forked from pytorch/pytorch.github.io
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtensor_attributes.html
1136 lines (927 loc) · 74.1 KB
/
tensor_attributes.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="generator" content="Docutils 0.18.1: http://docutils.sourceforge.net/" />
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Tensor Attributes — PyTorch 2.0 documentation</title>
<link rel="canonical" href="https://pytorch.org/docs/stable/tensor_attributes.html"/>
<link rel="stylesheet" href="_static/css/theme.css" type="text/css" />
<!-- <link rel="stylesheet" href="_static/pygments.css" type="text/css" /> -->
<link rel="stylesheet" href="_static/pygments.css" type="text/css" />
<link rel="stylesheet" href="_static/css/theme.css" type="text/css" />
<link rel="stylesheet" href="_static/copybutton.css" type="text/css" />
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/katex.min.css" type="text/css" />
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/katex.min.css" type="text/css" />
<link rel="stylesheet" href="_static/katex-math.css" type="text/css" />
<link rel="stylesheet" href="_static/sphinx-dropdown.css" type="text/css" />
<link rel="stylesheet" href="_static/panels-bootstrap.min.css" type="text/css" />
<link rel="stylesheet" href="_static/css/jit.css" type="text/css" />
<link rel="index" title="Index" href="genindex.html" />
<link rel="search" title="Search" href="search.html" />
<link rel="next" title="Tensor Views" href="tensor_view.html" />
<link rel="prev" title="torch.Tensor.zero_" href="generated/torch.Tensor.zero_.html" />
<!-- Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=UA-117752657-2"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag(){dataLayer.push(arguments);}
gtag('js', new Date());
gtag('config', 'UA-117752657-2');
</script>
<!-- End Google Analytics -->
<script src="_static/js/modernizr.min.js"></script>
<!-- Preload the theme fonts -->
<link rel="preload" href="_static/fonts/FreightSans/freight-sans-book.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="_static/fonts/FreightSans/freight-sans-medium.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="_static/fonts/IBMPlexMono/IBMPlexMono-Medium.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="_static/fonts/FreightSans/freight-sans-bold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="_static/fonts/FreightSans/freight-sans-medium-italic.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="_static/fonts/IBMPlexMono/IBMPlexMono-SemiBold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<!-- Preload the katex fonts -->
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Math-Italic.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Main-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Main-Bold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Size1-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Size4-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Size2-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Size3-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/[email protected]/dist/fonts/KaTeX_Caligraphic-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="stylesheet" href="https://use.fontawesome.com/releases/v5.15.2/css/all.css" integrity="sha384-vSIIfh2YWi9wW0r9iZe7RJPrKwp6bG+s9QZMoITbCckVJqGCCRhc+ccxNcdpHuYu" crossorigin="anonymous">
</head>
<div class="container-fluid header-holder tutorials-header" id="header-holder">
<div class="container">
<div class="header-container">
<a class="header-logo" href="https://pytorch.org/" aria-label="PyTorch"></a>
<div class="main-menu">
<ul>
<li>
<a href="https://pytorch.org/get-started">Get Started</a>
</li>
<li>
<a href="https://pytorch.org/ecosystem">Ecosystem</a>
</li>
<li>
<a href="https://pytorch.org/mobile">Mobile</a>
</li>
<li>
<a href="https://pytorch.org/blog/">Blog</a>
</li>
<li>
<a href="https://pytorch.org/tutorials">Tutorials</a>
</li>
<li class="active docs-active">
<div id="resourcesDropdownButton" data-toggle="resources-dropdown" class="resources-dropdown">
<a class="resource-option with-down-orange-arrow">
Docs
</a>
<div class="resources-dropdown-menu">
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/docs/stable/index.html">
<span class="dropdown-title">PyTorch</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/audio/stable/index.html">
<span class="dropdown-title">torchaudio</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/text/stable/index.html">
<span class="dropdown-title">torchtext</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/vision/stable/index.html">
<span class="dropdown-title">torchvision</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/torcharrow">
<span class="dropdown-title">torcharrow</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/data">
<span class="dropdown-title">TorchData</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/torchrec">
<span class="dropdown-title">TorchRec</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/serve/">
<span class="dropdown-title">TorchServe</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/torchx/">
<span class="dropdown-title">TorchX</span>
<p></p>
</a>
<a class="doc-dropdown-option nav-dropdown-item" href="https://pytorch.org/xla">
<span class="dropdown-title">PyTorch on XLA Devices</span>
<p></p>
</a>
</div>
</li>
<li>
<div id="resourcesDropdownButton" data-toggle="resources-dropdown" class="resources-dropdown">
<a class="resource-option with-down-arrow">
Resources
</a>
