviewed uses of atomicAdd(),;
|
2016-05-05 11:02:21 -07:00 |
|
stdafx.cpp
|
License change
|
2016-01-18 09:36:17 +01:00 |
|
stdafx.h
|
renamed EsotericNodes.h to SpecialPurposeNodes.h, and CompositeComputationNodes.h to PreComputeNodes.h;
|
2016-01-22 10:45:53 -08:00 |
|
targetver.h
|
License change
|
2016-01-18 09:36:17 +01:00 |
e6c23fd7da