Граф коммитов

449 Коммитов

Автор SHA1 Сообщение Дата
duli1 42b2e68d87 Squashed commit of the following:
commit fb1ef6a11cd5d31484c0a229f2ec813ddc885a08
Author: duli1 <duli1@DU-LI-WS>
Date:   Fri Aug 25 10:02:23 2017 -0700

    Tensorboard image feature

commit a5584c245b6c8d6b4b31bacd8ac1dc3ed9242463
Author: duli1 <duli1@DU-LI-WS>
Date:   Fri Aug 25 10:02:23 2017 -0700

    Tensorboard image feature
2017-09-05 21:31:31 -07:00
Zhou Wang 364dd8f69f Update CPP Eval Examples; update UWP Eval tests; Add more examples in
C#; and various fixes

Details:

fix wrong link

use /Zo option instead of undocumented /d2Zi+; remove /d2Zi+ for debug build;

move EvalMultithreads.cpp into Tests/EndToEndTests/EvalClientsTests/CNTKLibraryCPPEvalExamplesTests as preparation for the new CPP examples

add new CNTKLibraryCPPEvalExamples

use stdout

add first sample in CPPEvalExamples

add parallel evaluation examples

complete cpp examples

adapt linux build

add CPP tests to UWP

enable more tests on UWP

add running new tests; update baseline

add new cpp exampels to CNTKLibraryExamples.sln; use stdout instead of stderr in legacy eval dll; add missing cs examples

add file types for UWP

flush stdout to make cygwin happy

update baseline file for CPPUWPEval

update baseline for CNTKLibraryCSEvalExamplesTest
2017-08-29 22:17:50 +02:00
Alexey Reznichenko 65471cdcd9 Add Proposal Layer as a native UDF 2017-08-19 10:03:10 +02:00
Mark Hillebrand 11e593d085 Remove the block multiplier code, as it isn't used right now.
Closes #1805.
2017-08-07 10:42:02 +02:00
Mark Hillebrand dd74af7e7e Bump version numbers 2017-07-31 17:55:45 +02:00
Eldar Akchurin 32cbf7071b Prohibiting infinite minibatch sources with training session 2017-07-26 14:37:50 +02:00
Alexey Reznichenko 93cc8c40db Add index caching
* Refactor index data structures and rewrite indexers (with most changes
    in the text index builder).
  * Add best effort caching: the cache is written out asynchronously in a
    separate thread, on restart the index builder tries to restore the
    index from cache (as long as the cache is not older than the input
    file) and goes back no normal indexing if that fails (i.e., the cache
    is corrupt).
  * Refactor and simplify MemoryBuffer (renamed to BufferedFileReader).
  * Use KMP patter-matching to simply sample counting with non-empty main
    stream (num samples in sequence = number of lines that contain main
    stream name).
  * Refactor and simplify MLFIndexBuilder (it now also uses
    BufferedFileReader)
  * Use 512KB chunks when loading index from cache for faster reading.
  * Add a number of unit tests for the indexing both with and without
    caching.
2017-07-25 10:11:25 +02:00
Eldar Akchurin d542921556 Adding python user deserializer 2017-07-24 15:10:59 +02:00
Ivan Rodriguez 84e74c899b addressing comments in CR 2017-07-19 10:34:11 +02:00
Wolfgang Manousek 2021b2bc3e remove compute_52 and sm_52 2017-07-13 17:34:02 +02:00
Wolfgang Manousek 8cd36ce385 more CR comments addressed 2017-07-13 17:34:02 +02:00
Wolfgang Manousek 4c2f93c658 addressed CR 2017-07-13 17:34:02 +02:00
Wolfgang Manousek ce44b81f70 minor improvement/fixes 2017-07-13 17:34:01 +02:00
Wolfgang Manousek bfc662de0a added compilation for compute 5.2,6.0 and 6.1 2017-07-13 17:34:01 +02:00
Wolfgang Manousek 2582b9ee91 addressed CR comments 2017-06-07 16:50:37 +02:00
Wolfgang Manousek 5ecc834722 updating links to old wiki - referencing now the doc site 2017-06-07 15:55:34 +02:00
Mark Hillebrand d6b55e28af Bump version numbers 2017-06-01 15:12:14 +02:00
Amit Agarwal 71b8a35771 CNTK v2 library: Add Binary convolution native user-defined Function library. 2017-05-27 22:52:08 -07:00
Mark Hillebrand 8dc56e2d7c Bump version numbers 2017-05-24 14:01:23 +02:00
Eldar Akchurin 98ac0be1c2 Ability to choose mbsize based on stream 2017-05-18 18:02:37 +02:00
Zhou Wang dd91e95512 remove -DSWIG from javabidning on Linux 2017-05-12 18:28:47 +02:00
Ratan Rai Sur 9dc60f2c80 add -DSWIG to java c++ wrapper compilation for compatibility with udf serialization 2017-05-10 00:20:19 -04:00
Ratan Rai Sur c6de4db35a Java Bindings (squashed)
adding some doc

use $$

add comments

automatically loadLibrary in java

semicolon

removing .iml

add more dependencies

move java static block code to cntk_java.i

remove also class files.