<div class="resources-dropdown-menu">
<a class="nav-dropdown-item" href="https://pytorch.org/features">
<span class="dropdown-title">About</span>
<p>Learn about PyTorch’s features and capabilities</p>
</a>
<a class="nav-dropdown-item" href="https://pytorch.org/foundation">
<span class="dropdown-title">PyTorch Foundation</span>
<p>Learn about the PyTorch foundation</p>
</a>
<a class="nav-dropdown-item" href="https://pytorch.org/#community-module">
<span class="dropdown-title">Community</span>
<p>Join the PyTorch developer community to contribute, learn, and get your questions answered.</p>
</a>
<a class="nav-dropdown-item" href="https://pytorch.org/community-stories">
<span class="dropdown-title">Community Stories</span>
<p>Learn how our community solves real, everyday machine learning problems with PyTorch.</p>
</a>
<a class="nav-dropdown-item" href="https://pytorch.org/resources">
<span class="dropdown-title">Developer Resources</span>
<p>Find resources and get questions answered</p>
</a>
<a class="nav-dropdown-item" href="https://pytorch.org/events">
<span class="dropdown-title">Events</span>
<p>Find events, webinars, and podcasts</p>
</a>
<a class="nav-dropdown-item" href="https://discuss.pytorch.org/" target="_blank">
<span class="dropdown-title">Forums</span>
<p>A place to discuss PyTorch code, issues, install, research</p>
</a>
<a class="nav-dropdown-item" href="https://pytorch.org/hub">
<span class="dropdown-title">Models (Beta)</span>
<p>Discover, publish, and reuse pre-trained models</p>
</a>
</div>
</div>
</li>
<li>
<a href="https://github.com/pytorch/pytorch">GitHub</a>
</li>
</ul>
</div>
<a class="main-menu-open-button" href="#" data-behavior="open-mobile-menu"></a>
</div>
</div>
</div>
<body class="pytorch-body">
<div class="table-of-contents-link-wrapper">
<span>Table of Contents</span>
<a href="#" class="toggle-table-of-contents" data-behavior="toggle-table-of-contents"></a>
</div>
<nav data-toggle="wy-nav-shift" class="pytorch-left-menu" id="pytorch-left-menu">
<div class="pytorch-side-scroll">
<div class="pytorch-menu pytorch-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
<div class="pytorch-left-menu-search">
<div class="version">
<a href='https://pytorch.org/docs/versions.html'>2.0 ▼</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>
<p class="caption" role="heading"><span class="caption-text">Community</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="community/build_ci_governance.html">PyTorch Governance | Build + CI</a></li>
<li class="toctree-l1"><a class="reference internal" href="community/contribution_guide.html">PyTorch Contribution Guide</a></li>
<li class="toctree-l1"><a class="reference internal" href="community/design.html">PyTorch Design Philosophy</a></li>
<li class="toctree-l1"><a class="reference internal" href="community/governance.html">PyTorch Governance | Mechanics</a></li>
<li class="toctree-l1"><a class="reference internal" href="community/persons_of_interest.html">PyTorch Governance | Maintainers</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Developer Notes</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="notes/amp_examples.html">CUDA Automatic Mixed Precision examples</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/autograd.html">Autograd mechanics</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/broadcasting.html">Broadcasting semantics</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/cpu_threading_torchscript_inference.html">CPU threading and TorchScript inference</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/cuda.html">CUDA semantics</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/ddp.html">Distributed Data Parallel</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/extending.html">Extending PyTorch</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/extending.func.html">Extending torch.func with autograd.Function</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/faq.html">Frequently Asked Questions</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/gradcheck.html">Gradcheck mechanics</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/hip.html">HIP (ROCm) semantics</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/large_scale_deployments.html">Features for large-scale deployments</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/modules.html">Modules</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/mps.html">MPS backend</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/multiprocessing.html">Multiprocessing best practices</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/numerical_accuracy.html">Numerical accuracy</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/randomness.html">Reproducibility</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/serialization.html">Serialization semantics</a></li>
<li class="toctree-l1"><a class="reference internal" href="notes/windows.html">Windows FAQ</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">torch.compile</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="dynamo/index.html">TorchDynamo Overview</a></li>
<li class="toctree-l1"><a class="reference internal" href="dynamo/installation.html">Installing TorchDynamo</a></li>
<li class="toctree-l1"><a class="reference internal" href="dynamo/get-started.html">Getting Started</a></li>
<li class="toctree-l1"><a class="reference internal" href="dynamo/guards-overview.html">Guards Overview</a></li>
