forked from Theano/Theano
-
Notifications
You must be signed in to change notification settings - Fork 0
/
HISTORY.txt
1775 lines (1566 loc) · 94.4 KB
/
HISTORY.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
.. _HISTORY:
=================
Old Release Notes
=================
Theano 0.7 (26th of March, 2015)
================================
We recommand to everyone to upgrade to this version.
Highlights:
* Integration of cuDNN for 2D convolutions and pooling on supported GPUs
* Too many optimizations and new features to count
* Various fixes and improvements to scan
* Better support for GPU on Windows
* On Mac OS X, clang is used by default
* Many crash fixes
* Some bug fixes as well
Theano 0.6 (December 3th, 2013)
===================================
We recommend that everybody update to this version.
Highlights (since 0.6rc5):
* Last release with support for Python 2.4 and 2.5.
* We will try to release more frequently.
* Fix crash/installation problems.
* Use less memory for conv3d2d.
0.6rc4 skipped for a technical reason.
Highlights (since 0.6rc3):
* Python 3.3 compatibility with buildbot test for it.
* Full advanced indexing support.
* Better Windows 64 bit support.
* New profiler.
* Better error messages that help debugging.
* Better support for newer NumPy versions (remove useless warning/crash).
* Faster optimization/compilation for big graph.
* Move in Theano the Conv3d2d implementation.
* Better SymPy/Theano bridge: Make an Theano op from SymPy expression and use SymPy c code generator.
* Bug fixes.
Change from 0.6rc5:
* Fix crash when specifing march in cxxflags Theano flag. (Frederic B., reported by FiReTiTi)
* code cleanup (Jorg Bornschein)
* Fix Canopy installation on windows when it was installed for all users: Raingo
* Fix Theano tests due to a scipy change. (Frederic B.)
* Work around bug introduced in scipy dev 0.14. (Frederic B.)
* Fix Theano tests following bugfix in SciPy. (Frederic B., reported by Ziyuan Lin)
* Add Theano flag cublas.lib (Misha Denil)
* Make conv3d2d work more inplace (so less memory usage) (Frederic B., repoted by Jean-Philippe Ouellet)
Committers since 0.5:
Frederic Bastien
Pascal Lamblin
Ian Goodfellow
Olivier Delalleau
Razvan Pascanu
abalkin
Arnaud Bergeron
Nicolas Bouchard +
Jeremiah Lowin +
Matthew Rocklin
Eric Larsen +
James Bergstra
David Warde-Farley
John Salvatier +
Vivek Kulkarni +
Yann N. Dauphin
Ludwig Schmidt-Hackenberg +
Gabe Schwartz +
Rami Al-Rfou' +
Guillaume Desjardins
Caglar +
Sigurd Spieckermann +
Steven Pigeon +
Bogdan Budescu +
Jey Kottalam +
Mehdi Mirza +
Alexander Belopolsky +
Ethan Buchman +
Jason Yosinski
Nicolas Pinto +
Sina Honari +
Ben McCann +
Graham Taylor
Hani Almousli
Ilya Dyachenko +
Jan Schlüter +
Jorg Bornschein +
Micky Latowicki +
Yaroslav Halchenko +
Eric Hunsberger +
Amir Elaguizy +
Hannes Schulz +
Huy Nguyen +
Ilan Schnell +
Li Yao
Misha Denil +
Robert Kern +
Sebastian Berg +
Vincent Dumoulin +
Wei Li +
XterNalz +
A total of 51 people contributed to this release.
People with a "+" by their names contributed a patch for the first time.
Theano 0.6rc5 (November 25th, 2013)
===================================
We recommend that everybody update to this version.
We plan to release 0.6 in one week if there is no problem introduced
with this release candidate.
Theano 0.6rc4 was skipped due to a problem with pypi
Highlights:
* Python 3.3 compatibility with buildbot test for it.
* Full advanced indexing support.
* Better Windows 64 bit support.
* New profiler.
* Better error messages that help debugging.
* Better support for newer NumPy versions (remove useless warning/crash).
* Faster optimization/compilation for big graph.
* Move in Theano the Conv3d2d implementation.
* Better SymPy/Theano bridge: Make an Theano op from SymPy expression and use SymPy c code generator.
* Bug fixes.
Committers for this rc5 only:
Frederic Bastien
Pascal Lamblin
Arnaud Bergeron
abalkin
Olivier Delalleau
John Salvatier
Razvan Pascanu
Jeremiah Lowin
Ludwig Schmidt-Hackenberg +
Vivek Kulkarni
Matthew Rocklin
Gabe Schwartz
James Bergstra
Sigurd Spieckermann +
Bogdan Budescu +
Mehdi Mirza +
Nicolas Bouchard
Ethan Buchman +
Guillaume Desjardins
Ian Goodfellow
Jason Yosinski
Sina Honari +
Ben McCann +
David Warde-Farley
Ilya Dyachenko +
Jan Schluter +
Micky Latowicki +
Yaroslav Halchenko +
Alexander Belopolsky
Hannes Schulz +
Huy Nguyen +
Robert Kern +
Sebastian Berg +
Vincent Dumoulin +
Wei Li +
XterNalz +
A total of 36 people contributed to this release.
People with a "+" by their names contributed a patch for the first time.
Installation:
* Canopy support (direct link to MKL):
* On Linux and Mac OSX (Frederic B., Robert Kern)
* On Windows (Edward Shi, Frederic B.)
* Anaconda instructions (Pascal L., Frederic B.)
* Doc Ubuntu 13.04 (Frederic B.)
* Better support of newer NumPy version(remove useless warning/crash) (Frederic B., Huy Nguyen)
Bug fixes:
* Scan: if a scan node was cloned (by theano.clone) with different inputs, and if both the initial and the cloned nodes are used in the function being compiled, the value of the outputs of one would be replaced with the outputs of the other one. (Pascal L.)
* Sparse: Disable the optimization that introduce the CSMGradC op as it doesn't work correctly with unsorted indices. (Frederic B.)
* Mac: Fix wrong result of GpuDownsampleFactorMaxGrad on Mac OSX. (Pascal L.)
* Mac: Auto-Detect and work around a bug in BLAS on MacOS X (Pascal L.)
* Mac: Work around bug in MacOS X. If 2 compiled modules had the same name, the OS or Python was not always the right one even when we used the right handle to it. (Pascal L.)
Use this hash in the Python module, and in %(nodename)s, so that different helper functions in the support code for different Ops will always have different names.
* Sparse grad: Fix ConstructSparseFromList.infer_shape (Pascal L., reported by Rami Al-Rfou')
* (introduced in the development version after 0.6rc3 release) (Frederic B.)
Reduction that upcasts the input on no axis (ex: call theano.sum() on a scalar when the original dtype isn't float64 or
[u]int64). It produced bad results as we did not upcasted the inputs in the code, we just copy them.
* Fix some cases of theano.clone() when we get a replacement of x that is a function of x. (Razvan P., reported by Akio Takano)
* Fix grad of Alloc when we unbroadcast the value and it isn't a scalar. (Frederic B., reported Ian G.)
