refs/heads/bowbao/onnx_matmul_gemm | 8e3431e | update documentation for arguments order change | 20 July 2018, 00:57:07 UTC |
refs/heads/bowbao/onnx_models_unittest | c039e45 | update precision to make mobilenetv2-1.0 pass | 17 August 2018, 21:40:43 UTC |
refs/heads/bowbao/onnx_slice | d58eb18 | fix Slice for onnx_backend_test | 20 August 2018, 18:55:09 UTC |
refs/heads/bowbao/onnxrt | 04088ae | nit | 02 January 2019, 22:53:09 UTC |
refs/heads/bowbao/onnxrt_stage | 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. | 14 December 2018, 18:16:57 UTC |
refs/heads/bowbao/onnxruntime_ci | 1b36e72 | preserve post onnxruntime-ci work | 26 October 2018, 00:50:19 UTC |
refs/heads/bowbao/onnxruntime_ci_stage | f6135ae | skip currently known issues | 22 October 2018, 20:10:44 UTC |
refs/heads/bowbao/overhaul_conv_export | b31b87d | update onnx_model_test with tests from onnx backend test | 07 September 2018, 00:33:58 UTC |
refs/heads/bowbao/preserve_work | bd40714 | add Maxunpooling export | 19 November 2018, 18:31:44 UTC |
refs/heads/bowbao/protobuf_version | d7656d4 | Print protobuf version before CNTKv2Python/ModuleTests | 11 September 2018, 21:43:51 UTC |
refs/heads/bowbao/seqconv | 476a60c | small tweak in seq conv to avoid additional gpu memory allocation and increase performance. | 02 July 2018, 17:47:37 UTC |
refs/heads/bowbao/seqconv_perf | 7ca2590 | adding profiling for seq conv v2 | 02 July 2018, 07:26:17 UTC |
refs/heads/bowbao/seqconv_profiling | 615a046 | Example: seq MNIST, and profiling | 28 June 2018, 01:40:27 UTC |
refs/heads/bowbao/seqconv_v2 | e237742 | Moving sequential convolution in python to a new high level api, to maintain compatibility with previous implementation (special case 1d sequential convolution). | 05 July 2018, 19:49:49 UTC |
refs/heads/bowbao/seqconv_v2_stage | 9a58f86 | nit | 11 July 2018, 00:50:00 UTC |
refs/heads/bowbao/update_onnx | 7ac3bf6 | Update export for one hot op. * Migrate from exporting onnx.ml.OneHotEncoder to onnx.OneHot. | 18 January 2019, 00:52:31 UTC |
refs/heads/bowbao/update_onnx_stage | 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 | 16 January 2019, 23:21:08 UTC |
refs/heads/bowbao/update_onnx_stage_cr | 66cd50f | 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. | 28 January 2019, 23:30:57 UTC |
refs/heads/bowbao/update_onnx_stage_cr_v14 | 8d2702d | enable more passing tests | 14 February 2019, 22:42:06 UTC |
refs/heads/broda/tolower | d567a8d | Correctly calling tolower or towlower | 17 August 2018, 16:05:07 UTC |
refs/heads/cazhan/lu-seq-reader | 52331f5 | License change | 18 January 2016, 08:33:07 UTC |
refs/heads/chanliu/AMTrain | 574bc97 | fix syntax again | 15 September 2017, 23:49:16 UTC |
refs/heads/chanliu/TSLearn | 31dfe78 | start newline when printing skipping information | 26 April 2017, 21:49:50 UTC |
refs/heads/chaojunl/lfmmi | e37c7dd | add usePrior option | 29 March 2018, 00:25:44 UTC |
refs/heads/chaojunl/smbr | 5a9268e | swap label and gamma | 09 December 2019, 05:10:54 UTC |
refs/heads/chazhang/faster_rcnn | da4f699 | Update faster rcnn code to the version that successfully runs on Philly. Revise random flipping code so flipping is decided per image rather than per-epoch. Enable gpu nms. Philly run shows that it has no impact in accuracy. Remove a redundant line in the code. Bug fix and revise learning rate schedule. Use config to specify nms, and bug fix. | 17 November 2017, 21:10:04 UTC |
refs/heads/chazhang/pretrain | 9273888 | Create a new Pretrained model folder. | 14 August 2017, 15:46:35 UTC |
