mxnet-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
Subject [incubator-mxnet] branch szha-patch-1 created (now 32c7f63)
Date Fri, 01 Feb 2019 08:52:50 GMT
This is an automated email from the ASF dual-hosted git repository.

zhasheng pushed a change to branch szha-patch-1
in repository

      at 32c7f63  Update NOTICE

This branch includes the following new commits:

     new ab64fe7  Initial commit
     new be75254  push in some note
     new c1180df  Update
     new 3e9e0c9  Update
     new 5120bb8  Update
     new b8b052b  1. add comments; 2. add LayerWithNArrayInterface; 3. add visual studio solution
     new c6390a0  incomplete operator interface
     new 7f7c6ee  move storage.h to include, simplify it
     new fd76356  refactor dagengine
     new c05dddc  refactor narray, keep original code for now
     new bc366d2  ok
     new b73bb48  clean
     new eeb3939  Update
     new cafb4e6  Update
     new d27ce8d  Update
     new 86ce32d  minor renaming
     new 3317180  activation op
     new e4e2178  chg
     new 3674aef  add op and extra opt
     new 062dba5  chg
     new 12b1c97  compile
     new 043bfff  act
     new 3d51583  chg readme
     new e595baf  add fullc op
     new 3a76b8d  minor
     new be5bf08  minor
     new 9c214c8  minor in act
     new 7a1296d  remove inplace check
     new 4d175e4  add property
     new df4227e  Merge pull request #1 from dmlc/dev
     new d4c1310  draft wrapper interface
     new f268663  draft wrapper interface
     new 1bd6858  conv op
     new 2bae9ac  Merge branch 'dev' of ssh:// into dev
     new f52a14e  pooling op
     new 275206c  reshape op
     new 4c3a9b6  dropout op
     new 4194c58  rename wrapper to api
     new 7924c25  minor fix
     new b55212f  api change
     new 7eb3ca7  Merge branch 'dev'
     new c7482d1  update on function registry
     new b0e41fd  api
     new 82a7f35  initial version of api
     new 0e30f9b  remove debug
     new 9947f97  python3 compatibility
     new ca89d4c  clean
     new 8424780  make narray working
     new 438ab2b  threaded engine draft
     new de37edb  Merge branch 'master' of
     new 59a7cee  refactor api
     new 480604f  minor fix
     new 65123c9  add travis
     new a89a2af  change to calloc
     new 0288356  try to fix gcc48 in travis
     new a312bb0  add doc
     new 028af64  fix doc
     new f8634d0  fix Doxyfile
     new 67f60bc  add api to doc
     new 6f93ae6  fix style
     new 1352684  minor
     new 23fa5ca  add nvcc
     new d483bb2  threadlocal
     new 6c6c013  minor change
     new df0f55a  Merge branch 'master' of
     new 9fbaa6b  fix doc
     new 7edd02b  fix style
     new 38747d0  Merge branch 'master' of
     new 6da99c9  Interface of Symbol
     new fec1c23  Merge branch 'master' of
     new 5a757ff  separate atomic_symbol from symbol
     new d239267  add implementation
     new cda674c  add symbol to makefile
     new a919707  narray operator
     new 0a9128b  fix a bit
     new cd6ffee  expose c api
     new 66d576c  symbol python api
     new c7d5e59  fix lint and doc
     new cb4b454  add template function to create atomic symbol
     new 007b15b  resolve some errors and lints
     new 93f0801  fix docs
     new 3a3003c  merge makefile
     new 7094ee1  rename
     new a5b8e48  mkfile
     new 8eb61be  Merge pull request #4 from winstywang/master
     new ada652b  docs
     new 8edf819  add type string to atomic symbol
     new 6214cd9  register AtomicSymbol
     new f137132  add virtual destructor
     new e3de4df  move definition to cc file
     new 6325cc4  fix doc
     new 31a3ad4  Merge pull request #5 from mavenlin/master
     new 3ae48a4  fc symbol
     new 1275470  move api_registry to registry
     new 5bd0fb8  move AtomicSymbol Registry to common Registry
     new f2a4bc7  repair some bugs
     new 573f238  Merge pull request #6 from mavenlin/master
     new 0ff57d3  Merge remote-tracking branch 'remote/origin/master'
     new 95116e9  static op wrapper
     new b8be6c3  fix fc sym
     new 05263cc  bind
     new b5c15e2  fix bind
     new 118acd2  fix
     new 4722b73  fix style and doc
     new b961bc1  Merge pull request #7 from winstywang/master
     new 504f735  static graph
     new 502dc47  move
     new 834ac6b  mv symbol def to static_operator
     new f5cf34c  move api to include
     new 975f489  chg
     new 35da51f  Merge branch 'master' of
     new 8f320c4  simplify static graph
     new d75d0ef  Merge remote-tracking branch 'remote/origin/master'
     new c516e06  move api
     new c815526  remove readme
     new 2e2885d  fix doc
     new 22ebc27  minor python
     new bccfbeb  fix static graph
     new 2c9091d  Merge pull request #8 from winstywang/master
     new fead856  fix travis
     new e50817d  new symbol interface
     new 84023c7  Merge pull request #10 from antinucleon/master
     new 6e538fc  static graph
     new 172b987  Merge pull request #11 from antinucleon/master
     new 5191b25  chg composite op interface
     new e4a7c5d  symbol switch to vector
     new 1abdf93  [storage] storage backends
     new d22b034  [storage] naive storage manager
     new c191d9b  [storage] enable c11 feature
     new 61ea32c  [storage] move define to top
     new a3cabcd  [storage] try change header
     new b2bc69f  [storage] use old function
     new 21674ef  [storage] use if
     new 2214ce3  [storage] storage managers
     new 84bf795  [storage] putting things together
     new 1ad28eb  [storage] fix Makefile
     new b33a8a0  reorg code
     new b5f75f1  Merge pull request #12 from antinucleon/master
     new cc2195f  chg interface
     new 4dd1924  out_data is necessary, e.g. sigmoid
     new 0cf889c  Merge pull request #13 from mavenlin/master
     new 270bfb2  discussed interface
     new b5f485f  Merge pull request #14 from antinucleon/master
     new b1127a7  merge
     new 3626c34  refactor as discussed
     new cbd23f7  Merge pull request #15 from antinucleon/master
     new 80a3d42  a bit more minor change
     new df28352  Merge pull request #16 from antinucleon/master
     new fcb45cb  Add activation op
     new ee2d7f1  infer shape
     new 8fe878e  minor change in act
     new a3cde4a  tmp save
     new 36cde60  [storage] refactor things a bit
     new 1b3163c  [storage] fix extra comma
     new dda24c2  [storage] a simple test for storage
     new 7c85007  [storage] Merge branch 'master' of
     new f932cd2  [storage] fix Makefile
     new 364349f  [storage] make Makefile right this time
     new 264c313  [storage] Merge branch 'master' into storage
     new 62f98a8  [storage] fix Makefile for dependency
     new 7cae8ce  [storage] add document to some common functions
     new 1fe4efb  [storage] disable Doxygen preprocessing so every doc works
     new f5358fe  [storage] document some more to pass lint
     new 3a8e071  [storage] Doxygen does not support scope operator?
     new 9f4d31c  Merge pull request #17 from dmlc/storage
     new a2a6795  adapt things to parameter
     new c78775a  implement backward
     new 83d89f3  Merge remote-tracking branch 'upstream/master'
     new d9748f1  add enum to activation
     new f5c6d67  add pooling op
     new 0448d67  graph exec as discussed
     new 7a1832c  update NArray
     new 910738d  MNIST is OK
     new 838da47  Merge pull request #20 from antinucleon/master
     new 72eb074  fix python call bug
     new c956ae7  Merge pull request #2 from mavenlin/master
     new addb2f2  fix travis
     new 83b8788  Merge pull request #21 from antinucleon/master
     new 4e0e413  Read docs
     new ef7bb06  simplify symbol creator as discussed
     new 0635103  Merge pull request #22 from antinucleon/master
     new 0f77b3e  return value does not need std move
     new 6b21c24  Read docs
     new ec0f128  simplify symbol creator as discussed
     new 140449b  return value does not need std move
     new 4d61ded  support backward of backward
     new bb4cb6a  format annotation
     new 6de67dd  Revert "support backward of backward"
     new 8b9489e  init sphinx
     new 85c51c0  doc
     new bd7ad19  format annotation
     new 985484b  try
     new ae64c5f  autodoc
     new 225fabd  add pythondoc
     new daf41e1  add cpp doc
     new 661cfd4  refactor reg
     new f8d4e0d  minor fix
     new 50dc62f  change registry
     new 25bc1ef  remove
     new c2570ec  add
     new c993bd0  change
     new 6a541b6  merge
     new 2cbe804  wtf
     new a8e73a8  fix lint
     new e6c87da  Merge pull request #25 from antinucleon/master
     new a726bae  fix docs
     new 6cf4929  fix doc
     new daeda6e  remove typo
     new b9f8d01  Merge pull request #26 from antinucleon/master
     new fae71c2  add doxygen
     new 8d95f52  fun
     new 74c41d7  finish merge remote master
     new 63ff6d0  built in python, start polishing new feature required
     new 065b061  finish old version registry in C
     new 7a3fb78  [simple-engine] cherry pick Minjie's refactoring on interface
     new 6bfcc7d  [simple-engine] fix typo and unmark virtual
     new e149942  [simple-engine] WIP, need to refactor interface and remove inheritance
     new 2dbdc0f  dfs visit use const shared_ptr<Node>&
     new 2cc53b4  arg mismatch code
     new d8bcd68  agg arg_grads
     new 686c619  symbolic gradient
     new 01362f8  compile debug
     new 1e0459c  fix docstring
     new 03f05d4  modify to dmlc registry
     new cfe80c6  [simple-engine] A not so simple engine that should somehow work Missing parts: 1. Workers (this should be easy since all conflicts have been resolved) 2. Runtime context (I have totally no idea)
     new 005db3a  [simple-engine] lint
     new e940668  add nosetest
     new 7144cb4  pass python mnist test, begin cleaning
     new 9ffdae3  minor
     new 4f6ef5d  clean io interface
     new 6812764  Merge pull request #27 from mavenlin/master
     new 12b3d22  finish merge remote master
     new f673f7d  built in python, start polishing new feature required
     new a8dcbd1  finish old version registry in C
     new 7f36bbb  modify to dmlc registry
     new 299e1d8  pass python mnist test, begin cleaning
     new 7cf5506  clean io interface
     new 994bc41  modify to pass travis
     new 819e5f8  merging latest master
     new 74d9936  Merge remote-tracking branch 'upstream/master'
     new be11e16  revert to gcc 4.8
     new 7f93905  Merge pull request #29 from antinucleon/master
     new 9028805  add convolution op
     new defbdbc  ToStaticGraph keep backward source
     new 5dfc759  [simple-engine] remove unnecessary files
     new 6b948fd  [simple-engine] that goes worker
     new c7146a7  [simple-engine] passed some tests
     new fef28d2  [simple-engine] some document
     new 761e0e1  [simple-engine] implement missing functions
     new 11432db  [simple-engine] fix order
     new 50bf69b  [simple-engine] fix concurrency
     new 1ad9132  [simple-engine] fix a concurrency bug
     new 4662b4e  [simple-engine] lint
     new 4bd535f  remove 2d requirement in fullc
     new 1045c18  conv is able to work
     new c057b56  Merge pull request #31 from antinucleon/master
     new 5e01b2a  more checks on shape
     new d59a125  [simple-engine] Merge branch 'master' into simple-engine
     new 783051d  user/developer guide sketch
     new 35e1d83  Merge pull request #32 from mavenlin/master
     new e71055c  [simple-engine] @antinucleon Sorry I had to turn this on to pass Doxygen
     new be8312d  [simple-engine] heavier test on engine
     new 3a15c13  [simple-engine] add tests to Makefile
     new e9ab2a2  [simple-engine] use macro for debugging
     new a3dbb93  [simple-engine] fix engine bug
     new 6b1b9a1  [simple-engine] even more tests
     new a1c02d9  chg
     new 3eec9ad  merge
     new 46844a1  change conv wmat shape
     new 39f2dfc  reshape and flatten
     new 1cc0ccb  Merge pull request #33 from antinucleon/master
     new 2ff2193  minor
     new 3e406ff  finish merge remote master
     new c81b60b  built in python, start polishing new feature required
     new 0ebf6aa  finish old version registry in C
     new 7511c12  modify to dmlc registry
     new c03d62b  pass python mnist test, begin cleaning
     new 08dcaf7  clean io interface
     new bcd2652  finish merge remote master
     new 97a88cb  built in python, start polishing new feature required
     new 6915299  finish old version registry in C
     new 305d2af  modify to dmlc registry
     new 04f9888  pass python mnist test, begin cleaning
     new c6ae534  clean io interface
     new 3da4ac0  modify to pass travis
     new f4fd400  finish refactoring io code
     new b6ebf94  add in doc
     new 1c57396  [simple-engine] Merge branch 'master' into simple-engine
     new 6e1b7dd  merged latest master
     new 8021478  resolve rebase problem
     new d576bb3  minor
     new 2ad67a3  Merge pull request #34 from antinucleon/master
     new 1b19f9d  finish merge remote master
     new 0c24dbf  built in python, start polishing new feature required
     new a2f03a7  finish old version registry in C
     new 9d4d98e  modify to dmlc registry
     new 1544258  pass python mnist test, begin cleaning
     new 36cff22  clean io interface
     new 3ee7f24  finish merge remote master
     new 88d2db7  built in python, start polishing new feature required
     new 2ff4ff0  finish old version registry in C
     new 9b002de  modify to dmlc registry
     new c29793f  pass python mnist test, begin cleaning
     new 56b4129  clean io interface
     new e48e400  modify to pass travis
     new 49728eb  finish merge remote master
     new cb99a1e  built in python, start polishing new feature required
     new 24ad6f5  finish old version registry in C
     new 3931343  modify to dmlc registry
     new b1e2bd9  pass python mnist test, begin cleaning
     new 0e99411  clean io interface
     new 51b7208  finish merge remote master
     new 358a623  finish old version registry in C
     new 94c6872  modify to dmlc registry
     new 8caa4ac  pass python mnist test, begin cleaning
     new 3fc60b1  clean io interface
     new af12892  finish refactoring io code
     new 08f12b8  add in doc
     new 9bdcebd  merged latest master
     new e68452f  add local io into test
     new 32d10ae  use local io in test
     new a4e70b2  merge master
     new f1b9764  meet python3 strict
     new 982f18d  can not train in python3
     new 602a436  [simple-engine] Merge branch 'master' into simple-engine
     new 25e3363  list zip for py3, no reset in iter
     new 5e9b53b  Merge pull request #28 from sneakerkg/master
     new e85fb0d  [simple-engine] fix concurrency bug
     new 5d4189e  [simple-engine] obj pool raw implementation
     new bb16dc4  [simple-engine] Merge branch 'master' into simple-engine
     new 8628138  [simple-engine] fix counter
     new 5b609f3  gitsubmodules
     new 0341d2c  update readme for submodule
     new 296e942  Merge pull request #35 from thirdwing/master
     new 711f715  update python guide
     new e3945ec  add auxiliary input in interface
     new 33b536a  narray saveload
     new 46a743d  add bn, not test yet
     new 705ab71  Merge remote-tracking branch 'upstream/master' into op
     new 8642c1e  update submodule
     new 61aa571  Merge pull request #30 from dmlc/simple-engine
     new 32e0b7c  merge
     new a498f6e  various changes
     new 4cbf9de  Merge pull request #36 from antinucleon/op
     new 5644490  Delete
     new 4a43f0d  merging the code, not compiled
     new 2251812  add image rec and associate files in
     new 37b986e  add narray scalar op
     new 36b7e8f  Merge pull request #37 from antinucleon/narray
     new 38931a6  pass compilation, not tested
     new c4cdd27  add graph
     new a83b6ac  compile on mac osx
     new 27f8241  add set/get item to narray
     new bb17a4a  reload iop
     new 2ca56c9  checkin naive engine back
     new 75eabd4  yMerge branch 'master' of ssh://
     new 6f95b02  fix gpu
     new 0ab1e18  update
     new beb6702  Merge branch 'master' of
     new a5a54e2  Merge pull request #39 from antinucleon/master
     new e214034  [engine-refactor] object pool optimization
     new 0eb739d  [engine-refactor] DAG_ENGINE_DEBUG macro
     new 25b9ee6  [engine-refactor] rename
     new 152b6de  [engine-refactor] rename to ThreadedEngine
     new f23bc01  [engine-refactor] remove redundant lines
     new a5c19c9  [engine-refactor] switch engine in
     new 5bdb007  [engine-refactor] switch back to threaded engine
     new 73f0110  [engine-refactor] remove executable permission
     new 94c70ed  [engine-refactor] refactor using var queue abstraction
     new 5b06ae3  [engine-refactor] lambdas are faster if inlined
     new d6f8459  [engine-refactor] reindent
     new ffa6444  update numeric type
     new 3c769f1  Merge pull request #38 from thirdwing/master
     new a5c7fd3  add rdiv/rsub
     new 990b798  Merge remote-tracking branch 'upstream/master'
     new c1603d7  merge augmenter, modify param attribute
     new 63a0043  add elementwise sym
     new d6ceed3  call augprocess in base iter
     new 23910e5  add set/unary
     new 6da7787  Merge branch 'master' of
     new e6e654e  remove lhs
     new 16d8569  Merge pull request #42 from antinucleon/master
     new 1ad68a2  improve bind and doc a bit
     new 8aa7255  init ps.h
     new 8b59d33  refactor ps::worker
     new d615746  Add var args in operator
     new 15c17ea  add concat op
     new 44c21aa  Merge branch 'master' of
     new 75ee58a  update ps::worekr/server
     new 022e6a4  add ps::node
     new 78ac41b  more doc
     new 8bb57e8  minor
     new da9b0cd  Merge pull request #43 from antinucleon/master
     new dd97baa  simplify ps.h
     new da48024  tiny
     new 2c3d9c2  add cifar10 draft
     new 1fa43df  Merge pull request #44 from antinucleon/master
     new 88dfa9c  mnist gpu
     new a655d30  Merge pull request #45 from antinucleon/master
     new aa91f4b  [engine-refactor] Merge branch 'master' into engine-refactor
     new 9794621  [engine-refactor] fix storage concurrency issue
     new ad97494  [engine-refactor] DAGEngine -> Engine
     new 261c757  [engine-refactor] renaming [fixes #40]
     new 933b86f  [engine-refactor] rename in test file
     new 06637a3  recio works
     new ccfba89  merging the code, not compiled
     new a54f6c9  add image rec and associate files in
     new 9cad774  pass compilation, not tested
     new 2566b3e  merge augmenter, modify param attribute
     new bb60063  call augprocess in base iter
     new 0913273  recio works
     new 36aab11  checked cifar, work when nthread=1
     new 4d08d0c  checked in imagerec
     new ff58b24  do not run auto test_io, just visualize test
     new 528b5a4  one path bug
     new 3f199e5  add python test
     new 13b4aba  update
     new 8644884  add sync copy, change numpy to as numpy
     new 7790181  fix
     new 726817f  Merge pull request #46 from sneakerkg/master
     new 5f4fd55  Merge remote-tracking branch 'upstream/master'
     new 69576b8  minior
     new 089cce4  update test
     new e25b129  Merge pull request #47 from antinucleon/master
     new b3e9ecb  add .bak in unused .cc files
     new d02672e  simplify makefile, automatically generate dependency
     new 13f2936  fix tiny bug in makefile when USE_CUDA=1
     new 39f1d8e  Merge pull request #48 from dmlc/refactor-makefile
     new c68a818  rename simple_engine to pass compiling
     new d64a83f  engine developer doc
     new 938aa2c  [engine-refactor] Merge branch 'master' into engine-refactor
     new 29b20f4  [engine-refactor] Merge commit 'd64a83f' into engine-refactor
     new 0a0d814  [engine-refactor] doc
     new 6beaeff  Merge branch 'master' into ps
     new 00ff283  tempspace resource
     new 3402997  dfs->postorder
     new b6e71a8  toposort->postorder
     new 482f17b  update subproject
     new 8a7ca40  cifar inception like net is OK
     new acd7ed1  Merge remote-tracking branch 'upstream/master'
     new 8e4d2e2  [engine-refactor] stream manager and travis script
     new ba28bd8  Merge pull request #50 from antinucleon/master
     new 3f5334d  kvstore_local
     new 430cca7  passed
     new 2d5ae84  update kvstore
     new 6f09066  [engine-refactor] integrate stream manager
     new f38bca7  [engine-refactor] optimizations: inplace operation and waiting returns if ready
     new a202a30  [engine-refactor] less log
     new d5475d5  [engine-refactor] optimize for IO operations
     new 56579ae  [engine-refactor] lint
     new 8fef335  [engine-refactor] Merge branch 'master' into engine-refactor
     new 8addb28  [engine-refactor] bump submodule version
     new d4e14ac  [engine-refactor] fix compiling issues
     new 3e8e1e0  simple bind
     new 5c5d5db  refactor kvstore
     new cdabe63  fix bug in init agg_buf
     new dd503d7  Merge branch 'master' into ps
     new dfb5b70  key-value list for kvstore in c api, better test_kvstore
     new 3c12463  rename mlp_multi_dev
     new 2c57ccb  pass lint
     new 44bd80a  Merge pull request #53 from dmlc/ps
     new 1b2ebf5  python callback for kvstore is working
     new 7e1cd36  update doc
     new 44fa3b9  Merge pull request #54 from mli/master
     new b0e515a  cudnn conv
     new 4a85448  [engine-refactor] Merge branch 'master' into engine-refactor
     new 036644a  [engine-refactor] fix compiling issue with new master
     new 5e81c98  Merge pull request #55 from dmlc/engine-refactor
     new b60a4eb  Merge pull request #1 from dmlc/master
     new c660d2c  pinned cpu memory
     new 1d0ae53  change kvstore interface
     new f341b03  simplified kvstore_local
     new 5810c83  switch to a pure list(key,value) interface for kvstore
     new 1b3fe0f  impl
     new 68150d6  python interface
     new 34fd684  passed
     new 2b150f5  pass gpu test in
     new 2f738b7  fix lint
     new 8b7f300  fix for nose
     new 3b07140  Merge pull request #57 from mli/master
     new d9f2e3d  cudnn activation, pooling
     new 948c6cd  merge
     new c6d6679  Merge pull request #58 from antinucleon/master
     new 1c9f1c3
     new 24ff8d0  Merge pull request #59 from mli/master
     new 144d538  fix bug in
     new e0e8ff9  Merge pull request #60 from mli/master
     new a28a83a  add stream wait to all pushes, change kvstore to redef pinned memory
     new 9ef1ec6  Merge pull request #62 from tqchen/master
     new 04517cc  Change engine callback to old style; support multiple streams in Naive. Make PushSync inline; Use registry and env variable to select engine
     new 72bab4c  update doc, fix test
     new 47bdb1d  only unit test threaded engine
     new 1c35dc6  Merge pull request #63 from tqchen/master
     new d747857  minor switch to kwargs
     new acc8025  Merge pull request #64 from antinucleon/master
     new 65d5110  rename narray->ndarray
     new da7c7c6  update examples
     new 99c7975  fix lint and compile
     new 4471fc8  rename returns/heads -> outputs
     new cba8358  fix lint
     new 451641c  almost add prefetcher
     new 3d53ac7  Merge pull request #65 from tqchen/master
     new d5cebc8  Merge pull request #2 from dmlc/master
     new d73a486  rename executor.heads() to executor.outputs in example
     new d963d76  disable printing all ops in graphc_executor
     new 61432b0  measure time in
     new 5eb04f8  add LRN
     new 5b57755  start mult-gpu cifa
     new b4febff  add more detailed comments on threaded engine, move dispatcher logic outside the threadedvar complete
     new ea98a4c  update cifa mulgpu
     new 7673db0  Merge pull request #67 from tqchen/master
     new 11c6ce68 seperate scheduling and running policy in threaded engine
     new 67588f9  Bugfix for infershape with aux_shape.size()!=0
     new ab64cc3  move write dispatch inside
     new c18ea83  Merge pull request #68 from tqchen/master
     new f8543e2  move write trigger out correctly
     new 2e4594f  fix
     new 006b188  add leaky relu
     new 26df426  add prefetcher, need merge to compile
     new 7740723  Merge remote-tracking branch 'upstream/master'
     new 220a27c  Merge pull request #69 from tqchen/master
     new 6482c5c  update cifar10 mulgpu
     new 38f4a5f  Merge branch 'master' of git:// into mavenlin-master
     new 6f1b56d  Merge branch 'mavenlin-master'
     new d98901c  Keep things up to date
     new 7e270bf  Merge remote-tracking branch 'upstream/master'
     new 82a3a86  New resource allocation interface, fix bug caused by old allocation interface
     new 89fd652  simplify mlp mulgpu
     new da63410  Merge remote-tracking branch 'upstream/master'
     new b696009  bug fix in cifar mulgpu
     new 63e5775  tiny changes
     new e467e47  merge with min's pR
     new 253cd79  add threaded engine per device
     new 6c4dc57  fix lint
     new 4ab44b18 tiny fix
     new daff15c  Add Per Device Threaded Engine Policy, Explicit use copy
     new 4c1016f  Merge pull request #71 from mli/master
     new 294b35e  fix compile
     new f6e32d9  Merge remote-tracking branch 'upstream/master'
     new 91d420d  Merge pull request #72 from tqchen/master
     new a29ded0  use sphinx_rtd_theme in default
     new bd85aab  Merge remote-tracking branch 'upstream/master'
     new 090918c  fix gpu code dependencies
     new aae9501  Bugfix operator temp space need to change as stream changes
     new d7402c4  mark todo
     new ef9a10a  Merge pull request #73 from mli/master
     new 46e4c6b  Merge pull request #74 from tqchen/master
     new ef2c5dc  Merge remote-tracking branch 'upstream/master'
     new 24a59b3  debugging on ndarray delete
     new cb31917  Enable resource allocator, allocation policy can be configed via env variable
     new f048bd1  remove debug log
     new 146b0a7  Merge pull request #75 from tqchen/master
     new fb0be72  refactor doc/python
     new 92e41e6  more doc
     new 0b45bd5  Merge branch 'master' of
     new 7c80eef  Merge pull request #76 from mli/master
     new c926bb1  Merge pull request #77 from mli/master
     new a7e3009  update doc
     new 297396c  modify bug in getnarray
     new 7e9e1e7  update python_guide on kvstore
     new 6b424e3  update
     new b42da79  import kvstore as kv
     new 41f8b07  Merge pull request #78 from mli/master
     new 9def4a7  Merge remote-tracking branch 'upstream/master'
     new 4a48736  BUGFix in ThreadedEnginePerDevice
     new 02cdd66  Merge remote-tracking branch 'upstream/master'
     new dcacd67  Merge pull request #79 from mli/master
     new ceefa96  update symbol in python_guide
     new 7c6eb9a  fix threaed engine per device
     new f4207b5  Merge pull request #80 from tqchen/master
     new 9c8ed93  Merge remote-tracking branch 'upstream/master'
     new 0f9a849  Merge pull request #3 from dmlc/master
     new b487685  Switch to resource mgr
     new 0a880da  fix tiny bug in iter_image_rec, and tiny refactor
     new 7cbfd57  Merge pull request #82 from mli/master
     new fbc713a  Merge remote-tracking branch 'upstream/master'
     new 58d1553  fix 2D case
     new 81318a1  Merge pull request #81 from antinucleon/master
     new e4e8f9d  prefetcher can work but has uncertainty
     new b83b587  check in new resource manager
     new 262da66  Add random generation engine.
     new 46e776b  allow graph allocator to use random
     new bc8ffb2  Merge pull request #83 from tqchen/master
     new d91a5fc  change temp space allocation to dynamic size
     new 43afd11  Merge pull request #84 from tqchen/master
     new 2512e2a  Add rrelu, dropout
     new 82450ad  Merge pull request #85 from antinucleon/master
     new 4070166  minor chg on dropout
     new 8c87c35  Merge pull request #86 from antinucleon/master
     new f2a96a0  Add graph coloring based temp space allocation.
     new 9514226  Merge pull request #87 from tqchen/master
     new 50aaffb  Enable color based heurestics in node allocator
     new 4f1441e  Merge pull request #88 from tqchen/master
     new c91e0db  make
     new ee1a6ef  merge latest master
     new 1a82035  prefetcher take iter
     new 84350f3  Add finalize
     new fbb1418  more on finalize
     new 828bda0  sanity test passed
     new a9d5227  Merge pull request #89 from tqchen/master
     new 0482b57  use tensor copy, uncertainty solved
     new fabac55  Merge branch 'patch-0917' of git:// into Iroul-patch-0917
     new b74d1f2  Merge branch 'Iroul-patch-0917'
     new 3e61d8f  Finish Up JSON serialization of symbol.
     new 1d9e9c9  Merge pull request #91 from tqchen/master
     new e36351e  use TBlobBatch for intermediate data
     new 9c444a5  Merge remote-tracking branch 'upstream/master'
     new e7c38c0  upgrade mshadow
     new 5de9233  Merge pull request #92 from tqchen/master
     new 9316bc6  add opencv gate
     new dc4ff9d  remove warning
     new cb40e00  Merge pull request #93 from tqchen/master
     new 88645e7  Merge remote-tracking branch 'upstream/master'
     new c21bbf1  minor fix
     new 07dd781  Merge pull request #94 from tqchen/master
     new 608a7a0  some style problem
     new 0dbb4c3  Update
     new 5ae2c70  modify iter param in tests
     new 571a0d7  merged remote master
     new 6c4a04b  minor fix
     new d5d6177  Update
     new a55d6db  Merge branch 'master' of
     new 990823f  Merge pull request #95 from sneakerkg/master
     new 504c777  compile
     new c1e8174  Merge pull request #96 from antinucleon/master
     new 41ce311  new documents
     new 2e2e710  Merge pull request #97 from tqchen/master
     new 422a9a6  minor fix
     new 4ebfc47  draft
     new 24c26d9  Merge remote-tracking branch 'upstream/master'
     new 93f21f1  update document
     new 38a869f  Merge pull request #98 from tqchen/master
     new 3481bd8  Trim the features
     new b558520  Merge pull request #99 from tqchen/master
     new 3941f9e  Update
     new 4229747  operator doc
     new 69ac3e1  Update
     new 2657094  Update
     new 8d40480  Update
     new 559bbb6  make a change
     new 1a9d167  remove finalize, ignore cuda deriver shutdown, add release to pinned storage
     new 84df51a  fix lint, move finalize function into destructor
     new dafe1ee  Merge pull request #100 from tqchen/master
     new 1e27491  more draft
     new ec620a5  merge
     new 9c07172  tmp remove lint
     new b7627d4  Merge pull request #101 from antinucleon/master
     new 912694b  change def of context, change pinned to special device
     new 5b96c3d  fix symbol
     new 5bc10be  Merge pull request #102 from tqchen/master
     new 2248ab5  add, need more work on
     new 4e237e6  markdown modify in
     new a4ddd4b  modify back some explanations for IO param, suggest manually write
     new 9873400  refactor model, only checks test_mlp for now, hand over to bing.
     new ef2a2de  remove print
     new cdc6259  alexnet draft
     new e6cc80f  Merge pull request #103 from tqchen/master
     new 86dbb94  Merge branch 'master' of
     new 935c513  merge
     new ded4555  Merge pull request #104 from antinucleon/master
     new 88a1e68  add showme example and im2rec, modify io,
     new a8c5ed1  Merge branch 'master' of
     new fd201b3  Update model to add save/load and period checkpoint. Removed wait for All at threaded Engine. Still do not know the cause of WaitforVar stalling.
     new f7574e8  add composite symbol
     new bd36bbc  [BUGFIX] Engine CompleteWriteDependency. to_delete_ stale value.
     new 3401ac5  Merge pull request #105 from tqchen/master
     new 5985f34  Merge remote-tracking branch 'upstream/master'
     new 17227a0  Pull bugfix of threaded iter from dmlc-core
     new 509a33b  Merge pull request #106 from tqchen/master
     new 8dfc213  Merge remote-tracking branch 'upstream/master'
     new c8b8513  [IO] Change prefetcher content to NDArray, remove some of parameters
     new a0facaf  Merge pull request #107 from tqchen/master
     new 49c9c9e  Merge remote-tracking branch 'upstream/master'
     new 510ba1b  [BUG FIX] model prediction and model in python3
     new 1c57a28  Merge pull request #109 from antinucleon/master
     new 61dfc92  [KVStore] make kvstore non-singleton
     new 758e281  add slice to ndarray
     new 94f4f55  Merge pull request #108 from tqchen/master
     new 38c406a  [IO] remove default for data_shape, make it required
     new 7f52e0d  [SYMBOL] move static graph to internal
     new a48a9e0  Merge pull request #112 from tqchen/master
     new 0bb9546  Make model training multiple device
     new a5941d8  Merge pull request #113 from tqchen/master
     new ccd0abd  Allow partial positional arguments of input symbol
     new eb3e87d  Merge pull request #1 from dmlc/master
     new 339085a  Merge pull request #115 from tqchen/master
     new d9df556  [BUG FIX] Slice in float type
     new 8149248  Merge pull request #2 from dmlc/master
     new f69c194  Merge pull request #116 from antinucleon/master
     new 74bd455  Merge pull request #3 from dmlc/master
     new 99d5925  Merge pull request #117 from antinucleon/master
     new 8ad680b  move mnist to ualberta server
     new 33e1cbb  add name manager
     new 10739e0  fix optimizer under multi device
     new 02ca41b  Merge pull request #118 from tqchen/master
     new d4f2674  Merge pull request #4 from dmlc/master
     new a898fcd  use mult-gpus in
     new 8155a6a  Merge branch 'master' of
     new 6f7002a  multi-gpus in cifar10
     new 88d4545  BUGFIX
     new 3e625ae  Merge pull request #119 from tqchen/master
     new b60b496  Merge pull request #5 from dmlc/master
     new c1897a7  Merge branch 'master' of
     new de81af2  cifar-recipe draft
     new c5900cb  Merge remote-tracking branch 'upstream/master'
     new 56051cc  More on model doc
     new d857997  [KVStore] BUGFIX GPU merge
     new c0b0093  Merge pull request #120 from tqchen/master
     new 522aa64  Merge pull request #121 from tqchen/master
     new a026814  Merge pull request #6 from dmlc/master
     new ffab7ec  Merge branch 'master' of
     new 632fa73  update draft
     new 6707db1  Merge remote-tracking branch 'upstream/master'
     new f480aaf  add results for multi-gpu cifar
     new 4c2381d  use 1 gpu in default
     new 4576f75  Merge pull request #122 from mli/master
     new 8e8fbcb  Merge remote-tracking branch 'upstream/master'
     new aae9e46  enable callable to metric
     new 4b14d95  Update error handling
     new c411ab6  Merge pull request #123 from tqchen/master
     new 16524b2  [SYMBOL] BUGFIX in DFSVisit
     new d315c92  Merge pull request #124 from tqchen/master
     new 7ca0e38  Merge remote-tracking branch 'upstream/master'
     new f7c039a  [SYMBOL] Add get internals
     new 6558a8a  Merge pull request #125 from tqchen/master
     new 26b1b71  save
     new d29a5de  Merge remote-tracking branch 'upstream/master'
     new f6327ec  refactor NumpyIter
     new 90f8a45  ndarray doc
     new ee5da8d  update build doc
     new d890669  use USE_CUDA_PATH/bin/nvcc if possible
     new cf97948  update
     new 3acbbf2  [ENGINE] update on error handler
     new 77110a7  [ENGINE] some changesto make it more careful
     new 0300da2  Merge pull request #127 from mli/master
     new 90c0921  Merge pull request #128 from tqchen/master
     new a461772  Merge pull request #129 from mli/master
     new 7862179  Merge pull request #7 from dmlc/master
     new c4084cf  update example/mnist
     new 85d5eae  Merge pull request #130 from mli/master
     new bae4324  [ENGINE] More robust termination
     new f9d5272  Update documents
     new 0cb9cb6  Merge pull request #131 from tqchen/master
     new d95fdb0  finish notebook
     new 99dc389  merge
     new bb7d2c5  update mnist exmaple
     new 4430ae1  [MODEL] Allow extra params
     new ece3d50  Merge pull request #133 from tqchen/master
     new 90ca9dd  update the colormap in
     new 69478aa  try mlp.svg
     new 79d4462  remove mlp.svg
     new c854fd4  tiny
     new ccdb65a
     new 56a7f89  update mnist example
     new 2a5f713  tiny
     new 924aa2e  Merge pull request #134 from mli/master
     new eaed425  more conservative cpu request
     new 64aa3a1  Merge pull request #135 from tqchen/master
     new 7b4a6b5  move batch to init
     new bd3dbef  Merge remote-tracking branch 'upstream/master'
     new a9b66fb  update
     new 1d6c25a  Merge pull request #132 from antinucleon/master
     new 017a478  Merge pull request #126 from sneakerkg/master
     new a78516e  [SYMBOL] Use engine operator
     new d8821f7  Merge pull request #136 from tqchen/master
     new db97532  update cifar perf on cudnn v3
     new b2dff70  add include/lib path in makefile for cudnn
     new 4683971  comment out USE_CUDNN_PATH
     new ea18aeb  [example] update mnist
     new b6e8eb9  Merge pull request #137 from mli/master
     new ae06f8d  [unittest] bug fix
     new b71157c  [cpptest] refactor
     new 539a1af  [ENGINE] Remove additional state, more invariant check. [BUGFIX] Conditional variable modification need to be mutex locked
     new 5a08e98  [cpptest] test engines by rand sum expression
     new 5caa221  Merge pull request #139 from mli/master
     new 6cb752c  draft
     new 583f26e  Merge remote-tracking branch 'upstream/master'
     new 5338de8  Merge pull request #138 from tqchen/master
     new 2a1a8fd  [doc] add logo
     new 516f71c  [doc] update logo
     new 75aa956  [doc] update logo
     new 4739c20  [doc] update logo
     new b48b05c  [doc] update logo
     new f0772f3  [doc] update
     new c31fc26  [doc] update
     new ca8bf57  Merge pull request #141 from mli/master
     new 41c9f6f  [DMLC-CORE] update to the priority queue and refactored threadediter
     new 597793c  Merge pull request #140 from tqchen/master
     new c2d997c  [IO] fix python3 bug
     new 1797180  Merge pull request #142 from antinucleon/master
     new 38108a6  fix travis
     new 14c3c14  Merge pull request #143 from antinucleon/master
     new e684b5c  [example] alexnet for imagenet
     new a1a74ae  [example] update alexnet
     new f326831  Merge pull request #144 from mli/master
     new 52ff3b6  [example] minor fix for alexnet
     new 30d5ecd  Merge pull request #147 from mli/master
     new 5c85e3f  [OSX] add travis build.
     new 594773f  [NDArray] add clip op
     new 914d8ca  Merge pull request #145 from dmlc/osx
     new 33dc603  Merge pull request #149 from antinucleon/master
     new 3b517f1  [KVStore, Engine] add priority for copy operations, optimized the reduction.
     new f2f3d4e  Merge pull request #146 from tqchen/master
     new 8857d37  Update
     new bad86cb  Update
     new 0e02bd7  [IO] Refactor, make it more consistent with CXXNet IO
     new d8c5afe  [IO] Fix, make shuffle false by default
     new 82248e9  Merge pull request #155 from tqchen/master
     new c36760e  [IO] skip the affine transformation when parameters not set
     new 1dd16bd  Merge pull request #156 from tqchen/master
     new 6f7873a  [PARAMETER] Remove default setting of op, make resource allocation conservative
     new e459646  Merge pull request #157 from tqchen/master
     new 71edc49  [ENGINE] Change obj pool to normal struct instead of union, to support more compiler
     new a801b3e  Some update
     new f7fecc9  Merge pull request #158 from tqchen/master
     new 22f84b2  [STORAGE] BUGFIX of memory pool release
     new b58725d  [NDARRAY] removed inline from the op to be registered
     new 01b7559  Merge pull request #159 from tqchen/master
     new 011bde4  [python] minor
     new 6f04dec  Merge pull request #160 from tqchen/master
     new c575347  Merge pull request #161 from antinucleon/master
     new 1ef70d7  [IO/MODEL] move reset to end of function, add numpy
     new ddd108e  Update submodule
     new c69d1f9  Merge pull request #162 from tqchen/master
     new 180ab22  Update CIFAR-100.ipynb
     new 4b95999  Update DMLC-CORE
     new e09c435  Merge pull request #163 from tqchen/master
     new 48f2617  [CTYPES] Fix python interface to be more consistent with C API
     new ce56e35  Merge pull request #164 from tqchen/master
     new 81761ee  [PY_METRIC] Make it consistent with numpy sklearn style, support numpy custom function
     new e6b1973  [IO] fix pad in numpy iter
     new 6d989e5  FIX
     new 555bc6b  Merge pull request #165 from tqchen/master
     new 9a74874  More robust im2rec
     new 3b514df  Merge branch 'master' of
     new a9ccf54  [python] add logloss metric
     new c320104  Merge remote-tracking branch 'upstream/master'
     new 9f0f7cd  Update .travis.yml
     new f92c496  [OSX] Make OSX Travis work
     new bbb45c2  Merge pull request #166 from dmlc/osx
     new 11de65d  [DOC] add program model discussion
     new e3c23a4  Merge branch 'master' of ssh://
     new 6b218f4  Merge pull request #167 from tqchen/master
     new d3129e7  Update
     new 7b6bf2a  Update
     new ec73910  Update
     new 676fd47  remove python error when path incorrect
     new 88a1bad  Merge pull request #168 from sneakerkg/master
     new c22c8fe  [PY] Add more executor functions to make it easy, improves the docs
     new dc50586  Minor updates on doc
     new e4193cf  Merge pull request #170 from tqchen/master
     new 3faa483  [Example] Inception-BN
     new f415404  Merge pull request #171 from antinucleon/master
     new 49d419e  Update
     new cf13244  [MODEL] Support update on kvstore
     new c85b4e1  Merge pull request #172 from tqchen/master
     new 06ad14d  [STORAGE] Change storage to pure interface, move lock to pooled manager
     new 6ea13e6  Merge pull request #173 from tqchen/master
     new a36338e  [MSHADOW] sync to head
     new cd142c4  Merge pull request #174 from tqchen/master
     new 138d058  remove logloss
     new 2e70300  merge
     new e373355  Merge pull request #175 from antinucleon/master
     new e335a26  auto select update on kvstore
     new f623a5b  REMOVE logloss content directly
     new 927c102  Merge pull request #176 from tqchen/master
     new a50027b  Update
     new cbe09c6  Update
     new b38e54f  Update
     new 5169414  Update
     new 25437dc  [DOC] fix dead link
     new 27f6552  Merge remote-tracking branch 'dmlc/master'
     new f42f2a0  Merge remote-tracking branch 'upstream/master'
     new aecef55  [DOC] Move figures to dmlc.github, fix typos
     new fc2f408  minor
     new a0b1ee7  Merge pull request #178 from tqchen/master
     new 943d6d9  [DOC] Add design overview
     new 63ba667  Add Symbolic in pics doc
     new 86405c2  minor
     new 1af6ba6  Merge pull request #179 from tqchen/master
     new cae71b4  Update
     new 0d1feb5  Update
     new ada0c29  Update
     new 10af9b9  [example] simple_bind
     new b3c0988  Update
     new cdeb822  Merge pull request #180 from antinucleon/master
     new 2358618  Make python package install more easily
     new 516a90a  Change NDArray function registry to function pointers
     new 7738279  Merge pull request #181 from tqchen/master
     new eaf8cc4  fix the license icon
     new 189fbc1  [PYTHON] update init
     new cf0a609  Merge pull request #187 from antinucleon/master
     new 3900530  [Python] Remove debug string
     new 1080e1f  Update
     new 32b04f6  fix doc
     new 2df2ce7  [EXAMPLE] make example run on any folder
     new 8b9e1c6  Merge pull request #190 from antinucleon/master
     new 6155880  Merge pull request #191 from tqchen/master
     new 3fafcd9  visual studio compatible macros
     new c8bacb1  Update
     new f4e9f8e  add contributor
     new d61e601  Merge pull request #193 from happynear/msvc
     new 5136c4f  typo
     new 18566dc  Update README
     new 145f5a3  Merge pull request #196 from antinucleon/master
     new dd8074e  [KVStore] BUGFIX inderminism
     new bbf5f31  [IO] Make IO deterministic
     new 97651f8  [KVStore] support new kvstore type, Change epoch_end_callback style to enable logging result periodically
     new 1484f78  FIX MNIST IO
     new a8e604f  Merge pull request #197 from tqchen/master
     new a13d97b  Update
     new 92c0202  Merge pull request #195 from donglixp/patch-1
     new 75e835b  [EXAMPLE] numpy
     new 87217f8  Merge pull request #200 from antinucleon/master
     new 6568864  initial empty commit
     new 1b1dcf9  MXNet.jl generated files.
     new 3e0abce  MXNet.jl regenerated files.
     new 8f38c7f  fix opencv local path
     new e25bf7e  basic API testing, create NDArray
     new 11029be  a placeholder README
     new 7c02f88  Merge pull request #201 from lipiji/master
     new 8fb8b6a  Update
     new 222c393  fix typo causing unary ndarray function being defined with fallback generic function
     new 38c5055  typo for a tuple
     new 41c1bb7  Merge pull request #202 from pluskid/master
     new 3b0afb8  update gitignore
     new 122dfc5  [kvstore] add ps-lite
     new b57e70f  [kvstore] kvstore_dist.h
     new 0fd2a3f  package skeleton. I need to think more before going on.
     new 2365d9d  Rcpp module seems a better idea
     new d7f011c  move gc to handle finalizer
     new 28531c8  define binary dn array functions
     new e426493  copy ndarray to cpu for debuggin
     new 0062c9d  define unary ndarray functions
     new 0ee41be  use a prefix instead of submodule for dynamically imported functions
     new f3184ad  a more general way of defining mx imported functions
     new e3a4119  only the generic way of defining function is enough
     new 307a8b7  [kvstore] push & pull for kvstore_dist
     new 99cafdc  [kvstore] add barrier on kvstore
     new 5b4039a  [ENV] Enable MXNET_EXEC_NUM_TEMP to control space replicas, update guidelines
     new 539cde1  it seems i need to get function pointers, but i can't. maybe i should try another method.
     new 61f3f15  [kvstore] more api in kvstore.h
     new 5ed258f  common array-like interface functions
     new a8cb0c7  Merge pull request #204 from tqchen/master
     new d0b4017  [kvstore] more
     new 59dcb7e  inplace operator for +
     new e276284  [kvstore] update
     new bf62571  weired segfault. Maybe it is better to use C API directly.
     new 6231434  [kvstore] server handle
     new 38dd7bd  fix stupid mistake and finally know what should be done
     new 24184b7  save and load
     new 97bfb3f  [NDArray] clip
     new c6c22f2  [kvstore] barrier
     new 92a83dc  Update engine NOTE
     new 82feff0  Merge pull request #209 from tqchen/master
     new c5e54d2  [kvstore] almost done c++ part
     new c22da90  Merge pull request #208 from antinucleon/master
     new 72dc8a8  [NDArray] dot
     new 565d26c  Update
     new c05a471  Update
     new 90e40e4  Merge pull request #210 from antinucleon/master
     new 514d417  [kvstore] start python interface
     new 678f6f2  basic test for ndarray
     new 525c5b2  comment out release testing until v0.4 is released
     new e0a0bbe  fix copy from and to CPU
     new c9a705f  Update
     new 9e139d7  Update
     new 6adc2e1  simplifies the title
     new e243343  Merge pull request #211 from tqchen/master
     new 11cd8f1  [kvstore] more on python interface
     new 39eaad9  [kvstore] bug fix
     new 8cd296f  [kvstore] bug fix
     new 451addc  Re-organize based on KK's code
     new a694501  [kvstore] split pull to several calls to engine
     new 39ac76b  Merge pull request #1 from tqchen/R
     new 9785c63  [kvstore] bug fix
     new 90b7370  [R] MINOR Change
     new 6294c93  ndarray plus and minus unit-tests
     new 113435e  __neg__ should return new array instead of changing self
     new 78cee9c  Merge pull request #213 from pluskid/master
     new 242c396  Update
     new 3d51d4b  Merge pull request #216 from dmlc/hotpxl-patch-1
     new 6781426  Merge pull request #2 from tqchen/R
     new c6d2a84  add cmake
     new 164e9e3  Merge branch 'master' of into addcmake
     new de32e1e  [Operator] Change Concat, add slice channel
     new 4c1126d  More formal cleanup according to Rcpp
     new d70b1dc  Merge pull request #215 from antinucleon/master
     new 0e91a89  fix warning
     new c13a7ca  Merge pull request #218 from antinucleon/master
     new 3569cc1  [kvstore] make all push and pull on the same key sequential on the single machine
     new 7d8559d  [kvstore] bug fix
     new e323ba3  [kvstore] add Wait and WaitAll in kvstore
     new 49070e9  [kvstore] bug fix
     new 5b6efe4  [kvstore] send command to servers
     new d145f56  [kvstore] try to send updater to servers
     new cce16b9  add lint in cmake
     new f2bd7df  [R] basic NDArray, maybe need change RObject to support operators
     new be460fb  Merge pull request #220 from tqchen/R
     new 6839852  Merge pull request #1 from dmlc/master
     new 5d9da0f  [R] allow operator overloading
     new 5ce672d  Merge pull request #221 from tqchen/R
     new 94206ae  Merge remote-tracking branch 'upstream/master'
     new 40391bd  update the const
     new d6a1367  use more meaningful name for sub! and add!
     new d33dff8  assignment operator for NDArray
     new b2acdc7  test ndarray elem-wise mul
     new 4667583  unittest ndarray div
     new 12ca85a  require julia v0.4 to run
     new 65cc788  basic symbol type
     new 4170e37  possibility to add dynamic doc string to imported functions
     new d84de29  Merge pull request #222 from thirdwing/master
     new 6490928  try setting up travis CI
     new 573eedd  fix travis script
     new b1ca062  more fix for travis script
     new 7a66ef4  enable osx build on travis
     new b22c83f  fix build script
     new f9b88b0  fix typo
     new bdd798a  properly switch directory
     new 7713589  fix sed option
     new 4a5d4f6  [R] Runtime auto load of libmxnet
     new c1cf178  Merge pull request #224 from tqchen/master
     new 8950df5  disable openmp for osx build
     new 2f4fb6d  [R] enable formals and move
     new 8ebe20b  Merge pull request #225 from tqchen/master
     new 694f0bb  add badges
     new d56aadd  [kvstore] clossure
     new e83622c  overload other operators
     new ed1bb2c  fix style
     new 32e5c15d one more style fix
     new 44d14b7  add as.array
     new 79ba32e  cleanup load/save works well
     new 3084cde  caffe converter
     new 2a4bcf7  [kvstore] ask the main thread to execute python codes
     new 6537294  Merge pull request #226 from hetong007/master
     new 4a3373f  [kvstore] python updater works now
     new e422085  Merge remote-tracking branch 'upstream/master'
     new 145ca5a  reorganize R script
     new 6acf718  clean up converter
     new 95aa981  Merge pull request #228 from thirdwing/master
     new 856c321  [kvstore] update doc
     new 51a1d97  [kvstore] update gitmodleus
     new 26ff9f0  [kvstore] update modules
     new ab3907a  adaptive data and output
     new a643a00  [kvstore] add ps-lite
     new 06be47a  converter readme
     new af896d1  update converter readme
     new 8cbe423  Merge branch 'master' into master
     new 349560e  Merge branch 'master' of into dmlc-master
     new 1475289  Merge branch 'dmlc-master'
     new 4c97284  [kvstore] update makefile
     new 0d594f9  update mshadow submodule ref
     new f9ced4f  [kvstore] fix lint
     new 2ed52c9  add tests that cannot pass yet
     new 3e04fb0  Merge pull request #229 from winstywang/master
     new 05f57c5  fix style
     new 7e3310d  fix #231
     new 39413ab  Merge pull request #233 from thirdwing/master
     new efb0b90  clean up built script
     new 0ac9e9f  [kvstore] make something in be private
     new 7e65723  [kvstore] fix
     new 7626fe7  list symbol creator functions
     new 33bb247  open test for matrix
     new 029e28b  [R] Add symbol, more todos
     new 84795cd  Merge pull request #234 from tqchen/master
     new fad9a2d  add tests that cannot pass yet
     new 46fd374  fix style
     new aa0c45e  open test for matrix
     new db61cd8  fix confliction
     new 8b696d5  [kvstore] fix for python 3
     new a940a77  cmake works on linux
     new 569cbf6  Update
     new ea257f8  Merge branch 'addcmake'
     new 72a1ead  fix lint
     new 6a9712a  Update
     new e0c22ce  Merge pull request #232 from hetong007/master
     new 627a8e5  name manager
     new 5fea004  symbol constructor
     new cf852bc  [kvstore] minor refactor
     new f4359ee  Merge branch 'master' of
     new 48120f1  [kvstore] add mxnet_ps_node.h
     new d57147d  [R] Almost finish the symbolic part
     new 773395e  Merge pull request #236 from tqchen/master
     new eac43b3  Merge pull request #230 from mli/master
     new fa77d17  [SUBMODULE] update
     new da6e967  Merge pull request #237 from tqchen/master
     new 2e6f833  [Symbol] fix slice
     new 05c9fd9  Merge remote-tracking branch 'upstream/master'
     new 44f039c  add test
     new e3c486a  Merge pull request #238 from antinucleon/master
     new ce68e59  symbol basic test
     new 8701e0a  symbol test_internal
     new 320706f  symbol test_compose
     new 16976cc    # This is a combination of 2 commits.
     new 0e1504c  [travis] learn
     new f724f6f  [TEMPFIX] Executor Remove deleter of executor o
     new 49bc380  [travis] move travis to tests/travis
     new 98e5cdf  [FIX] Executor deletion order
     new 272d1df  Merge pull request #239 from tqchen/master
     new fbf0251  [travis] update
     new f420b82  [make] fix make doc warnings
     new b1582d4  [travis] more
     new eb98667  [travis] update travis
     new 455dbe0  [travis] update
     new ddb0dd6  [R] most of symbol part done
     new 7b35de2  [travis] update
     new df3c879  Merge pull request #241 from thirdwing/master
     new dad98a2  [travis] hope to fix mac os x
     new 122f654  [travis] update
     new eb768c6  [make] delete lib and bin, try to fix travis python
     new 6638981  [travis] fix python
     new 67376a7  [travis] fix mac os x python
     new bcea942  [Rcpp] move and fix rcpp is
     new eef0401  [travis] mac os x again
     new c477ba9  [travis] update osx
     new 631a965  test infer shape
     new a9b6afe  Merge pull request #243 from tqchen/master
     new 56d4d5d  [travis] python
     new fca4789  test infer_shape with error
     new 09b9520  [travis] fix
     new 8060dd5  [travis] python
     new cd16d3c  [travis] fix cache
     new c05e6a5  try out julia v0.4 precompilation
     new dd1922d  executor
     new a1fde1c  [travis] update
     new 11818fe  [travis] update
     new 9b46790  [R] Finish basic executor, more todos
     new ea6e44e  [travis] remove
     new 0617541  [travis] install graphivz
     new 3e889a6  Merge pull request #246 from tqchen/master
     new 09a0e34  [travis] train by python3
     new 7572b3e  Merge pull request #247 from mli/master
     new dcca266  Update DESCRIPTION
     new 7994941  [R] Finish IO Module
     new 9390c72  Merge pull request #251 from tqchen/master
     new 148479b  [IO] add back num_parts and part_index for imgrec
     new 7fbdc3d  [python] support dist kvstore
     new 0a04f25  executor bind forward
     new 2074d12  executor backward
     new 2006810  [IO] multiple part support in mnist
     new e3353ab  [kvstore] add
     new 140f742  [R] Support basic kvstore
     new 43d6b48  [cmake] fix for MSVC
     new 812e503  Merge pull request #253 from tqchen/master
     new a9499ec  add kvstore balance greed alloc startry
     new 0973234  Merge pull request #254 from happynear/cmake-vs
     new f6edc38  Update
     new dbe316b  [R] Minor change, add default context
     new d8a619a  Merge pull request #257 from tqchen/master
     new 815cf30  [R] Random and initialization, need more docs
     new ab3621e  Merge pull request #258 from tqchen/master
     new f9fcf8d  [doc] multi node
     new a76453d  [doc] multi-node
     new facfcde  [doc] mn
     new 5b95f34  [doc] multip
     new 8151433  [doc] multi
     new 829f143  [doc] update
     new 591c99e  [doc] update
     new d6590b2  [doc] update
     new 5aa0598  Merge pull request #1 from dmlc/master
     new 4c47c27  [doc] update
     new 46ec30e  simple modification
     new 76c8724  [kvstore] refactor the kvstore types
     new d022489  [kvstore] lint
     new 782cd27  [doc] for multipe machines
     new 52b7cf1  [kvstore] bug fix
     new 709ad2b  [python] add cpu_pinned on context
     new 76420eb  add ps-lite
     new e7efd31  temp sync
     new 0ccab70  [kvstore] update
     new 49a0ff6  Merge pull request #263 from mli/master
     new a777402  modify tests and docs, have segfault
     new 0a285c7  simple modification
     new bbe25d5  temp sync
     new f5ca758  modify tests and docs, have segfault
     new b7c4401  Merge pull request #2 from dmlc/master
     new 2be7fd8  add checks
     new f3804ea  solve conflict
     new 150936d  pass test
     new aaf231f  Merge pull request #268 from hetong007/master
     new a5065f7  [python] add multi-node test
     new 6dc48bb  simple_bind
     new 9dd26e5  random module
     new 3fdf02a  [kvstore] more tests
     new 28f0155  [test] forgot
     new 357a105  default initializers
     new 27ca314  [test] for multi-device
     new db0bcd4  [test] multidevices
     new f664f9d  LSTM init
     new 0125990  Merge pull request #271 from mli/master
     new f58ae32  [R] Basic training of neuralnet
     new f79ed2b  Merge pull request #272 from tqchen/master
     new 11bae50  add kvstore balance greed alloc startry
     new 9d30698  fix
     new a1933c5  Merge branch 'master' of
     new 4f35791  [python] add multi-node test
     new 25f0067  [kvstore] more tests
     new 8764740  [test] forgot
     new 7c8cca4  [test] for multi-device
     new 52f56d9  [test] multidevices
     new bd84740  [R] Basic training of neuralnet
     new 9eb65c3  [doc] add visual studio build instruction, without zip archive.
     new b9bf958  merge
     new 2ad91f9  [kvstore] remove wait and wait_all
     new e76f29c  Fixed a typo
     new 4ce9672  Merge pull request #274 from loveisp/patch-1
     new f3e4e95  non-unicode name for initializer parameters
     new 2ffdfe6  [kvstore] update
     new c476b24  add kvstore balance greed alloc startry
     new 2567421  fix
     new 2f99bd4  refactor
     new 2eedd2d  Merge branch 'master' of
     new 2956764  [kvstore] dist_sync is done
     new fe076b2  [kvstore] fix lint
     new f044f0e  [R] Basic Model interface done, seems training is slow, might need to move some part back to Rcpp
     new 2df3c3c  Merge pull request #277 from tqchen/master
     new 67f3774  Merge pull request #276 from mli/master
     new 8fe0e05  fix typo
     new 6ffa1c0  fix a typo
     new b441134  Merge pull request #281 from hjk41/master
     new 18c746b  Merge branch 'master' into master
     new 5b4f976  Merge branch 'addcmake' of into yajiedesign-addcmake
     new 8bb8564  Merge branch 'yajiedesign-addcmake'
     new 5fcb4fa  Update submodule
     new 086612b  Merge pull request #280 from feiga/master
     new b47a0e9  Merge pull request #256 from junranhe/master
     new 1a0a970  Update
     new b14ae7b  [kvstore] replace ndarray by std::vector on server nodes
     new c388d51  [DOC] Fix typos and grammar errors in top level and first part
     new 0ee4f2c  Merge pull request #259 from bryandeng/note-engine
     new c1a3dae  [python] change the place of initialization optimizer
     new 4c336d2  [python] use batch_size * num_workers in for dist_sync
     new f35c021  [kvstore] test case for multi-machine
     new b33e283  [kvstore] more dist_sync test cases
     new d790b64  check writable bit in NDArray
     new 5f027ab  Merge branch 'master' into master
     new 7f35875  make clear the meaning of row-major vs. column-major in array conversion
     new 23c14fc  check writable bit in NDArrays
     new 7481dab  untested io.jl
     new e2a16e5  Merge pull request #285 from pluskid/master
     new 956a220  unittest MNIST data provider
     new d7cc10f  [R] Refactor NDArray to be simple S3, for speed reason
     new 6f38bcc  Merge pull request #289 from tqchen/master
     new 80ef4eb  Merge pull request #284 from mli/master
     new 7e75175  [R] make executor mutable, R now runs as fast as python
     new 3113256  kvstore init / push / pull
     new a405101  Merge pull request #290 from tqchen/master
     new 8e4102f  kvstore test aggregration
     new 6a38026  [R] KVStore, enable allreduce kvstore
     new 3661bbb  [KVStore] Add updater handle to kvstore C API, to enable R
     new 08ed562  Merge pull request #291 from tqchen/master
     new 73b1a23  [R] Make prediction ready, native to R
     new ef4a88f  Merge remote-tracking branch 'upstream/master'
     new c8a38ea  Merge pull request #292 from tqchen/master
     new ac4ebef  mxnet_generate
     new 875a5ee  Merge pull request #293 from thirdwing/master
     new 89211f7  fix submodule version
     new 8e2b1e3  Merge pull request #294 from tqchen/master
     new 48a2b20  eval metric
     new dd02638  shape should be translated between Julia<->C in infer_shape
     new c237ef0  [R] switch to roxygen, man is removed for now need to group by rdname
     new 4abcb95  Merge pull request #296 from tqchen/master
     new e656879  add begin_round set in checkpoint
     new dbada7a  some bug fix
     new bf3a121  fix some bug with vs2013 the lambda capture error and initialization
     new 4cb4da5  fix lint
     new ac7d504  fix lint
     new ea2221a  fix lint
     new 13d853b  fix some bug with lint
     new ebcd9a5  Merge pull request #299 from yajiedesign/bugfix-1
     new 3e7919c  Merge pull request #300 from yajiedesign/addcmake
     new 51911e4  [R] Update the R for roxygen
     new 23b5c56  Merge pull request #301 from tqchen/master
     new e9991d4  [R] Checkin Rd files
     new 7d9f781  Merge pull request #302 from tqchen/master
     new 37d7505  Add support for eval_data to be numpy.ndarray/NDArray/list pair. Add contributor.
     new fc47f72  [DOC] update guideline
     new 40cea29  Merge pull request #278 from II-Matto/master
     new d698185  Merge pull request #303 from tqchen/master
     new 4faab08  [DOC] contribute guideline
     new eeb0d5b  Merge pull request #304 from tqchen/master
     new ca2afb6  update the docstring
     new 7e4f28c  [EXECUTOR] Fix rare case when action ndarray is not allocated
     new 68c4dc0  [R] Add example of image classification
     new be6b77f  Merge pull request #305 from tqchen/master
     new dc4c066  Merge pull request #306 from tqchen/master
     new e42ab7a  Merge branch 'master' of into begin_round
     new 6fc3a01  [R,DOC] Enable Rmarkdown knitr
     new 015937b  Merge pull request #307 from tqchen/master
     new edc7e98  Merge branch 'master' into begin_round
     new 88b1c9d  [DOC] minor fix
     new 7975165  Merge pull request #308 from tqchen/master
     new 31e1a3e  Merge branch 'master' into begin_round
     new d061a18  optimizer and estimator
     new 8a0a43b  fix initializers
     new 92ec499  fix executor
     new 158b602  temp doc modification
     new 8c0e465  bug fix
     new 614de3a  mnist mlp example
     new 61fa00c  lenet example (not tested)
     new b2ab9e8  add two tutorials
     new 8cd7ce7  Merge pull request #310 from hetong007/master
     new 383fd98  fix kvstore closure
     new 53faab1  Merge branch 'master' into begin_round
     new bad46c1  Upgrade mshadow
     new 3ae0c80  keep a shared_ptr to objectpool from threaded-engine
     new eda0498  more shared pointers needed to keep from crashing
     new 2f3789a  remove debug prints
     new 6d8f72b  cleanup
     new 7f8612e  cifar10 example (pushing to GPU server for testing)
     new edf0871  working cifar10 example
     new d5ca8e0  Merge pull request #298 from yajiedesign/begin_round
     new 845c4d8  Merge pull request #311 from tqchen/master
     new f3b1165  fix lint error
     new 23aacc6  Merge pull request #312 from pluskid/master
     new 55948a7  [R] Fix the shuffle
     new ced5486  [DOC] Detail build info
     new dd37e67  Merge pull request #313 from tqchen/master
     new fa782b0  Merge pull request #4 from dmlc/master
     new 3737bbf  [OP] Add Regression operator for binary and linear regression
     new c0186b3  Merge pull request #315 from tqchen/master
     new 9c26111  add 5 min tutorial and fix docs
     new 2b08350  Merge pull request #5 from dmlc/master
     new 4e7dafc  wait for predict fixed
     new 92365d4  change cmake  out put name to libmexnet.dll
     new 0979504  Merge pull request #316 from yajiedesign/addcmake
     new 2896779  [R] Change License to BSD
     new f11c343  Merge pull request #317 from tqchen/master
     new 28deec0  Update
     new 87cd7b8  add one more doc, modify metric
     new 34857f3  Merge pull request #318 from hetong007/master
     new 6a71e36  update r logo
     new 685c6f1  update
     new 9c9a263  Merge pull request #319 from mli/master
     new 9f77880  [R, DOC] update rmarkdown doc
     new 7d54e1a  Merge pull request #321 from tqchen/master
     new 36a4ada  Update
     new fd1525d  Update
     new ab62f70  Merge pull request #6 from dmlc/master
     new b9363ba  fix docs
     new 1817d3e  Merge pull request #322 from hetong007/master
     new ad70dee  Update
     new bc0dd8d  Update
     new ddb7b4b  Update
     new d6dfa5e  Merge pull request #323 from XD-DENG/patch-1
     new 016e049  Update
     new dd0d703  [R] windows compatiblity
     new 8dce815  Merge pull request #326 from tqchen/master
     new 59e64d4  add regression metrics MAE and RMSLE
     new ed64233  Merge pull request #2 from dmlc/master
     new 97f5d4f  Merge pull request #327 from road2stat/master
     new 29c7b1c  Update
     new 31fb966  small typo
     new b969aad  Error in R install command
     new ac3cd26  Update
     new aa18e98  basic doc building setup
     new cf0e6fb  [kvstore] switch to mli/ps-lite
     new 849a6e1  Merge pull request #330 from pommedeterresautee/patch-1
     new ce42f16  [kvstore] refactor to new ps-lite
     new 55694dd  [kvstore] bug fix
     new ff055b5  test rtd hook
     new 4a25b00  add rtd badge
     new 34a1c89  use rtd theme
     new 4eff47d  fix some doc markdown formatting
     new a2f4750  [kvstore] bug fix
     new d13ddc6  clean up Context
     new 11f6922  [R] switch mem layout, and update examples
     new 598795a  [kvstore] refer to the dev branch of ps-lite
     new 70579c4  use light-weighted mkdocs for documentation instead
     new 0257055  one-sentence readme.
     new 62684d1  Merge pull request #333 from tqchen/master
     new a5f9b62  Merge pull request #334 from mli/master
     new 71f0906  [R] Add auto layout selection to detect layout for rowmajor in simple MLP case.
     new 6fb3d0e  Merge pull request #339 from tqchen/master
     new 1eb7ebc  use to simplify travis OSX build
     new 9223721  clean up CI build script
     new 7df27e0  Merge remote-tracking branch 'upstream/master'
     new 6c66a76  [kvstore] async test
     new bc89f4d  [kvstore] update ps-lite
     new 8e4a7d5  [NDARRAY] Add sqrt, square, choose_element operator
     new 349474d  LSTM base code
     new 0f4c6fb  Merge pull request #343 from tqchen/master
     new 12cb967  Merge remote-tracking branch 'upstream/master'
     new 280a423  [NDArray] Add onehot_encode
     new 28f7ee4  [kvstore] more test scripts
     new 90d3047  Merge pull request #345 from tqchen/master
     new 2c945ea  docstring for imported symbols
     new 32185b3  Merge remote-tracking branch 'upstream/master'
     new 1da331c  installation guide (or what we wanted to be...)
     new 684bad2  part of overview
     new d68fe53  Update LSTM
     new 6044c19  Merge pull request #342 from antinucleon/master
     new d96fd9e  Update char_lstm.ipynb
     new 306246b  Update
     new c9d1eab  fix doc typo
     new ce7f994  Merge pull request #351 from pluskid/master
     new e56e154  [OP, Refactor] Enable register mshadow unary op in only one line
     new 0dc9818  Merge pull request #346 from mli/master
     new df6613d  shape inference doc
     new 9e5de63  Merge pull request #350 from tqchen/master
     new 4ac5e7b  allow keyword arguments in bind.
     new af56456  part of the doc on NDArrays
     new 83e46c2  fix typo
     new 7159fdf  Merge pull request #353 from pluskid/master
     new 7d7e5c4  add installation guide for pre-built windows binary
     new cf963c1  fix dll name on windows
     new a29137a  fix windows linkage problem by adding dllexport
     new 708d4fc  disable ps-lite on windows for now
     new 5f3e9ac  add installation guide for pre-built windows binary
     new 52f11c3  fix dll name on windows
     new f606d86  Update char_lstm.ipynb
     new 7dcd8c4  change MXAPI to MXNET_API
     new 6c815c1  Merge branch 'master' into prebuiltdoc
     new 016c4ac  Merge branch 'master' of
     new 7d0f29d  fix windows linkage problem by adding dllexport
     new 3ec9344  change MXAPI to MXNET_API
     new a37f485  disable ps-lite on windows for now
     new 9828ad805 Merge pull request #356 from hjk41/prebuiltdoc
     new 69bbea3  Merge pull request #358 from hjk41/disableps
     new 6270fd1  Merge branch 'master' into dllexport
     new f672a9d  [OP] Allow register symbolic and ndarray unary operator in one place
     new 89dda75  minor
     new d01562a  Merge branch 'master' of into dllexport
     new 92557aa  Merge branch 'dllexport' of into dllexport
     new cf9a06a  getindex operator for NDArray
     new 3a26544  ndarray doc
     new 3e91cc9  Merge pull request #357 from hjk41/dllexport
     new 343d5c9  [OP] Fix reshape
     new de366df  Merge pull request #360 from tqchen/master
     new 8e797f7  Merge pull request #361 from antinucleon/master
     new 0c3d066  API variable -> Variable; group -> Group
     new 9665cb1  update doc: variable -> Variable
     new 210aba8  add libmxnet.dll into library search list
     new b1de0cb  required interface of optimizer should be present in the base class to avoid confusion
     new 478568c  doc for mnist
     new 5f18796  Use factory pattern to create optimzers. Creator tracks all (direct and indirect) subclasses of Optimizer by class name. We currently silently allow overriding existing names, but maybe giving a warning is better?
     new 65f4893  Merge
     new e4e58ab  Optimizer factory is now case insensitive
     new d3feb83  fixed lint error
     new 3273874  fixed python3 compatibility
     new 30bfd5f  fix more py3 compatibility...