* In some cases (I think most cases), there was an exception raised in the theano.tensor.grad() method.
But in theory, there could be bad shapes produced in the unbroadcasted dimensions.
Interface Deprecation (a warning is printed):
* The mode ProfileMode is now deprecated, use the Theano flag profile=True to replace it.
* New theano.sparse_grad() interface to get the sparse grad of a_tensor[an_int_vector]. (Frederic B.)
This can speed up the sparse computations when a small fraction of a_tensor is taken.
Deprecate the old interface for this. (Frederic B.)
Interface Changes:
* Interface change subtensor and take are not in tensor.basic anymore. They were available from tensor.* and are still available from there. (Frederic B., Matthew Rocklin)
* This lowers the basic.py size to 191k, so under 200k for github search.
* Add -m32 or -m64 in the module cache key and add the python bitwidth in the compiledir path. (Pascal L.)
* mrg.normal now has the parameter size mandatory. It was crashing with the default value of None. (Olivier D.)
* Remove the deprecated passing of multiple modes to theano function. (Frederic B.)
* Change FunctionGraph Features interface of the {on_prune(),on_import()} call back to take a reason. (Frederic B.)
* FunctionGraph now clone the input graph by default. (Frederic B.)
* Added a parameter to optionally not do this cloning.
* This was needed to speed up compilation
New Interface (reuses existing functionality):
* Add hostname as a var in compiledir_format (Frederic B.)
* Add a new Theano flag: compute_test_value_opt. It takes the same values as compute_test_value. It enables compute_test_value during Theano optimization. Only useful to debug Theano optimization. Also small changes to some optimization to work correctly in that setup. (Frederic B.)
* Add the value pdb to the Theano flag: compute_test_value and compute_test_value_opt. (Frederic B.)
* Add the Theano flag: optimizer_verbose. Default False. When True, we print all the optimization being applied.(Frederic B.)
* Add Op.c_init_code() to allow running the code when the c cmodule is imported (Pascal L.)
* Allow theano.tensor.ones(3) to support scalar and not just list of scalar as numpy.ones (Jeremiah Lowin)
* Make the memory profiler print the FLOPS used for the ops that know how to compute it. (Frederic B.)
New Features:
* Make tensor.{constant,as_tensor_variable} work with memmap. (Christian Hudon, Frederic Bastien)
* compilation work on ARM processor (Raspberry Pi, Vincent Dumoulin)
* Add numpy.random.choice wrapper to our random number generator (Sigurd Spieckermann)
* Better SymPy/Theano bridge: Make an Theano op from SymPy expression and use SymPy c code generator (Matthew Rocklin)
* Move in Theano the Conv3d2d implementation (James Bergstra, Frederic B., Pascal L.)
* First version of the new GPU back-end available (Arnaud Bergeron, Frederic B.)
* Not all Ops have been converted to this new back-end.
To use, use Theano flag device=cudaN or device=openclN, where N is a integer.
* Python 3.3 compatible (abalkin, Gabe Schwartz, Frederic B., Pascal L.)
* A new profiler (Frederic B.)
The new profiler now can profile the memory with the Theano flag profile_memory=True.
The ProfileMode now can't profile memory anymore and prints a message about it.
Now we raise an error if we try to profile when the gpu is enabled if we didn't set
correctly the env variable to force the driver to sync the kernel launch.
Otherwise the profile information are useless.
The new profiler supports the enabling/disabling of the garbage collection.
* Adds tensor.tri, tensor.triu, and tensor.tril functions that wrap Numpy equivalents (Jeremiah Lowin)
* Adds tensor.nonzero, tensor.flatnonzero functions that wrap Numpy equivalents (Jeremiah Lowin)
* Adds tensor.nonzero_values to get around lack of advanced indexing for nonzero elements (Jeremiah Lowin)
* Make {inc,set}_subtensor work on output of take. (Pascal L.)
* When device=cpu and force_device=True, force that we disable the gpu. (Frederic B.)
* Better Windows 64 bit support for indexing/reshaping (Pascal L.)
* Full advanced indexing support (John Salvatier, seberg)
* Add theano.tensor.stacklist(). Recursivly stack lists of tensors to maintain similar structure (Matthew R.)
* Add Theano flag value: on_opt_error=pdb (Olivier D.)
* GpuSoftmax[WithBias] for bigger row. (Frederic B.)
* Make Erfinv work on the GPU (Guillaume Desjardin, Pascal L.)
* Add "theano-cache basecompiledir purge" (Pascal L.)
This purges all the compiledirs that are in the base compiledir.
* A_tensor_variable.zeros_like() now supports the dtype parameter (Pascal L.)
* More stable reduce operations by default (Pascal L.)
Add an accumulator dtype to CAReduceDtype (acc_dtype)
by default, acc_dtype is float64 for float32 inputs,
then cast to specified output dtype (float32 for float32 inputs)
* Test default blas flag before using it (Pascal L.)
This makes it work correctly by default if no blas library is installed.
* Add cuda.unuse() to help tests that need to enable/disable the GPU (Frederic B.)
* Add theano.tensor.nnet.ultra_fast_sigmoid and the opt (disabled by default) local_ultra_fast_sigmoid. (Frederic B.)
* Add theano.tensor.nnet.hard_sigmoid and the opt (disabled by default) local_hard_sigmoid. (Frederic B.)
* Add class theano.compat.python2x.Counter() (Mehdi Mirza)
* Allow a_cuda_ndarray += another_cuda_ndarray for 6d tensor. (Frederic B.)
* Make the op ExtractDiag work on the GPU. (Frederic B.)
* New op theano.tensor.chi2sf (Ethan Buchman)
* Lift Flatten/Reshape toward input on unary elemwise. (Frederic B.)
This makes the "log(1-sigmoid) -> softplus" stability optimization being applied with a flatten/reshape in the middle.
* Make MonitorMode use the default optimizers config and allow it to change used optimizers (Frederic B.)
* Add support for ScalarOp.c_support_code in GpuElemwise. (Frederic B.)
* Also make the Psi function run on GPU. (Frederic B.)
* Make tensor.outer(x,y) work when ndim != 1 as numpy.outer.
* Kron op: Speed up/generalize/GPU friendly. (Frederic B.)
(It is not an op anymore, but reuses current op)
* Add gpu max for pattern (0, 1) and added all gpu max pattern for gpu min. (Frederic B.)
* Add GpuEye (Frederic B.)
* Make GpuCrossentropySoftmaxArgmax1HotWithBias and GpuCrossentropySoftmax1HotWithBiasDx work for bigger inputs (Frederic B., reported by Ryan Price)
* Finish and move out of sandbox theano.sparse.basic.true_dot (Nicolas Bouchard, Frederic B.)
And document all sparse dot variants.
* Implement the mode ignore_borders for GpuImages2Neibs (Frederic B.)
* Make many reduction functions accept a numpy scalar as axis (Jeremiah Lowin)
* Allow numpy.asarray(cuda_ndarray, dtype=...) (Frederic B.)
* theano-cache cleanup now remove cached module old version of code. (Frederic B.)
