Merge pull request #6434 from notoraptor/prepare-release-0.10.0beta3

.jenkins/jenkins_pr_core.sh

@@ -11,6 +11,10 @@ rsync -a $HOME/cache/ $HOME/.theano/pr_theano
 
 echo "===== Testing theano core"
 
+# Allow subprocess created by tests to find Theano.
+# Keep it in the workspace
+export PYTHONPATH=$PYTHONPATH:${WORKSPACE}
+
 # Test theano core
 PARTS="theano -e gpuarray"
 THEANO_PARAM="${PARTS} --with-timer --timer-top-n 10 --with-xunit --xunit-file=theanocore_tests.xml"

.jenkins/jenkins_pr_gpu.sh

@@ -51,6 +51,10 @@ export PYTHONPATH=${PYTHONPATH}:$LIBDIR/lib/python
 
 python -c 'import pygpu; print(pygpu.__file__)'
 
+# Allow subprocess created by tests to find Theano.
+# Keep it in the workspace
+export PYTHONPATH=$PYTHONPATH:${WORKSPACE}
+
 # Testing theano (the gpuarray parts)
 THEANO_GPUARRAY_TESTS="theano/gpuarray/tests \
                        theano/scan_module/tests/test_scan.py:T_Scan_Gpuarray \

HISTORY.txt

@@ -5,6 +5,133 @@
 Old Release Notes
 =================
 
+=============
+Release Notes
+=============
+
+Theano 0.10.0beta2 (7th of September, 2017)
+===========================================
+
+This release contains new features, improvements and bug fixes to prepare the upcoming release candidate.
+
+We recommend that every developer updates to this version.
+
+Highlights:
+ - Support NumPy ``1.13``
+ - Support pygpu ``0.7``
+ - Added conda recipe
+ - Optional faster optimization step with new destroy handler
+ - Added documentation for RNNBlock
+ - Bug fixes, crash fixes, warning improvements and documentation updates
+
+A total of 67 people contributed to this release since 0.9.0, see list below.
+
+Interface changes:
+ - Added new parameter ``target`` for MRG functions
+
+Convolution updates:
+ - Added unshared convolutions
+ - Added 3D separable convolutions
+ - Added 3D grouped convolutions
+ - Removed old ``conv3d`` interface
+ - Deprecated old ``conv2d`` interface
+ - Updated ``conv`` documentation
+
+GPU:
+ - Added a meta-optimizer to select the fastest GPU implementations for convolutions
+
+ - cuDNN:
+
+   - Official support for ``v6.*`` and ``v7.*``, support for ``v5.*`` will be removed in next release
+   - Added spatial transformation operation based on cuDNN
+   - Updated and improved caching system for runtime-chosen cuDNN convolution algorithms
+   - Support cuDNN v7 tensor core operations for convolutions with runtime timed algorithms
+   - Restricted cuDNN reductions to contiguous inputs
+   - Automatic addition of cuDNN DLL path to ``PATH`` environment variable on Windows
+
+New features:
+ - Added ``tensor6()`` and ``tensor7()`` in ``theano.tensor`` module
+ - Added boolean indexing for sub-tensors
+ - Added covariance matrix function ``theano.tensor.cov``
+ - Added new Theano flag ``pickle_test_value`` to help disable pickling test values
+
+Others:
+ - Kept stack trace for optimizations in new GPU backend
+
+Other more detailed changes:
+ - Moved all C code files into separate folder ``c_code`` in every Theano module
+ - Improvements for Jenkins tests
+
+Commiters since 0.9.0:
+ - Frederic Bastien
+ - João Victor Tozatti Risso
+ - Arnaud Bergeron
+ - Steven Bocco
+ - Mohammed Affan
+ - amrithasuresh
+ - Pascal Lamblin
+ - Reyhane Askari
+ - Alexander Matyasko
+ - Simon Lefrancois
+ - Shawn Tan
+ - Gijs van Tulder
+ - Thomas George
+ - Vikram
+ - Andrei Costinescu
+ - Faruk Ahmed
+ - Boris Fomitchev
+ - Zhouhan LIN
+ - Aleksandar Botev
+ - jhelie
+ - xiaoqie
+ - Tegan Maharaj
+ - Matt Graham
+ - Cesar Laurent
+ - Gabe Schwartz
+ - Juan Camilo Gamboa Higuera
+ - Tim Cooijmans
+ - Anirudh Goyal
+ - Saizheng Zhang
+ - vipulraheja
+ - Florian Bordes
+ - Sina Honari
+ - Yikang Shen
+ - erakra
+ - Chiheb Trabelsi
+ - Shubh Vachher
+ - Daren Eiri
+ - Joseph Paul Cohen
+ - Laurent Dinh
+ - Mohamed Ishmael Diwan Belghazi
+ - Jeff Donahue
+ - Ramana Subramanyam
+ - Bogdan Budescu
+ - Dzmitry Bahdanau
+ - Ghislain Antony Vaillant
+ - Jan Schlüter
+ - Xavier Bouthillier
+ - fo40225
+ - Aarni Koskela
+ - Adam Becker
+ - Adam Geitgey
+ - Adrian Keet
+ - Adrian Seyboldt
+ - Anmol Sahoo
+ - Chong Wu
+ - Holger Kohr
+ - Jayanth Koushik
+ - Lilian Besson
+ - Lv Tao
+ - Michael Manukyan
+ - Murugesh Marvel
+ - NALEPA
+ - Zotov Yuriy
+ - dareneiri
+ - lrast
+ - morrme
+ - wyjw
+
+
 Theano 0.10.0beta1 (9th of August, 2017)
 ========================================
 