     new ed9cd63  fixed more lint
     new 948af9e  [OP] update flatten
     new c0f4860  Merge remote-tracking branch 'upstream/master'
     new a962a9e  Merge pull request #367 from antinucleon/master
     new 637a1c2  refactored optimizer factory for py3 compatibility
     new bae6bf7  lint
     new d15bc77  add mx.chain macro
     new f14b670  Merge branch 'master' into master
     new e43f1e5  add compatibility for old style caffe prototxt
     new 2a05f0e  Merge branch 'master' of
     new e9088f5  rename init.jl -> base.jl
     new bed3a02  refactor optimizer options
     new a75a8ef  simplify mnist mlp example
     new bf6a604  [kvstore] async test
     new ef7212a  [kvstore] update ps-lite
     new 3a09872  [kvstore] more test scripts
     new 2ce65ed  [OP, Refactor] Enable register mshadow unary op in only one line
     new 0a6e761  required interface of optimizer should be present in the base class to avoid confusion
     new e1d4b6e  Use factory pattern to create optimzers. Creator tracks all (direct and indirect) subclasses of Optimizer by class name. We currently silently allow overriding existing names, but maybe giving a warning is better?
     new 2abcd07  fix typo
     new 30e995b  Update char_lstm.ipynb
     new a030aef  fix windows linkage problem by adding dllexport
     new 18adb83  change MXAPI to MXNET_API
     new 9369151  minor
     new 27d67e4  add installation guide for pre-built windows binary
     new 2bde47d  fix dll name on windows
     new 6929299  disable ps-lite on windows for now
     new b5596bf  [OP] Allow register symbolic and ndarray unary operator in one place
     new cd1fb67  [OP] Fix reshape
     new d54c277  Optimizer factory is now case insensitive. Changed to explicit registration because metaclass grammar is different between python 2 and python 3.
     new c512801  add compatibility for old style caffe prototxt
     new e4e7738  [OP] update flatten
     new 9ffe9fb  Merge branch 'master' of
     new 43b2486  clean up lenet example
     new 195586d  fixed naming
     new 00bcab3  Move the link to the figure to web-data
     new 514acea  Merge pull request #368 from tqchen/master
     new f0af4a7  Merge pull request #366 from piiswrong/master
     new 02e4af3  hygiene in chain macro
     new 1fb4946  use @chain in lenet example.
     new 52f02d9  [Example] Fast sign adversary
     new 3548a5f  lenet tutorial
     new 7c43f2c  update readme.
     new 483c32c  Merge pull request #370 from antinucleon/master
     new 3b43bcb  first version of deps/build.jl that is not working b/c opencv
     new ad385c0  Fix num_pad of io.NDArrayIter.
     new e77fdd0  default building without opencv.
     new 263e602  update install guide
     new cf84037  [EXAMPLE] check in memcost
     new 50134a1  Merge branch 'master' of ssh://
     new d7ed372  Merge pull request #372 from II-Matto/hotfix_io_pad
     new 3c4f424  Note memory
     new dc501d3  Merge pull request #373 from tqchen/master
     new 68a4b77  Update
     new 7934f67  add release notes for v0.0.1
     new 0b595e5  Add news about julia binding release
     new ff6db3c  Merge pull request #374 from pluskid/master
     new dbb94c9  Update
     new 0a32996  fix build script (don't exit Julia)
     new 48f4230  update news for bugfix v0.0.2
     new 5511fb4  Update
     new dc04567  Merge pull request #382 from turinglife/master
     new 8dea6d1  Update
     new 44f5a49  [MSHADOW] Upgrade get_with_shape
     new 6cc65db  [kvstore] bugfix for not allowing pull immediately after init
     new 6714558  Merge pull request #385 from mli/master
     new a4cc3f3  Merge pull request #383 from tqchen/master
     new 66bec9b  Update
     new 9a4484a  [API] Add predict API
     new a6bd422  a python script for generating image list required by im2rec and an option center crop that allow im2rec to output rectangular images
     new d5dd419  fixed get_with_shape calls to comply with changes in mshadow
     new b8e1085  Merge pull request #388 from tqchen/master
     new 7f63766  multi output softmax
     new 3a227b4  Merge pull request #387 from piiswrong/master
     new 18ab5d9  proper debug flag (-g -G) for nvcc
     new 119fac1  update kvstore doc.
     new d738fba  NDArray save and load API
     new 4e5e152  gradient clipping
     new 511c041  also do grad clipping when momentum == 0
     new 9d138da  save and load for symbol
     new 30f89a8  unit test for symbol save and load
     new 86b2c12  multi softmax fix and test case
     new 7e86c8f  improve defstruct macro
     new 00bd457  callback interface
     new 06879c6  fix estimator API
     new f551c91  add speedometer to cifar10
     new cb067c3  fix speedometer
     new f7d7ce3  save checkpoints in callbacks
     new 3301c27  Merge pull request #393 from pluskid/master
     new 1d83343  Merge pull request #392 from piiswrong/master
     new 1e2bd62  [Python] Fix numpy IO
     new 5ea6b8c  array iterator and (part of) the unittest
     new 9e243b1  Merge pull request #390 from antinucleon/master
     new 7bcbf24  [example] Move
     new 5730e49  [DOC] Add imagenet full
     new 92a0d25  Merge pull request #395 from antinucleon/master
     new c39de26  Update
     new b4cc05c  more test on array data iterator
     new 2c9d9ae  more tests on array data provider
     new dc8ca18  fix README example
     new eb84402  test shuffling of array data provider
     new 1538340  rename estimator -> model
     new 2a2b425  Merge pull request #6 from yeesian/patch-1
     new 6d87215  Merge branch 'master' of git://
     new 28791db  use joinpath in README
     new 1baeabc  expose init_model API
     new 2298858  fix compile issue in VS
     new f3c6299  merge origin master
     new 2e73ea9  Merge branch 'dllexport'
     new 9af3181  get rid of __cdecl
     new 7fa1511  model predict API and demo
     new 99fb2fc  Merge pull request #396 from hjk41/master
     new f1ac64d  add ps-lite in cmake with ps-lite in dev branch
     new bb4620e  model checkpoint loading
     new f8a01bf  allow callbacks in predict
     new 0f632cc  add pretrained Inception model
     new 5bcd859  add example using pre-trained model to do prediction.
     new 5a80bf3  fix #8, more informative error when libmxnet not found
     new 8bb82e4  use sphinx as documentation system
     new bcc90ef  switch from mkdocs to sphinx
     new 8d56a9f  a dead simple embedded sphinx doc system (cf #3)
     new 549bd78  [OP] Add norm
     new f1f869f  API document for NDArrays
     new 7033d17  Update mshadow
     new ac39fd6  libmxnet API for ndarrays
     new e7ec661  automatic build API for symbols
     new e13dbd3  fix
     new 9c87c61  fix mal-formatted param type in sphinx syntax
     new 091c234  add title to symbol doc page
     new 54e621e  prepare for v0.0.3
     new db1a5b5  Merge pull request #405 from tqchen/master
     new 87c2492  add exp, log for ndarray
     new cadee3a  better type annotation in docs.
     new 891f4f5  change to UseIN_
     new 1b62ddf  more docs.
     new d19f28f  docs for built-in callbacks
     new e88187b  fix lint issue
     new 93a036d  UseOut for exp, fix grad for log
     new 064cab6  Add Docker images to installation docs
     new 7542da6  [OP] update convolution stride
     new d34a7be  delete exp_grad
     new 58fd8e6  Merge pull request #412 from antinucleon/master
     new 7b55bb4  Merge pull request #410 from Kublai-Jing/master
     new 053ba92  Delete
     new ef16e8a  add namespace
     new 0ebe2a1  Merge branch 'master' into master
     new c051c2a  fixed joinpath problem in example in
     new de566eb  Merge pull request #9 from cbecker/master
     new 0132f19  added ADAM SGD
     new c29e18a  Merge pull request #10 from cbecker/master
     new 6e0c5c7  some doc for symbol
     new 069a6ac  fix some complie bug with ps-lite in windows
     new 07fb686  Merge branch 'master' of into addcmake
     new 975ece6  fix lint
     new d06a672  Merge pull request #409 from Kaixhin/master
     new 84405e6  [Name] Convention
     new ce26751  fix bug with lambda
     new 762c45e  refactoring and naming convention dmlc/mxnet#407
     new a2beae8  refactoring of optimizer
     new fe36340  refactor adam optimizer
     new 1f39e1d  more learning rate scheduler
     new b4007a1  Merge pull request #415 from antinucleon/min-net
     new e821f95  update equation in sphinx doc
     new 85225c8  [API]  make predict more standalone without relying on filesys
     new 04bde14  fix accidental notebook changes
     new ed1aa84  [OP] Fix Operator norm
     new 26da63b  Merge pull request #418 from tqchen/master
     new dcb2a1e  [EXECUTOR] Fix destructor
     new 4aa1b8d  Merge pull request #419 from tqchen/master
     new 40600f1  quality option for im2rec
     new d8129fd  python interface for RecordIO
     new 052471e  move recordio interface to
     new 7fb08dc  lint fix
     new 6119a82  Merge branch 'master' into addcmake
     new 4393a12  Add deconvolution operator, and the corressing implement using cudnn_v3
     new ee9150c  update name
     new 1dc944e  first step of refining the data provider API.
     new fb77af2  Merge pull request #408 from piiswrong/master
     new b2a1d13  [Example] fix convention
     new 49a5acb  import doc of built-in data provider.
     new 76fee6b  refine mxdataiter api
     new 39d5f36  Merge pull request #425 from antinucleon/min-net
     new 8d0ef5d  Merge pull request #7 from dmlc/master
     new a1f8ce7  fix epoch and batch naming
     new 0da58d1  Merge pull request #428 from hetong007/master
     new 5e03d27  fix unittest for new data IO api
     new ceaa3eb  fix mlp example to use the updated data io API
     new 993be59  simplify model a bit with data IO api
     new 64d26e9  Merge branch 'master' into addcmake
     new f30aedc  part of the executor API doc
     new 169a05f  Merge remote-tracking branch 'mxnet/master'
     new 62feefb  Merge branch 'master' of
     new 5611825  fix lint
     new 7c93e2f  Update
     new 1ceeb10  update iteration to epoch
     new bdde0f5  Merge branch 'master' into master
     new cee2587  update Convention
     new 4b3f3fc  Merge remote-tracking branch 'mxnet/master'
     new 2128250  Merge branch 'master' of
     new d48dd6f  fix lint
     new 2be14f7  Merge pull request #399 from yajiedesign/addcmake
     new 9d743aa  fix num_round to num_epoch
     new 388e91d  fix mnist doc links
     new c2727a2  Merge pull request #422 from tornadomeet/master
     new ace906c  Isolate Naive Engine for prediction only
     new 8e9d08c  Update
     new 4810f98  Merge pull request #438 from tornadomeet/patch-1
     new 3974bca  Merge pull request #437 from antinucleon/min-net
     new ae9f202  Update
     new b0909a7  Update
     new 724de45  [kvstore] fix compile warnings
     new c35f6ff  python operator
     new 93732fc  osx compatibility fix
     new 3e64fde  rename pythonop to nativeop(c++side) and numpyop(python side)
     new 3dc32a9  fix
     new 9dc3e5d  numpyop based softmax example
     new c07f52c  lint
     new f699bd2  Merge pull request #436 from piiswrong/master
     new 1bb9d24  Update git
     new 4cb754c  Add viz function
     new 526e634  [OP] lift restrictions of slice of concat and slice channel
     new 1fc8885  Merge pull request #441 from tqchen/master
     new 5c2408c  disable the PS build in windows by default
     new 99dc8e5  Merge pull request #442 from yajiedesign/addcmake
     new c4398b5  Update the choose_element api
     new d9523b3  add test
     new 2badb2a  Merge pull request #443 from tqchen/master
     new dced853  [Rename] Softmax->SoftmaxOutput
     new 137ff4c  Merge pull request #444 from antinucleon/min-net
     new edad1d8  Rename
     new 7d7bdc8  Merge pull request #445 from antinucleon/min-net
     new 7db0bf8  add amalgamation
     new 5d8b489  remove cp
     new 023f81d  fix MXNET_EXTERN_C undefined
     new 4f4fec2  Merge pull request #448 from jdeng/predict_api_fix
     new b8a6302  Merge pull request #446 from jdeng/amalgamation
     new 9a90aea  [R] Update terminlogy
     new 274ca31  Merge pull request #450 from tqchen/master
     new 5005252  Update
     new e24f3bb  Merge pull request #451 from tqchen/master
     new dba8e5a  [AMALGAMATION] Cleanup
     new 7054c23  Update
     new 5e895d4  [OPERATOR] fix leaky relu
     new ee9db70  Merge pull request #452 from antinucleon/min-net
     new f7d17f0  Merge remote-tracking branch 'upstream/master'
     new 87a5def  fix bugs in function calls
     new 4152164  Merge pull request #453 from pommedeterresautee/master
     new eae2511  Update
     new 3044377  [AMALGAMATION] Allow a very slow mininum build with no dep
     new bdca8e9  Merge pull request #454 from tqchen/master
     new 9e35598  Fix Warning
     new 68a25b2  Merge pull request #455 from tqchen/master
     new fb350d4  [R]Added basic model test
     new bdc52e4  [R] set up Travis tasks for R tests (+1 squashed commit) Squashed commits: [dc52eb2] t (+12 squashed commits) Squashed commits: [266218e] test [2ea3089] fixes [b53c301] test [4def042] Added Rcpp [974a111] test [31d674c] test [c649a59] test [80950a6] test [9dd63a1] test [824100d] test [f6994d4] test [264d388] Set up R tests on Travis
     new 075ca59  Added and R lib dependencies(+5 squashed commits) Squashed commits: [796abcb] Added to print out the failure log (+1 squashed commit) Squashed commits: [b7dd59f] Added to print out the failure log (+5 squashed commits) Squashed commits: [89349f0] fixed dependencies [3302caf] Fix namespace [8bb3182] try to fix Travis R install [9987ef9] Added sh to print out Tt Travis failure log [72fc6e2] Detailed Travis failure log [fc0 [...]
     new 9832672  Merge pull request #457 from terrytangyuan/test
     new 4790171  travis
     new 35c53f6  Merge pull request #8 from dmlc/master
     new 5e49eb9  add mlp interface
     new f66c18b  pass R unit test
     new df71dc7  fix namespace
     new e9b55fa  Merge pull request #462 from thirdwing/master
     new 1d17181  numpy op example readme
     new 8dab2af  update dataiter to support any number of source arrays
     new f789dc8  add hard_reset for ndarrayiter
     new d007fc5  use roll_over for training
     new 8eef36d  fix dropout and softmaxOutput
     new 319c5f3  rename Softmax -> SoftmaxOutput (mxnet/mxnet#434)
     new 6389094  update doc Softmax -> SoftmaxOutput (dmlc/mxnet#434)
     new e937a51  [R] minor change
     new 922fc85  Reenable 0.4 testing.
     new 4acbce9  Add coverage.
     new ebb573a  Merge pull request #13 from invenia/coverage
     new dc775e0  Merge branch 'master' into master
     new 98b68b3  Merge pull request #466 from thirdwing/master
     new 6b920f7  Merge branch 'master' into master
     new b1c7e69  Merge pull request #460 from hetong007/master
     new f676211  [example] num_round -> num_epoch
     new a989950  remove constraint on num_arg in elementwise_sum
     new 381ea4c  Merge pull request #469 from Kublai-Jing/master
     new 5f21337  Merge pull request #456 from piiswrong/master
     new 428ed60  Merge pull request #9 from dmlc/master
     new 2feb6c5  add mx.mlp to tutorial
     new e4ccd2b  Merge pull request #471 from hetong007/master
     new c1d663d  run mlp.jl on travis CI
     new 3d447a4  Merge branch 'master' of git://
     new b1e882b  fix env variable for detecting Travis CI
     new f823510  only run 1 epoch for MNIST on travis CI
     new 23378f1  fix error due to API changes of subarray in Julia nightly
     new d26c36b  a specific mlp script for Travis CI regression test.
     new 1786bdb  wrap mnist test into a module
     new 3bf4364  fix travis test error
     new 7208a3d  regression_output now can have arbitrary output size
     new 5dd2cae  refine tutorial
     new 53b16a3  Merge pull request #10 from dmlc/master
     new d0dadad  pull-back only when update_on_kvstore
     new 069a570  [multi-node] results
     new 75b3e68  SVMLight Data Provider
     new f6cbba1  add description of SVMLight Provider
     new 0f22cfa  add readme for plugins
     new df9637f  regression output label shape compatibility
     new 3739e4f  Merge pull request #474 from piiswrong/master
     new 1ca010f  update mistake
     new f8e6127  Merge pull request #481 from tornadomeet/update-mistake
     new 76286a6  Merge pull request #475 from hetong007/master
     new ab78837  Update
     new a6ce993  fix model.FeedForward with Optimizer param
     new 39a35d0  modify MXDataIter to to provide data/label info
     new 27dda3a  modify model for extended data API (un-tested)
     new ca95176  fix model for MNIST MLP example
     new 671e332  a working MNIST MLP training example
     new 677a705  clean up and fix for update_on_kvstore
     new a73f40b  prediction API to use extended dataiter API
     new a0cf820  NDArrayIter and mlp_numpy
     new ddcff58  recover fit API
     new 10828f8  fix lint errors
     new 77b22ef  fix original interface of model
     new dfc293a  clean up
     new b260c69  fix examples for updated data API
     new 7b31c93  fix lint
     new 9178a3c  io unit-test
     new b93dd83  fix io unit-test
     new bdbac7f  fix unit-test (different version of python)
     new 0297e6a  modified gradient clip norm & learning rate policy
     new 9e8fc7e  fix unit-test error (python version)
     new e00d03f  update Makefile
     new 311b93a  update travis
     new b7f33e7  [example] update dist results
     new 67a36e9  [example] udpate dist exp
     new 3cf82af  Merge pull request #483 from chenjx1005/optimizer
     new 119be0f  Update
     new fa28d03  fix
     new 03e3811  Merge branch 'master' into multi-data
     new dee85d9  clean up
     new 8b72a8c  fix regression tests
     new 72dfafb  Merge branch 'master' into master
     new dee39a8  fix lint on
     new 3ae790c  fix python 3 compatibility
     new 9e5e81c  Merge branch 'master' of into dmlc-master
     new aff6b56  Merge branch 'dmlc-master'
     new af09198  Merge pull request #485 from thirdwing/master
     new 3ca13be  [kvstore] use static_cast
     new 5ebb98d  [python] add an optional epoch_size in for synced kvstore
     new 9192375  Merge branch 'master' into multi-data
     new f47a7f9  Merge pull request #486 from mli/master
     new fa3e200  Merge branch 'master' into multi-data
     new 931912c  [kvstore] bug fix
     new d10344d  [example] more distributed examples
     new e1875da  [example] update
     new 9de25b5  add payload void* pointer to NativeOp
     new ea9b1a0  change backend in numpy_softmax to cpu
     new b93c146  Merge pull request #490 from vchuravy/vc/jl_callbacks
     new 65eed0b  [example] update
     new deba1bd  [example] update
     new b36d956  [test] update testcases
     new 30a84ece Merge pull request #493 from mli/master
     new 42a46fd  Update
     new 7ed591d  add libatlas-base for cblas
     new b1e9c33  modified gradient clip norm
     new 431098e  Merge branch 'master' into rnn
     new 7e89a78  Merge pull request #17 from vchuravy/vc/travis
     new ee855ec  fix cifar10 example
     new 08d9f2a  python: Fix assert
     new ba76e4c  [python] support float epoch_size
     new dda8e83  [example] udpate
     new d39758c  [example] update
     new 54f839c  [example] update
     new 2d68b52  add tools/
     new e5d96c2  update building instruction for R package. fix #500
     new f0f89d2  [example] update
     new 488feec  Merge pull request #501 from thirdwing/master
     new 2156898  fix deconvolution kernel size check
     new f71f1b1  add enable image index
     new 0bbff0c  add index
     new 97b3d09  fix lint
     new 90d0139  Merge pull request #497 from ry/ry
     new b3af9df  Merge pull request #503 from tornadomeet/fix-deconv
     new 19a3389  Update
     new a4a1922  Merge pull request #484 from yang1fan2/rnn
     new 176db28  modified gradient clip norm & learning rate policy
     new 07b83de  modified gradient clip norm
     new f9834be  native_op shape check
     new 3fc2c2c  mlp_numpy example bug fix
     new 202671a  [example] update
     new baa30a2  Merge pull request #505 from piiswrong/master
     new 691d45e  [example] update
     new 2014f8a  Merge branch 'master' into multi-data
     new b387195  recover from some ancient commit
     new bbf1d00  remove explicit call to reset in mxdataiter
     new 37e78d1  [ps-lite] link libprotobuf.a and libzmq.a
     new 01fe0f9  [ps-lite] tiny fix in makefile
     new 62da23b  recover epoch_size commit
     new 6cc3ee3  fix epoch_size loop
     new 7a4dde0  [doc] dist train
     new a1a6f65  Merge pull request #468 from pluskid/multi-data
     new b687fa1  [tools] update
     new e60cf95  [makefile] link tracker
     new b9df59d  [exmaple] update
     new 015cf61  Merge branch 'master' of
     new a8bafe8  individual learning rate multipler for each parameter in optimizers
     new 729e2b5  auto encoder example
     new 169a686  lr mult update
     new 6a9d9c9  [python] changes the order of data.reset() to fix a strange bug when reading a big s3 file
     new 726badd  update doc for updated data iter API
     new 8486b97  Merge branch 'master' of git://
     new 637bf0c  avoid calling mxdataiter reset
     new 91e5c9c  [doc] more doc
     new 1c71a1d  python3 fix
     new 0c29661  Merge pull request #506 from piiswrong/master
     new b8877ad  add ScalarOp
     new 2af8d5b  Merge branch 'master' of into scalar_symbol
     new 224e88a  fix lint
     new ee0583f  remove unused variable
     new 0d61c21  add example of Graph viz to R documentation
     new 7b15a87  Merge pull request #499 from pommedeterresautee/master
     new 1c3c598  numpy op doc
     new f00b520  link
     new 3e5128c  link
     new fc54d6e  Merge pull request #515 from piiswrong/master
     new c06562c  Update
     new 98b8c9f  del old
     new d8cb825  add enable image index
     new 278dfad  add index
     new 9505323  fix lint
     new a84c749  [doc] update
     new 6341fac  [doc] update
     new f34044c  [doc] update readme
     new f0912a8  Merge branch 'master' of into dmlc-master
     new 35bb719  update dmlc-core
     new b35f796  [python] fixbug for tiny epoch_size
     new 4efe0e3  Merge branch 'dmlc-master'
     new 315c1a4  simplify a bit io test
     new 580c14e  add swapaxis operator into mxnet
     new 722eb36  fix array data iter provide size not properly reported (see #14)
     new ff5a1ff  puth using namespace statement into function.
     new 25c0f19  change struct to namespace.
     new edd9b9c  change c_api index interface
     new b3249d4  Merge branch 'master' of
     new 07a0de7  fix lint
     new f8187a6  fix coredown
     new 33b0d03  fix bug with inst_index is null
     new 652e3b2  fix io index
     new 189e755  prepare for v0.0.4
     new de8d9d1  add ElementwiseBinaryScalarOpProp
     new cec2de4  Merge branch 'master' of into scalar_symbol
     new d9979ea  fix dmlc/mxnet#521
     new 58cf383  Merge pull request #525 from pluskid/master
     new 4f27fae  [doc] update
     new 76261c7  refactoring: Symbol -> Node
     new 955b6d2  refactor: Node -> SymbolicNode
     new aab8784  upate doc for refactoring
     new c13f88f  add nn-factory
     new 3bc475a  add doc for nn factory
     new 41c20ae  fix travis CI build error due to refactoring
     new 3e8627d  document clean up
     new be8cbda  more test for IO
     new 4348575  Merge branch 'master' of into scalar_symbol
     new b49dc62  update doc (Symbol -> SymbolicNode)
     new 0bbc9ea  lint fixed
     new 0c2d228  Merge branch 'master' into master
     new 9d1abbf  Merge pull request #519 from starimpact/master
     new e986468  [example] bug fix
     new a059d6a  checkout Makefile
     new fa0e313  Merge branch 'master' of
     new 650e839  [doc] replace by two cp
     new e659062  [example] partition validation data for imagenet in distributed example
     new 91a5703  [doc] update
     new 503443f  add enable image index
     new 775d97b  add index
     new 72d21ad  fix lint
     new e175967  [doc] update
     new 120cf4b  add enable image index
     new 7df73de  add index
     new 318ca2a  fix lint
     new c17affd  change c_api index interface
     new baf3684  fix lint
     new 909ca3e  fix coredown
     new 2fc16a3  fix bug with inst_index is null
     new 04a9a79  fix io index
     new 6492418  Merge branch 'master' of
     new 90b034d  Android build
     new 4ca3d1b  [doc] update
     new 5c4922e  [doc] update
     new f6a951f  [doc] update
     new 591871c  add doc for metrics
     new 2b85c9c  share NDArray data with Julia Array.
     new c2200be  add test
     new 1f8a0a0  Merge branch 'master' of into scalar_symbol
     new 86129f7  small fix
     new 2055727  nd_as_jl macro
     new 1722b49  Update
     new ea90b55  use nd_as_jl to re-write accuracy
     new b1540ea  Merge pull request #526 from Iroul/master
     new 7773549  Update
     new b7a7487  Update
     new fa83a5b  autoencoder example update
     new beee0f0  [doc] remove kvstore impl
     new 406b84c  Merge pull request #527 from mli/master
     new aad61c5  Merge branch 'master' of
     new 5447c78  [example] add model checkpoint and better log format for distributed train
     new 65e7cd5  Update
     new acd8bc4  Update
     new 1e31689  Update
     new 965a7e6  Update
     new 4701568  some change
     new f1ac7cd  Merge branch 'master' of into scalar_symbol
     new 7ab8c33  fix python
     new d637439  move
     new df538d4  Merge pull request #534 from antinucleon/master
     new 4c1d325  Update predict-with-pretrained-model.ipynb
     new b5c0d96  reformultate accuracy with multi_output in mind
     new 8d282fe  Merge branch 'master' of into scalar_symbol
     new 69ffc2f  add doc
     new 323d1d1  Update
     new fb9d496  Update
     new 102e7cf  update news
     new c9b3248  Merge branch 'master' of
     new 1e62913  update
     new 8322b63  Calculate accuracy based on size of pred.
     new 7e0ec56  input from image
     new ec344c0  fix exception
     new edada8e  fix image mean type
     new bd71565  fix exception name
     new c300971  remove sub from accuracy
     new d51d2af  add comments for accuracy and rebuild docs
     new d160624  update
     new 689d04c  Merge pull request #536 from antinucleon/master
     new 3c4d006  Update
     new bc935ca  Merge branch 'master' of into scalar_symbol
     new 8f7a2e0  updata test_scalarop
     new 210d2d8  fix example predict path changed
     new 7947313  Merge pull request #23 from vchuravy/vc/multiout_acc
     new 4362525  add nd_as_jl unit test
     new e884692  update ipython notebook BGR -> RGB
     new 4ea9188  update IO doc
     new 593dc44  Merge pull request #543 from pluskid/master
     new 55acccd  Merge pull request #537 from jdeng/master
     new 05ee640  Merge pull request #541 from isnowfy/master
     new 07e46c5  Merge pull request #514 from yajiedesign/scalar_symbol
     new fd4b034  Update
     new 30dd954  Update
     new 6b68854  Update
     new b0c34dc  Update
     new 8013c79  Update
     new 513da47  Add waitall
     new 8c66622  Merge pull request #548 from antinucleon/min-net
     new 4511d84  fix CustomMetric
     new fa5976d  regression label shape fix
     new 292b839  autoencoder example update
     new eb1e347  Merge pull request #550 from piiswrong/master
     new b8a684c  Merge remote-tracking branch 'upstream/master'
     new 3226713  complete doc for symbolic-node
     new e9a6477  Fixed the trailing L problem of hex(...)
     new 20f905c  fix num_round in imagenet
     new dc2f5c5  fix bn momentum default value
     new d3bd2d5  Merge pull request #504 from junranhe/master
     new f5f2777  Adds Mean Squared Error evaluation metric
     new 94e17a6  Merge pull request #25 from Andy-P/pull-request/f5f27779
     new d5d1d72  Merge pull request #555 from sxjscience/pr_branch
     new 0b538d1  make notes on MSE metrics
     new 645091c  char-ltsm temp commit
     new 5ee654d  Inconsistency for dimension between R and backend. fix #544
     new 703d78f  char-ltsm example, not tested
     new eea7d2c  typo ltsm -> lstm
     new e35b929  fix macro error reporting
     new 5a9d56a  fix some errors
     new 487d5e8  fix kvstore refactoring bug
     new e4b6306  fix databatch
     new 7fa8802  Merge pull request #560 from thirdwing/master
     new 6ddbecd  add speedometer
     new 00c7bff  neg-log-likilihood eval metric
     new d685f77  tweak training log formatting
     new 69cc430  reset! for NLL calculation
     new 42b7c82  Revert "Merge pull request #560 from thirdwing/master"
     new c5c1e05  basic network visualization
     new e0857a2  add visualization doc
     new c4abb97  fix visualization bugs
     new a9f4a91  Merge pull request #563 from thirdwing/master
     new 0e1515d  fix lstm learning rate
     new c3bee16  Merge branch 'master' of
     new 6193ef4  add lr decay to lstm training
     new 7159e79  [OP] Special fix regression for backward compatiblity
     new 7e8ad14  Add Gitter badge
     new 585b907  Merge pull request #565 from gitter-badger/gitter-badge
     new 7cab461  Update
     new d998be8  Add uneven_load training feature
     new f3467cc  Merge pull request #564 from tqchen/master
     new b129ced  Merge branch 'master' into master
     new b078fec  lstm sampler (not tested)
     new 0c3f32d  tweak learning rate
     new b624089  Merge branch 'master' of
     new 6770702  Fix lint 'line-too-long' problem
     new 5e7df90  Merge branch 'master' of
     new 0f1fbaa  fix sampler
     new 6db1aae  Update
     new a6886c2  copy states over for sampler
     new f915865  use adam optimizer
     new a153c3c  Merge branch 'master' of
     new db7ae38  fix LSTM sampler
     new 386990a  Merge remote-tracking branch 'upstream/master'
     new ce95a6f  Remove the hack from and add work_load_list to
     new 8829912  add basic readme for lstm example
     new 5da45f3  use a more sensible default argument in favor of #24
     new 3bc4fb5  fix path in char-lstm example outputs
     new 0e6402d  add simple visualization example of char-lstm
     new a392e04  tweak visualization
     new b1f036d  fix visualization bug
     new 6e385b6  Merge pull request #549 from winstywang/master
     new 3c6f315  Update
     new 3aad289  adding data loading design doc
     new 75c32cd  Add gitter link to dmlc/mxnet
     new f9852fd  Merge branch 'master' into master
     new 5b9df89  update README for LSTM example.
     new a9242f4  Merge branch 'master' of
     new 7c5feb2  [R] update the testing scripts, so the code in vignettes will be checked on travis
     new a0a850a  Merge pull request #566 from XericZephyr/master
     new 52c2dfe  Update
     new 293fc4c  Merge branch 'master' into master
     new 4cf0d21  Merge pull request #569 from pluskid/master
     new a503f4b  fix mx.nd.load
     new 8192673  fix doc
     new 0494f64  html check
     new 9a35c61  Merge remote-tracking branch 'upstream/master'
     new 168a713  improve error message and comments
     new 0e680e0  prevent workspace from being bigger than anticipated
     new a2b1e2b  adding myself to contributors
     new cc649ef  Merge branch 'master' into master
     new 30e930e  Merge pull request #577 from sneakerkg/master
     new dc732b4  Merge pull request #574 from thirdwing/master
     new 632fb52  Merge pull request #567 from vchuravy/vc/fix_conv
     new ae4583b  Update
     new 3bc1e5d  Merge remote-tracking branch 'refs/remotes/dmlc/master'
     new 3230734  Merge remote-tracking branch 'refs/remotes/dmlc/master'
     new 0c4ae7b  tune the data loading
     new 53fc4f7  Merge pull request #580 from tqchen/master
     new a4772da  Update
     new 1305fe3  [R] mx.model.FeedForward.create can accept R list as validation set
     new 1bd4619  Merge branch 'master' into master
     new fa58a39  update concat
     new bd397e5  Merge pull request #582 from antinucleon/min-net
     new 5c72688  Merge branch 'master' into master
     new 379de2b  part of LSTM tutorial
     new b2b903c  Merge branch 'master' of git://
     new 7d8844c  Merge pull request #581 from thirdwing/master
     new c804be9  char-LSTM tutorial
     new b595394  call visualization in test
     new f099670  prepare for v0.0.5
     new ce09b1f  Added power
     new 37b48f1  fix the sample code in README
     new a35c709  Merge pull request #27 from bicycle1885/patch-1
     new a2f6e5f  more glossy intro
     new a9174be  Merge pull request #585 from tqchen/master
     new b5ad9ea  Update
     new 648b7a6  Merge pull request #584 from lukemetz/lm/add_pow
     new f3ece7e  Update
     new 1b8ca3c  numpyop garbage collection fix
     new d6575f8  Merge branch 'master' into master
     new e8ab01d  tweak visualization style.
     new c2d4b1e  Merge branch 'master' of git://
     new aea096f  use a smaller network for visualization.
     new 4850e7b  save training/evaluation metric during training. close #511
     new d3e39b0  Add multi-label usage and fix num_round typo in API doc
     new baf50d3  Merge pull request #590 from thirdwing/master
     new a6b4baf  Merge pull request #591 from XericZephyr/master
     new 4c9915e  Merge branch 'master' into master
     new bf988bf  Merge pull request #587 from piiswrong/master
     new 5fc1920  [R] update for callback functions and doc
     new 4f09440  implements variants of Xaiver
     new ae19877  Update
     new 56bd4f7  Merge pull request #28 from vchuravy/vc/fix_xavier
     new 89cf70d  update doc
     new 9c6a6a0  Merge branch 'master' into master
     new 0740bac  Merge pull request #596 from thirdwing/master
     new 1a935de  rename Xaiver to Xavier
     new 5dd8664  rebuild docs
     new 6e791eb  fix backward no aux states
     new 0b38010  Merge pull request #30 from vchuravy/vc/rename
     new be16edb  fixed paths
     new 6f03ea2  Update
     new d3612d7  Merge pull request #607 from zygmuntz/master
     new 515642e  Merge pull request #608 from mli/master
     new e263936  [example] refactor image
     new 0a7efc5  [python] add batch_end_callback to create()
     new 61894ef  [example] mnist
     new 989042a  [example] cifar
     new 6bb3623  fix weight initialization for new layers in finetuning
     new 0a1a63e  fix bug for when self.arg_params is none
     new 52793a1  Merge pull request #611 from huazehuang/master
     new 376d262  [example] refactor
     new 29546a5  [example] update gitignore
     new ee6699d  [example] update
     new 63e1ae6   Changes to be committed: 	modified:
     new aa7b09d  Merge pull request #613 from wangg12/master
     new 93ba9de  Update
     new f9436b9  Update
     new d5c74ea  Merge pull request #5 from wangg12/dev
     new 4675043  Merge branch 'master' into master
     new 5139e02  Merge pull request #618 from wangg12/master
     new d552456  make caffe converter tool available without caffe installation
     new 375b4c6  fix bug in optimizer epoch counter
     new 3ee0910  Adam optimizer
     new 30c557d  Merge pull request #622 from jermainewang/master
     new 10b00946 Merge branch 'master' into master
     new 8d8298d  Merge remote-tracking branch 'upstream/master'
     new 906a28e  Merge branch 'master' of
     new 9342514  [example] update
     new c4eae82  [example] update readme
     new a0d3799  Adam optimizer
     new e304fc0  Merge pull request #624 from mli/master
     new f1a7664  fix Adam
     new 36665f4  lint issues with Adam
     new b149c37  Merge branch 'master' of
     new 11226be  Merge remote-tracking branch 'upstream/master'
     new 955f66c  # This is a combination of 3 commits. # The first commit's message is: Adam optimizer
     new a1693d5  Merge branch 'master' of
     new 5b0e1ac  remove grad_scale parameter from optimizers (duplicated with lr)
     new 4fc66b4  update documents for optimizers
     new e14098f  [example] rename
     new 13e9b95  [example rename]
     new 1d90555  Merge branch 'master' into master
     new 620b86c  [python] add mxnet/
     new dedfe61  [python] better server logging format
     new c9089e0  rm caffe.proto and update readme.
     new 932887a  Merge branch 'master' of
     new 6869fdc  remove debug info
     new b98c0c6  [python] fix lint
     new f3b1250  Merge branch 'master' into master
     new b33a582  fix lint
     new 99d1ec7  Merge branch 'master' of
     new a9e5461  Merge pull request #620 from dongguosheng/master
     new ef84a84  lint
     new 4a7763c  Merge branch 'master' into master
     new f77848e  Merge pull request #626 from winstywang/master
     new b502851  sparse reg op based on KL
     new 07f21ff  Merge branch 'master' of
     new 74caf0d  Merge branch 'master' into master
     new ebfa686  [R] fix two demos (close #631)
     new de4da8f  replace != with is not None
     new 884c7f8  more arithmetic operations on symbolic nodes
     new b42f63e  [Example] num_round->num_epoch
     new 9d97743  [tool] add tools/
     new 8bea73c  dec example
     new 210e786  dec_example
     new 0e7cd6d  [example] update distributed training on image_classificaiton
     new 31f26bb  [example] update readme
     new 0e08b17  [example] update readme
     new 598fbc7  mshadow uninitialized shape fix
     new 616e223  [example] clean cifar10/mnist/imagenet
     new 09ac5a7  Merge pull request #627 from mli/master
     new c8f0149  Merge branch 'master' into master
     new 6e8b6bc  Merge pull request #643 from piiswrong/master
     new a651d35  [example] test embeded table
     new 0d08dc5  [example] test
     new 08e69d3  [example] add figure
     new b136aa4  Temporary fix to validation accuracy when train batch_size < val batch_size
     new 90c069e  concat supports arbitrary dimension
     new 0566d4b  Update composite_symbol.ipynb
     new 62f05a5  Merge pull request #652 from wangg12/patch-1
     new 55b01c7  Update Xavier
     new 4aa7d3b  Merge pull request #610 from antinucleon/min-net
     new 62c9703  rework Xavier to be more flexible
     new 6081fce  xavier: rebuild documentation
     new 7e42002  rm auto generated; add caffe.proto; add make tools option
     new 4a36dfd  add make tools option
     new 63b9302  update readme
     new bdfbf05  update readme
     new ea85774  Merge pull request #32 from vchuravy/vc/xavier
     new 845e373  Merge branch 'master' into master
     new 3a5a844  Merge pull request #650 from yang1fan2/NewBranch
     new a079fd3  Merge pull request #637 from thirdwing/master
     new 7e001b1  [Operator] OnehotEmbedding
     new 3e580e0  Merge remote-tracking branch 'upstream/master' into min-net
     new ca13003  [EXECUTOR][Fix] Fix hanging when symbolic input have duplications
     new 56f3eef  update README
     new e848e8d  spelling :bee:
     new 948966a  Merge pull request #665 from tqchen/master
     new d2a1f91  onehot->embedding
     new 6fb9ca8  merge
     new 03eb95f  Merge pull request #33 from yeesian/update-readme
     new 0b5c072  Merge pull request #654 from antinucleon/min-net
     new 158d581  Update LSTM
     new 510b409  [OP] Enable asynchronize op
     new bd8fd49  Update activation-inl.h
     new 0f70cf9  Merge pull request #666 from antinucleon/min-net
     new d57ca8b  Update
     new d21a7b1  Merge pull request #667 from tqchen/master
     new d804496  Update
     new a6dbf81  Update
     new 41f4fcd  Merge pull request #649 from XericZephyr/master
     new 45bae3f  [example] add lr-factor for imagenet
     new 6f1ce2a  Merge branch 'master' of
     new 3545f2d  merge _update_count in Adam
     new f728603  use fixed learning rate when using wd in Adam
     new 399b91d  [tool] update parse_log
     new 12dd601  [example] update readme
     new d2fb53b  Merge branch 'master' of
     new 32609ca  [example] update readme
     new da2810f  [example] update
     new 94fe551  [example] remove distributed-training
     new f47ca41  Merge pull request #662 from dongguosheng/master
     new 4852b1c  Merge pull request #674 from mli/master
     new d161c57  epoch->num_update
     new ae5c692  rename scalar_on_right to scalar_on_left
     new 99bdc51  abs add in unary_function-inl.h
     new 16b4c04  Merge pull request #623 from udibr/master
     new 31f630e  Merge pull request #653 from yajiedesign/scalar_symbol
     new 7d8de6b  fix API change: scalar_on_right -> on_left
     new 2c72446  Merge branch 'master' of git://
     new 3e5aa53  [OP] Enable asynchronize op
     new 931d489  Update activation-inl.h
     new dabc5ba  Update
     new 6f3bc62  Update
     new 503c45a  Temporary fix to validation accuracy when train batch_size < val batch_size
     new 1a56670  runtime cuda compile using nvrtc
     new 449cf37  fix multi device
     new 790bafd  python interface
     new ac21f57 not found fix
     new f097a3f  mxrtc fix
     new b002a92  naming
     new 13440af  Merge branch 'master' into master
     new ff787ef  fix typo in doc/
     new bac5ddb  Merge pull request #668 from piiswrong/master
     new e83154a  fix doxy check code
     new 65fa289  Merge pull request #682 from tqchen/master
     new 81235b6  [R] update R installation instruction. close #459
     new 7d41114  Merge pull request #683 from thirdwing/master
     new 7dc5526  Merge pull request #677 from pengli09/master
     new 7acd201  [example] fix
     new c551ff4  [example] fix googlenet
     new d0fb0d3  [SYMBOL] enable attributes in graph node
     new c0bf9d6  ndarray operator
     new 83f0d08  doc
     new 681c377  Merge pull request #678 from piiswrong/master
     new 87a28ca  Merge pull request #685 from tqchen/master
     new 82849fb  Merge branch 'master' of into addops
     new 4eb698e  change abs_grad to sign
     new ff0de38  add an option to plot network in other shapes
     new 46257d2  basic interface for setting and getting attributes
     new 47628ff  fix typo in
     new 95d0d95  Merge pull request #691 from wangg12/master
     new 05bcddf  Merge pull request #35 from vchuravy/vc/symbol_attr
     new e19ca8e  add warning to discourage users from calling set_attr directly
     new d7684c1  Merge branch 'master' into addops
     new e281296  Merge pull request #634 from yajiedesign/addops
     new 7a3c920  Merge pull request #688 from mli/master
     new e969a36  Merge pull request #693 from austingg/fix-typo
     new b6a76e4  Update mshadow with Shape5 helper function for prospective volumetric operators
     new 8f7cfd7  Merge pull request #696 from korneil/master
     new beb611a  ndarray op is now async
     new 9f5ec2c  fix macro name shapce
     new 1f90b6c  ndarray op backward dependency
     new 2586821  ndarray op backward dep and error handling
     new 0937d94  documents
     new 86d3c28  engine help information
     new 1cdd484  Merge pull request #689 from piiswrong/master
     new 80597d8  [doc] update
     new c287fe7  [make] use `BLAS=apple` if osx is detected
     new 20043f3  [doc] update
     new f9ace02  [doc] tiny
     new 90c98b5  Merge pull request #698 from mli/master
     new 266d946  attr interface for SymbolicNodes
     new 247ec44  attrs: Doc rebuild
     new 1f4bc8b  Merge pull request #36 from vchuravy/vc/attr
     new b6794ff  [EXECUTOR] Enable bind multiple devices in one graph, beta of model parallel
     new f9970de  update example
     new 93498c0  fix tutorial link
     new b395f7f  Merge pull request #690 from tqchen/master
     new c9794bf  add sign op
     new f88a900  add miss lib nvrtc cuda
     new 51d1cc0  Merge pull request #702 from yajiedesign/addops
     new bb7d759  Merge branch 'master' into addcmake
     new 3954033  numeric gradient
     new 4f67a7d  Merge pull request #703 from yajiedesign/addcmake
     new c05a8ee  Merge branch 'master' of into lm/numeric_grad
     new ef627f7  [doc] minor fix
     new 0f545bb  fix list aux_states
     new 5f6ff08  Merge remote-tracking branch 'upstream/master'
     new cd28bef  Merge branch 'master' of
     new 839a692  remove warning
     new 8205b47  [example] update init, add clip gradient
     new 893243d  add round ceil floor
     new 05cb0d3  Adds ELU to LeakyReLU activation layer
     new 3375f7a  Merge pull request #718 from vchuravy/vc/elu
     new c5c8a20  Merge pull request #713 from antinucleon/master
     new c66531f  adds LearningRate.Inv from Mocha
     new bb7e542  Merge pull request #37 from vchuravy/vc/lr_inv
     new 61feebf  [example] ignore clip_gradient if not set
     new e2a294f  Merge pull request #724 from mli/master
     new 15965e2  Merge pull request #711 from chuan92/master
     new 3897d54  Update
     new c05915a  fix autoencoder eval
     new bc4c323  ndarray op interface fix
     new df134a6  executor callback
     new fa98c2a  executor call back add name
     new d6ab7f4  monitor
     new 23311b8  Merge pull request #700 from piiswrong/master
     new eb42c51  Merge branch 'master' into master
     new 24aae83  some minor fix to document
     new 717ea32  correct link in
     new 5e62997  Merge pull request #730 from wangg12/fix_imagenet_full_doc_typo
     new b5061e1  [tool] update caffe_converter
     new 0f34012  remove gradient function
     new a6da221  Merge branch 'master' of into addops
     new 60b3871  Merge pull request #717 from yajiedesign/addops
     new 59afc61  Merge branch 'master' of
     new fc68265  fix compilation error of null-initializing std::function on windows
     new cfc9949  [example] init commit of neural-style
     new 7b575fd  [example] remove model downloader
     new cceb031  Merge pull request #731 from hjk41/master
     new 9de49f6  [example] update readme
     new 1490ed3  minor fix for doc
     new e7a459c  update example/notebooks/simple_bind.ipynb
     new 74cc5ba  Fixes usuage of floor
     new 91dd556  Merge pull request #734 from wangg12/iter_bug_fix
     new 0d11ebc  Merge pull request #42 from vchuravy/vc/fix_ops
     new 9664b73  [example] = -> [:] =
     new fd15bc9  Add softplus activation function.
     new 6fc7c17  Merge pull request #739 from mli/master
     new eaedfb6  pass memo parameter to deepcopy method
     new 4c28b5b  Merge pull request #738 from sveitser/fix_symbol_copy
     new e23b6c6  fix cifar10-recipe error
     new 5cea34e  [Operator] add nearest neighboor up sampling
     new 390fcc5  rename
     new 30d886f  Merge pull request #737 from wangg12/fix_cifar10-recipe_error
     new 6e584cd  Rename softplus to softrelu.
     new 3310f58  Merge branch 'master' into softplus
     new 2bf9d6b  Merge pull request #740 from nowozin/softplus
     new e73fc6a  Merge pull request #742 from antinucleon/master
     new 17ea996  [Operator] Bilinear upsamping
     new f4b5366  fix shape
     new 6872180  add init bilinear weight
     new 0723472  Merge branch 'master' of
     new 7d3b9fa  Merge branch 'master' into master
     new 26639e3  Merge branch 'master' of
     new 9313b01  naming
     new 920219a  add
     new f094a11  add missing _init_default
     new 9ab41d4  fix
     new d6f5086  expose ndarray interface for c++ dll, and add ndarray code example
     new 8ce7dfa  fix lint
     new 30e742b  update pylint version
     new 8007864  Merge pull request #746 from antinucleon/master
     new 5712539  [Operator] fix doc of upsampling
     new d6e24c3  Merge branch 'master' into master
     new c656d20  Merge branch 'master' into master
     new 8db09dc  Merge pull request #754 from antinucleon/master
     new 8fb0810  lint
     new 6198c38  Merge branch 'master' of
     new 85c01f3  Merge branch 'master' into master
     new 2d7eee8  add rsqrt cos sin
     new 0c4da94  Merge branch 'master' of into addops
     new 2e1fd25  fix lint
     new 562ded3  Merge pull request #633 from winstywang/master
     new 6c1dd42  Merge branch 'master' into addops
     new fcf51b3  fix lint
     new b3f02c0  Merge pull request #687 from lukemetz/lm/numeric_grad
     new e043615  Merge branch 'master' into addops
     new 6b14c95  Merge branch 'master' into master
     new b2eb4bf  Merge pull request #750 from hjk41/master
     new 34bbfc6  Merge branch 'master' into addops
     new 0fb336d  Merge pull request #755 from yajiedesign/addops
     new bcd1374  Update
     new dfb7f0c  fixes workspace being bigger than anticipated
     new 2ff90c3  imports operators into the namespace to prevent issues
     new fdf536c  Merge pull request #46 from vchuravy/vc/ops
     new 171754a  Merge pull request #761 from rinuboney/patch-1
     new c8db9df  fix 3 minor mistakes 1. 'rectangle' -> 'square'
     new a9ac086  Merge pull request #770 from suhangpro/patch-2
     new d42e2c6  Update
     new 67c8bb2  prepare for v0.0.6 (c.f. #42)
     new c8d561c  [doc] add learningsys paper link
     new 206a711  [doc] update
     new 8247056  fix ae && add sparse to sigmoid
     new 982c71b  Merge pull request #778 from winstywang/master
     new 3fa6244  Minor corrections to
     new 3ef8f41  Merge pull request #776 from mli/master
     new 37186bd  Merge pull request #785 from sennendoko/patch-1
     new 465f665  Update
     new c56a673  [doc] intro cn
     new a60fbfb  [doc] update
     new 12f0655  update
     new 4fc786e  [doc] intro
     new 5826fb4  [doc] move
     new 80dd2a9  [doc] add index
     new 86c8c34  Merge pull request #795 from mli/master
     new 280ef37  Update
     new 3456bfd  Fix a couple of typos
     new 1567263  python interface for creating image record binary
     new 366dce0  hypercolumn nearest neighbor upsampling
     new 6b844d5  opencv3 compatibility
     new 526c26b  monitor fix
     new 87580b5d record io fix
     new 135bff7  Update
     new bbfa834  Merge pull request #779 from piiswrong/master
     new aa1b600  backport @defstruct from SimpleStructs.jl
     new f913f5b  add rmsprop optimizer
     new 02452b7  Merge pull request #801 from terrytangyuan/patch-1
     new f7aa84f  Merge pull request #800 from DODOLONG/patch-1
     new dc0229b  fix format
     new a6624ae  remove unnecessary code
     new 03ba419  remove unnecessary code
     new a519efd  修正几处笔误
     new 1b629b9  Enable MXNetJS
     new 990bbd1  fix makefile
     new b22ffa8  minor
     new 8e287c8  Merge pull request #815 from tqchen/master
     new 3bfe14d  some common error for mac user
     new 09cd084  fix
     new b98c639  Merge pull request #819 from CodingCat/README_fix
     new e70cef2  Merge pull request #813 from qiaohaijun/patch-1
     new ce74693  Merge branch 'master' into troubleshooting
     new f292c81  Merge pull request #818 from CodingCat/troubleshooting
     new 2668ba7  Fix RGB mean value
     new 26874f4  Fix #820: Order of RGB channel
     new 60f0067  Merge pull request #823 from zhouye/master
     new 58575c2  [EXEC] Enable partial forward for easy display of progress
     new 6cc09e6  Merge pull request #824 from tqchen/master
     new fabb55e  Update c_predict_api.h
     new 34c7fe5  Update
     new 146c99f  Update
     new 59e0cfa  Update
     new dd4ad39  Merge pull request #811 from Kublai-Jing/master
     new 83ddd0e  Merge remote-tracking branch 'upstream/master'
     new 1023f5c  fix rmsprop lint
     new bc73705  Merge pull request #828 from antinucleon/master
     new 5e67d11  Create
     new b513898  use download instead of wget
     new 5cc456c  update gitignore
     new 0321b96  More info about CUDA in library requirements
     new 8d3a9a0  Merge pull request #48 from SimonDanisch/patch-1
     new 9427212  Merge pull request #834 from CodingCat/master
     new 2ba474e  [LSTM] fix Char-RNN example
     new 24cc09d  Merge pull request #839 from tqchen/master
     new b317aa8  Merge pull request #733 from vchuravy/vc/fix_ws
     new 42556d5  Merge branch 'master' into patch-2
     new a60e848  Merge pull request #835 from HeNine/master
     new 7d8d257  add test_deconvolution()
     new 001459d  fix deconvolution workspace
     new 23ef4ff  fix deconvolution workspace
     new d5a2467  fix lint
     new e8cb37e  Merge pull request #741 from tornadomeet/master
     new ecb1e13  add inception-bn-full to choice of network
     new f16c7a3  upsampling fix and add test
     new 552c064  prefetching data iter
     new a6afcc9  recordio add reset
     new 2c3d3b5  L1 metric
     new 53c0512  directly compute the workspace size
     new 737c795  allow workspace = 0, to automatically determine workspace size
     new a0335b8  ELU: change greater than equal to strict greater.
     new a73682e  [update bn] moving aux to backward
     new 81f835e  choose the NIC
     new 86fdb64  prefeching iter refactor
     new a1a25ba  Merge pull request #856 from qiaohaijun/patch-4
     new 2090ab7  Merge pull request #849 from qiaohaijun/patch-3
     new c2f6dd7  Merge pull request #854 from oist/vc/fixup_elu
     new 07448b6  Merge branch 'master' into master
     new 799a5dc  Merge pull request #853 from antinucleon/master
     new 2a18964  Merge branch 'master' into master
     new bf91ea7  Merge pull request #832 from piiswrong/master
     new fccfff8  Fix bugs of grad_scale in softmaxOutput. When grad_scale is > 1.0, it should also be multiplied by.
     new 2ed4c99  Add grad_scale parameter in RegressionOutput to support scaling the gradient
     new d612faf  Accuracy metric added to epoch callbacks..
     new a5c6693  Accuracy metric added to epoch callbacks.
     new 9344979  笔误修改
     new 49a92d1  Accuracy metric added to epoch callbacks.
     new 239c06e  Merge pull request #874 from lyttonhao/master
     new 9286bcc  Merge pull request #50 from kasiabozek/acc_callback
     new db530e1  Merge pull request #875 from taiyun/patch-1
     new f8123cd  mean absolute error regression
     new 66c6a54  add min max sum
     new 212b991  c++ optimizer
     new 3e51c33  reshape implicit dimension
     new e9e99a7  convolusion workspace fix
     new 7a420d0  update bn
     new 46eb3d9  pythonop garbage collection fix
     new 2c502a4  fix
     new 7cae19a  Merge pull request #882 from antinucleon/master
     new 51be106  [EXECUTOR] Experimental Mirror option to squeez mem
     new 635f5ef  add maximum minimum ops
     new 29ab2a6  Merge branch 'master' into master
     new 4e0a026  Merge pull request #838 from yajiedesign/addops
     new e9b1572  Merge branch 'master' into master
     new 4163530  Merge pull request #873 from piiswrong/master
     new 1dd5d3c  Create mlp.cpp
     new a3851d3  Merge branch 'master' into master
     new acaa0df  Makefile
     new dc0141b  Update mlp.cpp
     new 7264098  Update Makefile
     new e94e7f3  Update mlp.cpp
     new c720377  Fix issue #893
     new 0c99995  remove duplicate var definition
     new 869f7ac  Merge pull request #899 from zhangchen-qinyinghua/master
     new a08acc7  Additional installation instruction
     new b51b352  Merge pull request #900 from srand99/patch-1
     new 7f4a851  Merge pull request #896 from javelinjs/master
     new ea0fe33  Merge pull request #901 from terrytangyuan/patch-1
     new 2b09255  Update
     new 841bd10  Update
     new 9ba8f79  Update mlp.cpp
     new 9b058ac  Update Makefile
     new f26801d  Merge branch 'master' into master
     new f43d8f7  Merge pull request #906 from zhangchen-qinyinghua/master
     new 71b4945  Fix compilation errors for julia 0.4.2
     new 8e879d7  Fix compilation errors for julia 0.4.2
     new 32c8954  Merge pull request #53 from BigEpsilon/patch-4
     new 5ed54d8  try to fix build-script for installing cblas
     new f91c200  add Juliapkg badge
     new 7c1f832  Handle grayscale images. * Add a color parameter in the tool im2rec * Decode rec dataset without forcing color mode. * Substract mean on 1 channel only for grayscale image
     new 8e890e4  add RegressionOutputParam of MAERegression for MXNET_REGISTER_OP_PROPERTY
     new 8b81bb2  update data_iter example
     new 71a7128  Merge pull request #914 from lyttonhao/master
     new 4efaa42  Merge pull request #859 from u1234x1234/master
     new 9692929  corrected some minor typos
     new 4921980  Merge branch 'master' into master
     new 349a8c1  Merge pull request #925 from rinuboney/master
     new b19e140  Merge pull request #929 from mli/master
     new 7b80326  [doc] rename from chn to zh
     new e3bdb12  [doc] update
     new 1c1dfbc  Merge pull request #930 from mli/master
     new 4f93a7c  [python] disable kvstore for single gpu to improve performance
     new e442734  [python] bugfix
     new 0169981  try to use the built-in bindeps for building in travis CI
     new dfe66f0  Merge branch 'master' of git://
     new b0ec8b5  [doc] update
     new d8e5065  fix travis CI building script
     new 7a7cfbe  trying to fix travis CI gcc-4.8 thing
     new 99e7e66  still trying to fix travis CI gcc issue...
     new 3861edc  clean up travis CI script
     new 3a134e8  test whether bindeps works well on travis CI to install cblas
     new ce57a39  test failed, travis CI is configured to disable sudo
     new 8edb94b  prepare for v0.0.7
     new 627289f  Merge pull request #932 from mli/master
     new 1d7f887  [doc] add link to neurl art
     new 432c22e  [doc] update FAQ
     new 54f8eb8  [doc] update build instruction
     new da04a56  Update
     new 378d2f2  Update
     new 2d06f0b  Update
     new 973ba88  Merge pull request #852 from oist/vc/conv_mem
     new 0bb567c  Update
     new 627d0ff  Merge pull request #933 from wangxianliang/master
     new 6946d19  Update
     new ad8649e  Revert "[RFC] Exactly calculate nstep from workspace."
     new 46ed123  Merge pull request #934 from dmlc/revert-852-vc/conv_mem
     new dbc1a80  update ps-lite
     new 021f700  Merge pull request #829 from qiaohaijun/patch-2
     new 4ba4f50  Merge pull request #936 from mli/master
     new 18404b1  fix compile error with visual studio
     new 56a16cb  Fix a typo
     new 677e248  Merge pull request #939 from bigtan/patch-1
     new e0577cf  Merge branch 'master' into addops
     new de25459  refactor _train_multi_device
     new e58eb06  directly compute the workspace size
     new f4ebd05  Merge pull request #944 from oist/vc/conv_mem2
     new 3c51e8d  Merge branch 'master' into addops
     new fee8175  Update
     new d5eb996  Moved mxnet-scala to main repo
     new b4f1359  Added scala-specific gitignore
     new 7f61cbe  Trigger Travis
     new 72d4900  Minor changes
     new a998399  Merge pull request #950 from yzgysjr/master
     new 47ef937  Merge pull request #935 from yajiedesign/addops
     new 087e198  Update
     new fbe0228  Changed scala sub-dir name
     new 305752a  Corrected relative libmxnet path
     new a8fc194  Setup travis test for scala package
     new bfaae3c  Temporarily disable other tests and exported correct JAVA_HOME (+2 squashed commits) Squashed commits: [479f427] trigger [9bb7bf0] Temporarily disable other tests and exported correct JAVA_HOME
     new 6882cdc  Merge pull request #4 from terrytangyuan/master
     new e2cc21b  add linux maven settings for scala-package
     new c5a358a  add scala-package ignore settings
     new 84c1772  fix compile error with visual studio
     new d8fcdf7  Update
     new 721cb54  Update
     new cd471ba  Merge pull request #1 from yanqingmen/scala
     new 4746a3f  add inception-v3 symbol
     new 6977ac7  Merge pull request #959 from antinucleon/master
     new 805a542  modify maven linux settings
     new 686fcc7  Merge pull request #2 from yanqingmen/scala
     new 5b7790e  Add integration test for linux
     new 2d87baf  Merge pull request #3 from terrytangyuan/terry
     new 05431f5  Merge pull request #5 from terrytangyuan/master
     new fd1a3b8  NDArray array() and ones()
     new 19011c6  NDArray toScalar and test
     new ff4f019  fix the chunk_size
     new 3f28d18  Add image classification predict example for C++
     new 0333370  Merge pull request #964 from qiaohaijun/patch-6
     new 7bb03ac  Update
     new 91507e9  Merge branch 'master' into master
     new a41c2e8  add IO classes
     new ea453a5  Merge pull request #891 from skylook/master
     new d038bad  Some test cases
     new 72e0f6e  Remove wrong implementation of array() and fixed typo in test
     new b5280c7  Test for * and /
     new 1102345  Test case for size and shape
     new 96ae2d8  Revises per comments on PR (+1 squashed commit) Squashed commits: [f65005c] Revises per comments on PR
     new 30bc343  Update
     new 6d5bc04  Merge pull request #4 from terrytangyuan/terry
     new 60f58a5  Update
     new ce49c4e  Merge pull request #969 from mli/master
     new 3a48d26  [EXECUTOR] mirror fix of mirror direction
     new d1e279e  [OP] Fix BatchNorm when requirement is not all writeto, more memory efficient BN
     new 398278c  fix nvrtc problem on windows
     new d1fb532  [EXEC] fix g++ 4.9
     new e4d769b  resolve conflicts
     new d75f9ec  add ant-run to involve blas dependency
     new f0028c5  linux-x86_64 builds successfully
     new 8d9ee47  Update
     new c661161  fix NDArray.toScalar
     new 7d8bfb6  KVStore clean code & add test cases
     new ac7a933  KVStore clean code
     new 2f8ea33  NDArray remove useless codes
     new 1cf40ad  Merge pull request #973 from xjlc/patch-1
     new 6194cbf  Softmax Activation Layer
     new 5b9f40d  various small improvements
     new bf172a0  support nd cudnn activation
     new ea25998  make ccsgd the default and add number of batch to predict
     new 5a5fb9e  Merge pull request #915 from piiswrong/master
     new e316055  Merge pull request #909 from lukemetz/lm/model_refactor
     new afe2add  add property for c++ compiler, fix linux build failure on travis
     new aae94f0  Merge pull request #5 from javelinjs/scala-package-cc
     new 615fc2f  add cudnn v4 with CUDNN_MAJOR auto switch v3/v4
     new 2bb51f9  fix lint
     new ea1414f  Merge from upstream dmlc/mxnet
     new 087c6e6  Merge pull request #884 from tqchen/master
     new 8b76e06  Merge pull request #974 from yajiedesign/cudnnv4
     new ebba851  KVStore setOptimizer, with JavaSerializer
     new 8ac0849c fix cudnn selection
     new 53aca71  Merge pull request #983 from antinucleon/master
     new 0c14a89  Merge pull request #912 from Nzeuwik/master
     new 8968d1b  [SYMBOL] Fix multiout check error
     new fc0f1c0  change size_t -> int in c_predict_api.h
     new de143cb  Merge branch 'master' of
     new 96cd865  add KVStore rank & numWorkers
     new bf90d43  Merge pull request #986 from tqchen/master
     new f189797  Update
     new df355a5  update fix gamma for bn
     new f4ae0e3  Merge pull request #984 from antinucleon/master
     new e78cc31  Merge pull request #6 from javelinjs/scala-package-cc
     new 4649936  Merge remote-tracking branch 'upstream/master'
     new ea45469  Add error detector for scala tests on Travis
     new 19b296b  corrected path to new sh file
     new b343122  Merge pull request #971 from hjk41/master
     new fb06e5f  revert to sgd
     new 6227db2  Merge pull request #988 from antinucleon/master
     new 9de9af3  add SGD optimizer but not tested yet
     new b294fe8  update ps-lite
     new 4535b1f  more matlab
     new 39731d0  Update
     new 3a55502  Merge pull request #996 from xgastaldi/master
     new aad3468  print message and content when failing (+2 squashed commits) Squashed commits: [99ccdac] fix path [b32020a] fixed paths of using test scripts
     new d37fe4c  Merge remote-tracking branch 'upstream/master'
     new 5197470  Merge branch 'terry'
     new b6c971f  Added class for learning rate scheduler
     new 16bc2a2  fix implicit-conversion-failure-from-initializer-list
     new 15790ba  Documentation for LRScheduler
     new f980f05  Fixed type mismatch
     new 68dbd8e  Change error throw to require
     new aad6806  break src/symbol/static_graph.h line 186
     new a2edc89  Change to Float and make member vars protected
     new 19863ea  Merge pull request #998 from shiyemin/master
     new a6c9ead  Merge pull request #7 from terrytangyuan/terry
     new 8846e7a  [EXECUTOR] Experimental Mirror option to squeez mem
     new e395eb0  Handle grayscale images. * Add a color parameter in the tool im2rec * Decode rec dataset without forcing color mode. * Substract mean on 1 channel only for grayscale image
     new bb18f47  [EXECUTOR] mirror fix of mirror direction
     new aefe734  [OP] Fix BatchNorm when requirement is not all writeto, more memory efficient BN
     new e3c3473  fix nvrtc problem on windows
     new 510ca83  [EXEC] fix g++ 4.9
     new 4485363  add cudnn v4 with CUDNN_MAJOR auto switch v3/v4
     new 821bf3e  fix lint
     new 15d7af2  fix cudnn selection
     new 982fa87  [SYMBOL] Fix multiout check error
     new 3e6cc62  Update
     new 832a198  update fix gamma for bn
     new ebe1001  Add error detector for scala tests on Travis
     new 587266c  corrected path to new sh file
     new f6e767d  revert to sgd
     new 6359d2a  update ps-lite
     new 1d45e2d  Update
     new 617222b  print message and content when failing (+2 squashed commits) Squashed commits: [99ccdac] fix path [b32020a] fixed paths of using test scripts
     new 4be04b1  Added class for learning rate scheduler
     new 5e5aa59  Documentation for LRScheduler
     new 436062a  Fixed type mismatch
     new 770afc2  Change error throw to require
     new b170bf2  Change to Float and make member vars protected
     new 8c35e55  merge from tangyuan/mxnet
     new 0ad6840  cudnn batch norm
     new 76ef038  [IO] Make batch loader generic to accept all incomes
     new 60aac18  Merge branch 'master' of into terrytangyuan-master
     new 6f5f61e  Merge branch 'terrytangyuan-master' into scala
     new 71c46cb  add a shell to get ptb data
     new afddc97  correct train's output
     new 3b5e7a3  add MXDataIter
     new b10a6b2  add SGD optimizer, add java friendly inteface for NDArray
     new 8fe6905  NDArray copy from java array
     new c14ed50  remove NDArray.apply because it is really confusing
     new 13e9ff8  add a shell scripts to download ptb dataset
     new d417d9a  Merge pull request #1006 from qiaohaijun/patch-8
     new 6cf2f64  Merge pull request #1005 from qiaohaijun/patch-7
     new 4b87b58  Update
     new b6dbddb  [IO] Add CSV Iter
     new 524c369  matlab
     new 67abdee  Merge pull request #8 from javelinjs/scala-package-cc
     new c0919d3  Added Adam Optimizer
     new 9580eca  change from char* to void* in predict_api for matlab binding
     new 63be467  bugfix
     new f37b1f6  bugfix
     new 5083a33  update readme
     new f8c4403  correct chunk_size
     new 63f311e  Finished Adam optimizer
     new edaeeec  Minor fix to pass the test
     new ef46e26  Merge remote-tracking branch 'upstream/master' into terry
     new 5ac77c6  Merge pull request #1015 from qiaohaijun/patch-9
     new 029ba78  Merge pull request #1019 from tqchen/master
     new 44e5dc7  ndsb
     new 217fc72  Merge remote-tracking branch 'upstream/master'
     new 602a10d  Merge pull request #1021 from antinucleon/master
     new 86084f8  Update
     new 9c79179  Update
     new f9b68fd  add IO funcs
     new 9af4f65  Handle grayscale, RGB, RGBA in im2rec and data loader
     new ee417ef  add NDArray sqrt function, but failed to load the function on osx
     new 30ca7f2  Merge pull request #9 from terrytangyuan/terry
     new 3f66118  Merge pull request #1013 from mli/master
     new 6dcc3e5  Update
     new bcd1dc4  Merge pull request #1025 from walterreade/patch-1
     new 4dc3a45  [doc] faq for build
     new 93061b1  [doc] update
     new ed7eea2  [doc] update
     new 1de91ab  Merge pull request #999 from piiswrong/master
     new 045aac0  Merge pull request #1026 from mli/master
     new d7d4cff  force load on osx
     new 0f113c4  add CuDNNBN to list not re compute
     new b839db5  Python2 compatibility for implicit integer divide
     new 4f142e7  change clang++ to g++ on osx
     new fd4aace  link libgomp on osx
     new 96ca3f8  sync with changes in C++ end
     new a73b51b  Merge pull request #1030 from thirdwing/sync
     new 143d60a  Merge pull request #1029 from udibr/master
     new ddf93c9  Merge pull request #1028 from antinucleon/master
     new 4e868d0  Merge pull request #1000 from Nzeuwik/feature/generic_input_for_data_loader
     new 5571bf1  Update
     new a2fa3df  EvalMetric and Accuracy class
     new f856aac  metrics: MAE and RMSE (WIP)
     new 78e3240  Custom metrics and docummentation
     new cb1833e  override name in CustomMetric
     new bdb6d70  override name in CustomMetric (+1 squashed commit) Squashed commits: [cb1833e] override name in CustomMetric
     new ac06bee  quick fix for pack
     new 53fc1dc  output should use os.str
     new 3e1f8bb  fix libgomp ld error on osx
     new 7b83f94  Merge pull request #10 from javelinjs/scala-package-cc
     new 97c7068  Merge pull request #1033 from winstywang/master
     new 528bf7c  Merge pull request #1034 from qiaohaijun/patch-10
     new a7970fc  add --log-file --log-dir argument to let output write to log file
     new 949a611  add IO jni code
     new 31d4f01  fix image augmenter bug
     new 55c890b  misc update
     new 35c277c  Merge pull request #1043 from mavenlin/augmenter
     new e3696b3  Fix conflict
     new c78e495  Merge branch 'master' into master
     new a67fe1b  Added scalastyle plugin
     new fd499d5  Merge pull request #1044 from thirdwing/master
     new 5a52ab4  Added config file for scalastyle
     new cf442ca  Correct error spotted by @winstywang
     new baaa9ae  Merge pull request #1046 from Nzeuwik/hotfix/Correct_PR#1000
     new 6b5f48a  Merge pull request #1039 from qiaohaijun/master
     new 203ab2f  [R] operators between mx.symbol and scalar
     new 17a0a0c  Merge pull request #1050 from thirdwing/master
     new ba33cc9  [IO] fix odd image load
     new d8c8a0c  Merge pull request #1054 from antinucleon/master
     new fc436a2  fix toPyString
     new 53a85bd  Merge pull request #1059 from thirdwing/master
     new dcd0d7c  [R] ndsb2: training part (Score: 0.039011)
     new ac8f5c8  Merge pull request #1062 from thirdwing/kaggle
     new 2cc45fc  fix TabError: inconsistent use of tabs and spaces in indentation python3.3
     new d8b1e77  batchnorm gradient test
     new 383a6a0  [doc] add titan-x perf numbers into image-classification
     new bbcfeba  add CreatePartialOut in c_prediction_api
     new 737e117  [matlab] support feature extraction
     new a5e51e4  Merge pull request #1070 from mli/master
     new 1f0bb48  Initial code for Monitor and abstract class for Executor
     new ceae2c3  Monitor.toc, tocPrint and doc-strings
     new 76a7f55  Fixed indentation for scalastyle_config
     new c5d630e  Merge pull request #1066 from qiaohaijun/patch-11
     new 5b751a2  update makefile
     new bba1dbc  Merge pull request #1077 from antinucleon/master
     new d7fbdbe  Changed to use mutable.Queue for performance
     new 3916e8e  Merge pull request #11 from terrytangyuan/terry
     new 1bbde69  add IO MINISITer test
     new 7778d2d  Gracefully fail if EmbeddingOp is instantiated with input of incorrect shape
     new 1d02f4b  Merge pull request #1079 from SkidanovAlex/master
     new ac38dc4  Merge pull request #1061 from lukemetz/lm/batchnorm_grad
     new d4e397f  Merge remote-tracking branch 'upstream/master'
     new 8d5249f  Merge pull request #7 from terrytangyuan/master
     new ee03dae  add IO ImageRecordIter test
     new cfb7162  Merge branch 'scala' of into scala
     new 884f14b  add Model.scala with createKVStore & initializeKVStore
     new 2ce0e51  some Model functions, add (key, array-of-values) pull & push to KVStore
     new 9e3fe87  Executor::forward and other helper functions
     new ffd819c  Executor class functions
     new 781a0a7  add Executor::splitInputSlice, fix build break
     new e720e70  Executor static functions
     new 7f89e07  merge Monitor and fix some type mismatch
     new ed4077d  change Monitor callback function to trait to make it java-friendly
     new 7cc514a  complete Monitor.scala, add NDArray::norm
     new 52733fc  typo fix
     new a03c228  link cblas for linux profile
     new c89e4ab  add util scriptsfor IO test
     new 356ff47  force load libblas in travis
     new ff983c7  [R] misc updates: (1) makefiles for Windows pkg; (2) README and comments for ndsb2; (3) sync with C++ side
     new 213cd10  linux travis scala-test should be success this time, FSM bless me
     new ee51492  Merge pull request #12 from javelinjs/scala-package-cc
     new 2645bac  Merge pull request #1086 from thirdwing/master
     new 9a979ea  Random.uniform and normal
     new c12eb5e  Random.seed and doc-strings
     new d9915ad  merge from terrytangyuan/mxnet
     new ffb0df0  Merge branch 'terrytangyuan-master' into scala
     new 27b892c  Initializer abstract class
     new 6c2702f  Uniform, Normal, Xavier
     new 9575156  Constructor for Initializer and handle exception
     new 794d75f  abort on any error in debug build
     new df0993e  Merge pull request #1087 from SkidanovAlex/debug_abort
     new 654840d  Added missing period and fix typo
     new 7f7c1c2  Merge remote-tracking branch 'upstream/master'
     new 24426c3  upsampling with summing over
     new 9a58ffe  Load initializer and mixed initializer
     new 8c64c44  monitor with pattern matching and sorting
     new 794232c  extra operator makefile fix
     new b3a08e9  cuda 6.5 compatibility fix
     new 4917155  fix Adam optimizer
     new fe41ffc  Merge pull request #1048 from piiswrong/master
     new bed8eaf  Merge pull request #14 from javelinjs/scala-package-cc
     new d5304a6  Adam bug fix
     new d9202e3  clean up gitignore
     new e01f6fa  modify linux settings
     new b7d510b  modify linux settings
     new 2a53b83  Merge pull request #15 from javelinjs/scala-package-cc
     new 84e00fd  modify linux settings
     new 8449c89  recover Makefile
     new c75baf0  merge from terrytangyuan/mxnet
     new 933d3d4  Merge branch 'terrytangyuan-master' into scala
     new c685425  Avoid scope in ifelse and for, changed var to val
     new 111158c  Fix error detector
     new 29c0e64  Merge pull request #10 from terrytangyuan/master
     new 2a671ba  Use apply() for Initializer instead
     new 1d988d5  Added note in
     new 539e6e3  Merge branch 'terry'
     new 48172b2  Used arr.set instead of use = to change reference, explicit return type for init methods
     new baf1279  Removed arguments in Initializer and its subclasses, add the arguments to init methods instead
     new acb82e2  Changed to set()0f without initializing a new array first
     new aea7c13  Merge pull request #13 from terrytangyuan/terry
     new 19f445e  rm debug codes and fix typo
     new 14f314c  add hsl color space aug
     new 7054f54  fix
     new 1f239ee  small change
     new 4710c7c  Merge branch 'scala' of into scala
     new 3a017c5  smll change
     new d91c39c  change NDArray writable flag and rm log txt
     new e74d226  Merge pull request #16 from yanqingmen/scala
     new 4a9a783  let `make clean` do clean_all
     new 202d0c3  Merge pull request #1109 from dmlc/mli-patch-1
     new 98cd435  add fcn-xs example for image segmentation refactor the fcn-xs example
     new cc0bc33  fix attribute set problem
     new f890b79  fix num_filter
     new 3a66df2  fix lint
     new bf90f63  update name
     new b8b9700  update for pr review
     new d91b5b7  fix HSV
     new 621573d  Delete test result file when finished
     new 5830796  update pr review2 fix description
     new 0826a1f  Merge pull request #975 from tornadomeet/seg
     new 5737ad7  Merge branch 'master' into master
     new 69b47c9  Merge pull request #1103 from winstywang/master
     new 397a96b  Format Scala codes according to scalastyle-config.xml (mostly follows Apache Spark)
     new aa82d39  Fix issue #1116
     new bc48a43  remove personal path
     new def4e70  scala-lint in travis
     new 14f1ca1  Merge pull request #1119 from tornadomeet/seg
     new 0927a28  Merge pull request #1117 from recastrodiaz/fix_char_lstm_example
     new ab87ca9  scalastyle add return check (warning level)
     new ce20e5a  Merge pull request #17 from javelinjs/scala-package-style-check
     new 2de3d28  add dilate for convolution
     new afdfac3  make Executor::SetMonitorCallback take std::function as parameter
     new bd84b2c  example/autoencoder fix optimizer param in layerwise_pretrain & finetune
     new def03c2  make MXExecutorSetMonitorCallback take an additional pointer as argument, allow frontend packages bind their closure objects
     new b82c741  Remove originally tmp error detector
     new 0b94a6a  fix wd in sgd
     new 79860f8  Merge remote-tracking branch 'upstream/master'
     new b24f6df  Merge pull request #1094 from javelinjs/callback-func-api
     new 2f29df1  Merge pull request #1113 from antinucleon/master
     new d586e5f  Remove one scala_lint in osx (same as tested in linux)
     new 887a6b7  Update .travis.yml
     new 930cd91  Symbol functions list, creator not implemented yet
     new 88ab916  Symbol function build with attr scope
     new ec16277  NameManager. AttrScope withScope move from object to class
     new 12b11a7  Symbol creator partially finished. AttrScopeSuite using Symbol.Variable
     new a41016e  Symbol plus tested, but not assert final result
     new f61d05c  Merge remote-tracking branch 'upstream/master'
     new e1aa386  FullyConnected, Activate and Group Symbol creation
     new c54d52b  fix lint error
     new 725d5a8  fix wd
     new 95f6ed9  fix deconvolution dilate default setting
     new 599df31  Merge branch 'master' into master
     new cfdcd9c  Reorder ldflags
     new 9e999f9  Include CMakeParseArguments
     new af4d53d  Merge pull request #1141 from wistone/master
     new b96749c  Merge remote-tracking branch 'upstream/master'
     new 72a2cde  Merge branch 'master' of
     new 0aa5ea1  Merge pull request #1142 from antinucleon/master
     new fbd3813  fix scale in Xavier initializer
     new 3f6db09  Merge pull request #1148 from lyttonhao/master
     new 222ee2d  CC to CXX
     new 82883fe  Merge branch 'master' into master
     new b02df00  Add Interpolation Options
     new b7da3d8  Merge pull request #1153 from superzrx/master
     new 396c895  add accnn tool
     new 0b66c47  Merge pull request #1154 from sodabeta7/feature-accnn
     new 841c6e7  5x Speedup of - lost ability to pass lambda function though
     new 8655b4f  imported dill for passing lambda function
     new 1e5e8c0  set argument '--gpus' to '0'
     new cb83007  Merge pull request #1165 from qiaohaijun/master
     new 5cd328c  [DOC] minor fix in installation guide
     new 35aee8d  Merge pull request #1166 from bryandeng/blas-typo
     new 47a19e3  Merge branch 'master' into master
     new 6116c93  Merge pull request #1145 from hiraditya/master
     new 1ab9104  Merge pull request #18 from javelinjs/scala-package-cc
     new 9430f20  NDArray handle ptr change to pure long
     new 35e898e  [example] kaggle ndsb-1
     new ffb269b  Merge pull request #1170 from antinucleon/master
     new 9e59f7b  Update
     new f127a69  Update
     new ee49338  Merge pull request #1171 from Lodewic/patch-1
     new 722c7a8  Merge pull request #19 from javelinjs/scala-package-ndarray-fix
     new f3b7a6a  change type of DataIterHanle and DataIterCreator to Long
     new 45f9f7a  Fix a spelling error in test script
     new d42f8eb  Put the onehot_encode test to main
     new feafd75  Add test for the new `fill_element_0index` operator
     new 14b6dd2  Add `TernaryOp` that takes three input: lhs, mhs, rhs. Also, add  a ternary operator `fill_element_0index` , which fills the specific element in each line of the `lhs` according to the index in `rhs` and value in `mhs`
     new 2e678a9  Add `TernaryOp` that takes three input: lhs, mhs, rhs. Also, add  a ternary operator `fill_element_0index` , which fills the specific element in each line of the `lhs` according to the index in `rhs` and value in `mhs`
     new 8274e60  Merge branch 'master' into pr_branch
     new e0f1252  small change for IOSuite
     new a0ad1e0  Merge pull request #20 from yanqingmen/scala
     new e70338b  fix bug
     new d9aeb99  Fixing ConcatOp when one of the inputs doesn't have gradient
     new d717c42  Merge pull request #1185 from sodabeta7/master
     new 0a1b901  Merge branch 'master' into fix1130
     new d249c90  Update Submodule
     new ea8e575  style fixes - spacing
     new 592b78e  ndarray type support.
