forked from pytorch/pytorch.github.io
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathnn.html
7467 lines (7105 loc) · 621 KB
/
nn.html
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
<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>torch.nn — PyTorch master documentation</title>
<link rel="stylesheet" href="_static/css/theme.css" type="text/css" />
<link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Lato" type="text/css" />
<link rel="stylesheet" href="_static/css/pytorch_theme.css" type="text/css" />
<link rel="index" title="Index"
href="genindex.html"/>
<link rel="search" title="Search" href="search.html"/>
<link rel="top" title="PyTorch master documentation" href="index.html"/>
<link rel="next" title="torch.optim" href="optim.html"/>
<link rel="prev" title="torch.Storage" href="storage.html"/>
<script src="_static/js/modernizr.min.js"></script>
</head>
<body class="wy-body-for-nav" role="document">
<div class="wy-grid-for-nav">
<nav data-toggle="wy-nav-shift" class="wy-nav-side">
<div class="wy-side-scroll">
<div class="wy-side-nav-search">
<a href="index.html">
<img src="_static/pytorch-logo-dark.svg" class="logo" />
</a>
<div class="version">
<a href="https://pytorch.org/docs/versions.html"> 0.2.0_1 ▼</a>
</div>
<div role="search">
<form id="rtd-search-form" class="wy-form" action="search.html" method="get">
<input type="text" name="q" placeholder="Search docs" />
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
</div>
</div>
<div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
<p class="caption"><span class="caption-text">Notes</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="notes/autograd.html">Autograd mechanics</a><ul>
<li class="toctree-l2"><a class="reference internal" href="notes/autograd.html#excluding-subgraphs-from-backward">Excluding subgraphs from backward</a><ul>
<li class="toctree-l3"><a class="reference internal" href="notes/autograd.html#requires-grad"><code class="docutils literal"><span class="pre">requires_grad</span></code></a></li>
<li class="toctree-l3"><a class="reference internal" href="notes/autograd.html#volatile"><code class="docutils literal"><span class="pre">volatile</span></code></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="notes/autograd.html#how-autograd-encodes-the-history">How autograd encodes the history</a></li>
<li class="toctree-l2"><a class="reference internal" href="notes/autograd.html#in-place-operations-on-variables">In-place operations on Variables</a></li>
<li class="toctree-l2"><a class="reference internal" href="notes/autograd.html#in-place-correctness-checks">In-place correctness checks</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="notes/broadcasting.html">Broadcasting semantics</a><ul>
<li class="toctree-l2"><a class="reference internal" href="notes/broadcasting.html#general-semantics">General semantics</a></li>
<li class="toctree-l2"><a class="reference internal" href="notes/broadcasting.html#in-place-semantics">In-place semantics</a></li>
<li class="toctree-l2"><a class="reference internal" href="notes/broadcasting.html#backwards-compatibility">Backwards compatibility</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="notes/cuda.html">CUDA semantics</a><ul>
<li class="toctree-l2"><a class="reference internal" href="notes/cuda.html#best-practices">Best practices</a><ul>
<li class="toctree-l3"><a class="reference internal" href="notes/cuda.html#use-pinned-memory-buffers">Use pinned memory buffers</a></li>
<li class="toctree-l3"><a class="reference internal" href="notes/cuda.html#use-nn-dataparallel-instead-of-multiprocessing">Use nn.DataParallel instead of multiprocessing</a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="notes/extending.html">Extending PyTorch</a><ul>
<li class="toctree-l2"><a class="reference internal" href="notes/extending.html#extending-torch-autograd">Extending <code class="docutils literal"><span class="pre">torch.autograd</span></code></a></li>
<li class="toctree-l2"><a class="reference internal" href="notes/extending.html#extending-torch-nn">Extending <code class="docutils literal"><span class="pre">torch.nn</span></code></a><ul>
<li class="toctree-l3"><a class="reference internal" href="notes/extending.html#adding-a-module">Adding a <code class="docutils literal"><span class="pre">Module</span></code></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="notes/extending.html#writing-custom-c-extensions">Writing custom C extensions</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="notes/multiprocessing.html">Multiprocessing best practices</a><ul>
<li class="toctree-l2"><a class="reference internal" href="notes/multiprocessing.html#sharing-cuda-tensors">Sharing CUDA tensors</a></li>
<li class="toctree-l2"><a class="reference internal" href="notes/multiprocessing.html#best-practices-and-tips">Best practices and tips</a><ul>
<li class="toctree-l3"><a class="reference internal" href="notes/multiprocessing.html#avoiding-and-fighting-deadlocks">Avoiding and fighting deadlocks</a></li>
<li class="toctree-l3"><a class="reference internal" href="notes/multiprocessing.html#reuse-buffers-passed-through-a-queue">Reuse buffers passed through a Queue</a></li>
<li class="toctree-l3"><a class="reference internal" href="notes/multiprocessing.html#asynchronous-multiprocess-training-e-g-hogwild">Asynchronous multiprocess training (e.g. Hogwild)</a><ul>
<li class="toctree-l4"><a class="reference internal" href="notes/multiprocessing.html#hogwild">Hogwild</a></li>
</ul>
</li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="notes/serialization.html">Serialization semantics</a><ul>
<li class="toctree-l2"><a class="reference internal" href="notes/serialization.html#best-practices">Best practices</a><ul>
<li class="toctree-l3"><a class="reference internal" href="notes/serialization.html#recommended-approach-for-saving-a-model">Recommended approach for saving a model</a></li>
</ul>
</li>
</ul>
</li>
</ul>
<p class="caption"><span class="caption-text">Package Reference</span></p>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="torch.html">torch</a><ul>
<li class="toctree-l2"><a class="reference internal" href="torch.html#tensors">Tensors</a><ul>
<li class="toctree-l3"><a class="reference internal" href="torch.html#creation-ops">Creation Ops</a></li>
<li class="toctree-l3"><a class="reference internal" href="torch.html#indexing-slicing-joining-mutating-ops">Indexing, Slicing, Joining, Mutating Ops</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="torch.html#random-sampling">Random sampling</a></li>
<li class="toctree-l2"><a class="reference internal" href="torch.html#serialization">Serialization</a></li>
<li class="toctree-l2"><a class="reference internal" href="torch.html#parallelism">Parallelism</a></li>
<li class="toctree-l2"><a class="reference internal" href="torch.html#math-operations">Math operations</a><ul>
<li class="toctree-l3"><a class="reference internal" href="torch.html#pointwise-ops">Pointwise Ops</a></li>
<li class="toctree-l3"><a class="reference internal" href="torch.html#reduction-ops">Reduction Ops</a></li>
<li class="toctree-l3"><a class="reference internal" href="torch.html#comparison-ops">Comparison Ops</a></li>
<li class="toctree-l3"><a class="reference internal" href="torch.html#other-operations">Other Operations</a></li>
<li class="toctree-l3"><a class="reference internal" href="torch.html#blas-and-lapack-operations">BLAS and LAPACK Operations</a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="tensors.html">torch.Tensor</a></li>
<li class="toctree-l1"><a class="reference internal" href="sparse.html">torch.sparse</a></li>
<li class="toctree-l1"><a class="reference internal" href="storage.html">torch.Storage</a></li>
<li class="toctree-l1 current"><a class="current reference internal" href="#">torch.nn</a><ul>
<li class="toctree-l2"><a class="reference internal" href="#parameters">Parameters</a></li>