<li class="toctree-l1"><a class="reference internal" href="dynamo/custom-backends.html">Custom Backends</a></li>
<li class="toctree-l1"><a class="reference internal" href="dynamo/deep-dive.html">TorchDynamo Deeper Dive</a></li>
<li class="toctree-l1"><a class="reference internal" href="dynamo/troubleshooting.html">TorchDynamo Troubleshooting</a></li>
<li class="toctree-l1"><a class="reference internal" href="dynamo/faq.html">Frequently Asked Questions</a></li>
<li class="toctree-l1"><a class="reference internal" href="ir.html">IRs</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Language Bindings</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="cpp_index.html">C++</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/javadoc/">Javadoc</a></li>
<li class="toctree-l1"><a class="reference internal" href="deploy.html">torch::deploy</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Python API</span></p>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="torch.html">torch</a></li>
<li class="toctree-l1"><a class="reference internal" href="nn.html">torch.nn</a></li>
<li class="toctree-l1"><a class="reference internal" href="nn.functional.html">torch.nn.functional</a></li>
<li class="toctree-l1"><a class="reference internal" href="tensors.html">torch.Tensor</a></li>
<li class="toctree-l1 current"><a class="current reference internal" href="#">Tensor Attributes</a></li>
<li class="toctree-l1"><a class="reference internal" href="tensor_view.html">Tensor Views</a></li>
<li class="toctree-l1"><a class="reference internal" href="amp.html">torch.amp</a></li>
<li class="toctree-l1"><a class="reference internal" href="autograd.html">torch.autograd</a></li>
<li class="toctree-l1"><a class="reference internal" href="library.html">torch.library</a></li>
<li class="toctree-l1"><a class="reference internal" href="cuda.html">torch.cuda</a></li>
<li class="toctree-l1"><a class="reference internal" href="mps.html">torch.mps</a></li>
<li class="toctree-l1"><a class="reference internal" href="backends.html">torch.backends</a></li>
<li class="toctree-l1"><a class="reference internal" href="distributed.html">torch.distributed</a></li>
<li class="toctree-l1"><a class="reference internal" href="distributed.algorithms.join.html">torch.distributed.algorithms.join</a></li>
<li class="toctree-l1"><a class="reference internal" href="distributed.elastic.html">torch.distributed.elastic</a></li>
<li class="toctree-l1"><a class="reference internal" href="fsdp.html">torch.distributed.fsdp</a></li>
<li class="toctree-l1"><a class="reference internal" href="distributed.optim.html">torch.distributed.optim</a></li>
<li class="toctree-l1"><a class="reference internal" href="distributed.tensor.parallel.html">torch.distributed.tensor.parallel</a></li>
<li class="toctree-l1"><a class="reference internal" href="distributed.checkpoint.html">torch.distributed.checkpoint</a></li>
<li class="toctree-l1"><a class="reference internal" href="distributions.html">torch.distributions</a></li>
<li class="toctree-l1"><a class="reference internal" href="_dynamo.html">torch._dynamo</a></li>
<li class="toctree-l1"><a class="reference internal" href="fft.html">torch.fft</a></li>
<li class="toctree-l1"><a class="reference internal" href="func.html">torch.func</a></li>
<li class="toctree-l1"><a class="reference internal" href="futures.html">torch.futures</a></li>
<li class="toctree-l1"><a class="reference internal" href="fx.html">torch.fx</a></li>
<li class="toctree-l1"><a class="reference internal" href="hub.html">torch.hub</a></li>
<li class="toctree-l1"><a class="reference internal" href="jit.html">torch.jit</a></li>
<li class="toctree-l1"><a class="reference internal" href="linalg.html">torch.linalg</a></li>
<li class="toctree-l1"><a class="reference internal" href="monitor.html">torch.monitor</a></li>
<li class="toctree-l1"><a class="reference internal" href="signal.html">torch.signal</a></li>
<li class="toctree-l1"><a class="reference internal" href="special.html">torch.special</a></li>
<li class="toctree-l1"><a class="reference internal" href="torch.overrides.html">torch.overrides</a></li>
<li class="toctree-l1"><a class="reference internal" href="package.html">torch.package</a></li>
<li class="toctree-l1"><a class="reference internal" href="profiler.html">torch.profiler</a></li>
<li class="toctree-l1"><a class="reference internal" href="nn.init.html">torch.nn.init</a></li>
<li class="toctree-l1"><a class="reference internal" href="onnx.html">torch.onnx</a></li>
<li class="toctree-l1"><a class="reference internal" href="onnx_diagnostics.html">torch.onnx diagnostics</a></li>
<li class="toctree-l1"><a class="reference internal" href="optim.html">torch.optim</a></li>
<li class="toctree-l1"><a class="reference internal" href="complex_numbers.html">Complex Numbers</a></li>
<li class="toctree-l1"><a class="reference internal" href="ddp_comm_hooks.html">DDP Communication Hooks</a></li>
<li class="toctree-l1"><a class="reference internal" href="pipeline.html">Pipeline Parallelism</a></li>
<li class="toctree-l1"><a class="reference internal" href="quantization.html">Quantization</a></li>
<li class="toctree-l1"><a class="reference internal" href="rpc.html">Distributed RPC Framework</a></li>
<li class="toctree-l1"><a class="reference internal" href="random.html">torch.random</a></li>
<li class="toctree-l1"><a class="reference internal" href="masked.html">torch.masked</a></li>
<li class="toctree-l1"><a class="reference internal" href="nested.html">torch.nested</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"><a class="reference internal" href="testing.html">torch.testing</a></li>