     new f29723e  restore softmax grad scaling
     new 5da8558  symbolic backend type support
     new 18e68fe  Merge remote-tracking branch 'upstream/master'
     new 25d71a8  Merge pull request #1160 from cshaley/master
     new 61d0fdc  Merge pull request #1206 from sxjscience/submodule_update
     new a0b74a9  [IGNORE] add ipynb and input text
     new 1bead5b  [OP] Fix load of scalar binary op
     new 9be3284  Merge pull request #1215 from tqchen/master
     new d288851  python front end symbolic type support and cast op
     new a5ff977  Merge branch 'master' into type
     new e48aa60  Merge branch 'master' into fix1130
     new 446ff07  Merge pull request #1184 from piiswrong/type
     new 29d8638  Merge branch 'master' into fix1130
     new 48c736c  Fix lint
     new 73f6c46  Merge branch 'master' into pr_branch
     new e7514db  Merge pull request #1181 from SkidanovAlex/fix1130
     new b9301fc  fix issue 1224
     new 0f523e5  Merge pull request #1227 from piiswrong/type
     new b2b8cde  [CONTRIBUTOR] Add Alexander Skidanov
     new aec5714  Update
     new 1f19c86  [CONTRIBUTOR] Add Lodewic van Twillert
     new 6b86893  [CONTRIBUTOR] Add Aditya Kumar
     new 3638d8b  Update
     new eb4a571  Fix the previous building error
     new aea4ec6  Merge branch 'master' into pr_branch
     new 3130031  change KVStore handle to long
     new 63cd3c6  change Executor handle to long
     new 797352e  add resnet example for cifar10
     new 8242772  [R] documents update; force users to use latest version of DiagrammeR
     new b63cf23  Merge pull request #21 from javelinjs/scala-package-l
     new 3bde778  Merge pull request #1233 from thirdwing/master
     new 520e5c2  Merge pull request #1230 from freesouls/master
     new 10acc9e  python front end symbolic type support and cast op
     new c39f852  torch ndarray function backend
     new 903c6ef  more torch ops
     new f5c1b8d  torch criterion
     new 34b7807  fix random crash caused by torch_module.list_arugments
     new 968785b  Merge remote-tracking branch 'upstream/master'
     new a9d1423  Merge pull request #1213 from piiswrong/torch
     new 22bb168  Fix bug preventing from working when data-dir is an S3 uri.
     new 8978b30  remove MXNET_API in narray.h
     new 68d64e5  Merge pull request #1244 from hjk41/master
     new 08466f8  Merge pull request #1241 from svohara/mnist_example_bug_fix
     new e6d0c83  NDArray save & load
     new 120dd4b  minor fix network choice
     new 69c76ba  minor fix function name for AttributeError
     new e4b1e55  Merge pull request #1250 from qiaohaijun/patch-18
     new d9ec30b  Merge pull request #1249 from qiaohaijun/patch-17
     new 42dd1e4  ENH python metrics added, fixes
     new 02bc987  Merge pull request #1167 from jmschrei/python_metrics
     new ce16f37  remove current cpp train examples, will be rewritten with the cpp wrapper
     new 1d5a5cc  minor fix donot to don't
     new 7e9962f  Merge pull request #1258 from qiaohaijun/patch-19
     new 25f5a26  [io] sframe iter
     new 5e28159  Merge pull request #1251 from antinucleon/master
     new 28db218  NDArray all functions finished
     new e6a44da  rename
     new 4d085f1  [Symbol] fix bug in post order dfs
     new 883c7fa  Merge pull request #1261 from antinucleon/master
     new ee98fb7  Merge branch 'master' into master
     new 45ddeab  move plugin rules to sub makefile
     new 27f38ae  [op] fix
     new 59b6f6e  Merge pull request #1257 from hjk41/master
     new 161ab4a  Merge pull request #1265 from antinucleon/master
     new 07f726d  Use additional ptr to implement KVStore Updater callback
     new ac0435a  Merge branch 'master' into torch
     new f1bb68e  Merge pull request #1259 from piiswrong/torch
     new fb14c0b  Merge pull request #1256 from LinJiayn/master
     new 759a07e  [Torch] Fix compatibility for torch modules with view output
     new 0013f85  Merge pull request #22 from javelinjs/scala-package-l
     new 9d8a7f1  Correct typos.
     new 0c65e80  Merge pull request #1270 from piiswrong/torch
     new 6e494a9  Merge remote-tracking branch 'refs/remotes/dmlc/master' into pr_branch
     new 874e892  Update ternary function header
     new 949630e  Merge branch 'pr_branch' of into pr_branch
     new 187a288  fix OpInfo for concat and softmaxOutput
     new 1b98427  add script for download test data
     new 647bdb5  fix OpInfo for upsamplingOp
     new 4038077  Merge pull request #1273 from hjk41/master
     new 5f75b31  Merge pull request #1271 from weijianwen/correct-typos
     new 3e24441  [TRAVIS] merge doc check into lint
     new 5fccdb5  Merge pull request #1278 from tqchen/master
     new 262fe4e  Add LD_LIBRARY_PATH into search paths for
     new 98b2e77  Merge pull request #1279 from weijianwen/python-ldpath
     new c064769  add train_dataset and val_dataset to the argument
     new 13c88b5  Merge branch 'master' into pr_branch
     new f9b8e3e  fix jni typo
     new c27b705  Merge pull request #1282 from qiaohaijun/patch-20
     new 4c52eb8  workaround for #55
     new 724007a  add notes about AWS GPUs
     new 7fbfb55  del "?" from comment
     new 0fbf7f1  Update license year to range
     new 380996f  [PYTHON] Fix when ld_library_path do not exist
     new 54dcff0  Merge pull request #1292 from pra85/patch-1
     new 0d56423  Merge pull request #1291 from qiaohaijun/patch-21
     new c141013  [DOC] Add installation guide
     new e854a3f  Merge pull request #1294 from tqchen/master
     new 3bfce30  torch document update
     new 2c2e183  torch include fix
     new 2d068d8  Merge pull request #1295 from piiswrong/torch
     new 3319e5c  Symbol::bind roughly finished
     new ac777fb  merge upstream
     new a063e6f  Update
     new 2501611  Update
     new 2208243  Symbol::bind finished, not tested yet
     new 942dd88  Added prediction and generation of submission.csv. Minor changes to training to be in the same format as image-classification examples
     new 800b4eb  Merge pull request #1316 from Gelu74/kaggle_pred
     new 422957f  Adam(Optimizer) arg fix
     new ee21c52  Merge remote-tracking branch 'origin'
     new 620687d  adam try 2
     new 84551bd  Merge pull request #1318 from rfarouni/master
     new 54259df  add data-shape to argume
     new 145be05  doc for interface and impl of Theano scan
     new f8653ce  doc for api/impl for Lasagne RNN
     new 7da835d  torch lstm/rnn interface design doc
     new 13ca3f9  doc for cntk and tensorflow
     new 7bac143  l2 norm op && default value of NumOutputs
     new 4007afd  Merge pull request #1329 from qiaohaijun/patch-22
     new 1028d80  Merge branch 'master' into master
     new 3bec1c2  Add partial shape inference for recurrence support.
     new 959c1ee  rename rnn doc for naming convention.
     new 9633bd4  Merge remote-tracking branch 'upstream/master'
     new 4d257a5  add shape inference to symbol unit test
     new 427b87a  Symbol::bind and Executor finished (with unit tests)
     new e2d4a3d  fix lint errors
     new 4886598  Merge pull request #1310 from xylcbd/master
     new c6b31c6  fix devtools update
     new d5d48fd  Merge pull request #1346 from thirdwing/devtools_fix
     new 43e281f  fix #1118
     new f0bf6d4  Merge pull request #1349 from thirdwing/master
     new 145af38  Merge remote-tracking branch 'upstream/master'
     new b4c91dd  add orthogonal init
     new 10f1d5a  Merge pull request #23 from javelinjs/scala-package-l
     new a18fdfb  fix some typos, comment mismatch, and small code optimization
     new 0a7c230  Merge pull request #11 from terrytangyuan/master
     new c357c4c  Fixed mean calculation in cpp example for single channel images
     new 461e534  small change
     new 49cd715  Merge pull request #1348 from javelinjs/py-typo-fix
     new d5c77b2  Merge pull request #1351 from antinucleon/master
     new 70c8e2c  small change
     new 6739dee  Added graph
     new 9629f22  Corrected identation
     new 9844ebe  Merge pull request #1350 from Gelu74/evaluation
     new f9c1f09  Merge remote-tracking branch 'upstream/master'
     new 6d54253  Merge branch 'master' into fix_single_channel_cpp
     new 33f4432  Copy-edit aws documentation file.
     new 5b041d2  Merge pull request #1357 from dansbecker/first-docs-revision
     new 8a4a75e  Update
     new e53c3ea  add provideData provideLabel for MXDataIter
     new 5aa75be  fix bug with zip in mixed initializer
     new 70ddbaf  Merge pull request #1361 from ppwwyyxx/patch-1
     new ee5eefc  Fix NDArrayIter.__init__()
     new 20f7523  Merge branch 'master' into master
     new a54e6b9  Update
     new 0c52322  add ImageRecorderIter
     new 10993a7  merge from master
     new cc54bca  small change from scalastyle
     new d86cc42  small change for IOSuite
     new c15d2c4  Edit main README and doc/ for grammar and style
     new 120f8af  Merge pull request #1367 from dansbecker/master
     new 665de43  Merge branch 'master' into master
     new 3933be5  Merge with upstream
     new 9979ca5  Minor modification of CONTRIBUTORS
     new 9a60d20  Merge pull request #24 from yanqingmen/scala
     new b649974  Update
     new 101c68b  Merge branch 'master' into master
     new c1fb6b6  Merge branch 'master' into pr_branch
     new 805edf2  Merge pull request #1180 from sxjscience/pr_branch
     new 050c572  Tested conv net on MNIST (cpu).  (Train-accuracy=0.8868562, Time cost=57.118s) v.s. python (Train-accuracy=0.942483, Time cost=57.206s). Probably need to adjust the SGD params to achieve the same training accuracy.
     new 276fc9d  Fix model training stop condition
     new 926edbb  add Symbol::toJson, symbol clone tested
     new 1111deb  Merge branch 'master' into master
     new 2d9e6de  Merge pull request #25 from javelinjs/scala-package-l
     new 8d8188b  Fix training bug, sgd bug, initializer bug
     new 85d2aeb  Replaces 'stytole' with 'systole'
     new cd28d4b  Adds Woolsey's preprocessing suggestion to allow reprocessing in case of crash, from\#post103854
     new b6d3bb5  Merge pull request #26 from javelinjs/scala-package-l
     new 5bd3db5  fix typos and spelling errors
     new e51699a  fix typo & better path handling
     new 9fd9d6c  Merge pull request #1386 from darxriggs/fix-spelling
     new d4a80d4  Merge pull request #1387 from ppwwyyxx/patch-2
     new 4616e8b  Merge pull request #1384 from fionnan/kaggle-dsb2-edits
     new 0482813  Merge branch 'master' into master
     new dd868c6  Change optimizer interface, and make FeedForward take Optimizer as param
     new 0119c5b  finished trainMultiDevice & fit
     new 6ff3a49  enable registering operator return type
     new 2021b8e  Merge branch 'master' of
     new a411589  enable specifying operator return type
     new a03ee60  add default value for return_type in MxFuncGetInfo so it is compatible with old code
     new f1e175d  Merge pull request #1364 from Vesnica/master
     new 9df79b0  Merge branch 'master' into master
     new 184e446  fix api problem for R and python
     new 89fac44  Merge branch 'master' of
     new 4cf895c  FeedForward::predict finished
     new 13ac8c6  Minor formatting in doc/
     new dbddb81  Merge pull request #1397 from dansbecker/install_docs
     new b22a459  Merge branch 'master' into master
     new 5397524  fix lint
     new 61b6e5c  Merge branch 'master' of
     new f7b9d3d  fix lint
     new 3899265  minor
     new eeecc52  Merge pull request #1390 from hjk41/master
     new 0b597c8  Merge pull request #27 from javelinjs/scala-package-l
     new a04c8dd  update use italic title to better align logo
     new 95aae93  add jenkins status to
     new 9cd308f  add gcc-5 and openmp support for osx, but ld error for x86_64
     new bc0d868  remove openmp for mac os x due to opencv problem
     new 810bc43  add gcc-5 and openmp support for osx, but ld error for x86_64
     new 200ebd3  remove openmp for mac os x due to opencv problem
     new 53d13af  add Awesome MXnet page
     new 077a482  ccsgd optimizer pickle support and variable wd
     new e75c504  CuDNNBatchNorm init parameter
     new f8e3ecf  fix initializer and metric
     new 800f740  enable evaluation on gpu
     new 581a2bf  add fix_gamma for cudnnbatchnorm
     new 2716364  ndarray image decoding
     new e1d5484  better error message
     new b3dae09  Include locals() in BatchParam
     new 30de1ca  update mshadow
     new ca1029b  remove eval_iters
     new 58ac3ae  lint
     new 7d2255e  Merge pull request #1405 from phunterlau/master
     new 343e560  Merge branch 'master' into master
     new c449161  Merge pull request #1309 from piiswrong/master
     new ee5bda7  [GraphExecutor] Shared Memory bind and runtime reshaping
     new 2ba9356  add test cases
     new cc8f91f  make dummy color constant
     new c2905fd  Merge pull request #1408 from piiswrong/master
     new c0654e2  small change for DataIter's Iterator implementation and add DataPack
     new 77e07c0  fix cudnn batch norm
     new 590a7b7  Merge pull request #1425 from piiswrong/master
     new b8c3ec1  small changes and reorganized io classes files
     new 5e5616f  fix bugs
     new 5239d6a  Merge pull request #1426 from piiswrong/master
     new 77e79c9  Merge pull request #28 from yanqingmen/scala
     new 504564c  Merge remote-tracking branch 'upstream/master'
     new 208e266  fix typo
     new 18c6122  fix compile errors due to submodule versions
     new c5cbf24  fix seg-fault due to upstream API change
     new 175d399  workaround for jenkins CI
     new f277cd7  revert last commit (not useful)
     new 30b7142  fix one travis CI test error
     new 706e508  Merge pull request #1334 from pluskid/master
     new 1e4f9f0  fix error due to upstream API changes on symbolic node function
     new 579b881  fix for breaking changes in Julia nightly
     new 04ea797  Change NDArray shape's type from Array to Vector, add Random unit test cases
     new 0d683bb  add NDArray equals, add unit tests for ElementWiseSum and Concat
     new a0559bc  Implement NDArray hashCode
     new ce3f6d3  another attempt to fix julia v0.5 compatability
     new 938b236  update doc for initializer
     new 06019e2  stricter type anotation for julia v0.5
     new 1c707a2  fix typo
     new ac709d1  prepare for v0.0.8
     new e63925d  executor reshape bug when grad_req='null'
     new fca833f  Merge pull request #1433 from lukemetz/lm/reshapebug
     new 33aa94d  fix bilinear initializer
     new 09c3144  fix opencv is unavailable message
     new 6e24a34  refactor concat and slice op
     new 27a3382  fix convolution / deconvolution grouping
     new f14ca37  fix
     new d90cd95  fix obsolete docker info
     new 16b8f70  enable testing of the Julia bindings
     new 9cf857b  add Symbol::SwapAxis, tested
     new f475b45  Merge pull request #1379 from oist/vc/test_julia
     new 67dc44c  Merge pull request #29 from javelinjs/scala-package-l
     new 18487d9  add to
     new 3dad814  Merge pull request #1441 from dmlc/mli-patch-1
     new 82eaeb5  Update
     new f07172b  1 lua_State per Operator to fix GC issues
     new 9660a9e  Merge pull request #1430 from lukemetz/lm/torch
     new 6eef9f2  Merge branch 'master' into master
     new bc111b5  Merge pull request #1435 from piiswrong/master
     new dd7349f  cudnn v4 production compatibility
     new bd639ac  put executor manager in a separate file
     new a4a7ad7  refactor executor manager to allow multiple executor groups
     new 575c5f6  sanity check with bucketing method using a trivial MNIST example
     new 21c8f85  add bucket-key to existing DataBatch API
     new 2b9dcfa  also share arrays for data&label if possible
     new 1c09699  fix some lint errors
     new c5dfa83  fix lint error
     new 5596b65  more doc for executor group
     new fc69387  actually use bucketing......
     new fa32210  lstm ptb example
     new 024e591  lstm-ptb bucketing example (to be tested)
     new 81a7076  Merge branch 'master' into master
     new 0b387f7  add further information to install.rst
     new d23605f  Merge pull request #1449 from piiswrong/master
     new 8f6ba0d  AdaDelta optimizer
     new 77bb9f0  Merge branch 'master' into master
     new b8f795d  Merge branch 'master' into master
     new 7f2ca60  some debugging code
     new d67743f  Merge branch 'master' of
     new d8cde1e  mnist sanity check use different batch size for different bucket
     new ba1a761  Merge branch 'master' of
     new e231846  import utility functions back into
     new 3b5826a  check shape with shared aux-states.
     new 2c520a2  fix bug in ptb bucketing data generation
     new 7d547e5  fix lint error
     new 6df263f  Merge branch 'master' of
     new 42ca1ab  Merge pull request #1450 from pluskid/master
     new 2cd4cc2  scripts for nightly test
     new ad429e6  Merge pull request #1460 from mli/master
     new 2b946a1  add inception-v3 network to choice
     new 7d6e354  add inception + cifar10
     new 5d506ae  not only cudnn v3 is supported
     new 30cc5a3  Merge pull request #1463 from qiaohaijun/patch-23
     new adb963b  add symbol scalar & pow operators. Tested
     new bf3c0ef  remove previous debugging code
     new adebc8f  doc for bucketing
     new b1905aa  [test] use 200 epoch and 256 batch for inception
     new 57883d1  add distributed test to nightly, remove tests/python/multi-node/*
     new 66c46c4  Merge pull request #1469 from pluskid/master
     new 983b91f  Merge pull request #1465 from mli/master
     new c6a5f11  Merge branch 'master' into patch-24
     new 3fa86ed  workaround bug in VS2013. VS2013 is not able to deduce the type of the parameter passed to make_shared in this case
     new 46e28cc  Merge pull request #1475 from hjk41/master
     new 00e830e  Merge branch 'master' into patch-24
     new b46677f  add some unary symbol operators, tested
     new 40f7131  Merge pull request #1438 from tengpeng/patch-2
     new 42ae53e  Merge pull request #1466 from qiaohaijun/patch-24
     new 02adbec  fix cudnn layers destruction
     new ef8b108  refactoring: merge symbol argument with sym_gen in FeedForward
     new ed9721b  update doc on gen_sym
     new 6ff1047  Merge pull request #1481 from piiswrong/master
     new afc5977  Merge branch 'master' into master
     new 0136c24  add /MP to speed up compilation on Windows
     new d6bd397  Merge pull request #1453 from Azure-Vani/master
     new 37ed3c4  Merge branch 'master' into master
     new 71ba7cd  Merge pull request #1484 from hjk41/master
     new 33d79ae  add mutliprocessing
     new 1220139  presever image order for multiprocessing
     new 9dc35ad  fix lint error
     new a0fed04  fix typo
     new 16b99f8  Merge pull request #1493 from nebw/patch-1
     new ec3c446  fix lint error
     new 4b941e9  Merge remote-tracking branch 'upstream/master'
     new c21566f  Merge pull request #1483 from pluskid/master
     new 1f9df8a  Fix possible macro expansion error
     new d020d43  add mutliprocessing
     new 71f28ce  presever image order for multiprocessing
     new 8c232e2  Merge branch 'master' into master
     new 823e614  Merge branch 'master' of
     new 1884eb7  Symbol operators finished
     new 737c840  update docs. close #1498
     new 07e907b  Merge pull request #1500 from thirdwing/master
     new 1a111d5  Merge pull request #1497 from sxjscience/bach_norm_macro_expansion
     new c3180fb  Add AppVeyor script
     new 358e21e  Merge branch 'master' into master
     new 0ec25d3  Merge pull request #1495 from hjk41/master
     new 37fa56b  refine the symbol create APIs
     new f26ad96  Update install.rst
     new c04cc39  Merge branch 'master' into master
     new a1b0748  Merge pull request #1488 from junranhe/master
     new ebb43f8  add BlockGrad, Crop, ... operators for Symbol
     new 164ef82  Merge pull request #30 from javelinjs/scala-package-l
     new 45c2ffe  Merge remote-tracking branch 'upstream/master'
     new 3e15f62  fix travis scala test config
     new 3f441c3  There are two places which need to change to the python2 syntax
     new 2452d87  integrate scala-package compiling into ROOT/Makefile
     new c2ff0f0  Merge pull request #31 from javelinjs/scala-package-l
     new 180ae64  add ModelParallelSuite to test model on multiple devices
     new 8cc802a  Merge pull request #1511 from zhmz90/ndsb2
     new 0475e9b  Update readme
     new 28c60b7  Merge pull request #1515 from rupeshs/patch-1
     new c8568c1  Implement MAE, MSE, RMSE, CustomMetric
     new d8bcbae  Add example module. Run scalastyle check in verify stage
     new 5cdb5b0  Make maven failure return -1. Fix linux link problem
     new b545a96  Merge pull request #32 from javelinjs/scala-package-l
     new 09cb2bd  MXNet learning notes in chinese
     new fe49114  Merge branch 'master' into master
     new 94b4c52  Enhance example/TrainMnist. Add missing mxKVStoreIsWorkerNode implement
     new 744b4a2  update with intel-mkl 2016.1.30.450722
     new 6a1d96c  Add learning note in chinese link
     new 8249275  example/TrainMnist finished, tested on multiple cpus
     new 70d6124  Merge pull request #33 from javelinjs/scala-package-l
     new bc6c8b1  add mkl version > 2016.2.062 as a separate notice
     new fcdcafb  Change Shape from Vector to an independent class
     new fc4d397  Merge pull request #1519 from zhubuntu/master
     new 40d348b  Merge pull request #1527 from liubenyuan/patch-1
     new 1f84bc2  Merge pull request #34 from javelinjs/scala-package-l
     new 4366e9c  Merge remote-tracking branch 'upstream/master'
     new 5ffa9b7  add finalize() for Symbol & KVStore
     new eb0d973  Create the Bayesian-Methods example folder and add SGLD + BDK ipython notebook examples. Also, an SGLD optimizer is added to ``.
     new b52dc11  fix the type error in autocoder example
     new 96a1beb  Fix missing DeleteLocalRef. Get JNIEnv from global jvm
     new c5cf6b9  add notify shutdown
     new 35b2607  jvm shutdown hook
     new c740d65  Merge pull request #1499 from sxjscience/bayesian-dark-knowledge
     new 96fbc08  include stdint.h since uint64_t is not defined in c_api.h
     new cd2fbc9  Merge branch 'master' of
     new febd9ab  Merge pull request #1544 from hjk41/master
     new ffa6ab4  Merge branch 'master' into master
     new 2d7e4fe  add destroy method to NDArray, Executor, Symbol, KVStore. And destroy useless objects automatically during training
     new bc9fbe0  fix memory leak in SGD. destroy updater when training finishes
     new 223fd62  Merge pull request #1540 from shallyan/master
     new ed98764  remove dependency on stdint.h in c_api.h
     new 00f1d8e  fix a typo in comment
     new 789ac5e  Add javadocs for native memory dispose. Release kvstore at the end if it is created in fit().
     new 4b4e7b5  fix some typos in doc
     new 6c37230  Merge pull request #1555 from ziyeqinghan/master
     new 2956813  Merge pull request #35 from javelinjs/scala-package-mem-leak-fix
     new 1995984  Fix #1323.
     new 0777032  JNI code lint
     new 7351176  Merge pull request #36 from javelinjs/scala-package-jni-lint
     new 3c53488  Minor change for friendly screen-printing. See PR #1554
     new 27d0f9d  Merge pull request #1554 from Puriney/master
     new d9190ad  Update
     new 5268737  Merge pull request #1572 from terrytangyuan/patch-1
     new 0671941  update adagrad optimizer
     new 71721ec  modify the style in adagrad optimizer
     new 587f2a3  assembly package
     new 60a89c4  add Build,Usage parts to scala-package README
     new 9e5137f  Merge pull request #37 from javelinjs/scala-package-l
     new 6c38241  Added prod
     new 2bfca5f  Merge pull request #1587 from Gelu74/Xavier
     new 3f40cd4  Fix MXDataIter NPE bug. Throw NoSuchElementException when calling next() at the end of the dataItet, follow the same way as java
     new 16012f0  Merge branch 'master' into master
     new 0f50e29  Monitor callback
     new 15864ff  Merge pull request #38 from javelinjs/scala-package-l
     new bbc4e0e  Merge remote-tracking branch 'upstream/master'
     new 9965357  trigger all test cases
     new dbce1b3  Merge pull request #1592 from javelinjs/scala-package-l
     new 2688b68  allow directly use eval_metrix=mx.metric.Accuracy() for
     new 7b1dbd6  fix bug for MAE,MSE,RMSE
     new 9aea38a  fix for metric MAE, MSE, RMSE: should not scale with different batch sizes
     new afd0a79  Modified R-package clip_gradient to use same ctx
     new eade183  Modified clip grad ctx device R-package clip_gradient
     new b428bcb  Merge pull request #1589 from Gelu74/clip
     new 1e4811f  Merge branch 'master' into master
     new e4c8e29  Merge branch 'master' into fix-metric-error
     new f2e7cca  Update
     new 6f699d0  Update
     new c54e792  Create
     new 5d02f56  Merge branch 'master' into patch-2
     new 4b8d989  Merge pull request #1597 from wac81/patch-2
     new f91b2e5  Update
     new e20de1a  Merge branch 'master' into master
     new 5c7cb65  Added unittest for convolution dilate
     new 5c05f3f  Fix for Issue #1457 (Convolution with Dilute is erroneous)
     new 32375b8  Merge pull request #1580 from linmx0130/master
     new 17f0910  Update
     new bdfcb77  Builder constructor for FeedForward
     new db208db  Merge remote-tracking branch 'upstream/master' into scala-package-l
     new 850719a  Merge pull request #1606 from wac81/patch-3
     new ff15fae  Merge pull request #1582 from taoari/fix-metric-error
     new 40a8d5e  Merge pull request #1607 from javelinjs/scala-package-l
     new 6df6d58  Fix typo. Ref:
     new 0e2af19  Merge pull request #1610 from Puriney/master
     new aea9324  add myself to contributors
     new c7a58fd  add CrossEntroy loss
     new a179a58 MultiFactorScheduler
     new 67b5cf6  Merge pull request #1616 from taoari/fix-metric-error
     new 3cd7e9c  [Scala] + AdaGrad optimizer
     new b465399  [Scala] + RMSProp optimizer
     new 0d6d154  [Scala] + AdaDelta optimizer
     new a9934e8  [Scala] * Fixed forward reference for grad
     new e2d3a18  [Scala] * Fixed potential memory leak
     new 827541a  Merge pull request #1612 from terrytangyuan/optimizer
     new 49396a7  fix travis
     new 7967563  cnn text classification example using python api
     new 5f63111  update readme typos
     new 2ea129b  add corpus download infomation.
     new 10ecd0e  fix a bug in array correspondence in bucketing (dmlc/mxnet#1615)
     new ad12f52  Merge pull request #1624 from pluskid/master
     new 5f41e36  rename the folder & update example readme
     new df0328b  Merge branch 'master' into master
     new d17c360  Merge pull request #1621 from qcl6355/master
     new aafc767  Fix typo.
     new ca49c27  Merge pull request #1629 from heyf/patch-2
     new a3aebfc  DisposeDeps for NDArray. Fix memory leak for AdaDelta & AdaGrad. Fix implement bug for AdaGrad
     new 29b95af  Adam memory leak & impl bug fixed
     new 9eadd2f  fix RMSProp memory leak
     new 6cd1c0a  more memory for travis scala-test
     new b146967  Merge pull request #1640 from javelinjs/scala-dispose-dep-nd
     new bad7bfe  allow FactorScheduler factor=1
     new 036bd09f Merge branch 'master' into cross-entropy
     new f74d74d  Merge pull request #1617 from taoari/cross-entropy
     new b39226b  add ResizeIter and NDArrayIter
     new 96128ea  Merge branch 'master' into scala
     new 771a2d6  Merge pull request #1646 from yanqingmen/scala
     new 0c08aca  fix amalgamation. close #1660
     new d54ab27  Merge pull request #1661 from thirdwing/master
     new 34b36cf  [Doc] fix broken links in doc. close #1663
     new e0363e0  Merge pull request #1664 from thirdwing/master
     new 3d29439  new document under docs/
     new d2505a3  remove wrongly commit README
     new 945e6e9  lint .cu codes
     new 8f32b90  Separate load and save model prefixes
     new 6d8ef44  Support evaluating multiple metrics in model training
     new 1ec085a  [docs] fix no search results bug
     new c5cbb51  [docs] add package documents
     new 0d25d75  [doc] fix python docstring
     new fc922f0  Merge pull request #1668 from javelinjs/lint-cu
     new b0d3255  Fix Adagrad Optimizer (Add lr_scheduler and gradient_clipping)
     new 8bf8f71  Update
     new 569c486  added a verbose option to `mx.mlp` in R package. close #1608
     new bca3780  Merge pull request #1679 from ziyeqinghan/patch-1
     new bdbd217  Merge remote-tracking branch 'upstream/master'
     new ef47a24  Merge pull request #1633 from futurely/multiple-metrics
     new e688fe8  Merge pull request #1352 from matthill/fix_single_channel_cpp
     new b230159  Merge branch 'master' into master
     new 52ea0f0  Merge pull request #1648 from futurely/load-save-prefixes
     new 0e542e9  Merge pull request #1670 from mli/master
     new fef710e  Merge pull request #1678 from sxjscience/fix_adagrad
     new 53cc2a1  improve data-iter performance for bucketing-rnn example (dmlc/mxnet#1644)
     new ad878ac  fix some pylint warning
     new 522a591  fix the rest of pylint warning
     new 8881061  Merge branch 'master' into master
     new 62bc5c5  Merge pull request #1680 from thirdwing/master
     new dc6c326  [R] update mx.symbol.Concat API. close #1578
     new 2bb4034  set deconv layer no_bias default to true
     new 50d7715  Merge branch 'master' into master
     new fdd5381  Merge pull request #1681 from pluskid/master
     new 8a02cdb  add install doc for Julia to "get started"
     new fd0d4fe  Merge branch 'master' of
     new 2711191  Merge remote-tracking branch 'upstream/master'
     new 81a8bca  Merge pull request #1689 from pluskid/master
     new 76a2e73  add batch callback parameter to image training examples
     new 7d8c378  script for testing mirroring memory consumption
     new abb526b  [OP] Fix AddTo data in ConvOP
     new 00d3016  [EXAMPLE] Fix model prefix
     new 4c8ad0e  fix mshadow
     new 58d6aa7  Merge pull request #1696 from tqchen/master
     new 633a178  [EXECUTOR] Efficient gradient memory aggregation for RNN
     new 8f29787  [DOC] Add env var doc
     new 0ddfd10  Fix lr_scheduler of AdaGrad and RMSProp. Also add clip_gradient to the base class. See the discussion in
     new 19f46ef  Merge pull request #1699 from sxjscience/fix_lr_scheduler_grad_clipping
     new d0c88e3  Merge pull request #1697 from tqchen/master
     new f48599d  Fix addto for embedding gradient
     new 9b92123  Merge branch 'master' into master
     new c8276c8  a convenient interface for getting attr
     new ff49ad5  use attributes to decide whether mirroring is enabled
     new bfc32f7  use attributes to set force_mirroring
     new 518683f  Merge pull request #1682 from thirdwing/master
     new 6069b9a  some comments on the cifar10-mirroring test case
     new 8032474  Merge branch 'master' of
     new e486efb  Merge remote-tracking branch 'upstream/master'
     new 5224f6f  Merge pull request #1707 from chaoyuaw/master
     new d5e091e  createoperatorEx
     new 4c61cef  multi-type convolution layer
     new d05e44f  fully connected layer type
     new 5892d40  activation typed
     new a841a57  Merge branch 'master' into master
     new e4fa6ff  fp16 layer refactor
     new f1afe73  Merge pull request #1675 from piiswrong/master
     new 19edcf7  Add support for multiple outputs in
     new 6590ee3  add a multi-task example
     new 844da7f  fix lint
     new 95e8750  fix the implementation bug of "mae" and "mse".
     new 05e5ac2  fix implementation bug of mae and mse
     new abec5c3  Merge branch 'master' into master
     new d81adfa  Merge pull request #1711 from Ldpe2G/master
     new 45ee47c  update contributors
     new 5456e86  Merge pull request #1719 from javelinjs/master
     new e24c892  Merge branch 'master' into master
     new c07d899  Merge branch 'master' into develop
     new efdb19c  merge PR #125 by hand
     new cb73777  [docs] update index
     new 05c764a  added missing source file include.
     new 7e93232  split cpu & gpu version
     new 48503af  plugins for deploy to oss
     new ad2e47f  deploy for scala core module ready
     new c393172  deploy scala full jar
     new c2c8540  change proj version to 0.1.0-SNAPSHOT. dev tools for change scala version & proj version
     new bbc043b  maven pom headers
     new db9f03f  libmxnet-scala artifactId with platform
     new 2f2b428  force generate javadoc & source.jar, release 0.1.0
     new a870484  try find os-xpu libname
     new c1c2484  [Scala] Exclude example module from deploy. Update
     new 949d815  [Scala] disable GC limit for travis test
     new 1d3e8c2  [Scala] disable /etc/mavenrc in travis
     new c7732da  [Scala] Release 0.1.1
     new e712926  Merge branch 'master' into scala-mvn-repo
     new f3c392e  Merge pull request #1703 from javelinjs/scala-mvn-repo
     new b5e17a8  Merge branch 'master' into master
     new 4902462  [docs] update index
     new b88d011  [docs] update get_started
     new ebd95bd  Merge branch 'master' of
     new 4a5781e  Merge pull request #1726 from mli/master
     new c866df6  [docs] add docs/system/
     new 6545537  Merge pull request #1728 from mli/master
     new 312c17d  Merge branch 'master' into patch-1
     new 8278a14  Merge branch 'master' into master
     new 18d1898  Merge pull request #1709 from pluskid/master
     new c0daa19  [docs] add zh docs
     new b89832b  [docs] add favicon and fix index html title
     new c7af796  [Scala] add scala-doc.jar for deploy
     new 4f680ce  Merge pull request #1733 from javelinjs/scala-mvn-repo
     new a284644  Fix broken link in bucketing document page
     new 5ff545f  Merge pull request #1736 from eric-haibin-lin/master
     new e6a7faf  add Scala document #1727
     new 7566335  Merge pull request #1737 from javelinjs/docs
     new ae63658  Merge pull request #1732 from mli/master
     new 27a66bc  Merge pull request #1627 from zhaw/master
     new 41335ea  Merge pull request #1723 from haria/patch-1
     new d02e03d  [EXECUTOR] Fix mirror strategy by enforcing mirror of attr
     new 0b1ba5b  Merge pull request #1738 from tqchen/master
     new 0f621a9  Fix SoftReLU action when compiled with cudnn
     new f6736ce  remove print shape form get_output
     new 1edea31  Merge pull request #1745 from haria/patch-2
     new 9d31d8f  Merge pull request #1744 from sxjscience/fix_softrelu_cudnn
     new 9fcf7a2  GTC Tutorial
     new 8668c30  tutorial update
     new 026432e  add EXTRA_OPERATORS with cmake add /EHsc with msvc
     new 81988b4  cudnn v5
     new a72e39a  Merge pull request #1741 from piiswrong/tut
     new 68890c2  Merge pull request #1742 from antinucleon/master
     new ce89474  Merge pull request #1641 from yajiedesign/addcmake
     new 715b570  Small Change in R Coding Style
     new de965ba  Merge pull request #1757 from XD-DENG/master
     new 147dfad  Update
     new 940a93f  Fix potential macro expansion problem on Visual Studio
     new 1bcb5dd  rnn
     new 971f2df  Merge pull request #1756 from sxjscience/fix_macro_expansion
     new b045f3f  Merge pull request #1760 from antinucleon/master
     new 1fc858e  neuralart ipython notebook
     new 5a6dc78  fix
     new 487fe4f  Fix Crop layer bug which only takes 2 inputs, with improved doc
     new c26cb3e  Fix small problem in RNN example
     new 4067145  Update RNN Readme
     new 7fe4b32  add elewise add/sub/mul/div function for ndarray
     new 26d30e8  add full operation
     new ae1b621  add transpose operator
     new 6f808c8  fix subtract
     new 521659f  [sum-partial] Summation wrt row and col.
     new e76d04c  [sum-partial] Rename and use runtime check (no C++11 support in CUDA).
     new 0f6e64c  [sum-partial] Add sum_mid semantics.
     new efd8d58  [sum-partial] Do sum.
     new e0b7c8e  [broadcast] Broadcast single axis.
     new ba4a00d  [broadcast] Add C API.
     new 1ae4a87  [broadcast] Fix bug and broadcast.
     new ff6970d  broadcast for arithmetic operations
     new 5566db5  rewrite broadcast into ndarray function
     new bd07554  [master] Compatible with Python 3.
     new 5abcb78  1. allow left arith with numpy numerical value 2. fix broadcasting semantics
     new 717ce39  Acoustic model training example. python_wrap: Kaldi wrapper. io_func: IO function for speech features. utils: utility functions. Now the example is TIMIT, need to change to AMI (free).
     new f786095  Merge pull request #1770 from sxjscience/smallfix_rnn_example
     new a249cfa  fix lint, and update submodules
     new 30e93e9  sync ps-lite submodule
     new 753f0e5  Merge pull request #1761 from piiswrong/tut
     new e75660f  Merge pull request #1659 from lyttonhao/develop
     new 66bd728  [EMBED] improve speed
     new 9ea9692  Merge pull request #1776 from tqchen/master
     new f6d425a  change sum_internal back to sum; add workaround for name duplication
     new 0e576a1  update py2nb
     new 8724bab  Merge remote-tracking branch 'upstream/master'
     new e0b420b  Merge pull request #1777 from antinucleon/master
     new d4d14d6  fix lint
     new 087db50  Merge branch 'master' into merge
     new e04ba1d  Merge pull request #1774 from jermainewang/merge
     new 116e94e  [MODEL-PAR] Check in original code to allow diff
     new 68513c0  [MODEL-PAR] Add example test code for model par lstm
     new 37af0ac  fix bug in shape inference of FullyConnected (dmlc/mxnet#1778)
     new ccea824  Merge pull request #1782 from pluskid/master
     new 677f067  add arg.params,aux.params for mx.model.FeedForward.create function [#1543]
     new c1e49e6  Merge pull request #1571 from ziyeqinghan/iss1543
     new 93a8c87  add function mx.exec.update.grad.arrays to clear gradient array
     new a12ead8  Merge pull request #1673 from ziyeqinghan/lstm
     new b0bb51d  Merge pull request #1781 from tqchen/master
     new 07e1000  fix typo
     new 5bb759e  a basic SymbolModule
     new c25315d  update parameter for modules
     new 759d9b6  rename mod to module
     new 127af36  a simple test file for module
     new 22df12f  normal training results on MNIST
     new 4d1764d  sync parameters from device
     new b4c3ff9  high-level API fit.
     new ca1ddd9  high-level API score
     new 9e9c418  high-level pred API
     new 4b29cd4  bucketing module
     new b780f30  an (untested) example of lstm-bucketing using module
     new 72bed5a  move MNIST module testing to example directory
     new 3f6b2f2  use data from existing rnn directory
     new 846fa7a  fix bugs in ptb bucketing example with module
     new 002a72a  cifar-10 example
     new 40ebab9  load and save models in module API
     new d85326e  makes logging information similar to before
     new 1455254  fix model loading and saving
     new 7a50357  MXNet exposes a transpose function
     new fbcd12c  Merge pull request #71 from oist/vc/imports
     new 948bbdf  add the mnist example in R
     new 173ea1e  Merge pull request #1562 from ziyeqinghan/master
     new 451dc6a  Fix sync-dir option
     new 1961b0d  [doc] remove dmlc/forum on readme
     new 43d4288  Merge pull request #1803 from dmlc/mli-patch-1
     new a739de1  Update
     new 1a61815  Merge pull request #1805 from dmlc/piiswrong-patch-1
     new f898837  add function to get debug_str, similar to python
     new ee1d3c6  fix the wrong requirement of FactorScheduler in LRScheduler.scala (#1807)
     new f7562e9  Merge pull request #73 from oist/vc/executor
     new bb6f1b0  Merge pull request #1800 from nitnelave/patch-1
     new acd2a74  output TempSpace allocation
     new 34efa24  Merge pull request #74 from oist/vc/temp_space
     new aecb367  Merge pull request #1765 from ZhreShold/master
     new 2a06c45  update tools/ to use launcher=local if no hostfile is given
     new e1ba0c2  bug fix in kvstore when value size is larger than MXNET_KVSTORE_BIGARRAY_BOUND
     new e6c767c  add kvstore test with 4 gpus at tests/nightly
     new cf55e26  Make ccSGD behaviour match documentation w.r.t clip_gradient
     new c431458  Serializer test for optimizers. Link libps.a
     new 391ddee  helper handle for kvstore server controller
     new b84d461  dist training tested with version 0.1.2-SNAPSHOT
     new 0a5f9ea  update scala doc
     new 89a675f  Merge pull request #1786 from javelinjs/scala-dist
     new 5fc7595  Merge pull request #1817 from dmlc/mli-patch
     new eecdd47  Merge pull request #1820 from oist/vc/ccsgd
     new cb0ce87  add test case for training with various kvstore w/ muli-gpus
     new 34c9e3f  fix typo
     new 64e7a54  a basic SymbolModule
     new 2253cfa  update parameter for modules
     new 94482b2  rename mod to module
     new d3cc9ee  a simple test file for module
     new 072ae87  normal training results on MNIST
     new 710f116  sync parameters from device
     new 7d40280  high-level API fit.
     new fdf484c  high-level API score
     new 92aba1b  high-level pred API
     new bd0c9d6  bucketing module
     new 05ec8fb  an (untested) example of lstm-bucketing using module
     new 9405afd  move MNIST module testing to example directory
     new 9035088  use data from existing rnn directory
     new 0bd4e35  fix bugs in ptb bucketing example with module
     new 9e442e8  cifar-10 example
     new 75875d7  load and save models in module API
     new b8f4dfd  makes logging information similar to before
     new 9538b58  fix model loading and saving
     new cd50e96  doc for DataParallelExecutorGroup
     new a8ca724  doc for BaseModule
     new 419ecab  docs for Module
     new 562de61  complete doc for current module
     new a47a752  fix lint errors
     new a63e9d7  fix lint error for executor_group
     new 6a9af81  add out_grads to backwards of modules
     new 958c52c  draft module API
     new 2016775  fix mnist example
     new 5362a0a  update API doc
     new 6565aeb  fix more lint errors
     new 18cc1ce  specify all APIs at BaseModule
     new a8f9e34  refactor Module for the latest module API
     new aea9374  mnist now runs on latest module API
     new ecfe83f  re-organization and fix lint errors
     new 92ea2ce  fix lint errors for BucketingModule
     new ba0617a  remove temp doc on module
     new c3d3fba  bucketing module in a separate file
     new 315a30e  Merge branch 'module' of into module
     new 4e774d8  fix cifar and lstm examples with module
     new 5f864e0  fix a lint requirement of docstring of an internal function
     new 88471a2  delete legacy code
     new 3f3cc6b  Merge pull request #1787 from pluskid/module
     new 4643c31  allow slice channel to squeeze the sliced dimension
     new 749007a  Fix Shape of Fullc, update dmlc-core
     new fee081c  allow multi-dimension for embedding layer
     new ac9d899  fix bug in embedding layer with multi-dimension
     new 7ef4326  Merge pull request #1824 from tqchen/master
     new d28f43e  temp push to server for testing
     new 1cfe746  fix lstm example
     new de175b7  modify executor binding for better memory sharing under new lstm example
     new c7f7cb2  also change module system to bucketing memory sharing
     new ac974e8  adapt lstm-inference for modified lstm_unroll
     new 6295c75  update mshadow (+2 squashed commits) Squashed commits: [f43c089] update mshadow (+7 squashed commits) Squashed commits: [ac974e8] adapt lstm-inference for modified lstm_unroll [c7f7cb2] also change module system to bucketing memory sharing [de175b7] modify executor binding for better memory sharing under new lstm example [1cfe746] fix lstm example [d28f43e] temp push to server for testing [ac9d899] fix bug in embedding layer with multi-dimension [fee081c] allow multi-dim [...]
     new 524d97d  add neural-style to scala example
     new 68be0d0  fix pom.xml
     new 258049c  add run script to scalapkg examples/bin and fix some coding style
     new 90595ba  add run srcipt to scalapkg examples/scripts
     new 8681ad3  fix
     new 1dced30  Merge pull request #1821 from Ldpe2G/master
     new 079a9ed  fix lint error (+1 squashed commit) Squashed commits: [78028fd] fix travis CI errors
     new df79905  [OP] Add Simple registration mechanism
     new 6497c75  Merge branch 'master' into module
     new 36eb685  Merge branch 'module' of into module
     new 6a9fcdd  add module to registered packages (#1841)
     new f7b4f6f  try out SwapAxis in lstm example
     new 5bbb62e  allow slice channel to squeeze the sliced dimension
     new 32a2452  allow multi-dimension for embedding layer
     new 8af9652  fix bug in embedding layer with multi-dimension
     new 66ac8e5  temp push to server for testing
     new 4fc518e  fix lstm example
     new 9af9305  modify executor binding for better memory sharing under new lstm example
     new 17500cf  also change module system to bucketing memory sharing
     new b5b3f11  adapt lstm-inference for modified lstm_unroll
     new e13e8c8  update mshadow (+2 squashed commits) Squashed commits: [f43c089] update mshadow (+7 squashed commits) Squashed commits: [ac974e8] adapt lstm-inference for modified lstm_unroll [c7f7cb2] also change module system to bucketing memory sharing [de175b7] modify executor binding for better memory sharing under new lstm example [1cfe746] fix lstm example [d28f43e] temp push to server for testing [ac9d899] fix bug in embedding layer with multi-dimension [fee081c] allow multi-dim [...]
     new ee6fd62  fix lint error (+1 squashed commit) Squashed commits: [78028fd] fix travis CI errors
     new 0c99cb1  add module to registered packages (#1841)
     new 4bc9d9a  try out SwapAxis in lstm example
     new 045e68b  Merge branch 'module' of into module
     new 3dffb4a  sequential module container
     new 9053470  Merge pull request #1833 from pluskid/module
     new acdec84  add sanity check for meta keys
     new 7279622  temp commit before I delete some code...
     new 75fafde  auto-wiring of modules
     new e52544b  auto rewiring
     new 97108b8  example for seg-fault debugging
     new ac6651d  fix lint errors
     new f369668  Update
     new bc73c88  Merge pull request #1849 from piiswrong/master
     new 8b8c9c2  cleanup Accuracy
     new 9182148  add average cross-entropy metric
     new 3170c4e  update documentation
     new f6b7f31  Merge pull request #79 from oist/vc/accuracy
     new 0992e7d  Add comments about enums
     new 2a7626a  Merge pull request #1839 from tqchen/master
     new 6e1723d  engine info: add more info
     new 20e4f66  fix lint error
     new 000a844  Merge pull request #1853 from tqchen/master
     new 147a0cf  add equality comparison for contexts
     new ec94885  make sure out-grads lives in correct contexts
     new 4a7fa88  a very simple model-par + data-par example
     new 73ad1a4  sequential module container
     new 6bcd793  add sanity check for meta keys
     new 8b89aef  temp commit before I delete some code...
     new 9341cdd  auto-wiring of modules
     new 2242f11  auto rewiring
     new 36c6b39  example for seg-fault debugging
     new b6f964a  fix lint errors
     new 13029b1  fix lint error
     new e64baf9  add equality comparison for contexts
     new 03da234  make sure out-grads lives in correct contexts
     new 564921a  a very simple model-par + data-par example
     new 6a6355e  Merge branch 'container' of into container
     new 0a273d2  lstm demo using our newly available transpose operator
     new 4f9e83d  concatenate of ndarray in 1st dimension
     new 27cf268  try to avoid unnecessary copying when possible
     new 4600db8  [OP] Move elementwise binary op to simpleOp
     new fee5b3c  fix lint error
     new e63f7a6  Merge pull request #1855 from tqchen/master
     new 094a0b0  Update
     new 01e162d  Update
     new 30ae558  [OP] Move Elementwise binary scalar ops to SimpleOp Registry
     new 98b665f  Merge pull request #1826 from dmlc/mli-patch
     new 692e54a  [ENGINE] Add more info
     new 19b3e08  Remove old registry
     new be550d2  Merge remote-tracking branch 'upstream/master' into container
     new 1479f19  fix whitespace error in lint
     new 540b27b  fix bug when binding module as not-for-training (#1861)
     new 12ab2dc  Merge pull request #1843 from pluskid/container
     new 2ead050  [Scala] pass env variables to ps-lite and do distributed training
     new 4b76301  add an example to define loss function in Python
     new 4152764  fix lint errors
     new c3b8302  test python loss on gpu
     new 79bfdbe  Merge pull request #60 from ultradian/patch-1
     new 5bdf8d4  Merge pull request #1842 from javelinjs/scala-dist
     new 83e0944  fix lint error for dangerous default value (+3 squashed commits) Squashed commits: [960414a] simplify the user interface for defining python loss module [dbb797e] test python loss on gpu [f5d94a8] fix lint errors (+1 squashed commit) Squashed commits: [5d7c7ea] add an example to define loss function in Python
     new aa3747d  fix lint error
     new ec8302b  Merge pull request #1868 from pluskid/container
     new d85549b  adds MultiACE or ACE per class
     new 3bff587  Merge pull request #81 from oist/multi_aces
     new 45b5b7c  [DOC] Fix most of the pydoc
     new df02ece  [DOC] Fix howto
     new 8f6b7b1  [DOC] Fix howoto, get started and R docs
     new 07fdbc1  [DOC] Fix system
     new 2246b96  [DOC] include system
     new 156654f  [DOC] Fix tutorials
     new 25c3529  [DOC] enabe readthedocs build
     new 5ed01ac  [DOC] Fix links
     new 8c4b332  Merge pull request #1876 from tqchen/doc
     new 68b44ac  fix LSTM example evaluation refactoring
     new d86d0a3  [DOC] Remove primary domain
     new 71a8980  Merge pull request #1877 from tqchen/doc
     new c5e76a3  make the LSTM example backward compatible with saved checkpoints (#1875)
     new d14cf84  update ipython notebook
     new 452d355  Merge branch 'container' of into container
     new 3d0d337  merge conflits
     new 4e25010  auto bucketing
     new db5469e  fix typo in python loss module
     new f113b4b  Merge pull request #1879 from harouwu/master
     new ae9a2df  furtuer fix of dangerous default values in PythonModule
     new ec22546  Fix links in new doc
     new a6a3bb9  Update R doc link to new docs
     new 1f41b6a  Update scala doc link to new doc
     new 3e7f579  Merge branch 'master' into container
     new 34543bf  Auto indexing for generated module docs.
     new 56033d0  add autoindexing
     new dd16c19  differentiate different modules in the same file
     new 5c261c3  add doc for the javascript
     new 383add2  Merge pull request #1878 from pluskid/container
     new a8bc975  Merge branch 'master' into master
     new c16d1b1  Merge pull request #1880 from pluskid/master
     new 5129d47  fix broken link
     new 8ef8cf4  fix broken links
     new 1c64b11  more detailed doc for embedding layer (#892)
     new 2f2e80b  Merge pull request #1884 from haria/patch-3
     new 76c978d  Merge branch 'master' into master
     new 1a73dda  Use openblas from Julia, instead of installing yet another BLAS library
     new 6a78b03  AddTo support
     new 0e7d516  Merge pull request #1886 from pluskid/master
     new 0c5086a  Simplify linking
     new 7be95f6  Bundle generic cblas.h header for users that don't have libblas-dev installed already
     new 0ec6dca  Oh, it really does call itself `` on non-linux platforms too
     new d1c21b2  Merge pull request #1888 from antinucleon/cudnn_conv
     new 60c16e2  Cut down cblas.h to only the stuff we need, make this work on ILP64
     new b86adfc  Modification based on the PR suggestion.
     new 3cc7cdf  Merge remote-tracking branch 'upstream/master'
     new 7f445b9  Fix typo
     new 9474561  More intelligent parallelized build
     new 3a198e2  [docs] bugfix for anchor text is covered by the navbar
     new c4f76b2  Merge pull request #1890 from mli/master
     new 41fe8ab  Merge remote-tracking branch 'upstream/master'
     new 39786c5  Just remove parallel build, it's making Travis angry
     new 45bc9a2  [RESOURCE] Refactor resource to use storage allocator for temp space
     new dfa1f9a  ps-lite: update dependency, to get the new Environment
     new d53f941  Merge pull request #1775 from yzhang87/master
     new 184d439  Merge pull request #1891 from tqchen/doc
     new eb5cb2e  Merge pull request #1892 from nitnelave/master
     new 5caf710  [docs] bugfix for the search form
     new bdfad62  Merge pull request #1894 from mli/master
     new f046696  update file group in cmake
     new cf870f0  fix code problem
     new 089bdd5  collect attribute from symbolic graph
     new 19573b7  updater refactor
     new e1d30cd  optimizer refactor
     new cbff7b6  [scala]operands change for SimpleOp Registry
     new 32b9cd4  Update
     new bd86903  [OP] Add softmax_cross_entropy loss
     new 532aca4  Merge pull request #1883 from piiswrong/opt
     new d3869a5  Merge pull request #1856 from yajiedesign/addcmake
     new 208ce14  Merge pull request #1858 from yajiedesign/cleancode
     new 967cc3c  Merge pull request #1 from javelinjs/simpleop
     new 914406d  Merge pull request #1896 from tqchen/doc
     new 30e15e3  cudnn v5 compatibility
     new 0bc2171  update submodule
     new a2164ae  Merge pull request #82 from staticfloat/sf/openblas
     new 0d09e31  better error message
     new 2389cc0  Update
     new bbfcafb  Update
     new 90755e3  Merge pull request #1908 from Orange15/patch-1
     new 91b736c  Merge branch 'master' into master
     new fcac086  Merge pull request #1901 from piiswrong/master
     new 14dadf1  Update
     new b38e2cc  update docs
     new f2f88d5  fix bug (typo)
     new fd63f9c  add existing module docs
     new 085584f  fix a bug in vocabulary size reporting in
     new b244def  Merge pull request #1924 from pluskid/container
     new 74e0ae9  Merge branch 'master' into master
     new 4121648  try to add a script to determine whether core of libmxnet is changed
     new b58558d  now ready to go (+3 squashed commits) Squashed commits: [6d43d15] a dry run [734a273] here is the actual implementation [4890f13] more travis CI inspecting
     new 667dbac  add dockerfiles
     new abf4193  fix format of docker doc
     new 2b9008e  Merge pull request #1925 from pluskid/container
     new ed8283a  Update docker/
     new d1ef24a  Merge pull request #1926 from dmlc/docker
     new 3086a3e  Merge pull request #1923 from pluskid/master
     new 736402d  Merge pull request #1921 from Orange15/master
     new bafe679  Update
     new 23750e9  Added parentheses for print statements
     new 6e87d9c  fix breaks
     new 8457257  fix lint
     new ccf75d1  Merge pull request #1929 from dmlc/antinucleon-patch-1
     new ce255b0  change FullyConnected num_hidden only lower bound
     new 30932a5  Merge pull request #1934 from yajiedesign/addops
     new 65d7631a add ignore_label support for 2D softmax
     new 814369f  update gitignore (trigger travis CI)
     new 60d6383  update mshadow
     new a488d2c  Merge branch 'master' into patch-1
     new 753ea69  iterator API over predictions
     new 4f7e8dc  Merge pull request #1938 from pluskid/container
     new 769a4f7  Merge pull request #1927 from rldotai/patch-1
     new 8660a9d  Update
     new fc0fffd  Change rdiv back to be ndarray first, fix R package problem
     new c7b3ba5  Merge pull request #1857 from tqchen/master
     new 45189b3  iterator API over predictions
     new 912170c  remove unused arguments
     new 078e16d  Merge pull request #1939 from pluskid/master
     new d414009  1) Fix a typo in use context instead of ctx when reshape. 2) Add one more parameter to customize metric: can modify both metric and number of samples per minibatch. 3) using module RNN. 4) optimize frame accuracy (from 50% to 60%), now similar to state of the art number on timit.
     new d7fe833  Fix inconsistence of comments and actual code in optimizer (learning_rate) for Adam optimizer. Adding learning_rate into the example to make sure it generate reasonable results.
     new 234d1ef  Merge pull request #1930 from yzhang87/master
     new bbe3151  fix for upstream changes dmlc/mxnet#1857
     new 19c2fe2  [OP] Fix context check on cpu
     new 1419330  Merge pull request #1940 from tqchen/master
     new d51d58d  [OP] Move dot and transpose to new registry
     new d856a89  [OP] Add grad for mx.sym.sum
     new 7e22c75  add roi pooling operator
     new ca28c35  Merge pull request #1941 from tqchen/master
     new 1784d37  Merge pull request #1928 from precedenceguo/master
     new 60e9d6f  some docs for module API
     new d95d7d1  Merge pull request #1947 from pluskid/master
     new c27f406  init moving_var with 1
     new 89d81ad  cudnn's exponetialAverageFactor is 1 - momentum
     new ffca322  training accuracy (cumulative -> real time)
     new 8b1b2b9  Merge pull request #1948 from mavenlin/master
     new e036b4c  Merge remote-tracking branch 'upstream/master'
     new 23ce497  Merge branch 'master' of
     new 4bf4037  check if their are predefined weights and use them
     new 01202b1  lint
     new 4c7a066  be more lenient in model loading
     new 1124e17  Merge pull request #1950 from mavenlin/log
     new 539a070  Merge pull request #86 from oist/vc/transfer_model
     new f71bae0  Minor edits in documentation.
     new f317c96  Merge pull request #1954 from chrisdubois/minor-text-edits
     new fb98bd5  [IO] Refactor augmeneter to support additional ones and chains of augmenters
     new c6a4f01  update mshadow to latest
     new 5cb8cf2  fix doc indexing for module doc
     new 51365c0  fix lint
     new b5c8974  torch module fix
     new aa2008a  Merge pull request #1957 from pluskid/master
     new e88c6ce  Merge pull request #1956 from tqchen/master
     new 0924d16  fix create operator ex
     new 00fd7fc  fix
     new 2fe09e0  add dato
     new 27e32cc  Merge branch 'master' into master
     new ac9a97b  Merge pull request #1959 from piiswrong/master
     new 98b7cf2  add SymbolListAttr
     new 4c3b83c  [STYLE] make pure symbolic impl, add tv loss
     new 53c57c0  Fix c_api for ListAttr
     new b0fc16f  get idx that need to be frozen
     new ce599a9  skip parameter update on frozen weights
     new f62817e  Merge branch 'master' into master
     new f5f18b2  Merge pull request #1331 from winstywang/master
     new 9880bdf  set GRAD_NOP for all frozen params
     new 68180f7  ids derived from arg_names and param_names are not relatable.
     new 4d70ef4  Merge pull request #1965 from oist/vc/list_attr
     new dff7625  Merge pull request #1964 from tqchen/master
     new 5ba27eb  Merge pull request #87 from oist/vc/finetuning
     new deb021d  (temporarily) disable Julia nightly test
     new edd2a47  add keep_highest to reshape op
     new 709733d  Merge remote-tracking branch 'upstream/master'
     new 109aa22  Update
     new 795d479  Replaced iteritems() with items() for Python 2/3 compatibility
     new b6f625a  Minor change to make BaseModule.predict more like FeedForward.predict
     new 9caa418  Update
     new 8f099c3  Merge pull request #1973 from winstywang/master
     new 49dfe70  Merge pull request #1977 from rldotai/master
     new 9044119  reset training statistics after each training batch
     new 2fc8f0b  fix module predict outputing numpy instead of ndarray (#1985)
     new f3f7280  fix lint warning
     new 459cd71  reset eval_metric only each epoch
     new 4fbc8ff  Merge pull request #1989 from pluskid/master
     new d645e6f  [R] add get started docs for R
     new d1a01cc  Merge pull request #1997 from thirdwing/master
     new f2938a7  update recent examples, kaggle winning solutions and tools with MXnet
     new 5a01dd8  add TOC
     new 275b070  Merge pull request #2000 from phunterlau/master
     new 074e113  [speech] revert epoch if val performance drops
     new 6a49e9a  enforce a lower bound for lr decaying
     new 4e3219b  Adding decoder part.
     new 1db84f9  do not reset bind by default
     new 9adc1d0  Merge remote-tracking branch 'pluskid/speech'
     new cde5497  Merge remote-tracking branch 'pluskid/speech'
     new a96da50  keep params if rebind after params_initialized
     new 7db07fd  [OP] enable use_global_stats in batchnorm
     new 1f457a3  [docs] try to add python examples for symbol  (#2002)
     new 12345b8  [docs] update readme (#2005)
     new 72a5f10  Fix gcc5.3 + nvcc 8.0 (#2004)
     new 4229801  Merge branch 'speech' of into speech
     new fada86e  add a guard to avoid copying to itself in setting parameters
     new 0b974ec  Merge branch 'speech' of into speech
     new d116a76  Merge pull request #2003 from tqchen/master
     new 0b3582d  [R] fix R travis
     new 0a5182d  Print validation error
     new 5119ddc  Merge remote-tracking branch 'yzhang87/master' into speech
     new 53b4973  Merge remote-tracking branch 'upstream/master' into speech
     new a7b6c38  refactor speech demo
     new d1637ef  fix bugs in refactored speech training demo
     new 8712de1  remove the old LSTM bucketing script
     new 947d689  Fixing some typos.