<li class="toctree-l2"><a class="reference internal" href="#containers">Containers</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#module"><span class="hidden-section">Module</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#sequential"><span class="hidden-section">Sequential</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#modulelist"><span class="hidden-section">ModuleList</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#parameterlist"><span class="hidden-section">ParameterList</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#convolution-layers">Convolution Layers</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#conv1d"><span class="hidden-section">Conv1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#conv2d"><span class="hidden-section">Conv2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#conv3d"><span class="hidden-section">Conv3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#convtranspose1d"><span class="hidden-section">ConvTranspose1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#convtranspose2d"><span class="hidden-section">ConvTranspose2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#convtranspose3d"><span class="hidden-section">ConvTranspose3d</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#pooling-layers">Pooling Layers</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#maxpool1d"><span class="hidden-section">MaxPool1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#maxpool2d"><span class="hidden-section">MaxPool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#maxpool3d"><span class="hidden-section">MaxPool3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#maxunpool1d"><span class="hidden-section">MaxUnpool1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#maxunpool2d"><span class="hidden-section">MaxUnpool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#maxunpool3d"><span class="hidden-section">MaxUnpool3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#avgpool1d"><span class="hidden-section">AvgPool1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#avgpool2d"><span class="hidden-section">AvgPool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#avgpool3d"><span class="hidden-section">AvgPool3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#fractionalmaxpool2d"><span class="hidden-section">FractionalMaxPool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#lppool2d"><span class="hidden-section">LPPool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#adaptivemaxpool1d"><span class="hidden-section">AdaptiveMaxPool1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#adaptivemaxpool2d"><span class="hidden-section">AdaptiveMaxPool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#adaptiveavgpool1d"><span class="hidden-section">AdaptiveAvgPool1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#adaptiveavgpool2d"><span class="hidden-section">AdaptiveAvgPool2d</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#padding-layers">Padding Layers</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#reflectionpad2d"><span class="hidden-section">ReflectionPad2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#replicationpad2d"><span class="hidden-section">ReplicationPad2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#replicationpad3d"><span class="hidden-section">ReplicationPad3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#zeropad2d"><span class="hidden-section">ZeroPad2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#constantpad2d"><span class="hidden-section">ConstantPad2d</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#non-linear-activations">Non-linear Activations</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#relu"><span class="hidden-section">ReLU</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#relu6"><span class="hidden-section">ReLU6</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#elu"><span class="hidden-section">ELU</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#selu"><span class="hidden-section">SELU</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#prelu"><span class="hidden-section">PReLU</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#leakyrelu"><span class="hidden-section">LeakyReLU</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#threshold"><span class="hidden-section">Threshold</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#hardtanh"><span class="hidden-section">Hardtanh</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#sigmoid"><span class="hidden-section">Sigmoid</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#tanh"><span class="hidden-section">Tanh</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#logsigmoid"><span class="hidden-section">LogSigmoid</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#softplus"><span class="hidden-section">Softplus</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#softshrink"><span class="hidden-section">Softshrink</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#softsign"><span class="hidden-section">Softsign</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#tanhshrink"><span class="hidden-section">Tanhshrink</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#softmin"><span class="hidden-section">Softmin</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#softmax"><span class="hidden-section">Softmax</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#logsoftmax"><span class="hidden-section">LogSoftmax</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#normalization-layers">Normalization layers</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#batchnorm1d"><span class="hidden-section">BatchNorm1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#batchnorm2d"><span class="hidden-section">BatchNorm2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#batchnorm3d"><span class="hidden-section">BatchNorm3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#instancenorm1d"><span class="hidden-section">InstanceNorm1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#instancenorm2d"><span class="hidden-section">InstanceNorm2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#instancenorm3d"><span class="hidden-section">InstanceNorm3d</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#recurrent-layers">Recurrent layers</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#rnn"><span class="hidden-section">RNN</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#lstm"><span class="hidden-section">LSTM</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#gru"><span class="hidden-section">GRU</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#rnncell"><span class="hidden-section">RNNCell</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#lstmcell"><span class="hidden-section">LSTMCell</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#grucell"><span class="hidden-section">GRUCell</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#linear-layers">Linear layers</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#linear"><span class="hidden-section">Linear</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#dropout-layers">Dropout layers</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#dropout"><span class="hidden-section">Dropout</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#dropout2d"><span class="hidden-section">Dropout2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#dropout3d"><span class="hidden-section">Dropout3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#alphadropout"><span class="hidden-section">AlphaDropout</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#sparse-layers">Sparse layers</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#embedding"><span class="hidden-section">Embedding</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#embeddingbag"><span class="hidden-section">EmbeddingBag</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#distance-functions">Distance functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#cosinesimilarity"><span class="hidden-section">CosineSimilarity</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#pairwisedistance"><span class="hidden-section">PairwiseDistance</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#loss-functions">Loss functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#l1loss"><span class="hidden-section">L1Loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#mseloss"><span class="hidden-section">MSELoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#crossentropyloss"><span