NEWS.txt

@@ -3,58 +3,41 @@ Release Notes
 =============
 
 
-Theano 0.10.0beta2 (7th of September, 2017)
-===========================================
+Theano 0.10.0beta3 (20th of September, 2017)
+============================================
 
 This release contains new features, improvements and bug fixes to prepare the upcoming release candidate.
 
 We recommend that every developer updates to this version.
 
 Highlights:
- - Support NumPy ``1.13``
- - Support pygpu ``0.7``
- - Added conda recipe
- - Optional faster optimization step with new destroy handler
- - Added documentation for RNNBlock
- - Bug fixes, crash fixes, warning improvements and documentation updates
+ - conda packages now available and updated in our own conda channel ``mila-udem``.
+   To install it: ``conda install -c mila-udem/label/pre theano pygpu``
 
-A total of 67 people contributed to this release since 0.9.0, see list below.
+ - Improved elemwise operations
 
-Interface changes:
- - Added new parameter ``target`` for MRG functions
+   - Speed-up elemwise ops based on SciPy
+   - Fixed memory leak related to elemwise ops on GPU
 
-Convolution updates:
- - Added unshared convolutions
- - Added 3D separable convolutions
- - Added 3D grouped convolutions
- - Removed old ``conv3d`` interface
- - Deprecated old ``conv2d`` interface
- - Updated ``conv`` documentation
+ - Improved pickling and tests in debug mode
+ - Fixed pygpu detection
+ - Bug fixes, crash fixes, warning improvements and documentation updates
 
-GPU:
- - Added a meta-optimizer to select the fastest GPU implementations for convolutions
+A total of 69 people contributed to this release since 0.9.0, see list below.
 
- - cuDNN:
+Interface changes:
+ - Removed op ``ExtractDiag`` from ``theano.tensor.nlinalg``, now only in ``theano.tensor.basic``
 
-   - Official support for ``v6.*`` and ``v7.*``, support for ``v5.*`` will be removed in next release
-   - Added spatial transformation operation based on cuDNN
-   - Updated and improved caching system for runtime-chosen cuDNN convolution algorithms
-   - Support cuDNN v7 tensor core operations for convolutions with runtime timed algorithms
-   - Restricted cuDNN reductions to contiguous inputs
-   - Automatic addition of cuDNN DLL path to ``PATH`` environment variable on Windows
+Convolution updates:
+ - Added dilated causal convolutions for 2D
 
 New features:
- - Added ``tensor6()`` and ``tensor7()`` in ``theano.tensor`` module
- - Added boolean indexing for sub-tensors
- - Added covariance matrix function ``theano.tensor.cov``
- - Added new Theano flag ``pickle_test_value`` to help disable pickling test values
+ - Added ``unravel_index`` and ``ravel_multi_index`` functions on CPU
+ - Implemented ``max()`` and ``min()`` functions for booleans and unsigned integers types
 