<li class="toctree-l1"><a class="reference internal" href="benchmark_utils.html">torch.utils.benchmark</a></li>
<li class="toctree-l1"><a class="reference internal" href="bottleneck.html">torch.utils.bottleneck</a></li>
<li class="toctree-l1"><a class="reference internal" href="checkpoint.html">torch.utils.checkpoint</a></li>
<li class="toctree-l1"><a class="reference internal" href="cpp_extension.html">torch.utils.cpp_extension</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="jit_utils.html">torch.utils.jit</a></li>
<li class="toctree-l1"><a class="reference internal" href="dlpack.html">torch.utils.dlpack</a></li>
<li class="toctree-l1"><a class="reference internal" href="mobile_optimizer.html">torch.utils.mobile_optimizer</a></li>
<li class="toctree-l1"><a class="reference internal" href="model_zoo.html">torch.utils.model_zoo</a></li>
<li class="toctree-l1"><a class="reference internal" href="tensorboard.html">torch.utils.tensorboard</a></li>
<li class="toctree-l1"><a class="reference internal" href="type_info.html">Type Info</a></li>
<li class="toctree-l1"><a class="reference internal" href="named_tensor.html">Named Tensors</a></li>
<li class="toctree-l1"><a class="reference internal" href="name_inference.html">Named Tensors operator coverage</a></li>
<li class="toctree-l1"><a class="reference internal" href="config_mod.html">torch.__config__</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Libraries</span></p>
<ul>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/audio/stable">torchaudio</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/data">TorchData</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/torchrec">TorchRec</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/serve">TorchServe</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/text/stable">torchtext</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/vision/stable">torchvision</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/xla/">PyTorch on XLA Devices</a></li>
</ul>
</div>
</div>
</nav>
<div class="pytorch-container">
<div class="pytorch-page-level-bar" id="pytorch-page-level-bar">
<div class="pytorch-breadcrumbs-wrapper">
<div role="navigation" aria-label="breadcrumbs navigation">
<ul class="pytorch-breadcrumbs">
<li>
<a href="index.html">
Docs
</a> >
</li>
<li>Tensor Attributes</li>
<li class="pytorch-breadcrumbs-aside">
<a href="_sources/tensor_attributes.rst.txt" rel="nofollow"><img src="_static/images/view-page-source-icon.svg"></a>
</li>
</ul>
</div>
</div>
<div class="pytorch-shortcuts-wrapper" id="pytorch-shortcuts-wrapper">
Shortcuts
</div>
</div>
<section data-toggle="wy-nav-shift" id="pytorch-content-wrap" class="pytorch-content-wrap">
<div class="pytorch-content-left">
<div class="rst-content">
<div role="main" class="main-content" itemscope="itemscope" itemtype="http://schema.org/Article">
<article itemprop="articleBody" id="pytorch-article" class="pytorch-article">
<section id="tensor-attributes">
<span id="tensor-attributes-doc"></span><h1>Tensor Attributes<a class="headerlink" href="#tensor-attributes" title="Permalink to this heading">¶</a></h1>
<p>Each <code class="docutils literal notranslate"><span class="pre">torch.Tensor</span></code> has a <a class="reference internal" href="#torch.dtype" title="torch.dtype"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.dtype</span></code></a>, <a class="reference internal" href="#torch.device" title="torch.device"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.device</span></code></a>, and <a class="reference internal" href="#torch.layout" title="torch.layout"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.layout</span></code></a>.</p>
<section id="torch-dtype">
<span id="dtype-doc"></span><h2>torch.dtype<a class="headerlink" href="#torch-dtype" title="Permalink to this heading">¶</a></h2>
<dl class="py class">
<dt class="sig sig-object py" id="torch.dtype">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">torch.</span></span><span class="sig-name descname"><span class="pre">dtype</span></span><a class="headerlink" href="#torch.dtype" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<p>A <a class="reference internal" href="#torch.dtype" title="torch.dtype"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.dtype</span></code></a> is an object that represents the data type of a
<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.Tensor</span></code></a>. PyTorch has twelve different data types:</p>
<table class="docutils align-default">
<thead>
<tr class="row-odd"><th class="head"><p>Data type</p></th>
<th class="head"><p>dtype</p></th>
<th class="head"><p>Legacy Constructors</p></th>
</tr>
</thead>
<tbody>
<tr class="row-even"><td><p>32-bit floating point</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.float32</span></code> or <code class="docutils literal notranslate"><span class="pre">torch.float</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.*.FloatTensor</span></code></p></td>
</tr>
<tr class="row-odd"><td><p>64-bit floating point</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.float64</span></code> or <code class="docutils literal notranslate"><span class="pre">torch.double</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.*.DoubleTensor</span></code></p></td>
</tr>