class="hidden-section">CrossEntropyLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#nllloss"><span class="hidden-section">NLLLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#poissonnllloss"><span class="hidden-section">PoissonNLLLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#nllloss2d"><span class="hidden-section">NLLLoss2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#kldivloss"><span class="hidden-section">KLDivLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#bceloss"><span class="hidden-section">BCELoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#bcewithlogitsloss"><span class="hidden-section">BCEWithLogitsLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#marginrankingloss"><span class="hidden-section">MarginRankingLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#hingeembeddingloss"><span class="hidden-section">HingeEmbeddingLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#multilabelmarginloss"><span class="hidden-section">MultiLabelMarginLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#smoothl1loss"><span class="hidden-section">SmoothL1Loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#softmarginloss"><span class="hidden-section">SoftMarginLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#multilabelsoftmarginloss"><span class="hidden-section">MultiLabelSoftMarginLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#cosineembeddingloss"><span class="hidden-section">CosineEmbeddingLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#multimarginloss"><span class="hidden-section">MultiMarginLoss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#tripletmarginloss"><span class="hidden-section">TripletMarginLoss</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#vision-layers">Vision layers</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#pixelshuffle"><span class="hidden-section">PixelShuffle</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#upsample"><span class="hidden-section">Upsample</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#upsamplingnearest2d"><span class="hidden-section">UpsamplingNearest2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#upsamplingbilinear2d"><span class="hidden-section">UpsamplingBilinear2d</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#dataparallel-layers-multi-gpu-distributed">DataParallel layers (multi-GPU, distributed)</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#dataparallel"><span class="hidden-section">DataParallel</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#distributeddataparallel"><span class="hidden-section">DistributedDataParallel</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#utilities">Utilities</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#clip-grad-norm"><span class="hidden-section">clip_grad_norm</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#weight-norm"><span class="hidden-section">weight_norm</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#remove-weight-norm"><span class="hidden-section">remove_weight_norm</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#packedsequence"><span class="hidden-section">PackedSequence</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#pack-padded-sequence"><span class="hidden-section">pack_padded_sequence</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#pad-packed-sequence"><span class="hidden-section">pad_packed_sequence</span></a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="#torch-nn-functional">torch.nn.functional</a><ul>
<li class="toctree-l2"><a class="reference internal" href="#convolution-functions">Convolution functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#id16"><span class="hidden-section">conv1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id17"><span class="hidden-section">conv2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id18"><span class="hidden-section">conv3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#conv-transpose1d"><span class="hidden-section">conv_transpose1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#conv-transpose2d"><span class="hidden-section">conv_transpose2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#conv-transpose3d"><span class="hidden-section">conv_transpose3d</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#pooling-functions">Pooling functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#avg-pool1d"><span class="hidden-section">avg_pool1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#avg-pool2d"><span class="hidden-section">avg_pool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#avg-pool3d"><span class="hidden-section">avg_pool3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#max-pool1d"><span class="hidden-section">max_pool1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#max-pool2d"><span class="hidden-section">max_pool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#max-pool3d"><span class="hidden-section">max_pool3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#max-unpool1d"><span class="hidden-section">max_unpool1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#max-unpool2d"><span class="hidden-section">max_unpool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#max-unpool3d"><span class="hidden-section">max_unpool3d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#lp-pool2d"><span class="hidden-section">lp_pool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#adaptive-max-pool1d"><span class="hidden-section">adaptive_max_pool1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#adaptive-max-pool2d"><span class="hidden-section">adaptive_max_pool2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#adaptive-avg-pool1d"><span class="hidden-section">adaptive_avg_pool1d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#adaptive-avg-pool2d"><span class="hidden-section">adaptive_avg_pool2d</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#non-linear-activation-functions">Non-linear activation functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#id19"><span class="hidden-section">threshold</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id20"><span class="hidden-section">relu</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id21"><span class="hidden-section">hardtanh</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id22"><span class="hidden-section">relu6</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id23"><span class="hidden-section">elu</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id24"><span class="hidden-section">selu</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#leaky-relu"><span class="hidden-section">leaky_relu</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id25"><span class="hidden-section">prelu</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#rrelu"><span class="hidden-section">rrelu</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id26"><span class="hidden-section">logsigmoid</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#hardshrink"><span class="hidden-section">hardshrink</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id27"><span class="hidden-section">tanhshrink</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id28"><span class="hidden-section">softsign</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id29"><span class="hidden-section">softplus</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id30"><span class="hidden-section">softmin</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id31"><span class="hidden-section">softmax</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id32"><span class="hidden-section">softshrink</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#log-softmax"><span class="hidden-section">log_softmax</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id33"><span class="hidden-section">tanh</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id34"><span class="hidden-section">sigmoid</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#normalization-functions">Normalization functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#batch-norm"><span class="hidden-section">batch_norm</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#normalize"><span class="hidden-section">normalize</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#linear-functions">Linear functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#id35"><span class="hidden-section">linear</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#dropout-functions">Dropout functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#id36"><span class="hidden-section">dropout</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#alpha-dropout"><span class="hidden-section">alpha_dropout</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id37"><span class="hidden-section">dropout2d</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id38"><span class="hidden-section">dropout3d</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#id39">Distance functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#pairwise-distance"><span class="hidden-section">pairwise_distance</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#cosine-similarity"><span class="hidden-section">cosine_similarity</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#id40">Loss functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#binary-cross-entropy"><span class="hidden-section">binary_cross_entropy</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#poisson-nll-loss"><span class="hidden-section">poisson_nll_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#cosine-embedding-loss"><span class="hidden-section">cosine_embedding_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#cross-entropy"><span class="hidden-section">cross_entropy</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#hinge-embedding-loss"><span class="hidden-section">hinge_embedding_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#kl-div"><span class="hidden-section">kl_div</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#l1-loss"><span class="hidden-section">l1_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#mse-loss"><span class="hidden-section">mse_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#margin-ranking-loss"><span class="hidden-section">margin_ranking_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#multilabel-margin-loss"><span class="hidden-section">multilabel_margin_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#multilabel-soft-margin-loss"><span class="hidden-section">multilabel_soft_margin_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#multi-margin-loss"><span class="hidden-section">multi_margin_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#nll-loss"><span class="hidden-section">nll_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#binary-cross-entropy-with-logits"><span class="hidden-section">binary_cross_entropy_with_logits</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#smooth-l1-loss"><span class="hidden-section">smooth_l1_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#soft-margin-loss"><span class="hidden-section">soft_margin_loss</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#triplet-margin-loss"><span class="hidden-section">triplet_margin_loss</span></a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="#vision-functions">Vision functions</a><ul>
<li class="toctree-l3"><a class="reference internal" href="#pixel-shuffle"><span class="hidden-section">pixel_shuffle</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#pad"><span class="hidden-section">pad</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#id42"><span class="hidden-section">upsample</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#upsample-nearest"><span class="hidden-section">upsample_nearest</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#upsample-bilinear"><span class="hidden-section">upsample_bilinear</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#grid-sample"><span class="hidden-section">grid_sample</span></a></li>
<li class="toctree-l3"><a class="reference internal" href="#affine-grid"><span class="hidden-section">affine_grid</span></a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="#torch-nn-init">torch.nn.init</a></li>
<li class="toctree-l1"><a class="reference internal" href="optim.html">torch.optim</a><ul>
<li class="toctree-l2"><a class="reference internal" href="optim.html#how-to-use-an-optimizer">How to use an optimizer</a><ul>
<li class="toctree-l3"><a class="reference internal" href="optim.html#constructing-it">Constructing it</a></li>
<li class="toctree-l3"><a class="reference internal" href="optim.html#per-parameter-options">Per-parameter options</a></li>
<li class="toctree-l3"><a class="reference internal" href="optim.html#taking-an-optimization-step">Taking an optimization step</a><ul>
<li class="toctree-l4"><a class="reference internal" href="optim.html#optimizer-step"><code class="docutils literal"><span class="pre">optimizer.step()</span></code></a></li>
<li class="toctree-l4"><a class="reference internal" href="optim.html#optimizer-step-closure"><code class="docutils literal"><span class="pre">optimizer.step(closure)</span></code></a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="optim.html#algorithms">Algorithms</a></li>
<li class="toctree-l2"><a class="reference internal" href="optim.html#how-to-adjust-learning-rate">How to adjust Learning Rate</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="autograd.html">torch.autograd</a><ul>
<li class="toctree-l2"><a class="reference internal" href="autograd.html#variable">Variable</a><ul>
<li class="toctree-l3"><a class="reference internal" href="autograd.html#api-compatibility">API compatibility</a></li>
<li class="toctree-l3"><a class="reference internal" href="autograd.html#in-place-operations-on-variables">In-place operations on Variables</a></li>
<li class="toctree-l3"><a class="reference internal" href="autograd.html#in-place-correctness-checks">In-place correctness checks</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="autograd.html#function"><span class="hidden-section">Function</span></a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="multiprocessing.html">torch.multiprocessing</a><ul>
<li class="toctree-l2"><a class="reference internal" href="multiprocessing.html#strategy-management">Strategy management</a></li>
<li class="toctree-l2"><a class="reference internal" href="multiprocessing.html#sharing-cuda-tensors">Sharing CUDA tensors</a></li>
<li class="toctree-l2"><a class="reference internal" href="multiprocessing.html#sharing-strategies">Sharing strategies</a><ul>
<li class="toctree-l3"><a class="reference internal" href="multiprocessing.html#file-descriptor-file-descriptor">File descriptor - <code class="docutils literal"><span class="pre">file_descriptor</span></code></a></li>
<li class="toctree-l3"><a class="reference internal" href="multiprocessing.html#file-system-file-system">File system - <code class="docutils literal"><span class="pre">file_system</span></code></a></li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="distributed.html">torch.distributed</a><ul>
<li class="toctree-l2"><a class="reference internal" href="distributed.html#initialization">Initialization</a><ul>
<li class="toctree-l3"><a class="reference internal" href="distributed.html#tcp-initialization">TCP initialization</a></li>
<li class="toctree-l3"><a class="reference internal" href="distributed.html#shared-file-system-initialization">Shared file-system initialization</a></li>
<li class="toctree-l3"><a class="reference internal" href="distributed.html#environment-variable-initialization">Environment variable initialization</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="distributed.html#groups">Groups</a></li>
<li class="toctree-l2"><a class="reference internal" href="distributed.html#point-to-point-communication">Point-to-point communication</a></li>
<li class="toctree-l2"><a class="reference internal" href="distributed.html#collective-functions">Collective functions</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="legacy.html">torch.legacy</a></li>
<li class="toctree-l1"><a class="reference internal" href="cuda.html">torch.cuda</a><ul>
<li class="toctree-l2"><a class="reference internal" href="cuda.html#communication-collectives">Communication collectives</a></li>
<li class="toctree-l2"><a class="reference internal" href="cuda.html#streams-and-events">Streams and events</a></li>