     new 528b949  Merge branch 'container' into speech
     new 834aa0d  update README
     new 67740d6  fix lint error
     new 6546198  fix typos
     new 98f5395  experimental truncated-bptt
     new e35a7a1  truncated-bptt
     new 4d46f90  pass padding information through data batch
     new 35b5a97  fix bugs in truncated-bptt
     new 58c8ad8  overwrite options from command line
     new 40a3b03  add default.cfg
     new 64bf2ef  print to stderr in argparsing
     new e502e2d  Merge pull request #2020 from pluskid/speech
     new 9d900d5  Remove unrelated files
     new d5ac0b0  Merge remote-tracking branch 'pluskid/truncated-bptt'
     new 896fba9  Softmax Activation (#2022)
     new a67507a  Change init scale for lstm.
     new b926bf9  add state-forwarding in scoring of bptt
     new fb0cc15  Merge remote-tracking branch 'yzhang87/master' into truncated-bptt
     new d3a4bb1  add clip gradient and initializer
     new 5587c76  update readme
     new 918c14a  Fix cudnn softmax activation break (#2023)
     new 7becd08  Bug fix for truncated bptt
     new 8ba74b0  fix typo
     new 8ed5746  [R] export C++ functions to R side
     new 0c08a3c  pad with zeros for bptt IO
     new f11058d  split too-long sentence for bucket-io
     new cc555cb  Move utils to io_func.
     new 6224218  Merge remote-tracking branch 'yzhang87/master' into truncated-bptt
     new c680519  Merge branch 'truncated-bptt' of into truncated-bptt
     new df4265f  fix io bugs in bucket spliting
     new 9ff5f0c  Correct LD_LIBRARY_PATH (#2013)
     new 8752858  lets don't do transformation on labels
     new 5d549db  more flexible shape inference for softmax output
     new 7644f89  lstm-unroll without transformation on label
     new 1afccc6  prevent creating cudnnPoolingOp with unsupported Pooling method (#1986)
     new 25580a3  torch related fix (#2016)
     new 5dd4041  Add operator Smooth L1 Loss (#1953)
     new f58a578  Change "has_key" to "in" in python/mxnet/ (#2038)
     new e2a475f  Bug fix for BPTT. Remove the rescale_grad in Module (too confusing).
     new 8cc9ed5  use effective sample count to rescale lr
     new 041365c  Merge remote-tracking branch 'yzhang87/master' into truncated-bptt
     new 9e242bc  a separate config_util module
     new 19a92ab  don't rescale grad here
     new 67c3691  Merge remote-tracking branch 'upstream/master' into truncated-bptt
     new 16f3212  reset optimizer after rewinding an epoch
     new 0f2631d  fix lstm_unroll bug
     new ca53828  Move repeat code into config.
     new 51899ed  fix lint errors
     new a58c86e  Merge branch 'truncated-bptt' of into truncated-bptt
     new 2b97e40  Update
     new 27b18c0  allow specifying shape information for variable (ref #2039)
     new 624c9a7  test overwriting shape
     new f6e9aa0  Merge pull request #2054 from pluskid/speech
     new aa47d52  Update
     new b2faafd  add weight decay to gamma
     new 96cda5f  [abi-compatible] Fix GCC 5.1 ABI change issue. Refers to dmlc/mxnet#2060
     new 121989d  [abi-compatible] Fix another std::string usage which might cause SIGSEGV.
     new adbb485  nesterov sgd
     new 1a5ebb9  Merge pull request #2062 from hotpxl/abi-compatible
     new 76767aa  Merge pull request #1949 from mavenlin/optim
     new 49e470e  [DOC] Fix doxygen
     new 312d603  [DOC] remove old doc
     new b5a74cc  Merge pull request #2065 from tqchen/master
     new d53cd61  custom op
     new e1ce2f2  support aux for custom op
     new 1fafa06  Changed convolution Operator not to use CuDNN if dilate param is used
     new 65a1257  Merged latest changes from mxnet master
     new 34beff3  Fixed indentation
     new 85a8413  Moved test for convolution with dilate to
     new 4a1434f  Merge remote-tracking branch 'pluskid/truncated-bptt'
     new 26454d0  Adding decode script for acoustic modeling.
     new 7bb4608  Merge remote-tracking branch 'pluskid/truncated-bptt'
     new f856fe1  Merge remote-tracking branch 'yzhang87/master' into speech
     new 2ddc902  Merge remote-tracking branch 'upstream/master' into speech
     new c948dc3  Added 7z alternative for mnist
     new e4c0d25  Fix bug during merge stage.
     new 56cf95e  custom op doc and example
     new 54b9524  updates
     new 013b88b  Change where set_attr is called in _define_atomic_symbol_creator
     new 69e9898  Merge pull request #90 from Mike43110/win/zip
     new 2b0d825  Merge pull request #91 from Mike43110/list_attr
     new c594440  Fix SoftmaxActivation Gradient (#2071)
     new 35328be  allow list attr without traveling all descendents
     new 463f202  add support for ListAttrShallow
     new 3961e84  Merge pull request #2072 from pluskid/master
     new e7cd135  Merge pull request #92 from vchuravy/vc/fixup
     new 8993cc0  Merge branch 'master' into master
     new baa9c2a  revert #91
     new fc17743  Update
     new 083eb16  Merge pull request #2051 from piiswrong/master
     new f59e82b  Exactly reproduce 56 layers ResNet on CIFAR10 (#2046)
     new c304c2b  Removed superfluous and python3 incompatible print statements
     new cb514b6  Merge branch 'master' into feature/convolution-dilate
     new ab03224  Updated mshadow submodule to point to latest version
     new f38a188  Fixed whitespace lint errors
     new ec4c270  Merge pull request #2069 from kadeng/feature/convolution-dilate
     new 229153b  dcgan example
     new e51d60c  Merge pull request #2082 from piiswrong/master
     new 0c11d1f  Merge to Adding a speechSGD method in optimizer (1% better than SGD on AMI).
     new 0f77c1a  Update the readme file.
     new 15bfb67  decrease font size in docs
     new 6e23505  Merge branch 'master' into speech
     new cab4121  Move the speechSGD outside the optimizer
     new 72a6d90  Add speechSGD back.
     new eab7bc0  Merge remote-tracking branch 'upstream/master'
     new b7c14ba  Merge remote-tracking branch 'yzhang87/master' into speech
     new a582bb8  Update
     new 844d342  Merge remote-tracking branch 'yzhang87/master' into speech
     new 75e2295  update some typos
     new b6eed87  change truncated-bptt to default.
     new 697e65a  Update default parameters.
     new b54258c  Merge remote-tracking branch 'yzhang87/master' into speech
     new 2145e0d  Merge pull request #2083 from pluskid/master
     new e236fa0  Merge remote-tracking branch 'upstream/master' into speech
     new 8fbc539  fix lint error
     new 9bbb887  fix css issue
     new be7220a  add function sum to Symbol
     new b8f2911  Merge pull request #2073 from jingpengwu/patch-1
     new d412d54  fix typo and manual trigger travis CI strange server error
     new 7e2688e  fix typo
     new bac908d  temp disable core-change detection script due to strange travis failing
     new bb636c9  Merge remote-tracking branch 'upstream/master' into speech
     new 7c83e2c  Merge pull request #2086 from pluskid/speech
     new b8354fb  Merge pull request #2094 from g329/fix-typo
     new 3b34185  Merge pull request #2077 from Ldpe2G/test
     new 44bfa54  fixed MXExecutorBindEX
     new b9947a0  Merge pull request #2098 from sbodenstein/master
     new 64a63f7  Deduplicate and generalise metric update!
     new 2cff740  [Op] Fix RoI pooling (#2109)
     new 6d99054  Bidirectional lstm example #2093 (#2096)
     new 0ea4369  Merge pull request #94 from oist/vc/generic_metric_update
     new b5a3693  [OP] new mode to reshape (#2114)
     new c267469  fix typo
     new 81dd47a  update dmlc-core (#2132)
     new 4b1ce16  [OP] disable LRN for CUDA 7.0 (#2134)
     new 5fda113  Make Travis CI more stable (#2140)
     new b369483  [amalgamation] update (#2135)
     new 648a66a  update scala example neural-style, [STYLE] make pure symbolic impl, add tv loss
     new 42bc7e1  Merge pull request #2100 from Ldpe2G/master
     new cdc3780  Fix workspace calculation introduced in #944
     new 24fa38b  Add deconvolution workspace param to upsampling
     new f81bee2  Add backward for sum_mid_internal
     new 37a9b14  Merge pull request #2104 from oist/vc/deconvolution
     new 0c7024b  fix custom operator python3 bug
     new 99f8eb6  Merge pull request #2143 from JeffDong/fix_custom_operator_python3
     new 2eb8559  add documentation for new operator interface in operator_util.h
     new ad737df  add example to the operator_util doc
     new 403bd16  add links for the SimpleOp docs
     new 7841253  Merge pull request #2136 from precedenceguo/master
     new d6cba73  Merge branch 'master' into patch-25
     new aa512fa  doc for symbol attributes and naming convention (#2070)
     new bd800b1  Merge pull request #2110 from pluskid/master
     new e4f7d18  Merge pull request #2129 from qiaohaijun/patch-25
     new be48d7d  [SYMBOL] Enable group2ctx in simple_bind, for multiple device executor
     new 5a99014  add multi task tutorial
     new 691f1d3  Merge pull request #2152 from haria/patch-4
     new 45377da  [R] update to latest imager. close #2150 (#2162)
     new f834971  fix attribute doc
     new a833098  Merge pull request #2159 from pluskid/master
     new 9217027  Update (#2163)
     new 5b9b2a8  Merge pull request #2151 from tqchen/master
     new 5b35751  Merge pull request #2141 from sxjscience/add_backward_sum_mid_internal
     new ede5fbf  Add Fast R-CNN object detection example and fix ROIPooling cuda problem (#2167)
     new 8f9bf01  Fix typo.
     new 442d791  Merge pull request #2169 from darabos/patch-1
     new b54504b  fix typo and warning
     new dcacd74  Merge pull request #2172 from tornadomeet/fix1
     new 5eadb90  custom op fix
     new afb3661  Merge pull request #2183 from piiswrong/master
     new 00a7d06  [op] add broadcast plus
     new 3669ed1  Fix Bug in Reshape Operator (#2188)
     new 0aa3f86  Merge pull request #2165 from sxjscience/broadcast-op
     new 65d7dca  Fix small network visualization bug (#1970)
     new afb92db  Fixed typo in Readme (#2191)
     new 153507b  Update news (#2193)
     new 3922cbf  Fix library name and type in Linux/OS X. (#2198)
     new ca6c7be  fix typo in (#2199)
     new ec5538b  Add broadcast_div (#2196)
     new 6523a7a  suport dual stride for pooling (#2202)
     new 3a99dff  [Resource] Allow get host memory as temp space in any context
     new 37c2b28  Merge pull request #2230 from antinucleon/master
     new 4ef6972  [Op] Add softmax norm option (mimic caffe)
     new 3a013b8  fixup expr check for v0.5
     new e6d07f0  fix deprications on v0.5
     new 55f2a42  fix overwriting of inner function
     new 48b723a  fix ByteString deprication and deprications in tests
     new 7cea0c3  - added state support for Custom Operator API functions. (#2181)
     new 4292245  Update random seed in (#2238)
     new 5f0a403  Update (#2239)
     new 2d97cac  Merge pull request #2231 from antinucleon/softmax_caffe
     new 34feac3  Update c_api.h
     new 7213e3d  Remove debug print (#2245)
     new 6fbad56  Update
     new 2eb532f  move PostDFSVisit to graph_algorithm.h
     new cbe23ac  Merge pull request #2252 from mavenlin/dfs
     new 1313609  Update mshadow PR #120 (#2255)
     new 118b37e  [RELASE] tag 0.7 release (#2247)
     new afd9baa  enable travis for 0.5
     new d1bf894  Merge pull request #98 from oist/vc/0.5
     new 179ca3a  [Op] global pool option (#2243)
     new c9cd76f  fix doc build (#2265)
     new 9b925f4  Copy output in BaseModule.predict to avoid overwritten by following batches (#2226)
     new 0c6b42f  [Scala] Spark module sketch
     new e6b9031  [scala-spark] worker, server, scheduler initialize success
     new 448e555  [scala-spark] LabeledPointIter seems to work
     new ffd3a14  [scala-spark] train on spark single worker
     new 35728f0  [scala-spark] Tested 2 workers on spark standalone mode (multiple workers on one node)
     new 0fe38c6  [scala-spark] start ps scheduler through runtime exec
     new 5e883f4  [scala-spark] use ParameterServer to launch ps scheduler/server
     new 9b7e8db  [scala-spark] success to make ps server & scheduler terminate after training finishes
     new 4a46dc3  [scala-spark] descide scheduler ip & port automatically
     new b6c821a  [scala-spark] change input dimension from Int to Shape
     new 9c443f1  [scala-spark] take MXNetParams for job
     new 2525b1e  [scala-spark] serialize symbol to json and load from worker
     new 4da5383  [scala-spark] dump training time
     new fb91899  [scala-spark] hotfix for databatch shape mismatch & LabelPointIter dispose
     new 4dfae48  [scala-spark] support training on multi devices
     new 5d5a9e3  [scala-spark] test spark ann
     new f0bd547  [scala-spark] adjust epoch size in each executor, switch to async training
     new 3f54908  [scala] serialize & deserialize NDArray
     new 2d2b595  [scala-spark] MXNetModel finish
     new 6a87a83  [scala-spark] Serializable NDArray for spark & special long-living DataBatch for cached dataiter
     new 1c05765  [scala-spark] add
     new 5a23b0a  [scala-spark] make spark-mllib version configurable
     new c3edd8d  [scala-spark] remove useless debug logger
     new abd57a0  Merge pull request #2256 from javelinjs/spark
     new 95296ff  Fix makeloss (#2263)
     new c71bb7e  Create
     new 9103e92  Merge pull request #2273 from qiaohaijun/patch-27
     new 3081e28  Update
     new 5e000e2  Merge pull request #2275 from mhzdlm/patch-1
     new 0c77f9f  update text_cnn_classification (auto learn word embedding)
     new 7b5bb42  add chinese word segment module using lstm model
     new e3e26b5  Fix a typo include/symbolic.h (#2281)
     new a19340e  fix a bug, you should save param first then reset
     new 9c502e1  fix cmake problem and enable it to find libopenblas.dll.a on windows
     new 99e64a8  Merge branch 'master' of
     new beddf15  revert back changes in c_api.h
     new 60d3b2a  update build doc for windows
     new 6a68ab5  Merge pull request #2290 from MonkeyTang/master
     new 8f35cc5  translate docs/system/ to chinese (#2291)
     new f9ca6df  Fix the link to model API and some grammars
     new b04df11  Merge branch 'master' into master
     new 62f527d  Merge pull request #2299 from koallen/master
     new 8ced10f  Merge branch 'master' into master
     new e0ae42b  support both openblas.lib and libopenblas.dll.a
     new 53bdb96  Merge branch 'master' of
     new b9d8391  Merge branch 'master' of
     new 674a973  minor fix
     new 3b35435  Merge pull request #1548 from hjk41/master
     new 3ff8c5d  Create (#2286)
     new 2e9b96b  Update (#2303)
     new 751f07c  Add a InferType to SimpleOp in order to support other DTypes.
     new 32bb3cf  Merge pull request #2300 from oist/vc/infertype_so
     new 427de15  add function reshape to NDArray in scalapkg
     new 157cbe9  fix code style
     new d36d152  add test for reshape and fix some problems
     new ba48ffc  align the indent
     new e6af30a  Merge pull request #2288 from Ldpe2G/master
     new 86719b0  fix pooling typo (#2311)
     new 6fbd0a1  fix typo
     new 5823240  [spark] automatically upload and distribute native jars to all workers
     new ce701e1  [spark] update example and document for ExecutorJars
     new 78e6641  Add axis support for broadcasting and sum. Also add max_axis and min_axis (#2292)
     new 7b5644e  Merge branch 'master' into spark
     new ddb8b0f  Merge pull request #2309 from javelinjs/spark
     new 8d58780  Update (#2312)
     new 23e69ae  Merge branch 'master' into patch-31
     new 28c400d  Lazy Refactor Native OP (#2272)
     new 063da7f  Revert "Lazy Refactor Native OP" (#2319)
     new 5a5f399  Merge pull request #2313 from qiaohaijun/patch-31
     new d81eb7d  new operators transpose, crop, flip
     new 511930d  custom op memleak fix and ndarray at indexing
     new 69dd8bb  lint
     new 88a5a16  lint
     new ae07452  fix
     new c00a4f7  doc for dependency engine's implementation(Chinese) (#2332)
     new 0e04a61  Merge branch 'master' into master
     new 5d1430e  capture by value->reference (#2335)
     new 77e1750  Merge branch 'master' into master
     new a77d47d  Correct init idx2name on mult-CPU/GPU training
     new b1eef81  Merge pull request #2337 from Answeror/bug/idx2name
     new 6444fde  update examples of OCR, LSTM sort, FastRCNN etc
     new 39d11d4  Merge pull request #2338 from phunterlau/master
     new 1cfd6d6  scalapkg executor, shared memory bind and runtime reshaping
     new 9c9e762  Initial support for DType and TypeFlags
     new a282586  add typed empty, ones, zeros
     new c161fe1  make arithmetic work with DType and extend arithmetic tests on all types
     new 2916a8f  turn macro into helper function
     new dc405e7  add type_infer
     new 4a4ea61  make bind tests use isapprox and don't run division tests for Integer types
     new a76ab7c  add test case for executor reshaping
     new 8e9772a  replace long with int32
     new 3aff180  Merge pull request #2341 from Ldpe2G/master
     new 419f34b  Merge branch 'master' into master
     new f123a64  Merge pull request #2297 from piiswrong/master
     new 26f9dfd  reshape (#2342)
     new 91ba6a6  Amalgamation and readme update for Android, JS and iOS support (#2331)
     new 4ebd786  Fixed minor typo (#2340)
     new 79c1565  fix build warning (#2346)
     new 45abd15  [DOC] translate docs/system/ to chinese  (#2314)
     new 35ea182  fix python optimizer wd
     new ae3f43b  Merge pull request #2352 from JeffDong/fix_python_optimizer_wd
     new 6545a48  extend makeloss to arbitrary input tensor shape (#2353)
     new a80ae98  Merge pull request #97 from oist/vc/ndarray
     new affebe6  Exclude division test for Float16
     new 9a827f7  fix blas_vendor for v0.5
     new f2684a6  Model parallel with bucketing (#2323)
     new 1940a68  Merge pull request #102 from oist/vc/test
     new 55ced9e  refactor lstm model and add documentations in R (#2329)
     new cbc61a7  add example of building GRU (#2348)
     new 944504d  Fix stream of output gradient (#2362)
     new de12a63  fix a typo (#2368)
     new b627e92  fix python3 monitor bug (#2364)
     new 2beee67  Scala package, add transpose to Symbol
     new 8a5754a  Merge pull request #2377 from Ldpe2G/master
     new 0c81842  Patch for (#2376)
     new e4d7842  translate docs/system/ to chinese (#2378)
     new db6f6a9  Correct init idx2name on multi-CPU/GPU training for FeedForward (#2366)
     new 8e7515b  enable other dtype in deconvolution (#2322)
     new 259f6d1  Fix minor issue that causes batch data in Iterator to be skipped (#2219)
     new 004c238  Enable training fast rcnn with multiple GPUs (#2358)
     new a994183  add CallbackFunctionTutorial Chinese Translation
     new a015dc0  Merge branch 'master' into master
     new 0b70563  add stop_factor_lr in FactorScheduler (#2211)
     new 06f99f0  Add shuffle option for im2rec (#1666)
     new 7484444  Improve Caffe Importer (#2345)
     new a8ea8d4  Merge branch 'master' into master
     new c672a91  Merge remote-tracking branch 'dmlc/master'
     new 30be63f  Translate fiveMinutesNeuralNetwork into Chinese
     new 52ccd25  Merge pull request #2379 from harryprince/master
     new a07cd0c  Update Docker instructions (#2388)
     new a32ba41  fix dtype of workspace in cudnn convolution (#2321)
     new b400f9a  add scala example cnn-text-classification
     new 3d298f8  Merge pull request #2387 from Ldpe2G/master
     new 110684f  fix test acc metric error
     new ba25a9d  Merge pull request #2394 from Ldpe2G/master
     new 9e8faf6  don't intermingle type switch and preprocessing directives in (#2385)
     new 9f2a5d3  Fix a typo
     new a4be1e6  [docs] try to fix search in the doc webpage
     new 4cf2086  Merge branch 'master' into patch-1
     new 786a977  Merge pull request #2397 from Atry/patch-1
     new a1c101f  automatically adjust num of threads used by make (#2320)
     new 8cd4cc2  update contributors (#2402)
     new d7fdf40  1. temporarily fix gcc 6.1 and cuda problem; 2. add power to ndarray
     new 9a42e15  add ufunc_helper that helps all broadcasting issues; fix div operator in python3
     new eb8b3ec  add maximum operator
     new 1a0d914  fix lint and python2 division overload
     new 28b9188  fix clang unsupported option -fopenmp issue
     new 1314fb1  Merge pull request #2407 from qiaohaijun/patch-34
     new edb32aa  enable other DTypes in Concat
     new f5bd5bb  enable other DTypes in UpSampling
     new 0c56be2  add tests for multiple dtypes for Concat and UpSampling
     new a825c57  enable other DTypes in Reshape
     new 22ca494  enable other DTypes in BlockGrad
     new 4605518  enable other DTypes in SwapAxis
     new 7b847e5  enable other DTypes in ElementWiseSum
     new 05a696b  Merge pull request #2380 from oist/vc/dtypes
     new be8266d  Update submodules
     new 232e6b9  Merge pull request #2416 from sxjscience/update_submodule
     new 967e07d  Revert "Enable training fast rcnn with multiple GPUs (#2358)" (#2425)
     new 8fd4d16  Intergrate with baidu warpctc
     new 2279cbf  Merge pull request #2326 from xlvector/warpctc
     new 498a0e9  Update (#2433)
     new 6e85eb2  Saving folder, log, removed some import, fix typos (#2406)
     new 887491d  create mxnet/docs/zh/packages/python/ (#2434)
     new 4b88c19  Support ndim up to 7 for binary broadcasting operators + Accelerate reducing OPs by calling reduce_except_dim if possible. + Add `/bigobj` to CMakeList (#2418)
     new e721aa5  fix some typos (#2450)
     new 8d57211  [OP] Support softmax with probability label (#2456)
     new 1afa354  add warpctc config to (#2451)
     new 72121a5  update reshape using shape (#2446)
     new 7a2b67e  fix worker node could not exit (#2440)
     new 86d71e9  don't run char rnn test in R (#2459)
     new 23d428d  [OP] add slice_axis (#2458)
     new 7cb0474  add target_shape and adj to deconv for friendly using (#2445)
     new 3bd0c61  Add RNN model in R (#2417)
     new 68bda2a  Update warpctc-inl.h (#2437)
     new 7693b93  Deleted, fixed a typo in (#2463)
     new 30b97b6  start work on the docs (#2431)
     new 2d0ca16  add type="text/css" for mxnet.css (#2466)
     new 8c4c005  rm lstm-model in this branch
     new c902299  Merge branch 'master' into master
     new ed8fa53  Merge pull request #2469 from qcl6355/master
     new 560c312  Add minimum operator
     new ba57aa9  Merge pull request #2473 from jermainewang/master
     new 0475454  allow partial set paramter in module (#2482)
     new c39abe4  [OP] Fix target_shape infer in deconv (#2483)
     new 3655bd5  create mxnet/docs/zh/packages/python/ (#2487)
     new 8e88cb7  end 2 end neural art example (#2490)
     new 5544a05  [OP] Add expand_dims (#2488)
     new 2373e6b  [scala]bug fix for example. more logs for native library loading (#2493)
     new cde4e0d  Upload VGG19 symbol (#2502)
     new 91421fc  add WARPCTC plugin compile with cmake (#2477)
     new 2262bb2  Update MShadow (#2500)
     new 1934de5  Enable Bulk Execution of a Subgraph For Faster GPU Serving (#2496)
     new bf7798d  caffe_converter: fixed bug parsing caffemodel files on windows (file should be opened in binary mode) (#2504)
     new be9b08e  create mxnet/docs/zh/packages/python/ (#2479)
     new 88151ca  create mxnet/docs/zh/packages/python/ (#2480)
     new 721fd80  Add ability to checkpoint at lower rates (#1669)
     new 388a2a9  [OP] Move sample op to operator_util, support source function (#2511)
     new a717570  softmax output label shape
     new 9b6d4d2  opencv plugin
     new db34c50  Merge pull request #2457 from piiswrong/master
     new f10e709  Sum with axis full support (#2518)
     new 107f879  [RFC] Embedding DTypes (#1) (#2516)
     new caee4b4  Add GRU in R (#2491)
     new d388008  [scala] disable kvstore for single device
     new 5daba1a  [scala] fix memory leak
     new b85d2e8  Merge pull request #2517 from javelinjs/bugfix
     new d6540eb  add cmake command for Mac OS and edit related file
     new e2336aa  Merge pull request #2523 from jennyzhang0215/Mac_cmake
     new b5313c1  Implement broadcast_to in the C++ side (#2525)
     new 90f28b4  (1) export more C++ functions; (2) R docs for CUDA (#2520)
     new 4ad8c9b  [scala] change NDArray._random_uniform to NDArray._sample_uniform. support kwargs for GenericNDArrayFunctio (#2529)
     new f857c1e  Add broadcast_power (#2534)
     new 3778990  finish PrefetchIterator
     new 7ec93d2  add some check to prefetchIter
     new 4095656  fix code style
     new 03bd1d7  fix some problems
     new 1bf1337  Merge pull request #2526 from Ldpe2G/develop
     new 0ad34c9  cudnn auto tuning (#2538)
     new 6272e76  KV: enable aggregate on GPU for distributed kv (#2545)
     new 474f74b  [EXEC] More generic segment training api (#2551)
     new 6c27c3b  update (#2563)
     new 6df4df2  Improve caffe converter (#2550)
     new fd4c560  [OP]Spatial Transformer Network (#2547)
     new ca800df  [SYMBOL] Fix stackoverflow on big chain destruction (#2586)
     new 13bb5c1  Fixed a minor typo in markdown that breaks a link to file in github.
     new 048546d  Merge pull request #2594 from tjrileywisc/master
     new 97fa49f  fixed typo in warpctc install doc which caused compile error (#2552)
     new 2a9ca98  softmax shape fix
     new b40ec35  module monitor
     new d7bdcc1  fix batch norm
     new d5abf21  monitor for all modules
     new eaa5bda  Merge pull request #2606 from piiswrong/master
     new 5e70cb9  start transitioning to Documenter.jl
     new c06dfbd  1. Support numpy.ndarray as input for exe.forward() 2. Add batch_dot
     new 596d003  Merge pull request #2615 from sxjscience/batch_dot
     new ce16bfe  split import_ndarry functions up
     new f09362c  add as a dependency
     new 411a08b  simplify _import_ndarray
     new 7b65547  implement _import_ndarray functions as macro
     new 89e2c8e  implement a function cache
     new 7ff4903  Allowed for CPU usage.
     new 5b2bd3e  Refactor _import_atomic_symbol to be a macro
     new 2f089b8  expose GetAtomicSymbolName in header (#2627)
     new 5c07a91  batchnorm fix
     new 1ad9a1a  transition context to see how it turns out
     new ca949d7  get rid of the most of the old docs
     new 504aec6  do some housekeeping and start moving things to markdown
     new 1a2cb9e  Merge branch 'master' into master
     new 144e53d  making adam more in-place
     new 7a485bb  Merge pull request #2628 from piiswrong/master
     new bfd1df5  Merge branch 'master' into adam-in-place
     new 7659d3b  Merge pull request #2629 from danielgordon10/adam-in-place
     new c4d5868  Merge branch 'master' into master
     new dfb1443  Merge pull request #2626 from asmith26/master
     new 2feb559  Add correlation op
     new 1d9d6b1  Merge pull request #2655 from dsqx71/master
     new acdae67  Warpctc mult label length (#2650)
     new 70fe575  fix bugs of rcnn example (#2651)
     new 5532511  [OP] 3d conv + 3d pool (#2585)
     new 7acf93b  update mshadow to fix the docsystem (#2665)
     new f1c6efe  fixed python3 compatibility (#2668)
     new 733d46d  setup api docs with autodoc
     new b7df20d  fixes (#2667)
     new 8590200  Update
     new 9113b21  rough first pass .rst -> .md
     new d8de742  improve autogenerated documentation
     new 597c736  convert rst to md with pandoc
     new d7d4d90  finish mkdocs support
     new 449e88e  fix build issues (#2679)
     new 3f95364  Merge branch 'master' into master
     new 58dec8b  Merge pull request #2674 from wshenx/master
     new b496758  fix lint
     new 47d8d89  Merge pull request #2692 from piiswrong/master
     new 44daeb4  fix bug of image resize in
     new a4ddf49  Merge pull request #2701 from zlmzju/master
     new aa18766  SVM Output cpu implementation + test + example (#2708)
     new f3cc5b1  Update (#2721)
     new 88fd2ff  Use a smaller threshold (#2726)
     new 0ef25de  Update link in  #2729
     new 4c4d51c  Merge pull request #2730 from Devinsuit/master
     new 3790c12  fix convolution speed
     new 5f58c7a  1. Accelerate embedding using the new kernel
     new 5cf7729  fix 3d conv shape inference
     new cf6a5ae  Merge pull request #2728 from sxjscience/acc_embedding
     new dba99a2  add RepIterator
     new 92a488d  add Img2Vector tools
     new 160f814  del RepIterator
     new 4f0c8c9  add scala example lstm (#2737)
     new 3f6a69e  Merge branch 'master' into master
     new 5ab62db  Merge pull request #2736 from piiswrong/master
     new 1daf89a  fix some code style
     new bc61824  Merge pull request #2747 from Ldpe2G/develop
     new 72ea6f9  Refactor symbol ndarray register mechanism (#2745)
     new 7e98334  Add MSRAPrelu initialization (#2647)
     new 3c3cba4  Enable auto-broadcasting for ndarray.broadcast_to. We can broadcast (50,) to (20, 50)
     new 92323dd  Merge pull request #2755 from sxjscience/auto_broadcasting
     new d415076  scala-pkg, fix DataParallelExecutorManager
     new 1b31841  Merge pull request #2769 from Ldpe2G/develop
     new cbcdceb  Add RMSProp optimizer in R
     new 6760cf5  Merge pull request #2748 from ziyeqinghan/optimizer
     new 88e7207  formatted code style
     new 6b7f912  update rcnn example to Faster R-CNN
     new 03f3c28  Merge pull request #2780 from precedenceguo/master
     new 3c7bdcb  Revise Comment + Fix typo
     new 7fd4c57  Merge pull request #2790 from sxjscience/fix_typo
     new ff550af  support cpu (#2793)
     new 28d1860  fix cudnn auto tuen
     new b852065  turn off auto tune by default
     new 8a3424e  Merge pull request #2791 from piiswrong/master
     new 407871f  Infershape_fix
     new 70fa289
     new 6fa5813  Merge pull request #2802 from Godricly/Infershape_fix
     new fde63e8  Update (#2798)
     new a1b404c  DataBatch support gpu (#2797)
     new 3fb29a3  Fix amalgamation compiling issue. (#2814)
     new 2fa4f39  Accelerate batch_dot using gemmbatched
     new 06841a0  Merge pull request #2816 from sxjscience/acc_batch_dot
     new 0460049  [Op] cuDNN RNN Symbol (#2795)
     new fc0bb39  extend Module to fix_param_names that saves gradient computation and update faster rcnn example (#2794)
     new 5af3696  replace IOException with NoSuchElementException & drop object RepIterator
     new e969d45  Add Adam optimizer in R
     new 095d742  Add AdaGrad optimizer in R
     new 3b9e0b9  Add AdaDelta optimizer in R
     new 49b141b  revoke import scala.collection.mutable.ArrayBuffer
     new 7159d47  add javadoc
     new 611800f  Merge branch 'master' into RepIterator
     new 1767f0b  Merge pull request #2822 from ziyeqinghan/optimizer
     new 977ac39  Merge branch 'master' into RepIterator
     new 2a2fbb3  Merge branch 'master' into Img2Vector
     new 2196588  Merge pull request #2742 from liyuance/RepIterator
     new 4247f52  Merge branch 'master' into Img2Vector
     new 95ef531  Merge pull request #2744 from liyuance/Img2Vector
     new d45ad51  scala-pkg, add makeloss and smooth_l1 to Symbol, and test case
     new e343416  Caffe Plugin (#2746)
     new fd6ce60  Revert "Caffe Plugin" (#2852)
     new 36d4f5b  Merge pull request #2849 from Ldpe2G/develop
     new 1d8fa0b  Caffe-plugin (#2854)
     new 99b04d2  redirect PS or scheduler log to parent process's
     new 0f0098f  fix bug when redirectlog
     new 152ee6b  add InputStream OutputStream import
     new c2d08e4  Merge pull request #2868 from liyuance/redirectLog
     new d517612  fix typo (#2872)
     new 01d899e  fix graph executor
     new 6061c87  Merge pull request #2880 from dmlc/piiswrong-patch-1
     new da81506  Update makefile to depend on sub directories
     new f3d6938  Add indexed record io
     new c8989e8  fix test
     new fdd3524  Merge pull request #2887 from piiswrong/master
     new 14f90a6  nce loss
     new 5f27880  Merge pull request #2756 from xlvector/nce-loss
     new 3dbec0f  cleanup docs
     new c74594f  enable travis to build docs
     new 85ec111  finalize mkdocs setup
     new 8dad674  Merge pull request #106 from oist/vc/docs
     new 789bb6b  setup Documenter.jl
     new 2c651ff  fix documenter config bug
     new f442c49  switch doc to new location.
     new 79f51e6  fix URL to Windows building instruction (#111)
     new 8cc197c  CSS for docs: removing inherit for font properties on most elements. (#2904)
     new ea24b38  [OP] Generalize SoftmaxActivation (#2833)
     new c49ef1a  PS node recovery support (#2671)
     new 2042271  Update (#2885)
     new 9ee093c  [TENSOR] Move TBlob and TShape to mxnet (#2906)
     new b54620b  Fix broken documentation link.
     new 8058ff7  Merge pull request #112 from slundberg/patch-1
     new c37dbb4  [fcn-xs]fix offset and init bug & update readme (#2926)
     new e2d8e1b  fix info
     new 1e5337a  Merge pull request #2931 from tornadomeet/fix-info
     new 4ed3f3b  [Storage] Temp space directly release to CUDA (#2927)
     new ac53ca2  Dockerfile for building and previewing docs locally.
     new aa8e911  [caffe-plugin] Fix Tshape/TBlob. Remove setblob. Update (#2941)
     new f2b6403  Merge branch 'master' into master
     new ceecc63  Merge pull request #2943 from leopd/master
     new 2951dbb  update contributors (#2946)
     new 985553c  Caffe-plugin: 1) add listoutputs 2) fix code style & warning (#2967)
     new 3968eac  DTypeDropout (#2812)
     new ce3e11a  DType Pooling (#2824)
     new d54a2e6  Revert "KV: enable aggregate on GPU for distributed kv" (#2985)
     new a78de2b  fix the output errors about composed_net arguments (#2839)
     new a7b1c65  fix deprecations in build.jl
     new 0408e3e  fix String deprecations
     new 3145973  Change back to dict comprehension syntax for v0.4
     new 016efd6  improve docstrings
     new 5088242  Don't override help for already defined methods
     new 3bc93fe  improve documentation formatting for SymbolicNodes
     new d98e823  fix caffe stream (#2984)
     new b328f24  fix deprecation for call
     new 2c36a5d  nightly is v0.6
     new a0a4a4e  Merge pull request #115 from oist/vc/deprecations
     new 0d6e097  Change the multiprocessing relevant code (#2978)
     new 63e6063  fix Inception URL #113
     new e597cd0  Update revised for python3 compatible (#3004)
     new 13facf0  [R] default ctx in the pkg env not in global env (#3010)
     new 5d8413c  Caffe-plugin Sync cpu data/diff to gpu (#3017)
     new 450141c  [IO] Support multi dimenional label for record io (#3005)
     new 15d1727  Updata to adapt Windows (#3028)
     new 3a05d99  OP ROIPooling CPU fix and DType support (#3011)
     new d08d87f  broadcast_mask for rnn (#3016)
     new 8cc2560  DType regression (#3018)
     new 513c030  fix link in doc
     new 3a7913f  fix some formatting errors in docs
     new 63772b8  fix typos in doc
     new 08864e0  fix speech demo build error (#2963) (#3040)
     new 83dc0b4  network symbol in R (#3044)
     new acaa0e9  fix backward global_pool (#3035)
     new 916dc96  fix plot_network bottom blob drawing (#2712)
     new 50a2946  documentation cleanup
     new 6032927  Fix data number computation in NDArrayIter (#2974)
     new 71eefbb  Remove `` and :class:
     new 13257ce  scalapkg, add recordio and test case (#3060)
     new 8c22211  Merge pull request #116 from vchuravy/vc/doc
     new ab5211f  [doc] minor fixs (#3067)
     new ccc6f58  Add TuSimple (#3069)
     new ba301d1  Update index.html
     new 288c22d  Storage manager fix (#3075)
     new 33065cc  Update
     new 221611a  Get rid of sub depwarn
     new e672e42  import Compat.view
     new 08ba2d1  Update REQUIRE with min version of Compat
     new 6a8dba2  fixed bug for lstm: a possibly non existant output was being derefenced (#3063)
     new ce92c10  Revert "DType regression" (#3098)
     new de77bce  Merge pull request #117 from ranjanan/ranjanan-depwarn
     new f6fa98d  update if yarn.submit without hostfile (#2991)
     new e3b7935  reorder docs for io.jl and bring them into the markdown age
     new c126895  use `!!! note` syntax (this breaks documentation creation on v0.4)
     new 3bbd866  build documentation on 0.5
     new 84866f4  fix documentation being attached to the wrong binding:
     new da6d9e4  Merge pull request #118 from oist/vc/io_doc
     new 56673f5  [doc] steps to build deps from source (#3119)
     new 053ab42  fix overseen typo
     new b928ca9  [doc] update faq (#3121)
     new d79f9df  Merge pull request #119 from oist/vc/typo
     new b12f957  Update Julia Package doc link (#3134)
     new 73a0f6e  fix batch norm (#3049)
     new 1bb1feb  fix resources (#3077)
     new 8d21e2c  Fixing broken links (#3153)
     new e697960  add grad function to do autodiff
     new 68989b6  [Scala], add EvalMetric TopK, F1 and Optimizer NAG, SGLD, ccSGD (#3149)
     new 4403ef3  add MultiMetric and reorder metric docs
     new 12a87e5  add one-dimensional ACE
     new ab19db8  update docs for ACE
     new 0621b13  Merge pull request #124 from vchuravy/vc/multi_metric
     new a01a11b  Merge pull request #125 from vchuravy/vc/symbol_grad
     new 764a569  Caffe data layer to MXIter converter (#3103)
     new cec6e4b  scalapkg, add example multitask (#3186)
     new 637e378  cleanup docs a bit more
     new abc8ffa  Merge pull request #127 from vchuravy/vc/clean_docs
     new 3fac768  associate docs with method and not binding
     new f0b758b  move call docs into type docs
     new 1ba3896  update default dataset URL (fix test failing of #123)
     new 57aa885  Merge pull request #128 from vchuravy/vc/method_docs
     new 6840119  handle kwargs for ndarray functions
     new e0b7668  special case transpose
     new f5e80af  add special casing of transpose to the symbolic layer
     new dc43bfe  Merge pull request #123 from vchuravy/vc/ndarray_funcs
     new 9cf17b9  RNN cell demo with ptb LSTM language model (#3197)
     new 92a2d03  Bulk lint fix (#3211)
     new ae1d45b  preparing for v0.0.9
     new 9e3a60a  Positional arguments in symbolic API need type restriction. (#134)
     new 8949dbb  remove mistaken + in REQUIRE (#135)
     new 5ca890f  Fix little bug on context (#3202)
     new ad5b6ab  add PennTreeBank Language Model using lstm model in R (#2659)
     new 7a90e59  [TENSOR] Add FlatTo1D for all elementwise ops (#3238)
     new f06d994  fix special handling of dot for SymbolicNode (#123)
     new 6e2c1ae  add readme for developers
     new ebdafe8  update libmxnet SHA
     new 545332d  keep the master branch to track libmxnet master
     new 39bc643  make is_windows compatible with Julia v0.4
     new 835dd87  Add function 'print_summary' and some revise (#3161)
     new ba2c999  fix build script
     new 8f43961  prepare for v0.1.0
     new 86362f3  update workflow for creating release
     new 1a0b440  Make build.jl aware of other blas vendors besides OpenBLAS (#137)
     new d4278ca  remove unnecessary @compat call
     new 60bffbe  [Scala] Code generation for Symbol (#3217)
     new 1ff0975  replace pipe with pipeline
     new 37e715a  fix build script
     new f70970c  Revert "fix travis scala test config" (#3246)
     new 9798c97  improve buildscript
     new eb4a3fd  fix build script.
     new 7b2c67f  fix spelling errors (#3258)
     new 873b928  fix typo in (#3260)
     new 44d659a  Fixes deps/build.jl on Travis (#142)
     new e218d85  fix for skipped v0.0.9
     new d67964c  Copy slice along arbitrary axis (#3259)
     new 3e27dc5  Enable concatenation for dim-1 vectors (#3264)
     new dd7f6e1  fix PReLU backward computing (#3277)
     new ae22b6b  Add `reverse` option in Reshape (#3280)
     new f14ce88  add scala example, end2end neural-style (#3267)
     new 9dfb354  Improve multi-GPU performance (#3241)
     new 1b6b0a5  when set use_global_stats then do not use cudnn (#3289)
     new 7e53771  update dmlccore (#3293)
     new 20d870b  Fix newer version of gtest and cpptest (#3294)
     new cb6e86e  Fix req+reserve_space in cudnn_rnn (#3274)
     new ef68dae  add cudnn off option in Convolution (#3270)
     new e7514fe  add support for building on power (#3302)
     new 6bd3951  try to automatically find CUDA
     new 6267856  add recent examples, collect some missing tutorials (#3340)
     new 992a6e2  Improving docs & utilities for distributed training example. (#3341)
     new 7a2432c  ignore vscode project files
     new e6b0508  ndarray ops
     new d16b654  ndarray op test
     new 03509ae  temp commit, symbolic not passing due to ambiguity
     new b3c4d25  disable SSE for arm hardware e.g. Raspberry Pi (#3346)
     new 2d0e3ac  Add channel_ to Shape2D calculation (#3181)
     new 623fbaf  disambiguate nd op and sym op
     new d718cfc  fix API changes in random
     new 0dcde6b  Fix for LRN Layer (#3366)
     new 590055b  change examples for new symbolic calling conventions (nnvm, #146)
     new bcc28df  Documentation improvements especially on symbol.simple_bind, and notebook example (#3384)
     new 066373a  Correct a spelling mistake in /docs/packages/python/ file (#3387)
     new 4935989  Update Caffe Readme (#3412)
     new b0f0b81  Add Set Phase (#3414)
     new 06583ee  Feature/caffe importer improvements (#3397)
     new 51dce9b  Change the readthedoc link to in (#3418)
     new 286d751  Add files via upload
     new 45c1a6c  Merge pull request #147 from mcreel/master
     new eedd513  [Scala], add network visualization (#3357)
     new 21d4b68  fixed a typo in CHECK_EQ. (#3428)
     new 0cca9db  Move all the test_utils to the main package. (#3429)
     new 1d06b81  Add MXNET_GPU_MEM_POOL_RESERVE option to reserve memory for kernel launch etc (#3430)
     new 8453c0a  Update regression-example.jl (#148)
     new d7ce792  Sequence Layers (#3338)
     new 4e7803c  Try to fix the doc (#3451)
     new a5aeb0c  Update (#3413)
     new 9059c00  Merge pull request #144 from vchuravy/vc/cuda
     new 4664f3b  Add some docs to `eachbatch` (#149)
     new d09c080  Stochastic Depth Training (#3468)
     new c1dd65e  bug with model loading in R (#3453)
     new 147c6dc  fix PrefetchingIter for provide_data/label (#3393)
     new 611eedf  [faster-rcnn] support joint end2end training (#3327)
     new e0be39d  Spelling correction across all mxnet docs pages (#3496)
     new 5c01943  Renamesystemtoarchitecture (#3503)
     new 71942c2  Update Inception-BN pretrained model and correponding examples (#3506)
     new 0b18e0a  Add DQN/Double DQN example (#3512)
     new d3ba518  DQN4mxnet (#3511)
     new d97e360  Renaming Packages to Supported Languages for more clarity (#3502)
     new 23bf60c  Time major data layout support for RNN (#3497)
     new f8e8a72  capped lr scheduler in R to a minimum limit like in python (#3522)
     new 259b017  Reorganizing docs high level structure and adding github fork me ribbon (#3526)
     new 80363a1  Use executor.reshape for testing. (#3495)
     new 037a598  add train_cifar10.R and fix inception-bn in R (#3530)
     new 67613fa  Doc (#3513)
     new 8bcbebf  support global shuffling for image data (#3490)
     new 2e9d9e6  jenkins test (#3534)
     new 6505983  Doc for activation (#3538)
     new 00b5245  Cleaning up getting started, install, architecture and community chapters (#3536)
     new 6dfad3d  fix HitChunkSize in image data iter (#3540)
     new 16fd777  fix lr_scheduler for continuing training (#3251)
     new 4d9ac5b  [Scala] Bucketing API Support (#3521)
     new 1c49718  Revamped tutorials and API reference pages. Adding examples from mxnet repo to tutorials sections  (#3545)
     new cf00ca0  Crop assign (#3547)
     new cd86d5e  Quick install script for MXNet with Python on Ubuntu. Fixing few broken links and header renaming (#3560)
     new f128e4e  fix #3553: legacy copy_slice_to (#3564)
     new fbef78d  fix #3542: move layout_mapper argument out of existing API (#3565)
     new af1f7d4  more docs for dropout (#3566)
     new c70d233  gpu tests (#3557)
     new ae3b210  jenkins test for scala (#3583)
     new 8e1e7f0  MKL2017 implemented layers integration to improve IA perf. on mxnet (#3581)
     new 51322f4  doc for embedding layer (#3596)
     new 37abdac  Upmaster (#3597)
     new eee5ca7  fix missing wget #150
     new 5880b77  Merge pull request #151 from dmlc/vc/fix_150
     new 1931cf5  fix cifar10-recipe.ipynb pool shape (#3599)
     new a7c1aad  add MKL2017 experimental support for high perf (#3589)
     new 8413b28  Revert "dtype support for variable" (#3606)
     new 6ba8224  Speech (#3620)
     new 32cb6bc  Fix css (#3617)
     new 03fa025  added multifactor scheduler in R (#3629)
     new 4a5602f  fix char-lstm dataset URL (#3626)
     new a8bc3d1  Bug fixing and code refine (#3632)
     new 73bb939  Fix error when get name from grouped symbol (#3635)
     new 5f5c84b  Api page issue (#3649)
     new 267cf34  [rcnn] remove a redundant check and an argument in test_rcnn (#3641)
     new 23cb175  Update (#3662)
     new 875754b  quick fix fcn-xs example (#3654)
     new 6f2f899  Bug fixing for mkl_lrn and link to right mkl lib with iomp5 (#3646)
     new c9402f1  Feature/pad (#3651)
     new f93e53f  Support negative axis for SliceChannel (#3655)
     new 2c8d209  Fix the numerical gradient checker in `test_utils` (#3663)
     new e714d6b  Fix typo. (#3671)
     new de41c73  Update MShadow (#3678)
     new 50923b7  Feature/instance normalization (#3687)
     new 369d509  [R] Accuracy topk and update roxygen (#3714)
     new 8e05f6a  add tutorial for class activation maps (#3720)
     new c3615de  Solve Python3 compability issue (#3694)
     new b50317b  fix MKL warning in ubuntu (#3699)
     new 349b865  Add introduction of "class active maps" in example/ (#3722)
     new 5409dde  [scala] add example Gan (#3733)
     new 9107fc3  Feature/pad3d (#3742)
     new 266e439  Fix linkand head issue (#3744)
     new 290728a  fix incorrect assert (#3750)
     new f6076ab  add ssd as example (#3674)
     new 3d10b9b  Fixing adverserial example notebook. (#3619)
     new 1be257c  Fix setup toc (#3757)
     new 0e2f8ff  update README to include document for lstm-ocr prediction. (#3602)
     new 80d19ca  move julia test to jenkins (#3769)
     new 0096936  Updates setup instructions for mac to include pkg-config installation (#3768)
     new d5ae4d8  CloudFormation template to launch a cluster of EC2 GPU instances based on the AWS Deep Learning AMI (#3770)
     new d5b3d42  Fix bugs when running the code in Windows (#3775)
     new 90fcaf1  Fixing broken links. Fix format issue in set up guide. Making how to contribute page more accessible by adding links in multiple relevant pages. (#3780)
     new cd43a66  Modifying FindMKL to respect i386 and allow ILP64 (#3762)
     new 638d784  Add USE_MXNET_LIB_NAMING for DLL irregularities (#3763)
     new 807ec97  Apply AWSDoc Css (#3783)
     new 0071324  Fix make/ for aarch64 (#3765)
     new 251634f  add NNPACK support for high convolution inference perf (#3666)
     new 1e51314  Typo: gpu to cuda (#3774)
     new 65db4ca  API class and function formatting (#3789)
     new 333298b  Update CloudFormationTemplate and add Readme (#3790)
     new 465fac4  Update AMI ids in cloud formation template (#3795)
     new 94f67a6  Making mxnet introduction more detailed. (#3798)
     new 09519dc  doc changes: MKL readme edits, fixing 404s (#3802)
     new 76ac4b8  Fix heading in intro page. Fix broken link. Bring Python main API reference headings to top header under API (#3804)
     new 5257652  Less redundant Reshape spec (#3781)
     new b341bfb  [Scala] add rtc (#3788)
     new 5c255f2  spelling/typo fixes (#3815)
     new 6f38450  DType support for data iterator (#3640)
     new 599b8c1  Update new AMI Ids (#3826)
     new 7c56813  add links recent examples and notebook, correct author github info for license plate recognition (#3828)
     new f30d640  reorded cv tutorials and added descriptions that reflect content and … (#3825)
     new 46d0d1b  Docs: tutorials page reformatting TOC (#3838)
     new aa2520b  fix im2rec index (#3839)
     new fba4e01  Fixing API reference docs for parameters by adding blank lines to docstrings. (#3841)
     new bc25127  update url (#3842)
     new 01cde15  Fixing bug in NDArrayIter.provide_data (#3844)
     new e4479d2  [doc] build mxnet-notebooks into (#3845)
     new 39dce6d  Only enable SSE2 flags when detected (#3854)
     new 9f7dd02  fix example ssd variance vector initialization (#3840)
     new bfa7332  Making grammar correction, spell checks and language tone improvement. (#3834)
     new 2562070  Cleaned up tutorial index, linked new content, moved language specific feature walkthru to the how_to section  (#3859)
     new 36bc437  [doc] Build R reference pdf manual (#3862)
     new 0fe6e0c  fixing links (#3863)
     new 428da08  Pass CMAKE_C_FLAGS to CMAKE_CXX_FLAGS (#3851)
     new ea72849  Expandable toc (#3866)
     new 635bafd  Add me to contributors. (#3867)
     new 0dd490c  Fix decoder for speech demo (#3865)
     new 08e065d  Grammar and language correction to pages under architecture. Fixing broken links. Adding redirection place holders for old packages and system sections. (#3872)
     new 146f373  spelling fixes on tutorials page. removing bad link to mxnet notebooks (#3874)
     new d5b2294  url fixes (#3875)
     new 9a6ebe1  refactor test util (#3861)
     new bc8074d  fix glog (#3881)
     new 3871108  Updated URL (#3885)
     new 672ed36  Grammar and language correction under API section. Restructuring Getting Started Page (#3887)
     new d3c1b6c  module grad_req fix (#3880)
     new b8f1e95  add aws logo in the front-page (#3889)
     new cddc830  R tutorials; link to tutorials for generated tutorials (#3890)
     new 7812647  adding scala API docs (#3879)
     new 28564a5  adding scala api docs to doc build (#3891)
     new 16060b9  filling out scala and r API pages (#3893)
     new ece16e5  pointing pretrained models link to predict_imagenet notebook. yet more url fixes (#3897)
     new d7a8b7d  mx.viz.plot_network renders all inputs not just first (#3896)
     new 718c885  [Scala] update Visualization.plotNetwork (#3900)
     new a48daee  Improve sidebar (#3898)
     new 38b8334  Update
     new 66b263b  Update (#3903)
     new e971c0c  [R] ImageNet training with ResNet-v2, Inception-resnet and others for R (#3832)
     new 2fefed4  Fix broken links. Add link to stack overflow. Make supported language claim consistent and generic. (#3905)
     new 28ffed1  fix PrefetchingIter doc string error (#3909)
     new fdf7374  fix #3892 - adding MNIST scala tutorial (#3911)
     new 9e767a2  Adding more details about supported programming languages. (#3922)
     new 486adee  Adding links to AWS blogs talking about using mxnet on AWS resources preconfigured with mxnet. (#3923)
     new 703bf70  Regression fix (#3925)
     new 46b3e71  adding link to mxnet-notebooks to generated tutorials (#3926)
     new 314835d  Fixed comments in README for R package (#3928)
     new 560eb4e  CMake: Add preprocessor defines for CUDA/NVRTC (#3929)
     new 2559e53  added to contributors (#3931)
     new bcd4a33  RL examples (#3935)
     new eaa6253  Adding quick installation script for installing MXNet with Python on Amazon Linux (#3927)
     new b52fe32  [R] implemented speedometer for R (#3932)
     new e452664  Better error messaging when infer_shape fails (#3848)
     new 3332950  Better error messaging and some docs improvements (#3827)
     new d8e3f49  Tutorial to Notebooks  (#3943)
     new 2205df6  Better error messaging in tool (#3942)
     new cfa779d  checking for correct input to sequence layers (#3940)
     new 37f61d6  fix readme (#3945)
     new 35b5346  [R] Add early stop callback function (#3938)
     new 2397ab4  [R] Script installation for R  (#3803)
     new 311bcf9  Sidebar auto expand (#3956)
     new 58970f5  Update copyright (#3957)
     new 00735b2  fix error: bash: cd: ~/mxnet: No such file or directory (#3960)
     new 405d36a  remove .lst before rename in (#3958)
     new c507056  add Wei Wu to contributor list (#3962)
     new 33ce751  Added MSE in R (#3968)
     new abbb5b9  Fixed an inconsistency between the load_param return value and the received value (#3966)
     new 8aad0b1  fix-python-path-in-setup-utils (#3963)
     new 4ac6730  fix a few dependence issues in (#3964)
     new c6db878  New callback interface for training visualization. (#3849)
     new 8efc3c4  added custom mask type (#3916)
     new 550ae83  typo error of relu name (#3977)
     new ddc27e0  MXNet benchmark scripts (#3971)
     new e9983ac  checking if args has a "benchmark" attribute (#3993)
     new 487c22a  [Scala] add Mixed Initializer & fix some spell error (#3998)
     new fb1ed65  Remove unused backward dependencies. (#3681)
     new 8d5a9ab  Update
     new 04a0779  a major refactor of example/image-classification (#3999)
     new 22ca86e  Add example with sub-word units representation for word2vec. (#3990)
     new da9dd6f  Recommender system examples (#4021)
     new 1fa148b  [doc] Website search engine (#4015)
     new 241020a  Fix nvidia logo (#4023)
     new 9c284f9  Update train_imagenet arguments in benchmark scripts (#4024)
     new 3a60c95  [doc] update readme in example/image-classification (#4025)
     new a3a928c  [example/image_classification] run benchmark on 2 power number of nodes. (#4030)
     new 5a6ba9c  fix a missing word. (#4039)
     new 22c8640  [example] add inception-resnet-v2 into image-classification (#4032)
     new ccb1c97  fix compile error with vs2015 and new dmlc-core (#4034)
     new 5296003  add a fine-tune link in how-to (#4035)
     new 0867f6a  Add HKUST LOGO (#4066)
     new d647e58  add burness into (#4063)
     new 6442b05  Fix confusions between comments in variables and their descriptions. (#3835)
     new b9b5e45  Fix typo  (#4073)
     new e905f8e  Fix computer_vision/ typo (#4072)
     new 9d066ee  [doc] Added additional AMI option to Cloud Setup (#4069)
     new a1cc8f0  [kvstore] no warning if p2p is already enabled (#4075)
     new feb8762  [example] Python2/3 Support for example/image-classification (#4081)
     new 54b3dc4  [tools/bandwidth] fix, add (#4076)
     new 6666f64  [example] fix ddpg bug (#4078)
     new 377e70a  [example/image-classification] add two pre-trainede models (#4086)
     new 2c02e89  [example] moved notebook to dmlc/mxnet-notebooks (#4092)
     new 87f218a  [doc] Update What's New in Readme (#4096)
     new 6868a66  [doc] add "model zoo" in the navigation bar (#4122)
     new 6900c40  [doc] Add a model zoo page (#4123)
     new 1bb4ae9  [doc] fix format in modelzoo/
     new 0ac4b8f  [doc] convert markdown tables into rst before building (#4125)
     new a2cc2d7  [docs] also convert md sentencies in a table into rst (#4126)
     new 6a7a892  [docs] remove \n\r in pandoc converted results (#4127)
     new 56d3092  [example/image-classification]  ResNeXt symbol (#4108)
     new fc874df  fixed main() of (#4131)
     new 092ff44  MKL feature enhance (#4128)
     new 3f6cc9d  Add CUDA 7.5 and 8.0 Dockerfiles (#4114)
     new d31bd99  Fix amalgamation scipt. (#4135)
     new bab6d61  fix error when enable monitor for rcnn (#4136)
     new c05efbe  Improve build system to avoid calling pkg-config many times (#4111)
     new 8601f25  Update (#4142)
     new d1eb318  Bug fix updating the ACE loss function. Included a test case. Also provided an example.
     new a457228  Add add_stn parameter to avoid error (#4148)
     new 327132d  Create
     new 2829115  [example/image-classification] add distributed scalability results (#4167)
     new 29143dd  change default quality from 80 to 95 for bigger and deeper network (#4164)
     new 7907e45  fix segmentation faults in c api (#4158)
     new 55f728d  [Python] Python2/3 Support for examples:autoencoder, bayesian, bi-lst… (#4106)
     new 1e7332b  Model zoo style (#4174)
     new c06b211  Merge pull request #156 from JockLawrie/pull-request/d1eb3183
     new 85844f1  Fixed bug in MSE (#4179)
     new 5b6a0ee  [DOC][R] simple docs for using own loss function (#4181)
     new 9622714  Add a note for warpctc's example 'python' (#4133)
     new cffa7d0  Reorganizing setup and installation guide by OS and Programming Language. Fixed errors in Python for Ubuntu, Amazon Linux and Mac installation (#4218)
     new ebc9615  Change model zoo table style (#4239)
     new d201b16  Adding How To for visualizing network graphs. Updating install guide to install graphviz and jupyter (#4240)
     new 87c6404  [Scala] add CustomOp support (#4118)
     new 4d6b24b  Improve Navbar style (#4251)
     new 8f2d58b  [example/image-classification] scripts to benchmark scoring (#4084)
     new e3a41ae  More modification for navbar (#4252)
     new 93680f1  [DOC][R] add example code to read the original MNIST data set (#4254)
     new 917d759  Repository is missing (#4109)
     new c9437cd  Adjust navbar (#4263)
     new b158ebd  update info for resnext50 (#4266)
     new 041fcce  Fixing LDFLAGS issue in Mac setup (#4265)
     new ffac62d  Fix memory leak bug in Monitor (#4269)
     new 31ede10  fix a typo in docs/tutorials/computer_vision/ (#4268)
     new 1d532fc  add path checking in example/image-classification/common/ (#4270)
     new d187d14  Explicitly exported the call to the context, so it can be easily modified for use on machines without GPU support (e.g. OSX) (#4262)
     new 1036277  Adding description to How To links and Tutorials. (#4284)
     new ee61ac1  Add some Julia docs in Getting Started (#4291)
     new e62fbe1  [example/image-classification] add  convergence rate (#4292)
     new 1446bfc  [image-classification] default use min_random_scale=1, add data/ (#4295)
     new 221846b  [Doc]Add MXNET_BACKWARD_DO_MIRROR in MXNet environment description (#4300)
     new 15d32ca  fix resnet-v2 segmenet fault problemwq (#4237)
     new 8efcfee  [image-classification] add resnet* into (#4309)
     new 074a9ef  fix awesome link and typo (#4311)
     new 7135d67  [doc] Fix link to C++ example (#4320)
     new 68971d3  Add get started backto nav (#4326)
     new 1d2ddff  fix cmake build with vs2015+cuda8.0+win7 (#4334)
     new 1a71e3b   sync aux params across devices after each epoch (#4327)
     new 1610b4f  [Scala] NDArrayIter constructor fix for null (#4308)
     new 2931dc4  Add go binding project reference (#4336)
     new d05e82e  'lst'->'.lst', means format like '.rec' (#4349)
     new 266828e  Adding a link to setup guide from How To page. (#4324)
     new 0a03417  Typo in (#4343)
     new 1ff820a  Fix underflow bug in GPUPooledStorageManager::Alloc(). (#4356)
     new 9448efe  Update (#4381)
     new dc1a360  Add mxnet.notebook into (#4378)
     new 1ae2be5  [doc] update how_to/ (#4353)
     new f92ebf1  add support to monitor aux params (#4363)
     new 4586bc5  [doc] GPU installation for mac and ubuntu (#4325)
     new 67bee19  	[RELEASE] v0.8 Release (Last release before NNVM refactor (#4409)
     new 04f49aa  Update kvstore_dist.h (#4413)
     new 6546722  refactor test util
     new 3e0210b  forward test
     new bcc7547  NNVM Refactor (#3194)
     new e59cb57  [NNVM] Imperative Invoke (#3208)
     new b2d222c  [scala] link libnnvm.a (#3214)
     new a541a1a  [PYTHON] Optional Cython Module for Symbols (#3242)
     new 0052500  [EXEC] Support fcompute (#3249)
     new 43528de  [OP] Add alias support (#3261)
     new 43ff947  Fix path in (#3276)
     new b4f744f  [WIP] Element wise op refactor (#3245)
     new b261bc9  update nnvm version (#3290)
     new e027f1c  Fix breaking changes after pull master (#3291)
     new fbc6910  add nnvm cmake with windows (#3255)
     new fe57039  [CYTHON] Cython module for NDArray (#3292)
     new d94dc9e  [NNVM] change of attr to set_attr (#3303)
     new 725ced4  Update
     new 47507c9  [WIP] binary broadcast wip (#3301)
     new 40b335a  x (#3308)
     new 7637627  [IO] Python based ImageIter and Augumenter (#3227)
     new 6f0dcda  [OPT] NNVM Optimizer (#3314)
     new 758d529  fix cpython in windows (#3309)
     new 550fff1  Add Mathematical functions (#3317)
     new dfe1db5  fix image io
     new 32bc762  add hypot degrees radians cosh sinh tanh arcsinh arccosh arctanh (#3335)
     new 96d4ae4  Fix metric &
     new 42a3147  [Scala] Nnvm ops for NDArray & Symbol (#3361)
     new fc63979  [R] Fix the R interface (#3334)
     new 85478cb  Locate compiled library on Windows (#3369)
     new da3cbc3  Fix metric & (#3375)
     new 61482dc  Update legacy op FBackwardInGradIndex (#3376)
     new fb9f96d  add backward index
     new 30e4e34  include some special functions (#3337)
     new c01abbd  fix kv build (#3385)
     new a210bd6  close# 3343 (#3421)
     new e927fc3  move sample ops to nnvm (#3436)
     new f9a672f  Fix Shape inference problem for dtype (#3456)
     new 2ed210b  [EXEC] Enable inplace sum optimization (#3470)
     new d012d3c  move matrix op to nnvm (#3472)
     new ef61c51  [OP] Move elemwise sum op to tensor fcompute, fix broadcast_to backward, fix roi pooling. (#3476)
     new a02a1e7  Fix hanging when kNullOp exists, safter checking in executor (#3482)
     new 0e1cd57  MXNet Profiler (#3163)
     new 4131c0c  [scala][nnvm] add more nd functions and dtype support (#3481)
     new 1e2a00d  Nnvm (#3485)
     new 5df21a9  revert remove simple op (#3501)
     new 7bb72db  fix type (#3519)
     new 1d06fa2  fix sample op (#3551)
     new fcb4a19  fix test
     new a73afab  Nnvm scala fix (#3590)
     new 346d6de  [Scala] DType support (#3658)
     new d7a8a52  Rebase master
     new 5c7cb8a  log10, log2, rint, fix (#3672)
     new 1fd167f  Update
     new 4bc3497  Update
     new f5cde41  Add zeros, ones, fix compiler warnings (#3679)
     new e719292  fix after rebase
     new 11456ba  [WIP] json versioning (#3894)
     new 34c640d  Formalize broadcast and elemwise ops (#3907)
     new af1a640  [profiler] add failure information (#3959)
     new cfd3075  enable hidden arg (#3939)
     new f755dc4  Upgrade nnvm to use automatic gradient correspondence guessing (#3973)
     new c2581d6  Fix ccAdam in nnvm, add ccoptimizer testcase (#3986)
     new 627e71f  only return visible outputs for imperative (#3996)
     new aa4c477  refactor logging (#4005)
     new 7ba8307  Add simple equal operator (#4027)
     new be816bd  allow using numpy dtypes for symbol and ndarray functions (#4006)
     new 22f3ce9  post rebase fix
     new e7a2735  [EXEC/NNVM] Enforce sum order for addto optimization (#4093)
     new 07c564d  refactoring EmbeddingOp based on NNVM (#4097)
     new 62ce534  [OP] Move reshape and flatten to nnvm (#4101)
     new ca535da  	fix operators and improve error message (#4132)
     new a44d6c4  deprecate model (#4138)
     new 6b91837  fix compatibility (#4144)
     new 5ee2c21  close #2185 and close #4095 (#4160)
     new 941e9d3  	split tensor operator files (#4150)
     new d5ec111  add example of using pipe (#4175)
     new 209a4ac  Fixed bug in ReduceAxesBackwardUseInOut that caused finding the gradient (#4171)
     new 241af0b  fixed error in as.matrix.MXNDArray (#4185)
     new dc0270b  post rebase fix
     new c03e1d2  [Keras] Add support for element wise comparison operators #4182 (#4190)
     new ad4d1d7  Temporarily disable julia_test
     new 9dc0403  custom iterator tutorial for R (#4203)
     new 2e1b286  fix executor destruction (#4208)
     new 1b782ca  Add test case for RHS and LHS scalar operators. (#4229)
     new d40c2ac  fix cython ndarray error handling (#4209)
     new 5bfbd17  Docs for NNVM new operator (#4278)
     new 2d03fb1  refactor elementwise ops (#4227)
     new 8d630c8  [OP] Implementations for prod, nansum and nanprod (#4207)
     new fb58771  fix example (#4210)
     new e18ff28  fix print_summary and add one test case (#4321)
     new a9ec282  Convert to new programing API (#4307)
     new 55cea1d  [CALL FOR TESTING] Single script to build and install mxnet for Windows with Python (#4256)
     new 02cd2da  fix (#4212)
     new c91f948  fix compilation under gcc-5 (#4346)
     new 9feab1c  Variable patch (#4345)
     new 28a1cb2  fix ndarray bcast (#4352)
     new 6706168  Layout (#3498)
     new c4113f3  0.9 release note (#4293)
     new e52bd5c  fix recordio (#4344)
     new 3861568  add back lr_scheduler in classification example (#4393)
     new 835d690  fix cpu ctx (#4394)
     new 6f95b74  Fix the index check in ndarray. (#4398)
     new 82f53b7  Introduce a developer mode option in Makefile (#4364)
     new 12d51e8  reduce default lr to 0.05 for mnist (#4400)
     new 528dee0  [Module] add save load optimizer states (#4408)
     new 04dc056  fix hidden keys (#4414)
     new f52cf30  Fix segment fault when profiling (#4405)
     new c2cc269  fix
     new 986b736  [Release] v0.9.1 pre-release (NNVM refactor)
     new b81b26c  New Optimizers (#159)
     new c1aa01b  fix docs (#4429)
     new fd0f9b7      mkl fast path patch update (#4433)
     new 53d1ead  fixing r pdf docs -- generating Rd files (#4432)
     new 84e5155  [Scala] update Visualization.plotNetwork (#4423)
     new 21409ac  doc improvement (#4435)
     new 9cb8d5e  Improve profiler (#4434)
     new c48fdfb  Improve OSX install experience (#4351)
     new ecb0bd7  fixing typo in the R manual command (#4445)
     new acd6ff4  fix unordered map (#4443)
     new d0728ca  fix memmonger (#4441)
     new 4920480  fix build warning (#4448)
     new 7a04bb1  fix complie error with vs2013 and cuda7.5 (#4439)
     new 2adf449  fix dropout and block8 (#4458)
     new 440b637  Document the enviroment variabls of the profiler (#4452)
     new 9fc2246  Update (#4463)
     new 4bb8037  Update (#4467)
     new 0556c78  update submodule (#4459)
     new cad90b6  Update (#4473)
     new 342b2d9  add inst, src/*.o src/*.so to clean rule for R-package (#4472)
     new c3c2171  Enable warning as error (#4451)
     new a21f0fe  Update
     new 36deae0  Use mean data for prediction (#4475)
     new 53622f8  Fix syntax error with C compiler in mxnet/c_api.h (#4476)
     new cb710d9  add mxnet install script for fedora python (#4478)
     new 7eb4cec  add access to attributes of R symbols (#4474)
     new 7d79906  add profiler to R Interface (#4484)
     new e9fc00a  add new Chinese translation 'docs/zh/tutorials/computer_vision/' (#4483)
     new b5988bd  update ps-lite (#4461)
     new 4ef5e49  Build SSD operators in jenkins (#4481)
     new 5b14a0e  Fix JQuery version (#4501)
     new ad2e0d1  bug in scala package: scalar times symbol (#4496)
     new d4af261  Fix Jquery 1.11.1 (#4503)
     new 6e285ef  Revise .gitignore (#4506)
     new f161928  test_rmsprop (#4507)
     new b82ac92  [test] Wrote Amazon Linux CPU CI setup (#4430)
     new 4861309  Automatically download data (#4510)
     new fbb6885  Add transpose kwargs for dot (#4498)
     new f4b8317  A3C example (#4516)
     new ceb9f01  Update (#4517)
     new a9efceb  fix image_file in line 133 (#4518)
     new 50a3a31  Pop up version number in ubuntu-r setup script (#4464)
     new 512bb8d  fix custom op (#4528)
     new e06c55d  Convert iterator to mx.img API (#4442)
     new ec78752  close #4527 (#4549)
     new 1ae2905  logging when save params and optimizer states (#4534)
     new 6b6ed56  [kvstore] use CPUPinned only if there is more than one GPUs (#4550)
     new b67a445  Fix predict c api (#4537)
     new 12cdde6  Adding missing commands to pull MXNet source code in installation steps of OSX, Ubuntu and Amazon Linux (#4556)
     new 999aa3f  PythonInterp (#4542)
     new a424d24  Reduce the overhead of profiler (#4544)
     new 803ba5b  Docs fix: Recursive -> recurrent in RNN. Downplaying model zoo until … (#4555)
     new 1ea44c9  Fix hygiene bug (#164)
     new 4d0aa87  Fixed visualize (#163)
     new ab80048  ENH: added verbosity training option to control printouts (#162)
     new 6ef3058  [EXEC] Specify external memory directly in memory planning (#4558)
     new b24c6c5  fix warpctc plugin work with new version warpctc (#4530)
     new 3a820bf  Fix search function (#4570)
     new 9f9c135  [OP] Topk and arange + Update submodules (#4565)
     new 614fd94  Fix search for sphinx 1.5.1 (#4579)
     new 29307c2  [nnpack] update && support more op (#4519)
     new 98fcce0  fix compile error (#4590)
     new 8159ad6  [OP] take (#4538)
     new c34a023  enable transpose flags for batch_dot (#4595)
     new 2d21179  Usage example for module (#4428)
     new 2393d4a  Fix scalar operations in NDArray (#165)
     new d824f27  Update (#4604)
     new 761b161  fix typo as mentioned in #4588 (#4605)
     new 29ba590  fix imageiter and add profiling (#4598)
     new c5e59cb  support profier for native engine (#4513)
     new f7e5a64  Enabled loading of saved NDArray with GPU context in CPU only environment (#4608)
     new e1cafff  Scala docs (#4615)
     new 093844e  fix channels not convert from BGR to RGB when len(layer_blobs[0].shape.dim) == 0 (#4625)
     new d8e64a6  update docs && fix typos (#4624)
     new 54ce28e  Fixing missing XML R package and installation issue for R on Ubuntu (#4620)
     new d4457ed  Added epsilon to `CrossEntropy` loss to avoid divison by zero errors. (#4594)
     new 837fe9b  fix custom op for naive engine (#4585)
     new 0907d98  Python scala topic broken links fixed (#4628)
     new 88e6106  Revert "[OP] take (#4538)" (#4627)
     new dc859fb  fix when image_encode failed in (#4636)
     new 2123ac4  fix cpu/gpu copy (#4626)
     new 0e9f56a  Fix operator arguments (#4575)
     new d6328a5  Add predict sample code for lstm+ctc ocr. Also update it's (#4365)
     new 541d109  Add prelu layer support for caffe convert tool (#4277)
     new feafcb0  fix error (#4643)
     new aa04648  Adding Mac quick install script for mxnet with Python (#4642)
     new 596ffd4  Minor CMake build changes and fix a couple of signed/unsigned warnings (#4641)
     new 5656c86  Add DC-ASGD optimizer (#3614)
     new d5238ac  Update cudnn_convolution-inl.h
     new 179daa5  Fix of bugs in nnvm branch (#169)
     new 1b09fb4  Merge branch 'nnvm'
     new 4d0ee6d  Update (#4670)
     new 6763b2e  unplug a warning under VC compile (#4658)
     new 030990a  BUG: fix two small mistakes with verbosity control (#168)
     new 1f976e9  Update docs/how_to/ (#4678)
     new e482d7e  Update neural art tutorial (#4677)
     new a0ec859  Create ImageIter from a in-memory list (#4612)
     new 22a3c72  Drop support for v0.4 and Deprecations for v0.6 (#171)
     new ae37519  fix inception-resnet-v2 example (#4686)
     new 72dd135  Fixed reshaping of input, was erroring out (#4687)
     new cfe4b76  [Release] v0.9.2 MinPy compatible release
     new 6d05979  Copy3 (#4680)
     new cd25fb4  Fix Amazon Linux Installation Title (#4701)
     new d49d566  Reorganzie get_started page content (#4703)
     new 8abf132  add new OP take (#4715)
     new eaae080  Update
     new 571b9a4  Update
     new 25a056c  Update (#4723)
     new 949300d  Update example/fcn-xs/ (#4706)
     new 387b8df  Refactor zero gradient + Revise test of binary logic ops (#4732)
     new 2916b4e  Update rcnn example with accleration and support for module, resnet, coco and nnvm (#4730)
     new 438211a  added fix for nonsquare stride during backprop (#4704)
     new 885cdc3  [Scala] fix error when run example ExampleCustomOpWithRtc.scala (#4718)
     new 36a6658  Add Index2d (#4702)
     new 612a3ff  Improve index layout (#4740)
     new bb27c6c  Add other cc projects into CMakefile (#4657)
     new 2a66ea3  Update build.jl (#174)
     new f96fb86  Fixes mxnet additonal deps: python-matplotlib is an apt package, not a pip package, and the pip package for scimage is scikit-image (#4748)
     new 1069f28  fix gpu_num_ in cpu only no initialization error. (#4747)
     new c141542  fix image-classification import error with python3 (#4681)
     new e793ecb  fix a replicated beta params bug in cudnn_convolution (#4649)
     new 1bdf415  Fix cuda error when multi process is used (#4695)
     new 7444084  update MKL_README (#4599)
     new 5ca5a05  argument error (#3265)
     new df81803  fix conv / deconv bug on non-square stride (#2365)
     new 1b3d6d1  Fix warning when CPU only (#4751)
     new 339270b  add symbol clip (#4744)
     new 94bfe49  update
     new 0ea996e  update submodules
     new ead4be4  pip package starter (#4754)
     new 04be299  Enhencement on build (#4760)
     new 5f3ae0e  Fix launcher log message (#4762)
     new 0964409  Fix build on OSX (#4759)
     new b11d3a2  [RELEASE] v0.9.3 official release (#4767)
     new c414f7e  Keras support (#4690)
     new 2413a8e  Making spelling and general grammar corrections (#4771)
     new 6df126b  Add tutorial link to index. (#4784)
     new ac5364b  add  docs/get_started/ (#4788)
     new 01ae4df  Update Makefile
     new d7f4be3  fix (#4797)
     new b45b843  Fix voc_ap issue in rcnn test path. (#4796)
     new 9a757ba  [Scala] Add default name to NDArrayIter data and labels (#4781)
     new 62e91d5  [OP] Negative axis, begin, end and end=None for slice_axis (#4778)
     new 340f965  New AddTakeGradLargeBatch and SortByKey for GPUs (#4782)
     new ba89c0f  Jenkins notebook test (#4799)
     new 3aaf94f  Doc (#4802)
     new 3cfd2ba  Modify the caffe_converter to use the new api. (#4655)
     new 8659b75  Update viz.graph for R (#4776)
     new c07fe6f  Update Makefile
     new 64bdd25  use shallow for cub (#4807)
     new 3ad5819  update
     new 117587a  Imporve two Ops (#4808)
     new d2fae31  Fix cmake for cub (#4810)
     new 14781c1  Add docs/get_started/ (#4790)
     new e9c3029  add BatchNorm data gamma argument (#4786)
     new 7c197d6  Fix amalgamation on Mac
     new 7165476  [pip] script for making python wheel for mxnet, CPU-only version (#4814)
     new cc59358  Enable addto for deconvolution (#4812)
     new 8d044b1  refactor windows support
     new 15c7052  Merge pull request #186 from dmlc/vc/windows
     new e086a1f  Clean/add test environments and scripts (#4813)
     new b039771  Using WebRequest function to download the JSON file
     new 6335eca  fix vectorized abs deprecation
     new d9f9551  don't include code from within a function
     new a6968a6  deprecation of filter and Array creation
     new a79d002  readall deprecation
     new a476d3e  MXNet.jl on v0.6 is currently subtlety broken due to broadcasting.