Adding .java files to jar

changing test location

typo

ignore DeviceDescriptorVector size constructor

DRY in Main.java

use List interface instead of ArrayList implementation

moving test location, expanding tests to gpu, fixing comments

move linux java tests

updating baseline.txt
2017-05-10 00:20:18 -04:00
Alexey Reznichenko 5bc3661988 Improve UDF serialization
* Add a base class for User-Defined Functions
  * Add support for native udf serialization
  * Change API to accept std::function callbacks
2017-05-09 10:08:23 +02:00
Mark Hillebrand a1e10ff9e7 Bump version numbers 2017-04-21 15:58:00 +02:00
Eldar Akchurin ed2f97311c Fixing compilation of CPUMatrix on Windows: heap overflow fixed 2017-04-19 10:59:53 +02:00
Amit Agarwal c7c2547f55 CNTK v2 library: Fixed handling of references of existing netowrk matrix storage handed out from Forward/Backward 2017-04-14 01:33:24 -07:00
Amit Agarwal 1f23f6e161 CNTK v2 library: Add ability to register and instantiate native C++ user-defined functions from python 2017-04-14 01:33:24 -07:00
Project Philly 5a25d3ecce Integrate jqian/wip/gdr into master 2017-04-13 16:23:44 -07:00
Junjie Qian bc04458bda Do gradients allreduce aggregation with GPUDirect RDMA
1. Iallreduce with cuda-aware not supported
ref: https://www.open-mpi.org/faq/?category=runcuda#mpi-apis-no-cuda
2017-04-13 09:16:37 -07:00
Mark Hillebrand 3542533708 Py 3.6 build & test 2017-04-13 14:19:19 +02:00
Zhou Wang 071367f8cb further reduce # instance of version defintions 2017-04-12 15:08:56 +02:00
Eldar Akchurin 0ec0a2af87 Some memory optimizations and refactoring of HTK deserializer 2017-04-07 15:56:47 +02:00
Eldar Akchurin 9af17fa354 Introducing chunked mlf deserializer 2017-04-06 16:52:51 +02:00
Mark Hillebrand 861bfbd714 Bump versions 2017-04-03 18:16:31 +02:00
Eldar Akchurin ca3ba7fc62 Splitting Trainer and Evaluator 2017-03-27 09:32:47 +02:00
Friedel van Megen 20422fc5a3 removed comment 2017-03-22 14:50:05 +01:00
Friedel van Megen c40639c3ed update latest comments 2017-03-22 14:19:14 +01:00
Friedel van Megen e27fa18d0a fix typo in makefile 2017-03-21 20:39:28 +01:00
Friedel van Megen f599fe0619 renaming cntk.library to cntk.core
(thats what other apps link to)
2017-03-21 17:45:46 +01:00
Friedel van Megen e8cf030d72 fix names based on Eldar's comments. 2017-03-21 17:34:43 +01:00
Friedel van Megen 45421e403c adding version to makefile as well 2017-03-21 00:15:57 +01:00
Friedel van Megen 07915f01d0 stringify 2017-03-20 16:56:30 +01:00
Friedel van Megen 9760a34f92 escaping " 2017-03-20 16:48:54 +01:00
Friedel van Megen 6242f6cfaf adding version to every compile 2017-03-20 16:43:43 +01:00
Friedel van Megen 87914900c5 modifying linux build file 2017-03-20 15:37:15 +01:00
Alexey Reznichenko 66c922b2e0 Fix second gpu detection in GPUMatrixTests 2017-03-09 18:11:04 +01:00
Project Philly b0246324a1 Integrate qiwye/asgd-fix into master 2017-03-08 06:52:02 -08:00
PengchengHe 0df54d8333 Enable python setup in debug bits 2017-03-08 01:38:37 +00:00
Qiwei Ye 00b0b0c2d1 Merge branch 'master' into asgd-fix 2017-03-06 23:40:44 +08:00