<li class="toctree-l2"><a class="reference internal" href="cuda.html#nvidia-tools-extension-nvtx">NVIDIA Tools Extension (NVTX)</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="ffi.html">torch.utils.ffi</a></li>
<li class="toctree-l1"><a class="reference internal" href="data.html">torch.utils.data</a></li>
<li class="toctree-l1"><a class="reference internal" href="model_zoo.html">torch.utils.model_zoo</a></li>
</ul>
<p class="caption"><span class="caption-text">torchvision Reference</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="torchvision/torchvision.html">torchvision</a></li>
<li class="toctree-l1"><a class="reference internal" href="torchvision/datasets.html">torchvision.datasets</a><ul>
<li class="toctree-l2"><a class="reference internal" href="torchvision/datasets.html#mnist">MNIST</a></li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/datasets.html#coco">COCO</a><ul>
<li class="toctree-l3"><a class="reference internal" href="torchvision/datasets.html#captions">Captions</a></li>
<li class="toctree-l3"><a class="reference internal" href="torchvision/datasets.html#detection">Detection</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/datasets.html#lsun">LSUN</a></li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/datasets.html#imagefolder">ImageFolder</a></li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/datasets.html#imagenet-12">Imagenet-12</a></li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/datasets.html#cifar">CIFAR</a></li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/datasets.html#stl10">STL10</a></li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/datasets.html#svhn">SVHN</a></li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/datasets.html#phototour">PhotoTour</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="torchvision/models.html">torchvision.models</a></li>
<li class="toctree-l1"><a class="reference internal" href="torchvision/transforms.html">torchvision.transforms</a><ul>
<li class="toctree-l2"><a class="reference internal" href="torchvision/transforms.html#transforms-on-pil-image">Transforms on PIL.Image</a></li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/transforms.html#transforms-on-torch-tensor">Transforms on torch.*Tensor</a></li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/transforms.html#conversion-transforms">Conversion Transforms</a></li>
<li class="toctree-l2"><a class="reference internal" href="torchvision/transforms.html#generic-transforms">Generic Transforms</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="torchvision/utils.html">torchvision.utils</a></li>
</ul>
</div>
</div>
</nav>
<section data-toggle="wy-nav-shift" class="wy-nav-content-wrap">
<nav class="wy-nav-top" role="navigation" aria-label="top navigation">
<i data-toggle="wy-nav-top" class="fa fa-bars"></i>
<a href="index.html">PyTorch</a>
</nav>
<div class="wy-nav-content">
<div class="rst-content">
<div role="navigation" aria-label="breadcrumbs navigation">
<ul class="wy-breadcrumbs">
<li><a href="index.html">Docs</a> »</li>
<li>torch.nn</li>
<li class="wy-breadcrumbs-aside">
<a href="_sources/nn.rst.txt" rel="nofollow"> View page source</a>
</li>
</ul>
<hr/>
</div>
<div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
<div itemprop="articleBody">
<div class="section" id="module-torch.nn">
<span id="torch-nn"></span><h1>torch.nn<a class="headerlink" href="#module-torch.nn" title="Permalink to this headline">¶</a></h1>
<div class="section" id="parameters">
<h2>Parameters<a class="headerlink" href="#parameters" title="Permalink to this headline">¶</a></h2>
<dl class="class">
<dt id="torch.nn.Parameter">
<em class="property">class </em><code class="descclassname">torch.nn.</code><code class="descname">Parameter</code><a class="reference internal" href="_modules/torch/nn/parameter.html#Parameter"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Parameter" title="Permalink to this definition">¶</a></dt>
<dd><p>A kind of Variable that is to be considered a module parameter.</p>
<p>Parameters are <a class="reference internal" href="autograd.html#torch.autograd.Variable" title="torch.autograd.Variable"><code class="xref py py-class docutils literal"><span class="pre">Variable</span></code></a> subclasses, that have a
very special property when used with <a class="reference internal" href="#torch.nn.Module" title="torch.nn.Module"><code class="xref py py-class docutils literal"><span class="pre">Module</span></code></a> s - when they’re
assigned as Module attributes they are automatically added to the list of
its parameters, and will appear e.g. in <a class="reference internal" href="#torch.nn.Module.parameters" title="torch.nn.Module.parameters"><code class="xref py py-meth docutils literal"><span class="pre">parameters()</span></code></a> iterator.
Assigning a Variable doesn’t have such effect. This is because one might
want to cache some temporary state, like last hidden state of the RNN, in
the model. If there was no such class as <a class="reference internal" href="#torch.nn.Parameter" title="torch.nn.Parameter"><code class="xref py py-class docutils literal"><span class="pre">Parameter</span></code></a>, these
temporaries would get registered too.</p>
<p>Another difference is that parameters can’t be volatile and that they
require gradient by default.</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
<li><strong>data</strong> (<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><em>Tensor</em></a>) – parameter tensor.</li>
<li><strong>requires_grad</strong> (<a class="reference external" href="https://docs.python.org/2/library/functions.html#bool" title="(in Python v2.7)"><em>bool</em></a><em>, </em><em>optional</em>) – if the parameter requires gradient. See
<a class="reference internal" href="notes/autograd.html#excluding-subgraphs"><span class="std std-ref">Excluding subgraphs from backward</span></a> for more details.</li>
</ul>
</td>
</tr>
</tbody>
</table>
</dd></dl>
</div>
<div class="section" id="containers">
<h2>Containers<a class="headerlink" href="#containers" title="Permalink to this headline">¶</a></h2>
<div class="section" id="module">
<h3><span class="hidden-section">Module</span><a class="headerlink" href="#module" title="Permalink to this headline">¶</a></h3>
<dl class="class">
<dt id="torch.nn.Module">
<em class="property">class </em><code class="descclassname">torch.nn.</code><code class="descname">Module</code><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module" title="Permalink to this definition">¶</a></dt>
<dd><p>Base class for all neural network modules.</p>
<p>Your models should also subclass this class.</p>
<p>Modules can also contain other Modules, allowing to nest them in
a tree structure. You can assign the submodules as regular attributes:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">torch.nn</span> <span class="k">as</span> <span class="nn">nn</span>
<span class="kn">import</span> <span class="nn">torch.nn.functional</span> <span class="k">as</span> <span class="nn">F</span>
<span class="k">class</span> <span class="nc">Model</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="nf">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="nb">super</span><span class="p">(</span><span class="n">Model</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">20</span><span class="p">,</span> <span class="mi">5</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv2</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">20</span><span class="p">,</span> <span class="mi">20</span><span class="p">,</span> <span class="mi">5</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv1</span><span class="p">(</span><span class="n">x</span><span class="p">))</span>
<span class="k">return</span> <span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv2</span><span class="p">(</span><span class="n">x</span><span class="p">))</span>
</pre></div>
</div>
<p>Submodules assigned in this way will be registered, and will have their
parameters converted too when you call .cuda(), etc.</p>
<dl class="method">
<dt id="torch.nn.Module.add_module">
<code class="descname">add_module</code><span class="sig-paren">(</span><em>name</em>, <em>module</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.add_module"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.add_module" title="Permalink to this definition">¶</a></dt>