     new 11deaab  Merge pull request #188 from bmharsha/Fix-Windows
     new ba3c9fb  Merge pull request #187 from oist/vc/travis
     new fecfcdd  fix use of Symbol in symbolic-node.jl, fixes #189
     new fb2a0d0  Merge pull request #190 from dmlc/vc/fixup
     new cac5625  refactor sgd.jl to exploit multiple dispatch
     new d0d7963  remove BaseTestNext and bump Compat
     new d59b214  Merge pull request #192 from dmlc/vc/review_comments
     new abf4b77  [OP] add BilinearSamplingOp and GridGeneratorOp
     new 6b6bf26  try fixing bilinear sampler + grid generator
     new 9cc4d01  Update
     new 0ee926e  remove usr/setupenv.cmd because it is too invasive
     new 79d80b6  Merge pull request #196 from dmlc/vc/fix_win
     new deb803b  [scala-package] remove confusing warning logging (#4829)
     new 3ffbc35  fix typo (#4828)
     new 28b4540  [scala-package] add apache license header and corresponding scalastyle config entry (#4830)
     new 0467dc2  fix reshape for 0 shape (#4834)
     new 0f6d583  fix leakyrelu shape (#4835)
     new 5568641  fixes loading of voc evaluations for empty classes (#4840)
     new f9f0c99  Adding a tutorial template jupyter notebook to enable easy and standard contributions (#4843)
     new dec952a  refactor cast (#4842)
     new ae4c14d  Fix typo in error message (#4855)
     new 46de615  fix broken url (#4859)
     new e6ccbd5  call mkdir modified (#4848)
     new 254e70a  added rpi setup instructions (#4692)
     new 30df58c  close #4870 (#4874)
     new bc710bd  RMSProp Update as NNVM op (#4873)
     new 5572027  Fix lots of typos in documentation (#4875)
     new 8b9d909  [WIP] Tentative RNN Interface (#4618)
     new eabf31c  Accuracy Function For Speeding Up LSTM+CTC Convergence (#4865)
     new 90a0327  Update
     new a273078  Refactor initializer (#4644)
     new 75060bc  Fix for symbolic node (#199)
     new 32dc3a2  Refresh viz.graph (#4892)
     new d94b9cb  [R] switch to Apache-2.0 (#4895)
     new bf92d7a  incremented dmlc commit (#4904)
     new 23f7e39  bug fix for python 3 support in python/mxnet/module/ (#4903)
     new c069c77  Remote use of __shfl_xor() operation to keep compatiblity with sm20 (#4905)
     new 430ee91  Add repeat and tile ops (#4863)
     new 277994b  Enhance infershape for concat & dot  (#4907)
     new 5e24eec  fix amalgamation (#4913)
     new e9da515  Fix rmsprop on GPU (#4914)
     new 2f222cf  fix ssd demo (#4890)
     new 912a7eb  Don't permit to use gpu context in CPU only build (#4915)
     new cf28f92  Fix bilinear Upsampling (#4923)
     new 39a2c0c  fix ListOutputs in slice_channel op (#4920)
     new 6f08c0c  Removing two confusing line from CustomMetric (#4938)
     new d753d45  Fix gpu consistency test of upsampling (#4937)
     new aea0d51  Fix some doc typos (#4949)
     new 00dd356  Fix wine-detector page (#4958)
     new 9ebd906  Fix some doc grammar issues (#4959)
     new a5b7193  Fixing double cloning issue in osx install script. Adding manual install guide for python. (#4930)
     new 70df5ad  Add one hot op (#4950)
     new 45a2a4a  change recursive make commands per gnumake manual recommendation (#4983)
     new f4c6d55  Scala documentation (#4979)
     new 3d78151  Improve numerical gradient and fix tests (#4986)
     new 47ba7b9  suport MKL ml 02/05 new feature (#5000)
     new e622438  fix l2normalization (#5011)
     new 3b9d82d  Fix bug in CPU memory allocation (#5017)
     new ba51c0a  Fix output_shapes property of BucketingModule (#5030)
     new 9f54daa  update dmlc-core (#5027)
     new a4060f7  fix index bug caused by mx.img.ImageIter (#5025)
     new 5e9f4e3  [Doc] Fix doc of optimizer ops + fix doc of dtype (#5012)
     new 0fc54d3  MKL compile update to remove full mkl pack dependency for blas=mkl (#5036)
     new 4b0665e  [Scala] add Profiler support (#4854)
     new 22673b6  add fused rnn cell (#5004)
     new 15396a4  Enable R tests in Jenkins (#4976)
     new 2d473f2  update perf. number on CPU (#5049)
     new 898c4e7  Fix 3d conv&pooling infer shape check failures for valid kernel sizes (#5048)
     new b4960cf  fix bilinearsampler unittest bug (#5044)
     new 0aeddf9  fix speech demo, add timit demo (#4952)
     new 4ed66ee  Add where op (#4984)
     new 9f41ba2  Optimized broadcast and reduce CUDA kernels (#5018)
     new 1d61acb  fix tests
     new aa375bf  [scala-package] generate source jars correctly and a large cleanup of pom files (#4936)
     new 9ab7d32  fix rmsprop optimizer (#5043)
     new 3e61db1  Add ckomaki to for #5048 (#5053)
     new bfc75a2  fix repeat and tile error (#4935)
     new 23fff3f  Update
     new a416ed0  [WIP][Scala] Module API Support (#4583)
     new 94de855  fix crash (#5088)
     new 2a7d6f2  Run tests in Docker as current user (#5086)
     new 6a0db69  Perl5 interface to MXNet (#5022)
     new df9f612  Fix to alamgamation so that JS build target works using latest emscripten. (#5090)
     new 305f753  fix mx.img.resize arguments order (#5072)
     new 8e995ae  refactor slice (#4974)
     new 8328664  Doc spell checker (#4985)
     new c83624c  fix name creation in ListArgument functions (#4970)
     new 9a6d8c3  Importing for all collobarators log on mxnet home page to mxnet repo (#5106)
     new 252ac62  updating aws logo reference from svg to png (#5107)
     new 7e3f5e5  Update (#5111)
     new 2f826dd  fix indent error (#4966)
     new 3439f97  Update
     new f33f1dc  get_name (#206)
     new 7fc3db5  Support incomplete shape inference for slice_channel and convolution (#4898)
     new f03da63  Fp16 training for Alexnet (#5019)
     new 8294809  add infer stage for lstm_ocr. (#3386)
     new 38a33ef  Amalgamation JNI bug for Java/Android prediction (#5109)
     new 5298aa6  Update
     new 5e5f19a  [MERGE_MINPY] add, a more verbose customed logger (#4911)
     new 33930c6  Update Matlab demo script and get it working again with the HEAD of MXNet. Improve readme documentation with information on downloading model and running the demo. (#5148)
     new c9e252f  fixed run the several times and will get different results (#5143)
     new dc47e06  add api symbol.get_children (#5141)
     new fd7fe25  Fix FusedRNNCell input to be time-major if passed a list of (batch-major) tensors as inputs (#5166)
     new fa80891  add states to modules (#5104)
     new 984ebd4  Fix char RNN dataset URLs (#5168)
     new 0ce9c69  	partial set params (#5167)
     new 25e9665  fix conv addto mode test failed (#5144)
     new eeeab1c  remove redundancy head in storage (#5160)
     new 884b50a  Memory allocator bug fix #5035 (#5133)
     new 7757b6d  bugfix: set correct default values to 3d pooling padding and stride (#5096)
     new 7688982  solved issue with multiple outputs in visualize in python (#5183)
     new 941ba5c  Fix the incorrect unit test for cpu naive storage (#5175)
     new 675731b  fix Profiler timestamp when using VC12 (#5173)
     new a9751fa  add svm output gpu code (#4541)
     new 17a5442  fix data URL error (#204)
     new 4961c9a  add data_shapes check (#5190)
     new 7510ef1  Correct handling of weight decay in Adam and RMSProp, uncentered version of RMSProp, weight clipping for RMSProp (#5137)
     new 67eac20  [op][keras]add reverse (#4973)
     new 1fd940e  Add stacked fused RNN example (#5186)
     new 65a7f3e  add parallel actor critic algorithm (#5199)
     new 1ac3536  Update (#5207)
     new 7cd0610  Add sudoers permission to docker user (#5179)
     new b4e8743  Added ResNet support to Caffe converter
     new be38c5b  Syncing up perl-package to current state of the Python interface (#5165)
     new 9d19242  Fix 3d conv inter shape check failures for valid kernel sizes (#5221)
     new 5a3fe7c  Last post we did for text classification with CNN (#5227)
     new 2605f86  Fixes bug in GPU broadcast_sub (#5216)
     new 34eb88c  [doc] refactor python ndarray api (#5151)
     new 39813a5  Doc (#5230)
     new c58eb46  fix for issue #207 (#208)
     new 95d9b59  Update nnvm submodule (#5232)
     new 187e605  [Scala] fix scalatest failed problem (#5248)
     new 8f27d7f  fix and scripts in tools (#5257)
     new 6567468  Fine-tune in R: end-to-end example (#5253)
     new 31101bb  [doc] fix frontpage style (#5252)
     new d0491c6  Fix broken amalgamation (#5236)
     new a23608f  Fixed previous weight handling for DCASGD optimizer. (#5140)
     new 3f410c2  gcc-4.8.5 error reported by pfc and fixed by sxj (#5269)
     new d9d2da9  fix str full-width characters (#5268)
     new 104bb1e  fix typo (#5267)
     new 55bb4cd  [doc] refactored api/python/ (#5265)
     new 8d0a3d1  Fixed docs (#5288)
     new cf35676  Visual Studio nvcc crash workaround (#5308)
     new c356e25  Added GRU cell from examples to new RNN API (#5215)
     new fa25ad6  metric: intro NMSE
     new f23b0a9  doc: update CDN link of MathJax
     new 64f21f0  Cleanup compilations & tests (#5309)
     new 4ae1cf7  Added Pipeline transformer capabilities with MXNet (#5277)
     new b0fc714  symbol not serializable (#5319)
     new 96191ea  metric: add test cases for NMSE
     new f442b18  Improve Caffe Converter (#5247)
     new 9bc8a78  Add valgrind checks to cpu tests (#5234)
     new 7908831  float16 training for resnet (#5312)
     new d9d798b  [CI] add script to build and run docker (#5325)
     new 9afce67  Add __str__ method to EvalMetric to display the metric values (#5317)
     new 3ba709d  [Scala] module save & load checkpoint (#5152)
     new 1816c86  Remove static keyword from ilog2 (#5335)
     new a857141  Fix spell checker (#5333)
     new ba97d93  docs: add an explanation for normalized MSE
     new 09d8f27  Merge pull request #211 from iblis17/nmse
     new 31530dd  Replace concatenate with concat when merging from ctxs (#5305)
     new d953b40  mkl spring update 1 (#5339)
     new 9414d51  [Scala] add SequentialModule (#5336)
     new e669bdc  fix nested macro on VS (#5352)
     new cf7f32b  Fixed data attribute
     new abf4c76  BidirectionalCell and SequentialRNNCell Unroll (#5337)
     new 2abd2f6  Fix docstring (#5359)
     new 8b6c859  Fix plus icon position (#5358)
     new 8102d5b  fixed a typo. (#5363)
     new 50c6733  Add unfuse() to FusedRNNCell (#5362)
     new 8861ccc  Update
     new 96513b5  [CI] Jenkinsfile with CPU/GPU build and unittest (#5364)
     new b446aef  fix the link of operation engine documentation (#5367)
     new 93d0040  [CI] safe incremental make (#5377)
     new 6732d5e  [CI] Add amalgamation build (#5393)
     new 72a107c  Fix bug in instance norm op (#5388)
     new fb9bbd5  [Scala] add optimizer DCASGD (#5386)
     new dcb3048  module api and mxnet scala through ide doc added (#5379)
     new 67b3bff  add 'export' to benchmark variable settings (#5397)
     new ad92b84  added dockerfile for crosstool chain (#5403)
     new 3a48185  [CI] add timeout to each build (#5406)
     new 3dc5a2e  Improvements to PTB bucketed LSTM (#5400)
     new 6dd2ddf  List string metrics on the documentation of (#5410)
     new ec19e4d  Padding GPU allocations (#5276)
     new 2f201fa  fix caffe importer (#5415)
     new 99cd4ad  debugging mxnet c++ operator (#4926)
     new e638aba  include GTEST_PATH for compilation option (#5241)
     new 8b28b8c  Fix out of bounds error in iterator (#5402)
     new e4d5e60  fix typo (#5430)
     new 3c9e5fc  [docs]Update invalid notebook links. (#5434)
     new 3b7b630  Scalable build and warning fixes (#5396)
     new 1e1bbc4  refactor ImperativeInvoke (#5423)
     new b5f7f81  [CI] Add MKLML build and test (#5419)
     new ffb692e  Fix small api style issue (#5439)
     new 4dbb54a  [CI] Add brief doc in (#5443)
     new 2faac0e  Fix api style (#5445)
     new 430ea7b  [CI] Add integration tests (#5450)
     new 9203e38  Bulk execution (#5220)
     new 6b74dcd  Improve convolution (#5284)
     new 9db513f  [doc] make api document more compact  (#5462)
     new 1ccbd6e  [doc] add virtual padding before methods (#5464)
     new 3639172  [doc] css hotfix
     new 01e9d1c  Changing according to new checkpointing method changes for Module API (#5438)
     new 2735414  fix MKL dropout issue (#5463)
     new 39776a3  Fix typo (#5474)
     new 47b3b7a  Added eval(...) to Symbol class (#5300)
     new d2fc1e6  Refactoring about None, spacing and indent. (#5477)
     new 2e4be17  [doc] new api document (#5478)
     new c9eb71b  switch to Plots.jl from pyplot (#217)
     new 2e3d28b  Fix pooling-inl.h: 3D pooling InferShape typo: pool_type == pool_enum::kValid (#5483)
     new 7478f3f  Use print function for python3 compatibility (#5481)
     new c93bf31  softmax and log_softmax (#5497)
     new dcdb35a  [doc] update io api (#5492)
     new ca53cb4  [tools/cfn] redirect cfn to (#5461)
     new 5e9e3d0  fix np.concatenate error (#5503)
     new f2e0f6a  Add Prefix NameManager use case in python tutorial (#5510)
     new 289ff56  add contrib package (#5499)
     new 4870ce9  Add pad param to DataBatch returned by BucketSentenceIter (#5518)
     new 0d8a63c  	修改:         example/reinforcement-learning/a3c/ (#5473)
     new dd51ca3  Synchronize parameters to CPU before creating kvstore (#5522)
     new cefc144  add forget bias option to LSTM cell (#5395)
     new 21b2da0  Integrate cpp package (#5251)
     new 181cddd  Improve pooling (#5519)
     new 2ad9a29  refactor normalize sequence (#5523)
     new 16df941  [scala-package] fix the "too-wide" accessibilities of scala classes (mxnet-core) (#5485)
     new 9f1f60e  Fix markdown syntax error (#5531)
     new b674c2c  Add __str__ method to DataBatch (#5512)
     new 8b91053  A better thread-safe profiler constructor and dump profile method (#5490)
     new 51c9a0c  Heavily revised & docstrings (#5507)
     new 2649276   MKLML build in scalable build, warnings, rnn and ssd operators (#5533)
     new 968adea  Fix doctests for symbols example and ndarray cpu version example (#5527)
     new ebff838  Update
     new ff8700c  Update
     new eb21472  Update
     new d7f8cd3  Update
     new 8f8d834  Add snapcraft packaging (#4852)
     new e4f73f1  fix model to fit the new contrib structure, fix links (#5543)
     new 662ae23  Fix docstring test for and examples (#5549)
     new 0267254  fix initializer (#5547)
     new f2d33e2  Pick (#5553)
     new b276a9d  Skip the line in lst, if the format is incorrect (#5561)
     new 1550f17  Add tensorboard support in Speedometer. (#5345)
     new caeabaa  Fix slicechannel type check (#5558)
     new cfe2848  Update .travis.yml
     new 0339433  added crosstool changes (#5564)
     new c057bea  update fc with nnpck (#5570)
     new 3d16fe6  add dockerfiles and hosted docker images for python/julia/r/scala (#5574)
     new 7a142b0  fix some installation script problem; add a dependency lib in ubuntu r installation (#5562)
     new 62057e6  Made test pass deterministically in the presence of cuDNN algo find. (#5584)
     new 9db10f8  Fixing _normalize_sequence call in FusedRNNCell unroll to use valid length input instead of None (#5586)
     new b36ce1b  Adding MXnet Keras integration tests for functionality and performance benchmarking (#5579)
     new c0f5105  [doc] updated get started page (#5526)
     new 6e6e151  [CI]  windows build and test (#5583)
     new 0012731  fix pad constant grad omp error (#5593)
     new f3366be  Making Perl API visible on (#5595)
     new 0cb7f63  Added perl docs to the navbar and disabled slow linux travis tests. (#5604)
     new 4abd4b7  Bugfix: remove get engine in initialization (#5598)
     new f0a450d  [doc] add track numbers (#5608)
     new 84f1f2e  add infer type to custom op (#5589)
     new 8ca8dee  zoneout (#5567)
     new 00b2b47  [cpp-package] Fix multiple definition issue (#5545)
     new 9a14c0a  Inference in ndarray Reshape. (#5607)
     new 05a101e  Fix pslite (#5600)
     new 395152c  Update
     new 49f7984  OpWrapperGenerator not Py3 compatible (#5544)
     new d814bb4  ADD python before for correct multi-devices bash command (#5616)
     new ce5e7e7  [Scala] add inferType to CustomOp (#5618)
     new 1781290  Update ACE metric (#220)
     new 9474d95  provide an optional context in mx.load_checkpoint (#221)
     new c3bb066  Fix the behaviour of ndarray.T (#5625)
     new 0df3bf6  Extensive revision of Python API docs  (#5611)
     new 7b34a09  contrib fft,ifft,countsketch (#5563)
     new 7f688f7  Revert "contrib fft,ifft,countsketch (#5563)" (#5634)
     new 46440e3  Added initializers for Spatial Transformer Networks (STN) layers (#223)
     new a5abd79  Caffe without the patch, cpp-package fixed also with Caffe plugin (#5573)
     new d7f233c  add env path repare for win pip (#5641)
     new 53765f1  [doc] update (#5639)
     new a5fd6f8  Debug command can't kill remote processes (#5646)
     new 71b5cc0  [doc] updated get_started index page (#5637)
     new c3fc511  [doc] update local build instructions for docs
     new 62d8a5b  fix NDArrayGetData to return to use the right datatype (#5629)
     new ea71c35  win build docs (#5654)
     new 447fef0  Change mkl fully connected operator to stateless. (#5587)
     new 1f850e1  Adds mx.nd.moveaxis (#5640)
     new 9dd3291  docs: update renamed `every_n_batch`
     new 1383446  modify mx.nd.reshape documentation (#5609)
     new fc2974f  Merge pull request #225 from iblis17/docs
     new 5c02fcb  Check if key stride exists in the node info when producing graphviz output. Fixes #213
     new b35148e  Example ssd, major update  (#5643)
     new 867b1c5  Merge pull request #226 from facundoq/master
     new d2f40d8  Implement 3D deconvolution (cuDNN) (#5615)
     new e39df3e  simple changes to mxnet.image.ImageIter to make it more newbie friendly (#5278)
     new 37d32e4  Temporarily Revert to fix formatting issues "modify mx.nd.reshape documentation (#5609)" (#5673)
     new fa37125  compile with macOS support (#4260)
     new b57237f  New CUDA kernels for binary broadcast, for: add, sub, mul, div, eq -operations (#5651)
     new 8f2fe4c  add link to 512 model (#5675)
     new 316dd3e  Update the (#5670)
     new cb28222  Revert "Revert "contrib fft,ifft,countsketch"" (#5642)
     new 1181178  [WIP] Rewriting image IO pipeline to improve performance (#4947)
     new 5a14458  Adds GetDType() signature to ndarray.h and fixes typo in implementation. (#5680)
     new 911f747  Fix data type issues of one_hot, take, and pick and minor fix for conv/pool (#5619)
     new 5c69214  Update
     new 9a65a2e  fix blockgrad (#5678)
     new 8edd97c  fix compilation error for USE_OPENCV=0 (#5689)
     new 6614a58  [doc] update main-nav mxnet.css
     new ad5f01d  Typos in the comment examples (#5700)
     new e3bb3d4  fix argument parsing (#5693)
     new 006c3a5  [Scala] fix bug of DataParallelExecutorGroup (#5698)
     new 974d7da  Update (#5701)
     new 6e81d76  Move autograd into master branch (#5652)
     new 01b808b  reverse previous change on script/edit markdown (#5711)
     new d99a7bf  fix formatting issues with Reshape documentation (#5686)
     new 401443b  Update
     new 0f6b6b3d [doc] build .so with default for doc (#5688)
     new ebcdecc  mxnet spring update 2 (#5721)
     new bbcc102  fix build error for USE_DIST_KVSTORE (#5727)
     new 3480297  fix mkl experimental and faster rcnn (#5732)
     new a18bc2c  Update (#5730)
     new d4e8a46  enable layout for rnn_state + enable additional attribute via kwargs … (#5729)
     new 39c643f  added mxnet.js tests (#5719)
     new af23470  fix MSHADOW_USE_CUDNN check (#5708)
     new 990f663  Fix (#5738)
     new f234226  ZoneoutCell, nd inferred reshape and moveaxis, cosmetic changes to Image iter. (#5740)
     new 723966c  add pretrained resnext-101-64x4d model. (#5741)
     new 9158ade  improve parallel_actor_critic: stop an episode when all envs finish(instead of stopping when any env finishes) (#5743)
     new 22482ae  Add error for NDArray bool casting (#5745)
     new 31b95fb  Update (#227)
     new 6ed3135  Change argument name from src to data for ops that don't have FListInputNames properly set (#5757)
     new 0f7a292  Use Pre-installed EC2 GPU Instance (#5709)
     new 3d8e823  Resnext101 64x4d (#5770)
     new c4d5a74  Remove MKL's fully connected layer for the convergence (#5768)
     new b3b21e4  Python API revisions (#5687)
     new c186705  Fix path issue in jenkins nightly build. Add execute permissions (#5765)
     new 7896e43  Copy code button (#5766)
     new f89f965  Fix executor monitor bug (#5752)
     new a0726a1  [Scala] make func start with '_contrib_' public (#5750)
     new caa3253  Debug error caused by missplaced import find_mxnet in googlenet (#5784)
     new cc40619  Fix initialization problem (#5771)
     new 7ac6fc0  Quick fix installation errors (#5785)
     new ced646a  refactor autograd (#5764)
     new 086d578  Update
     new 6c05943  Changed from numpy to ndarray for Perplexity (#5786)
     new 6179974  Perf md update for AWS instances (#5788)
     new 259fb15  fix ndarray split (#5791)
     new 62ecb60  Add 'argnum' for autograd (#5787)
     new 1be1b87  fix conflict with Images
     new 7a6120a  Merge pull request #228 from dmlc/vc/images_subtype
     new c296ea6  Ndarray api docs (#5703)
     new 4ede00e  Wait for ndarray to readble in dist kvstore push (#5794)
     new 8edb701  Griddim (#5797)
     new d20aa92  Minor wording edits to the documentation (#5795)
     new c1641cd  fix warning for test : UserWarning: Data provided by label_shapes don… (#5792)
     new d371d4b  New docs (#5803)
     new f6a7499  Add R mirror (#5777)
     new bfa966a  fixes bilinear initializer following approach in #34
     new 5f765d4  [doc] Copy-edits on landing pages (#5810)
     new 8947ead  adds test for bilinear initializer
     new 44a6c36  Merge pull request #229 from dmlc/vc/bilinear_initializer
     new 558adb7  add SeqMetric to apply different metrics for multi-output
     new b4ece81  add NullMetric
     new ced3f73  fix pointer(NDArray) for Float64
     new 1200759  Update footer.html (#5817)
     new 9efb2d9  Add brew/osx search path for openblas (#5821)
     new 104162b  added new distributions to sampling operators. added operators for sa… (#5726)
     new 86e3fa0  Merge pull request #231 from dmlc/vc/pointer
     new 3263b95  Merge pull request #230 from dmlc/vc/seqmetric
     new d0253f1  XCode8 supports thread_local in C++
     new 68b67e8  Merge pull request #234 from dmlc/vc/mac_travis
     new 1697522  update MSE to be agnostic to dimensions.
     new e12052c  convert single outputs to array so that inference has it easier
     new def129f  convert eps to eltype of array
     new b0e2df8  index clipping for embedding + testing (#5832)
     new c2044d5  adds THTT to enable metrics to support NDArray or Array
     new 349d2a7  convert MSE to use NDArray as an example for async
     new 7033cae  Head color and list padding (#5831)
     new 17bfe18  Fix docs for some NDArray functions (#5808)
     new 4675ace  Merge pull request #233 from dmlc/vc/metrics
     new c4a2658  ensure that new submodules are initialized
     new 42102b5  Add AWS Seoul Summit demos to ipython notebook examples (#5846)
     new 9aac7e1  Update callback.jl (#235)
     new 4948c30  [R] fix message error on Jupyter notebook (#5854)
     new 59d0670  [R] fix image classification examples ( close #5080 ) (#5855)
     new 1722b08  [R] update docs (close #5683) (#5856)
     new 781544a  [CI] use sudo to make clean
     new 437baa4  fix typo in (#5860)
     new 689188d  Fix type warning when dev=1 (#5859)
     new 9958690  [Scala] refactor EvalMetric, add Perplexity and CompositeEvalMetric (#5823)
     new beba421  refactor executor dedup logics (#5857)
     new 6b715d2  BatchNorm: Document CUDNN_BN_MIN_EPSILON (#5824)
     new d4e611b  Debug data shape related error of toy_ctc example (#5864)
     new 093a920  Small fix (#5871)
     new 53d3c63  Add cpp_package build option in CI (#5844)
     new 78e75d9  rm a duplicate line in doc (#5877)
     new ff5897d  Dockerfile fix for scala (#5869)
     new 3939512  Update doc for some NDArray functions (array, arange and full) (#5833)
     new 362c641  Fix documentation (#5882)
     new 47ee5e8  Update doc for log_softmax function. (#5887)
     new 4b03a6a doc strings clean up (#5801)
     new 6c4268e  Compiles against cuDNNv6 with new support for dilated convolution. (#5606)
     new 405c7c7  Fix spelling errors in recordio module (#5893)
     new 7a82076  Fix spelling error in recordio module description (#5892)
     new a022a0b  update doc for math, concat, slice/crop functions (#5716)
     new bffdaf8  refactor ndarray function parsing & add feed_dict to bind for autograd (#5883)
     new 3e4be71  improve infer_type error msg (#5841)
     new d8febd1  resnet v1 in fp16 (#5897)
     new 161d9d2  Fix NDArray operator docs (#5884)
     new 76241fc  implement ftrl (#5830)
     new a05980e  SGD optimization & added test (#5903)
     new ff64892  fix UnicodeDecodeError (#5908)
     new 4fa0c8e  optimized perplexity metric. (#5906)
     new bb6d4e0  [Doc] Improve search (#5919)
     new 51611b0  [Doc] Fix search result url (#5921)
     new f613e29  fix UnicodeDecodeError (#5920)
     new ac85f64  Update (#5922)
     new 49d57b1  Compute perplexity by averaging over the document (#5905)
     new bb7ad9d  Fixed argparse usage for --exts to accept one or more arguments (a list) instead of specifying the type=list (which stores a list of lists and does not work). (#5928)
     new f780b08  Improve Search UI (#5933)
     new 7bbcbef  Avoids running unneeded backprop kernels. (#5902)
     new 955f6be  Update
     new 4855bd2  Use dict.items() instead of dict.iteritems() for Python 3 compatibility. (#5956)
     new ff7589c  ADD speech_recognition example (#5954)
     new 10e38a4  Python kvstore doc fixes (#5929)
     new 2297695  Add test for pip installations (#5916)
     new 37f4c8c  option to not reset eval metric with Speedometer (#5827)
     new 48033dc  Update documentation for mxnet.initializer.Mixed (#5937)
     new 9bd0f89  Update documentation for SVMOutput. (#5931)
     new 1134d92  Symbol docs fixes (#5930)
     new e433862  Adding install instruction for Ubuntu-CPU-Python (#5885)
     new c293318  edit ndarray API docs (#5806)
     new 2ea2f23  More API Doc Edits (#5886)
     new 6d26d2d  Update documentation for mxnet.image.imdecode (#5957)
     new b772469  Fix script by adding path to Dockerfile (#5958)
     new e6fc294  Improve the doc of pick + Update dmlc-core (#5946)
     new 7008200  Update doc for Custom operator. (#5875)
     new b13223d  ADD missing Libri_sample.json, FIX minor bugs in speech_recognition example (#5962)
     new 517069c  [cpp-package] Add C++ basic tutorial and build instruction (#5971)
     new 58e3346  Image docs modified (#5973)
     new 02671eb  [KVStore] Add support for other data types (#5818)
     new 56e712e  Convert caffe AbsVal to mx.symbol.abs in caffe converter (#5984)
     new 1a9db4e  [Module] fix input_grads order (#5980)
     new 22aac01  Update documentation for (#5977)
     new 49a6307  Getting rid of identity (#5935)
     new 1711297  Activation ops (#5938)
     new 6b029fb  update env_var doc (#5964)
     new 3d4d845  Correction to LSTMCell docstring (#5986)
     new 5efd91a  [Quantization] ops: 'quantize', 'dequantize' (#5843)
     new 7a530fa  Fixed SpatialTransformerNetwork loc_bias initializer. Added lenet-stn example. (#238)
     new 80ef44f  Fix NDArray delay allocation inconsistency (#5990)
     new 232d837  Fix for scala test (#6001)
     new f8c6df2  Update doc for mxnet.ndarray.elemwise_add (#5898)
     new a7f429e  Add attrs for backward ops (#6003)
     new fdadaca  dmlc-core submodule updated (#6002)
     new b2080da  Adding installation instruction for Ubuntu GPU Python. (#5988)
     new 1974be4  Doc updates to ensure method descriptions are in 3rd person declarative (#5989)
     new 3775c9c  [Documentation] Ndarray instancenorm doc modified (#6008)
     new c44f7c4  fix float16 random (#6009)
     new 99716de  Fix indentation in LSTMBias docstring (#6014)
     new b5903c3  update docs for lrn (#5913)
     new 06f1c4d  Fix take op kNullOp issue (#6011)
     new 15b24ca  [doc] remove Chinese docs (#6017)
     new d5706de  Clarify order of data and labels in DataBatch (#6026)
     new 5d10cb3  [doc] remove tutorials/computer_vision (#6022)
     new 703e8ee  [doc] remove docs/packages and docs/system (#6029)
     new 96bc20b  update doc for pooling_convention (#6037)
     new 06010b8  explicitly specify 7z (#239)
     new 6f73cba  Update documentation for 'mxnet.initializer.register' (#5995)
     new 0bee1de  Update documentation for Normal, Uniform, One and Zero initializers. (#6016)
     new 1460b16  update doc for pad (#5934)
     new f0bfbfa  Change to CUDNN_CALL (#6048)
     new 3378bf5  fix utf8 (#6045)
     new fd43d68  Sync with python, Ftrl optimizer, new examples, (#6042)
     new 1171e7f  update viz for py to support 3d net (#6036)
     new 35aa93d  Update osx installation script to move ~/mxnet if present (#5974)
     new bbfd8c0  Cures SEGV seen in when DEBUG=1. (#6049)
     new 88deb1c  [Scala] update Visualization.scala (#6051)
     new 5d7f4f0  update doc for Leaky ReLU (#5872)
     new 35eebfd  Toc Restructure (#6061)
     new e57786c  revised English on these documents (#6060)
     new 98b4f53  update mshadow (#6062)
     new d8dbe42  Fix search function for new toc (#6065)
     new 70a0eeb  More How-Tos Revised (#6071)
     new a3045e2  CTC Loss Operator (#5834)
     new c1f19d2  Fix memory error in threaded engine (#6084)
     new 3b4e605  Fixing typos (#6089)
     new cb00132  Typo dose->does (#6081)
     new de272ae  Test get-started installation guide (#6070)
     new 452db0d  Context, CSVIter doc modified (#6028)
     new 287490c  Update documentation for mxnet.metric.create (#5991)
     new c33649e  Fix tutorial links (#6090)
     new e9f281a  Some more symbol docs modified (#5987)
     new 400498d  Doc sample change (#6094)
     new ecc3dfa  Update (#6101)
     new 11fe466  Update doc for mx.random.seed (#6122)
     new f94f449  fix shape order bug (#6136)
     new 9ed5492  Fixed imdecode crash bug when flag=0 (#6134)
     new 82b43bb  Fix (#6131)
     new 6606821  Fix NDArray bool checking (#6130)
     new 4069e14  Docs for MXRecordIO, MXIndexedRecordIO modified (#6013)
     new fa704d7  Update documentation for mx.callback.Speedometer. (#6058)
     new c15fc27  update doc for Load (#6092)
     new b0aa761  Installation instructions for MacOS and Cloud (#6012)
     new 28efd59  TOC click unfold (#6133)
     new ba8b4c5  [doc] new sphnix plugin  (#6105)
     new 3cca692  [doc] use debug mode to build (#6151)
     new 59bc589  move ctc loss to contrib (#6154)
     new a3b7a55  Fix for invalid numpy float indexing (#6144)
     new efd0a11  Fix python3 compatibilities (#6143)
     new f5e11c4  [example]ADD practical functions and options for speech_recognition example (#6141)
     new 1d917b0  [doc] small changes to tutorials (#6164)
     new ff27414  [doc] Fix left toc link (#6162)
     new db300ca  Added reflection padding (#6123)
     new dc1078d  [Scala] Change version to 0.9.5-SNAPSHOT (#6173)
     new 6cc9e99  add verbosity option to predict()
     new e8c95d8  Fixing LICENSE file and adding NOTICE (#6172)
     new ae611ed  doc improvement - softmax, metrics, and initializer (#5945)
     new b798eca  DataBatch and NDArrayIter doc modified (#6091)
     new 03f4279  Explicitly specify quiet in R install_version (#6171)
     new c08503a  API doc improvement Dropout and SoftmaxActivation (#6088)
     new 0a52456  Update documentation for mx.callback.do_checkpoint (#6059)
     new 5dfb461  Update documentation for plot_graph. (#6098)
     new ae22836  Restruct get started (#6167)
     new bdeb09c  Update documentation of Initializer.dumps() (#6128)
     new a883b79  Doc Improvement - RMSProp and RMSPropAlex (#6107)
     new cda17f5  Docforcs,fft,ifft (#6145)
     new fb40b8e  Improve executor API documentation (#6110)
     new 28de9e2  Adding back all other OS install guide. (#6193)
     new 940db17  Improve howto modeling (#6182)
     new 4823d81  Docbash (#6138)
     new e383fd8  Docs changes to BucketSentenceIter, Optimizer, ImageIter (#6112)
     new 5d65519  update cudnn version (#6197)
     new 77e5233  [DOCS] Add docs for several APIs (#6111)
     new 973c2bf  Update documentation for scale_down and resize_short in python/mxnet/ (#6074)
     new 4cd5f3f  Merge pull request #244 from rdeits/predict-verbosity
     new e0f0bf0  fix batchnorm has no effect on fully connected layer in speech_recognition example (#6203)
     new 3dfb231  update kvstore docs (#6109)
     new 5ef672c  updated docstring for set_lr_mult and set_wd_mult (#6120)
     new ce01e3c  Enabling other language options on install page (#6212)
     new 5b5aba9  Added documentation for random_crop (#6119)
     new 9e14daf  Half2 implementation (#6165)
     new 5d16305  explicitly specify 7z
     new 02c7661  Merge pull request #246 from musm/patch-1
     new 8918f1b  update appveyor script (#195)
     new 60f5306  improve GPU detection.
     new efff124  Add appveyor badge
     new 07777d6  switch to vc14
     new 2ba41bf  Fixed typos (#6209)
     new 5d07679  Correct typo in set_params doc (#6230)
     new 1ac8ca2 issue fix (#6229)
     new 28344a4  actually copy the mxnet.dll to the right place on windows
     new 96bdfa7  Prebuilt binaries only work on windows 64bit
     new b0053060 Pre nn patch (#6201)
     new aae9114  Merge pull request #194 from dmlc/vc/gpu_detection
     new 38f7c55  Fix RMSProp update rule (#6235)
     new 2ecf2c1  update
     new 410c839  fixing english (#6241)
     new 07e9cc5  Fix grammar and typos (#6246)
     new 5fc074e  [Perl] Autograd, bugfixes for the visualization of convolution.  (#6217)
     new 2d43676  Update dc gan (#6249)
     new d5c2463  Rearranged and clean-up of "" doc (#6183)
     new d1f61c5  Fix a spelling mistake. (#6256)
     new 228df95  fix prediction bugs in speech_recognition (#6253)
     new f40bd36  Improvements to predict image tutorial. (#6247)
     new 87927ee  MNIST tutorial improvements. (#6232)
     new cde5361  Correctly import Caffe BatchNorm (#6176)
     new 9786c93  update resnet of faster-rcnn configuration (#6004)
     new 8300fbe  Batch Norm rewrite without mshadow, 1D, 2D, 3D, float16, float32, float64 as well as operator gtest framework (#5936)
     new 1e3c3b5  fix bug in update_metric (#6263)
     new 74efa2b  Add .clang-format to gitignore (#6262)
     new c9e8c93  Fixed Broken links at various Documentation files (#6270)
     new c73aa88  added ability to set omp threads at runtime (#6125)
     new 565609c  getLogger api documentation changes (#6129)
     new 36420de  Added ResidualCell (ModifierCell) for vertical connections in stacked RNNs (#6267)
     new 0cdc1d1  Adding mxnet slack and apache email list details (#6273)
     new 3d545d7  Lint for Caffe converter (#6264)
     new e3f5f96  Updated (#6276)
     new 88bdca1  mshadow updated to include half2 inline change (#6274)
     new 9c70922  Update CI URLs to point to new caching infrastructure
     new 6df38a9  Fix a broken link (#6285)
     new bc6e4b4  Fix typos in cnn text classification example (#6286)
     new 75bc7c0  Cleaning up NNPACK installation and broken links (#6295)
     new 46d2433  Modified device build guides (#6292)
     new aeae370  Misc docs improvements (#6289)
     new 2d6299c  [doc] enable doc build in ci (#6282)
     new 3634549  Fixed: MXNET R Reference manual broken link (#6283)
     new 3c6759b  Fixing review comments for mxnet social channels (#6279)
     new 3a61a64  Add install btn (#6275)
     new 4544c58  [cpp-package] Add monitor, optimizers, metrics and bug fixes (#6033)
     new dfc9e7a  Added comments for getting data and set default data/model path (#6251)
     new cbf3cc7  add mshadow lint (#5970)
     new 3f75c41  Fix typo (#6306)
     new 9feaeae  Using the right image url in tutorial (#6310)
     new fb4e1e7  Updated (#6315)
     new d6b3157  Update (#6318)
     new 46782c3  Fix typos and correct syntax for several examples. (#6293)
     new fd91bff  fixed some APIs documentation used in Tutorial (#6227)
     new bdb4ba9  Fix the errors in fcn-xs when using "cut_off_size" mode (#6321)
     new 88b4c60  remove hack (#6316)
     new 0891312  mnist tutorial: fix image rendering in docs + fix toc rendering (#6311)
     new 4dfcdc9  Docs fixes (#6308)
     new 50b2e39  Fix several typos (#6322)
     new f659ef5  Added documentation for imdecode (#6118)
     new f884002  Adding docstring for IRHeader (#6114)
     new 479538e  Cleaning up tuning mxnet for performance (#6320)
     new edc9c31  Tutorial for large scale image classification (#6280)
     new afd665a  Fixing broken links (#6331)
     new 2e89352  Minor edits to architecture notes. (#6328)
     new bd7a05e  Improve API Style (#6330)
     new 29cb243  LR regression tutorial fixes (#6337)
     new 99e9144  Delete common/thread_local.h (#6335)
     new 557cfb6  Changes in tutorials/how-to index (#6334)
     new ed13b70  Improvements to module tutorial. (#6236)
     new 6241ef0  Add ci test for c++ package and rename enum values (#6341)
     new 94a4294  Module api page cleanup (#6258)
     new f27ef04  Add example for (#6261)
     new 033bf92  DataIter documentation. (#6291)
     new 978f53b  Improved documentation of test util functions (#6314)
     new a16f809  Added brief intro to Python API (#6350)
     new efea8e9  Enabling CPP doxygen API docs build (#6346)
     new 0f9aebc  Update IO Tutorial (#6221)
     new 4a33a00  LinearRegressionOutput doc update (#6327)
     new a4c7265  Update document for zeros_like, ones_like, softmax_cross_entropy, c_array, c_str (#6021)
     new 0784f62  BaseRNNCell doc changes to make some method docstrings more readable. (#6108)
     new 594f9f0  Add large scale image classification tutorial to tutorials index. Also fix some formatting issues in the tutorial. (#6348)
     new 8edc8ad  Fixed few typos in the NDArray tutorial (#6355)
     new 14867a8  Symbol tutorial improvements (#6196)
     new e6dae17  Documentation for image.center_crop api. (#6116)
     new 9997eef  Improve symbol api doc (#6225)
     new 1804bdd  Add documentation for S3 integration (#6352)
     new 7c39f84  Fix ci_test (#6353)
     new 42b4561  Code title background color (#6359)
     new 2892fd0  ci test debug (#6363)
     new 57027aa  Fix ci test (#6364)
     new e0578ed  Update
     new 3abdbf8  update doc for DataDesc (#6358)
     new 653f547  Fixed few typos in the Module tutorial (#6370)
     new 35ef490  Added linear regression to tutorials list (#6357)
     new 548800d  update tutorials index, minor formatting fix (#6372)
     new 9dac416  Cleanup Python API pages redundant content. (#6215)
     new 82b5de4  Fixing a dead link in the python Readme. (#6373)
     new 13fc92f  Documentation update for ImageRecordIter (#6354)
     new bc3ee60  Added OSX classpath support for running charrnn example (#6374)
     new 57d4eaa  Add LocationNet and Multimedia Commons Dataset (#6378)
     new 7ac046c  [DOC] Fix smooth_l1 example display (#6377)
     new d35fc76  add num_group argument in convert_symbol (#6380)
     new 691cf98  Fixed few typos in the Symbol tutorial (#6369)
     new d1545f1  fix a typo (#6388)
     new de5acf0  Fix install rendering issue (#6391)
     new e1b7556  Documentation update for tutorials/index.html. (#6394)
     new 2b638ff  Moving Raspberry Pi and TX2 setup to common install page (#6393)
     new cc140ad  add get_mnist in (#6389)
     new 7460401  fix cython build (#6356)
     new fc0e8a5  Fixed few typos and wordings in the Iterators tutorial (#6400)
     new b63fd9b  Fix #242 (#247)
     new 5116381  Update
     new 29712c3  RNN API doc index page (#6255)
     new 2782d5e  Bugfix in PandasLogger.  Fixing broken link in Docs (#6403)
     new ee918e0  module api formatting fixes (#6406)
     new b5e386d  Force render all pages with no index in (#6407)
     new 7f77bab  Update
     new 32be9dc  Update
     new 5de9bdf  Fixing typos (#6414)
     new 8c34bb9  Doc Improvement - SoftmaxOutput - Add ignore_label use example.  (#6199)
     new 0947af7  Removed deprecated configs. (#6417)
     new 301f712  Replace ReduceToAssign with Reduce (#6415)
     new 2cd42a7  Improve progress bar doc and log the bar instead of stdout (#6409)
     new 35d5e54  Removing unnecessary copies when copying data to device (#6390)
     new 69d0adb  Tools of caffe_convert : support 1X7 convolution kernel and pad_w,pad_h params (#6139)
     new 54cada5  Add NNVM .cc .h deps to Makfile (#6416)
     new afca489  Optimizer formatting fixes (#6427)
     new 454772c  Fix download link (#6431)
     new 82efa6a  Add release note (#6434)
     new 624b2d1  Fixing tutorials. (#6436)
     new a4ce7f2  Formatting fixes (#6433)
     new 132544d  doc bash 2-5, for pack, unpack, pack_img and unpack_img (#6140)
     new 7708d80  fixing the early stop for maximize = T  (#5915)
     new c86f27a  Improve style (#6445)
     new 4ce3f9f  Correction (#6444)
     new 754480a  Update documentation for MXNetDataIter in (#6000) (#6113)
     new 6b9afd8  fix member variable name: make them end with underline (#6438)
     new 4506323  Fix minor issues with api pages. (#6410)
     new 5f625f2  Update documentation for mxnet.ndarray.GridGenerator. (#6430)
     new 780009c  Update documentation for deconvolution operation. (#6184)
     new f46a0bc  skip lines that have %matplotlib (#6451)
     new 67508af  Fixing some more broken links before v0.10 release (#6449)
     new cbf8a82  close #4838 (#6452)
     new f2241e5  Fix linear regression (#6432)
     new 996d90f  Pre-trained model tutorial fixes. (#6453)
     new 573253e  Nightly test tutorial (#6447)
     new c7032f0  Merge pull request #249 from staticfloat/ci_url
     new c051db2  [R] captcha example (#6443)
     new 3f19697  skip lines that have %matplotlib (#6459)
     new 775073b  Fix cudnn_deconv not guarding no_bias (#6456)
     new 57693ba  fix batchNorm cpp example (#6454)
     new ed7fdd5  Fixing up issues in install guide (#6463)
     new 809c6de  Fixing copy code functionality for bash command (#6465)
     new e39cf47  Residual unroll (#6397)
     new 9a66cf1  Linear regression Tutorial link (#6468)
     new 65d3dcb  bump up version number for release (#6462)
     new 8713d25  [R][DOC] update R installation guide (#6457)
     new 1f63eb9  Use sphinx==1.3.5 in Dockerfile.doc (#6470)
     new 4f27d23  Add 0.10 release info to and (#6471)
     new b35dc56  Update (#6473)
     new 05e0728  Change Interface of  NDArray & TBlob for DLPack Compatible (#6345)
     new 172af60  Use `Sys.CPU_CORES` not `nprocs()` when doing a parallel `make` (#250)
     new 4fb4a20  update faster rcnn example with logging and cpu make (#6486)
     new 1a10c2d  Remove horizontal scroll bar (#6491)
     new 215ae4a  [R][DOC] fix R tutorials (#6472)
     new 32ced38  backward headgrads and detach (#6332)
     new 2a09f16  Revert "[R][DOC] fix R tutorials (#6472)" (#6499)
     new 0f2c2cd  Fixed few typos in the Python tutorials (#6509)
     new 36520dc  Remove BUILD_TAG in DOCKER_IMG_NAME in (#6513)
     new d73f2dc  Fix for build (#6511)
     new bc83786  Test for macOS (#6231)
     new fa61571  Revert "Test for macOS (#6231)" (#6520)
     new 98cb4e1  Fix Search Description (#6521)
     new 26b1cb9  Remove duplicate new operator how-to (#6519)
     new 2b26bc2  update for Julia v0.6 (#248)
     new 6a00ba4  hide _move_var _move_mean _beta _gamma for visualization (#6523)
     new 1f7a7cd  add link to style guide (#6529)
     new 2cbab7b  Module tutorial improvements and metric API doc linked (#6532)
     new fd91cde  add requests, graphviz used test_utils and visualization to python setup (#6533)
     new 4c8aa79  [R] compatibility with R 3.2.0. close #6525 (#6534)
     new 841e9ee  update regression example for julia v0.6(#254)
     new 19f842c  Fix Nightly Tutorial Test (#6544)
     new e41ab86  [R] more tweaking for R 3.2.0 (#6542)
     new 551996a  fix typo on BatchNorm for mirroring (#6541)
     new 3959d1e  dlpack directory added (#6550)
     new 728d632  [R] CGAN R demo scripts. close #6040 (#6551)
     new 13c1463  Small fix (#6552)
     new 673a6db  [Scala] hide _move_var _move_mean _beta _gamma for visualization (#6555)
     new 10af1c7  Improve symbol bindings (#5870)
     new f4e4731  add random multinomial sample (#6553)
     new 5426154  [R] save/load MXNet model with RData format. close #362 (#6494)
     new deff013  [Scala] Optimizer support lr_mult and wd_mult (#6557)
     new da08c92  Some Changes to NDArray Interface (#6561)
     new b2e7c3a  Update
     new 788c280  batchnorm specify channel axis and performance optimizations for batchnorm (#6411)
     new ba2d9f6  fix `out` option for mx.nd.zeros and mx.nd.ones" (#6589)
     new 5978186  Fixed caffe_converter and improved test_converter (#6489)
     new 2428e00  Add instructions to install OpenCV. (#6582)
     new fa709b2  Fixing obsolete installation guide (#6585)
     new 7d572e6  fix rnn doc (#6595)
     new 00151e9  tools/ add 'from functools import reduce' for python3 compatibility (note: needs python 2.6) (#6593)
     new 0985c2e  Clearer error message for simple_bind failure (#6597)
     new 5e3f3e6  Changed make to support more gpu archs, multiple toolkits, reduce lib size. (#6588)
     new 8147e7e  Update to version 1.7.0 of CUB (#6607)
     new e885c1b  Create .gitmodules
     new c078b7a  data_shapes are extracted from dict, so its sequence is not determined. (#6610)
     new 49b1513  Remove references to deprecated API in how to docs (#6616)
     new 4feb759  fix pylint (#6618)
     new c965235  Add extra check on CUDA library detection
     new 05a04c8  update cub url (#6625)
     new 64b9c88  remove Python file from tutorials (#6623)
     new 25e719f  Fix couple of print statements that weren't working on Python 3. (#6604)
     new 0da55a7  [CI] remove COPY * in dockerfiles (#6515)
     new 8b3a56d  add doc build into ci (#6636)
     new ffc90ee  Website improvement (#6609)
     new 18547b8  Merge pull request #255 from joshbode/cudalib_check
     new 5375a37  Update Tutorial prerequisites (#6639)
     new 514d721  fix python debug cpp markdown format (#6641)
     new d75ef8e  fix doc build on ci (#6638)
     new cba4245  Added ResNet v1 fp32 and double buffering of input data to both resnet (#6661)
     new f7d7438  [Scala] add eq, ne, gt, ge, lt, le to NDArray and Symbol (#6665)
     new c09520d  Add backward infer shape in expand_dims (#6654)
     new e39607f  Spellcheck for pre-trained model tutorial (#6669)
     new 6eb8bd1  Truncate operator implementation (#6640)
     new 3db2533  Change method and attribute (#6676)
     new 17332e5  Ssd hotfix (#6675)
     new c43c901  updated version used of mshadow submodule (#6681)
     new e852036  Add BLAS3 and LAPACK routines (#6538)
     new 28f5843  round batch should start with first element. (#6684)
     new 5a9c3c0  Minor grammatical errors in 3 of the tutorials. (#6671)
     new b48ceae  Fix for broken link (#6690)
     new dc23888  sync with python. (#6655)
     new 5aeffd3  Use pinned memory in IO iterator to avoid unnecessary memory copies (#6660)
     new 17ea9e3  fix cpp-package charRNN shape (#6626)
     new 7fcaf15  Update (#6571)
     new a98e502  [R] fix accessing variables in environment. close #4282 (#6688)
     new fc32a5a  Fix a typo (#6709)
     new 5434612  [R] fix the predict function and optimizer (#6711)
     new eeec2c8  Fix data layout handling in BucketSentenceIter (Fixes #5509) (#6707)
     new 9073bc8  macOS install script: Support $MXNET_HOME; Print MXNet version; Added few package dependencies (#6670)
     new 790328f  train text cnn using mx.module (#6717)
     new 02634c9  Add python-opencv installation to Docker image (#6699)
     new aacfef4  Fix docs typo from gpu0 -> cpu (#6727)
     new ce2bca6  Add operators for Deformable ConvNets/DFF (#6298)
     new 56bd806  add verbose option to initializer (#6712)
     new 2debc91  fix bug in optimizer because state is not synced with weight context (#6731)
     new 423490c  enable use of lapack by default (#6704)
     new 46e5dfb  example/svm_mnist: use module instead of model (#6738)
     new c281bd0  Remove executable bit from header and cu files in src/operator (#6748)
     new 5126bd5  Fix typos in installation guide (#6746)
     new 36b6fe5  Fix pad example error (#6745)
     new 9d0508e  Update (#6744)
     new 82e8bda  Update (#6742)
     new 4d79eda  Enable support for float64 dot operator. (#6756)
     new 63392cf  Update (#6754)
     new 63ac793  mod (#6698)
     new e7d1b62  Added cblas_[ds]tr[sm]m to cblas include file (#259)
     new 94d8e18  set __layout__  attribute for FusedRNNCell output states (#6747)
     new 3384d20  Reformulate some comments in executor_manager and executor_group (#6740)
     new 3c38aed  Env variables doc modified (#6755)
     new 0df68e8  MKL 1D ReLU fix (#6700)
     new 3ceb6d2  refactor cachedop to specialize (#6735)
     new e4c7bfc  Multi precision SGD (#6428)
     new 0f7bdfa  Fix MacOS install instructions for venv and python. closes #6776 (#6777)
     new bf4d774  Support LSTM callbacks (#6659)
     new 0ba8809  Added script to test caffe converted models layer by layer (#6680)
     new 3fb169f  Update (#6786)
     new 073f803  [R] align parameters with original paper (#6781)
     new 61e2197  add -Werror to NVCCFLAGS (#6774)
     new 77b20c9  Deadlock and crashes during shutdown (#6773)
     new 1621b76  support str key type in kvstore (#6765)
     new d6d6191  allow extra params in module set params (#6753)
     new 10ad90b  Added MXNET_COMMIT env variable for upstream commit/tag (#260)
     new eb09f44  [R] switch order of LRN and pooling layer (#6795)
     new c160976  Removing unnecessary copies from backward pass of add and add_n (#6800)
     new 5f2f12d  fix typos in the docstring of resnets (#6808)
     new 906c358  [R] use ctx to be consistent. close #6818 (#6819)
     new 984aaa1  Change community to github for small screen (#6820)
     new cc62ade  Better ssd (#6827)
     new 9e66154  Add mxnet issue and roadmap link (#6825)
     new 5acc2c9  [Scala] support str key type in kvstore (#6829)
     new 459efce  fix numerical issue in softrelu (#6834)
     new ff96822  Module forward reshape resolving conflicts (#6805)
     new 17620fe  Fixed broken SpatialTransformerNetwork example (#5798)
     new 49cd2e6  Don't use __CUDACC_VER__ (#6844)
     new cb69e15  Documentation and tutorials are not 2 separated links. (#6845)
     new 4b1dd22  lstm_bucketing now shows more information if the file is not present.wq (#6862)
     new 8d3aa29  add Adamax & Nadam (#6784)
     new 78b1208  [R] instruction to build GPU-enabled pkg on Windows. close #3902 (#6841)
     new e906f88  [R] add R test into jenkins; add rpkgtest into makefile (#6733)
     new 52abdab  [R] R >= 3.3.0 (#6840)
     new b9e02cf  Updated installation instruction (#6861)
     new 49018a2  [R] doc update for regression and classification (#6826)
     new 09e08f5  [R] use AWS S3 to host all pre-built binary packages. close #6543 (#6722)
     new 8c81ee4  * extended caffe to mxnet converter and improved converter test (#6822)
     new aec0460  update mklml path in jenkins docker (#6876)
     new bfb79bd  [R] fix concat for symbol and NDArray. close #6650 (#6696)
     new a4fe4b4  fix warnings (#6881)
     new a94d0e2  fix pinned mem USE_CUDA=1 on host with no device (#6864)
     new 4a2bae2  create inception-v4 (#6847)
     new d79488a  Update deconvolution-inl.h. Change the condition control statement in InferPad. Merge differences between v0.8 and v0.9. (#6019)
     new 9f1df68  Fix 404 links on rnn page (#6900)
     new 04617df  Link to install XCode from the App Store has been added. (#6901)
     new d79e0e0  Add inplace identity (#6896)
     new bdcb45f  Change home page and api subtitle color (#6899)
     new b760eb1  add allow_extra parameter to module (#6903)
     new c8289d1  added docs for output of ctc loss layer (#6906)
     new 2e9943b  [R] fix the operations between MXSymbol and scalar. close #4994 (#6758)
     new 00b0f95  Fix a bug in deformable convolution operator (#6911)
     new 998378a  Fix Python 3 compatibilities (#6908)
     new 5fb0832  fix label bug which harms voc accuracy (#6849)
     new e9cc791  [R] To use the latest R in the installation script for ubuntu. (#6897)
     new 2c7f1d8  convert to long int for direct comparison (#6869)
     new a21cab7  [R] add "fixed.param". close #3906 (#6922)
     new de5b0fe  Add ConvRNN Cell, ConvLSTM Cell (#6832)
     new c159029  reworked cachedop. (#6910)
     new d7852e8  Fix smooth_l1 comment (#6929)
     new e09cc7c  Fix for WarpCTC conflict (#6905)
     new 8003ee1  Fix Mistake (#6940)
     new 45bcc58  Page title for installing from source has been udpated. (#6942)
     new 9af2734  Failing with a more descriptive error message when infer_shape does not get passed a tuple. (#6893)
     new 202de02  Fix Python 3 compatibilities (#6927)
     new 92428fb  Fix for pinned memory never being used (#6954)
     new 16bff59  [R] negate and more operators for ndarray. close #6768 (#6886)
     new 44b8c07  Fix broken link in example/svm_mnist/ (#6972)
     new f205ffd  ctc example (#6949)
     new 9aa91e0  Mention mxnet version in tutorial prereq (#6983)
     new f5cdd42  Fix linear regression tutorial (#6978)
     new c5b78ae  Change slave labels and add nightly Jenkinsfile (#6957)
     new 3a50dca  Fix im2col.h (#6986)
     new b1cb2cd  [R] custom iter in model training; MF demo in R (#6673)
     new 22f9a0d  add bucketing/batchnorm and improved performance for speech_recognition example (#6971)
     new 66b00b5  Fix tutorial test (#6988)
     new a8a2505  CMake based compilation for Mac OS X. (#6870)
     new 96902b7  Implemeneted reciprocal operator (#6719)
     new 5aa3034  [R] update the cats vs dogs example. close #6788 (#6935)
     new 65d0d47  Fix for codes not using pinned memory (#7001)
     new ed19095  Refactor Stateful operator and custom op (#6928)
     new b88e0d4  nn interface (#6501)
     new 3c18f84  resnetv1 (#6517)
     new e52b6d9  actor_critic (#6559)
     new c8bad62  reorg code to temporary namespace foo (#6568)
     new 1e27169  [WIP] Foo (NN) API reference doc (#6583)
     new 11796c3  error messages for ndarray-only methods (#6591)
     new 120a00c  split Layer.params into Layer.params and Layer.all_params() (#6598)
     new a34ec18  add language model example (#6621)
     new abebd3b  extend tranpose to 6dim (#6634)
     new 334df7d  super resolution (#6633)
     new e37727f  OnlyImplementedInNDArray (#6635)
     new 98792cc  add auto infershape (#6637)
     new 92ce9b7  Create
     new 9120d26  foo doc and tutorial (#6708)
     new c892ba8  Fix parameter initialization (#6728)
     new 944a725  [RFC]HybridLayer (#6772)
     new a887e11  nn fixes (#6792)
     new 9cdae4d  input output shapes added and formatting fixes (#6824)
     new 4b3dc7b  refactor rnn layers (#6823)
     new 4ed646b  add hybrid docs (#6859)
     new 82c3e76  nn mnist tutorial (#6879)
     new 615c377  rename load_data to split_and_load (#6934)
     new 9220680  rename nn.Layer to foo.Layer (#6959)
     new 3f74e8e  rename Foo to Gluon (#6980)
     new 652fc13  Add Icon to Homepage
     new 100a108  Add python-pip to the build from source install prerequisites (#6989)
     new 908b3c5  Remove setting of CUDNN_AUTOTUNE in example, since 1 is the default (and (#7010)
     new 128c655  Change Gluon API style (#7012)
     new eb93083  interpret negative values in ReduceAxes ops (#6992)
     new 953c375  Add python-pip for GPU installation steps too (#7017)
     new 853f334  [doc] Evaluate the python blocks on markdown files and append the results after (#7020)
     new b66ca44  [R] To ignore R-pkg when releasing on github (#7007)
     new e1d97cc  Use pinned memory in benchmark mode (#7028)
     new 3a7eb55  super-resolution fix (#7016)
     new 70ec3ce  add pearsonr as metric (#6966)
     new dad8381  Some fixes for gluon (#7013)
     new 2f53cb0  fix mkl (#7030)
     new 903ba3a  Dcgan fix (#7032)
     new 54e4614  fix (#7033)
     new 1529c5c  kldiv loss (#7014)
     new 7e24755  gluon docs (#7038)
     new 1bb9e45  fix ndarray slicing (#7035)
     new cac919b  similarity tree lstm model (childsum) (#7022)
     new 897cc55  revert (#7043)
     new 4a16fa7  Fix tutorial notebook names (#7050)
     new 3410e49  Remove empty commit file (#7049)
     new 845b62f  [R]fix optimizer with multi-GPU. close #5296 (#7056)
     new 4747c6b  revert module changes (#7055)
     new c70d549  [Scala] add arange to Symbol and NDArray (#6904)
     new 1ae1fbe  avoid runtime crash caused by error context of gradient storage when using context group (#7042)
     new 5af56bb  Improve PTB results (#7059)
     new 8144361  pretty-print structure/parameters of gluon layers (#7060)
     new a8804a1  Optimized sequence reverse operator (#6946)
     new 70a781b  Update documentation for correlation operation. (#7071)
     new e9aa011  Fix a spelling mistake (#7085)
     new add7e43  fix custom op and add tutorial (#7076)
     new bd5df7c  Add h5py support to NDArrayIter (#6790)
     new 4fd0891  Fix argmax_channel method and add preqs in scala mnist tutorial (#7101) (#7102)
     new bfbff25  Fix a spelling mistake (#7100)
     new e3fd434  add missing files for speech_recognition example (#7099)
     new c40442a  [R] Image segmentation example and test. close #5003 (#7096)
     new ef85876  added casts and overrides to fix compiler warnings (#7109)
     new fada5c1  Fix (#7036)
     new 59d5934  Updating dmlc-core (#7114)
     new 4210d85  Resolve a problem when import mxnet in python : ‘cannot import name gluon’ (#7119)
     new 7606864  [R] add "allow.extra.params". Update related documents. (#7125)
     new 46679f1  Adding absolute tolerance to few tests (#7133)
     new 4f1fad3  New algo for batchnorm for cuDNN 7 (#7131)
     new f45f091  Volta arch support in Makefile and cooperative groups (#7130)
     new a150a74  recursively include modules in (#7128)
     new 376279b  Remove gluon tutorials from index (#7140)
     new 799ed45  Update (#7143)
     new 37c1823  add save_params and load_params to Block (#7097)
     new 266cf3a  Streamlined fp16 examples (#7150)
     new f3a4b7d  Fix a typo (#7151)
     new 4b6f217  Khatri-Rao product (#6567)
     new a242e9f  Mx image2 (#6961)
     new f34cd23  remove where= from find_packages() (#7167)
     new 83828cb  bugfixes for linear algebra operators (#7175)
     new c0377a5  [scala-package] improve the readability of Spark-Mxnet implementation  (#7141)
     new 84ce29b  Gluon data pipeline (#7155)
     new 238b890  Fix a spelling mistake (#7187)
     new 4fef6fd  Fix 2 spelling mistakes (#7186)
     new d1d4ea5  Set default value for dtype for alexnet (#7193)
     new f3f8a99  fix ndarray setitem (#7195)
     new c79ec5d  Fix incorrect reference. (sym -> symbol) (#7199)
     new 500ea39  Compilation warnings (#7198)
     new 1ed8b19  [R] RNN update. close #6723 (#7211)
     new f8d2bf8  Jenkins tests fix (#7213)
     new 66f9b33  fix misleading in how_to (#7208)
     new 72d00fd  Fix broken amalgamation. Script was copying a declaration from a header file but missing out the definition, causing a runtime undef symbol error. (#7222)
     new b736999  Add note about about Apache migration (#7238)
     new 5c536f7  Add Keras Installation Link (#7241)
     new eb23742  Add post-build email notification & fix Windows CPU build (#7225)
     new 959ac39  Use different array comparison function for useful output (#7205)
     new 424143a  [R] allow users to use other names than "label". close #7126 (#7232)
     new b996dd1  [R] lstm bucketing example (#7237)
     new 14ba07e  Modify announcement (#7242)
     new cac9c25  Quiet regression stacktrace pr (#7156)
     new bcdde36  Update (#7248)
     new 66df7c8  Correct Python Docs about tensorboard path (#7250)
     new 13bcb5a  gluon models (#7023)
     new 37e40be  fix pretty print (#7254)
     new dcceb2f  Update (#7249)
     new 42544ed  add reset_ctx (#7221)
     new 1ed5f9f  Fixes for gluon RNN [WIP] (#7258)
     new 414a96c  Fix a spelling mistake (#7266)
     new 024525f  Fix gluon zero grad (#7263)
     new e830179  [Doc] Add the new doc link to to (#7271)
     new c521fa5  add getitem to container layers (#7265)
     new 92a93b8  [R][DOC] make sure all code in vignettes can run without error (#7274)
     new f965542  Fix a spelling mistake (#7277)
     new f187d5b  Now Jenkins correctly executes clean command when incremental build failed. Fix for #7272. (#7275)
     new 5393002  Gluon RNN fixes for seqlen 1 (#7260)
     new 43ea355  Update cub for CUDA 9 (#7270)
     new 1784f62  fix random sized crop (#7173)
     new 4c4aa56  fixed dcgan to use integer division so it won't die horribly in Pytho… (#7287)
     new 583fd9e  change random sized crop lower bound to 0.08 (#7278)
     new 4762785  fix dataloader length bug (#7283)
     new 861e929  Properly check argument types in to avoid segfaults. (#6894)
     new af56a77  add support for port package manager on macOS (#7276)
     new 1d1d6c2  Require Mouse v2.1.0 with explicit ‘v’ (#7171)
     new 1e2dfa8  Fix typoes (#7297)
     new 202373f  fix initialization warning by g++ (#7294)
     new 3f3d0fa  finetune now includes short hint how to run the script on windows (#7298)
     new 5aab4ad  Update (#7296)
     new 6333096  Fix mix of tabs and spaces (#7311)
     new 7578e35  Change variable/function names: defered -> deferred (#7301)
     new 5976973  [R] RNN bucketing with multiple devices. (#7315)
     new e0639eb  lstm crf example (#7253)
     new 8e0f627  add doc for gluon, sym/nd contrib (#7284)
     new e8ed548  Fix install error for jupyter (#7308)
     new cffbc2c  reduce model zoo test size (#7318)
     new e6a1139  add backward(is_train=False) and always mode for dropout (#7303)
     new 5a286b2  Fix module tutorial (#7324)
     new 1f0b813  assert size eq between shared_module.execs and context (#7233)
     new 4939dc2  Add document for BilinearSampler Op (#7203)
     new 2fe7aa4  Fix bug in symbolic RNN (#7282)
     new c7d18e0  fix py3 compatibilities (#7305)
     new 82a3d21  Added sparsity functionality, with tests (#7138)
     new 9add5ae  improve add support to SUM with coeff (#7120)
     new 8519eaf  add mobilenet (#7121)
     new 76dee53  [cpp-package] add lr scheduler (#6885)
     new c9440ba  update RCNN example for BaseModule::init_params (#6813)
     new d65d363  [caffe] support convert mtcnn and MobileNet model (#6956)
     new be7e791  Fixed visualization code error for bi-directional lstms (#6674)
     new b848c24  fix example/rnn: Speedometer(..., auto_reset=False) (#6679)
     new edbb4ff  Test installation of pip --pre (#7314)
     new dd4512f  [R] update docs from mx.symbol.MakeLoss. close #2922 (#7325)
     new feebe77  Add KEYS file
     new 51ec484  Mobilenet (#7330)
     new 14b83fc  [Scala] Make Module Api sync with Python interface (#7246)
     new c9aacaa  fix cpplint (#7332)
     new c8db271  Add Autograd doc to gluon page (#7327)
     new 79231e2  Fixes improper deconv workspace alloc (#7326)
     new b230d17  Attempting to add Perl interface to Apache CI. (#7170)
     new 43d1d2c  Build versioning website (#7340)
     new 15ffc82  Add DISCLAIMER and lxn2 GPG keys (#7344)
     new 0d8d27e  Fix gluon bottleneck v2 (#7339)
     new b2360f4  [R] fix mx.symbol.min. close #7219 (#7342)
     new be0579c  refactor gluon trainer (#7338)
     new b0c8c6c  [Perl] Fix for CI (#7343)
     new 2f257e8  fix build (#7358)
     new c198572  Fix data tutorial (#7329)
     new ef383c2  disable vgg-16 resnet converter check (#7369)
     new 63ae4c9  Small fix for versioning doc build (#7371)
     new 7e24097  Seed numpy random (#7372)
     new ada6d4e  Update (#7373)
     new d08abf5  Docs for GAN (#7378)
     new 1a617fa  fixes the incorrect retrieval of the batch size in the RNN forward pass. (#7385)
     new 07661ae  decouple record/train and add state readers (#7356)
     new 4c13b96  Update (#7388)
     new 251ae71  Add license header (#7379)
     new 44c2bfe  Clarify definition of cross-entropy metric in the documentation (clean up PR #7291) (#7365)
     new 8fb6407  Update (#7391)
     new 89e3ee3  [R] im2rec in R. close #7273 (#7389)
     new bd845aa  Drafted documentation for autograd.  (#7395)
     new a33ded7  fix [#7368] Caffe converter test fails, causing CI to halt for all PRs  (#7381)
     new f674bc4  1) Fixes for ImageIter (#7357)
