sort by:
Revision Author Date Message Commit Date
e939648 disable fp16 test for poolWithSequenceAxis (#3810) 31 March 2020, 15:55 UTC
e1467a7 Merge pull request #3806 from microsoft/liqun/wrap_maxlpool_with_reshape support Pooling ops with Sequence axis 27 March 2020, 21:42 UTC
a2055f6 fix a typo 27 March 2020, 21:40 UTC
80e9f79 update with reviewers' comments 27 March 2020, 21:35 UTC
9a7dd4c add comments 25 March 2020, 19:37 UTC
280ec14 support Pooling ops with Sequence axis 25 March 2020, 18:20 UTC
c7bc93f Fix Windows Test Dockerfile 06 August 2019, 15:30 UTC
9a4bbfe Update Windows Test Dockerfile According to https://techcommunity.microsoft.com/t5/Containers/Windows-Server-2019-Now-Available/ba-p/382430, 'latest' tag is no longer supported. Using '1607' (windows server 2016) 31 July 2019, 17:49 UTC
bda8fcb address CR 23 July 2019, 00:24 UTC
c3d7440 add FutureValue Op's custom attributes into Scan node's description 22 July 2019, 23:48 UTC
91b78a7 Merge branch 'yanchen/load-from-memory' Signed-off-by: Yang Chen <yanchen@microsoft.com> 20 July 2019, 01:19 UTC
f843f91 Merge branch 'yanchen/rm-asserts' Signed-off-by: Yang Chen <yanchen@microsoft.com> 19 July 2019, 02:23 UTC
3ea8507 Merge pull request #3722 from microsoft/yanchen/output-mapping added a mapping from onnx outputs to cntk outputs into the graph's description 18 July 2019, 22:11 UTC
d01ef8f remove an assert that seems to be too conservative 18 July 2019, 17:22 UTC
cfe8f6c Added a new API to load models from memory * added a new C api CNTK_LoadModel_FromArray that loads a model from memory * also enabled loading ONNX models from buffer 18 July 2019, 07:57 UTC
f7f1e8d added a mapping from onnx outputs to cntk outputs into the graph's description 18 July 2019, 07:39 UTC
99a272b Fix hard tabs 05 July 2019, 23:18 UTC
a4621a8 Export custom attributes on Times node to ONNX file (#3709) 05 July 2019, 20:33 UTC
13f3b5f Merge branch 'master' into liqun/scan14 27 June 2019, 17:26 UTC
abf0985 transpose batch with dir before reshape to [seq, batch, dir*hidden] for RNN ops 27 June 2019, 17:25 UTC
ac7a549 Fix V2LibraryCSTests end to end test 26 June 2019, 06:20 UTC
9756a11 Fix V2LibraryCSTests end to end test 25 June 2019, 19:16 UTC
ef67317 set shape bacth dim to 0 for reshape op after LSTM to make batch size editable after ONNX export. 21 June 2019, 21:50 UTC
03a8c41 default keep dim is true 19 June 2019, 16:24 UTC
1fe0bf1 reduction keep dim edge cases 19 June 2019, 16:17 UTC
ad1709c fix reduce op keepdim setting 18 June 2019, 22:16 UTC
248ff78 onnx exporter for past/future value: fix for offset > sequence length cases; output delayed initial values 11 May 2019, 00:03 UTC
8ed7050 ONNX convolution with more than 1 non-feature/channel dimensions 01 May 2019, 21:06 UTC
22a86bf onnx export: fix cases when scan input needs to be broadcasted 29 April 2019, 18:58 UTC
0e172db Fix build 24 April 2019, 18:20 UTC
82b7a39 merge release 2.7 19 April 2019, 18:28 UTC
6b72436 Adds CNTK libs folder to LD_LIBRARY_PATH at CNTK init routine on Python 09 April 2019, 19:54 UTC
b8edaab Rpengms/fp16 bs bugfix (#3632) Update CNTK.core.bs to add support for precision overwrite for every node 09 April 2019, 16:25 UTC
9688ac6 fp16 - brain script add to CNTK.core.bs (#3617) 01 April 2019, 17:14 UTC
764c8c4 Fix CNTK Tutorial 303 25 March 2019, 18:08 UTC
49d4c2d Update OOBE GPU Dockerfile to CUDA 10 25 March 2019, 04:39 UTC
f48e11c Update conda to >= 4.3.30 22 March 2019, 22:12 UTC
3b7e3b9 merge fp16_bs fix (#3609) fp16 brain script fix accumulator 22 March 2019, 09:07 UTC