<dd><p>Adds a child module to the current module.</p>
<p>The module can be accessed as an attribute using the given name.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.children">
<code class="descname">children</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.children"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.children" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns an iterator over immediate children modules.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.cpu">
<code class="descname">cpu</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.cpu"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.cpu" title="Permalink to this definition">¶</a></dt>
<dd><p>Moves all model parameters and buffers to the CPU.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.cuda">
<code class="descname">cuda</code><span class="sig-paren">(</span><em>device_id=None</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.cuda"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.cuda" title="Permalink to this definition">¶</a></dt>
<dd><p>Moves all model parameters and buffers to the GPU.</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>device_id</strong> (<a class="reference external" href="https://docs.python.org/2/library/functions.html#int" title="(in Python v2.7)"><em>int</em></a><em>, </em><em>optional</em>) – if specified, all parameters will be
copied to that device</td>
</tr>
</tbody>
</table>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.double">
<code class="descname">double</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.double"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.double" title="Permalink to this definition">¶</a></dt>
<dd><p>Casts all parameters and buffers to double datatype.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.eval">
<code class="descname">eval</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.eval"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.eval" title="Permalink to this definition">¶</a></dt>
<dd><p>Sets the module in evaluation mode.</p>
<p>This has any effect only on modules such as Dropout or BatchNorm.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.float">
<code class="descname">float</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.float"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.float" title="Permalink to this definition">¶</a></dt>
<dd><p>Casts all parameters and buffers to float datatype.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.forward">
<code class="descname">forward</code><span class="sig-paren">(</span><em>*input</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.forward"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.forward" title="Permalink to this definition">¶</a></dt>
<dd><p>Defines the computation performed at every call.</p>
<p>Should be overriden by all subclasses.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.half">
<code class="descname">half</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.half"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.half" title="Permalink to this definition">¶</a></dt>
<dd><p>Casts all parameters and buffers to half datatype.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.load_state_dict">
<code class="descname">load_state_dict</code><span class="sig-paren">(</span><em>state_dict</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.load_state_dict"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.load_state_dict" title="Permalink to this definition">¶</a></dt>
<dd><p>Copies parameters and buffers from <a class="reference internal" href="#torch.nn.Module.state_dict" title="torch.nn.Module.state_dict"><code class="xref py py-attr docutils literal"><span class="pre">state_dict</span></code></a> into
this module and its descendants. The keys of <a class="reference internal" href="#torch.nn.Module.state_dict" title="torch.nn.Module.state_dict"><code class="xref py py-attr docutils literal"><span class="pre">state_dict</span></code></a> must
exactly match the keys returned by this module’s <a class="reference internal" href="#torch.nn.Module.state_dict" title="torch.nn.Module.state_dict"><code class="xref py py-func docutils literal"><span class="pre">state_dict()</span></code></a>
function.</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>state_dict</strong> (<a class="reference external" href="https://docs.python.org/2/library/stdtypes.html#dict" title="(in Python v2.7)"><em>dict</em></a>) – A dict containing parameters and
persistent buffers.</td>
</tr>
</tbody>
</table>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.modules">
<code class="descname">modules</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.modules"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.modules" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns an iterator over all modules in the network.</p>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p>Duplicate modules are returned only once. In the following
example, <code class="docutils literal"><span class="pre">l</span></code> will be returned only once.</p>
<div class="last highlight-default"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">l</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">net</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span><span class="n">l</span><span class="p">,</span> <span class="n">l</span><span class="p">)</span>
<span class="gp">>>> </span><span class="k">for</span> <span class="n">idx</span><span class="p">,</span> <span class="n">m</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="n">net</span><span class="o">.</span><span class="n">modules</span><span class="p">()):</span>
<span class="gp">>>> </span> <span class="nb">print</span><span class="p">(</span><span class="n">idx</span><span class="p">,</span> <span class="s1">'->'</span><span class="p">,</span> <span class="n">m</span><span class="p">)</span>
<span class="go">0 -> Sequential (</span>
<span class="go"> (0): Linear (2 -> 2)</span>
<span class="go"> (1): Linear (2 -> 2)</span>
<span class="go">)</span>
<span class="go">1 -> Linear (2 -> 2)</span>
</pre></div>
</div>
</div>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.named_children">
<code class="descname">named_children</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.named_children"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.named_children" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns an iterator over immediate children modules, yielding both
the name of the module as well as the module itself.</p>
<p class="rubric">Example</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="k">for</span> <span class="n">name</span><span class="p">,</span> <span class="n">module</span> <span class="ow">in</span> <span class="n">model</span><span class="o">.</span><span class="n">named_children</span><span class="p">():</span>
<span class="gp">>>> </span> <span class="k">if</span> <span class="n">name</span> <span class="ow">in</span> <span class="p">[</span><span class="s1">'conv4'</span><span class="p">,</span> <span class="s1">'conv5'</span><span class="p">]:</span>
<span class="gp">>>> </span> <span class="nb">print</span><span class="p">(</span><span class="n">module</span><span class="p">)</span>
</pre></div>
</div>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.named_modules">
<code class="descname">named_modules</code><span class="sig-paren">(</span><em>memo=None</em>, <em>prefix=''</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.named_modules"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.named_modules" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns an iterator over all modules in the network, yielding
both the name of the module as well as the module itself.</p>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p>Duplicate modules are returned only once. In the following
example, <code class="docutils literal"><span class="pre">l</span></code> will be returned only once.</p>
<div class="last highlight-default"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">l</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">net</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span><span class="n">l</span><span class="p">,</span> <span class="n">l</span><span class="p">)</span>