 Others:
- - Kept stack trace for optimizations in new GPU backend
-
-Other more detailed changes:
- - Moved all C code files into separate folder ``c_code`` in every Theano module
- - Improvements for Jenkins tests
+ - Added ``R_op()`` for ``ZeroGrad``
+ - Added description for rnnblock
 
 Commiters since 0.9.0:
  - Frederic Bastien
@@ -68,9 +51,9 @@ Commiters since 0.9.0:
  - Alexander Matyasko
  - Simon Lefrancois
  - Shawn Tan
+ - Vikram
  - Gijs van Tulder
  - Thomas George
- - Vikram
  - Andrei Costinescu
  - Faruk Ahmed
  - Boris Fomitchev
@@ -86,10 +69,10 @@ Commiters since 0.9.0:
  - Tim Cooijmans
  - Anirudh Goyal
  - Saizheng Zhang
+ - Yikang Shen
  - vipulraheja
  - Florian Bordes
  - Sina Honari
- - Yikang Shen
  - erakra
  - Chiheb Trabelsi
  - Shubh Vachher
@@ -103,8 +86,10 @@ Commiters since 0.9.0:
  - Dzmitry Bahdanau
  - Ghislain Antony Vaillant
  - Jan Schlüter
+ - Nan Jiang
  - Xavier Bouthillier
  - fo40225
+ - wyjw
  - Aarni Koskela
  - Adam Becker
  - Adam Geitgey
@@ -123,4 +108,4 @@ Commiters since 0.9.0:
  - dareneiri
  - lrast
  - morrme
- - wyjw
+ - naitonium

NEWS_DEV.txt

@@ -5,7 +5,7 @@ DRAFT Release Notes
 ===================
 
 git log -p rel-0.9.0... |grep Merge|grep '#[0123456789]' |cut -f 8 -d ' ' | sed 's\#\* https://github.com/Theano/Theano/pull/\'
-git log -p rel-0.10.0beta1... |grep Merge|grep '#[0123456789]' |cut -f 8 -d ' ' | sed 's\#\* https://github.com/Theano/Theano/pull/\'
+git log -p rel-0.10.0beta2... |grep Merge|grep '#[0123456789]' |cut -f 8 -d ' ' | sed 's\#\* https://github.com/Theano/Theano/pull/\'
 
 # Commit count per user
 git shortlog -sn rel-0.9.0..
@@ -19,6 +19,8 @@ TODO: better Theano conv doc
 # NB: Following notes contains infos since 0.9.0.
 
 Highlights:
+ - conda packages now available and updated in our own conda channel ``mila-udem``.
+   To install it: ``conda install -c mila-udem/label/pre theano pygpu``
  - Support NumPy ``1.13``
  - Support pygpu ``0.7``
  - Added conda recipe
@@ -28,6 +30,11 @@ Highlights:
  - Removed old GPU backend ``theano.sandbox.cuda``. New backend ``theano.gpuarray`` is now the official GPU backend
  - Support more debuggers for ``PdbBreakpoint``
 
+ - Improved elemwise operations
+
+   - Speed-up elemwise ops based on SciPy
+   - Fixed memory leak related to elemwise ops on GPU
+
  - Scan improvements
 
    - Speed up Theano scan compilation and gradient computation
@@ -40,6 +47,7 @@ Highlights:
 
    - Added documentation for RNNBlock
 
+ - Improved pickling and tests in debug mode
  - Many bug fixes, crash fixes and warning improvements
 
 Interface changes:
@@ -60,6 +68,7 @@ Interface changes:
    - ``pycuda.init``
 
  - Changed ``grad()`` method to ``L_op()`` in ops that need the outputs to compute gradient
+ - Removed op ``ExtractDiag`` from ``theano.tensor.nlinalg``, now only in ``theano.tensor.basic``
 
 Convolution updates:
  - Removed old ``conv3d`` interface
@@ -69,6 +78,7 @@ Convolution updates:
  - Added unshared convolutions
  - Implemented separable convolutions for 2D and 3D
  - Implemented grouped convolutions for 2D and 3D
+ - Added dilated causal convolutions for 2D
  - Automatic addition of cuDNN DLL path to ``PATH`` environment variable on Windows
 
 GPU:
@@ -105,6 +115,7 @@ GPU:
    - Started to use ``float32`` precision for computations that don't support ``float16`` on GPU
 