<tr class="row-even"><td><p>64-bit complex</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.complex64</span></code> or <code class="docutils literal notranslate"><span class="pre">torch.cfloat</span></code></p></td>
<td></td>
</tr>
<tr class="row-odd"><td><p>128-bit complex</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.complex128</span></code> or <code class="docutils literal notranslate"><span class="pre">torch.cdouble</span></code></p></td>
<td></td>
</tr>
<tr class="row-even"><td><p>16-bit floating point <a class="footnote-reference brackets" href="#id3" id="id1" role="doc-noteref"><span class="fn-bracket">[</span>1<span class="fn-bracket">]</span></a></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.float16</span></code> or <code class="docutils literal notranslate"><span class="pre">torch.half</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.*.HalfTensor</span></code></p></td>
</tr>
<tr class="row-odd"><td><p>16-bit floating point <a class="footnote-reference brackets" href="#id4" id="id2" role="doc-noteref"><span class="fn-bracket">[</span>2<span class="fn-bracket">]</span></a></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.bfloat16</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.*.BFloat16Tensor</span></code></p></td>
</tr>
<tr class="row-even"><td><p>8-bit integer (unsigned)</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.uint8</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.*.ByteTensor</span></code></p></td>
</tr>
<tr class="row-odd"><td><p>8-bit integer (signed)</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.int8</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.*.CharTensor</span></code></p></td>
</tr>
<tr class="row-even"><td><p>16-bit integer (signed)</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.int16</span></code> or <code class="docutils literal notranslate"><span class="pre">torch.short</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.*.ShortTensor</span></code></p></td>
</tr>
<tr class="row-odd"><td><p>32-bit integer (signed)</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.int32</span></code> or <code class="docutils literal notranslate"><span class="pre">torch.int</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.*.IntTensor</span></code></p></td>
</tr>
<tr class="row-even"><td><p>64-bit integer (signed)</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.int64</span></code> or <code class="docutils literal notranslate"><span class="pre">torch.long</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.*.LongTensor</span></code></p></td>
</tr>
<tr class="row-odd"><td><p>Boolean</p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.bool</span></code></p></td>
<td><p><code class="docutils literal notranslate"><span class="pre">torch.*.BoolTensor</span></code></p></td>
</tr>
</tbody>
</table>
<aside class="footnote brackets" id="id3" role="note">
<span class="label"><span class="fn-bracket">[</span><a role="doc-backlink" href="#id1">1</a><span class="fn-bracket">]</span></span>
<p>Sometimes referred to as binary16: uses 1 sign, 5 exponent, and 10
significand bits. Useful when precision is important.</p>
</aside>
<aside class="footnote brackets" id="id4" role="note">
<span class="label"><span class="fn-bracket">[</span><a role="doc-backlink" href="#id2">2</a><span class="fn-bracket">]</span></span>
<p>Sometimes referred to as Brain Floating Point: use 1 sign, 8 exponent and 7
significand bits. Useful when range is important, since it has the same
number of exponent bits as <code class="docutils literal notranslate"><span class="pre">float32</span></code></p>
</aside>
<p>To find out if a <a class="reference internal" href="#torch.dtype" title="torch.dtype"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.dtype</span></code></a> is a floating point data type, the property <a class="reference internal" href="generated/torch.is_floating_point.html#torch.is_floating_point" title="torch.is_floating_point"><code class="xref py py-attr docutils literal notranslate"><span class="pre">is_floating_point</span></code></a>
can be used, which returns <code class="docutils literal notranslate"><span class="pre">True</span></code> if the data type is a floating point data type.</p>
<p>To find out if a <a class="reference internal" href="#torch.dtype" title="torch.dtype"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.dtype</span></code></a> is a complex data type, the property <a class="reference internal" href="generated/torch.is_complex.html#torch.is_complex" title="torch.is_complex"><code class="xref py py-attr docutils literal notranslate"><span class="pre">is_complex</span></code></a>
can be used, which returns <code class="docutils literal notranslate"><span class="pre">True</span></code> if the data type is a complex data type.</p>
<p id="type-promotion-doc">When the dtypes of inputs to an arithmetic operation (<cite>add</cite>, <cite>sub</cite>, <cite>div</cite>, <cite>mul</cite>) differ, we promote
by finding the minimum dtype that satisfies the following rules:</p>
<ul class="simple">
<li><p>If the type of a scalar operand is of a higher category than tensor operands
(where complex > floating > integral > boolean), we promote to a type with sufficient size to hold
all scalar operands of that category.</p></li>
<li><p>If a zero-dimension tensor operand has a higher category than dimensioned operands,
we promote to a type with sufficient size and category to hold all zero-dim tensor operands of
that category.</p></li>
<li><p>If there are no higher-category zero-dim operands, we promote to a type with sufficient size