<span class="gp">>>> </span><span class="k">for</span> <span class="n">idx</span><span class="p">,</span> <span class="n">m</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="n">net</span><span class="o">.</span><span class="n">named_modules</span><span class="p">()):</span>
<span class="gp">>>> </span> <span class="nb">print</span><span class="p">(</span><span class="n">idx</span><span class="p">,</span> <span class="s1">'->'</span><span class="p">,</span> <span class="n">m</span><span class="p">)</span>
<span class="go">0 -> ('', Sequential (</span>
<span class="go"> (0): Linear (2 -> 2)</span>
<span class="go"> (1): Linear (2 -> 2)</span>
<span class="go">))</span>
<span class="go">1 -> ('0', Linear (2 -> 2))</span>
</pre></div>
</div>
</div>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.named_parameters">
<code class="descname">named_parameters</code><span class="sig-paren">(</span><em>memo=None</em>, <em>prefix=''</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.named_parameters"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.named_parameters" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns an iterator over module parameters, yielding both the
name of the parameter as well as the parameter itself</p>
<p class="rubric">Example</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="k">for</span> <span class="n">name</span><span class="p">,</span> <span class="n">param</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">named_parameters</span><span class="p">():</span>
<span class="gp">>>> </span> <span class="k">if</span> <span class="n">name</span> <span class="ow">in</span> <span class="p">[</span><span class="s1">'bias'</span><span class="p">]:</span>
<span class="gp">>>> </span> <span class="nb">print</span><span class="p">(</span><span class="n">param</span><span class="o">.</span><span class="n">size</span><span class="p">())</span>
</pre></div>
</div>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.parameters">
<code class="descname">parameters</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.parameters"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.parameters" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns an iterator over module parameters.</p>
<p>This is typically passed to an optimizer.</p>
<p class="rubric">Example</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="k">for</span> <span class="n">param</span> <span class="ow">in</span> <span class="n">model</span><span class="o">.</span><span class="n">parameters</span><span class="p">():</span>
<span class="gp">>>> </span> <span class="nb">print</span><span class="p">(</span><span class="nb">type</span><span class="p">(</span><span class="n">param</span><span class="o">.</span><span class="n">data</span><span class="p">),</span> <span class="n">param</span><span class="o">.</span><span class="n">size</span><span class="p">())</span>
<span class="go"><class 'torch.FloatTensor'> (20L,)</span>
<span class="go"><class 'torch.FloatTensor'> (20L, 1L, 5L, 5L)</span>
</pre></div>
</div>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.register_backward_hook">
<code class="descname">register_backward_hook</code><span class="sig-paren">(</span><em>hook</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.register_backward_hook"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.register_backward_hook" title="Permalink to this definition">¶</a></dt>
<dd><p>Registers a backward hook on the module.</p>
<p>The hook will be called every time the gradients with respect to module
inputs are computed. The hook should have the following signature:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">hook</span><span class="p">(</span><span class="n">module</span><span class="p">,</span> <span class="n">grad_input</span><span class="p">,</span> <span class="n">grad_output</span><span class="p">)</span> <span class="o">-></span> <span class="n">Tensor</span> <span class="ow">or</span> <span class="kc">None</span>
</pre></div>
</div>
<p>The <code class="xref py py-attr docutils literal"><span class="pre">grad_input</span></code> and <code class="xref py py-attr docutils literal"><span class="pre">grad_output</span></code> may be tuples if the
module has multiple inputs or outputs. The hook should not modify its
arguments, but it can optionally return a new gradient with respect to
input that will be used in place of <code class="xref py py-attr docutils literal"><span class="pre">grad_input</span></code> in subsequent
computations.</p>
<p>This function returns a handle with a method <code class="docutils literal"><span class="pre">handle.remove()</span></code>
that removes the hook from the module.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.register_buffer">
<code class="descname">register_buffer</code><span class="sig-paren">(</span><em>name</em>, <em>tensor</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.register_buffer"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.register_buffer" title="Permalink to this definition">¶</a></dt>
<dd><p>Adds a persistent buffer to the module.</p>
<p>This is typically used to register a buffer that should not to be
considered a model parameter. For example, BatchNorm’s <code class="docutils literal"><span class="pre">running_mean</span></code>
is not a parameter, but is part of the persistent state.</p>
<p>Buffers can be accessed as attributes using given names.</p>
<p class="rubric">Example</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="bp">self</span><span class="o">.</span><span class="n">register_buffer</span><span class="p">(</span><span class="s1">'running_mean'</span><span class="p">,</span> <span class="n">torch</span><span class="o">.</span><span class="n">zeros</span><span class="p">(</span><span class="n">num_features</span><span class="p">))</span>
</pre></div>
</div>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.register_forward_hook">
<code class="descname">register_forward_hook</code><span class="sig-paren">(</span><em>hook</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.register_forward_hook"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.register_forward_hook" title="Permalink to this definition">¶</a></dt>
<dd><p>Registers a forward hook on the module.</p>
<p>The hook will be called every time <a class="reference internal" href="#torch.nn.Module.forward" title="torch.nn.Module.forward"><code class="xref py py-func docutils literal"><span class="pre">forward()</span></code></a> computes an output.
It should have the following signature:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">hook</span><span class="p">(</span><span class="n">module</span><span class="p">,</span> <span class="nb">input</span><span class="p">,</span> <span class="n">output</span><span class="p">)</span> <span class="o">-></span> <span class="kc">None</span>
</pre></div>
</div>
<p>The hook should not modify the input or output.
This function returns a handle with a method <code class="docutils literal"><span class="pre">handle.remove()</span></code>
that removes the hook from the module.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.register_forward_pre_hook">
<code class="descname">register_forward_pre_hook</code><span class="sig-paren">(</span><em>hook</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.register_forward_pre_hook"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.register_forward_pre_hook" title="Permalink to this definition">¶</a></dt>
<dd><p>Registers a forward pre-hook on the module.</p>
<p>The hook will be called before <a class="reference internal" href="#torch.nn.Module.forward" title="torch.nn.Module.forward"><code class="xref py py-func docutils literal"><span class="pre">forward()</span></code></a> is invoked.
It should have the following signature:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="n">hook</span><span class="p">(</span><span class="n">module</span><span class="p">,</span> <span class="nb">input</span><span class="p">)</span> <span class="o">-></span> <span class="kc">None</span>
</pre></div>
</div>
<p>The hook should not modify the input.