refs/heads/chenta/add_keras_learner | 774238c | add keras learner | 17 May 2017, 05:50:05 UTC |
refs/heads/chenta/align_axis_indices | 05396ba | align axis under keras mode | 15 May 2017, 19:51:37 UTC |
refs/heads/chenta/auto_broadcast_binary_op | 3415b9b | skip some sequence operator add ut | 18 October 2017, 21:53:41 UTC |
refs/heads/chenta/crash_investigate | 4d6cc0f | roll back test | 02 November 2017, 22:40:42 UTC |
refs/heads/chenta/disable_omp | 559eb94 | comment out omp code | 16 June 2017, 05:15:17 UTC |
refs/heads/chenta/distPerf | 216a07f | no omit frame pointer and other changes | 12 June 2017, 22:49:47 UTC |
refs/heads/chenta/expose_gradient_api | 20dd264 | remove hard tab | 12 April 2017, 20:21:37 UTC |
refs/heads/chenta/fix_conv_crash | accd9c7 | add unit test | 17 May 2017, 05:36:36 UTC |
refs/heads/chenta/fix_crash | 58c0308 | fix nightly build crash by call reconcile directly | 07 November 2017, 22:13:50 UTC |
refs/heads/chenta/fix_gather_bug | 00bd68a | add ut | 03 August 2017, 22:40:40 UTC |
refs/heads/chenta/fix_nightly_build_crash | 28032f1 | add test to check does the reconcile dynamic axes op skip the autobroadcast | 03 November 2017, 18:24:50 UTC |
refs/heads/chenta/fix_slice_free_dimension | 24027b5 | fix a bug when slice with free dimension use fliplr for old numpy version | 25 September 2017, 16:40:57 UTC |
refs/heads/chenta/free_dimension_pooling | f07a05d | remove check for free dimension on pooling add ut add test with strides input | 08 September 2017, 18:47:07 UTC |
refs/heads/chenta/freedimension_test | f42f85e | ignore shape check for free dimension in times | 14 August 2017, 21:04:54 UTC |
refs/heads/chenta/keras-final | 4967e82 | update interface name. update python api to access lambda with parameters and gradients | 26 May 2017, 23:17:01 UTC |
refs/heads/chenta/keras_learner | 601a648 | align axis test code | 18 May 2017, 23:53:19 UTC |
refs/heads/chenta/lotus_test | 963019c | temp fix | 31 October 2017, 04:30:25 UTC |
refs/heads/chenta/nightly_crash | 1ec402b | fix nightly build by call reconcile dynamic axis directly | 07 November 2017, 00:00:29 UTC |
refs/heads/chenta/padding | 69fac7a | fix cs binding break | 12 August 2017, 01:28:06 UTC |
refs/heads/chenta/print_omp_thread | 75ca9ae | add set thread back | 16 June 2017, 00:05:24 UTC |
refs/heads/chenta/profile_test | 8e44ff4 | temp profiling code | 17 April 2017, 22:13:48 UTC |
refs/heads/chenta/reverse_support | 6351dde | Merge branch 'chenta/reverse' into chenta/reverse_support | 20 July 2017, 04:20:12 UTC |
refs/heads/chenta/splice | a3188a4 | support splice with free dimension axis | 14 August 2017, 22:19:53 UTC |
refs/heads/chenta/test | 626f9fd | add node level log | 29 April 2017, 00:46:53 UTC |
refs/heads/chenta/times | 01148e7 | fix merge error | 21 August 2017, 17:05:16 UTC |
refs/heads/chenta/univeral-learner | 75453bb | add keras learner | 05 May 2017, 23:09:43 UTC |
refs/heads/chenta/unpack_batch_grad_fix | 7261048 | fix comments in CR | 01 September 2017, 20:50:30 UTC |
refs/heads/chenzhehuai/lfbmmi-negstream | b273a7b | proc nan in assign | 16 June 2017, 02:29:23 UTC |
refs/heads/chenzhehuai/lfbmmi-negstream-sil | e7afedf | to make jenkin extremely happy | 18 June 2017, 22:32:07 UTC |
refs/heads/clemensm/cntkmkl | 38f0218 | Switch off OpenMP for Math | 03 June 2016, 12:33:58 UTC |
refs/heads/clemensm/cntkmkl-evalx | 715a476 | Merge branch 'master' into clemensm/cntkmkl-evalx | 21 June 2016, 19:45:45 UTC |