Speed-ups:
* Optimizer speed up. (Frederic B.)
* Fix warning on newer llvm version on Mac. (Pascal L., reported by Jeremiah Lowin and Chris Fonnesbeck)
* Allow pickling of more Ops to allow reusing the compiled code (Pascal L., Frederic B.)
* Optimize more cases of dot22 and scalar when we can't make a gemm (Pascal L., Frederic B.)
* Speed up GpuJoin with c code (Ludwig Schmidt-Hackenberg, Frederic B.)
* Faster GpuAdvancedIncSubtensor1 on Fermi GPU (and up) on matrix. (Vivek Kulkarni)
* Faster GPUAdvancedIncSubtensor1 in some cases on all GPU (Vivek Kulkarni)
* Implemented c_code for AdvancedSubtensor1 (abalkin)
* Add the equivalent of -march=native to g++ command line. (Frederic B., Pascal L.)
* Speed up compilation with Scan (Jan Schluter)
* Merge more Scan nodes together (Pascal L., Yao Li).
* Add MakeVector.c_code (Frederic B.)
* Add Shape.c_code (Frederic B.)
* Optimize Elemwise when all the inputs are fortran (Frederic B.)
We now generate a fortran output and use vectorisable code.
* Add ScalarOp.c_code_contiguous interface and do a default version. (Frederic B.)
This could optimize elemwise by helping the compiler generate SIMD instruction.
* Use ScalarOp.c_code_contiguous with amdlibm. (Frederic B.)
This speeds up exp, pow, sin, cos, log, log2, log10 and sigmoid when the input is contiguous in memory.
* A fix that removes a local_setsubtensor_of_allocs optimization warning and enables it in that case. (Frederic B., reported by John Salvatier)
* Make inv_as_solve optimization work (Matthew Rocklin)
Crash/no return fixes:
* Fix scan crash in the grad of grad of a scan with special structure (including scan in a scan) (Razvan P., Bitton Tenessi)
* Fix various crashes when calling scan() with inputs specified in unusual ways. (Pascal L.)
* Fix shape crash inserted by Scan optimization. The gradient of some recursive scan was making the PushOutSeqScan optimization insert crash during the execution of a Theano function. (Frederic B., reported by Hugo Larochelle)
* Fix command not returning with recent mingw64 on Windows (Pascal L., reported by many people)
* Fix infinite loop related to Scan on the GPU. (Pascal L.)
* Fix infinite loop when the compiledir is full. (Frederic B.)
* Fix a shape cycle crash in the optimizer (Pascal L., Frederic B., reported by Cho KyungHyun)
* Fix MRG normal() now allow it to generate scalars. (Pascal L.)
* Fix some GPU compilation issue on Mac (John Yani, Frederic B.)
* Fix crash when building symbolic random variables with a mix of symbolic and numeric scalar in the "size" parameter. (Pascal L., Reported by Wu Zhen Zhou)
* Make some Op.grad() implementions not return None (Pascal L.)
* Crash fix in the grad of elemwise about an DisconnectedType (Pascal L, reported by Thomas Wiecki)
* Fix local_gpu_multinomial optimization handling of broadcast information. (Frederic B., reported by Caglar)
* Fix crash with change introduced in NumPy 1.7.1 (Pascal L., reported by Thomas Wiecki)
* Compilation failure with complex (Pascal L., reported by autumncat)
* Gpu reduction on all dimensions of a 4d tensor. (Frederic B., reported by Arjun Jain)
* Fix crash for a combination of grad of dot and dimshuffle when only one of the inputs for a corresponding dimensions was knowing that it was broadcastable. (Frederic B., reported by Micky Latowicki)
* AdvancedSubtensor1: allow broadcasted index vector. (Frederic B., reported by Jeremiah Lowin)
* Fix compute_test_value for ifelse (Olivier D., reported by Bitton Tenessi)
* Fix import error with some versions of NumPy (Olivier D.)
* Fix Scan grad exception (Razvan P., reported by Nicolas BL)
* Fix compute_test_value for a non_sequence when calling the gradient of Scan (Pascal L., reported by Bitton Tenessi).
* Crash fix in Scan following interface change in 0.6rc2 (Razvan P.)
* Crash fix on Scan (Razvan P.)
* Crash fix on Scan (Pascal L., reported by Sina Honari and Sigurd)
* Fix crash in Scan gradient related to compute_test_value (Frederic B., reported by Bitton Tenessi)
* Fix a scan optimization warning/error depending of Theano flags (Frederic B.)
* Fixed crash for unimplemented elemwise gradient (Olivier D., reported by Michael McNeil Forbes)
* Fix crash in the elemwise python code for some big shape with power of 2. (Sina Honari, Pascal L.)
* Fix compile and import errors on Windows including for the GPU. (Bogdan Budescu)
* Fix GPU compilation on Windows (XterNalz)
* Fix local_abs_merge optimization crash (Pascal L., reported by Jeremiah Lowin)
* Fix import theano crash when g++ isn't there (Olivier D.)
* Fix crash related to rebuild of Theano graph (Pascal L., reported by Divine Eguzouwa)
* Fix crash during compilation (David Ward-Farley)
* Crash fix in the grad of GPU op in corner case (Pascal L.)
* Crash fix on MacOS X (Robert Kern)
* theano.misc.gnumpy_utils.garray_to_cudandarray() set strides correctly for dimensions of 1. (Frederic B., reported by Justin Bayer)
* Fix crash during optimization with consecutive sums and some combination of axis (Frederic B., reported by Caglar Gulcehre)
* Fix crash with keepdims and negative axis (Frederic B., reported by David W.-F.)
* Fix crash of theano.[sparse.]dot(x,y) when x or y is a vector. (Frederic B., reported by Zsolt Bitvai)
* Fix opt crash/disabled with ifelse on the gpu (Frederic B, reported by Ryan Price)
* Fix crash in optimization involving dot22, (Pascal L., reported by @micklat)
* Prevent shape optimizations from introducing cycles in the graph (Frederic Bastien, Pascal Lamblin, reported by Kyunghyun Cho)
Others:
* Update/Fixes/Typo/pep8 documentation and/or tutorial (Olivier D., David W.-F., Frederic B., Yaroslav Halchenko, Micky Latowicki, Ben McCann, Jason Yosinski, reported by Arnaud Bergeron)
* Doc how to make a sparse Op. (Frederic B.)
* Doc compatibility guide (abalkin)
* Fix problem in remove_constants_and_unused_inputs_scan. (useless warning and maybe slow down) (Pascal L.)
* Fix rop dot.(Razvan P., reported by Jeremiah Lowin)
* Raise better error related to pydot bug. (Frederic B., reported by Jason Yosinski and Ludwig Schmidt-Hackenberg)
* Fix to Theano tutorial examples. (reported by Ilya Dyachenko)
* Fix SharedVar.value property to make it raise an exception (Frederic B., reported by Drew Duncan)
* Fix verification with compute_test_value in grad() (Frederic B.)
* Theano flags are now evaluated lazily, only if requested (Frederic B.)
