https://github.com/Microsoft/CNTK

sort by:
Revision Author Date Message Commit Date
3cc806c Insertion boost and length normalization during beam search pruning 23 March 2020, 01:52:38 UTC
3b38b66 N-best rescoring 14 February 2020, 21:37:35 UTC
b4088e8 N-best rescoring 14 February 2020, 21:36:16 UTC
ffa471d Implement joint decoding 09 November 2019, 00:52:34 UTC
67a4260 New branch for RNNT sequence T/S 06 November 2019, 02:03:01 UTC
2ddfea2 force expand blank 24 September 2019, 22:56:55 UTC
4c67aae add Vadim's code 06 September 2019, 17:05:07 UTC
b80f6f0 beam search for merged linear 05 September 2019, 21:34:18 UTC
94d2581 fix htkfeatio 20 July 2019, 01:07:17 UTC
2e96d5b support big output 17 July 2019, 20:47:25 UTC
eb20656 movecpu clean 24 June 2019, 20:47:43 UTC
2f5703c refine transfer from gpu to cpu and profile 24 June 2019, 04:23:36 UTC
3eb0f6f profile and transfer fix 19 June 2019, 21:56:52 UTC
4886bc5 greedy decoding for merged model 14 June 2019, 23:57:18 UTC
1088355 fast fb, clean 12 June 2019, 03:57:11 UTC
d819aff fast fb, other tries 12 June 2019, 01:02:24 UTC
3ea3c21 profile and MB change 05 June 2019, 18:49:41 UTC
fad3efb resize big maxtrix after back prop 04 June 2019, 00:29:07 UTC
fa70d36 merge further 30 May 2019, 08:59:24 UTC
0abf026 fix memory leak 30 May 2019, 00:36:14 UTC
ea3c0df no profile, save memory further 29 May 2019, 06:07:46 UTC
9bd9903 merge linear and loss 29 May 2019, 05:23:03 UTC
d0675c2 profile cost 22 May 2019, 17:52:47 UTC
35935a1 fix build bug 11 May 2019, 05:54:44 UTC
54b40f4 fix bias gradient 10 May 2019, 20:02:56 UTC
61ce131 revert mbchange 04 May 2019, 19:15:40 UTC
6f68c08 refine GPU gradient cal 04 May 2019, 04:50:19 UTC
ba31a0d minibatch change 02 May 2019, 00:11:00 UTC
0b85dd4 Fix build 01 May 2019, 20:46:45 UTC
2359708 rever minibatch change 30 April 2019, 01:19:32 UTC
73e4f59 change minibatch size 29 April 2019, 23:14:02 UTC
b256574 inplace softmax to save memory 21 March 2019, 00:53:15 UTC
cc90db5 add max label length limit 21 March 2019, 00:40:08 UTC
353d8dd fix empty word 17 March 2019, 06:11:14 UTC
26204b2 fix build error for test 16 March 2019, 18:49:55 UTC
e75616a fix build error 16 March 2019, 17:55:19 UTC
8895cb4 no print word 16 March 2019, 16:46:08 UTC
8c65fdf bias work, print words 16 March 2019, 16:44:04 UTC
d1aaaea bias node added, but not work 09 March 2019, 19:06:10 UTC
d4555f3 bias version 2 01 March 2019, 17:55:04 UTC
bce5346 bias version 1 21 February 2019, 21:19:06 UTC
36d9cd1 inplace softmax 20 February 2019, 22:56:02 UTC
944dd8d move softmax inside 31 December 2018, 21:34:45 UTC
4603a2e fix typename error 31 December 2018, 18:04:00 UTC
7485ad3 fix tab 30 December 2018, 06:32:50 UTC
e9beff5 Merge branch 'ruizhao/RNNT_830merge' of https://github.com/Microsoft/cntk into ruizhao/RNNT_830merge 30 December 2018, 05:42:01 UTC
0e0d6ee training eval 30 December 2018, 05:40:22 UTC
255573d training set evaluation 29 December 2018, 01:46:09 UTC
b318a2d gpuall and beam for plusbroadcast 28 December 2018, 22:57:39 UTC
6699320 greedy for merged sum 27 December 2018, 04:43:37 UTC
2770a3d greedy and beam search for first sum version 26 December 2018, 21:29:15 UTC
0917c97 train error evaluation 26 December 2018, 20:57:11 UTC
70ed5e5 add RNNErrorNode 21 December 2018, 19:15:25 UTC
cc72339 fix print error 21 December 2018, 01:39:35 UTC
f02a47f fix dimension mismatch 20 December 2018, 22:40:51 UTC
d7fc1b7 fix test build error 20 December 2018, 07:16:56 UTC
1fbbb0c remove greedy/beam decoder 20 December 2018, 01:49:03 UTC
5a49d31 fix build error 20 December 2018, 00:23:21 UTC
e99342c fix tab 19 December 2018, 23:31:29 UTC
b31cd8e support mixed add 19 December 2018, 22:57:33 UTC
843974b greedy word 17 December 2018, 23:04:46 UTC
292308e beam search v2, remove the duplicate forward prop for decoder 17 December 2018, 18:46:46 UTC
2cc7a04 beam search v1 16 December 2018, 04:07:09 UTC
d9da1d3 fix greedy search has 0 output 10 December 2018, 23:55:37 UTC
73ea15a greedy decoding 07 December 2018, 19:34:32 UTC
bed32cd global norm 01 December 2018, 00:31:50 UTC
7f8a645 norm gradient for decoder 20 November 2018, 23:54:31 UTC
1a31402 fix fabs error 2 17 November 2018, 06:00:09 UTC
0f07c2c fix fabs error 17 November 2018, 05:58:00 UTC
a05456e fast version norm 10 17 November 2018, 00:46:15 UTC
d28fdd0 sln change 15 November 2018, 01:17:00 UTC
93cc179 fix reader issue 15 November 2018, 01:12:24 UTC
c78c3c1 RNNT could run 09 November 2018, 00:18:37 UTC
e9ab9a7 before make python work 06 November 2018, 23:41:17 UTC
2dca1fb merge RNNT step2 25 October 2018, 23:09:35 UTC
69f2a95 merge RNNT step1 17 September 2018, 22:27:00 UTC
c01d4e4 test build 30 August 2018, 21:04:44 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
back to top