This function returns a handle with a method <code class="docutils literal"><span class="pre">handle.remove()</span></code>
that removes the hook from the module.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.register_parameter">
<code class="descname">register_parameter</code><span class="sig-paren">(</span><em>name</em>, <em>param</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.register_parameter"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.register_parameter" title="Permalink to this definition">¶</a></dt>
<dd><p>Adds a parameter to the module.</p>
<p>The parameter can be accessed as an attribute using given name.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.state_dict">
<code class="descname">state_dict</code><span class="sig-paren">(</span><em>destination=None</em>, <em>prefix=''</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.state_dict"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.state_dict" title="Permalink to this definition">¶</a></dt>
<dd><p>Returns a dictionary containing a whole state of the module.</p>
<p>Both parameters and persistent buffers (e.g. running averages) are
included. Keys are corresponding parameter and buffer names.</p>
<p class="rubric">Example</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">module</span><span class="o">.</span><span class="n">state_dict</span><span class="p">()</span><span class="o">.</span><span class="n">keys</span><span class="p">()</span>
<span class="go">['bias', 'weight']</span>
</pre></div>
</div>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.train">
<code class="descname">train</code><span class="sig-paren">(</span><em>mode=True</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.train"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.train" title="Permalink to this definition">¶</a></dt>
<dd><p>Sets the module in training mode.</p>
<p>This has any effect only on modules such as Dropout or BatchNorm.</p>
</dd></dl>
<dl class="method">
<dt id="torch.nn.Module.zero_grad">
<code class="descname">zero_grad</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/module.html#Module.zero_grad"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Module.zero_grad" title="Permalink to this definition">¶</a></dt>
<dd><p>Sets gradients of all model parameters to zero.</p>
</dd></dl>
</dd></dl>
</div>
<div class="section" id="sequential">
<h3><span class="hidden-section">Sequential</span><a class="headerlink" href="#sequential" title="Permalink to this headline">¶</a></h3>
<dl class="class">
<dt id="torch.nn.Sequential">
<em class="property">class </em><code class="descclassname">torch.nn.</code><code class="descname">Sequential</code><span class="sig-paren">(</span><em>*args</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/container.html#Sequential"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.Sequential" title="Permalink to this definition">¶</a></dt>
<dd><p>A sequential container.
Modules will be added to it in the order they are passed in the constructor.
Alternatively, an ordered dict of modules can also be passed in.</p>
<p>To make it easier to understand, given is a small example:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="c1"># Example of using Sequential</span>
<span class="n">model</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span><span class="mi">20</span><span class="p">,</span><span class="mi">5</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">(),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">20</span><span class="p">,</span><span class="mi">64</span><span class="p">,</span><span class="mi">5</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">()</span>
<span class="p">)</span>
<span class="c1"># Example of using Sequential with OrderedDict</span>
<span class="n">model</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span><span class="n">OrderedDict</span><span class="p">([</span>
<span class="p">(</span><span class="s1">'conv1'</span><span class="p">,</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span><span class="mi">20</span><span class="p">,</span><span class="mi">5</span><span class="p">)),</span>
<span class="p">(</span><span class="s1">'relu1'</span><span class="p">,</span> <span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">()),</span>
<span class="p">(</span><span class="s1">'conv2'</span><span class="p">,</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">20</span><span class="p">,</span><span class="mi">64</span><span class="p">,</span><span class="mi">5</span><span class="p">)),</span>
<span class="p">(</span><span class="s1">'relu2'</span><span class="p">,</span> <span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">())</span>
<span class="p">]))</span>
</pre></div>
</div>
</dd></dl>
</div>
<div class="section" id="modulelist">
<h3><span class="hidden-section">ModuleList</span><a class="headerlink" href="#modulelist" title="Permalink to this headline">¶</a></h3>
<dl class="class">
<dt id="torch.nn.ModuleList">
<em class="property">class </em><code class="descclassname">torch.nn.</code><code class="descname">ModuleList</code><span class="sig-paren">(</span><em>modules=None</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/container.html#ModuleList"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.ModuleList" title="Permalink to this definition">¶</a></dt>
<dd><p>Holds submodules in a list.</p>
<p>ModuleList can be indexed like a regular Python list, but modules it
contains are properly registered, and will be visible by all Module methods.</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>modules</strong> (<em>list</em><em>, </em><em>optional</em>) – a list of modules to add</td>
</tr>
</tbody>
</table>
<p>Example:</p>
<div class="highlight-default"><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">MyModule</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="nf">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="nb">super</span><span class="p">(</span><span class="n">MyModule</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">linears</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">ModuleList</span><span class="p">([</span><span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">10</span><span class="p">,</span> <span class="mi">10</span><span class="p">)</span> <span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">10</span><span class="p">)])</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="c1"># ModuleList can act as an iterable, or be indexed using ints</span>
<span class="k">for</span> <span class="n">i</span><span class="p">,</span> <span class="n">l</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">linears</span><span class="p">):</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">linears</span><span class="p">[</span><span class="n">i</span> <span class="o">//</span> <span class="mi">2</span><span class="p">](</span><span class="n">x</span><span class="p">)</span> <span class="o">+</span> <span class="n">l</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="k">return</span> <span class="n">x</span>
</pre></div>
</div>
<dl class="method">
<dt id="torch.nn.ModuleList.append">
<code class="descname">append</code><span class="sig-paren">(</span><em>module</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/container.html#ModuleList.append"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.ModuleList.append" title="Permalink to this definition">¶</a></dt>
<dd><p>Appends a given module at the end of the list.</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>module</strong> (<a class="reference internal" href="#torch.nn.Module" title="torch.nn.Module"><em>nn.Module</em></a>) – module to append</td>
</tr>
</tbody>
</table>
</dd></dl>
<dl class="method">
<dt id="torch.nn.ModuleList.extend">
<code class="descname">extend</code><span class="sig-paren">(</span><em>modules</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/nn/modules/container.html#ModuleList.extend"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.nn.ModuleList.extend" title="Permalink to this definition">¶</a></dt>
<dd><p>Appends modules from a Python list at the end.</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>modules</strong> (<em>list</em>) – list of modules to append</td>
</tr>
</tbody>
</table>
</dd></dl>
</dd></dl>
</div>
<div class="section" id="parameterlist">
<h3><span class="hidden-section">ParameterList</span><a class="headerlink" href="#parameterlist" title="Permalink to this headline">¶</a></h3>
<dl class="class">
<dt id="torch.nn.ParameterList">