* Fix test when g++ is not avail (Frederic B.)
* Add manual instructions for OpenBLAS on Ubuntu by (Jianri Li )
* Better/more error messages (Frederic B., Pascal L., Ian Goodfellow)
* Fix Error reporting with GpuConv (Frederic B., reported by Heng Luo and Nicolas Pinto)
* Now travis-ci tests with scipy the parts that need it (Frederic B.)
* Export some functions that work on CudaNdarray for windows (Frederic B.)
* If the user specifies a -arch=sm_* value in the Theano flags for the gpu, don't add one (Frederic B., Pascal L.)
* If a C thunk returns an error, check if a python exception is set. Otherwise, set a default one (Pascal L.)
* Crash fix introduced in the development version (Wei LI)
* Added BLAS benchmark result (Frederic B., Ben McCann)
* Fix code comment (Hannes Schulz)
* More stable tests (Frederic B.)
* Add utt.asset_allclose(a, b) to have better error message. (Frederic B.)
* Better error message with compute_test_value (Frederic, reported by John Salvatier)
* Stochastic order behavior fix (Frederic B.)
* Simpler initial graph for subtensor infer shape (Olivier D.)
The optimization was doing the optimization, but this allows better reading of the graph before optimization.
* Better detection of non-aligned ndarray (Frederic B.)
* Update MRG multinomial gradient to the new interface (Mehdi Mirza)
* Implement Image2Neibs.perform() to help debug (Frederic B.)
* Remove some Theano flags from the compilation key (Frederic B.)
* Make theano-nose work on executable '\*.py' files. (Alistair Muldal)
* Make theano-nose work with older nose version (Frederic B.)
* Add extra debug info in verify_grad() (Frederic B.)
Theano 0.6rc3 (February 14th, 2013)
===================================
Highlights:
* Windows related fixes.
* Speed-ups.
* Crash fixes.
* A few small interface changes.
* GPU memory leak fix.
* A few corner cases fixes without incidence.
* More Theano determinism
* tensor.{dot,tensordot} more complete/faster/GPU friendly.
* tensor.tensordot now support Rop/Lop
* tensor.dot support n-dimensional inputs as NumPy
* To support more NumPy syntax:
* Add theano.tensor.take()
* Add a_tensor_variable.{sort,dot,std,argmin,argmax,argsort,clip,conj,conjugate,repeat,round,trace,real,imag,take}
Commiters for this rc3 only:
Frederic Bastien
Ian Goodfellow
Pascal Lamblin
Jeremiah Lowin
abalkin
Olivier Delalleau
Razvan Pascanu
Rami Al-Rfou'
Vivek Kulkarni
Guillaume Desjardins
David Warde-Farley
Eric Hunsberger
Amir Elaguizy
James Bergstra
Bug fix:
* Fix memory leak on the GPU in some corner cases with the Theano flags `allow_gc=False`. (Frederic B., reported by Jonas Gehring)
* Fix copy of random state between graph. (Guillaume D.)
http://deeplearning.net/software/theano/tutorial/examples.html#copying-random-state-between-theano-graphs
* Fix wrong dtype in sandbox.linalg.ExtractDiag with shape of 0. (Frederic B., reported by abalkin)
* Correctly support array with more then 2*10e32 element in AdvancedSubtensor1. (Abalkin)
* Fix wrong broadcast dimensions of output of Repeat op. (Abalkin)
We where using the inputs broadcasting pattern in some cases when we shouldn't.
* Fix theano.sandbox.linalg.eigh grad that didn't always returned the right dtype. (Frederic B., Olivier D.)
New Features:
* More Theano determinism (Ian G., Olivier D., Pascal L.)
* Add and use a new class OrderedSet.
* theano.grad is now deterministic.
* Warn when the user uses a (non ordered) dictionary and this causes non-determinism in Theano.
* The Updates class was non-deterministic; replaced it with the OrderedUpdates class.
* tensor.tensordot now support Rop/Lop (Jeremiah Lowin)
This remove the class TensorDot and TensorDotGrad. It is the Dot/Elemwise ops that are used.
* tensor.dot support n-dimensional inputs as NumPy (Jeremiah Lowin)
Work on the GPU too.
* The Theano flag `nvcc.flags` now accept `-ftz=true`, `--prec-div=false` and `--prec=sqrt=false` as value. (Frederic B.)
To enable all of them, use the Theano flag `nvcc.flags=--use_fast_math`.
* New op theano.sparse.ConstructSparseFromList (Rami Al-Rfou' Vivek Kulkarni)
* Make Theano work with Anaconda on Windows. (Pascal L.)
* Add tensor_var.diagonal and theano.tensor.{diag,diagonal}. (abalkin)
* AdvencedSubtensor1 can now have a sparse gradient. (Rami Al-Rfou', Vivek Kulkarni)
* Implemented GpuContiguous.grad. (Ian G.)
Interface Deprecation (a warning is printed):
* theano.misc.strutil.renderString -> render_string (Ian G.)
* Print a warning when using dictionary and this makes Theano non-deterministic.
Interface Change:
* Raise an error when theano.shared called with a theano variable. (Frederic B.)
* Don't print warning for bug before Theano 0.5 by default. (Frederic B.)
* Theano functions now always have a field name, default to None. (Frederic B.)
* Theano function fct.fgraph have a copy of the Theano function name field. (Ian G.)
This is needed to allow the fgraph to know it.
* In the grad method, if it were asked to raise an error if there is no path between the variables, we didn't always returned an error. (Ian G.)
We returned the mathematical right answer 0 in those cases.
* get_constant_value() renamed get_scalar_constant_value() and raise a new exception tensor.basic.NotScalarConstantError. (Ian G.)
* theano.function raises an error when trying to replace inputs with the 'given' parameter. (Olivier D.)
This was doing nothing, the error message explains what the user probably wants to do.
New Interface (reuse existing functionality):
* tensor_var.sort() as a shortcut for theano.tensor.sort. (Jeremiah Lowin)
We where already doing this for argsort.
* Add theano.tensor.take() and a_tensor_var.take() to support NumPy syntax. (abalkin)
* Add a_tensor_variable.{dot,std,argmin,argmax,argsort,clip,conj,conjugate,repeat,round,trace,real,imag}. (abalkin)
New debug feature:
* DebugMode print more info when there is an error. (Frederic B.)
* Better profiling of test time with `theano-nose --time-profile`. (Frederic B.)
* Detection of infinite loop with global optimizer. (Pascal L.)
* DebugMode.check_preallocated_output now also work on Theano function output. (Pascal L.)
* DebugMode will now complain when the strides of CudaNdarray of dimensions of 1 are not 0. (Frederic B.)
Speed-ups:
* c_code for SpecifyShape op. (Frederic B.)
* cross-entropy optimization now work when specify_shape is used. (Pascal L.)
* The Scan optimization ScanSaveMem and PushOutDot1 applied more frequently. (Razvan P, reported Abalkin)
A skipped optimization warning was printed.
* dot(vector, vector) now faster with some BLAS implementation. (Eric Hunsberger)
OpenBLAS and possibly others didn't call {s,d}dot internally when we called {s,d}gemv.