refs/heads/clemensm/dynamicaxis | 2c2a2de | Merge branch 'clemensm/dynamicaxis' of https://github.com/Microsoft/CNTK into clemensm/dynamicaxis | 11 April 2016, 13:51:55 UTC |
refs/heads/clemensm/fixseqcla | 223a013 | Changed to sparse; Fixed parameter | 28 April 2016, 22:29:19 UTC |
refs/heads/clemensm/mblayouts | 6ace7b3 | Merge branch 'deepbliscore/vnext' into clemensm/mblayouts | 05 April 2016, 09:18:25 UTC |
refs/heads/clemensm/perf2 | 73b2371 | Review comments | 15 September 2016, 19:16:31 UTC |
refs/heads/clemensm/perf4 | e2e83eb | Fix Scatter for packed MBLayouts | 22 September 2016, 11:20:38 UTC |
refs/heads/clemensm/perf4b | 0ca523b | Renamed TemporaryMatrix to TemporaryMatrixPool and made it thread safe | 20 October 2016, 23:14:51 UTC |
refs/heads/clemensm/profiler | fc7e051 | Glitch | 20 July 2016, 21:32:17 UTC |
refs/heads/clemensm/quantmultiply2 | 0cb61d0 | Fix bug in extended eval with multiple inputs - Fix ordering of inputs in extended eval (it was mixing up the order returned from the network with the order [alphabetic] returned - Expose brainscript exceptions to .net, including stacktrace - Expose exceptions happening in StartForwardEvaluation rather than throwing SEHException - Some more error checking for forward pass - Enable retrieving brainscript errors rather than outputting to stderr. | 13 July 2016, 10:04:58 UTC |
refs/heads/clemensm/runtimeopt | e47fe13 | Correct AllocateAllMatrices | 08 July 2016, 08:14:25 UTC |
refs/heads/clemensm/s2s | 019b146 | adapted to franks mail | 25 April 2016, 08:51:33 UTC |
refs/heads/cntk_agg | 5df8651 | Increase error tolerance for BatchNormalization Spatial tests | 22 June 2018, 21:14:26 UTC |
refs/heads/cntkteam/new_onnx_without_batch_axis | a87d69a | Fix Hardmax/Softmax/LogSoftmax ONNX export. | 02 August 2018, 17:41:05 UTC |
refs/heads/cntkteam/onnx1.0 | b7811d3 | Fmeasure (#3343) * amended a bug, swapped precision and recall * added a bit more documentation | 10 August 2018, 15:50:42 UTC |
refs/heads/cntkteam/onnx1.2 | 7449613 | Revert "Temporarily skip several failing float16 ONNX tests." This reverts commit deef207a6201015d2b7dbc6649a4e35cb47cfe8a. | 19 June 2018, 17:20:00 UTC |
refs/heads/cntkteam/onnx_without_batch_axis | 88f822a | Merge branch 'cntkteam/onnx_without_batch_axis' of https://github.com/Microsoft/CNTK into cntkteam/onnx_without_batch_axis | 02 August 2018, 23:14:56 UTC |
refs/heads/csnugetquickfix | bee6947 | add doc for Clone | 31 October 2017, 21:48:10 UTC |
refs/heads/dapisani/center_loss | b39f729 | Revert Main | 26 May 2017, 18:18:30 UTC |
refs/heads/dapisani/cntk_converter | 0b384cb | Crosstalk Caffe Bug Fixes Resolved bugs due to: 1. Scaling in batch normalization was done twice 2. Splicing did not take into account a default axis 3. Did not support newer protobuf versions. 4. Stdout was not being ignored during validation | 09 October 2017, 17:10:15 UTC |
refs/heads/dependabot/nuget/Examples/Evaluation/CNTKAzureTutorial01/CNTKAzureTutorial01/Newtonsoft.Json-13.0.1 | 9f269cc | Bump Newtonsoft.Json Bumps [Newtonsoft.Json](https://github.com/JamesNK/Newtonsoft.Json) from 6.0.8 to 13.0.1. - [Release notes](https://github.com/JamesNK/Newtonsoft.Json/releases) - [Commits](https://github.com/JamesNK/Newtonsoft.Json/compare/6.0.8...13.0.1) --- updated-dependencies: - dependency-name: Newtonsoft.Json dependency-type: direct:production ... Signed-off-by: dependabot[bot] <support@github.com> | 22 June 2022, 20:21:28 UTC |