and category to hold all dimensioned operands.</p></li>
</ul>
<p>A floating point scalar operand has dtype <cite>torch.get_default_dtype()</cite> and an integral
non-boolean scalar operand has dtype <cite>torch.int64</cite>. Unlike numpy, we do not inspect
values when determining the minimum <cite>dtypes</cite> of an operand. Quantized and complex types
are not yet supported.</p>
<p>Promotion Examples:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">float_tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">ones</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">float</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">double_tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">ones</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">double</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">complex_float_tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">ones</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">complex64</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">complex_double_tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">ones</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">complex128</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">int_tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">ones</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">int</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">long_tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">ones</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">long</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">uint_tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">ones</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">uint8</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">double_tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">ones</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">double</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">bool_tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">ones</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">bool</span><span class="p">)</span>
<span class="go"># zero-dim tensors</span>
<span class="gp">>>> </span><span class="n">long_zerodim</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">long</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">int_zerodim</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">int</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">add</span><span class="p">(</span><span class="mi">5</span><span class="p">,</span> <span class="mi">5</span><span class="p">)</span><span class="o">.</span><span class="n">dtype</span>
<span class="go">torch.int64</span>
<span class="go"># 5 is an int64, but does not have higher category than int_tensor so is not considered.</span>
<span class="gp">>>> </span><span class="p">(</span><span class="n">int_tensor</span> <span class="o">+</span> <span class="mi">5</span><span class="p">)</span><span class="o">.</span><span class="n">dtype</span>
<span class="go">torch.int32</span>
<span class="gp">>>> </span><span class="p">(</span><span class="n">int_tensor</span> <span class="o">+</span> <span class="n">long_zerodim</span><span class="p">)</span><span class="o">.</span><span class="n">dtype</span>
<span class="go">torch.int32</span>
<span class="gp">>>> </span><span class="p">(</span><span class="n">long_tensor</span> <span class="o">+</span> <span class="n">int_tensor</span><span class="p">)</span><span class="o">.</span><span class="n">dtype</span>
<span class="go">torch.int64</span>
<span class="gp">>>> </span><span class="p">(</span><span class="n">bool_tensor</span> <span class="o">+</span> <span class="n">long_tensor</span><span class="p">)</span><span class="o">.</span><span class="n">dtype</span>
<span class="go">torch.int64</span>
<span class="gp">>>> </span><span class="p">(</span><span class="n">bool_tensor</span> <span class="o">+</span> <span class="n">uint_tensor</span><span class="p">)</span><span class="o">.</span><span class="n">dtype</span>
<span class="go">torch.uint8</span>
<span class="gp">>>> </span><span class="p">(</span><span class="n">float_tensor</span> <span class="o">+</span> <span class="n">double_tensor</span><span class="p">)</span><span class="o">.</span><span class="n">dtype</span>
<span class="go">torch.float64</span>
<span class="gp">>>> </span><span class="p">(</span><span class="n">complex_float_tensor</span> <span class="o">+</span> <span class="n">complex_double_tensor</span><span class="p">)</span><span class="o">.</span><span class="n">dtype</span>
<span class="go">torch.complex128</span>
<span class="gp">>>> </span><span class="p">(</span><span class="n">bool_tensor</span> <span class="o">+</span> <span class="n">int_tensor</span><span class="p">)</span><span class="o">.</span><span class="n">dtype</span>
<span class="go">torch.int32</span>
<span class="go"># Since long is a different kind than float, result dtype only needs to be large enough</span>
<span class="go"># to hold the float.</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">add</span><span class="p">(</span><span class="n">long_tensor</span><span class="p">,</span> <span class="n">float_tensor</span><span class="p">)</span><span class="o">.</span><span class="n">dtype</span>
<span class="go">torch.float32</span>
</pre></div>
</div>
<dl class="simple">
<dt>When the output tensor of an arithmetic operation is specified, we allow casting to its <cite>dtype</cite> except that:</dt><dd><ul class="simple">