4003c08 Merge fp16 brainscript work (#3606) * FP16 BrainScript - address code review comments * Remove Tab and fix debug build breaks * Fix Linux Build breaks * fp16 brain script - add _CRT_SECURE_NO_WARNINGS * fp16 brain script - fix NetworkTests * Update tests for model version change * Remove changes for InputAndParamNodes * Fix typo * Remove redundant code * Fix optional parameters 20 March 2019, 18:36 UTC
45ae386 disable Manual/Tutorial tests on Windows Python2.7 17 March 2019, 21:36 UTC
9f60d8f Remove tab which caused build breaks 15 March 2019, 17:47 UTC
19ccd9f Address code review comments 15 March 2019, 17:47 UTC
3a21b0a FP16 all reduce. Cast gradients to and from float16 before sending over NCCL to do all reduce. 15 March 2019, 17:47 UTC
4765f69 support onnx export of 2GB models 14 March 2019, 16:16 UTC
22ded46 Address code review comments 28 February 2019, 22:43 UTC
d46ad9d Update 28 February 2019, 22:43 UTC
773379e Add more comments 28 February 2019, 22:43 UTC
d2dc123 Fix BMUF fp16 to update local learner's fp32 cache 28 February 2019, 22:43 UTC
26e26e4 Fix BMUF fp16 bug. Local learner ResetSmoothedGradients should NOT wipe cached fp32 model/parameters. 28 February 2019, 22:43 UTC
e104129 Remove obsolete comments 28 February 2019, 18:09 UTC
92054e8 Remove debug messages 28 February 2019, 18:09 UTC
730a220 Add useV2Aggregator to SGD config 28 February 2019, 18:09 UTC
6ef0224 Merge branch 'pull/3431' Signed-off-by: Thiago Crepaldi <thiagofc@microsoft.com> 28 February 2019, 09:45 UTC
39ca0a2 Merge branch 'pull/3482' Signed-off-by: Thiago Crepaldi <thiagofc@microsoft.com> 28 February 2019, 09:04 UTC
93f875c Merge branch 'pull/3477' Signed-off-by: Thiago Crepaldi <thiagofc@microsoft.com> 28 February 2019, 06:49 UTC
8cbfbfe Merge branch 'pull/3458' Signed-off-by: Thiago Crepaldi <thiagofc@microsoft.com> 28 February 2019, 04:38 UTC
ef5b745 Merge branch 'pull/3455' Signed-off-by: Thiago Crepaldi <thiagofc@microsoft.com> 28 February 2019, 04:27 UTC
1c5f124 Merge branch 'pull/3486' Signed-off-by: Thiago Crepaldi <thiagofc@microsoft.com> 28 February 2019, 02:24 UTC
726e48e Merge branch 'pull/3491' Signed-off-by: Thiago Crepaldi <thiagofc@microsoft.com> 28 February 2019, 00:43 UTC
771d754 Merge branch 'pull/3492' Signed-off-by: Thiago Crepaldi <thiagofc@microsoft.com> 27 February 2019, 23:59 UTC
5ccf54a Merge branch 'pull/3514' Signed-off-by: Thiago Crepaldi <thiagofc@microsoft.com> 27 February 2019, 22:39 UTC
af42873 Merge branch 'pull/3561' Signed-off-by: Thiago Crepaldi <thiagofc@microsoft.com> 26 February 2019, 21:07 UTC
06469f7 Merge branch 'pull/3558' 25 February 2019, 18:46 UTC
a5fb3d6 add more onnx rnn model tests, enable optimizedRNN op test 19 February 2019, 20:48 UTC
00b8037 Update onnx submodule, support ConstantOfShape & OneHot * Support SequenceIs[First,Last] with ConstantOfShape * Update bypass load test in verify\_one\_input & add test for one hot op * Update export for one hot op. Migrate from exporting onnx.ml.OneHotEncoder to onnx.OneHot. Op fixes * Fix topk onnx\_op\_test * Support MVN export using ONNX function * Fix LayerNormalization * Skip tests for sequence slice float16: not supported in cntk * Support gather export & import with float16 - Fix cntk gather issue with float16 inputs. - Support exporting constant float16 tensor. - Support importing int32 indices input for gather. * Enable more passed op tests 15 February 2019, 23:12 UTC
aa82819 Merge branch 'liqun/remove-crosstalkcaffe-test' 14 February 2019, 21:07 UTC
3bef0f4 protobuf 3.6.0 to 3.6.1 13 February 2019, 19:08 UTC