refs/heads/dependabot/nuget/Examples/Evaluation/LegacyEvalDll/EvalDll_AzureTutorial/EvalDll_AzureTutorial/Newtonsoft.Json-13.0.1 | 014da9b | Bump Newtonsoft.Json Bumps [Newtonsoft.Json](https://github.com/JamesNK/Newtonsoft.Json) from 6.0.8 to 13.0.1. - [Release notes](https://github.com/JamesNK/Newtonsoft.Json/releases) - [Commits](https://github.com/JamesNK/Newtonsoft.Json/compare/6.0.8...13.0.1) --- updated-dependencies: - dependency-name: Newtonsoft.Json dependency-type: direct:production ... Signed-off-by: dependabot[bot] <support@github.com> | 22 June 2022, 20:16:50 UTC |
refs/heads/dongyu/addhypertune | 76ffa9e | fix BlockMomentumSGD.h | 11 January 2017, 06:07:19 UTC |
refs/heads/dongyu/htkreaderext | c4fb0e5 | Merge branch 'master' of https://github.com/Microsoft/CNTK into dongyu/htkreaderext Conflicts: Source/Readers/HTKMLFReader/HTKMLFReader.cpp Source/Readers/HTKMLFReader/htkfeatio.h Source/Readers/HTKMLFReader/utterancesourcemulti.h | 02 January 2017, 01:12:32 UTC |
refs/heads/dongyu/writerfix | 8fef21b | support minibatchSize (for larger models that will be out of memory without it) and frameMode (for RNNNode based Windowed RNN) for HTKMLFWriter (and HTKMLFReader). | 09 August 2016, 21:51:45 UTC |
refs/heads/duli/histogram | c65b2b6 | Integrate b5e92de45e82443a80dc729d8e9f1f40acaf5cb6 into master | 23 August 2017, 20:09:19 UTC |
refs/heads/duli/tensorboard_image | ee04a89 | Tensorboard image feature support. | 07 September 2017, 05:35:17 UTC |
refs/heads/duli/tensorboard_image1 | a62e793 | Merge branch 'duli/tensorboard_image1' of https://github.com/Microsoft/cntk into duli/tensorboard_image1 | 06 September 2017, 04:34:18 UTC |
refs/heads/duli/tensorboard_image2 | 832c3f3 | Merge branch 'duli/tensorboard_image2' of https://github.com/Microsoft/cntk into duli/tensorboard_image2 | 25 August 2017, 16:06:34 UTC |
refs/heads/duli/tensorboard_staging | b55ab9f | Tensorboard image feature | 25 August 2017, 16:36:05 UTC |
refs/heads/ebarsoum/cleanup | eda71ab | Merge branch 'master' into ebarsoum/cleanup For CR | 30 March 2017, 20:21:25 UTC |
refs/heads/ebarsoum/cntkv2doc | 4ebcca9 | Update convolution_map description. | 19 October 2016, 18:27:34 UTC |
refs/heads/ebarsoum/dilation_with_padding | 1cbccd0 | Merge branch 'master' into ebarsoum/dilation_with_padding Merge with master... | 21 September 2017, 20:03:28 UTC |
refs/heads/ebarsoum/emotion | 360ff1e | Switch to standard SGD. | 06 November 2016, 20:26:02 UTC |
refs/heads/ebarsoum/inceptionv3 | 4fcc6dc | Add more comments, remove tab and use 32 bit SGD. | 31 October 2016, 22:12:25 UTC |
refs/heads/ebarsoum/iteration | d007293 | Merge with master. | 14 September 2017, 01:29:10 UTC |
refs/heads/ebarsoum/layerconv | b2278ec | Remove SequenceOver. | 09 June 2017, 23:38:21 UTC |
refs/heads/ebarsoum/pixelcnn | 46aaf31 | Merge branch 'master' into ebarsoum/pixelcnn To get random OPs... | 30 May 2017, 18:07:30 UTC |
refs/heads/ebarsoum/png16 | cbcf555 | Update cv::imdecode | 01 March 2017, 18:29:17 UTC |
refs/heads/eldak/adaptiveLearningRate | 8e93cf6 | Some fixes | 12 April 2017, 09:43:42 UTC |
refs/heads/eldak/addingMLFTests | eafcd9e | Adding more mlf tests | 23 May 2017, 10:35:59 UTC |
refs/heads/eldak/addingPerDimMeanVarDeNormalization | 734b776 | Adding PerDimMeanVarDeNormalization node | 07 November 2017, 15:45:16 UTC |
refs/heads/eldak/alignedBundler | 4ffe3a1 | Some not yet working implementation | 28 February 2017, 13:58:44 UTC |
refs/heads/eldak/cesparsebs | 5d14950 | Implementation of sparse labels | 01 March 2017, 14:17:46 UTC |