https://github.com/Microsoft/CNTK

sort by:
Revision Author Date Message Commit Date
d33b7b4 update iteration plan 13 September 2018, 00:24:45 UTC
6a4ec05 Merge branch 'peykash/batchnorm_and_pooling_fixes' 13 September 2018, 00:13:46 UTC
8b48976 Adding CNTK 2.6 release work summary to current_iteration.md 12 September 2018, 18:12:10 UTC
b374e14 fixes on Batchnorm and Pooling for v1 pretrained models after removal of sequence axis from input 12 September 2018, 17:02:52 UTC
5897265 small patch on conv/pooling export - when pads are all zero, check if autopad is true. - when pads are all zero, check if ceilOutDim is true, and extra cells are needed. 11 September 2018, 15:52:22 UTC
61572e8 Update onnx_model_test skip list 11 September 2018, 15:51:45 UTC
0754b38 update onnx_model_test with tests from onnx backend test 09 September 2018, 20:53:26 UTC
2f52f22 update conv/convtranspose/pooling import. pad values are explicitly computed based on ONNX spec equations during import in the following cases: - case 1: when auto_pad is SAME_UPPER | SAME_LOWER for convolution, convolution transpose and pooling. - case 2: when output_shape is explicitly set for convolution transpose. note: output_shape in ONNX spec can have the two below format: 1. [X1 * X2 * ... * Xn] 2. [N * O * X1 * X2 * ... * Xn] 09 September 2018, 20:41:04 UTC
d877233 Make broadcast ops compitable between CNTK and ONNX, Enable ONNX export/import for optimizedRNN op, More ONNX support for Sequence ops 09 September 2018, 15:59:33 UTC
fcf9f48 Overhaul conv/convTrans/pooling pads value export - Update exporting of conv/pooling to always export pad values. - Enable correct exporting of multiple pretrained models (ResNet50/ResNet101/ResNet152_ImageNet_Caffe, etc). - Overhaul convtranspose pads exporting - Support conv weight export with omitted out channel axis (LRN). - Add tests in onnx_op_test to cover the above changes 06 September 2018, 18:46:14 UTC
e3a1acf Resolve dependencies and build issues -Temporary add importorskip around import onnx -bump up .yml matplotlib version 05 September 2018, 22:02:23 UTC
dc5e482 fix onnx average pooling export. - this fix solves the issue that ceilOutDim == true will enforce exporting auto_pad as true, even if autoPadding is explicitly set to false. 31 August 2018, 01:19:04 UTC
77a8c49 Temporarily skip onnx_model_test if import onnx fail 30 August 2018, 17:52:53 UTC
94a43ed Merge branch 'liqun/liqun/RNN2.6.Stage' 30 August 2018, 07:40:20 UTC
73cd53e fix nightly issues related to onnx dependencies - Windows OOBE (pip) tests & Linus OOBE tests: skip onnx_model_test. This test requires onnx to be installed. Skip Until we decide to add onnx dependencies to OOBE test environment. 30 August 2018, 00:19:19 UTC
18d9f39 skip dynamic axes wrapper, export onnx test cases, handle output op being Combine op, workaround a specific case bug of ONNX bidirectional broadcast shape inference 29 August 2018, 23:52:58 UTC
902f1a4 times export fix 29 August 2018, 17:09:07 UTC
b86fe1a Merge branch 'pull/3374' 28 August 2018, 07:17:54 UTC
1f2e42e Merge branch 'sptiwari/convtranspose_update7' 27 August 2018, 07:32:33 UTC
edc29f8 Packaging newly-added internal header files Recently, we added a couple of new header files into API/Internals. This patch includes them into our pre-built binaries. 27 August 2018, 04:36:04 UTC
b3c0fa2 Overhaul ConvTranpose to match ONNX 1.2.2. spec. 27 August 2018, 04:20:05 UTC
7dd9638 squash of the following changes: - fix flatten onnx export. - fix unsqueeze onnx export. - add comments on temporarily skipped tests. - adjust the importing of softmax, logsoftmax and hardmax with blockfunction - such that they could be exported as is back to onnx. - update reshape onnx export to pass mobilenet round trip test. 26 August 2018, 20:11:44 UTC
0e20836 CNTK splice allows broadcast. This case is handled in the change. For noop (identity) ops, its inputs and outputs types shall be set according to upstream ops. ToBatch/ToSequence and Unpack Batch/Sequence ops added during model importing need tp be skipped. Model import need to handle ops with multiple outputs 26 August 2018, 15:41:20 UTC
4a6238d reduction all axes export fix 24 August 2018, 17:02:34 UTC
d2ff412 temporarily disable 2 tests on Windows. - This is due to an issue on Windows CI introduced by adding onnx dependencies. These tests are temporarily disabled to not block CI while we investigate. - Disable CNTKv2Python/Tutorial/205 - Disable CNTKv2Python/Keras 24 August 2018, 08:27:45 UTC