     new e61fa7c  Tensorcore conv deconv support (#7347)
     new 85ad649  remove dmlc/mxnet logo from readme (#7256)
     new 46039c3  Update (#7405)
     new 613132e  [R][MISC] update Makefile/Jenkinsfile; use mx.ctx.default() in R test (#7401)
     new 59d7170  Hotfix mx.image documents (#7404)
     new 0d1407f  add verification to gluon dataset (#7322)
     new c55fc57  add Sequential compatibility to rnn layers (#7352)
     new aed297a  Add disclaimer and download link (#7402)
     new 1a3faa6  Initial commit of an MXNet converter. (#7413)
     new 8ae2970  Add autograd function (#7403)
     new a5373e6   Add more license files (#7429)
     new cee48d8  fix consistency of cpu/gpu in stn (#7374)
     new 89e29e2  Prepare for v0.11.0 release (#7432)
     new 16c8f96  MXNet -> Apple CoreML converter (#7438)
     new 606d3a9  Website fix (#7435)
     new 83078d7  cuda support for linalg-functions, restructuring of linalg interfaces (#7147)
     new 5e622d1  Small doc cleanups (#7439)
     new 8ad3c8a  clean up pom (#7440)
     new 4b8235b  broken link in readme (#7441)
     new 0142ea0  [R] vignette update (#7437)
     new 568b5a2  bce loss (#7304)
     new a13fa6f  Updating CoreML readme file (#7459)
     new 1470bf6  add Naveen's Code Signing Key (#7460)
     new 507e830  Fix toc (#7465)
     new b51515f  Fix apache link (#7468)
     new 09303cb  Update Jenkinsfile (#7466)
     new a21d3e0  Fix more broken links (#7480)
     new 7d6385a  fix autograd memory cost (#7478)
     new bca9c4c  add gluon resnet18_v2, resnet34_v2 models (#7484)
     new d634bea  Fix more links (#7485)
     new d7d31b2  add depthwise convolution's gpu version optimization (#7393)
     new 75ee597  fix a formula typo in doc (#7434)
     new c79d4e5  remove WaitToRead in dist-kvstore (#7489)
     new 1286809  Change git clone to specific tag for installation (#7502)
     new 462dee7  Fix description of argument parser (#7507)
     new 56eae58  Fixed Makefile so a null CUDA_ARCH is treated like an unset one. (#7515)
     new ff21e1f  Changed FullyConnected to use new linalg gemm, plus TensorCore if fp16 I/O. (#7505)
     new 6004e52  Modify pip install to specific tag (#7514)
     new d2dbffe  [scala-package][spark] fix example script (#7411)
     new 406bc19  V0.11.0 (#7518)
     new 5ee1cfe  Update for easy copy and paste
     new 0efc326  Fix a bug in SequentialRNNCell.reset() (#7449)
     new 686153e  Fix argsort + Update MShadow (#7535)
     new 32fc60b  Update Jenkinsfile (#7541)
     new ab14867  Move usage of persistent BN to cuDNN 7.0.3 (#7543)
     new 9796134  0.11.1 (#7542)
     new e0607bc  remove MXNet License from rcnn license (#7549)
     new f517d9d  Fix optimizer parms in + Don't repeatedly call slow script (#7545) (#7547)
     new 0b13631  Sparse Tensor: request for reviews (#7082)
     new 54b9240  add flatten option to fc (#7548)
     new 68cd9c9  Updating the LICENSE and NOTICE Files (#7563)
     new 491f81e  add resnet50_v2 pretrained (#7564)
     new 3932931  fixed minor typo (#7581)
     new 6d9b6a3  modify parameters counting of FC and CONV (#7568)
     new f68cb40  FP16-I/O conv/deconv to use pseudo-fp16, ignoring MSHADOW_USE_PASCAL. (#7527)
     new 050d85e  Set dev_id in streams, also update mshadow. (#7526)
     new d839abc  fix for amalgamation build with MIN=1 (#7597)
     new 9296907  Fix import error of broadcast max, min, mod in and add unit tests (#7572)
     new b0b4664  add ctx to begin_state in rnn_layer (#7580)
     new f489810  contrib ctc interface changes, cudnn7 CTC, and gluon CTC (#7442)
     new 0a3ee08  Fix symbol load json (#6420)
     new 7f90a39  Added and modified for bi-lstm-sort example (#6549)
     new c584b51  Fix RCNN multi-gpu bucketing warning which may cause OOM error. (#6965)
     new 3730f54  Changed next() method to use the seq_size attribute and not the global variable $seq_size. (#7521)
     new 125a126  Relaxing condition in slice (#7487)
     new d956d19  Fixing loss function code in tutorial (#7583)
     new 7f65a34  Add MXNet MKL pip install (#7598)
     new b34580e  add license to new file (#7599)
     new b710636  nightly build test mnist training and optimizer (#7559) (#7562)
     new 4b94360  make MXDataIter work without indices (#7456)
     new 97a15c2  NAG also has 'momentum' optimizer args (#7602)
     new bc468b0  Convert dot to linalg gemm (#7603)
     new 942e888  fix linalg_impl (#7611)
     new 3b3d824  set build status to success only after job ends (#7628)
     new c42453d  Fix build status of a test (#7629)
     new 50342a4  entire codebase build with mshadow_use_clas=0 (#7625)
     new b7fcd09  Update (#7630)
     new 2e6ef8c   unit test for csv iter, doc update for libsvmiter (#7623)
     new 1e48e12  gpu access of ndarray (#7496)
     new cb36058  refactor cudnn algo reg to no use string (#7561)
     new 39ff764  Update (#7634)
     new 9aa051c  fix tests (#7633)
     new e051297  [build] explicitly install JDK8 (#7574)
     new aceef5a  Add script to build doc files for all versions (#7636)
     new e845cec  add fashion mnist and move mnists to s3 (#7635)
     new 910b422  add doc for dataset (#7644)
     new 860dda2  Change apache package URL to https (#7622)
     new 4e11674  Pip installer for CoreML Converter: mxnet-to-coreml (#7624)
     new 12b244d  Parallelize windows unit tests of python 2 and 3 in jenkins (#7646)
     new 03b1d8d  Removed asset loaded insecurely and added the asset to be loaded from the origin securely (#7649)
     new ec1998c  skip failing test temporarily (#7648)
     new f6dd702  lower really high threshold to fix test failure (#7650)
     new 18c4eee  Doc updates for install and doc generation (#7647)
     new 5591f42  fluent (#7584)
     new 94a2c60  Refactor random linalg contrib namespaces (#7604)
     new a677b43  fix gluon fasionmnist dataset (#7655)
     new 62e6d2f  Parallelize Python 2 and 3 unit test cases in Jenkins CI. (#7658)
     new cb432a7  Change namespace and make logging functionality changes (#7627)
     new 546b917  update mklml and mkl mac support (#7587)
     new 9ed171e  [scala-package][spark] Resources running PS (role = server) should be explicit to Spark (#7571)
     new 583722d  Remove python function negative for rendering ndarray api in doc (#7657)
     new ec7cd6e  CSRNDArray from/to scipy csr_matrix; fix rand_shape_nd (#7638)
     new 23e0a35  fix ctc on softmax grad and req option (#7660)
     new ec71cff  ndarray.hpp needs to include op.h (#7495)
     new 470b564  Add string interface to updater to make it consistent with kvstore (#7585)
     new 6562a8c  1x1 convolution acceleration (#7613)
     new b7efd68  remove self-implemented speedometer (#7430)
     new cfdd081  [R] update tutorial link. close #6536 (#7461)
     new 62eaa36  add variational autoencoder example (#7190)
     new 42a43ba  Adding license header
     new de4c6e4  add -1 indexing to ndarray (#7668)
     new ee5bf32  [R] fix CI test (#7674)
     new 9cbcabc  sign compare warnings fixed (#7679)
     new f9a97ca  Add more gpu tests and fix nested try catch in Imperative Invoke  (#7676)
     new 2dbadf0  update epsilon to 1e-12 (#7670)
     new 4562ddd  Update docs/README link and folder name (#7673)
     new 0d294dd  [R] update Makefile. close #7675 (#7686)
     new 6d71577  Fix shape inference bug (#7682)
     new 31f27d6  Fix unit test for (#7693)
     new 7835864  [WIP] support autograd with sparse grad (#7695)
     new 93cd017  [R] fix Jenkins R testing. (#7694)
     new e095f5d  Adding opencv-python so import cv2 in the tutorial will actually work (#7700)
     new 7ca8b88  forbid setattr type changes and enable block replacement (#7667)
     new 783d12a  [scala-package] fix the bug in parent section of pom files (#7696)
     new 5b5723b  Use a big ndarray in gluon/data/vision (#7661)
     new 44feec0  Update (#7711)
     new 4e02553  updating a broken link in a tutorial (#7713)
     new d8e92df  Fix build error C2872 on VS2015 (#7715)
     new 9eafc1a  Fix Error (#7724)
     new f3087ff  gluon save/load optimizer states (#7494)
     new b81a039  Implementation of cbrt and rcbrt operators (#7041)
     new 802a61e  update caffe model converter for ssd models only (#7083)
     new 842c096  Fix for hang in depthwise convolution for CUDA 9 (#7685)
     new cdfe8a7  Isolated benchmarking support (#7662)
     new ef16cc2  Fix for UnicodeDecodeError on Python 3.x (#7728)
     new b3ad753  display type string instead of type code in error message (#7730)
     new d6cc2a8  update submoules with android fixes (#7416)
     new f39e3ed  Update (#7699)
     new da7c5bc  Clean amalgamation ws and skip failing test (#7678)
     new 0c66eff  modify syntax error (#7755)
     new 9a021c7  fix caffe model convert leakyrelu (#7702)
     new 92ee930  CoremlBugFixes:Input variable name can be something other than data; … (#7746)
     new b235cf7  Update doc for sparse related APIs (#7688)
     new 38381c7  Changing each task timeout from 60minutes to 120minutes (#7777)
     new 541158d  New operators linalg_syrk, linalg_gelqf. Numerical unit tests can run in float64 now (#7741)
     new 890e8f2  multinode dist_sync fails with mklml (#7776)
     new e242fd2  fix doc (#7780)
     new 84f8aac  Fix Jenkins R gpu compilation failure (#7779)
     new 5bd63f6  providing explicit null pointer for the provided_arg_stypes. (#7791)
     new 1267c6a  fix wrong dist-kvstore push/pull/rsp_pull (#7762)
     new 9bfd9e1  rm duplicated and unused code (#7764)
     new 087f96e  Fix Moderngpu usages in MXNet for CUDA 9 (#7789)
     new b30a056  Website redesign (#7787)
     new f81055f  Remove warnings about not used variable when compiling with CUDA 8 (#7769)
     new 2aa97c1  fix empty prefix case for namescope (#7794)
     new 10a72c8  Python symbol/ndarray api re-org (#7712)
     new 4807e73  Fix GPU/CUDNN memory allocation failure issues (#7795)
     new ec363d1  The order of hue composed transformation (#7756)
     new 09a4feb  Call vcvars for x64 before build on GPU (#7793)
     new 02e805f  Update doc README (#7797)
     new 3f742d2  Scala fix (#7782)
     new 0750c41  build: bump curr_win to 20170819, MXNet 0.11.0
     new 778529c  Merge pull request #268 from iblis17/win-prebuild
     new 3d35260 modified (#7600)
     new 5c95d29  Add fail flag for tutorial test (#7817)
     new bc30e54  [Perl] bugfixes for distributed training using tools/ (#7801)
     new ce20a60  [Scala] add BucketingModule (#7716)
     new 85a2d97  ndarray: add kInt8 and kInt64 into TypeFlag enum
     new 873715b  cblas: import all func prototype into header
     new 8bd307f  Merge pull request #270 from iblis17/nd-typeflag
     new 4e5ab13  travis : set make -j4 and travis_wait
     new d1bbcc4  base/build: using `Libdl.dlext` for searching lib
     new 0d54f4b  build: force enable LAPACK when Julia's OpenBLAS is available
     new 1dc29c0  Add back get_started index page (#7822)
     new 274424d  fix linalg_impl uninitialized error (#7819)
     new 2abee10  correct file path for API reference auto index (#7818)
     new 1175a1e  update dataset to use acceleration (#7821)
     new 4dcd96a  modify gluon image_classification for imagenet (#7626)
     new 17c7303  fix (#7671)
     new 4c2c1f5  Use GNUInstallDirs and improve CMake handling of shared/static mxnet library. (#6482)
     new bc206c2  update core (#7812)
     new 4808658  Change gluon link (#7690)
     new 8df3a5d  Update in example/classification (#7556)
     new 54cefb4  fix and add (#7754)
     new 45ebe53  Added C++ Option to file (#7830)
     new 2d2033c  Adversarial VAE (#7621)
     new a537257  Add sparse memory benchmark script (#7800)
     new fc2f708  Add learning rate properties to gluon trainer and optimizer (#7760)
     new f34643b  Merge pull request #271 from iblis17/fix-travis
     new b68ce5b  Merge pull request #265 from iblis17/ext
     new 21ccc56  Merge pull request #269 from iblis17/build-lapack
     new 957068b  Remove duplicate pylint (#7833)
     new 4044703  Add example of manipulating learning rates in gluon (#7834)
     new d3bd262  Refactor api table of contents (#7840)
     new e25074a  add option for segfault logger (#7836)
     new 23866bd  build: fix macOS LAPACK build
     new 85e95bb  fix condition (#7843)
     new 123deb0  change contrib CTC to zero-indexed, in log space. sequence mask for grad (#7727)
     new 2c5120d  Merge pull request #277 from iblis17/mac-lapack
     new 9d24cfd  only print if there are warnings (#7850)
     new 3f77f16  Disabling the test_CSVIter for now (#7829)
     new 4e439ed  build: remove FileRule on and fix CUDA config
     new 7cc66d1  executor: fix docstring of `bind`
     new 3ba3226  Optimizing the build (#7798)
     new dff605e  Adding developer keys for sandeep (#7482)
     new 2527657  Added a git clean to the Jenkinsfile (#7863)
     new 2372518  Update Jenkinsfile to sleep 2 seconds and change order of clean
     new 5d739da  ndarray: implement deepcopy via MXNDArrayGetDataNDArray
     new 2580dc0  Enabling persistent BN in cuDNN 7.0.2 (#7876)
     new ca9a383  Sparse operators for unary and binary elemwise NDArray operators. (#7577)
     new 9d56db6  Many loss functions  (#7605)
     new fd201d4  Fix CMakeLists.txt for unit tests (variable wasn't expanded) (#7882)
     new 14cfb5a  Fix missing of CUDA device on non-GPU host (#7855)
     new c560902  Fix Symbol Index (#7902)
     new 69d915f  Elementwise Sum (add_n) for rowsparse on GPU (#173) (#7832)
     new e81a3a8  [sparse] add ftrl optimizer for sparse (#7720)
     new 19af91b  Small improvement to IndexTensorToVector (#7919)
     new f713212  Remove unnecessary sequence layer size restriction (#7923)
     new 3c76e44  add warning to global norm clip (#7904)
     new 01ec708  add mobilenet to gluon model zoo (#7875)
     new f20534d  Merge pull request #273 from iblis17/issue-273
     new a51cba4  Merge pull request #281 from iblis17/bind-doc
     new 2e819b2  cuda support for new linear algebra operators (#7883)
     new 065adb3  add advanced indexing (#7910)
     new 16c2c5a  base: update GRAD_REQ enumeration
     new d824631  Merge pull request #275 from iblis17/file-rule-cuda
     new e3ad29d  travis: bring coverage report back
     new dcba690  build: CuDNN detection
     new 15a1eca  [Perl] reduce buffer copies in NDArray<->PDL conversions (#7815)
     new de41b88  Update (#7940)
     new e68298c  simplify CTC forward after namespace reorganize (#7936)
     new 9cdac22  R-package RNN refactor (#7476)
     new ec9f82e  Fix computation of num_inputs for Python API create_operator_entry for custom operators with 0 arguments. (#7967)
     new a7fe0d3  Increase the tolerance (#7937)
     new 67c0ad7  Add Kaggle w. gluon pipeline and k fold cross validation (#7915)
     new 6f764e9  update doc icon (#7969)
     new 3d4adfb  ndarray: `reshape` share the same API with Base
     new 812ee67  more informative error message for deferred init (#7976)
     new 3e3ac4c  Changed How To String to FAQ and added Discuss Tab to the Navigation bar (#7950)
     new 0111ec4  Faq discuss branch (#7983)
     new dc27f13  suppress wanrings for sparse tests (#7971)
     new 5db5da9  account for batch padding when updating metrics (#7949)
     new 6934ce2  Update (#7991)
     new 37a6516  Second order gradient and Subgraph execution (#7698)
     new 1306ef4  allow sequential add for multiple blocks (#7877)
     new 22bea88  add math fluent functions (#7992)
     new 98c4b29  gluon conv rnns (#7264)
     new ae975e5  allow users to specify repo url for downloadable contents (#7892)
     new 4aaefa0  Update (#8003)
     new 3e1ad58  Update (#8002)
     new 6ff309b  Fixed broken URL by renaming FAQ folder name to faq (#8009)
     new e632310  fix elemwise_sum test script (#8008)
     new b83a03b  user friendly image loading (#7974)
     new 6c3ca32  Revert "Many loss functions  (#7605)" (#8010)
     new ebf1bf9  Add an argument to omit row_sparse_push for end to end benchmarking (#7799)
     new bbd0e66  Merge pull request #285 from iblis17/cov-report
     new b68ca2e  Merge pull request #272 from iblis17/issue-272
     new e33da52  Update
     new 0b9bf78  sym-node: support `print` for debugging
     new 4e89c91  sym-node: extract a until function _create_atomic_symbol
     new 63a34cc  sym-node: implement reshape api as Base
     new 68d6ff7  sym-node: update deprecation for reshape
     new a2e270e  sym-node: update NEWS for reshape API
     new bb88d79  Merge pull request #283 from iblis17/issue-283
     new 9e0844a  sym-node: add reverse keyword arg
     new 55ccbfa  sym-node: more test cases for reshape
     new 3595403  Merge pull request #280 from iblis17/build-cudnn
     new df13ddd  Merge pull request #276 from iblis17/print-symbol
     new 6d1ba53  Merge pull request #279 from iblis17/sym-reshape
     new ff0556b  random: rename srand! to srand
     new 5b3c507  random: update NEWS and deprecated.jl
     new 32c588f  fix cached op (#8031)
     new 2e0ffdb  	Revert "account for batch padding when updating metrics (#7949)"  (#8032)
     new ed44e51  Issue #7751: MXNet 0.11.0 Release Feedback: and (#8035)
     new 09345aa  MXCallbackList is a struct (#8028)
     new 50564bd  fix bgr channel error (#8036)
     new fa7c654  Issue #7748: MXNet 0.11.0 Release Feedback: NOTICE file (#8043)
     new a36bf57  Add cuda rtc module (#8017)
     new e040610  Issue #7748: Update the Copyright years in NOTICE file (#8046)
     new 0ad14b6  Random refactor (#7939)
     new 08c15fc  Fix a mistake in example/ssd/ (#8022)
     new 3ba0c22  fix example (#8006)
     new 4c0df62  Kill running PR builds when a new build is triggered for the same PR (#8034)
     new f65da2c  Update (#7995)
     new 494a642  fix symbolblock (#8050)
     new 68181ef  Operator linalg_syevd: Symmetric eigendecomposition (#7966)
     new fb4760b  add cnn + highway network architecture for Chinese text classification example (#8030)
     new e727cdc  Disable the test test_batchnorm_training and test_cached (#8045)
     new 6452fe6  More sparse related docs (#7911)
     new 862e78c  Ignore Perl GPU failure
     new bc388cb  more python3 compatibility fix (#8064)
     new b5648a4  rename (#8056)
     new 3059030  Skip adding control_deps if nodes are the same (#8055)
     new 0449cc6  Revert "Kill running PR builds when a new build is triggered for the same PR (#8034)" (#8078)
     new d21de24  New tutorial of implementing operators in MXNet backend (#7828)
     new 4203133  fix path expand in model zoo (#8086)
     new 198df2f  Update (#8083)
     new b1d65e9  Fix faq breadcrumb (#8094)
     new 7bc48cd  Update the threshold for diff (#8095)
     new 429e2a4  [Perl] Fix for Perl GPU CI build (#8089)
     new b48124c  Fix link (#8102)
     new d887a0e  upadte the python version for nosetest (#8087)
     new ccfef53  variational dropout cell (#7067)
     new 22b2897  Fixed a bug in (#8099)
     new 752be1d  fix superresolution example (#8106)
     new 974ac17  Refactor infer storage function for sparse operators. (#7947)
     new db374b1  doc update (#8088)
     new 41c7690  Removing WaitToWrite in dist kvstore (#8116)
     new 53c6fa0  Add ability to query MXNet version to C API (#8013)
     new 5235bb3  OSX CMake fixes for tests binary (#8026)
     new 36043b9  Enhancement for distributed sparse linear regression example  (#7864)
     new 0d37d5d  LibsvmIter Doc Updates (#8111)
     new 283378a  correct boolean values for ndarray (#8104)
     new 44d72e7  Removed redundant hash of out_shape in algoreg key. (#8122)
     new 1637011  Autograd with multiple devices (#8124)
     new 0013ca1  [Perl] Adding Gluon interface to Perl, miscellaneous changes in order to sync with Python (#8015)
     new cac948b  Issue 7264:  MKL supports 2D/3D/4D input tensor only for concatenation (#7868)
     new 57d59ac  Update (#8131)
     new 236d1c2  Updating code owners (#8128)
     new 8a4221b  Update (#8134)
     new 251a8b3  random: upstream api change
     new ebb03d0  Merge pull request #288 from iblis17/upstream-random
     new 077a8cb  build: improve cuda detection
     new 032fad8  remove unnecessary flatten in model zoo (#8144)
     new b8e98f1  add storage type logging to graph executor (#8138)
     new 7b7f9ae  Add three sparse tutorials (#7921)
     new f6ecafa  changed url references from dmlc to apache/incubator-mxnet (#8147)
     new 759f509  update sparse LR example (#8141)
     new 0a53778  Merge pull request #282 from iblis17/srand
     new 92e5198  update arg params in (#8162)
     new 3a698da  Add gluon losses (#8149)
     new bbf1c0b  stable sum (#8136)
     new 0e8ad6e  Add infer storage type function for sparse slice operator (#8154)
     new ac7bbd0  base: support MXGetVersion
     new 17a22a3  Minor enhancement for the auto-encoder example (#8167)
     new 73b7d27  documentation fix: mx.nd.random.uniform --> mx.nd.random_uniform (#8172)
     new f56aaa9  Fix input/output format for gluon layers (#8173)
     new ee97715  Making mixed precision work with all optimizers (#7654)
     new 63e5ed2  Add diagnose script in tools (#8181)
     new fbfff66  shared module bug fix (#8185)
     new 06806f5  hotfix delete error (#8166)
     new b6b8da0  pr template (#8115)
     new 573a010  Delete workspace before build. (#8195)
     new 7232b12  improve diagnose script (#8200)
     new 904eb20  rand: docstring
     new aae016c  ndarray: pretty printing
     new a48991f  Zeroes CSR still needs a valid row_pointer array. (#7935)
     new 19082ef  Merge pull request #289 from iblis17/mx-get-ver
     new 9d90506  Merge pull request #295 from iblis17/doc-rand
     new 15def2c  Merge pull request #290 from iblis17/nd-pretty-print
     new d22373b  Gpu samplers (#8179)
     new 70b3dd4  Fix warnings (#8208)
     new 4286780  Fix segfaults and tests for distributed kvstore (#8207)
     new 2ff9584  style transfer example (#7579)
     new 506dd28  Enable MKL support for full pooling (#7810)
     new fa47c61  Get bz2 data fix (#8020)
     new 956e79b  update issue template (#8082)
     new c5d60b5  fix typo of gluon hybird tutorial (#8220)
     new 667932d  Solve problem in kvstore init (#8187)
     new c1beac5  increase number of training epochs. (#8214)
     new d43d7ec  Fix pip-test, install libgfortran3 on docker (#8233)
     new b2ce812  SSD Example: Fixed training bug (#7962)
     new 0092fa2  Fixing the Caught error (#8199)
     new a05ac49  Temporarily disable some unit tests to fix the build (#8253)
     new 49ef7b5  Update documentation links to point to
     new ddc6c39  add export to gluon (#8212)
     new 63ba460  ReleaseFeedback: License Files  (#8247)
     new 69a93bc  Sequential aug (#8243)
     new 43234d0  Basic CPU Kernel OMP selection based upon whether GPU has been used (#7854)
     new 3b36217  Multiplatform docker based builds (#7792)
     new 46ec178  Operators for sum(csr, axis=0) and sum(csr, axis=1) (#8174)
     new ffa6e45  Fix arange (#8268)
     new 65b2587  [WIP] New faster version of the RecordIO iterator (#7152)
     new d4120ef  Prepare for 0.12.0 Release (#8263)
     new 70c7868  Feature branch gluon (#8209)
     new 484d4c6  OMP num threads 0->1 (#8267)
     new b1d8e7c  scipy style sparse ndarray constructor and misc sparse fixes (#8261)
     new 01ae629  GPerftools update, also include include/mxnet/*.h as sources for CLion (#8232)
     new efef7b7  Temporarily disable 'test_row_sparse_pull' on GPU. (#8265)
     new 9f97dac  Refactor AdaGrad optimizer to support sparse tensors + unary and binary refactoring for new infer storage type logic (#7903)
     new 9d3ce81  Proper float64 support for unary/binary elemwise operators (#8281)
     new 73390f6  trigger test
     new a8be0d2  fix
     new 99099fa  generate op frontend code for IDE auto-complete (#8244)
     new 1c1c788  [CMAKE] Fix windows cmake build (#8227)
     new 22ab185  Added my code signing key (#8293)
     new 421ef8a  remove usage of install command from code gen (#8304)
     new f3df462  Negative begin and end support for csr slice (#8241)
     new 8c66b96  Preparing for 0.12.0.rc0: Final changes before RC (#8301)
     new 9cf41fb  Enable smoothing in softmax operator (#8125)
     new e797f45  v0.12 regression: Fix registration of children for Block (#8277)
     new 6a8d0e6  Revert "[CMAKE] Fix windows cmake build" (#8311)
     new 9034855  fixed broken links. https was pointing to http for (#8300)
     new 2e5160d  Update (#8320)
     new dd3853f  fluent methods for missed ops (#8329)
     new 349ebc3  update ps lite (#8327)
     new f9e25b4  Fix unused type warning (#8316)
     new 050d304  Misc fixes for sparse distributed training (#8345)
     new 4332ecd  Fix the Readme (#8369)
     new 72b0051  Allow test to converge (#8351)
     new ad20d91  Update cudnn_algoreg-inl.h (#7988)
     new d6062f7  [Perl] emulate Python zip() for Perl (#8192)
     new 590d7b0  add profile option for frontend profiling to image script (#8171)
     new 725a542  Fix Typo (classification) (#8376)
     new c7837b5  Timing output for test_factorization_module when Verbose enabled (#8363)
     new c82c555  Use omp_get_max_threads() when OMP_NUM_THREADS environment variable is set (#8379)
     new 87068e6  CPU optimization for ActivationOp (#8296)
     new 81b134f  ndarray: implement `fill` and `fill!` (#297)
     new fa5126f  Sparse operator performance improvement (#8412)
     new 171fd83  dtype default to source_array.dtype for sparse ndarrays (#8403)
     new 6c3aee4  fix using default mean pixels (#8352)
     new 1df79b4  fix (#8353)
     new a79e33c  Merge pull request #287 from iblis17/issue-286
     new 71f2d40  base: merge _julia_to_mx_param into dump_mx_param (#296)
     new 3824df0  Fix mx.chain var reference in macro (#299)
     new 840b9f1  upgrade MKL (#8378)
     new 9ca5caf  Lint fix (#8402)
     new d6f6416  [CMAKE] Cmake changes, upgrade training test so it converge (#8343)
     new 5073171  fix diagnose if hashtag not found. (#8346)
     new 5ebf682  rename and restructure model zoo models (#8446)
     new 971c712  C++ Framework and validation tests for operator validation and performance (#8435)
     new 3552b95  Optimizations to set_to, copy, fill ops (#8340)
     new eb08012  Memset/memcpy/omp profiling tester (#8380)
     new 68038ad  fix place device (#8450)
     new c5afa1f  bugfix and GPU support for syevd (#8426)
     new 37def24  gluon lambda block (#8418)
     new fc0293b  Fixes #8292. Bug on Slice accessing uninitialized memory in param.beg… (#8400)
     new 3bec7f1  fix topk with big shape (#8398)
     new 1787138  Fix arange bug on ARM (#8472)
     new a6fa7b4  fix condition when CSRNDArray is used in NDArrayIter with shuffle=True (#8374)
     new 100eb88  Fix urllib bug affecting Python 3.x in the finetune tutorial (#8229)
     new 41f8f55  visualize: fix node_attr on upstream master (#301)
     new 78b0a1e  commiting v12 changess (#8478)
     new c1846ce  Simplified unary/binary math operators (#8361) (#8361)
     new 8ac6ec9  Add Scala package dev tools for deploy (#8498)
     new 467a008  Getting read of maybe_uninitialized warnings (#8318)
     new e22d993  Fix of log10_grad, log2_grad (#8502)
     new 8592e1c  add default hash to ndarray (#8476)
     new 29a6a75  bump up version (#8488)
     new 9e0432a  fix makenonlossgrad bug (#8508)
     new 23f5af3  fix expand_dims if axis< 0 (#8489)
     new 396943e  Correct Initialization Description in Finetune Tutorial. (#8517)
     new c02c6b1  fill parameter shape (#8528)
     new b537830  Fix typo in gluon l1loss docstring (#8325)
     new b8745e9  Fix a description (#8501)
     new dd278f3  Tweak a markdown list in (#8533)
     new 90ba62a  Move OpenMP calculations into its own class (#8475)
     new 52d809c  Fix for issue #8491, elemwise_mul nan behavior (#8492)
     new 53e891b  fixed practioners to practitioners (#8534)
     new ba74540  rename output layer of resnet (#8536)
     new 63d5394  Add asscipy feature (#8362)
     new d275701  Update for 0.12.1 (#8544)
     new 990bab8  Fix libgfortran3 error, Add flag -y (#8535)
     new 9fcab40  symbol: add (#302)
     new 3f37577  * [cpp-package] fix for issue #7725 (#8551)
     new 4f182ee  ndarray: more Base-like APIs (#303)
     new a3317f1  ndarray: implement rdiv (#292)
     new cf2545f  Fix unused var warning in release build (only used in assert on debug builds) (#8561)
     new 8cc5c1c  util: docstring of _sig_checker (#305)
     new 5c84c30  Engine reserves cores from OMP.  Set some defaults for dynamic and recursion (#8553)
     new a488d7a  ndarray: intro TakingBroadcastSeriously and broadcast unfusing (#300)
     new 938c25f  csr ndarray iter with shuffle (#8550)
     new c64c307  div_scalar for sparse (#8540)
     new 4862c41  cpu sparse embedding op (#8460)
     new c2358f9  Operators for mean(csr, axis=0) and mean(csr, axis=1) (#8264)
     new bbf8b3e  Update mshadow to 2d7780c3f2eefe4453fa419862d1b2089bedb8d5 (#8516)
     new bf2336c  [sparse] slice for csr on two dimensions, cpu implementation (#8331)
     new 57cc677  ndarray: elementwise power (#293)
     new eea128a  ndarray: fix `add` broadcasting and more tests (#308)
     new f839be1  Fix deprecations (#307)
     new 70b68b1  slice operator supporting arbitrary values of step (#8558)
     new 8dc09ec  gluon with multiple data type (#8522)
     new 6da0472  reimplement cross-entropy operator in recommender example with NDArray (#8588)
     new 4919273  ndarray: getindex/setindex! linear indexing (#294)
     new 1fc03f2  ndarray: elementwise power for irrational (#310)
     new 00d61d2  executor: return `outputs` on forward (#312)
     new 3f93ffc  git: union merge policy for (#309)
     new 678b49e  ndarray: regularize fill! API (#311)
     new a9f11a0  file not exist (#8596)
     new e083e5f  doc: cleanup useless enc file (#315)
     new 97e7015  Update PR & Issue Template (#8555)
     new c9bde1b  expose group2ctx to module (#8539)
     new 00ddcf4  support for lapack functions with mkl (#8577)
     new 582e14c  Fix for "name epochs is not defined" (#8399)
     new 5c8981a  [R] Initializer fix and adjustments to RNN API (#8121)
     new 35ceea7  use first class cuda with cmake 3.9 and cuda9.0 support (#8572)
     new 9572340  generalize array dataset (#8612)
     new 23a9294  optimize broadcast (#8566)
     new 8ba5de8  Pull more optimize and simplification changes from tuner branch (#8599)
     new 79b00cd  Fix python 3 docstring (#8616)
     new c11a608  Fix changed OMP call in LaunchEx, add broadcast perf test (#8622)
     new 09b9718  build: prevent CUDA_HOME KeyError (#316)
     new 9304e6e   sym: broadcast unfusion (#314)
     new f8e1938  Remove Compat and src/compat.jl (#318)
     new a19fc93  test: replace `reldiff` with `isapprox` (#321)
     new 526857e  docs: address docstring override warning (#320)
     new 8071186  travis: enable 'build branch updates' on master/stable only (#319)
     new e4b6c0e  docs: update examples in overview (#317)
     new d8d2c39  Continued Work on Advanced Indexing (#8246)
     new 399ac03  [EXPERIMENT] increasing timeout to 24hrs. (#8613)
     new a51d9b5  docs: fix more example rendering in overview (#322)
     new 710bf4e  check_format of sparse ndrray (#8259)
     new 25ef9bb  Fix HybridLambda repr (#8565)
     new e27f054  fix custom op error when using auxiliary states (#8637)
     new ca4b683  Imperative bulk execution (#8520)
     new 73b856b  Update README badge (#324)
     new 079fc45  add nightly test for dist-device-sync (#8597)
     new 1dfb5b3  Cut a release for mxnet v0.12.0 (#304)
     new b56c44e  Bump upstream to v0.12.1 (#326)
     new 9efedbc  Update releasing note (#327)
     new 6b69608  fix bug in TryCrop() (#8644)
     new 938eda9  Support sparse for custom python operators (#8620)
     new c8f7dce  sparse embedding operator, gpu implementation (#8647)
     new 85d4bd2  multi processing and fork fix (#8677)
     new 4d05bff  ctypes speed improvement (#8642)
     new a4aab01  docs: minor changes in overview (#330)
     new 57b50d4  build: set upstream version back to master (#328)
     new 3107326  Refreshed branch quick_pr (#8643)
     new 33698ea  Fix sparse dot test failure due to launching kernel when nnr = 0 and bug of square_sum (#8470)
     new 257ad1d  Doc updates for sparse operators (#8641)
     new bdd14a8  mac build fix (#8692)
     new a8f79bc  fix random docs (#8676)
     new a499f89  2bit gradient compression (#8662)
     new d2a856a  Restored some copyright attribution that were accidentally removed. (#8688)
     new bf7a0ff  Invert environment check (#8706)
     new 6ca944b  disable test causing mklml-cpu to fail (#8713)
     new 38a032c  [COREML] Update the json getter (#8698)
     new 9dcdf57  Fixing the monitor callback of the bucketing module. (#8696)
     new f79d22d  optimization for dot(csr.T, dense) = rsp (#8611)
     new 7f7e13d  replace `has_key` by `in` (#8317)
     new 504f42f  Revert "2bit gradient compression (#8662)" (#8711)
     new 7fc0396  fix group2ctx with null reqs (#8717)
     new defd3c3  Remove experimental warning on Gluon and add Gluon tutorials (#8710)
     new fd45517  Doc src and fix (#8718)
     new 8564f19  executor: add and Base.print (#337)
     new ceb7fbf  travis: enable caching for reducing build time (#329)
     new 1a7887c  ndarray: add outer constrcutor for AbstractArray (#334)
     new 010ea3c  sym: implement get_children (#333)
     new 1f50a14  sym: fix printing of symbol generated via get_internals (#332)
     new 4dc8933  Fix docstring format (#8734)
     new 1f63a9a  Revert "Support sparse for custom python operators (#8620)" (#8733)
     new cd47962  2bit gradient compression (#8728)
     new 068b589  Kernel operator tuning (#8686)
     new cace29f  NCCL integration (#8294)
     new bbe078e  [cpp-package]Update readme#8655 (#8746)
     new 5c328f1  comment typo: there are there (#8753)
     new 12c3fb1  fix lint with cmake (#8752)
     new 1852e2f  Add cast to Block and Parameter. Implicit dtype casting is removed. (#8735)
     new 9d48f74  add code signing key (#8743)
     new eb72d72  Added a security best practices doc (#8744)
     new ec6144f  a user friendly way to use g2c in module and an example of g2c (#8632)
     new 8df20a2  [scala] EvalMetric sumMetric is now a Double instead of a Float (#8297)
     new abf01b4  Remove spureous std::move, fix warning regarding RVO being prevented (#8755)
     new b0556e6  Update regression-example.jl (#339)
     new 5518986  Explicitly convert float value (#8758)
     new 91a410e  test: minor style changes for mlp-test (#340)
     new f126482  sym: handy macro `mx.var` for creating `mx.Variable` (#338)
     new 89d2a71  Updating ps-lite submodule (#8769)
     new e835777  Fix doc link (#8776)
     new 5ad6eab  [CMAKE] Fix cmake library path when installing python package (#8680)
     new 1f6d586  [ImageIO] Fix image io for opencv3.3 (#8757)
     new be18794  Make make lint compatible with python3 (don't call python2 explicitly) (#8401)
     new a0031ac  cast scalar value in invoke to float (#8778)
     new 1264313  Add Intel openmp as a submodule and build for x86 architectures (#8730)
     new 3953e79  Fix weird hang bug due to cuInit sometimes calls fork (#8790)
     new 0bf76a8  Fix weird hang bug due to cuInit sometimes calls fork (#8804)
     new 5088477  remove meaningless type qualifier (#8774)
     new 3c32f76  fix build (#8788)
     new 65b57b1  Fix a typo (#8809)
     new cb06a9a  build: bump max number of make jobs to 32 (#343)
     new 2ccc6b6  fix a typo (#8811)
     new 8bd65fd  Fix warnings (#8813)
     new fdc0766  multi-device support (#8812)
     new 05047ad  Tune without Launch specialization macros (#8719)
     new c462d34  Python 3 compatiblity during optimizer serialization. (#8334)
     new 2353229  fix building warnings below (#8820)
     new 91ffd69  Update (#8819)
     new 45d6279  ndarray: add `, MIME"text/plain")` (#347)
     new 3226bce  mxent link fix (#8824)
     new e0f625a  example: fix batchsize config in regression-example (#342)
     new 935eb35  ndarray: make _minus type stable (#345)
     new 2a5a284  model: refine and test cases for FeedForward constructor (#346)
     new ce4b921  fix (#8838)
     new 7aea2ef  fix search (#8841)
     new 632c897  Refreshed branch fix_omp_link (#8842)
     new 2f8c1e8  add inline for cached op & fixed a bug when calling backward on variable (#8701)
     new 5b00a5f  change default arc to Common (#8624)
     new 47a71a6  Profiler: set cpu/gpu num during execution (#8722)
     new 0df7953  Moved cub to 3rdparty (#8878)
     new 6d6c4ea  fixed broken links on master (#8879)
     new 89c4db7  Use https for Jquery file (#8855)
     new cabbcff  Fix typo. (#8888)
     new 2aa506a  Move legacy operators under the nn directory (#8858)
     new 319494f  Change the initialization for default constructor to avoid using explicit constructor (#8854)
     new ae0fe77  fix reshape check (#8893)
     new cb042fd  parametric NDArray (#331)
     new d7da05b  fix race when temp space is used in copy & fix instance overwrite in g2c (#8867)
     new f13ce52  doc improvements for default param 'TShape()' (#8906)
     new 3b3016e  Caffe to MXNet code translator (#8782)
     new 914fab6  ndarray: minor style changes (#349)
     new c43d0dd  model: fix test cases macro (#355)
     new 1dc4aea  fix multiprocessing too many open files (#8904)
     new 09ee1f4  ndarray: implement `size(x, dims...)` (#350)
     new 8a19651  ndarray: implement endof (#351)
     new fd7fb79  ndarray: make _plus type stable (#352)
     new ddb4782  Update (#8760)
     new b9855c4  ndarray: type convert of _mul_scalar (#356)
     new ad57be9  ndarray: type convertion of _plus_scalar (#360)
     new 642b17b  ndarray: type convertion of _rdiv_scalar (#358)
     new 7639e2b  fix documentation homepage typo (#8931)
     new 2a85cd4  Add myself as contributor. (#8929)
     new 14cd740  Remove finalizers from Scala API (#8887)
     new 95a226d  Fix unit test build (#8946)
     new 6c51629  rsp[:] = const_value, used in constant initializer for rsp weight (#8891)
     new 6f403c0  Initial Prep for 1.0: bump up version and add 0.12.1 changes to master (#8704)
     new 5f8efb0  NDArray Indexing tutorial and Gradient Compression FAQ (#8766)
     new b909769  [R] various small changes for efficiency and robustness in R package (#8118)
     new ce0e237  doc: sort out table for ndarray api (#362)
     new 43d534d  Small first iteration for profiler: use nonblocking queue (#8803)
     new 7484437  [Do not merge] Move to new CI (#8960)
     new f8d4f62  ndarray: type convertion of _minus_scalar (#354)
     new c965c8b  Change type of 'lr_refactor_ratio' to 'float' or else the learning-rate-refactor will never work (#8982)
     new 14f9748  ndarray: do not auto-import _full, _ones and _zeros (#359)
     new 5c11ecd  Fix race condition in engine start/stop (#8995)
     new c94b104  Implement assignment operator on Timing 'info' struct (#8996)
     new f154101  Some fixes for example/reinforcement-learning/parallel_actor_critic (#8991)
     new b652a20  Fix multi task (#8987)
     new 075ef68  Fix 'from_layers' of resnet101 by changing names of extract-layer from base network (#8998)
     new c9e440d  Reduce the verbose warnings in CI (#8981)
     new 24ea5b0  Factorization machine example & sparse example folder re-org (#8767)
     new 9d4bb9c  Minor changes to Caffe Translator (#8939)
     new d77d937  Fix __repr__ for gluon.Parameter (#8956)
     new 918fac3  Update (#9003)
     new 3d34f2e  [CI] Fix indentation of Jenkinsfile, no code changes (#8988)
     new e692d58  Updates for R readme and example (#8984)
     new 7d610f8  Correct the alexnet url. (#8975)
     new 01b7ae3  change workspace allocation in random samplers (#8935)
     new 1c22bc7  [clang][warning] fix warnings about braced initializer (fixes warning in concurrenc… (#8870)
     new 84fde4d  Locate liblapack.a in /usr/lib64 (#8073)
     new ac4369c  update readme (#9005)
     new 3ee1ca2  Fix Example example/dec/ (#8990)
     new daf787c  ndarray: make trigonometric functions dot-call only (#361)
     new 233fcfc  ndarray: change internal api of plus to help autograd (#364)
     new eb819b0  ndarray: change internal api of mul/div to help autograd (#366)
     new 2ca5565  ndarray: change internal api of minus to help autograd (#365)
     new 12198f0  ndarray: copy(AbstractArray, context) (#367)
     new 2d7cdc6  Reexporting NDArray and SymbolicNode (#306)
     new db09528  example: fix parametric NDArray in lstm example (#370)
     new 8a8e770  Pylint fixes (#9013)
     new 910c835  Delete (#9009)
     new 20666c3  Add wide and deep model into sparse example (#8180)
     new 2700ddb  Symbol __getitem__ using list_outputs() is too expensive (#8989)
     new 27c66ec  sym: overload `bind` (#372)
     new 5284513  Update for MXNet 1.0. PEP8 fixes to code and misc improvements. Tested under Python 2.7 and Python 3.6 on Sagemaker. (#9036)
     new 06ed4dc  Add truncated bptt RNN example using symbol & module API (#9038)
     new 3901f5f  Use backtrace from dmlc-core with demangled C++ symbols on segfault (#8851)
     new db15049  Remove redundant or unused imports (#9021)
     new 167871a  fix small memory leak of sparse embedding (#9025)
     new 087a906  Adding .png to the list of allowed extensions in the (#9043)
     new ca79cfc  Fix Gan tutorials with symbolic interface (#9030)
     new 8f79651  update autograd.Function docstring for usage (#9039)
     new 6d1f07d  Remove defunct NLP tutorial page (#9045)
     new 9e2bdeb  Disable test_optimizers:test_sgd (Temp fix #9000) (#9001)
     new 73fa65f  Add cscope filenames to gitignore (#9019)
     new 85a5b8e  fix VGG error in benchmark_score (#8664)
     new 41394a4  Disable test_operator_gpu.test_svmoutput_with_type (#9059)
     new fdd41a8  Add googletest as a 3rdparty library (#9016)
     new d43f4f9  improve image io example (#9061)
     new 1f70a48  Updating the build documentation with instructions to build mxnet with NCCL (#9058)
     new c4a76da  Implement Khatri-Rao operator (#7781)
     new 49c60fd  [example] fix neural-style end2end example (#9063)
     new bfa0989  Fix examples of profiler and cpp-package (#9065)
     new 8623bab  add CapsNet example (#8787)
     new 392e7ac  Increase git timeout for CI (fixes #9070) (#9071)
     new a8862a7  Remove torch support (#9072)
     new 09ed385  square sum gpu impl (#9014)
     new 1c8e066  Restore submodules to dmlc (#9069)
     new 48f62cf  Fix Khatri-Rao test build (#9081)
     new bfb1cc4  ndarray: remap hyperbolic function (#374)
     new 6609616  ndarray: support transpose on 1D array (#375)
     new 881759f  example: fix MSE init (#376)
     new 5908d97  base: cleanup stale code in _defstruct_impl (#377)
     new a941f3a  ndarray: add modulo operator (#373)
     new 8135a63  ndarray: type convertion of _div_scalar (#357)
     new 5540ee9  Fix check against nonexistent variable (#9082)
     new 1e20f50  ndarray: protect from diving zero for non-inplace op (#378)
     new 24c7231  Modify Caffe example to use module interface instead of the deprecated model interface. (#9095)
     new 2a4505a  add hyper parameters alpha and resolution, add extra compact symbol with less lines of code (#9099)
     new d921225  doc/ndarray: add ref table for hyperbolic function (#382)
     new 4507598  ndarray: `cat`, `hcat`, `vcat` (#380)
     new 39ecb32  base: improve MXError display (#379)
     new 3b7134a  Some changes to numpy_ops examples -  (#9092)
     new cb443b7  ndarray: porting Python's autograd (#274)
     new d23b91d  added assertion to validate accuracy (#9107)
     new 5858d62  Remove defunct demo (#9060)
     new df25378  add shared storage in windows (#8967)
     new fb254c4  Updating the Python readme (#9075)
     new b53a13e  FCN example updates (#9066)
     new 5dcdd5d  Update (#9103)
     new c6d4e7f  image-classification example fix python3 compatibility (#9053)
     new b9b4532  More details to the windows build process (#8519)
     new 9b0a5ca  csr slice operator, gpu implementation (#8814)
     new 73dba33  [BugFix][CoreML Converter] Dense layers w/o bias. (#8684)
     new df55829  Fix compatibility of examples with py3 (#8993)
     new 16bd961  Usability fixes for examples (#9091)
     new d1d3e97  Fix example/reinforcement-learning/dqn (#9128)
     new 883cdd3  build: make appending lib to MXNET_HOME optional (#386)
     new 500ed5f  fix multi worker dataloader deadlock (#9126)
     new 9272d9c  Fix nadam (#9127)
     new ae70769  add zero-grad for rounding ops (#9040)
     new a38ab6b  Install docs - default to install CUDA 9 and cuDNN 7 (#9147)
     new b59943e  warning autotune show once (#9132)
     new 521825d  Fixed incorrect wording in mnist tutorial (#9125)
     new c71a2f3  Usability improvement bi lstm sort (#8944)
     new 080d29c  Fix float16 min and max (#9149)
     new 5c3acff  Add Wikitext for gluon/rnnlm (#9090)
     new bfeba81  ndarray: support matrix/tensor multiplication (#384)
     new 3787895  ndarray: remap expand_dims (#381)
     new ddec3cc  R RNN API fixes and Optimizer clip gradient on NDArray (#9022)
     new 813bfdc  random: Base-like APIs for rand, rand!, randn, randn! (#383)
     new c6cdf51  Merging version changes from 1.0.0 to master (#9168)
     new b91ccdc  example/nce-loss fixes and improvements (#9093)
     new f8af6f5  LicenseUpdates from 1.0.0 to master (#9170)
     new 710d0f6  updated master branch with 1.0.0 release on the home page (#9173)
     new bccf9fa  Remove duplicate material (#9080)
     new d90c1b7  bump up version to 1.0.1 (#9183)
     new ef85063  Dropped Sergey's explicit copyright note per (#9191)
     new 3219620  Mkl fix pr (fix an issue in and add test case) (#9153)
     new 17427d1  Make versions of python dependencies like numpy, deterministic. (#9169)
     new ad1ed7c  fix incude error in plugin/opencv/ (#9206)
     new 911c81e  Support negative axis in concat (#9204)
     new 01a2a40  fix parameter overwrite in _generate_symbol_function_code (#9180)
     new 02b1131  Rnn example updates (#9074)
     new f20a205  Install docs - update info about DLAMI (#9150)
     new be40d31  refactor graph exec (#9122)
     new 7f1dd77  Usability improvement for cnn_chinese_text_classification (#9027)
     new d3cdbf6  ndarray: remap activation functions (#387)
     new 2586c66  Modified nce-loss README to use image files in dmlc/web-data repo (#9212)
     new 34a5195  fix random generator: do not gen seed each time (#9119)
     new 35781a4  Add result sections in sparse tutorials (#9085)
     new d0388d0  Update mkl (#9218)
     new f11e591  fix autograd import path (#9221)
     new 4c15f58  fix link error for parallel rng (#9256)
     new 7e1c227  fix bayesian-methods example (#8958)
     new 9d60663  ndarray: refine signature of `clip` (#391)
     new e3e1f23  raise err in io.NDArrayIter for invalid usecase (#9228)
     new e4128ec  Fix website parsing error (#9260)
     new 25dd12d  Adding multiple pretrained models (#9215)
     new a4143e9  Bump :copyright: (#397)
     new ae425bf  doc: append 'Defined in' into operator docstring (#393)
     new 930090b  ndarray: fix 1D NDArray display (#399)
     new 79cb705  Lint enhancements to R demo scripts (#9270)
     new a1cef7f  ndarray: fix type convertion in setindex! (#398)
     new 3f10bd4  fix typo for assert messages in gluon/nn/ (#9277)
     new 26e81d6  Add documentation generation to PRs (#9281)
     new 19f8e2b  Add storage fallback msg to copyfromto (#9219)
     new 57d1692  Modify R package gitignores (#9269)
     new 4aff838  Update deadsnakes ppa name (#9268)
     new 5b99b25  1 d conv with cudnn (#9184)
     new 5ab629f  Add timeouts to windows tasks in Jenkinsfile (#9280)
     new dc48eb5  Update to include results of the code snippet (#9123)
     new ef82fd2  Update (#9287)
     new 4e3ac33  update front-page model zoo to gluon model zoo (#9286)
     new 12cb0d2  FTML optimizer implementation (#9262)
     new 728e5c2  fix trainer: init kvstore before load states (#9224)
     new 7e7cf3a  Contrib operators for object-detection bounding box related stuffs (#8603)
     new 996d548  update with details on typical usage of the scripts (#9108)
     new 600a163  Updating the README files and examples in "ctc" and "recommenders" folder. (#9073)
     new 9eff221  Usability improvement mxnet vae (#9008)
     new 9d51e1a  sparse output for binary scalar op with zero (#9227)
     new f79f0d4  Make numpy functions return Symbol instead of numpy object array (#9293)
     new 8505442  Add operator for dot(dns, csr) = csr (#8938)
     new 546b73a  Add clarification in sparse api doc (#9282)
     new 86ffb5d  ndarray: inplace modulo operators (#389)
     new 30852dd  ndarray: support modulo operation in macro inplace (#390)
     new 0ef5966  ndarray: add Base.ones(::NDArray) and Base.zeros (#363)
     new 2df29cc  Updating the PyPi package URL to be Apache and not DMLC (#9299)
     new 661bcef  bugfix for parallel rand generator on multi-gpu (#9300)
     new 469955d  Increase timeout for GIT on windows (Fix #9297) (#9298)
     new 5e587e5  fix w&d (#9303)
     new 4988535  add epoch parameter to export() of HybridBlock (#9290)
     new c3616ed  Add label_from_zero_one argument to LogisticLoss (#9265)
     new d918868  Fix the gradient of gather_nd (#9200)
     new 1602b12  Add Marco de Abreu to CODEOWNERS related to CI (#9309)
     new 9be50e0  Fix CODEOWNERS syntax (#9313)
     new df9f79a  standard update for sparse sgd_mom_update (#9189)
     new f9fd88b  fix utf8 error in odc (#9322)
     new b596cf6  Prepend LD_LIBRARY_PATH to dll_path in find_lib_path (#9311)
     new 004dead  Fix custom op multi-gpu scaling (#9283)
     new 898eab3  added assert for achieved accuracy (#9106)
     new 2d28593  example/autoencoder fixes for MXNet 1.0.0 and pylint and addition of README (#9097)
     new acff519  updated comments for mnist tutorial (#9104)
     new c3d6cf0  Fix example/module folder and remove duplicate examples (#8964)
     new 49399fc  ndarray: broadcasting along dimension on arith operators (#401)
     new deb25bc  Refactor scala code/tests/examples/docs to use org.apache.mxnet (#9324)
     new 3463967  change username javelinjs -> yzhliu so as to be more easily mentioned (#9351)
     new 8ad77f3  Temporarily disable test with failing http connection (#9333)
     new 8168292  fix hard-coded in_channels in gluon model zoo resnet (#9341)
     new 9543db4  fix path separator (#9352)
     new 243ea39  Fix CUDA auto-detect and cxx flags (#9362)
     new 50bfdc7  added updated examples to list and contrib info (#9356)
     new 492a5fd  add MSRAPrelu initializer for cpp-package (#9121)
     new f30bb4a  Use provided scope for log4j dependencies (#9365)
     new e79ce71  ndarray: broadcast comparison operators (#402)
     new 2d49d59  Fix warnings on cpp-package (#9089)
     new 46cf856  Fix unit test build (#9371)
     new fd5e5fe  Update dependencies / Makefile + add Maven profile to support build with Scala 2.12 (#9339)
     new a80245d  improve memory usage in (#8826)
     new 6c1f4f7  Add mxnet.text APIs (#8763)
     new 2b73aac  Add clang jobs to the CI (#9330)
     new 3eae3bb  Update MKLML to v0.12 (#9383)
     new d511d98  Remove using namespace mshadow from header file (bad practice) (#9321)
     new 1732460  Fix C++ unit tests (#9349)
     new 51ea815  Revert "Temporarily disable test with failing http connection (#9333)" (#9379)
     new 4600070  support regex of collect_params() (#9348)
     new 5106985  example/ctc improvements (#9345)
     new 68750bc  Update (#9386)
     new 5251b86  Signum optimizer (#9220)
     new 67cfbc8  Revert "Add mxnet.text APIs (#8763)" (#9401)
     new f129992  adding missing modules to DEC (#9407)
     new 47a3422  Update (#9402)
     new 2ab0584  ndarray: copy!(NDArray, AbstractArray) (#405)
     new e4957c3  resubmit #8763 #9394 to contrib (#9406)
     new b1ad161  Fixed memory leak (#9400)
     new 171d717  bitrot fix. (#9414)
     new 786e376  rsp push and rsp pull for comm device, used in kvstore('device') (#8732)
     new 134b205  Use a better formula to calculate sigmoid_bce and logistic_loss (#9404)
     new 6dcb0eb  move concurrent/identity blocks to contrib (#9427)
     new b623488  Fix warnings on unused variables (#9432)
     new dea19e3  Add recommendation to use RAII to the contributors guide (#9439)
     new 3454974  Add discuss link for asking questions to the README (#9437)
     new 1b5caf6  Fix undefined behaviour in, use of delete instead of delete[] to free arrays through unique_ptr. (#9435)
     new 63394de  header updates to clarify use of CODEOWNERS (#9426)
     new 52f5330  Update (#9424)
     new a4a09e5  Random seed setting (#9409)
     new 9cc8ea3  Re-enable all op segments when in batch mode (#9055)
     new 3ac5376  Batching improvements for GEMM/TRSM operators and full MKL usage docs. (#8846)
     new e36a945  raise warning when we detect Block inside nested list/dict (#9148)
     new 262c74c  Introduce the ENABLE_CUDA_RTC build option (#9428)
     new 23f973b  amalgamation library fixes for python  2.7 and 3.6 compatibility (#9346)
     new 25c7232  Update (#8825)
     new 11cb609  removes python path insert of tests folder for examples (#9151)
     new 894c6de  Revert "header updates to clarify use of CODEOWNERS (#9426)" (#9442)
     new 02b69db  Limit the to run on 8 GPUs (#9441)
     new c74cf1b  enabling multithreading in broadcast_reduce (#9444)
     new 8054290  Fix crash when opening an image, fix exception safety. (#9370)
     new 59700b0  np.fromstring: DeprecationWarning (#9445)
     new d35c3d8  Test backtrace demangling functionality (#9434)
     new 16746c1  Refactor dropout operator to use ParallelRandom generator and also react deterministically when seeding (#9366)
     new f960522  Sequence Operator Improvements (#9306)
     new b396bc3  text: use _contan, import modules, and keep code style consistent (#9446)
     new 89b4fa0  correct usage of bool arguments from command line (#8880)
     new a182aec  Update (#9447)
     new 9353c2e  Remove empty lines at the end of file (#9459)
     new 1c990bc  ndarray: new constructor NDArray(Type, AbstractArray) (#404)
     new b023bd5  Add CMake CUDA build job to the CI. (#9329)
     new 46c9f48  Rename kvstore/utils.* to kvstore/kvstore_utils.* (#9463)
     new a937d98  [CMake] Add im2rec executable (#9456)
     new 9ae63d1  standard adam update for sparse tensor (#9468)
     new 49846f0  Glossary takes token_indexer and token_embedding in its constructor (#9471)
     new 55ce97c  added mac gpu install; refactored windows install (#9353)
     new 92d79f2  refactor logging in infer storage pass (#9464)
     new 0fba766  ndarray: broadcast_min/max (#403)
     new 23b5657  fix hidden node for batch norm in gluon api[] (#9487)
     new 384b736  Prep script and Fixes for single shot detector (SSD) (#9480)
     new 1100eca  host files needed by caffe converter (#9503)
     new caa6250  [Build] Remove linker flags on compilation flags causing tons of warnings when using OPENCV. (#9397)
     new 4616737  add norm operator for sparse ndarray  (#9479)
     new 20fbda6  Fix broken link (#9510)
     new df88901  [ReviewRequired] License Fixes post 1.0.0 Release (based on Voting Thread Feedback) (#9462)
     new 0319588  Adding Apache Header to .m, .cfg, R and .mk files (#9499)
     new b1d62d2  [Review Required] License Fixes based on Apache RAT Feedback Part 2 (#9482)
     new 39beaaf  remap broadcast_to/axis/axes (#406)
     new be8134d  Updated proposal to use of codeowners file (#9448)
     new 3aef4e8  [WIP]Image Augmenter (#8633)
     new 992123d  [WIP]]Vision (#8649)
     new 1997319  add stub (#8650)
     new a0cff26  [WIP][vision]fix (#8657)
     new e81f2dd  fix image_random compile (#8665)
     new 7999c43  Add Gluon data transform (#8672)
     new 2957a33  image to_tensor (#8691)
     new 6b06bf2  [Image OP] Normalize (#8731)
     new 434b016  fix (#8736)
     new 6cfd5b4  add test script
     new b6698b8  Revert "add test script"
     new 29f6055  Refactor image operators (#8761)
     new 68fc799  [Image] add random lighting (#8779)
     new a15a6e7  image flip op (#8759)
     new 9dda920  [WIP]hue (#8678)
     new fb199fc  Vision (#8856)
     new f74f6e7  fix (#8857)
     new b9569ee  add comments and sanity check (#8901)
     new 62ffb92  [Vision] add test cases for flip, normalize, to_tensor (#8919)
     new ab0d1d5  lint
     new dae6cda  add use_global_stats in nn.BatchNorm
     new cb52708  Some more license adds (#9504)
     new bfe2097  Adding License to files (#9505)
     new afd8241  Subscribe to changes (#9522)
     new fbbb4ab  example/image-classification work with python3 (#9524)
     new 25f7617  [ReviewRequired] Adding License files for MXNet Dockerfiles. (#9500)
     new 8930d96  [Review Required] Fixing Licenses: Cleaning up the Top Level LICENSE file (#9484)
     new a840ae5  Docs changes for usage (#8000)
     new ec2b699  ndarray: remap hypot (#407)
     new 20253d5  Update (#9539)
     new 323fdcb  Update (#9549)
     new a1f1321  Text APIs: interface re-design with name changes (#9534)
     new 5f9dd2c  Bugfix: kAddTo support for swap_axes. (#9541)
     new 5166e57  Data-iterator tutorial made python3 compatible. (#9460)
     new e57426f  add three pretrained english fasttext embedding (#9551)
     new 542c3d7  Fix non-MSVC build problem (#9564)
     new 9aaa29d  fix CreateOp to accept dtypes of uint8 and int32 (#9561)
     new 52219e8  Fix example  example/reinforcement-learning/a3c (#9046)
     new 1fc46e0  Broken Link Fixes (#9562)
     new b647248  fixed the rm command that was not cleaning the versioned master folder (#9566)
     new 7ee8a1f  License Adds - some more (#9559)
     new b8a832e  Adding License Header to docs folder (#9556)
     new 32463f8  [Review Required] Adding Apache License header to the perl-package (#9554)
     new 6f3f171  Add checksum for test_consistency (#9570)
     new 55163bd  fix CreateOp of altering shape operations to accept dtypes of uint8 and int32 (#9569)
     new 7fb8154  fix vision datasets doc (#9576)
     new 671df77  Reverting namespace change since this needs a major version change and breaks backward compatibility. (#9579)
     new 50850af  replaced how_to with faq (#9575)
     new f1ae225  Force the dtype of data to be float32 for test_correlation (#9581)
     new 6a4671b  Tutorials index page update (#9535)
     new f5f1b91  use nd for accuracy calculation (#9583)
     new 45366a4  Remove unnecessary RuntimeError (#9578)
     new 59627e2  fixed mail:to and the install typo (#9595)
     new ee61078  update copyright info (#9597)
     new e465cfa  Fix flaky test R (#9598)
     new f18901a  Remove the gradle wrapper. Apache RAT complains about the jar file being present in repo. (#9596)
     new c6d97d6  refactor regression ops to nnvm interface (#9540)
     new 157b088  doc: update predict section of mnist tutorial (#411)
     new 3a1093b  Update (#9607)
     new 785690c  Added in Large-Batch SGD with a warmup, and a LARS startegy. Also add… (#8918)
     new 16d995a  host test dataset for libsvmiter (#9620)
     new 5a5cd64  fixed links that were missng ndarray folder path (#9618)
     new ed823b2  proper flatten in acc (#9619)
     new 8bdc806  Language Modeling Datasets and Sampler (#9514)
     new 5e0a0b0  Fix skipping error in docstr and API docs (#9626)
     new 8205e24  Fix typo in python/mxnet/module/ doc (#9627)
     new a0a52b3  Expand gpu-kernel-launch synchronous error checking. (#9560)
     new 87ba915  Fix the version on website mainpage (#9630)
     new 8e9798c  add where op with sparse condition (#9481)
     new 9f4f533  Optimizer module overhaul (#396)
     new be39bcf  [SUBMODULE] Upgrade ps-lite to the latest version (#9389)
     new c8241cb  Update (#9639)
     new a0c8224  Update (#9638)
     new d161818  Update shape.h (#9643)
     new 60ae6d4  sgld bugfix (#9641)
     new 3fe694e  avoid per-batch blocking in metric (#9636)
     new 4957c7c  add embedding learning example (#9165)
     new 1ac6dcd  Update (#9470)
     new 134c584  Allow user to pass specific environment variables for server/worker/both through (#9049)
     new 376d47f  Minor refactor: prevent string copying, shared_ptr by ref (#8872)
     new fb8e55b  add LSTNet example (#9512)
     new 2cc2aa2  Refactor operators & MKLDNN (#8302)
     new 8c7f883  Fix OSX build, std::max template resolution with different types. (#9654)
     new 0900f64  Update (#9659)
     new 846f269  Update (#9658)
     new afdb839  Revert "Refactor operators & MKLDNN (#8302)" (#9661)
     new 4640865  Enable CPP unit tests in CI (#9609)
     new 41395c5  [CMake] Fix OSX double compilation and add static library (#9615)
     new fb8f607  Fix lint (#9660)
     new 5f336c7  Update submodules: dmlc-core and mshadow (#9670)
     new 8ee3177  typo (#9683)
     new 51776aa  Fix test_sparse_operator failures due to out-of-bounds ref. (#9674)
     new 95095e4  Set the autotune default to 1 in docs (#9653)
     new ec76003  Fix warning in sequence_last-inl.h (#9687)
     new fe5b56e  Fix DataBatch.__str__ for cases where we don't have labels. (#9645)
     new d452162  instance norm and reflection padding (#7938)
     new 2c09c26  Update README: `predict` section (#412)
     new 4c55d06  Enable dp4a for the CI (#9684)
     new 6ac37f9  kvstore: refine and copy docstring from Python (#388)
     new 5e053b8  fix NAG (#9697)
     new 921fbe9  Remove '+=' in inception-resnet-v2 (#9694)
     new 3e79aef  Use argmax instead of argmax_channel in Accuracy to keep dimention (#8245)
     new d056bfd  Fixed 4 broken links (#9698)
     new 74d5c82  Update NOTICE (#9706)
     new 00a12f3  Bump version to 1.1.0 (#9707)
     new a80ea3f  Update KEYS (#9708)
     new c19f506  Squeeze op (#9700)
     new ad3f0d8  revert acc changes (#9731)
     new 5591cb7  PGP keys add liuyizhi AT (#9728)
     new 48dcbf2  Improve take documentation. (#9717)
     new a51cdb1  Add my key (#9736)
     new 60c6434  Specify lint-versions and fix docker build issues due to nvidia-384 (#9723)
     new d4645ce  fix FileNotFoundError error (#9739)
     new 3761f2f  Fallout from simplifying/fixing commandline boolean options. (#9652)
     new 3f42582  CI environment refinements (#9741)
     new 3e93e6e  Check padding size for global pooling (#9730)
     new ccb08fb  Add convenience fluent method for squeeze op (#9734)
     new 7fb34a1  Added comment for clarity that numpy.ndarray.flatten behaves differently than nd.flatten (#9748)
     new fbc786a  update navbar model zoo link (#9749)
     new 08fc71b  Don't use FIRST_CUDA on generators which don't support toolsets (#9761)
     new 95bd0fc  Gluon PReLU, ELU, SELU, Swish (#9662)
     new b73c57c  Fix active tab on website (#9757)
     new 088623e  Improve sparse.retain performance on CPU  (#9722)
     new 56dfd8d  add metrics tests for #9640 (#9759)
     new 4a619ba  remove the extra @register (#9769)
     new 83f6279  CMake CUDA fixes + NCCL (#9672)
     new e64b7b8  Update (#9778)
     new 81dc02e  add missed optimizer docs (#9779)
     new 8a7b18f  Add C++ modules to CODEOWNERS (#9786)
     new 9b7058d  remove error output (#9785)
     new f57073e  Using the "global_pool" option to avoid an unmatching input shape of (#9783)
     new 7b24137  Better Exception Handling for Operators (#9681)
     new 45e8555  Some small typo fixes. (#9788)
     new 581cf90  add activations doc (#9789)
     new 7d4b4c0  fix markdown hyperlink syntax (#9756)
     new 4038704  build: Propagate more build flags from ENV (#414)
     new d03182f  [MX-9588] Add micro averaging strategy for F1 metric (#9777)
     new c3e3a83  Refactor operators and add MKLDNN (#9677)
     new 44ead24  Add Support for int64 (#9738)
     new f33591f  Add seed management to python tests (#9791)
     new d048615  temporary solution for instancenorm, will refactor using backend (#9807)
     new 9348a3a  make array.reshape compatible with numpy (#9790)
     new 50f326f  Use get_bz2_data in test_utils (#9776)
     new f14b2eb  fix print_summary bug and add groups of convolution (#9492)
     new af0c3b4  minor spelling tweaks for docs (#9834)
     new 8748b3e  add MUSE embeddings (#9818)
     new 28bc3c2  Ability to not build examples during the build of the cpp package. (#9832)
     new 6d2621a  Bump version to 1.2.0 (#9840)
     new 2093048  Dockerfile to create Jetson TX1 and TX2 compatible builds (#9837)
     new 5ca6efe  MobileNetV2 (#9614)
     new ed21873  fix cmake exclusion (#9850)
     new 6d2b6e3  Improve, add license to individual files (#9849)