MKL was doing this.
* Compilation speed up: Take the compiledir lock only for op that generate c_code. (Frederic B)
* More scan optimization (Razvan P.)
* Opt to make RNN fast in Theano.
* Optimize some case of dot, by moving them outside of Scan.
* Move some sequences outside of scan too.
* Merge more scan inputs, mostly byproduct of other Scan optimizations.
* c_code for theano.sparse.AddSD. (Rami Al-Rfou', Vivek Kulkarni)
Crash Fixes:
* Fix crash about dimshuffle. (abalkin)
* Fix crash at compilation. (Olivier D.)
* Fix openmp detection. (Pascal L.)
Resulted in a crash with EPD on Windows.
* Fix for new BLAS interface in SciPy. (Olivier D.)
Fix crash with some development version of SciPy.
* GpuSum work with bigger shape when summing on the first dim on 3d tensor. (Frederic B., reported Chris Currivan)
* Windows compilation crash fix. (Frederic B.)
* Make CrossentropySoftmax1HotWithBiasDx and CrossentropySoftmaxArgmax1HotWithBias support uint* dtype. (Frederic B., reported by Mark Fenner)
* Fix GpuSoftmax and GpuSoftmaxWithBias crash on GTX285. (Frederic B.)
* Fix crash due to a race condition when importing theano. (Ian G.)
* Fix crash from path problem with `theano-nose --batch`. (Abalkin)
* Fix crash with tensor.roll(Var, iscalar). (Frederic B., reported by Jeremiah Lowin)
* Fix compilation crash with llvm on Mac. (Abalkin)
* Fix the grad of Scan that told wrongly that there is no connection between cost and parameters. (Razvan P.)
* The infer shape mechanism now force that broadcasted dimensions have a shape know to be equivalent to one during compilation.
Sometimes, we where not able knowing this before run time and resulted in crash. (Frederic B.)
* Fix compilation problems on GPU on Windows. (Frederic B.)
* Fix copy on the GPU with big shape for 4d tensor (Pascal L.)
* GpuSubtensor didn't set the stride to 0 for dimensions of 1. This could lead to check failing later that caused a crash. (Frederic B., reported by vmichals)
Theoretical bugfix (bug that won't happen with current Theano code, but if you messed with the internal, could have affected you):
* GpuContiguous, GpuAlloc, GpuDownSampleGrad, Conv2d now check the preallocated outputs strides before using it. (Pascal L.)
* GpuDownSample, GpuDownSampleGrad didn't work correctly with negative strides in their output due to problem with nvcc (Pascal L, reported by abalkin?)
Others:
* Fix race condition when determining if g++ is available. (Abalkin)
* Documentation improvements. (Many people including David W-F, abalkin, Amir Elaguizy, Olivier D., Frederic B.)
* The current GPU back-end have a new function CudaNdarray_prep_output(CudaNdarray ** arr, int nd, const int * dims) (Ian G)
Theano 0.6rc2 (November 21th, 2012)
===================================
Highlights:
* Fix for a few regressions introduced in 0.6rc1.
* A few new features.
* Speed-ups.
* Scan fixes.
* Crash fixes.
* A few small interface changes.
Commiters for this rc2 only:
Razvan Pascanu
Pascal Lamblin
Frederic Bastien
Ian Goodfellow
Jeremiah Lowin
Caglar Gulcehre
Jey Kottalam
Matthew Rocklin
abalkin
Regressions in 0.6rc1 fixed:
* Fixed the scan gradient dtype issue. In 0.6rc1, some upcast were inserted. (Razvan P.)
* Now grad() will do as before 0.6rc1 for float, i.e. the grad dtype will be the same as the inputs inside the graph. If you ask for the direct grad, it will return the computed dtype. (Pascal L.)
Wrong results fixes:
* Scan fix in some case didn't returned the good results. (Razvan P., reported by Jeremiah L.)
This happened if you had a state with only neg tap and the output of the state was a function of some sequence.
If you had multiple states, there was no problem.
* Fixed bug in Scan with multiple outputs,
where one output would sometimes overwrite another one. (Razvan P.)
* Clip.grad treated the gradient with respect to the clipping boundary as always 0. (Ian G.)
Interface changes:
* We do not support anymore unaligned ndarray in Python code. (Frederic B.)
We did not support it in C code and supporting it in Python code made
the detection harder.
* Now we only officially support SciPy 0.7.2 and NumPy 1.5.0 (Frederic B.)
We weren't and aren't testing with older versions.
* The theano.sparse.SparseType is available even when SciPy is not (Frederic B.)
* Fixed issue where members of consider_constant grad parameter
were treated differently from Constant variables. (Ian G.)
* Removed the parameter g_cost from theano.grad(). (Ian G.)
Use the new more powerful parameter known_grads instead.
NumPy interface support:
* theano.tensor.where is an alias for theano.tensor.switch to support NumPy semantic. (Ian G.)
* TensorVariable objects now have dot, argmin, argmax, clip, conj, repeat, trace, std, round,
ravel and argsort functions and the real and imag properties as numpy.ndarray objects.
The functionality was already available in Theano. (abalkin)
Speed-ups:
* A C version of the SoftMax op (Razvan P.)
There was C code for the softmax with bias code.
* Faster GpuIncSubtensor (Ian G.)
* Faster copy on the GPU for 4d tensor. (Ian G.)
* The fix of flatten infer_shape re-enables an optimization (Pascal L.)
* The bug was introduced in 0.6rc1.
* Enable inc_subtensor on the GPU when updating it with a float64 dtype. (Ian G.)
It was causing an optimization warning.
* Make DeepCopy reuse preallocated memory. (Frederic B.)
* Move the convolution to the GPU when the image shape and logical image shape differ. (Frederic Bastien)
* C code for the View Op (Razvan P., Pascal L.)
New Features:
* Added a monitoring mode "MonitorMode" as a debugging tool. (Olivier D.)
* Allow integer axes when keepdims==True (Jeremiah Lowin)
* Added erfinv and erfcinv op. (Jey Kottalam)
* Added tensor.batched_dot(). (Caglar Gulcehre)
It uses scan behind the scenes, but makes doing this easier.
* theano.get_constant_value(x) (Frederic B.)
This tries to have x as a constant int.
This does some constant folding to try to convert x into an int.
Used by some optimizations.
* Add theano.tensor.io.{MPIRecv,MPIRecvWait,MPISend,MPISendWait} (Matthew Rocklin)
Theano does not automatically use them. It is up to you to use them and split your computations.
* Added theano.sandbox.linalg.eig (abalkin)
* Started some support for Python3 (abalkin)
setup.py supports python3 now.
It calls 2to3 during the setup.
Python3 is not fully supported as we didn't update the C code.
Crash Fixes:
* Fix a crash related to scan.grad due to the new mechanism. (Ian G.)
* Fix an optimization warning. Now it gets optimized. (Frederic B.)
* Fix crash introduced in 0.6rc1 in theano.grad (Ian G.)
* Fix crash introduced in 0.6rc1 in the grad of scan (Razvan P.)