88d8882 Fix issue#3228 and #3363 which found when build with MSVC+ permissive- 24 August 2018, 05:56:36 UTC
a26e542 update conftest.py to resolve doctest issue. - newer version numpy has a different print format for arrays and scalars that would potentially break the doctests. 24 August 2018, 04:28:06 UTC
28ada96 update python doctest handling newer version numpy print format. 23 August 2018, 06:21:13 UTC
56ef694 add onnx_model_test.py 23 August 2018, 06:21:13 UTC
c0ff1da fix gemm, pooling export to onnx. 23 August 2018, 06:19:18 UTC
4244320 Adding support for exporting CNTK TimesTranspose 22 August 2018, 18:33:55 UTC
3d809bf Added several internal API header files In case other projects may use these header files, we added them into API/Internals. * ComputationGraphAlgorithms.h was moved from Source/ComputationNetworkLib * PrimitiveOpType.h and EvaluatorWrapper.h were moved from Source/CNTKv2Library * PrimitiveFunctionAttribute.h was extracted from PrimitiveFunction.h. It contains a new class PrimitiveFunctionAttribute which is the collection of all attribute names for PrimitiveFunction. This change actually had a subtle side-effect. We had a global static variable s_stateAttributes that depended on PrimitiveFunction::AttributeNameRngSeed and PrimitiveFunction::AttributeNameRngOffset. After we moved those static attribute-variables into another translation unit, s_stateAttributes can be initialized with empty wstring, because PrimitiveFunctionAttribute::AttributeNameRngSeed PrimitiveFunctionAttribute::AttributeNameRngSeedOffset were initialized after s_stateAttributes. Note that the initialization order of global static variables is not well-defined cross translation units. To fix the issue, we also moved s_stateAttributes into PrimitiveFunctionAttribute class, and renamed it to s_rngStateAttributes. I think it's reasonable to consider s_rngStateAttributes to be part of the PrimitiveFunctionAttribute class. * PrimitiveFunction.h was moved from Source/CNTKv2Library 22 August 2018, 17:47:18 UTC
8a244e6 fix onnx_op_test.py reduce ops. - remove bracket around data to match latest verification. 21 August 2018, 21:23:28 UTC
b36347b Fix legacy test 21 August 2018, 19:38:26 UTC
db41aeb Merge branch 'bowbao/onnx_reduce_allaxes' 21 August 2018, 17:05:14 UTC
4bbb245 Merge branch 'liqun/TopKStage' 21 August 2018, 03:47:21 UTC
1706da2 Inject TEST_TAG env var into test containers 20 August 2018, 23:57:38 UTC
0e57d20 fix reduce ops for onnx export/import with all_static_axes() and all_axes(). 20 August 2018, 21:48:42 UTC
ee2fa5e Correctly calling tolower or towlower 20 August 2018, 21:32:08 UTC
3328f82 update onnx_backend_test.py skipped models and ops. 20 August 2018, 18:57:52 UTC
3a35b6f pedantic option cannot be ignored -pedantic options is not an option that can be ignored. GCC 6.x and above issued warnings (and hence errors) while building CNTK: error: pedantic is not an option that controls warnings. 17 August 2018, 22:09:40 UTC
1a572f8 ONNX support of TopK 17 August 2018, 20:01:47 UTC
9040c9d Build c# test dlls in sequence 17 August 2018, 18:32:42 UTC
4b2b0d3 Merge branch 'peykash/batchnorm_float16_fixes' 17 August 2018, 09:54:13 UTC
d67eba8 Remove template definition 17 August 2018, 04:25:39 UTC
af7e33e Fix template compilation 17 August 2018, 04:25:39 UTC
8c83adb Some cleanup 17 August 2018, 04:25:39 UTC
648c99d Refactor common methods 17 August 2018, 04:25:39 UTC
086dd28 Update .gitattributes to mark MLF as binary file 17 August 2018, 04:25:39 UTC
cb3d08a Add E2E test for binary MLF 17 August 2018, 04:25:38 UTC
7c59a63 Further binaryMLF refactoring 17 August 2018, 04:25:38 UTC
931193d Refactor binary MLF 17 August 2018, 04:25:38 UTC
7816541 Remove a logging statement 17 August 2018, 04:25:38 UTC
e500b6e Fix binary stream reading 17 August 2018, 04:25:37 UTC
0ef680e add binary MLF reader to makefile 17 August 2018, 04:25:37 UTC
cd838ae Introduce model version 2 17 August 2018, 04:25:37 UTC
9ad1910 Optimize senone count lookup 17 August 2018, 04:25:37 UTC
ce95c1c First working version 17 August 2018, 04:25:36 UTC
0a00f8b Implement MLF binary parser 17 August 2018, 04:25:36 UTC
0198fad Binary index builder' 17 August 2018, 04:25:36 UTC
d531345 binary mlf index' 17 August 2018, 04:25:36 UTC
222efb8 Binary mlf reader 17 August 2018, 04:25:35 UTC