     new e1d8c66  Enable the reporting of cross-entropy or nll loss value when training CNN network using the models defined by example/image-classification (#9805)
     new 18816a5  Add CuDNN (v6) support to Jetson builds (#9861)
     new 6ebec87  Softsign Activation Function (#9851)
     new cea8d2f  Fix broken website build, add latest version number and made build scripts fail fast (#9863)
     new 9158352  Add contrib.rand_zipfian (#9747)
     new b23e0a9  New example of custom operator using RTC (#9870)
     new 106f97f  Profiling enhancements, python API, vtune and chrome tracing objects, etc. (#8972)
     new fbbc080  Fixed indentation and added test outputs to gitignore (#9875)
     new c6a7aba  add doc for gluon contrib (#9873)
     new 5c5a904  remove MKL_EXPERIMENTAL and update make files for MKL-DNN (#9810)
     new 4bcca8c  Set worker thread to use OMP when necessary (and not to when not nece… (#9801)
     new 88f763e  remove useless code (#9886)
     new db24ac1  eye operator, for default storage type (#9770)
     new a352d1e  added function for loading content of nd_array files from a buffer (#9883)
     new 5d49205  Fix doc (#9889)
     new ada45fb  Updated to build on the new release tag found (#9896)
     new 8ee8fa2  Docs build all versions refactor (#9878)
     new dc39343  Make sure build_doc only builds target version (#9900)
     new 3545697  TVM bridge support to JIT NDArray Function by TVM (#9880)
     new 6005ad5  Update website version dropdown on master (#9901)
     new 7a0509d  fix optimizer bug in CPP-Package (#9809)
     new 34af930  add infer_type for regression ops, fix issues#9847 (#9848)
     new b8ae967  [Metric] Accelerate the calculation of F1 (#9833)
     new 9761f21  updated version to 1.1.0 (#9895)
     new 48749a5  workaround for install page display issue (#9902)
     new 17a9c6a  Using "uniform" Xavier strategy to initialize the weight for VGG network (a trial solution to issue#9866) (#9867)
     new 94d9043  bug fix (#9910)
     new f191a33  Update build status link to new CI (#9908)
     new baa277d  Fixed Android build (#9898)
     new 8940fb3  Add constant parameter (#9893)
     new d87468e  set a lower success_rate (#9927)
     new eaaf72d  save profile outputs to different file for troubleshooting purposes (#9929)
     new 632fbc8  organized installations (#9718)
     new e29b2f5  fix word_language_model for python3 (#9926)
     new f9c2689  Fix a race condition in converting data layouts in MKLDNN. (#9862)
     new 40da6cc  Removed MXNET_USE_MKLDNN definition when option USE_MKLDNN is disabled (#9923)
     new 02c3094  Update mxnet install script for OSX. homebrew/science to homebrew/core (#9924)
     new b8b869f  Add force_deterministic option for sparse embedding (#9882)
     new c7f9331  Improve workspace in convolution/deconvolution (#9904)
     new 9757091  Cleaned up image classification cpp example (#9799)
     new bce00aa  Fix R build crash in CI (#9965)
     new 62b9d48  Fix typo (#9968)
     new fc9e70b  Improve sparse adagrad update (#9651)
     new dedfd2d  sparse regression operators (#9625)
     new 8780096  Gluon image-classification example improvement (#9633)
     new 3c8e9d3  build: propagate CC/CXX into (#419)
     new 94d3c06  Fix test_cast (#9951)
     new f973fe0  revert DiagrammeR compatibility to 0.9.0 (#9975)
     new 02dd89a  Non-blocking row_sparse_pull. Fix incorrect indices generated by device kvstore.row_sparse_pull (#9887)
     new 5984a5b  host doc on s3 (#9928)
     new 7d08810  Exception handling documentation (#9869)
     new e2b1a56  [fix issue#9976] The assignment problem in NDArray (#9981)
     new 40de6ab  Add axes support to Dropout for variational dropout in NLP (#9931)
     new 29f7f43  gluon dataset download refactor (#10008)
     new 2a9c7d9  [Perl] Sparse feature. (#9988)
     new 0fe04e9  add SameType as default type inference function in imperative mode (#9942)
     new 098c9d9  add name (#10007)
     new 117c509  Fix bug for Dropout with axes, also adding unit test (#10030)
     new 39c0fd8  Fix ndarray assignment issue with basic indexing (#10022)
     new 3df1214  build: propagate USE_JEMALLOC (#422)
     new 7b4594d  Update an incorrect description in the API doc.  (#10034)
     new c162c9d  update namespace (#10036)
     new c3f7961  correct resnet link in model zoo page (#10033)
     new f04edb5  [MXNET-31] Support variable sequence length in gluon.RecurrentCell  (#9934)
     new 649b086  add axes support for dropouts in gluon (#10032)
     new ccf0469  [MXNET-67] Sync master with v1.1.0 branch (#10031)
     new 884408a  Parallelization for ROIpooling OP (#9958)
     new c153a6c  comments to copy and copyto are corrected (#10040)
     new a7ba921  Bug Fix and performance optimized for rtc (#10018)
     new 85215b6  CI docker revamp; Add Jetson, Raspberry and CentOS 7 build [MXNET-42][MXNET-43] (#9995)
     new 279ccb1  [MXNET-58]Layer Normalization in C++ (#10029)
     new 17643e7  Update (#10050)
     new bf8658b  fixed calling error while preparing data in ssd (#10024)
     new 13ae4d1  Cpu lstm inference (#9977)
     new 1794a2a  [MXNET-38]add reshape predicator function to c_predict_api (#9984)
     new 94f68fc  Fixes for profiler (#9932)
     new 0bcf334  Adding back comments to that cause nightly test to fail (#10058)
     new a9c1717  [TENSOR] Fix DLTensor conversion for int64 (#10083)
     new dd1f21b  Support single array input for metric (#9930)
     new c9ec311  [MXNET-86] Revert to pre-profile-changes copy code (#10090)
     new 1dc01ef   [MXNET-47] Documentation for build_version_doc scripts folder (#10010)
     new faa4f9f  [MXNET-83] Fix CMake build issue with MKL. (#10075)
     new f202f10  [MXNET-74]Update mkldnn to the newest & Add clang build test with mkldnn. (#9918)
     new 93e0ceb  [MXNET-34] Onnx Module to import onnx models into mxnet (#9963)
     new f7fa49e  Compile MKLDNN in Mac. (#10115)
     new 4ad37d8  Added as a separate file (#9705)
     new b87c63d  Adjust MKLDNN compile flags order to resolve some corner case (#9936)
     new 24a8b78  Fix multi worker (#10096)
     new cf188b4  implement DepthwiseConv2dBackwardFilterKernel from tensorflow codebase, which is over 5x faster (#10098)
     new 574f469  support for any datatype for correlation operator (#10125)
     new b95ae7c  [MXNET-50] Scala Inference APIs (#9678)
     new 50c694d  Update PR Template (#9919)
     new 67af47c  [Doc/Test Fix] Fix documentation of embedding layer + Fix LayerNorm Test (#10136)
     new 6e1f58d  Update (#10153)
     new 926f2e2  add deprecated profiler_set_state() (#10156)
     new bfdac0a  [MXNet-50] Set threads created by MXNetHandler to Daemon threads so they exit when the main thread exits. (#10163)
     new f2e08cf  `gluon.Block.load_params` set the default `ctx` to cpu (#10160)
     new 048e362  [MXNET-100] Fix buggy type inference in Correlation (#10135)
     new 959da84  Update (#10129)
     new c23aa9a  fixed the order of adding the new version into tag list before updating the links (#10099)
     new dd85860  [MXNET-89] Bug fix for bucketing module (#10094)
     new 1b71ce1  [MXNET-92] Support float16 in L2Normalization operator (#10078)
     new 484af32  Update profiler example code (#10120)
     new 9bbdc16  [MXNET-68] Random shuffle implementation (#10048)
     new a196c2f  add label augmentation support in image recordio (#9960)
     new 6bc2d3f  [MXNET-40]add multi proposal operator (cpu version) and fix the bug in proposal op (gpu version) (#9939)
     new 067590b  Fix broken link (#10177)
     new 91c2daa  Fix error in ImageIter example in docs (#10113)
     new b0a6760  [MXNET-99] Upgrade CI to cuda 9.1 cudnn 7 (#10108)
     new f7f73a0  [MXNET-32] add clarifications for environment variables (#9953)
     new 13e9691  [MXNet-50] Generate documentation for infer package (#10181)
     new 57534bf  [MXNET-96] Language model with Google's billion words dataset (#10025)
     new b51cfbe  [MXNET-124] Upgrade examples to use newer profiler API (#10190)
     new 8ecdc49  Update script to use new im2rec command line arguments (#10192)
     new 46e47cb  [MXNET-105] Fix CuDNN performance after code refactor (#10116)
     new 1e4bd20  Add quadratic op for the tutorial of adding ops in backend (#10091)
     new 6ceed4f  Update (#9990)
     new a2875d1  Fixes: Cosine Loss Formula (#9263)
     new 3e2fb6a  [MXNET-16] Move mshadow/ps-lite/dlpack to 3rdparty #10132 (#10138)
     new 18a2f27  Pin gcc version (#10200)
     new d283c51  [MXNET-73] Add Armv6 ci build (#10172)
     new bcb4f9a  [MXNET-130]disable cpu/gpu consistency test for Proposal OP and Multi Proposal OP (#10201)
     new 10d7458  [MXNET-101] Support float16 in LeakyReLU operator (#10169)
     new 291db45  Fix mem alignment. (#10161)
     new 09281c7  [MXNET-121] Increase onnx version being installed to v1.1.1 (#10209)
     new 5f2c1fc  [Doc] Update windows installation doc (#10206)
     new 10bfc4c  fix plugin bug (#10221)
     new ce2ce9b  Fix failed test with opencv 3.4.1 (#10197)
     new f386d91  [MXNET-53]Image classifier for scala-infer package (#10054)
     new 5c7407d  [MXNet-111] Doc for (#10155)
     new 4950f66  [MXNET-116] Inference - ObjectDetector and SSDObjectDetectorExample (#10229)
     new 200f541  fix word language model script and readme (#10225)
     new 0ca6dd2  Fix the order of `path.join` and `path.expanduser` (#10236)
     new 16930c0  move dmlc-core & make" (#10231)
     new 66c6dda  [MXNET-133] Model Quantization with Calibration (#9552)
     new ed14faa  [MXNET-138]Create CentOS gcc4.8 & MKLDNN ci-test (#10218)
     new cfc0941  [MXNET-36] Update ndarray binary ops to use kernel launch instead of mshadow operations (#9029)
     new 892494d  Avoid MKLDNN in-place data conversion in NDArray synchronously (#10215)
     new 777089d  adding context parameter to infer api- imageclassifier and objectdetector (#10252)
     new a2c4b0a  add guide for implementing sparse ops (#10081)
     new bd9b9c8  [MXNET-114] Add the ability to exclude specific lines in tutorial notebooks generated from .md (#10165)
     new 11deb67  Fix a compile error. (#10256)
     new 46b795b  remove auto value for cmake var (#10214)
     new 0a92e38  [MXNET-144] Fix test_sparse_quadratic_function test (#10259)
     new e08e1fd  Iter (#198) (#10124)
     new 13bd603  Copy only one libmkldnn file (#10245)
     new 88f2bd0  [MXNET-116] Fix OSX native package and make ImageClassifier/ObjectDetector member private (#10254)
     new f0f745d  [MXNET-72] Improve sparse.adam_update  (#10062)
     new 8c6f6bb  [MXNET-106][ONNX-MXNET] Adding ONNX Model zoo tests. (#10118)
     new 6988316  support loading parameter in partially declared gluon params (#10287)
     new 0d3ec94  add more attributes. (#10296)
     new fdb5664  [MXNET-145] Remove the dependences of and mx.initializer on the numpy's global random number generator (#10260)
     new e5f6238  Refreshed branch aggregate (#10228)
     new 0ff490b  Fix list output names typo (#10300)
     new 048f4c2  [not merge]add OPT:REF,Reduce the size of the windows binary file. (#9985)
     new e91ff1b  enhance test for leaky_relu to address fp16 precision issues (#10255)
     new 7c28089  Script to install on a python virtualenv for development (#9685)
     new d07f59d  [MXNET-132] Change dead link in cpp-package  (#10210)
     new 9488911  [MXNET-255] Temporarily disable partial test until hang is fixed (#10342)
     new c69a071  [MXNET-118] Added tutorial for Data Augmentation with Masks (#10178)
     new 4ad0bba  fix crash when profiler not enabled (#10306)
     new 72091ba  [MXNET-87] Add JIRA info to docs (#10344)
     new 3f9f871  Fix RNN doc (#10348)
     new 1bcfc46  typo in tutorials index (#10318)
     new 201ec00  [MXNET-258] Improve flaky test_random.test_shuffle (#10313)
     new b5acbcf  R Metrics (#9803)
     new 73273cf  Fix error messages (#10333)
     new 5e8e8a4  string.format doesn't support %s (#10337)
     new 453e6e9  [MXNET-240] Fix CUDA compile error (#10302)
     new d56edcf  Fix Gluon Language Model Example (#10350)
     new a3c0e0d  fix typo (#10321)
     new 0291f95  [MXNET-257] Do not copy when casting as same type (#10347)
     new 121adb6  [MXNET-247] Always build profiler (#10308)
     new 46c90ac  typo (#10265)
     new 10c6af4  [MXNET-121] Docs page for ONNX module. (#10140)
     new 5a480f0  Update doc for ndarray (#10361)
     new 9a0d002  [MXNET-141] Add tutorial Gluon Datasets and DataLoaders (#10251)
     new e208f4c  Update docs for multiple functions (#10362)
     new fe6220b  [MXNET-103] Added tutorial on types of data augmentations. (#10039)
     new b37b3f5  Replace std::swap_ranges with memcpy (#10351)
     new e0df25c  Fix MKLDNN sigmoid/softrelu issue (#10336)
     new e0df199  Update LICENSE (#10370)
     new eaa954c  improve sparse.adagrad on GPU (#10312)
     new 224b0c9  [MXNET-242][Tutorial] Fine-tuning ONNX model in Gluon (#10283)
     new c3c2676  [MXNET-146] Docs build updates: added some deps; clarified developer builds (#10270)
     new 6dd85e2  Adding a file to LICENSE (#10373)
     new 5245ef6  [MXNET-72] Improve sparse sgd on GPU (#10293)
     new 0ccf8b4  build: add `` into lib search list. (#424)
     new 62a615d  [MXNET-235] add axis support and gradient for L2norm (#9740)
     new a13c46c  [MXNET-81] Fix crash with mx.nd.ones (#10014)
     new eb94089  [MXNET-256] Add CI Test for GPU (#10346)
     new 47d0b58  remove use_fast_math (#10364)
     new a157d17  [MXNET-128] added load from buffer functions (#10261)
     new b1b0521  [MXNET-249] Add inplace support to mkldnn sum (#10315)
     new 9fee414  extend ndarray in-place reshape (#10360)
     new 779faef  [MXNET-131] Refine, misc fixes, tweak commandline, add help (#10202)
     new aa3f065  [MXNET-116] Optimized functions with batch input (#10343)
     new f0f8af1  [MXNET-266] Fix cudnn_conv and cudnn_deconv deadlock (#10392)
     new f950702  [MXNET-136] Enabling USE_DIST_KVSTORE flag for CI (#10232)
     new 44f4ebd  fix nproc arg (#10404)
     new 0243dbf  R documentation (#10408)
     new 28cb133  Fix windows setup doc using VS 2017 (#10363)
     new 6d28d0a  Add mkldnn (#10412)
     new 2cd09a0  Add support for cast storage on same stypes (#10400)
     new 56c7f02  update-dmlc-core (#10420)
     new df659bb  Fix the build status icons. These icons were not changed when we moved to the new CI. (#10418)
     new ed416be  Fix interval sampler (#10411)
     new 4342bd1  [MXNET-160] Fix for issue 9062 (#10413)
     new 45dae4b  fix docstring for EvalMetric.update_dict (#10384)
     new 153dc77  loosen overly tight rtol to control flaky behavior (#10422)
     new 73a0eb0  [MXNET-117] Sparse operator broadcast_mul/div(csr, dense) = csr (#10208)
     new 74cd220  Add CNN - character level to the examples list (#10372)
     new 78b971a  [MXNET-248] Scala Infer API docs editorial pass (#10307)
     new 2ae0bd5  Fix typos in random number docs (#10423)
     new 4c803e9  [MXNET-102] Added tutorial on how to use data augmenters. (#10055)
     new f600940  [MXNET-241] Module API for distributed training w/ row_sparse weight (#10285)
     new c3a6f19  [MXNET-166] add unknown dtype to resolve segfaults when inputType is -1 and type … (#10428)
     new 62bf9ec  Update versions of python dependencies (#10430)
     new b4161b0  [MXNET-288] Fix ssd example (#10427)
     new 582ec40  [MXNET-287] ARMv6 build with 8-10 times bigger file size (#10439)
     new 4efbbf5  [MXNET-176] Support axis=None for sum operator (#10405)
     new 590ec3f  [MXNET-143] Change MXNet-Scala namespace (#10284)
     new 4f306c9  Reduce emails (#10447)
     new 6f6a1dc  [MXNET-291] Fix an ndarray indexing bug with index -1 (#10440)
     new 4be7be7  Update (#10454)
     new 432ec20  [MXNET-292] Added social media handles (#10448)
     new 1b2805c  Updated OpenBLAS release tag for armv6 build (#10441)
     new de7269b  fix string representation of rnn layer (#10457)
     new 4fb5241  [MXNET-155] Add support for np.int32 and np.int64 as basic index types (#10434)
     new 658bad6  MXNET-295 [Perl] Added support for linalg operators. (#10460)
     new 6abcdbb  [MXNET-171] Fix a bug that was causing training accuracy to be printed as nan sometimes (#10437)
     new c24765f  [MXNET-265] Update optimizer doc to clarify wd behaviors (#10388)
     new 694bf49  use correct prop_kind for mkl-dnn FC layer (#10442)
     new cfaca3d  Fix minor math equation rendering format in NDArray API doc strings (#10444)
     new 09b880a  [MXNET-108] Adding BilinearResize2D and AdaptiveAvgPool2d operators (#9688)
     new 43f0e4d  [MXNET-212] [R] Fix Random Samplers from Uniform and Gaussian distribution in R bindings (#10450)
     new 1e532bf  [MXNET-80] Fix average pooling kernel size assignment error (#10000)
     new f9a34f8  Revert "[MXNET-265] Update optimizer doc to clarify wd behaviors (#10388)" (#10471)
     new 649a0a1  CUDNN not training on backward pass similar to pytorch (#10470)
     new 81a3827  bug fix (#10443)
     new 7d513db  Change the cuda version for docs test script (#10475)
     new 76f67b1  [MXNET-48] update on setting up Scala with MXNet and the IntelliJ IDE (#10013)
     new 0fd2969  [MXNET-303] Update Jetpack to version 3.2  (#10482)
     new a1a07b2  setting a fixed seed for module tutorial to pass Jenkins build (#10479)
     new ed7077a  [MXNET-300] Fix ndarray assignment bug with index -1 (#10468)
     new cbb7a5c  [MXNET-283] Error handling for non-positive reps of tile op (#10417)
     new 8b172d7  [MXNET-294] Fix element wise multiply for csr ndarrays (#10452)
     new c7e8047  Refreshed branch issue_187 (#10375)
     new 3023c70  [MXNET-262] Implement mx.random.seed_context to seed random number generators of a specific device context (#10367)
     new 24362d0  [MXNET-37] tutorial for distributed training (#9152)
     new 656e352  [MXNET-93] Sparse support for Custom Op (#10374)
     new 33823d3  [MXNET-106] [ONNX_MXNet] Change parameter names in imported model (#10472)
     new 6db9dd4  [MXNET-250] CSS and .htaccess fix (#10286)
     new 5f94c9a  [MXNET-264] Improve performance of MKLDNN in small batch sizes. (#10317)
     new c4258a5  Initial PR for RAT License Check (#10390)
     new e42f7e0  [MXNET-173]fix acc metric shape miss match (#10446)
     new 65ff207  update imagenet training script links (#10478)
     new dbe5c14  fix warnings from np.fromstring (#10493)
     new a8f6d64  [MXNET-286] Removed OpenMP from armv6 builds (#10469)
     new 55e7435  Remove Fermi from cmake (#10486)
     new 6c17705  updated R docs (#10473)
     new 9a8c164  [MXNET-120] Float16 support for distributed training (#10183)
     new 041b4fc  Fix rat excludes (#10499)
     new 56649e6  MXNET-308 added missing license (#10497)
     new 19a97b3  refactored example (#10484)
     new e42965c  [MXNET-298] Scala Infer API docs landing page (#10474)
     new 6660be1  Fix infer storage type (#10507)
     new 549fd64  [MXNET-306] Add slice_like operator (#10491)
     new 8097059  Minor simplifications in ci/ (#10496)
     new ee6dac7  [MXNET-305] Scala tutorial table fix (#10488)
     new ceb810c  [MXNET-311] change test needs a docker with sudo, hence image changed (#10510)
     new dd0ff7a  [MXNET-307] Add tutorials to the CI + Fix them (#10495)
     new 923eb90  Minor spelling corrections in profile (#10526)
     new 4623c6d  [MXNET-185] Improved error message (#10424)
     new 4852a99  Prevent partially update of ParameterDict (#10515)
     new b96affb  Fix output names of nn operators. (#10410)
     new fb50257  render fix (#10535)
     new b929892  add naming tutorial (#10511)
     new f220ad0  Fix typo in (#10540)
     new debe3e9  Fix bug in mx.symbol.Group (#10548)
     new 5d63151  [MXNET-318] Allow dot for fp16 on GPU (#10538)
     new 38b23ac  fix sequential (#10554)
     new e3b1daf  Fix typo (#10556)
     new 06b3809  Remove empty file from examples (#10547)
     new f3c01d5  ensure kv store is initialized when saving states (#10559)
     new fb72ee7  fix NNPACK header file position error (#10525)
     new 12cbcea  [MXNET-307] Add utility to get path (#10565)
     new 0a874ef  [MXNET-284] Update release version text to 1.2.0 (#10415)
     new 2c6d631  [MXNET-307] Add .md tutorials to .ipynb for CI integration (#10537)
     new 32ec63d  Update index.html (#10570)
     new 8cbd106  Fix gluon error message. (#10573)
     new 210e6d4  [MXNET-299] Adding the Download page and linking it to the Install page (#10477)
     new 4e32967  fix (#10576)
     new 0f1bc29  [MXNET-315] Fix the version select in the navbar to be independent of order (#10534)
     new 61a93a5  [MXNET-327] Fix docker disk-space leak by rearranging ubuntu_build_cuda #10581
     new b85d00c  [MXNET-328] Reduce jenkins timeout from 24 to 2h and add timeout to all jobs #10582
     new 62a0ab6  [MXNET-326] fix filter layout in DepthwiseConv2dBackwardFilterKernel (#10578)
     new 8ff6c15  MXNET-296 [Perl] Parity with Python Gluon Loss classes. (#10555)
     new 095921e  Add missing functions to ndarray and symbol API doc (#10577)
     new 5446900  Add "next sections" to sparse tutorials (#10551)
     new 1b0afc0  Make frontpage nav fonts bolder (#10572)
     new 7661fee  fix MobileNetV2 structure according to 1801.04381 v3 (#10603)
     new 869b156  Remove pip upgrade step from install docs (#10587)
     new 19c5212  [MXNET-293] Navigation updates (#10449)
     new 3062122  [MXNET-329] support SparseEmbedding with dense weight (#10585)
     new 162c54b  MXPredReshape support Windows (#10612)
     new 5777a16  [MXNET-292] Add reddit to social media links (#10597)
     new 615392b  [MXNET-316] Remove empty buckets causing index errors (#10528)
     new 9ccd647  standard sgd_update (#10614)
     new 6fb7c85  [MXNET-338] Fix symbol boolean evaluation (#10618)
     new ba40121  Change Model url to S3 (#10617)
     new 6df1979  [MXNET-322] Elemwise_mul(row_sparse, dense) = row_sparse on CPU (#10557)
     new 73d879c  [MXNET-320] Support elemwise_add/sub between dense and csr tensors (#10550)
     new 057885d  MXNET-335 [Perl] Support for image operations on symbols and ndarrays. (#10620)
     new 6c3e1a4  [MXNET-307] Add tutorial tests to the CI (#10608)
     new 1ed94d9  [MXNET-330] fix cached op req (#10590)
     new daf821c  logical not operator and unit test (#10639)
     new e40ed88  Stop ignoring the given name of CompositeEvalMetric (#10635)
     new 0bcc2da  [WIP] [MXNET-334] fix dtype of Embedding and add dtype argument to Dense (#10615)
     new dc3ea81  [MXNET-313] Python: Resolve some of the undefined names (#10490)
     new 81d0113  Fix for the case where there are no detections (#9784)
     new f75138f  fix SSDClassifierExample (#10654)
     new 274fae8  fix race condition in csr cast_storage block kernel (#10643)
     new 007c0c5  [MXNET-307] Update train/sparse seed (#10659)
     new b847c6f  attrs key should be 'attrs' not 'attr' (#10671)
     new a8e1912  Fix typo in __setitem__ doc (#10657)
     new 8011f3b  New tutorial on how to create a new custom layer in Gluon (#10607)
     new fd55a2e  [MXNET-33] SSD example not working with mkl-dnn (#10021)
     new faffe56  move parameters to backend for hybridblock (#10632)
     new 7989782  [MXNET-304][RFC] Jenkins docs build (#10485)
     new d2d7d65  Update (#10681)
     new 4bf3d01  [MXNET-307] Extend tutorial instructions (#10658)
     new 16d8e16  [MX-307] Upgrade timeout to 10 minutes per test (#10682)
     new 9c601b0  update doc for grad in (#10683)
     new 50c8d6a  Update (#10680)
     new efe5314  hard sigmoid together with unit test (#10633)
     new 663df0f  Use default nvcc when configured nvcc not present. (#10566)
     new 0cd082b  fix for asymmetric padding (#10676)
     new a5d7893  invalidate outputs for imperative.
     new 860bfa8  Invalidate all MKLDNN memory for reused NDArray.
     new b428937  Revert "invalidate outputs for imperative."
     new 8b91bce  remove mkldnn memory invalidation in graph execution.
     new 8727cae  [MXNET-263] Support for dot(dns, csr) = dns and dot(dns, csr.T) = dns on GPU (#10371)
     new b0649f6  fix multi context hybrid
     new f3f2733  handle NDArray slice properly for mkldnn layout
     new f236687  add unittest for gluon mkldnn ndarray slice computation
     new 301ea63  Update mobilenetv2 symbol definition (#10595)
     new 0a664ef  test inference multiple times.
     new 76a1029  Fix a bug in GetMKLDNNData().
     new d62f138  Update comments.
     new a1cb6a6  Handle all cases for GetMKLDNNDataReorder
     new 3713ead  avoid unnecessary message.
     new 29cd215  Add C++ unit test for NDArray.
     new deeeea4  Fix a minor bug.
     new 7d6e8e0  Unit tests on GetMKLDNNDataReorder.
     new 9d1b9dc  Fix lint error.
     new 2bd6691  Add more test cases.
     new 99bd17f  add comments for the test code.
     new 9e90090  Reorganize test code.
     new 07819ad  test in jenkins.
     new 6085831  Fix jenkins.
     new e554999  Fix Jenkins.
     new 652a895  test jenkins.
     new 1d33fc7  test Jenkins.
     new 148a50a  Revert "[MXNET-351] Fix a bug in the MKLDNN integration." (#10718)
     new f76f65e  fix rnn layer repr
     new 32aff9b  Update
     new 3ace09b  invalidate MKLDNN memory for reused NDArrays. (#10706)
     new 01e305b  make the slice of sequential a sequential (#10630)
     new 6c917df  Fix for cmake 3.11 (#10711)
     new 82c33f6  fixed broken links and added jenkins info (#10728)
     new 97daed8  [MXNET-363] fix race condition in sparse dot(csr.T, dense) on gpu (#10713)
     new 2e40a16  Update (#10720)
     new f0d2776  [MXNET-362] ensure same mkldnn engine is used for consistency (#10616)
     new 92c373b  moving assets to faq; redirect should handle old links (#10729)
     new 23d933b  Add cuda version check to skip building quantization ops for versions less than 8.0 (#10710)
     new c090f95  add FResourceRequestEx interface (#10649)
     new 6bef58a  Profile doc (#10732)
     new 4044eb4  Add gpg keys (#10701)
     new 6d7c39e  Update mshadow submodule (#10730)
     new 36f3422  add warning on input data for RNN for layout NTC
     new 04e00c8  Update
     new 8fdcb85  Update
     new 4d7505a  Correctly set the device context to CPU at symbolic mode when --gpus not set (#10646)
     new 50cb694  Sync the mshadow submodule (#10751)
     new 1ab0733  Use quasiquote replacement (#10707)
     new ecdff56  Use Jetson specific config in build instructions. (#10758)
     new 147c83a  [MXNET-371] Fixing the broken links from the failure list
     new 9f8f042  [MXNET-358] support dense weight & sparse grad for adam, sgd and sgd_momentum (#10664)
     new 61f86fc  [MXNET-347] Logical Operators AND, XOR, OR (#10679)
     new 8585ef5  Improve resnet18 and resnet34 pre-trained models (#10752)
     new 1f42f2c  Add more gluon computation on MKLDNN with memory operation(slice, reshape etc.) (#10764)
     new 97da5e3  API calls are not consistent in the example (#10763)
     new defb8fe  add mobilenet v2 1.0 pretrained model (#10774)
     new 65df1ee  fix flaky test for hard_sigmoid (#10759)
     new 1822dac  [MXNET-372] Add build flag for USE_F16C in CMake and clarify flag in make (#10760)
     new 23934cf  Use numpy.arange in RandomSampler (#10768)
     new ebd8a6b  update perf. (#10761)
     new 1420697  [MXNET-359] fix checks on convolution parameters in MKLDNN. (#10666)
     new 9dce757  Update (#10685)
     new 20bbef5  Fix learning rate of ctc example (#10790)
     new 66b2944  add reverse option to ndarray inplace reshape (#10767)
     new 4ba436b  Fix a bug in getting MKLDNN memory (#10731)
     new 4e68b3b  Added more information to API Docs for Python and Gluon. (#10622)
     new 7395be8  Fix row_sparse_pull with single gpu (#10772)
     new 3c7afcc  [MXNET-367] update mkldnn to v0.14 and disable building test examples (#10736)
     new 66365ef  Revert "[MXNET-367] update mkldnn to v0.14 and disable building test examples (#10736)" (#10808)
     new 38ec93c  fix topk nms in multibox_detection operator (#10794)
     new aabaab4  Update (#10800)
     new e49fdae  Fix a mem error. (#10812)
     new ab6a25e  Reorder2Default: return directly for default format (#10810)
     new 3bba4c8  fix (#10814)
     new 97511ba  [MXNET-340] Updated tutorials page. (#10621)
     new 602523e  [MXNET-354] Support elemwise_add/sub between dense and row sparse tensors on CPU/GPU (#10645)
     new 1968bba  fix creating cpu sparse array from gpu data (#10830)
     new cfd62e0  Fix sign bug in spatial transformer interpolation. (#10741)
     new aea0fc3  [WIP] fix csv iter (#10829)
     new 2e935b9  fix the latex formula for LRN operator (#10802)
     new 31016c4  [MXNET-343]fix Mkldnn with msvc (#10629)
     new e17869c  [MXNET-314] Support Integer Type parsing in CSVIter (#10533)
     new 6cfc614  Fix quantized pooling doc (#10841)
     new 45bcc1c  clarify docs build (#10785)
     new 7464e33  hide kvstore api from api index (#10806)
     new b64b9a0  Finish prerequisites on MXNetTutorialTemplate (#10851)
     new e573097  fix a bug in (#10843)
     new 0fbbe19  [MXNET-367] update mkldnn to v0.14 and disable building test examples (#10819)
     new b2ec05b  [MXNET-408] [WIP] inplace ReLU activation (#10847)
     new 4aff7de  [MXNET-409] Temporary fix for ARM64 builds - switch to own dockerhub (#10849)
     new 33c58dc  fix gradient number and document for box_iou operator (#10823)
     new 25bede9  consolidated contribute and community info (#10855)
     new 0d5560d  fix thread contention caused by openmp (#10820)
     new 3ae01e4  Fix mxnet::Context hash for 32 bit architectures, where size_t is 32 bit. (#10746)
     new f02caae  update mshadow (#10853)
     new c5958f0  Add Util Function for Memory Plan Inspection (#10859)
     new 333e7fe  [MXNET-407] Better error handling of NDArray setitem autograd (#10844)
     new 36f8477  Adding max_area to random_size_crop (#10825)
     new 61c00ee  Op for scaled attention (#10795)
     new 8090056  Fix for cpp examples in mxnet/cpp-package. (#10726)
     new c3242f7  Update expected result in osx python install script (#10842)
     new 5088ca9  add mobilenetv2 pretrained models (#10879)
     new 3972861  support broadcast_add/sub between csr and 1D dense vector (#10714)
     new 1014c94  prevent setting grad_req for Constant (#10869)
     new 26c30af  Fixed divison by zero bug in DistanceWeightedSampling in gluon example (#10857)
     new 35fb9ea  fix docs for new broadcast operators (#10887)
     new 4967feb  [MXNET-412] Loss update performance (#10892)
     new 5858a4f  [MXNET-365] handle inplace in mkldnn FallBackCompute (#10591)
     new 645b4c5  Pin scipy to 1.0.1 (#10902)
     new 74f104a  [MXNET-329] deprecate contrib.SparseEmbedding (#10877)
     new 86ee3e1  Change class variables to thread local variables (#10833)
     new ca08c1a  Added nullptr check for LeakyRelu gamma parameter. (#10886)
     new 5569cc4  [MXNET-406] support init/pull dense weight, push row_sparse grad in kvstore (#10845)
     new e4eee74  Fix thread local. (#10888)
     new 0fcf191  Fix eye doc (#10908)
     new ce57b58  fix a bug in cudnn softmax activation. (#10918)
     new 8b53a3d  support L2Norm backward for GPU. (#10874)
     new b011ecc  [MXNET-357] New Scala API Design (Symbol) (#10660)
     new 1143e78  Fix typo "weights" (#10934)
     new 89ffab9  fix example in smooth l1 document (#10929)
     new ab232c1  split trainer.step into allreduce and update (#10861)
     new 98c5b38  Fix flakeyness of (#10891)
     new 275378a  [WIP][MXNET-107] Fused LSTM implementation for CPU (#10104)
     new 0fb57ff  fix symbolblock save_params (#10748)
     new 0442954  fix doc (#10943)
     new a572aa4  Fix _backward_norm op registration (#10913)
     new e103ea5  Correction of gluon documentation in (#10944)
     new 0805e4d  [MXNET-9704] An assertion check for invalid layout (#10915)
     new 57c8ca1  Fix engine stop/start  (#10911)
     new db4de44  Update performance page. (#10876)
     new 1214205  Expose the number of GPUs. (#10354)
     new 65061dc  fix rnn (#10954)
     new bae1332  handle fallback correctly for write inplace when the array is MKLDNN. (#10651)
     new b2ccd34  [Sparse-Gluon] embedding with sparse grad (#10924)
     new 8ebc050  [MXNET-416] Add docker cache (#10917)
     new b845147  Fix test_sparse_mathematical_core sensitivity to scipy v1.1 (#10961)
     new 7641759  [MXNET-309] [ONNX-MXNet] Model Metadata API (#10512)
     new 632f514  Sampling negative samples other classes only (#10980)
     new c4144f2  [MXNET-307] Fix flaky tutorial tests from CI (#10956)
     new 33226ab  Remove kvstore calls from FM example (#10946)
     new ed846c9  Minor doc fixes (#10963)
     new 034f24f  Add Apachev2 License for contrib (#10938)
     new 48d6090  removed script tags that are being interpreted as html (#10978)
     new 3677b6f  [MXNET-343] Avoid importing docker_cache if feature is not used. (#10975)
     new 8937488  Fix optimizer pickle (#10983)
     new 6abd654  [MXNET-423] Gluon Model Zoo Pre Trained Model tutorial (#10959)
     new 2567c90  Add Windows MKLDNN Building Instruction (#10613)
     new 5cddc2d  fix rnn layer kernel forward (#10982)
     new f0be910  [MXNET-375] Lp Pooling and Global Lp Pooling (#10780)
     new bea5fd1  unlink memory shared file immediately on linux (#10992)
     new 10ac529  Only allocate cudnn-rnn dropout memory if dropout p > 0 and acquire descriptors during initialization (#11004)
     new ac49ddb  [MXNET-439] Fix DQN example (#11014)
     new 05f9429  Removing rc from the release text in the home page (#11010)
     new b746632  speed up of topk-operator (#10997)
     new 38cc9c0  added ability to set search path for Accelerate library (#10481)
     new 0185f87  Fix Python error on --model-prefix parameter used and make the param can be periodically saved (#10922)
     new 022f238  add gluon model summary (#10989)
     new 85668cc  Add live object detection from camera device. (#9808)
     new 98e671e  [MXNET-411] Add ROI Align (#10852)
     new ab08ee8  added more detailed error messages for potri/potrf (#11035)
     new f7e5e91  add cpu_pinned in __init__ (#11030)
     new b0d632f  [MXNET-357] New Scala API Design (NDArray) (#10787)
     new 30ca4e3  NEWS and README update to master (#11017)
     new 704d218  move exec.reshape to backend (#10882)
     new 0d09449  Update dmlc-core (#11034)
     new 3f346aa  [MXNET-414] Tutorial on visualizing CNN decisions using Grad-CAM (#10900)
     new 166944c  SSD performance optimization and benchmark script (#10483)
     new 16216a8  Added Module vs Gluon API FAQ (#11038)
     new d497b37  Fix bugs in MKLDNN. (#10979)
     new feb0757  [MXNET-413] Fixing the broken links - Week of 5/7 (#10895)
     new 1ac2bce  [MXNET-386] NDArray Bug fix (#11063)
     new d528453  [MXNET-473] Fix for dist_sync_kvstore (#11058)
     new 79bcd9c  bump up version number to 1.3.0 to make nightly build to build with right source (#11070)
     new 5153370  [MXNET-471] Add Builder class for Scala Module and DataBatch to simplify construction (#11045)
     new 0df58ae  Added missing links. (#11067)
     new 63e43a3  Update (#11081)
     new 482e50b  [MXNET-374] handle row_sparse weight in parameter and trainer (#11001)
     new 9ab0d2b  Update documentation (#11082)
     new 4ac76c8  Support for axis parameter in linalg.gemm (#10864)
     new 805a71a  [MXNET-491] Use depthwise convolution by cuDNNv7 if available, updated version (#11076)
     new 33de266  fix parameters name inconsistent for Proposal OP and Multi Proposal OP (#10242)
     new 9514a1e  Fixing the xml markup (#11068)
     new 92286c9  Test/mkl dnn act (#11026)
     new 714e296  [MXNET-470] Gluon Style Transfer Example (#11044)
     new badd282  Update Makefile (#11092)
     new 005f677 and fix visualization in (#11051)
     new 5109b00  added ravel/unravel operators (#11025)
     new 9feecce  [MXNET-399] Elemwise_mul between dense and csr on CPU & GPU (#10894)
     new 8be4b8e  fix dot(csr.T, dns)=dns can't be called on cpu and gpu (#11087)
     new 2dbd143  [WIP] Do Not Merge. Static memory allocation for cached_op (#10817)
     new d6f9cd8  migrating website build to secure nodes (#11066)
     new 9854583  [MXNET-386] API Docs Generation (#10991)
     new fe153ac  Updating readme to fix broken docs status (#11102)
     new a27b52e  [MXNET-424] dtype option for multinomial (#10970)
     new 4554c76  [MXNET-477] CI engine test bug fix test (#11065)
     new 7eb78d8  update scala api to properly track reshaped size (#11009)
     new a068fae  make slice operator 20x faster on GPU (#11124)
     new 39b6bba  [MXNET-502] Fixing broken feature_extract cpp example (#11114)
     new 2a63114  [3rdparty] introduce tvm (#11130)
     new f107397  [MXNET-454] Move distributed Docker cache from S3 to Docker Hub (#11018)
     new b93f90f  fix doc of sym.RNN (#11135)
     new 21997a2  allow int shape in parameter (#11104)
     new f754498  [MXNET-310] [ONNX-MXNet] API to import ONNX models into Gluon. (#10605)
     new 106b040  [ONNX] Added Unsqueeze operator import support (#11106)
     new 247d579  don't set device for pinned mem (#11146)
     new 5dde19f  [MXNET-356] Fix for the search results issues (#10656)
     new a87de58  [MXNET-506] Fix missing titles in certain pages (#11132)
     new d552640  [MXNET-503] Website landing page for MMS (#11037)
     new 776b239  Make inner transform activation configurable for LSTMCell (#10957)
     new 24804e8  Revert "[MXNET-503] Website landing page for MMS (#11037)" (#11154)
     new 318c689  fix shared_storage free (#11159)
     new aafbd17  add input argument in warpctc layer (#11167)
     new 069026a  [MXNET-107]Fused GRU implementation for CPU (#10311)
     new 7511045  Removing tutorial tests (#11170)
     new 3f30b44  minor fixes to example/ssd (#11138)
     new 6fe9387  Add tqdm to support The Straight Dope CI (#11153)
     new 2e10857  [MXNET-501] Navbar community fix (#11105)
     new 6ef7a0f  [MXNET-433] Tutorial on saving and loading gluon models (#11002)
     new c83b3ec  [MXNET-115] USE_LAPACK is forced on all platforms with OpenBLAS and cmake (#11094)
     new afe0e49  Add frame-resize option to ssd demo to scale camera input (#11033)
     new 74b784e  Fix ci/ using remote cache locally (#11169)
     new 794074d  [MXNET-504] Add version select + queryString capabilities + C++ instructions to install page (#11128)
     new 0e5b79f  [MXNET-289] Fix bugs in image classification example  (#10435)
     new 975d249  Add valid_thresh to contrib.box_nms (#11162)
     new 6e4a50f  [MXNET-472] ccache for docker builds (#11151)
     new b434b8e  [MXNET-525] Add retry logic to download functions to fix flaky tests (#11181)
     new 2a45640  Enable CUDNN for conv1D (#11194)
     new 462c3c3  Fixes for CI #11214 (#11217)
     new 935fc55  Support for dot(dns, csr) = dns and dot(dns, csr.T) = dns on CPU (#11113)
     new 04eb7f1  [MXNET-394] concat of CSR NDArrays on first dimension (#11024)
     new 3db6307  [MXNET-420] broadcast_mul/div between csr and 1D dense on GPU (#10939)
     new 3eada3b  fix propagation of cpu shared context, issue #11160 (#11182)
     new 0dbba84  Replace the old adhoc method to iterate over gpu devices with new mx.context.num_gpus (#11227)
     new 715457d  [WIP] Gluon sparse block and sparse embedding (#11197)
     new 14275a5  [MXNET-530] Remove install page artifacts (#11191)
     new b0f4bbb  Support for data iterators returning lists of batches (#11112)
     new ed80ff2  [MXNET-62] add test against spark integration (#10462)
     new 4da51b2  Fix a bug in sparse embedding operator (#11231)
     new 0254f9e  Fix #11214 ccache compilation (#11244)
     new 2d761c9  fix loading params if ignore_extra is set (#11235)
     new 4818103  [MXNET-344] [ONNX-MXNet] Add new Operator Translations for ONNX import module (#11140)
     new 74479b8  [MXNET-386] ongoing maintenance on NDArray (#11126)
     new fc102c3  Fix when CCACHE_DIR is set. (#11226)
     new c54f9f7  [MXNET-534] Add examples to example list (#11225)
     new bf26886  gpu mem pool strategy (#11041)
     new eb95d7b  [MXNET-543] disable scalatest on Spark (#11264)
     new d79e1ad  [MXNET-290] MKLDNN support for model quantization (#10433)
     new 66ab27e  add import_ for SymbolBlock (#11127)
     new bbc7a22  Scala inference memory leak fix (#11204)
     new 60d7090  [MXNET-521] Add Facebook open-graph tag integration (#11155)
     new a40055e  [MXNET-546] Add unit test for MKLDNNSum (#11272)
     new ee1a10f  Improve hybridblock doc (#11236)
     new 4a81433  disable testcase test_gru_bidirectional temporarily (#11273)
     new 496b0f2  [MXNET-532] Clarify documentation of save_parameters(), load_parameters() (#11210)
     new 2eadfa2  Enable shared ccache and fix test (#11269)
     new b819fd9  Improve data transform for gluon data loader (#11183)
     new 17803b6  [ARM] improvements to ARMv7 based builds. (#11245)
     new a7952f0  [MXNET-538] Add XUnit to python tests (#11199)
     new feef067  [BUILD] Remove useless additional link opt (#11300)
     new ebf3777  [MXNET-57] Add android64 build (#11188) (#11055)
     new fb084cc  leaky relu speed (#11012)
     new 258e96d  [MXNET-541] Correcting the new install pages bugs (#11261)
     new e48a8fd  Revert "[WIP] Do Not Merge. Static memory allocation for cached_op (#10817)" (#11311)
     new 02e8a71  [MXNET-539] Allow Scala users to specify data/label names for NDArrayIter (#11256)
     new 5431e12  Static alloc for hybridblock (#11313)
     new 9b8eb56  Revert "Static alloc for hybridblock (#11313)" (#11318)
     new 7210c2c  bump up rtol for fp16 case of test_sgd (#11246)
     new 11dd56a  [MXNET-352] Document behavior of mx.initializer.Constant (#10637)
     new ef3169a  [MXNET-554] Add ccache reporting to CI (#11322)
     new 7975265  [MXNET-540] update python gpu build from source instructions on windows (#11276)
     new 16a06b6   Fix gperftools/jemalloc and lapack warning bug. (#11110)
     new 0910450  allow user to define unknown token symbol (#10461)
     new 92fde19  [MXNET-542] Fix mkldnn performance regression + improve test logging (#11262)
     new 0ac9f8b  Bring back MXNET_GPU_COPY_NTHREADS env variable (#11237)
     new 6b5d9f9  Fix the inaccruate instruction on installing nose (#11317)
     new f34dafe  [MXNET-380] count_include_pad argument for Avg Pooling (#11021)
     new c2714c0  MXNET_FORCE_ADDTAKEGRAD to disable AddTakeGradLargeBatchCaller (#11316)
     new 9226d91  [MXNET-514] Add clip_global_norm(row_sparse_grad). Fix Fix trainer init_kvstore (#11266)
     new 00681c3  [MXNET-547] Tutorial explaining how to use the profiler	 (#11274)
     new 47e2b89  [MXNET-553] Restructure dockcross dockerfiles to fix caching (#11302)
     new ccee176  Add standard ResNet data augmentation for ImageRecordIter (#11027)
     new 6307c00  Fix axis Bug in MKLDNN Softmax (#11335)
     new 12cb4b9  [MXNET-498] Test MKLDNN backward operators  (#11232)
     new a7ea976  Static alloc for hybridblock (#11320)
     new 7aa8a0c  Fix save load doc (#11345)
     new ba9784d  [MXNET-404] elemwise_add/sub between rsp and rsp on GPU (#11179)
     new 9ffc03c  Add seed_aug parameter for ImageRecordItr to fix random seed for default augmentation (#11247)
     new 7a83a75  [MXNET-555] Add subgraph storage type inference to CachedOp  (#11306)
     new bb105a1  [MXNET-405] Add 2 new pipelines to the Official CI and run nightly tests.  (#10827)
     new afc1496  [WIP]Update resnet pretrained models (#11327)
     new d6813ef  Fix typo in lrn (#11347)
     new 529ee85  [MXNET-561] Add test coverage reporting (#11344)
     new 5550c0a  Fix issue of test_gru_bidirectional #11219 and add robust code (#11333)
     new 579e376  [MXNET-558] Fix 'AttributeError: '_thread._local' object has no attribute 'value'' on distributed processing applications (#11332)
     new 10100b8  Removing the stale links from docs/ (#11372)
     new adec280  flaky test disable test_ImageRecordIter_seed_augmentation temporarily (#11381)
     new 9b27262  Enable support for dense weight and sparse grad Adagrad updates (#11355)
     new 7d91602  [MXNET-533] MXNet-ONNX export (#11213)
     new 225f71f  [MXNET-538] Add XUnit test result publishing to windows (#11348)
     new cdb01fc  Don't fail storing test results if test suite got aborted (#11363) (#11391)
     new 619e4bd  Fix #11353 (#11360)
     new 72fa6a9  fix recordfile dataset with multi worker (#11370)
     new 21ff36b  Fix flaky test test_operator.test_binary_op due to numerical errors (#11259)
     new ed7e360  [MXNET-349] Histogram Operator (#10931)
     new e494cee  [MXNET-593] Adding 2 tutorials on Learning Rate Schedules (#11296)
     new 0538ad9  add vRNN and dropout (#11399)
     new e4bf646  [MXNET-551] Test CreateMKLDNNMem/CommitOutput (#11308)
     new 4db5a83  Fix bi-lstm-crf to update crf weights (#11375)
     new ced119b  Fix flaky test test_operator_gpu.test_batchnorm_with_type (#11396)
     new 99af59b  [MXNET-139] Tutorial for mixed precision training with float16 (#10391)
     new 7052720  [MXNET-559] Scripts for running the  Broken link checker job (#11340)
     new a4f4e76  Define build target for mkldnn lib build to fix 'make clean USE_MKMLDNN=1' issue (#11090)
     new 2885370  [MXNET-601] Fix caffe converter test (#11425)
     new ccc1237  Add license (#11433)
     new f848b3e  Update email notifications (#11432)
     new e8f52b3  [Nightly Tests] Sync with latest Dockerfile changes and Enable KV store test (#11358)
     new 8fde4b7  Fix symbolic example (#11428)
     new 645057b  Disable flaky tests (#11440)
     new a97a475  [MXNET-23] add README for test directory (#11390)
     new 4d3edd3  changes to rat excludes post 1.2.1 (#11454)
     new 2fe5144  [MXNET-602] [MXNET-603] Disabling a number of tests to improve CI stability. (#11422)
     new 515404a  [MXNET-605] Hotfix for USE_BLAS=mkl compile bug (#11449)
     new b73b7b8  [MXNET-606] Ppa retries in CI (#11461)
     new 1b69e4b  [MXNET-561] Retry submission of coverage report on 4xx, 5xxs (#11460)
     new a7d858d  lower learning rate (#11467)
     new dc291d3  Enable test_autograd_save_memory (#11437)
     new 52ce0c3  Replace PTB dataset to Sherlock Holmes  (#11435)
     new c6c04d4  Fix #9743 (#11426)
     new f7a0025  Fix flaky test test_operator_gpu.test_spatial_transformer_with_type (#7645) (#11444)
     new 4cdc150  fix. (#11471)
     new 3ff7bf7  Update (#11447)
     new 044afa3  [MXNET-312] Added Matthew's Correlation Coefficient to metrics (#10524)
     new 2594fca  Document AdaGrad eps as initial history accumulator value (#11392)
     new 36f2aae  support shared parameter in summary (#11379)
     new b344d89  fix record io augmentation speed (#11474)
     new f02e627  fix regex which processes RAT check output (#11476)
     new a4054cd  [MXNET-607] Fix the broken reported by the new BLC (#11465)
     new a0669d0  [Nightly Tests] Disable some nightly tests for installation guide (#11452)
     new 1d9096f  Update CODEOWNERS (#11416)
     new 7ce746f  Disable flaky test test_sequence_last (#11485)
     new 2509e0c  [MXNET-610] Disable flaky test test_sample_multinomial (#11488)
     new c84a2f6  [MXNET-564] Fix the Flaky test on arange (#11377)
     new 4511086  [MXNET-603] Disable flaky test test_hybrid_static_memory (#11486)
     new e4575ea  Fix typo in Dockerfile description. (#11354)
     new 48fbc72  Enable test_conv (#11490)
     new 39ca251  fix flaky test_sparse_dot and incorrect handling of empty sparse tensors (#11389)
     new ca60b94  [MXNET-537] add_n(dense, csr, dense) = dense and add_n([dense, csr, rsp]*, dense, [dense, csr, rsp]*) = dense on CPU & GPU (#11330)
     new af4a600  [MXNET-486] Create CPP test for concat MKLDNN operator (#11371)
     new fdf2283  [MXNET-612] Optimize RAT setup and move it into PR stage (#11492)
     new c74a8e1  Make MXNetError in Scala Not Private (#11491)
     new 275cd8e  update docs: deps using CI scripts and other clarifications (#11431)
     new 3a62200  support int64 data type in CSVIter (#11446)
     new ebfc16e  test for new int64 type in CSVIter (#11499)
     new e892301  [MXNET-517] add sample ratio for ROI Align (#11145)
     new 355de66  [MXNET-548] fixed path for auto_module_index.js (#11287)
     new f783a66  [MXNET-379] L1 Normalization (#11229)
     new 33022f8  [MXNET-382] Shape and Size Operator (#10889)
     new cca0883  MXNET-336 [Perl] Major Gluon update for Perl API. (#11414)
     new 806b41b  Update large word language model example (#11405)
     new 9a0099f  [MXNET-595] Install page not loading properly in Internet Explorer (#11404)
     new 7e5d88d  Disable flaky test test_conv in gluon (#11507)
     new b22ce0b  Disable test_norm (#11510)
     new 45a1040  Removing the test, as behavior varies with system arch (Intel vs AMD) (#11463)
     new 92900f0  rnn_cell little bug fixed (#11003)
     new 12709ac  new NER example: MXNET-321 (#10514)
     new 06a18ab  [MXNET-615] Fix for flaky test_svmoutput_with_type (#11505)
     new 8917318  Clojure Contrib Package (#11205)
     new 37c8165  Add licenses to clojure package (#11512)
     new 5627bc3  [MXNET-531] MNIST Examples for Scala new API (#11250)
     new cba2b03  add test for #11003 (#11511)
     new c5035b6  batchnorm falls back for sparse arrays. (#11473)
     new 568cbaf  Re-enable kvstore test  (#11519)
     new 84b9b41  Disable test_pooling_versions (#11518)
     new cb4d310  Add redundant R key servers. (#11489)
     new d34452b  Add fallback for codecov (#11524)
     new 6bdd922  Fix #11504. Division by zero in callback with very fast speed (#11521)
     new 555dc30  [MXNET-57] Android ARMv7 support (#11382)
     new d6a8ca7  handle the case that inputs and outputs of a graph share NDArrays (#11436)
     new 99ac2f5  Compile with LAPACK shared library (#11483)
     new 490e951  Fix flaky test 10380 (#11529)
     new defd544  Replace save_params and load_params with save_parameters and load_parameters (#11527)
     new 7c74d1f  [MXNET-210]give warning for variables with same name in graph visualization (#10429)
     new 030fbc3  [MXNET-432] Add Foreach (#11531)
     new 0ae38c7  Update (#11532)
     new 34fafb4  Fix Flaky test test_multi_proposal_op as in #10186 (#11530)
     new e94146f  Move the auto-generation for the ndarray and symbol files to compile time (AOT) (#11534)
     new 552c715  updating installation info to have latest packages and more clarity (#11503)
     new 5df5f2e  Update (#11540)
     new 40ae851  [MXNET-617] - Improve Clojure styles with cljfmt tools revise (#11537)
     new 815eb60  [MXNET-319] Javadoc fix (#11239)
     new 01f1457  [MXNET-565]change wget into java download (#11383)
     new 92f0c51  [MXNET-531] GAN MNIST Examples for Scala new API (#11547)
     new 9e681f2  [MXNET-531] Maven demo project (#11451)
     new eb4b66c  Profiler is already enabled by default in 1.2.0 (#11558)
     new e870890  [MXNET-594] Added Learning Rate Finder tutorial (#11304)
     new 19ac41d  Manually check node existence in CachedOp (#11545)
     new a4c8d97  Updating Scala IntelliJ tutorial & installation instructions (#11152)
     new 1ca6622  Fix typo in clojure project dependencies. (#11582)
     new 52dcb19  [MXNET-392] Support for a bunch of unary functions for csr matrices (#11559)
     new ef820c1  [MXNET-613] Upgrade jetpack to version 3.2.1 (#11494)
     new fc2acb0  Add path_suffixes for finding cudnn.lib (#11513)
     new dd954b4  fix flaky test (ran 50000 times locally) (#11595)
     new 9652b39  Fix scala maven repos (#11604)
     new c922547  Add CODEOWNERS (#11598)
     new 6924390  CMake 3.12 will not detect a variable as empty string (#11584)
     new 0d5ebe1  [MXNET-497] fix bugs in MKLDNN operators to handle the kAddTo request (#11129)
     new 459a891  ColorNormalizeAug now accepts mxnet NDArrays as well as stated (wrongly) in the documentation. (#11606)
     new 266de6b  Improve sparse pull performance for gluon trainer (#11429)
     new 53e8afd  Disable flaky test: gluon test_export (#11617)
     new 5ff9b92  fix nccl compilation (#11623)
     new 1e698d5  fix xml display bug (#11624)
     new 0bec94b  [MXNET-620]Fix flaky test batchnorm training (#11544)
     new 7a6e8de  [MXNET-624] Clojure Package Api Docs for the Website (#11574)
     new c792638  Fix flaky test:test_sample_multinomial (#11366)
     new 43ad56c  Add stable nrm2 Reducer (#11573)
     new d814733  Added query for cuDNN BN min. epsilon. Enabled choice of BN impl. for ONNX import (#11380)
     new f665c13  [MXNET-652] Allow for multi-context input to load gluon models from onnx (#11637)
     new 4ae0cd9  Fix for flaky keyserver Fix #11601 (#11602)
     new b264f6f  Fix flaky test test_gluon.test_hybrid_static_memory_switching (#11577)
     new db5d350  Adding models to example list and ONNX API page (#11632)
     new 5798b6a  [example/sparse/factorization_machine/]add auc and fix the typo in re… (#10594)
     new 7dbbe59  Fix InferStorage for sparse fallback in FullyConnected (#11498)
     new d25fdb7  [MXNET-670] shape_array and size_array operator is non-differentiable (#11661)
     new ea24c1c  [MXNET-627] Fix Installation instructions for R bindings on Linux systems. (#11590)
     new b786ead  [MXNET-531] Integration Test for Scala (#11596)
     new 2dcb35e  Modernize Python 2 code to get ready for Python 3 (#11648)
     new 8240d3a  [CI] cleanup and script refinements (#11600)
     new e912314  New style exception for Python 3 (#11668)
     new 2fd7085  documentation enhancement for optimizers (#11657)
     new 5b63144  Fix build issue with USE_CUDNN=0 (#11470)
     new c766712  update rcnn example (#11373)
     new 6b2cbd4  xrange(10) --> range(10) for Python 3 (#11673)
     new 92366a9  Update (#11677)
     new 32d298b  [MXNET-323] Improve performance of broadcast ops backward pass (#11252)
     new f22ae1d  Fix TabError for Python 3 (#11669)
     new d611037  Some Python 3 fixes in ./tools (#11672)
     new 5b4d528  fix batchnorm problem with sparse matrices when fix_gamma=True (#11656)
     new 238a497  Disable batchnorm (#11746)
     new d5c5006  Update (#11679)
     new 0bb6e2b  Speed up GPU Windows builds in CI (#11745)
     new daa2a43  Revert "Some Python 3 fixes in ./tools (#11672)" (#11750)
     new de2edc7  MXNET-650 [Perl] Gluon ModelZoo, Gluon examples  (#11642)
     new 11cd73f  Fix dist kvstore for trainer and flaky dist kvstore test (#11633)
     new 3ae4331  [MXNET-614] Adding Synchronized Batch Normalization (#11502)
     new 444775b  Remove dead code in Jenkinsfile (#11762)
     new 01cba7b  Fix R installation in CI (#11761)
     new e77fa04  fix R quick install instructions (#11755)
     new 3ea67a7  fix eps value for test_batchnorm_fallback (#11759)
     new 2fdd902  Fix force_addtakegrad (#11764)
     new 5495e2a  [MXNET-531] CNN Examples for Scala new API (#11292)
     new f622a99  updating image reference and fixing additional text (#11754)
     new f235605  Fix nightly R installation (#11768)
     new c792c9f  add arange fix (#11766)
     new 716df26  add arange fix (#11766)
     new bcf4245  Some Python 3 fixes in ./example (#11671)
     new b531cbc  MXNet ONNX export - Dot op support (#11771)
     new e70f6ff  onnx split op import (#11593)
     new 007c288  [MXNET-622] Add timeout for BLC (#11554)
     new 9264755  Docs build improvements (#11670)
     new fe567f2  added new users list info; sync info with channels page (#11782)
     new ede98e5  [MXNET-682] Disable Caffe import integration test (#11781)
     new e994a35  fix rnn layer save (#11776)
     new 3051c49  [MXNET-381] Enhancement of take operator (#11326)
     new bcf3a88  fix rnn api docs (#11788)
     new ab30cf8  [MXNET-8230] test_operator_gpu.test_rms fails (#11749)
     new 4a02bde  Some Python 3 fixes in ./tools again (#11757)
     new 6abfc70  NPE in BucketingModule when providedLabel of DataBatch is not set (#11790)
     new ae8b84c  Update Scala-MXNet install instructions (#11787)
     new 896271b  fix test_countsketch flakiness (#11780)
     new 072dd5a  [MXNET-531] Custom Operator Example for Scala (#11401)
     new 904052d  fix BucketSentenceIter bug related to #11430 (#11580)
     new c2a1a7c  fix bug when prefetch_buffer=1 (#11793)
     new 039e68a  Disable kvstore test (#11798)
     new 6d927d1  [MXNET-681] Add retry mechanism for the docker build stage (#11779)
     new 1240577  Use feature detection in (#11803)
     new 80b7715  Update documentation for count_tokens_from_str (#11800)
     new 54632bc  [MXNET-626] Add while_loop (#11566)
     new 4b8ab63  MXNet docs change for 1.2.1 release (#11791)
     new 1031fe1  [MXNET-600][Scala] NDArray auto-collector (#11751)
     new 9b30af2  enable CPU kernel for all RNN layer forward (#11807)
     new bbcf083  Refactor armv8 builds and add to CI (#11579)
     new ee40f7c  Fix broken website build pipeline (#11813)
     new d86f954  Initial commit (#11823)
     new b1f2f44  [MXNET-560] Add temperature parameter in Softmax operator (#11466)
     new f15b1b8  Added the diag() operator (#11643)
     new 83ecf2d  Adding missing optimizers that use momentum. (#11826)
     new 2d7166f  Add trove classifiers for PyPI page (#11818)
     new 0147388   Some Python 3 fixes in ./tests (#11799)
     new 61a7695  New ecosystem website page (#11528)
     new b16f875  [MXNET-524] Broadcast like operator (#11820)
     new 1242769  detailed doc about global pool layers in Gluon (#11832)
     new 95de767  fix for bug #10868: _backward_softsign activation is incorrect (#11827)
     new cdb1297  [MXNET-489] MKLDNN Pool test (#11608)
     new e14b575  updated docs and hashes for vgg16,16_bn,19,19_bn (#11830)
     new 11b69f2  Flaky: Disabled broadcast_mod unit test (#11845)
     new 6798703  fix a bug in CachedOp. (#11675)
     new e80c89f  add link changes (#11828)
     new 049d048  [MXNET-531] improvement of adding source.jar to assembly (#11681)
     new 60b6ab6  Use new Symbol APIs in Profiler example (#11477)
     new 7c1a49b  Updating MultiTask example to use new infer api and adding test for CI (#11605)
     new a94fcf9  [MXNET-319] Reduce Scala Warning messages (#11789)
     new 65fee98  support 1D inputs in leaky relu (#11850)
     new b4156da  Update doc examples for MCC (#11852)
     new 38282e9  [MXNET-599] Partial shape infer for Slice (#11406)
     new 74d6692  Fix spatial transformer op (#11806)
     new 4bb141d  [MXNET-684] Add `cond` operator (#11760)
     new 55fef30  [MXNET-688] Fix quantization divide by zero errors (#11833)
     new f407345  [MXNET-645] Add flakiness checker (#11572)
     new 8f4b092  Enable control flow test (#11869)
     new 64d2e8b  [MXAPPS-581] Nightly Straight Dope tests. (#11814)
     new fe07d50  [MXNET-331] Single machine All Reduce Topology-aware Communication (Updated) (#11591)
     new fa935a8  Fix file name creation for Windows (#11765)
     new 8a21a06  update vgg pretrained model (#11860)
     new 07a9977  Add verify_ssl option to (#11546)
     new 424fafe  [MXNET-710] Change POM files to be able to regularly publish to Apache Release & Maven Central Repo (#11862)
     new 06f4ec7  Enable three retries for Docker build commands (#11877)
     new 0b8b939  Avoid Division by Zero (#11397)
     new fe1c7ab  making AddTakeGrad as default for backward of embedding and take to avoid nan (#11795)
     new be47870  [MXNET-563] Refactor R optimizers to fix memory leak (#11374)
     new 832a5fb  Add logistic regression tutorial (#11651)
     new 7cd01ff  Re-enabling randomized test_operator/test_operator_gpu.test_dot (#11888)
     new 302aae3  Fix non-determinism of dot(csr.T, dns) = dns with tests (#11825)
     new f5b95b0  Support integer type in ImageIter (#11864)
     new c13ce5c  [MXNET-378] Adding depth_to_space and space_to_depth operator(Updated) (#11587)
     new 2bddf6f  Fix mxnet ctc_loss bug (#11834)
     new 4bbf15c  [MXNET-344] Add more operators to onnx import (#11856)
     new a8c8737  make skiptest work (#11889)
     new bd3fc88  Fix flaky test test_deconvolution (#11630)
     new 011a0dc  Remove fixed seed for test_sparse_nd_save_load (#11920)
     new 196468d  Corrections to profiling tutorial (#11887)
     new 54ebc5d  Fix image classification scripts and Improve Fp16 tutorial (#11533)
     new ab5242c  [MXNET-711] Website build and version dropdown update (#11892)
     new 6372037  [MXAPPS-581] Fixes for broken Straight Dope tests. (#11923)
     new 83ae3a3  Disable flaky test: test_spatial_transformer_with_type (#11930)
     new 2cc5a42  Add linux and macos MKLDNN Building Instruction (#11049)
     new b2fd3b1  [MXNET-531] Add download util (#11866)
     new 024b5a9  [MXNET-11241] Avoid use of troublesome cudnnFind() results when grad_req='add' (#11338)
     new 478b4a1  Improving documentation and error messages for Async distributed training with Gluon (#11910)
     new 815f42d  [MXNET-641] fix R windows install docs (#11805)
     new 461ba07  a hot fix for mkldnn link (#11939)
     new 7ffb252  re-enabling randomized test_l2_normalization (#11900)
     new a56a569  [MXNET-651] MXNet Model Backwards Compatibility Checker (#11626)
     new 98a41af  [MXNET-531] NeuralStyle Example for Scala (#11621)
     new ed20304  [MXNET-750] fix nested call on CachedOp. (#11951)
     new 51f650e  extend reshape op to allow reverse shape inference (#11956)
     new 1eef070  Improve sparse embedding index out of bound error message; (#11940)
     new fc912f3  [MXNET-770] Remove fixed seed in flaky test (#11958)
     new 394e5cc  Update ONNX docs with the latest supported ONNX version (#11936)
     new eed7a34  Reduced test to 3 epochs and made gpu only (#11863)
     new c6a32b6  Fix flaky tests for test_laop_4 (#11972)
     new 061076d  Updating R client docs (#11954)
     new 31c5fbc  Fix install instructions for MXNET-R (#11976)
     new a93905d  [MXNET-751] fix ce_loss flaky (#11971)
     new 564e01a  [MXNET-769] set MXNET_HOME as base for downloaded models through base.data_dir() (#11636)
     new 6009b26  [MXNET-748] linker fixed on Scala issues (#11989)
     new 946e9d0  [MXNET-772] Re-enable (#11979)
     new 1bd9356  [MXNET-771] Fix Flaky Test (#11978)
     new 833de7e  remove mod from arity 2 version of load-checkpoint in clojure-package (#11808)
     new bcfab3a  Add unit test stage for mxnet cpu in debug mode (#11974)
     new c937277  Website broken link fixes (#12014)
     new 1818280  removed seed from flaky test (#11975)
     new 619700a  Disable ccache log print due to threadunsafety (#11997)
     new 2534164  Added default tolerance levels for regression checks for MBCC (#12006)
     new 32c2e15  Disable flaky mkldnn test_requantize_int32_to_int8 (#11748)
     new 1fa04f2  [MXNET-769] Usability improvements to windows builds (#11947)
     new 5628194  Fix import statement (#12005)
     new 3dd0003  Disable flaky test test_random.test_gamma_generator (#12022)
     new 3910c08  [MXNET-770] Fix flaky test: test_factorization_machine_module (#12023)
     new ae698f9  disable opencv threading for forked process (#12025)
     new 22c97ef  Bug fixes in control flow operators (#11942)
     new 8d4d5fa  Fix data narrowing warning on (#11969)
     new 09add8a  Fix flaky tests for test_squared_hinge_loss (#12017)
     new 43ed13f  Fix flaky tests for test_hinge_loss (#12020)
     new 067197f  remove fixed seed for test_sparse_ndarray/test_operator_gpu.test_sparse_nd_pickle (#12012)
     new 942ab8d  Removed fixed seed from , test_loss:test_ctc_loss_train (#11985)
     new 727c318  Removed fixed seed from , test_loss:test_sample_weight_loss (#11986)
     new 6f7ef57  Fix reduce_kernel_M1 (#12026)
     new 1b4a7c9  Update to remove fixed seed (#11995)
     new 4649bfa  [MXNET-23] Adding support to profile kvstore server during distributed training  (#11215)
     new deae58d  Re-enabling test_ndarray/test_cached (#11950)
     new 5474b08  make gluon rnn layers hybrid blocks (#11482)
     new 7261d8c  [MXNET-751] fix bce_loss flaky (#11955)
     new 70efd32  Doc fix for a few optimizers (#12034)
     new f3712b8  Remove stale Keras-MXNet tests from MXNet repo (#11902)
     new 9dd5edd  Disable flaky cpp test (#12056)
     new 82ca632  Adjusting tolerance level and removing fixed seed for tests: test_ifft, test_fft (#12010)
     new 9b7351a  [MXNET-774] Flaky test in (#12016)
     new 9b18af8  fix flaky test_quantization.test_get_optimal_thresholds (#12004)
     new cb94be9  removed fixed seed 1234 (#12072)
     new 403831a  improve error message of cudnn operators (#11886)
     new edc28af  Fix for undefined variable errors (#12037)
     new 5293336  Fix undefined_variable lint errors in examples (#12052)
     new 5bf77bb  MXNET-776 [Perl] Better documentation/bug fixes. (#12038)