<li><p>An integral output tensor cannot accept a floating point tensor.</p></li>
<li><p>A boolean output tensor cannot accept a non-boolean tensor.</p></li>
<li><p>A non-complex output tensor cannot accept a complex tensor</p></li>
</ul>
</dd>
</dl>
<p>Casting Examples:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># allowed:</span>
<span class="o">>>></span> <span class="n">float_tensor</span> <span class="o">*=</span> <span class="n">float_tensor</span>
<span class="o">>>></span> <span class="n">float_tensor</span> <span class="o">*=</span> <span class="n">int_tensor</span>
<span class="o">>>></span> <span class="n">float_tensor</span> <span class="o">*=</span> <span class="n">uint_tensor</span>
<span class="o">>>></span> <span class="n">float_tensor</span> <span class="o">*=</span> <span class="n">bool_tensor</span>
<span class="o">>>></span> <span class="n">float_tensor</span> <span class="o">*=</span> <span class="n">double_tensor</span>
<span class="o">>>></span> <span class="n">int_tensor</span> <span class="o">*=</span> <span class="n">long_tensor</span>
<span class="o">>>></span> <span class="n">int_tensor</span> <span class="o">*=</span> <span class="n">uint_tensor</span>
<span class="o">>>></span> <span class="n">uint_tensor</span> <span class="o">*=</span> <span class="n">int_tensor</span>
<span class="c1"># disallowed (RuntimeError: result type can't be cast to the desired output type):</span>
<span class="o">>>></span> <span class="n">int_tensor</span> <span class="o">*=</span> <span class="n">float_tensor</span>
<span class="o">>>></span> <span class="n">bool_tensor</span> <span class="o">*=</span> <span class="n">int_tensor</span>
<span class="o">>>></span> <span class="n">bool_tensor</span> <span class="o">*=</span> <span class="n">uint_tensor</span>
<span class="o">>>></span> <span class="n">float_tensor</span> <span class="o">*=</span> <span class="n">complex_float_tensor</span>
</pre></div>
</div>
</section>
<section id="torch-device">
<span id="device-doc"></span><h2>torch.device<a class="headerlink" href="#torch-device" title="Permalink to this heading">¶</a></h2>
<dl class="py class">
<dt class="sig sig-object py" id="torch.device">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">torch.</span></span><span class="sig-name descname"><span class="pre">device</span></span><a class="headerlink" href="#torch.device" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<p>A <a class="reference internal" href="#torch.device" title="torch.device"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.device</span></code></a> is an object representing the device on which a <a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.Tensor</span></code></a> is
or will be allocated.</p>
<p>The <a class="reference internal" href="#torch.device" title="torch.device"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.device</span></code></a> contains a device type (<code class="docutils literal notranslate"><span class="pre">'cpu'</span></code>, <code class="docutils literal notranslate"><span class="pre">'cuda'</span></code> or <code class="docutils literal notranslate"><span class="pre">'mps'</span></code>) and optional device
ordinal for the device type. If the device ordinal is not present, this object will always represent
the current device for the device type, even after <a class="reference internal" href="generated/torch.cuda.set_device.html#torch.cuda.set_device" title="torch.cuda.set_device"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.cuda.set_device()</span></code></a> is called; e.g.,
a <a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.Tensor</span></code></a> constructed with device <code class="docutils literal notranslate"><span class="pre">'cuda'</span></code> is equivalent to <code class="docutils literal notranslate"><span class="pre">'cuda:X'</span></code> where X is
the result of <a class="reference internal" href="generated/torch.cuda.current_device.html#torch.cuda.current_device" title="torch.cuda.current_device"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.cuda.current_device()</span></code></a>.</p>
<p>A <a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.Tensor</span></code></a>’s device can be accessed via the <a class="reference internal" href="generated/torch.Tensor.device.html#torch.Tensor.device" title="torch.Tensor.device"><code class="xref py py-attr docutils literal notranslate"><span class="pre">Tensor.device</span></code></a> property.</p>
<p>A <a class="reference internal" href="#torch.device" title="torch.device"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.device</span></code></a> can be constructed via a string or via a string and device ordinal</p>
<p>Via a string:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s1">'cuda:0'</span><span class="p">)</span>
<span class="go">device(type='cuda', index=0)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s1">'cpu'</span><span class="p">)</span>
<span class="go">device(type='cpu')</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s1">'mps'</span><span class="p">)</span>
<span class="go">device(type='mps')</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s1">'cuda'</span><span class="p">)</span> <span class="c1"># current cuda device</span>
<span class="go">device(type='cuda')</span>
</pre></div>
</div>