* Fix crash introduced in 0.6rc1 in the grad of clip (Ian G.)
Also implement the gradient on the min/max bound.
* Fix crash in the grad of tensor.switch for int (Ian G.)
* Fix crash when mixing shared variable on the GPU and sparse dot. (Pascal L.)
* Fix crash as sometimes sparse.dot would return a different dtype number
that is equivalent but not the one expected. (Pascal L., reported by Rami Al-Rfou)
* Better error msg (Ian G.)
* Move all sparse random functions back to sandbox as they don't have a state inside Theano. (Pascal L.)
They were moved outside the sandbox in 0.6rc1
* LoadFromDisk now is allowed to only support some memmap mode. (Pascal L.)
Otherwise, this was causing errors, segmentation faults or wrong results.
* Fix import problem on PiCloud (Jeremiah Lowin)
* You need to use the c|py linker with the default
environment. Otherwise, you need to create your own environment.
* Fix a crash during optimization when we take a subtensor of a constant with a non constant index. (Ian G.)
* Better handling and error message of gradients on integer. (Ian G.)
* Fixed a crash where Scan assumed all TypeErrors raised by the grad function were due to undefined gradients (Ian G.)
Other:
* Doc typo fixes, Doc updates, Better error messages: Olivier D., David W.F., Frederic B., James B., Matthew Rocklin, Ian G., abalkin.
Theano 0.6rc1 (October 1st, 2012)
=================================
Highlights:
* Bug fixes, crash fixes, CPU and GPU speed up.
* theano_var.eval({other_var: val[,...]} to simplify the usage of Theano (Ian G.)
* New default linker `cvm`. This is the execution engine that tells ops to run in certain orders.
It is now implemented in C and enables lazy evaluation of ifelse op.
* Faster theano.function compilation. (Pascal L., Ian G.)
* Big sparse submodule update and documentation of it. (Nicolas Bouchard)
* Use GPU asynchronous functionality (Frederic B.)
* Better Windows support.
Known bugs:
* A few crash cases that will be fixed by the final release.
Bug fixes:
* Outputs of Scan nodes could contain corrupted values: some parts of the
output would be repeated a second time, instead of the correct values.
It happened randomly, and quite infrequently, but the bug has been present
(both in Python and Cython) since April 2011. (Pascal L.)
* In Sparse sandbox, fix the grad of theano.sparse.sandbox.sp.row_scale.
It did not return the right number of elements. (Frederic B.)
* set_subtensor(x[int vector], new_value) when moved to the GPU
was transformed into inc_subtensor on the GPU. Now we have a correct
(but slow) GPU implementation.
Note 1: set_subtensor(x[slice[,...]], new_value) was working correctly
in all cases as well as all inc_subtensor.
Note 2: If your code was affected by the incorrect behavior, we now print
a warning by default (Frederic B.)
* Fixed an issue whereby config values were used as default arguments,
with those defaults then stuck at old values if the config variables were
changed during program execution. (David W-F)
* Fixed many subtle bugs involving mutable default arguments which may have
led to unexpected behavior, such as objects sharing instance variables
they were not supposed to share. (David W-F)
* Correctly record the GPU device number used when we let the driver select it.
(Frederic B.)
* Min, max with NaN in inputs did not return the right output. (Pascal L.)
* The grad of TensorDot, was returning the wrong shape for some combination of axes.
We now raise NotImplementedError in those cases. (Frederic B.)
* conv2d with subsample >2 returned wrong values. (Pascal L.)
* Fixed when mode==valid, disabled when mode==full
* theano.sparse.CSMGrad op (generated by the grad of CSM) didn't
handle unsorted input correctly and gradient that is sparser
than the input. In that case, a bad result was returned. But this could
happen only when a sparse input of a Theano function was not
sorted. This happens for example with sparse advanced indexing from
scipy. The conclusion is most of time Nan in the graph.
(Yann Dauphin)
* theano.sparse._dot(CSC matrix, dense) optimized version UsmmCSCDense didn't handle
correctly not contiguous inputs/outputs. (Pascal L.)
* Fix a corner case CVM updates case. (Pascal L.)
This happened if the update to a shared variable is itself after optimization.
The CVM was not used by default.
* Fix the view_map of sparse.Transpose and sparse.sandbow.sp.RowScale. (Frederic B.)
This probably didn't cause problem as there is only the UsmmCscDense op
(used call to Usmm with CSC matrix) that could interfere with them.
Deprecation:
* Deprecated the Module class (Ian G.)
This was a predecessor of SharedVariable with a less pythonic philosophy.
Interface changes:
* Now the base version requirements are numpy >= 1.5.0 and the optional scipy >= 0.7.2.
* In Theano 0.5, we removed the deprecated sharedvar.value property.
Now we raise an error if you access it. (Frederic B.)
* theano.function does not accept duplicate inputs, so function([x, x], ...)
does not work anymore. (Pascal L.)
* theano.function now raises an error if some of the provided inputs are
not part of the computational graph needed to compute the output, for
instance, function([x, y], [y]). You can use the kwarg
``on_unused_input={'raise', 'warn', 'ignore'}`` to control this.
(Pascal L.)
* New Theano flag "on_unused_input" that defines the default value of the
previous point. (Frederic B.)
* tensor.alloc() now raises an error during graph build time
when we try to create less dimensions than the number of dimensions
the provided value have. In the past, the error was at run time.
(Frederic B.)
* Remove theano.Value and related stuff (Ian G.)
This was a test of what ended up as SharedVariable.
* Renamed Env to FunctionGraph, and object attribute "env" to "fgraph" (Ian G.)
Deprecation warning printed when you try to access the "env" attribute.
* Renamed the FunctionGraph.nodes attribute to FunctionNodes.apply_nodes (Ian G.)
* Warn when we don't handle correctly the parameter in Theano flags `nvcc.flags`
(Frederic B.)
* Do not reorder the user flags passed to the compiler. They get set after other flags. (Frederic B.)
* Make setuptools optional (Ilan Schnell)
* We warn when a user tries to use an old GPU with which Theano is untested.
This could cause crash and will also be very slow. (Frederic B.)
* Make theano.grad able to differentiate between not implemented, undefined and disconnected grad.
Op.grad function should return theano.gradient.{grad_not_implemented,grad_undefined} or
something of DisconectedType (Ian G.)
* Make theano.grad expect to always receive a float or undefined
gradient and enforce that op with integer output values always
return 0. (Ian G.)
New memory output contract (was mentioned in the release notes of Theano 0.5):
* Now the output memory received can be preallocated by other stuff.
In the past it was always the previous output an Apply node allocated.
So this means that the shape and strides can be different from previous calls
and there can be links to this memory at other places.
This means it could receive preallocated output that is not c_contiguous.
But we don't do that now. (Pascal L.)
* New Theano flags to test this DebugMode.check_preallocated_output (Pascal L.)
* Updated a few ops to respect this contract (Pascal L.)
New Features:
* GPU scan now works (does not crash) when there is a mixture of float32 and other dtypes.