 New features:
+ - Implemented ``max()`` and ``min()`` functions for booleans and unsigned integers types
  - Added ``tensor6()`` and ``tensor7()`` in ``theano.tensor`` module
  - Added boolean indexing for sub-tensors
  - Added covariance matrix function ``theano.tensor.cov``
@@ -113,6 +124,7 @@ New features:
  - Added Scaled Exponential Linear Unit (SELU) activation
  - Added sigmoid_binary_crossentropy function
  - Added tri-gamma function
+ - Added ``unravel_index`` and ``ravel_multi_index`` functions on CPU
  - Added modes ``half`` and ``full`` for ``Images2Neibs`` ops
  - Implemented gradient for ``AbstractBatchNormTrainGrad``
  - Implemented gradient for matrix pseudoinverse op
@@ -129,6 +141,8 @@ Others:
  - Kept stack trace for optimizations in new GPU backend
  - Added deprecation warning for the softmax and logsoftmax vector case
  - Added a warning to announce that C++ compiler will become mandatory in next Theano release ``0.11``
+ - Added ``R_op()`` for ``ZeroGrad``
+ - Added decsription for rnnblock
 
 Other more detailed changes:
  - Removed useless warning when profile is manually disabled
@@ -143,6 +157,29 @@ Other more detailed changes:
  - Many improvements for Jenkins CI tests: daily testings on Mac and Windows in addition to Linux
 
 ALL THE PR BELLOW HAVE BEEN CHECKED
+* https://github.com/Theano/Theano/pull/6005
+* https://github.com/Theano/Theano/pull/6427
+* https://github.com/Theano/Theano/pull/6424
+* https://github.com/Theano/Theano/pull/6419
+* https://github.com/Theano/Theano/pull/6415
+* https://github.com/Theano/Theano/pull/6418
+* https://github.com/Theano/Theano/pull/5891
+* https://github.com/Theano/Theano/pull/6316
+* https://github.com/Theano/Theano/pull/6331
+* https://github.com/Theano/Theano/pull/6100
+* https://github.com/Theano/Theano/pull/6412
+* https://github.com/Theano/Theano/pull/6221
+* https://github.com/Theano/Theano/pull/6386
+* https://github.com/Theano/Theano/pull/6411
+* https://github.com/Theano/Theano/pull/6405
+* https://github.com/Theano/Theano/pull/6410
+* https://github.com/Theano/Theano/pull/6413
+* https://github.com/Theano/Theano/pull/6389
+* https://github.com/Theano/Theano/pull/6409
+* https://github.com/Theano/Theano/pull/6406
+* https://github.com/Theano/Theano/pull/6396
+* https://github.com/Theano/Theano/pull/6392
+* https://github.com/Theano/Theano/pull/6393
 * https://github.com/Theano/Theano/pull/6384
 * https://github.com/Theano/Theano/pull/6326
 * https://github.com/Theano/Theano/pull/6317

doc/index.txt

@@ -21,6 +21,8 @@ learning/machine learning <https://mila.umontreal.ca/en/cours/>`_ classes).
 News
 ====
 
+* 2017/09/20: Release of Theano 0.10.0beta3, new features and many bugfixes, release candidate to coming.
+
 * 2017/09/07: Release of Theano 0.10.0beta2, new features and many bugfixes, release candidate to coming.
 
 * 2017/08/09: Release of Theano 0.10.0beta1, many improvements and bugfixes, release candidate to coming.

theano/tensor/tests/test_mpi.py

@@ -8,6 +8,8 @@ from theano.gof.sched import sort_schedule_fn
 
 from theano import change_flags
 
+from nose.plugins.skip import SkipTest
+
 mpi_scheduler = sort_schedule_fn(*mpi_cmps)
 mpi_linker = theano.OpWiseCLinker(schedule=mpi_scheduler)
 mpi_mode = theano.Mode(linker=mpi_linker)
@@ -41,7 +43,7 @@ def test_can_make_function():
 
 def test_mpi_roundtrip():
     if not mpi_enabled:
-        return
+        raise SkipTest('MPI not enabled')
     theano_root = theano.__file__.split('__init__')[0]
     p = subprocess.Popen("mpiexec -np 2 python " + theano_root +
                          "tensor/tests/_test_mpi_roundtrip.py",