     new 27f8e42  Redesign Jenkinsfiles (#12000)
     new 181a2e2  fix unidirectional model's parameter format (#12055)
     new 1a222c3  Fix syntax errors in Jenkinsfiles (#12095)
     new c000930  [MXAPPS-581] Straight Dope nightly fixes. (#11934)
     new 9602b5f  Fix jenkinsfile syntax errors (#12096)
     new ef89138  remove fixed seed for test_triplet_loss (#12011)
     new 39a4ea4  got rid of fixed seed for test_optimizer/test_operator_gpu.test_ftml (#12003)
     new b9673a9  [MXNET-696] Fix undefined variable errors (#11982)
     new 23d0c06  got rid of fixed seed for test_optimizer/test_operator_gpu.test_nag (#11981)
     new d762755  Fix flaky test for elementwise_sum (#11959)
     new aa1be93  Re-enabling test_operator.test_binary_math_operators (#11712) (#12053)
     new ac5036d  update docs to explain CPU incompatibilities (#11931)
     new 3b449ef  removed fixed from test_optimizer.test_signum (#12088)
     new 5aef78f  Add missing object to tests/nightly/model_backwards_compatibility_check/JenkinsfileForMBCC (#12108)
     new 22bfbd7  Add GetName function in Symbol class for cpp pack (#12076)
     new 23b8f89  Add unique number of parameters to summary output in Gluon Block (#12077)
     new 470c3dc  Update documentation (#12097)
     new bb0f8a6  [MXNET-244] Update RaspberryPI instructions (#11562)
     new af15853  [MXNET-749] Correct usages of `CutSubgraph` in 3 control flow operators (#12078)
     new c053262  [MXNET-703] TensorRT runtime integration (#11325)
     new 787cd08  Decrease success rate to make test more stable (#12092)
     new 7e99621  Add Clojure to website nav (#12075)
     new f6215b1  Fix flaky tests for quantize and requantize (#12040)
     new 53703e2  [MXNET-703] Use relative path for symbol import (#12124)
     new 5a9c3af  Fix shared memory with gluon dataloader, add option pin_memory (#11908)
     new 6f7dee0  reduce a copy for rowsparse parameter.reduce (#12039)
     new 584c5c1  GPU Memory Query to C API (#12083)
     new 2fc4248  take custom dataset into consideration (#12093)
     new f499dc4  [MXNET-782] Fix Custom Metric Creation in R tutorial (#12117)
     new 89717d4  [MXAPPS-805] Notebook execution failures in CI. (#12068)
     new f7211b2  rm wrong infertype for AdaptiveAvgPool and BilinearReisze2D (#12098)
     new 1f8debb  Document MXNET_LIBRARY_PATH environment variable which was not documented explicitly. (#12074)
     new c44f16b  Generalized reshape_like operator (#11928)
     new 9fd45b8  fix cython nnvm include path (#12133)
     new 1b60478  CI scripts refinements. Separate Py2 and Py3 installs cripts. Fix perms. (#12125)
     new 6f7254c   zipfian random sampler without replacement  (#12113)
     new 096aeb6  update dmlc-core (#12129)
     new d076d10  Fix quantized graphpass bug (#11937)
     new 0241e2c  support selu activation function (#12059)
     new 1eafbfc  Fix flaky test test_operator_gpu:deformable_conv and deformable_psroi_pooling (#12070)
     new ec90f4a  [MXNET-767] Fix flaky test for kl_loss (#11963)
     new 2193819  [MXNET-788] Fix for issue #11733 pooling op test (#12067)
     new 0d68079  Do not show "needs to register block" warning for registered blocks. (#12130)
     new 9933d7a  Fix precision issue of test case test_rnnrelu_bidirectional (#12099)
     new 95dd95c  Accelerate the performance of topk for CPU side (#12085)
     new ee8755a  Remove unused TensorRT code (#12147)
     new a6ecb59  Disable test_io.test_CSVIter (#12146)
     new b5a2b98  Fix RAT license checker which is broken in trunk (#12148)
     new 92edca8  set bind flag after bind completes (#12155)
     new f3070fb  Fix MXPredReshape in the c_predict_api (#11493)
     new 525ead9  [Flaky Test] Fix test_gluon_model_zoo.test_models when MXNET_MKLDNN_DEBUG=1  (#12069)
     new 8506783  RAT check readme updated (#12170)
     new b95835a  update ndarray stack Doc for #11925 (#12015)
     new cd6b1cb  Add worker_fn argument to multiworker function (#12177)
     new b675d69  Remove fixed seed for test_huber tests (#12169)
     new 5b9251b  Removed fixed seed and increased learning rate and tolerance for test_nadam (#12164)
     new 57313e3  documentation changes. added full reference (#12153)
     new 391377a  Partially enable flaky test for norm operator (#12027)
     new 8e60586  add examples for slicing option (#11918)
     new da2ce8f  Module predict API can accept NDArray as input (#12166)
     new c220974  [MXNET-744] Docs build tools update (#11990)
     new 6b01a4f  [MXNET-696] Fix undefined name errors (#12137)
     new 8f61777  Fix profiler executer when memonger is used (#12152)
     new 0455a11  add handling for grad req type other than kNullOp for indices (#11983)
     new 0fbdff8  Fix a minor bug in deformable_im2col.cuh (#12060)
     new 7cb28f2  [MXNet-744] Fix website build pipeline Python 3 issues (#12195)
     new ab92fd8  Fix MKLDNNSum cpp test failure (#12080)
     new 64283a9  bump timeout on Jenkins for docs/website to 120 min (#12199)
     new 427167e  Fixing typo in python/mxnet/symbol/ (#12194)
     new 1d31776  Fix the topk regression issue (#12197) (#12202)
     new a2b0164  Drop dmlc-core commits that break MXNet build (#12189)
     new 23c80c4  Remove fixed seed from laop_2 test (#12206)
     new 6843914  MKLDNN can be turned off with env var (#12058)
     new 6876820  Disable flaky test: test_operator_gpu.test_depthwise_convolution (#12204)
     new f6d594b  Fix undefined name in __init__ (#12207)
     new 30f6c39  [MXNET-696] Simplify the logic behind the reload(mxnet) test (#12188)
     new 54ed3e5  [MXNET-696] Define cmp() in Python 3 for line 222 (#12191)
     new b793768  documentation fix (#12212)
     new 38bcb88  Fix typo. (#12211)
     new 599ffe4  Update (#12208)
     new 2b4d512  fix flasky unittest for deformable psroi pooling (#12178)
     new abbe283  [MXNET-689] add DataDesc type for the Scala Package (#11844)
     new fa18f51  Disable flaky test: test_operator_gpu.test_convolution_grouping (#12220)
     new 2ff1837  Revert "CI scripts refinements. Separate Py2 and Py3 installs cripts. Fix perms. (#12125)" (#12223)
     new 633bef3  [MXNET-791] Pick with negative indices (#12090)
     new cd9f9c8  Fix a bug in CachedOP. (#12184)
     new f737cac  Add test to check that binded is not set when exception thrown (#12167)
     new 24e42a0  Revert "[MXNET-696] Define cmp() in Python 3 for line 222 (#12191)" (#12231)
     new afb77f8  [MXNET-737]Add last batch handle for imageiter (#12131)
     new 54af0cc  [MXNET-806] Report error when shape mismatch in "where" operator (#12174)
     new 89fbb29  re-enable randomized test_bilinear_sampler (#12227)
     new 1355bd7  [MXNET-703]  Remove race condition in onnx build (#12225)
     new ca92e3d  Add new test case for topk to cover large size of matrix as input (#12217)
     new f804ef5  add new scala build function (#12236)
     new 46437b8  Don't override global warnings filter (#12245)
     new 338a40b  Disable flaky test deformable_psroipooling (#12246)
     new b74a516  Update
     new 224ae96  Typo in README
     new 605c569  Disabled flaky test: test_operator_gpu.test_bilinear_sampler (#12249)
     new b03227d  Pad Operator Type Support (#12035)
     new 7f22b78  Fix typo in graph_executor (#12252)
     new 14b1e48e Tweak ExecType descriptions (#12251)
     new aec7d2c  Python fixes for PyLint test "consider-using-in" (#12214)
     new 383a2d0  Fix flaky tests in control flow (#12192)
     new e3ff167  Tweaked comments for OpExecutor. (#12253)
     new c479eb2  fix potential floating number overflow, enable float16 (#12118)
     new 2899715  [MXNET-792] Fix for issue #9816 with dropout operator and RNG (#12091)
     new a4aced7  Temporarily disable ARMv7 builds (#12260)
     new d7b39f4  Add support for kAddTo in softmax backward (#11836)
     new 1abe632  [MXNET-848] Pin dockcross base images in CI (#12270)
     new 81bc69b  [MXNET-484] MKLDNN C++ test for LRN operator (#11831)
     new 09ee17c  [MXNET-696] 'make pylint' should run a current version of PyLint (#12200)
     new ba6f4e1  fix broken links to subscribe to dev mailing list (#12271)
     new 8dbbfad  [MXNET-839] Fix the broken link in the Scala package (#12256)
     new 332a664  add gigasquid (Carin Meier) to the Clojure language binding (#12198)
     new 38f80af  [MXNET-836] RNN Example for Scala (#11753)
     new bf1edaf  Fix JNI custom op code from deregistering the operator (#11885)
     new 78c88df  [MXNET-628] Fix example on text classification using LSTMs on IMDB dataset (#12263)
     new 250c4ff  Add fpic correctly in cmake (#12281)
     new 96ac3ae  [MXAPPS-581] Enable remaining tests in CI (#12165)
     new c692ffd  [MXNET-795] Fix a bug that CutSubgraph works only when each subgraph has its distinct name (#12106)
     new e04565a  [MXNET-703] Update onnx-tensorrt to most recent version (#12274)
     new c9274be  Enable OpenMP for armv8 builds (#12273)
     new c1a8948  Revert "Disable kvstore test (#11798)" (#12279)
     new a3add6a  [MXNET-850] Fix incorrect abs function call (#12262)
     new 902c579  Make the output format of control flow operators consistent with their UDFs (#12209)
     new b060a01  removed mentions and links to a deleted example (#12288)
     new adc57d1  fixed broken links (#12293)
     new 271fd02  fixed broken link (#12289)
     new 76bdc87  [MXNET-696] Fix undefined name and enable Pylint undefined variable (#12277)
     new 5a9949a  Make sure input symbol names are unique in control flow operators. (#12187)
     new 9ef919e  Fix broken anchor in doc (#12193)
     new 08f1c2d  [MXNET-507] Set arbitrary dtype for ret_indices in ordering ops (#12250)
     new 2f177d8  [MXNET-729] Use NDArrayCollector to fix memory leaks in Scala Examples (#12232)
     new d19d6c2  Update docs of LogisticRegressionOutput (#12299)
     new 1ffb3f5  Add randn to random module (#12156)
     new 67ba3c5  A binary RBM example (#11268)
     new 84b3b56  upload key (#12301)
     new 72efe41  [MXNET-855] Update clang to version 6 (#12305)
     new e1b2265  [MXNET-849] - Enable armv7 with pinned docker images. (#12272)
     new 8ee8357  re-enable randomized test_ndarray_pickle (#12292)
     new ac85d7a  re-enable randomized test_ndarray_elementwise (#12300)
     new 43581a7  Generalized broadcast_like operator (#11984)
     new 4664a30  [MXNET-696] Define cmp() in Python 3 again (#12295)
     new cc30fab  MXNet to ONNX export tutorial  (#12297)
     new 490cf99  Disable flaky test test_operator.test_dropout (#12330)
     new 5189495  add initializer test (#12196)
     new 7bfe427  Allow stop of arange to be inferred from dims. (#12064)
     new ac10e95  [MXAPPS-581] Disable a long test in the SD nightly. (#12326)
     new 2276bb0  Tighten up PyLint directives again (#12322)
     new 91730fd  Fix a bug in where op with 1-D input (#12325)
     new 4f8d39f  [MXNET-825] Fix CGAN R Example with MNIST dataset (#12283)
     new 4e6366c  set proper atol for check_with_uniform (#12313)
     new d2495d1  [MXAPPS-581] Disable a long test in the SD nightly. (#12326) (#12339)
     new 15e43c0  Fall back when sparse arrays are passed to MKLDNN-enabled operators (#11664)
     new 5b37cf6  explain the details for Scala Experimental (#12348)
     new 39d0d06  Add cloverage codecov report to CI for clojure (#12335)
     new ad34e05  Disable flaky test test_ndarray.test_order (#12311)
     new 54d5777  add activation information for mxnet.gluon.nn._Conv (#12354)
     new 7230bb9  MKLDNN Forward FullyConnected  op cache (#11611)
     new c88b8ee  fix flaky test: test_broadcast_binary_op (#11875)
     new 84665e3  [MXNET-422] Distributed training tutorial (#10955)
     new 48d2155  [MXNET-535] Fix bugs in LR Schedulers and add warmup (#11234)
     new 308ada1  Make check_isfinite, check_scale optional in clip_global_norm (#12042)
     new 2f73958  [MXNET-859] Add a clang-tidy stage to CI (#12282)
     new 6a7bfe9  Separate refactoring from #12276 in a prior PR (#12296)
     new 4e03087  adding apache conf promo to home page (#12347)
     new 9e48f70  Add a tutorial for control flow operators. (#12340)
     new 3643b27  [MXAPPS-581] Disable an additional long test in the SD nightly (#12343)
     new cb7dc7f  Update ONNX API docs references (#12317)
     new d9ea96a  Documentation update related to sparse support (#12367)
     new d234b32  [MXNET-690] Add tests for initializers in R (#12360)
     new ae5d60f  Update PyPI version number (#11773)
     new e2a3eef  A solution to prevent zombie containers locally and in CI (#12381)
     new 65c374d  Disabled flaky test: test_mkldnn.test_activation (#12378)
     new 1f0d6ba  Add tutorial Gotchas using NumPy (#12007)
     new ba8a9d1  support softmin operator with unit test (#12306)
     new 846086d  Revert "Revert "Disable kvstore test (#11798)" (#12279)" (#12379)
     new 5af3d39  Edit shape.array doc and some style improvements (#12162)
     new e456dc4  Fix speech recognition example (#12291)
     new 6ca9092  fix bug in 'device' type kvstore (#12350)
     new 32c9ca7  [MXNET-753] Fallback when using non-MKLDNN supported operators (#12019)
     new a64cf7d  Subgraph API for integrating accelerators with MXNet (#12157)
     new 3fb7ab5  fix search result 404s (#12414)
     new 4033cdd  Make the output of ci/docker/install/ less verbose (#12422)
     new 5ace01c  fixed docs/website build checkout bug (#12413)
     new b8ee84b  fix help in imread (#12420)
     new 10e94f8  fixed flaky test issue for test_operator_gpu.test_convolution_grouping (#12385)
     new 58560f6  fixed flaky test issue for test_operator_gpu.test_depthwise_convolution (#12402)
     new 307e62a  Add trigonometric operators (#12424)
     new d00cb42  adjust tolerance levels of test_l2_normalization (#12429)
     new e0498eb  Revert "fixed flaky test issue for test_operator_gpu.test_depthwise_convolution (#12402)" (#12441)
     new 8e4aeee  Updated tvm submodule head (#12448)
     new e290623  Sphinx error reduction (#12323)
     new 4e19a32  remove flaky test and add consistency test for stable testing (#12427)
     new de9a2e8  Fix flaky test test_operator_gpu.test_batchnorm_with_type (#11873)
     new d7111d3  Installation instructions consolidation (#12388)
     new 28cb75a  [MXNET-908] Enable minimal OSX Travis build (#12462)
     new 285adda  [MXNET-909] Disable tvm_bridge test (#12476)
     new 2bcfc08  Fix tvm dependency for docker (#12479)
     new 445967e  Fix flaky test: test_mkldnn.test_activation #12377 (#12418)
     new 4eb7626  allow foreach on input with 0 length (#12471)
     new 6043ef0   [MXNET-696][PYTHON][UNDEFINED NAME] import logging in ci/ (#12488)
     new d410de1  [MXNET-703] Static linking for libprotobuf with TensorRT (#12475)
     new 690cf7e  fix render issue on &lt; and &gt; (#12482)
     new 3406845  Adding python SVRGModule for performing SVRG Optimization Logic (#12376)
     new ac4ef21  ONNX export - Clip operator (#12457)
     new c8c3b04  [MXNET-853] Fix for smooth_l1 operator scalar default value (#12284)
     new 20a087f  update C++ example so it is easier to run (#12397)
     new d789101  Add Python API docs for test_utils and visualization (#12455)
     new acede67  fix subscribe links, remove disabled icons (#12474)
     new 9ec4879  Temporarily disable flaky tests (#12513)
     new 597a637  Change the way NDArrayIter handle the last batch (#12285)
     new 4ee866f  Add support for more req patterns for bilinear sampler backward (#12386)
     new 35ca13c  Temporarily disable flaky tests (#12520)
     new 3d83c89  Fix broken URLs (#12508)
     new 90599e1  further bump up tolerance for sparse dot (#12527)
     new 7ea0533  Revert "Fix flaky test: test_mkldnn.test_activation #12377 (#12418)" (#12516)
     new 619bc3e  Remove regression checks for website links (#12507)
     new 46a5cee  [MXNET-580] Add SN-GAN example (#12419)
     new ceabcaa  Removing the re-size for validation data, which breaking the validation accuracy of CIFAR training (#12362)
     new dcf4e12  [MXNET-807] Support integer label type in ctc_loss operator (#12468)
     new 6456687  Draw labels name (#9496)
     new 8ff50c9  Revert "Change the way NDArrayIter handle the last batch" (#12537)
     new cb6c72c  Consistent website theme and custom 404 (#12426)
     new e81c0e2  update apachecon links to https (#12521)
     new b8153f6  MXNET-873 - Bring Clojure Package Inline with New DataDesc and Layout in Scala Package (#12387)
     new 7735fa6  Fix/public internal header (#12374)
     new 741635a  MKLDNN Backward op cache (#11301)
     new 3316f04  Used jom for parallel windows builds (#12533)
     new f387df4  [MXNET-851] Test coverage metrics for R-package (#12391)
     new c684909  [HOLD] 1.3.0 release website updates (#12509)
     new 5dc8cde  [MXNET-703] Add a TensorRT walkthrough (#12548)
     new 0b91230  [MXNET-950] Enable parallel R dep builds in CI (#12552)
     new 52116d4  Fix lazy record io when used with dataloader and multi_worker > 0 (#12554)
     new e213286  Fix error in try/finally block for blc (#12561)
     new f8ed533  add cudnn_off parameter to SpatialTransformer Op and fix the inconsistency between CPU & GPU code (#12557)
     new 8209906  [MXNET-798] Fix the dtype cast from non float32 in Gradient computation (#12290)
     new 9a5c3ff  redirecting navigation items to latest info (#12540)
     new 6f93b03  Add license headers to R-package (#12559)
     new ff39cf1  Make the Ubuntu scripts executable after download. (#12180)
     new ff10e1a  Fix the installation doc for MKL-DNN backend (#12534)
     new 9032e93  Updating news, readme files and bumping master version to 1.3.1 (#12525)
     new d8c51e5  replacing windows setup with newer instructions (#12504)
     new 7ce58c3  [MXNET-703] Add TensorRT runtime Dockerfile (#12549)
     new 3c42447  [MXNET-860] Avoid implicit double conversions (#12361)
     new 7168254  [MXNET-860] Reduce redundant copies, check for regressions with clang-tidy (#12355)
     new f6d2bef  Disable installation nightly test (#12571)
     new 6dca0c6  Fix CodeCovs proper commit detection (#12551)
     new 53d786d  [MXNET-952] Check for correlation kernel size along with unittest (#12558)
     new 1e95bcd  [MXNET-882] Support for N-d arrays added to diag op. (#12430)
     new 5d07a07  Added comment to docs regarding ToTensor transform (#12186)
     new 1744b0c  [MXNET-968] Fix MacOS python tests (#12590)
     new acf309e  Infer dtype in SymbolBlock import from input symbol (#12412)
     new d8984e8  [MXNET-910] Multithreading inference. (#12456)
     new 96424b3  Speed up CI windows builds (#12563)
     new db295ef  [MXNET-953] - Add ASAN sanitizer, Enable in CI (#12370)
     new 61b17b7  [MXNET-908] Enable python tests in Travis (#12550)
     new d01fbf3  Pinned dockcross to a tag with fixed ABI for RPi (#12588)
     new 73d8897  add TensorRT tutorial to index and fix ToC (#12587)
     new ce6525a  Track epoch metric separately (#12182)
     new dd55b2b  add fluent methods for softmin (#12605)
     new d4adbd2  Refine the documentation of im2rec (#12606)
     new 1a9814e  Fixed typo in (#12601)
     new f945f3b  Fix typo in profiler.h (#12599)
     new d754703  Tweaked the copy in c_predict_api.h (#12600)
     new 97a7457  fix test_activation by lowering threshold + validate eps for check_numeric_gradient (#12560)
     new 7797584  Updated Symbol tutorial with Gluon (#12190)
     new 3401e6e  review require() usages to add meaningful messages. (#12570)
     new ba993d1  improve tutorial redirection (#12607)
     new 0f8d2d2  Remove pip overwrites (#12604)
     new 251e11d  Fixed NoSuchMethodError for Jenkins Job for MBCC (#12618)
     new c479664  update download links to apache distros (#12617)
     new 2becd76  [MXNET-860] Update to modern nullptr usage (#12352)
     new 33876c2  [MXNET-922] Fix memleak in profiler (#12499)
     new e82eef9  Enable gluon multi worker data loader test (#12315)
     new 1045140  Correct PR branch detection in code coverage (#12615)
     new cb6815c  Include missing import in TensorRT tutorial (#12609)
     new 9f5c96e  [MXNET-969] Fix buffer overflow in RNNOp (#12603)
     new d4991a0  [MXNET-360]auto convert str to bytes in img.imdecode when py3 (#10697)
     new dba9487  Update and modify Windows docs (#12620)
     new f73c1c7  update docs to list cmake required for build from source page (#12592)
     new c4975e7  update the distributed_training document (#12626)
     new 504d24c  Add docstring in (#12621)
     new 846bda4   Fixed param coercion of clojure executor/forward (#12627) (#12630)
     new 9f14a68  [MXNET-779]Add DLPack Transformation API (#12047)
     new 6614b0c  Fix version dropdown behavior (#12632)
     new de01c46  Fix reference to wrong function (#12644)
     new 3caf2ca  [MXNET-876] make CachedOp a normal operator (#11641)
     new c93c78e  allow custom path and static linking for custom mallocs in make (#12645)
     new ccba621  Enable C++ coverage (#12642)
     new 54fef4a  Gluon DataLoader: avoid recursionlimit error (#12622)
     new 76ae725  Updated download links and verification instructions (#12651)
     new 29ac191  [MXNET-867] Pooling1D with "same" padding (#12594)
     new 4236b98  Fix the location of the tutorial of control flow operators (#12638)
     new 769165b  [Doc] Change the description for pip packages (#12584)
     new 3042825  Change dependencies documentation opencv2-->opencv (#12654)
     new d4e202a  fix bug, issue 12613 (#12614)
     new 34d9e10  [MXNET-780] Fix exception handling bug (#12051)
     new 40e3d9c  fix bug in prelu , issue 12061 (#12660)
     new 0d29aec  add mentions of the gluon toolkits and links to resources (#12667)
     new 82c96f5  Remove fixed seed for test_ctc_loss (#12686)
     new 62d8fac  remove apachecon promo (#12695)
     new 1ae3217  Onnx version update from 1.2.1 to 1.3 in CI (#12633)
     new ca81535  [MXNET-833] [R] Char-level RNN tutorial fix (#12670)
     new 9ce5792  Add documents for two new environment variables for memory pool. (#12668)
     new c92a202  update mshadow for omp acceleration when nvcc is not present  (#12674)
     new 4da3001  fix for test order (#12358)
     new 063139d  [MXNET-951] Python dockerfiles built on pip binaries and build/release script (#12556)
     new 6522a2c  [MXNET-500]Test cases improvement for MKLDNN on Gluon (#10921)
     new 094661d  Enable test_gluon.test_export (#12688)
     new 454ec3c  Update (#12685)
     new c595f2d  Update (#12709)
     new 6fb81af  [MXNET-637] Multidimensional LSTM example for MXNetR (#12664)
     new 06e1d36  Add 'julia/' from commit '224ae962db8ae252ce1ac317c6d4a2cfe67a7d77'
     new 9f8425b  import Julia binding
     new 42cc960  Fix static / dynamic linking of gperftools and jemalloc (#12714)
     new 5c86143  Disable test batchnorm slice (#12716)
     new bcd24f8  [MXNET-860] Use emplace where helpful (#12694)
     new 811618a  [MXNET-953] Correct ASAN cflags flag (#12659)
     new ebe6ea8  [MXNET-860] Remove std::moves that have no affect (#12730)
     new 01dd703  add FListInputNames attribute to softmax_cross_entropy (#12701)
     new be0c6ef  Fix #12672, importing numpy scalars (zero-dimensional arrays) (#12678)
     new e41faf0  [MXNET-908] Speed up travis builds to avoid timeouts (#12706)
     new 931555c  Merge branch 'pr-10149' import Julia binding
     new e1fe7b1  Throw exception if MXSymbolInferShape fails. (#12733)
     new e93af41  [MXNET-716] Adding Scala Inference Benchmarks (#12721)
     new f9f7416  [MXNET-623] Fixing an integer overflow bug in large NDArray (#11742)
     new d6dfd41  Change numpy version to 1.15.2 in python and docker install requirements (#12711)
     new 857cea1  Reenable test_gluon.test_conv (#12718)
     new c2bb012  Refine mxnet python installation (#12696)
     new c993ef1  Update packages and tests in the straight dope nightly (#12744)
     new 836ba78  Fix failing GPU test on single GPU host (kvstore) (#12726)
     new 5314cf4  Add option for automatic downcasting dtype for cudnn to allow using Tensorcore for fp32  (#12722)
     new 7e46b5e  [MXNET-1026] [Perl] Sync with recent changes in Python's API (#12739)
     new 077253d  fix benchmark on control flow operators. (#12693)
     new 610d79c  [MXNET-982] Provide example to illustrate usage of CSVIter in C++ API (#12636)
     new a2bafe1  [MXNET-912] Refactoring ctc loss operator (#12637)
     new 167355b  Add new name to (#12763)
     new 3c3506f  Add resnet50-v1 to benchmark_score (#12595)
     new f0140b3  [MXNET-716][MIRROR #12723] Scala Benchmark Extension pack (#12758)
     new ad027ca  Implement mkldnn convolution fusion and quantization. (#12530)
     new 4968900  Introduction to Clojure-MXNet video link. (#12754)
     new bcbac41  [MXNET-915] Java Inference API core wrappers and tests (#12757)
     new c98b19e  Disabled flaky test: test_mkldnn.test_Deconvolution (#12770)
     new 064c87c6 Add mkl-dnn to docker install method (#12643)
     new 443ded4  Improve mkldnn fallback. (#12663)
     new ddddede  Fix regression in MKLDNN caused by PR 12019 (#12740)
     new c0cd583  Fixed broken link for Baidu's WARP CTC (#12774)
     new 1b72d45  Updated to include lebeg and gigasquid, moved mabreu to committers section (#12766)
     new d02c9eb  Use modern onnx API to load model from file (#12777)
     new d43b786  Update (#12702)
     new 865255a  fix cnn visualization tutorial (#12719)
     new 0bab6d5  [MXNET-979] Add fix_beta support in BatchNorm (#12625)
     new 15deabc  [MXNET-947] Expand scala imclassification example with resnet (#12639)
     new 822e59f  R fix metric shape (#12776)
     new 50d2313  Revert "[MXNET-979] Add fix_beta support in BatchNorm (#12625)" (#12789)
     new 213ab09  Updated tvm submodule head (#12764)
     new 06bf600  Adagrad optimizer with row-wise learning rate (#12365)
     new c90d16c  Fix mismatch shapes (#12793)
     new 3eff8e8  Make Gluon download function to be atomic (#12572)
     new 5c74e3a  Re-enables test_dropout (#12717)
     new efa7d3a  [MXNET -1004] Poisson NegativeLog Likelihood loss (#12697)
     new c9a9db6  Update - Added "apple" to USE_BLAS comment (#12819)
     new 698bbec  [MXNet-1002] Add GluonCV and NLP tookits, Keras, and developer wiki to navigation (#12704)
     new b89a36d  fixed symbols naming in RNNCell, LSTMCell, GRUCell (#12794)
     new 64b9ac7  [MXNET-984] Add Java NDArray and introduce Java Operator Builder class (#12816)
     new 5961dce  simplify mac mkldnn build (#12724)
     new 13030b6  Change the way NDArrayIter handle the last batch (#12545)
     new 815f36c  [MXNET-707] Add unit test for mxnet to coreml converter (#11952)
     new d096aa5  Add embedding to print_summary (#12796)
     new 527e6a0  Scala Docs - Replace old Symbol api usages (#12759)
     new eee72d9  [MXNET-892] ONNX export/import: DepthToSpace, SpaceToDepth operators (#12731)
     new 3b5b2b2  R install instructions update for macOS (#12832)
     new fe2c4d8  Fixed __setattr__ method of _MXClassPropertyMetaClass (#12811)
     new 5a52374  Fixed regex for matching platform type in Scala Benchmark scripts (#12826)
     new 8271005  Added context object to run TestCharRnn example (#12841)
     new 5a680fc  [MXNET-703] Show perf info for TensorRT during tests (#12656)
     new 89eb24b  Update Operator Implementation Tutorial (#12230)
     new 7463810  Fix broken links (#12856)
     new 1ebbf94  Fix Flaky Topk (#12798)
     new 775870f  Add Psroipooling CPU implementation (#12738)
     new 6376c86  ONNX export: Fully connected operator w/o bias, ReduceSum, Square (#12646)
     new 97e86ff  Undefined name: load_model() --> utils.load_model() (#12867)
     new 673e31f  ONNX export/import: Selu (#12785)
     new 2e04aab  Sparse support for logic ops (#12860)
     new c5a7331  add a tutorial for the subgraph API. (#12698)
     new 441fdb7  MKL-DNN Quantization Examples and README (#12808)
     new 42e7110  [MXNET-1033] Fix a bug in MultiboxTarget GPU implementation (#12840)
     new 3154ec3  [MXNET-1107] Fix CPUPinned unexpected behaviour (#12031)
     new 137b6f5  NativeResource Management in Scala (#12647)
     new 2bc818e  Java Inference api and SSD example (#12830)
     new 94f3665  NativeResource Management in Scala (#12647) (#12883)
     new 76d5197  add/update infer_range docs (#12879)
     new daada21  Fix __all__ in optimizer/ (#12886)
     new 9c2810e  Add index_copy() operator (#12810)
     new 0ba259f  sparse support for take(csr, axis=0)  (#12889)
     new 524d01f  Add more models to benchmark_score (#12780)
     new d8c7375  [MXNET-1025] Add Jetpack 3.3 support to Jetson (#12735)
     new 58f4117  Fix Batch input issue with Scala Benchmark (#12848)
     new d3d343c  fix type inference in index_copy. (#12890)
     new 0137483  Extending the DCGAN example implemented by gluon API to provide a more straight-forward evaluation on the generated image (#12790)
     new d1234a4  [MXNET-674] Speed up GPU builds in CI (#12782)
     new 5b86701  [MXNET-793] ★ Virtualized testing in CI with QEMU ★ (#12094)
     new 3c81b3f  [MXNET-1017] Updating the readme file for cpp-package and adding readme file for example directory. (#12773)
     new be9ca1b  Fail the broken link job when broken links are found (#12905)
     new fce5154  Fix typo in formula in docstring for GRU cell and layer and add clarification to description (gluon.rnn) (#12896)
     new 38e32bd  fix the paths issue for downloading script (#12913)
     new 0874677  Ignore generated scala files. (#12928)
     new 6b4df85  use ResourceScope in Model/Trainer/FeedForward.scala (#12882)
     new 7d0f7d6  Disabled flaky test: test_gluon_gpu.test_slice_batchnorm_reshape_batchnorm (#12768)
     new 58d4efb  Added unit tests for Resource Scope in Java (#12955)
     new af55104  Fix the operator API documentation (#12942)
     new f759984  Bumping down minimum java support from 8 to 7 (#12965)
     new 57176cd  fix indpt[0] for take(csr) (#12927)
     new 1b96fc9  getnnz operator  for CSR matrix (#12908)
     new 50028e9  fix broken docs (#12871)
     new 96df2c5  Add bytearray support back to imdecode (#12855, #12868) (#12912)
     new d93467e  Update tree lstm example (#12960)
     new a39152b  Update bilstm integer array sorting example (#12929)
     new ffe551e  Fix the bug of assigning large integer to NDArray (#12921)
     new fef9b5c  Refactor mkldnn test files (#12410)
     new ee5f699  CudnnFind() usage improvements (#12804)
     new 5aaa729  [MXNET-984] Java NDArray Documentation Generation (#12835)
     new ffeaf31  fix mac r install and windows python build from source docs (#12919)
     new 743301c  First pass at adding JavaDocs for new java api classes (#12963)
     new 2f6d224  enable batchnorm unit tests (#12986)
     new 148819b  Update (#12996)
     new 1555735  fix Sphinx errors for tutorials and install ToCs (#12945)
     new 1f971c2  [MXNET -1030] Cosine Embedding Loss (#12750)
     new a362df1  [MXNET-1173] Debug operators - isfinite, isinf and isnan (#12967)
     new 7e776c9  [MXNET-1160] add Java build/run example (#12969)
     new 12371fb  [MXNET-1111] Remove CPUPinned in ImageRecordIter (#12666)
     new a9ac9d1  Added/changed file_name, brief description comments in some files (#13033)
     new afbb72f  sample_like operators (#13034)
     new 7b787d3  [MXNET-1179] Enforce deterministic algorithms in convolution layers (#12992)
     new 71d6621  Add a deprecate message (#13042)
     new 03c4020  Disable flaky test test_operator.test_dropout (#13057)
     new 8209712  Disable flaky test test_prelu (#13060)
     new 49e6a7e  la_op_inline.h to la_op-inl.h for consistency (#13045)
     new 0844d4b  Improve clojure tutorial (#12974)
     new 13ad595  removed unused header (#13066)
     new 1e16cf6  Moves f16c autodetection to its own cmake module (#12331)
     new 5bd6f10  Set correct update on kvstore flag in dist_device_sync mode (#12786)
     new 0bea50e  ONNX export: Cleanup (#12878)
     new d242f3b  Maven Surefire bug workaround (#13081)
     new 24d4903  remove legacy installation of Roxygen2 5.0 and add R-specific clean target (#12993) (#12998)
     new 1ccffde  Gluon LSTM Projection and Clipping Support (#13056)
     new 4830744  fix readme (#13082)
     new 6535c06  [MXNET-1180] Scala Image API (#12995)
     new b78265b  [MXNET-793] Virtual testing with Qemu, refinement and extract test results to root MXNet folder (#13065)
     new 3e8a976  Updated / Deleted some examples (#12968)
     new 50f43f0  Fix variable name in tutorial code snippet (#13052)
     new 62d2800  Maven Surefire bug workaround (#13097)
     new 31ebb95  customized take forward for CPU (#12997)
     new 976fd00  Update module example (#12961)
     new f5cef48  ONNX export: Scalar, Reshape - Set appropriate tensor type (#13067)
     new 974a04c  Fix example for mxnet.nd.contrib.cond and fix typo in src/engine (#12954)
     new a91b364  Improve the Clojure Package README to Make it Easier to Get Started (#12881)
     new 53c5a72  [MXNET-918] Introduce Random module / Refact code generation (#13038)
     new b645f41  Fix a typo in operator guide (#13115)
     new 8f6efe3  [Issue #11912] throw mxnet exceptions when decoding invalid images. (#12999)
     new 722ad7a  Update adversary attack generation example (#12918)
     new 94676b0  Disable travis tests (#13137)
     new c169b14  Update Gluon example folder (#12951)
     new 91c1f82  Document the newly added env variable (#13049)
     new 6e6663b  Updated to include mxnet-label-bot  (#13048)
     new f8052e4  Fix docker cleanup race condition (#13092)
     new 6c4bbd8  Improved git reset for CI builds (#12784)
     new 36eabfa  Refactor L2_normalization (#13059)
     new 466c488  Fix variational autoencoder example (#12880)
     new 2df7a61  use ResourceScope in Model/Trainer/FeedForward.scala (#12882) (#13164)
     new 088ae79  Add documentation on GPU performance on Quantization example (#13145)
     new 5d6a7ac  [MXNET-1194] Reenable nightly tutorials tests for Python2 and Python3 (#13099)
     new 498e03d  Update dec example (#12950)
     new a32fa84  Update MKL-DNN dependency (#12953)
     new e9a590f  License header (#13178)
     new 68bc9b7  Updated capsnet example (#12934)
     new 012288f  Updates to several examples (#13068)
     new 149ea17  [MXNET-1187] Added Tutorial for Java under (#13183)
     new 55ee9b3  Fix Sphinx python docstring formatting error. (#13177)
     new d424c0e  [Doc] Fix repo paths in Ubuntu build doc (#13101)
     new 99534c9  Update scala intellij tutorial (#12827)
     new 7f1d53e  Improve cpp-package example project build files. (#13093)
     new 3bbbf6d  Fix Sphinx document parsing error. (#13195)
     new 3a6dcc7  Fix #13090, Add image.imread to python API doc. (#13176)
     new 0166793  Fix Sphinx docstring formatting error. (#13004, #13005, #13006) (#13175)
     new 34d24b4  Fix #12944, Fix Sphinx python docstring formatting error. (#13174)
     new 24b8de0  Fix #13013, Fix Sphinx python docstring error. (#13173)
     new 89f3091  update the README (#13186)
     new 6ab06e6  Fixed Sparse astype doc string formatting error (#13171)
     new 05dce1c  Fix problem with some OSX not handling the cast on imDecode (#13207)
     new b441628  Port of scala Image API to clojure (#13107)
     new 517ede1  Fixed Documentation issues (#13215)
     new c18ad7d  update the doc (#13205)
     new e402242  Fix Sphinx doc errors (#13170)
     new f971d64  Fix Sphinx python docstring error: initializer.InitDesc (#12939) (#13148)
     new 042472a  Fix Sphinx python docstring error: text contrib module (#12949) (#13149)
     new 2e4d6c8  Sphinx failure fixes (#13213)
     new fd3dedc  [MXNET-793] Virtualized ARMv7 with Qemu CI integration (#13203)
     new d8d2d6e  Refactor kvstore test (#13140)
     new 6c82829  remove unused variable rotateM_ (#10803)
     new 3c5fa16  Revert "Sphinx failure fixes" (#13230)
     new c00697c  [MXNET-953] Fix oob memory read (#12631)
     new 3664a7c  [MXNET-1202] Change Builder class into a better way (#13159)
     new 1bb5b7f  [MXNET-1041] Add Java benchmark (#13095)
     new e1f221b  update log4j version of Scala package (#13131)
     new 62dcc44  Disable Flaky test test_operator.test_clip (#12902)
     new 3cebaa2  Update multi-task learning example (#12964)
     new 1822cef  Update MKLML dependency (#13181)
     new 8b38528  Add --no-cache option to when building containers (#13182)
     new bdb4443  Tool to ease compilation and reproduction of test results (#13202)
     new 8c74974  support for upper triangular matrices in linalg (#12904)
     new 6d5e3cb  Fix Sphinx python docstrings (#13160)
     new 1cc8629  Implemented a regression unit test for #11793 (#12975)
     new ea6ee0d  Add Java API docs generation (#13071)
     new 7baad6f  Fix Sphinx error in ONNX file (#13251)
     new fb4cad9  [MXNET-918] [Introduce Random module / Refact code generation (#13038)][Cherry pick]  (#13242)
     new efd925e  Merge branch 'master' into java-api
     new e655f62  [Example] Fixing Gradcam implementation (#13196)
     new 7dfcc94  Fix test failure due to hybridize call in test_gluon_rnn.test_layer_fill_shape (#13043)
     new f79bb18  Addressed sphinx build issue (#13246)
     new f5ba267  Add gauss err function operator (#13229)
     new 6b39c6b  Fixed missing break statement (#13257)
     new b8e36e0  Add Turing and Volta support to arch_name (#13168)
     new 2eb76b5  Bugfix in ci/ (#13249)
     new 8cb73ef  Fix scaladoc build errors (#13189)
     new ead3af2  Add missing documentations for getnnz (#13128)
     new 100a4aa  Addressed ONNX module documentation warnings and added notes for short-form representation (#13259)
     new 6f940cf  Java Benchmark failure (#13258)
     new 7541021  Manually track num_max_thread (#12380)
     new 218a7a9  Addressing PR feedback for merging Java API into master (#13277)
     new cf991ff  adding unit test for MKLDNN FullyConnected operator (#12985)
     new 339e085  Doc fixes (#13256)
     new 52bead0  clean up the NDArray follow the comments (#13281)
     new 1ef83c9  fix train mnist for inception-bn and resnet (#13239)
     new e7f9770  Fix a bug in index_copy (#13218)
     new 0259254  Addressed doc issues (#13165)
     new 7d51241  [MXNET-1181] Added command line alternative to IntelliJ in install instructions (#13267)
     new 226f9cb  Force APT cache update before executing install (#13285)
     new 8ac7fb9  [Example] Gradcam consolidation in tutorial (#13255)
     new 97fdfd9  [MXNET-1203] Tutorial infogan  (#13144)
     new c78f89f  Remove obsolete memory cost example (#13235)
     new 3ec9030  add defaults and clean up the tests (#13295)
     new f52b9aa  [MXNET-1187] Added Java SSD Inference Tutorial for website (#13201)
     new 1c54aaa  Merge branch 'master' into java-api
     new bb7bbaf  [MXNET-1182] Predictor example (#13237)
     new d73e4df  stop gap fix to let website builds through; scaladoc fix pending (#13298)
     new 5369605  Fix Sphinx errors in box_nms (#13261)
     new be5bac3  Fix Sphinx errors (#13252)
     new 20a23ef  Sphinx errors in Gluon (#13275)
     new 21fc3af  Fix Sphinx python docstring formatting error. (#13194)
     new 5af975e  Visualization doc fix. Added notes for shortform (#13291)
     new 6ae5b65  Addressed "dumplicate object reference" issues (#13214)
     new ab8772c  Reducing the length of setup tutorial (#13306)
     new efff66b  Update (#13299)
     new 2983185  add url and license to clojure package project (#13304)
     new d3f5340  [Example] Add docstring for test optimizer and test score (#13286)
     new 2775006  [Example] Update cpp example README (#13280)
     new 96a2a09  [Example]update NER example readme on module prediction (#13184)
     new 0e9a1ff  [MXNET-1198] MXNet Java API (#13162)
     new 70f4e2c  enabling test_dropout after fixing flaky issue (#13276)
     new 46e870b  fix the flag (#13293)
     new 1aa6a38  Made fixes to and (#13305)
     new dfeb6b0  Fix descriptions in scaladocs for macro ndarray/sybmol APIs (#13210)
     new 3fc4106  [Example] Gradcam- Fixing a link (#13307)
     new ac57ce3  Updated the Instructions for use of the label bot (#13192)
     new dc3648b  [MXNET-33] Enhance mkldnn pooling to support full convention (#11047)
     new f87db9e  [MXNET-1213] add Cent OS build for Scala (#13279)
     new 98830d5  fix file lock issue (#13296)
     new 64657c2  modify code for working in gpu context. (#13302)
     new 5a83b6b  [MXNET-860] Use modernized ranged loops where possible (#12356)
     new 91c536d  [MXNET-483] C++ tests for mkldnn convolution/deconvolution operator (#11778)
     new b8d1c02  Fix ONNX export of keepdims param (#12924)
     new cf6ae59  adding unittest for MKLDNN Softmax operator (#12884)
     new 779bdc5  [MXNET-1215] Allow dynamic shape exists in imperative mode (#13283)
     new 5042ec3  Fix/link to mkl build (#13268)
     new de5267f  Update Requests version to 2.20.0 to handle security risk (#13266)
     new f838bb5  Fix nightly test issue and add docs build settings for 1.3.x (#13323)
     new daf3529  Revert "[MXNET-1198] MXNet Java API (#13162)"
     new 4f481f5  [Example]Refactor alexnet cpp example (#13278)
     new 8e888c1  [Example] Update C++ tutorials (#13316) (#13317)
     new afc4703  CMake: Do not touch CMAKE_GENERATOR_TOOLSET (#13321)
     new 2c33574  add Volta and Turing arch (#13069)
     new ef2e705  Bumped minor version to 1.4.0 as 1.3.1 will be continued in the v1.3.x branch (#13231)
     new 38eeb0c  Improve cmake flags (#13109)
     new d8030c5  Refactor Jenkinsfiles (#13344)
     new 578c547  Add documentation about reproducing test results on ARM (#13334)
     new 3c4a97d  Fix github status context if running inside a folder (#13350)
     new 3a50ae0  Fix sphinx docstring warnings. (#13315)
     new f6317c9  fix broken links and reorganize build from source page (#12962)
     new 60dd170  fix formatting for the CentOS installation instructions (#13204)
     new 1e4afd5  GEMM Tensor Core Support (#13336)
     new cfabc6b  Replace fileread with reading from stdout (#13355)
     new be76f85  adding test for softmax operator for inputs with large magnitude (#13328)
     new be26104  Add timing output to python tests (#13377)
     new cd0ce3b  Support full convention in quantized pooling (#13260)
     new aee0953  Add utility slave (#13383)
     new 388a5f4  A few operators on graphs stored as CSR (#13290)
     new 79857a4  Fixes #13386 - Refer Warnings (#13387)
     new 6b5cf4f  Updated the paths for images for java tutorial (#13361)
     new 3d64d15  Fix/env disable mkldnn cache map (#13324)
     new d412f4f  Initial website documentation for Java API (#13289)
     new b7b0d3f  Replace mxnetci dockcross with public dockcross due to missing image (#13402)
     new afa4c3a  Correct shapes of images in cifar10 and cifar100 (#13348)
     new 8b3fa78  Updated recommenders example (#13041)
     new 7b1e7a5  Improving multi-processing reliability for gluon DataLoader (#13318)
     new cb627bc  Onnx multi output (#13390)
     new bc2f13d  Change docker login (#13408)
     new e9c8db7  Fixing doc links and minor edits for Java API (#13405)
     new 14ae5d4  Fix repeated typo in mxnet_op.h (#13406)
     new f1de8e5  Use dynamic omp schedule for sparse dot with large matrix (#13398)
     new 48bfc8d  Added proper default value in cpp-package for optional<bool> (#13415)
     new 075120e  Fix infoGan Gluon tutorial errors. (#13416)
     new c542a7a  :memo: Fixes #13388 Adds Clojure to MXNet installation docs (#13393)
     new e14482d  Minor fixes to documentation (#13412)
     new ab71205  [Example] fix cpp example inception-bn and training acc issue (#13284)
     new 4f8aa09  [Example]Fix mlp_csv example (#13273)
     new 5c73fc3  Java doc (#13368)
     new e6fffe9  Adding Java to ubuntu setup install page and minor fixes to docs (#13420)
     new 8a94dbd  [MXNET-1029] Feature request: randint operator (#12749)
     new 7542b2b  Java demo file-path fix (#13358)
     new 2d1c627  Updated README and NEWS with 1.3.1 release information (#13423)
     new d6e42d8  Be more explicit about the exit status of the container (#13425)
     new 5111b18  [MKLDNN]Add quantized concat (#13297)
     new 212364b  Add ARMv7 builds to (#13432)
     new 9288527  [MXNET-1110] find the path to include header files (#13359)
     new a7d17a4  add subgraph adjacency operator. (#13396)
     new 0eacdb3  Java added to install page (#13404)
     new a29185a  1.3.1 website updates (#13444)
     new d7b9bea  remove redundant link (#13428)
     new f876461  [MXNET-886] ONNX export: HardSigmoid, Less, Greater, Equal (#12812)
     new 859d294  Correct Inception Reference for Pertained Model (#13360)
     new b34b961  exclude the error folder from sphinx toc (#13354)
     new 8b87825  Update MKL-DNN to fix LSTM perf regression (#13417)
     new dcd8950  Mitigate #13341 (#13343)
     new 6fd4384  parallelize NDArray::Copy<cpu, cpu> when data size is large (#12926)
     new e434251  fix property not updating bug (#13085)
     new 0038473  [MXNET-1222] Scala Inference enable different shapes input (#13330)
     new 8a3bd9b  Fix deconvolution  / PR 13421 (#13433)
     new c72a38b  Add DGL subgraph sampling op (#13392)
     new 07a4319  fix quantize_graph pass error when there're multiple outputs from a single node (#13000)
     new 9adf214  Get the correct include path in pip package (#13452)
     new f5d131e  Use ~/.ccache as default ccache directory so is not cache is not erased on reboot (#13431)
     new b5ea194  Skip flaky test (#13480)
     new 883d771  Rewrite dataloader with process pool, improves responsiveness and reliability (#13447)
     new 068c2c6  Fix errors in docstrings for subgraph op; use code directive (#13463)
     new 55acf56  [MXNET-1158] JVM Memory Management Documentation (#13105)
     new dabd688  Update row_sparse tutorial (#13414)
     new b58de74  Add resiliency to onnx export code (#13426)
     new 0d480fb  [MXNET-1185] Support large array in several operators (part 1) (#13418)
     new baeada4  [MXNET-1210 ] Gluon Audio - Example (#13325)
     new 77510d7  ONNX export: Instance normalization, Shape (#12920)
     new 2bc4430  Clarify dependency on OpenCV in CNN Visualization tutorial. (#13495)
     new 6990b7d  clarify ops faq regarding docs strings (#13492)
     new 4cd3355  Add graph_compact operator. (#13436)
     new 1f73c5d  Deprecate Jenkinsfile (#13474)
     new d91284b  update github location for (#13508)
     new b684c65  #13453 [Clojure] - Add Spec Validations to the Optimizer namespace (#13499)
     new 7d44deb  ONNX export: Logical operators (#12852)
     new 9979c3c  Fix cmake options parsing in dev_menu (#13458)
     new 8949989  Revert "Manually track num_max_thread (#12380)" (#13501)
     new 65edc95  Feature/mkldnn static 2 (#13503)
     new f2dcd7c  fix toctree Sphinx errors (#13489)
     new 3d499cb  Disabled flaky test test_gluon_data.test_recordimage_dataset_with_data_loader_multiworker (#13527)
     new 7dde0eb  [MXNET-1234] Fix shape inference problems in Activation backward (#13409)
     new d60f37b  Docs & website sphinx errors squished 🌦  (#13488)
     new 6ae647d  [MXNET-1235] Add a test for AdaMax optimizer (#13467)
     new 06245b1  Adadelta optimizer test (#13443)
     new f7192b2  Update java setup docs for 1.4.0 (#13536)
     new a3eca5f  Revert "Feature/mkldnn static 2 (#13503)" (#13540)
     new 0f85f5d  doc fix (#13465)
     new d2102fa  [MXAPPS-1020] Clean up some Sphinx warnings. (#13539)
     new 2f55488  [MXNET-1110] Add header files required by horovod (#13062)
     new 40db619  Bumped minor version from 1.4.0 to 1.5.0 on master, updated License file (#13478)
     new cb0db29  Fixing a 404 in the ubuntu setup doc (#13542)
     new 1c8972c  [MXNET-1249] Fix Object Detector Performance with GPU (#13522)
     new bd8e0f8  [MXNET-769] Use MXNET_HOME in a tempdir in windows to prevent access denied due t… (#13531)
     new f6f8401  Add a retry to qemu_provision (#13551)
     new f6b4665  Fix #13521 (#13537)
     new 8bbac82  Simplifications and some fun stuff for the MNIST Gluon tutorial (#13094)
     new e0ff3c3  Updated docs for randint operator (#13541)
     new cf6e8cb  Chi_square_check for discrete distribution fix (#13543)
     new 29885c5  Revert "Bumped minor version from 1.4.0 to 1.5.0 on master, updated License file" (#13558)
     new 8feb826   #13441 [Clojure] Add Spec Validations for the Random namespace (#13523)
     new f390f0c  Adding test for softmaxoutput (#13116)
     new 9c0d173  Add workspace cleaning after job finished (#13490)
     new 0011ab2  Fix flaky test test_random:test_randint_generator (#13498)
     new 7485a78  Update version to v1.5.0 including clojure package (#13566)
     new 4f61c32  License update  (#13565)
     new 7d74452  Fix use-before-assignment in convert_dot (#13511)
     new 186a746  fix the situation where idx didn't align with rec (#13550)
     new 9e3f974  Update MXNetTutorialTemplate.ipynb (#13568)
     new 636933d  ONNX import/export: Size (#13112)
     new 95f1e1c  fix link for gluon model zoo (#13583)
     new 7d2b804  Fix exception handling api doc (#13519)
     new f2ca66f  add cpp example inception to nightly test (#13534)
     new ba02bf2  Add notes about debug with libstdc++ symbols (#13533)
     new 75d1d4f  Add imresize and copyMakeBorder to mx.image (#13357)
     new 46a2990  [MXNET-1253] fix control_flow_op (#13555)
     new e25e18f  Add Intel MKL blas to Jenkins (#13607)
     new 449e17d   #13385 [Clojure] - Turn examples into integration tests (#13554)
     new 1f8bb26  fix the Float not showing correctly problem (#13617)
     new a4c97ec  [MXNET-1155] Add scala packageTest utility (#13046)
     new b242b0c  [MXNET-1224]: improve scala maven jni build and packing. (#13493)
     new 97e0c97  [MXNET-1225] Always use in make install instructions (#13364)
     new 9ce7eab  Fix warning in waitall doc (#13618)
     new 002e0bb  Optimize C++ API (#13496)
     new 25be62f  Merge branch 'java-api' into master
     new e36f888  fix quantize pass error when the quantization supported Op are excluded in the model (#13596)
     new d1e7b81  Merge branch 'revert-and-merge'
     new 439f167  Scripts for building dependency libraries of MXNet (#13282)
     new b45e127  add batch norm test (#13625)
     new 5d2b763  Set install path for dynamic lib on Mac OS (#13629)
     new afb6703  Fix the bug of BidirectionalCell (#13575)
     new 5bcf2bd  Feature/mkldnn static (#13628)
     new 7fe0409  build config for maven and pip (#13556)
     new 090f222  Fix for import mxnet taking long time if multiple process launched (#13602)
     new c4a619c  Add reshape op supported by MKL-DNN (#12980)
     new a1cc2ba  Improve dev_menu usability, local build and virtualenv (#13529)
     new aca93d8  [Clojure] Correct the versions in the README so they correspond to the latest release (#13507)
     new 98750fe  Optimization of metric evaluation (#13471)
     new 9206c40  Revert "Feature/mkldnn static (#13628)" (#13638)
     new 30655f9  support mkl log when dtype is fp32 or fp64 (#13150)
     new 77fe96e  [MXNET-1209] Tutorial transpose reshape  (#13208)
     new 97ce687  updated reference to Apache MXNet (#13645)
     new b9b038f  Complimentary gluon DataLoader improvements (#13606)
     new 4766034  Improve CCache handling (#13456)
     new fc5fce7  [MXNET-918] Random module (#13039)
     new 67bbb49  Fix incorrect delete in MXExecutorReshape exception handling (#13376)
     new 1128f8e  [MXNET-1251] Basic configuration to do static-linking (#13621)
     new a024a90  [MXNET-1195] Cleanup Scala README file (#13582)
     new 3626fd1  scripts for building libmxnet binary and wheel (#13648)
     new ed2cb76  [MXNET-1083] Add the example to demonstrate the inference workflow using C++ API (#13294)
     new 63ae3b3  Update (#13653)
     new 1eb3344  Support Quantized Fully Connected by INT8 GEMM (#12922)
     new aa240cb  add build fix for Scala/Java build (#13655)
     new 48e25c4  Fix Jetson compilation (#13532)
     new 655f1c6  remove omp which can cause ssd accuracy variance (#13622)
     new 3433776  Revert "[MXNET-43] Fix Jetson compilation" (#13665)
     new 86d5572  Fix Jetson compilation (#13666)
     new 67a9a81  turn on Sphinx warnings as errors (#13544)
     new 5e46db4  Update CODEOWNERS, add Pedro Larroy. (#13579)
     new 8dd2fb1  Revert "Revert "[MXNET-43] Fix Jetson compilation" (#13665)" (#13672)
     new defa614  Accelerate DGL csr neighbor sampling (#13588)
     new f85b17b  [MXNET-1252][1 of 2] Decouple NNVM to ONNX from NNVM to TenosrRT conversion (#13659)
     new 61744b5  fix unpicklable transform_first on windows (#13686)
     new 8413180  Move the debug output message into MXNET_MKLDNN_DEBUG (#13662)
     new ebd5f68 backport from v1.4.x to master (#13693)
     new 59f4395  Fallback to dense version for grad(reshape), grad(expand_dims) (#13599)
     new a4f2ed5  ONNX export: Add Flatten before Gemm (#13356)
     new 80ec46c  [MXNET-1164] Generate the document for cpp-package using Doxygen (#12977)
     new 3044f8c  Update git clone location to apache github (#13706)
     new 63fe849  Add timeout/retry logic to docker cache download (#13573)
     new f211588  Fix NDArray ToDLPack Bug (#13698)
     new b018019  Added javadocs and improved example instructions (#13711)
     new e8e18c9  Rearrange tests written only for update_on_kvstore = True (#13514)
     new d18a7d5  Update mshadow to support batch_dot with fp16. (#13716)
     new 48bbac5  Fix the quantization script to support Python2 (#13700)
     new fd34dc5  ONNX test code cleanup (#13553)
     new 93b8dc1  update social media section (#13705)
     new 5ea983a  script for installing gpu libraries and build tools (#13646)
     new b6b197a  Port of scala infer package to clojure (#13595)
     new 3bfcd93  update code owner (#13737)
     new 116d01e  AdamW operator (Fixing Weight Decay Regularization in Adam) (#13728)
     new be3d945  ONNX import/export: Add missing tests, ONNX export: LogSoftMax (#13654)
     new 2f202af  ONNX test code cleanup - part 2 (#13738)
     new 8132a5b  Update (#13732)
     new 991bf3b  ONNX import: Hardmax (#13717)
     new 9fbd7dd  gluon docfix (#13631)
     new 33cd0af  Fixes for trainer with update_on_kvstore=False (#13721)
     new 812b06a  Reorder module import orders for dist-kvstore (#13742)
     new bff0fdc  CMake: Enable installation of cpp-package headers (#13339)
     new a38278d  disable error checking when building old versions (#13725)
     new d7f9a07  Integrate MKLDNN Conv1d and support 3d layout (#13530)
     new 855f8b9  Making MKL-DNN default on MXNet master (#13681)
     new e30d973  use relative links; update links (#13741)
     new fe46cd9  [MXNET-1231] Allow not using Some in the Scala operators (#13619)
     new 24068c2  fix bug in profiler tutorial when using cpu (#13695)
     new 0165ef4  local docs build feature (#13682)
     new 16cb135  make ROIAlign support position-sensitive pooling (#13088)
     new e9a7aa4  add examples and fix the dependency problem (#13620)
     new 6a4bac6  Update Adam optimizer documentation (#13754)
     new 863fb86  Less cudaGet/SetDevice calls in Gluon execution (#13764)
     new ec4ad58  Scope requests so it's not needed for dev_menu (#13771)
     new 17c8b25  Fix USE_MKLDNN check in Makefile (#13775)
     new 752cd0c  fix c complier to clang (#13778)
     new ccc2201  Fixed mailing list addresses (#13766)
     new c63ef9a  [MXNET-1255] update hybridize documentation (#13597)
     new 6dae0bf  [MXNET-244] Work around likely compiler bug on nested inlines and temporary acces… (#13535)
     new 7b7fd15  Use curl to download sample data instead of wget. (#13761)
     new 5784363  fix bipartite match memory corruption (#13727)
     new 1f151fb  remove attributs clear on TRT nodes for GetOptimizedSymbol (#13703)
     new 4d657902 Add CPU test coverage and refine cmake builds (#13338)
     new dfc2604  add license (#13793)
     new 96439e6  [MXNET-862] Basic maven jenkins pipeline (#13450)
     new 9c24df2  Re-organize Scala maven build (#13626)
     new 44c21bd  Add extra header file to export for error checking (#13795)
     new 71740ae  fix redirection issues; set default version to master (#13796)
     new 240f7ed  [MXNET-898] ONNX import/export: Sample_multinomial, ONNX export: GlobalLpPool, LpPool (#13500)
     new 8ece68c  whitelist symbols for using MXNet error handling externally (#13812)
     new ed92b8d  fix for params with no dims in onnx (#13413)
     new 9fc5382  Remove semicolon in libmxnet.sym file (#13822)
     new d3bd5e7   Clojure example for fixed label-width captcha recognition  (#13769)
     new 55ffe25  Update LICENSE File with subcomponents (#13808)
     new 30b552d  Dockerfiles for Publish Testing (#13707)
     new 5fe2172  Add new Maven build for Scala package (#13819)
     new bc39183  ONNX ops: norm exported and lpnormalization imported (#13806)
     new 39a3bc1  remove useless code (#13777)
     new 9440962  Fixing a symlink issue with R install (#13708)
     new c65b1f5  fix minor indentation (#13827)
     new 70316d6  [MXNET-880] ONNX export: Random uniform, Random normal, MaxRoiPool (#13676)
     new d973ed4  onnx export ops (#13821)
     new ed7ca26  [MXNET-1260] Float64 DType computation support in Scala/Java (#13678)
     new 17836f5  [MXNET-1263] Unit Tests for Java Predictor and Object Detector APIs (#13794)
     new 5aeed01  Fix scala doc build break for v1.3.1 (#13820)
     new abb7ce5  Remove MXNET_STORAGE_FALLBACK_LOG_VERBOSE from (#13830)
     new e632386  Add Local test stage and option to jump directly to menu item from commandline (#13809)
     new 27846dd  Removes unneeded nvidia driver ppa installation (#13814)
     new 5086da1  Improve license_header tool by only traversing files under revision c… (#13803)
     new 5282cdd  Disabled flaky test (#13758)
     new 83b9165  change to compile time (#13835)
     new 35c3383  fix Makefile for rpkg (#13590)
     new 9c3253d  [CI] Prevent timeouts when rebuilding containers with docker. (#13818)
     new a6ed619  Code modification for  testcases of various network models in directory example (#12498)
     new c4761d0  Add copyrights for third party licenses to license file (#13851)
     new 0d76675  Fix Tree Reduction on new instance type p3dn.24xlarge (#13852)
     new c2110ad  [Clojure] package infer tweaks (#13864)
     new 0e57930  Modifying clojure CNN text classification example (#13865)
     new f554835  adding tolerance to flaky test (#13850)
     new d22b323  Julia v0.7/1.0 support and drop v0.6 support (#12845)
     new bf0185c  Fix cpp examples build on Mac. (#13826)
     new 0faa5b7  Fix launch bounds in spatial transformer (#13188)
     new ac25eca  Update example scripts classpath. (#13849)
     new 22e2ef8  [MXNET-1177]Adding Scala Demo to be run as a part of Nightly CI (#13823)
     new 6be0b9a  Add CODEOWNERS for Julia package (#13872)
     new 4fe5461  fix ssd quantization script error (#13843)
     new c4b4246  Fix permissions of ci/docker/install/ (#13840)
     new bc7ea31  Avoid adding SegfaultLogger if process already has sig handler. (#13842)
     new 634d527  fix the fetching GPU problem (#13889)
     new 078a6c2  Fix SN-GAN example doc (#13877)
     new 3a97487  update Spectral Normalization Code (#13868)
     new 9d42812  Temporarily disable website testing (#13887)
     new 19764ac  Fixed java benchmark failing error by fixing the classpath (#13891)
     new 13f7c58  Jenkins nightly maven with static build script and gpu (#13767)
     new b1c9c2a  Added logging to GitHub commit status publishing (#13615)
     new e682169  Add a test for SGLD optimizer with comparisons for set noise seeds. (#13762)
     new cc15d9a  [MXNET-703] Update to TensorRT 5, ONNX IR 3. Fix inference bugs. (#13310)
     new 5b011b3  Fix the order of error term's operands (#13745)
     new 2616275  Add mkldnn OP for slice (#13730)
     new 9314689  fix bug in nag optimizer (#13683)
     new 754ff76   #13813 examples with opencv4/origami (#13813)
     new 0f7d33d  Fix BatchNorm converter for CoreML when fix_gamma=True (#13557)
     new bedf18e  beta doc fixes (#13860)
     new 86e9180  Update profiler doc (#13901)
     new d187060  Fix for test always returning true (#13911)
     new 26f5580  Add error checking for cpp examples. (#13828)
     new 2a86978  add ccache to docs build (#13832)
     new f3774c3  Java install info update (#13912)
     new 34a2dc8  Static build instruction for MXNet in general (#13914)
     new 61847be  julia: fix `argmax` for NDArray (#13871)
     new 0c85665  Support populating errors back to MXNet engine in callback (#13922)
     new 791c5d5  Fix document build (#13927)
     new d07187b  test_ImageRecordIter_seed_augmentation flaky test fix (#12485)
     new de5948b  Version switching user experience improvements (#13921)
     new 1e607a9  Julia: fix filename quoting in docstring (#13894)
     new e8a2b8b  disable default MKLDNN for cross compilation (#13893)
     new 372222b  Julia: deprecate `mx.empty`, replace it with `UndefInitializer` (#13934)
     new 692a24a  Runtime feature detection (#13549)
     new 7b9d6ed  Reduce verbosity of container builds (wget output) (#13888)
     new b8bb63d  Add back R tests and fix typo around R and perl tests (#13940)
     new 3165262  fix doc of take operator (#13947)
     new eebdd5f  #13624 clojure nightly tests (#13624)
     new b86ccf1  Add erfinv operator for calculating inverse error function (#13811)
     new 4c88f30  Update project.clj file to use the snapshots repo to be able to pull (#13935)
     new ce8b083  Julia: add windows-cpu build (#13937)
     new 45d1a1e  split_v2 (#13687)
     new 402e985  Update autoencoder example (#12933)
     new f033aec  Static build for Python (#13916)
     new a19b0b9  Flaky maven binary download (#13974)
     new 0a45e1a  Aggregate SGD (#13346)
     new 183be8c  Gradient multiplier (contrib) operator (#13632)
     new dfb397e  Update (#13973)
     new b30fba6  Fixing the doc for symbolic version of rand_zipfian (#13978)
     new 4700b40  Gluon end to end tutorial (#13411)
     new 24412df  [MXNET-1293] Adding Iterables instead of List to method signature for infer APIs in Java (#13977)
     new 49e8c57  Use CPUPinned context in ImageRecordIOParser2 (#13980)
     new 577275d  Added optional parameters to BilinearResize2D to do relative scaling (#13985)
     new f1c063e  [MXNET-1301] Remove the unnecessary WaitAll statements from inception_inference example (#13972)
     new 0f334ae  [MXNET-1000] get Ndarray real value and form it from a NDArray (#12690)
     new 5dc138d  api change (#13903)
     new 25e915b  ONNX export: Add Crop, Deconvolution and fix the default stride of Pooling to 1 (#12399)
     new 28c20fb  Sample python bilinear initializer at integral points in y-direction (#12983)
     new 3df0917  [MXNET-703] Minor refactor of TensorRT code (#13311)
     new bc98c0d  Python BucketingModule bind() with grad_req = 'add' (#13984)
     new da5242b  MXNET-1295 Adding integer index support to Sequence* family of operators. (#13880)
     new 195d4f1  Disabled flaky test test_negative_binomial_generator (#13784)
     new c701be5  Fix website error pages (#13963)
     new a82094d  build docs with CPP package (#13983)
     new 7b9779a  Update scala-package gitignore configuration. (#13962)
     new c4080de  [MXNET-1232] fix demo and add Eclipse support (#13979)
     new dbbd252  fix compile error in debug mode (#13873)
     new 3a1a80a  Image normalize operator - GPU support, 3D/4D inputs (#13802)
     new b9e02ab  remove debug; add support for v1.4.x docs; fix publish bug (#14015)
     new e37ff53   Return value docs for nd.random.* and sym.random.* (#13994)
     new 36a3cb8  Julia: split ndarray.jl into several snippets (#14001)
     new c939c2d  float32 -> float16 cast consistency across implementations (#13857)
     new 89c7d57  Improve bulking in Gluon (#13890)
     new 47277b0  Fix MXNet R package build (#13952)
     new 8102606  Fix inconsistent handling for FResourceRequestEx for imperative and symbolic executor (#14007)
     new 9a3e4a0  [MXNET-1180] Java Image API (#13807)
     new 2a4634b  Export resize and support batch size (#14014)
     new 439377d  add NAG optimizer to r api (#14023)
     new f95e794  Now passing DType of Label downstream to Label's DataDesc object (#14038)
     new 32c7f63  Update NOTICE

The 9238 revisions listed above as "new" are entirely new to this
repository and will be described in separate emails.  The revisions
listed as "add" were already present in the repository and have only
been added to this reference.

View raw message