9ce6c1c Merge pull request #3550 from Microsoft/yaronwe/hide-protobuf hide protobuf library symbols 08 February 2019, 16:37 UTC
2ccdff4 enable onnx rnn test, name input with proper name 04 February 2019, 05:18 UTC
979c8bf Merge branch 'liqun/liqun/scan6Stage2' 31 January 2019, 04:27 UTC
9093a4d Cache CNTK dependencies at CNTK Azure Blob 30 January 2019, 03:47 UTC
6b2323c onnx support of sequential convolution 28 January 2019, 22:48 UTC
631ef0f Fix typo Ture -> True 28 January 2019, 15:02 UTC
4ef4417 Adding sequence models to CI and skipping failing ones 24 January 2019, 23:01 UTC
fc839bc Moving CI test dependencies to python environments yml files 23 January 2019, 00:24 UTC
d67a5ae Free nnzPerRowOrCol at end of SetValue 16 January 2019, 22:00 UTC
fea6e9d Merge branch 'vadimma/bmlfex' 12 January 2019, 00:01 UTC
62fa818 Enable frame mode for the binary MLF reader 11 January 2019, 02:02 UTC
0be4bcb Merge branch 'liqun/scan5Stage' 09 January 2019, 02:39 UTC
fe939aa support ONNX scan(9) 08 January 2019, 20:16 UTC
a5fcc80 Resolve Linux binary drop & windows oobe test conda dependency 07 January 2019, 23:07 UTC
2274e02 hide protobuf library symbols 06 January 2019, 10:04 UTC
b732aeb Fix several DeprecationWarning: invalid escape sequence Signed-off-by: Mickaël Schoentgen <contact@tiger-222.fr> 05 January 2019, 21:36 UTC
322293c Merge branch 'vadimma/adambs' 04 January 2019, 00:21 UTC
6c56fdd Revert "refactor" This reverts commit 47345c6c54e95ed4a4948d5af9f1205f6e7c3932. 03 January 2019, 23:45 UTC
47345c6 refactor 03 January 2019, 23:26 UTC
05435d0 Expose Adam in BS 03 January 2019, 20:58 UTC
08130db update linux binary drop cudnn path 03 January 2019, 18:57 UTC
e2d79d7 Submodule onnxruntime, and remove previous drop. * A few patches are required to build cntk_uwp. * Use proto from onnxruntime/protobuf instead of from onnx. * TODO: Some issues with onnx_op_test RNN and OptimizedRNNStack from shape inference. 03 January 2019, 01:09 UTC
254a336 Bump cuda version in nuget spec 21 December 2018, 18:40 UTC
bc70863 Add Maxunpooling export 14 December 2018, 19:24 UTC
d087acc Fix typo & update .yaml files * pyqt is required by matplotlib and various other libraries. In our .yaml files we default to install the latest version of pyqt. However the current latest version (5.9.2) is exhibiting some problems. Thus we downgrade to the stable version 5.6.0. 13 December 2018, 18:37 UTC
ef2f039 fix break for cub 1.8 API change 13 December 2018, 00:14 UTC
f178144 Support CUDA 10 * Move to support CUDA 10, cudnn 7.3, cub 1.8. * Fixed a bug related to "pointer to pin pointer is disallowed" #3063, which is exposed in newer version vctools. * Added workaround for a potential vs2017 15.9 bug with cntk Debug version. 13 December 2018, 00:10 UTC
93e1009 comment debug code 12 December 2018, 03:17 UTC
de15bda passed prod models and unfold test. handle delay op. 12 December 2018, 01:49 UTC
22e869e Add ONNX support for zeros_like, ones_like, and eye_like. 05 December 2018, 21:51 UTC
5fc6c2a sequence_isfirst_islast test fix for python 2.7 30 November 2018, 19:22 UTC
75b0141 Adding support for exporting CNTK's Sequence::IsFirst and Sequence::IsLast nodes 29 November 2018, 23:57 UTC
d1d1133 ConvTranspose asymmetric padding * Temporarily reverse the extra padding location in case of SAME_UPPER vs SAME_LOWER for convTranspose to match with onnxruntime. * In case of importing asymmetric padding convTranspose, use symmetric pads by alter the output_shape and pads, and attach a slice node afterwards to enable cudnn. * Fix a bug in slice/squeeze attribute axes export. 27 November 2018, 17:33 UTC
back to top