f5b481a copy HalfConverter.hpp for make_binary_drop HalfConverter.hpp is included from CNTKLibrary.h. Without including it, we cannot use binary drops. 16 August 2018, 20:05:34 UTC
4906208 Merge branch 'bowbao/onnx_imageScaler' 16 August 2018, 01:01:32 UTC
bf2b5ea Merge branch 'liqun/Py2FixTupleUnpacking' 16 August 2018, 00:12:53 UTC
c8cecaa Dump exception messages to stderr in Release mode It's useful to have excpetion messages dumped to std in Release mode. 15 August 2018, 21:32:54 UTC
bff3002 fix python 2.7 tuple unpacking error 15 August 2018, 20:47:56 UTC
4b4b9cd fix imageScaler export/import to onnx 15 August 2018, 17:22:47 UTC
a542186 BatchNormalization Float16 Fixes 15 August 2018, 17:01:23 UTC
ef836bc Merge branch 'liqun/NewSequenceSliceStage2Stage' 15 August 2018, 06:46:32 UTC
6547e2c Fix bug in import of Conv op in ONNX. 14 August 2018, 23:00:38 UTC
ae163b3 ONNX support for CNTK Sequence::Slice op. Due to a specific feature of CNTK Sequence::Slice op to handle zigged inputs (batch of variable length sequences), models converted from CNTK may not produce marching outcomes with zigged inputs. 14 August 2018, 18:17:36 UTC
ceaec56 fix squeezenet onnx_backend_test of issues with Softmax, hardmax, logsoftmax and flatten with freedimension on batchaxis. 14 August 2018, 17:02:07 UTC
541e210 Merge branch 'master' into sptiwari/model_fix_test 13 August 2018, 17:48:20 UTC
b689ebb Ignore TEST_TAG environment variable when not defined during testing 13 August 2018, 16:02:01 UTC
4ce6607 Disable fmeasure test and increase pytest verbosity 12 August 2018, 20:24:20 UTC
b73670c Removing support for user-provided node names and using uid for ONNX names. 10 August 2018, 23:53:58 UTC
e0b2656 Fixing batch axis handling bug in ReduceElements and Reshape export. 10 August 2018, 23:36:29 UTC
ff118cc Merge branch 'master' into sptiwari/onnx_merge_with_master 10 August 2018, 17:10:06 UTC
b7811d3 Fmeasure (#3343) * amended a bug, swapped precision and recall * added a bit more documentation 10 August 2018, 15:50:42 UTC
2a748e1 Merge branch 'vadimma/lc-blstm-sq' 09 August 2018, 02:28:38 UTC
a87d69a Fix Hardmax/Softmax/LogSoftmax ONNX export. 08 August 2018, 23:52:33 UTC
1ebdc54 Introduce latency-controlled BLSTM 08 August 2018, 22:41:51 UTC
e5006a4 fix parameter sharing bug 06 August 2018, 10:00:33 UTC
d2b4954 Add more logging into HTKDeserializer 03 August 2018, 19:44:51 UTC
5e871a1 treat sequence axis as static axis for ONNX RNN model import 02 August 2018, 17:51:05 UTC
ce7075b Updating input image size requirement to include lower and upper paddingin ConvolveGeometry. 01 August 2018, 19:00:54 UTC
917173d Merge branch 'pull/3331' 01 August 2018, 05:50:14 UTC
6f7ceed Merge branch 'pull/3284' 01 August 2018, 04:48:10 UTC
360fbd4 Update the check in convolution if kernel fill the entire input space. 01 August 2018, 01:19:45 UTC
279a466 Squash of the following 1. fix gather and prelu arguments order issue. 2. add typenameToTypeProto map initialization. 3. add CNTK native default lotus logger. 4. add CNTK unsqueeze op. 31 July 2018, 02:58:55 UTC
c78f40d Avoid asym pad for MKL, block channel axis padding. 31 July 2018, 02:50:35 UTC
dc65c2b Fix overlapping memcpy reported by valgrind in SmallVector Valgrind sometimes reports "Source and destination overlap in memcpy" on the overloaded = operator of SmallVector class. Whenever it is reported, the src address and the dest address of memcpy is actually identical. The solution in this commit is to check the src and dest address first, and do memcpy only if they are not identical. This should be enough to handle the issue, since the only case they overlap is when they are identical, provided that the m_data is a member of a class. 31 July 2018, 01:03:17 UTC
3d753db Enabled set random seed function and force deterministic api in c# 30 July 2018, 23:06:34 UTC
6ed9659 partially support matmul broadcast. 30 July 2018, 20:29:58 UTC
2964f1f batchnorm_without_batch_axis_fixes 30 July 2018, 19:42:35 UTC
ffc9e14 Merge branch 'sedymche/onnx-reshape' into cntkteam/onnx_without_batch_axis 28 July 2018, 01:43:42 UTC
56adb88 Merge branch 'sedymche/onnx-maxes' into cntkteam/onnx_without_batch_axis 28 July 2018, 00:59:58 UTC
e98b00d Adding support for StopGradient as a no-op. 27 July 2018, 23:21:46 UTC
b7f7b53 Update Flatten/Reshape ONNX conversions for Softmax/Hardmax/LogSoftmax. 27 July 2018, 21:43:52 UTC
back to top