* theano_var.eval({other_var:val[,...]} to simplify the usage of Theano (Ian G.)
* debugprint new param ids=["CHAR", "id", "int", ""]
This makes the identifier printed to be a unique char, the Python id, a
unique int, or not have it printed. We changed the default to be "CHAR"
as this is more readable. (Frederic B.)
* debugprint new param stop_on_name=[False, True]. If True, we don't print
anything below an intermediate variable that has a name. Defaults to False.
(Frederic B.)
* debugprint does not print anymore the "|" symbol in a column after the last input. (Frederic B.)
* If you use Enthought Python Distribution (EPD) now we use its blas
implementation by default. (Frederic B., Graham Taylor, Simon McGregor)
* MRG random now raises an error with a clear message when the passed shape
contains dimensions with bad value like 0. (Frederic B. reported by Ian G.)
* "CudaNdarray[*] = ndarray" works in more cases (Frederic B.)
* "CudaNdarray[*] += ndarray" works in more cases (Frederic B.)
* We add dimensions to CudaNdarray to automatically broadcast more frequently.
(Frederic B.)
* New theano flag cmodule.warn_no_version. Default False. If True,
will print a warning when compiling one or more Op with C code that
can't be cached because there is no c_code_cache_version() function
associated to at least one of those Ops. (Frederic B.)
* CPU alloc now always generate C code (Pascal L.)
* New Theano flag cmodule.warn_no_version=False. When True, warn when an op
with C code is not versioned (which forces to recompile it everytimes).
(Frederic B.)
* C code reuses preallocated outputs (only done by Scan) (Pascal L.)
* Garbage collection of intermediate results during Theano function calls
for Ops with C code (Pascal L.)
* Theano flag compiledir_format now supports the parameter "numpy_version" and "g++". (Frederic B.)
* Theano GPU variables, shared variables and constants now support <, <=,
> and >= similar to those not on the GPU.
* AdvancedIncSubtensor now supports the set_instead_of_inc parameter. (Eric L.)
* Added Advanced Indexing support to inc_subtensor and set_subtensor. (Eric L.)
* theano.tensor.{any,all,std,var,mean,prod,sum,argmin,argmax,min,max,max_and_argman}
have a new parameter keepdims (Eric L.)
This allows to broadcast it correctly against the input data to normalize it.
* The Updates objects now check that the keys are SharedVariable when we pass them
in the __init__ function. (Pascal L.)
* Set a Theano Variable name on transposed op when the input has one (Frederic B).
* The cvm linker now supports garbage collection (enabled by default). (James B. Arnaud B., Pascal L.)
* The cvm linker is now the default linker.
This makes the "loop" around the execution of apply node in C. So this lowers the overhead.
* theano_variable[numpy.newaxis] is now supported (James B.)
* Enable ifelse on the GPU. (Frederic B.)
* Correctly support numpy.memmap everywhere (Pascal L.)
We add partial support for them before. Just use the normal tensor operation
on them and it should work.
But be careful not to exhaust your computer memory! (we always generate normal ndarray)
* Add an optimization that stabilizes log(softmax(x)). (Ian G.)
* Re-enable the Images2Neibs grad. It was not broken, the problem was how we tested it. (Frederic B.)
* If `theano_fn.trust_input` is set to False, do not check if the inputs are good
when calling the theano function. (Frederic B.)
* Add theano.tensor.blas,gem{m,v} as shortcut.
* theano.grad(..., add_names=True). False for the old
behavior. Otherwise it tries to name the grad variables. (Ian G.)
* theano-nose (Pascal L.)
A wrapper around nosetests that adds needed extensions.
* --profile-time option, to print time spent in each test (Eric L.)
* --batch option, to allow to run tests in batch to lower memory requirement.
* m = mean(log(1 - sigm(x)))
x - scalar * theano.grad(m, x)
There is a stabilization optimization for this.
Now it is applied more frequently. (Pascal L.)
New Op/functions:
* Added element-wise operation theano.tensor.{GammaLn,Psi} (John Salvatier, Nicolas Bouchard)
* Added element-wise operation theano.tensor.{arcsin,arctan,arccosh,arcsinh,arctanh,exp2,arctan2} (Nicolas Bouchard)
* Added element-wise operation theano.tensor.{gamma,conj,complex_from_polar,expm1,deg2rad,rad2deg,trunc,gamma} (Nicolas Bouchard)
* Added theano.tensor.argsort that wraps numpy.argsort (Hani Almousli).
* Added theano.tensor.diff that wraps numpy.diff (Nicolas B.)
* Added theano.tensor.bincount that wraps numpy.bincount (Nicolas B., Pascal L, Frederic B.)
* Added theano.tensor.squeeze (Nicolas B.)
This removes broadcasted dimensions from the variable.
Theano-esque version of numpy.squeeze.
* Added theano.tensor.repeat that wraps numpy.repeat (Nicolas B. + PL)
* Added theano.tensor.bartlett that wraps numpy.bartlett (Eric L.)
* Added theano.tensor.fill_diagonal that wraps numpy.fill_diagonal (Eric L., Frederic B.)
* Added tensor.square that is an alias for tensor.sqr as NumPy (Ian G.)
* Added theano.tensor.load(path, dtype, broadcastable, mmap_mode=None) op
that allows to load a .npy file in a theano graph (Matthew Rocklin)
* theano.sandbox.linalg.kron.py:Kron op. (Eric L.)
Kronecker product
Speed up:
* CPU convolutions are now parallelized (Frederic B.)
By default use all cores/hyper-threads.
To control it, use the `OMP_NUM_THREADS=N` environment variable where N is the number of
parallel threads to use. By default it is equal to the number of CPU cores/hyper
threads that you have.
There is a new Theano flag `openmp` to allow/disallow openmp op.
If your BLAS library is parallelized, this flag won't affect it, but the
env variable will.
* Remove a corner case causing duplicated dot22/gemm in the graph. (Frederic B., Ian G.)
* Enable fusion of elemwise that have the same clients multiple times. (Frederic B.)
* New optimization: Remove reduction over broadcastable dimensions (James B., Frederic B.)
* Faster theano.function compilation. (Pascal L., Ian G.)
* Remove GPU transfer around specify_shape op. (Frederic B.)
* Implemented/tested MANY op.infer_shape method (Eric Larsen)
This allows Theano to make better shape inferance.
* Implement Solve.infer_shape (Matthew Rocklin)
* Scan memory optimizations now work more frequently. (Razvan P.)
There was a warning printed by the subtensor optimization in those cases.
* Faster rng_mrg Python code. (mostly used for tests) (Frederic B.)
Speed up GPU:
* Convolution on the GPU now checks the generation of the card to make
it faster in some cases (especially medium/big ouput image) (Frederic B.)
* We had hardcoded 512 as the maximum number of threads per block. Newer cards
support up to 1024 threads per block.
* Faster GpuAdvancedSubtensor1, GpuSubtensor, GpuAlloc (Frederic B.)
* We now pass the GPU architecture to nvcc when compiling (Frederic B.)
* Now we use the GPU function async feature by default. (Frederic B.)
Set the environment variable `CUDA_LAUNCH_BLOCKING` to `1` to disable this
for profiling or debugging.
* Faster creation of CudaNdarray objects (Frederic B.)
* Now some Max reductions are implemented on the GPU. (Ian G.)
Sparse Sandbox graduate (moved from theano.sparse.sandbox.sp):
* sparse.remove0 (Frederic B., Nicolas B.)
* sparse.sp_sum(a, axis=None) (Nicolas B.)
* bugfix: the not structured grad was returning a structured grad.
* sparse.{col_scale,row_scale,ensure_sorted_indices,clean} (Nicolas B.)
* sparse.{diag,square_diagonal} (Nicolas B.)
Sparse:
* Support for uint* dtype.
* Implement theano.sparse.mul(sparse1, sparse2) when both inputs don't
have the same sparsity pattern. (Frederic B.)
* New Ops: sparse.{expm1,deg2rad,rad2deg,trunc} (Nicolas B.)
* New Ops: sparse.{sqrt,sqr,log1p,floor,ceil,sgn,round_half_to_even} (Nicolas B.)
* New Ops: sparse.{arctanh,tanh,arcsinh,sinh,arctan,arcsin,tan,sin} (Nicolas B.)
* New functions: structured_{add,exp,log,pow,minimum,maximum,sigmoid} (Yann D., Nicolas B.)
* Optimized op: StructuredAddSV, StrucutedAddSVCSR (inserted automatically)
* New Op: sparse.mul_s_v multiplication of sparse matrix by broadcasted vector (Yann D.)
* New Op: sparse.Cast() (Yann D., Nicolas B.)
* Add sparse_variable.astype() and theano.sparse.cast() and
theano.sparse.{b,w,i,l,f,d,c,z}cast() as their tensor equivalent (Nicolas B.)
* Op class: SamplingDot (Yann D., Nicolas B.)
* Optimized version: SamplingDotCsr, StructuredDotCSC
* Optimizations to insert the optimized version: local_sampling_dot_csr, local_structured_add_s_v
* New Ops: sparse.{Multinomial,Poisson,Binomial} (Yann D., NB)
* Implement the CSMProperties grad method (Yann Dauphin)
* Move optimizations to theano/sparse/opt.py (Nicolas B.)
New flags:
* `profile=True` flag now prints the sum of all printed profiles. (Frederic B.)
* It works with the linkers vm/cvm (default).
* Also print compile time, optimizer time and linker time.
* Also print a summary by op class.
* new flag "profile_optimizer" (Frederic B.)
when profile=True, will also print the time spent in each optimizer.
Useful to find optimization bottleneck.
* new flag "cmodule.remove_gxx_opt" (Frederic B.)
If True, will remove -O* parameter passed to g++.
This is useful to debug in gdb module compiled by Theano.
The parameter -g is passed by default to g++.
* new flag cmodule.compilation_warning
if True, will print compilation warning.
* new flag `allow_gc` (Frederic B.)
When False, do not garbage collect intermediate results when they are not needed.
This uses more memory, but allocates memory less frequently so faster.
* new flag `vm.lazy` (Frederic B.)
Useful only for the vm linkers. When lazy is None,
auto detect if lazy evaluation is needed and use the apropriate
version. If lazy is True/False, force the version used between
Loop/LoopGC and Stack.
* new flag `cxx`. This is the C++ compiler to use. If empty do not compile C code. (Frederic B.)
* New flag `print_active_device` that defaults to True. (Matthew R.)
Documentation:
* Added in the tutorial documentation on how to extend Theano.
This explains how to make a Theano Op from a Python function.
http://deeplearning.net/software/theano/tutorial/extending_theano.html
(Frederic B.)
* New installation instructions for Windows using EPD (Pascal L.)
* New installation on Windows by using a Linux VM from ContinuumIO (Frederic B.)
* Revisions of Theano tutorial and addition of exercises to it. (Eric L.)
* New tutorial on Sparse variable. (Nicolas B., Sebastien Lemieux, Frederic Bastien
http://www.deeplearning.net/software/theano/tutorial/sparse.html
* Installation documentation for CentOS6 (Frederic B.)
* Installation documentation for Ubuntu (with GPU) (Frederic B., Matthias Zoehrer)
* Doc typo fixes, Doc updates, Better error messages: Olivier D., David W.F., Frederic B., James B., Matthew Rocklin, Ian G.
* Python Memory Management tutorial (Steven Pigeon, Olivier D.)
Proposal:
* Math framework for complex gradients (Pascal L.)
Internal changes:
* Define new exceptions MissingInputError and UnusedInputError, and use them
in theano.function, instead of TypeError and ValueError. (Pascal L.)
* Better handling of bitwidth and max values of integers and pointers
across platforms (Pascal L.)
* Made a few Ops with C code versioned to reduce compilation time.
(Frederic B, Pascal L.)
* Better deletion of files in the compiledir (Frederic B.)
* Safer import on sort op (Nicolas Pinto)
* hash_from_dict for elemwise op (Fredric B.)
* Renamed BadCLinkerOutput into BadThunkOutput. (PL)
* tensor.utils.shape_of_variables (Matthew R.)
* Add the numpy abi version and g++/nvcc version in the key of compiled code. (Frederic B.)
* env.replace_all_validate_remove (Frederic B.)
This allows global optimizer to ensure it removed some nodes from the graph.
This is a generic way to catch errors that would otherwise duplicate
computation.
* It was used for GEMM and Scan optimization (Frederic B., Razvan P.)
* Fix how exception are raised in GPU code (James B.)
* Made code respect pep8: OD, Fred, Pascal L., Nicolas Bouchard, Eric Larsen and others.
* TensorType and CudaNdarrayType now have a value_zeros method that call CudaNdarray.zeros or
numpy.zeros with the right dtype. (Pascal L., Olivier D.)
This allows to have the same code work with both types.
* Renamed FunctionGraph.extend function to FunctionGraph.attach_feature. (Ian G.)
* New exception MissingGXX when we try to compile but there is no cxx compiler. (Frederic B.)
* New fct theano.gof.utils.give_variables_names(...) that gives unique names to variables. (Matthew R.)
* Use most of the time the new NumPy C-API for later NumPy release. (Frederic B.)
* New theano.gof.sched.sort_apply_nodes() that will allow other execution ordering. (Matthew R.)
* New attribute sort_schedule_fn, a way to specify a scheduler to use. (Matthew R.)
Crash Fix:
* Fix import conflict name (usaar33, Frederic B.)
* This makes Theano work with PiCloud.
* Do not try to use the BLAS library when blas.ldflags is manually set to an
empty string (Frederic B., Pascal L.)
* When importing theano on a computer without GPU with the Theano
flags 'device' or 'init_gpu_device' set to gpu* (Frederic B., reported by Luo Heng)
* Optimization printed a useless error when scipy was not available. (Frederic B.)
* GPU conv crash/slowdown on newer hardware (James B.)
* Better error handling in GPU conv (Frederic B.)
* GPU optimization that moves element-wise Ops to the GPU. Crash happened in
a particular execution order of this optimization and the