<p>Via a string and device ordinal:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s1">'cuda'</span><span class="p">,</span> <span class="mi">0</span><span class="p">)</span>
<span class="go">device(type='cuda', index=0)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s1">'mps'</span><span class="p">,</span> <span class="mi">0</span><span class="p">)</span>
<span class="go">device(type='mps', index=0)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s1">'cpu'</span><span class="p">,</span> <span class="mi">0</span><span class="p">)</span>
<span class="go">device(type='cpu', index=0)</span>
</pre></div>
</div>
<p>The device object can also be used as a context manager to change the default
device tensors are allocated on:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="k">with</span> <span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s1">'cuda:1'</span><span class="p">):</span>
<span class="gp">... </span> <span class="n">r</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">r</span><span class="o">.</span><span class="n">device</span>
<span class="go">device(type='cuda', index=1)</span>
</pre></div>
</div>
<p>This context manager has no effect if a factory function is passed an explicit,
non-None device argument. To globally change the default device, see also
<a class="reference internal" href="generated/torch.set_default_device.html#torch.set_default_device" title="torch.set_default_device"><code class="xref py py-func docutils literal notranslate"><span class="pre">torch.set_default_device()</span></code></a>.</p>
<div class="admonition warning">
<p class="admonition-title">Warning</p>
<p>This function imposes a slight performance cost on every Python
call to the torch API (not just factory functions). If this
is causing problems for you, please comment on
<a class="reference external" href="https://github.com/pytorch/pytorch/issues/92701">https://github.com/pytorch/pytorch/issues/92701</a></p>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>The <a class="reference internal" href="#torch.device" title="torch.device"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.device</span></code></a> argument in functions can generally be substituted with a string.
This allows for fast prototyping of code.</p>
<div class="doctest highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="c1"># Example of a function that takes in a torch.device</span>
<span class="gp">>>> </span><span class="n">cuda1</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s1">'cuda:1'</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">((</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">),</span> <span class="n">device</span><span class="o">=</span><span class="n">cuda1</span><span class="p">)</span>
</pre></div>
</div>
<div class="doctest highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="c1"># You can substitute the torch.device with a string</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">((</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">),</span> <span class="n">device</span><span class="o">=</span><span class="s1">'cuda:1'</span><span class="p">)</span>
</pre></div>
</div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>For legacy reasons, a device can be constructed via a single device ordinal, which is treated
as a cuda device. This matches <a class="reference internal" href="generated/torch.Tensor.get_device.html#torch.Tensor.get_device" title="torch.Tensor.get_device"><code class="xref py py-meth docutils literal notranslate"><span class="pre">Tensor.get_device()</span></code></a>, which returns an ordinal for cuda
tensors and is not supported for cpu tensors.</p>
<div class="doctest highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span>
<span class="go">device(type='cuda', index=1)</span>
</pre></div>
</div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>Methods which take a device will generally accept a (properly formatted) string
or (legacy) integer device ordinal, i.e. the following are all equivalent:</p>
<div class="doctest highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">((</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">),</span> <span class="n">device</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">device</span><span class="p">(</span><span class="s1">'cuda:1'</span><span class="p">))</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">((</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">),</span> <span class="n">device</span><span class="o">=</span><span class="s1">'cuda:1'</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">torch</span><span class="o">.</span><span class="n">randn</span><span class="p">((</span><span class="mi">2</span><span class="p">,</span><span class="mi">3</span><span class="p">),</span> <span class="n">device</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span> <span class="c1"># legacy</span>
</pre></div>
</div>
</div>
</section>
<section id="torch-layout">
<span id="layout-doc"></span><h2>torch.layout<a class="headerlink" href="#torch-layout" title="Permalink to this heading">¶</a></h2>
<dl class="py class">
<dt class="sig sig-object py" id="torch.layout">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">torch.</span></span><span class="sig-name descname"><span class="pre">layout</span></span><a class="headerlink" href="#torch.layout" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<div class="admonition warning">
<p class="admonition-title">Warning</p>
<p>The <code class="docutils literal notranslate"><span class="pre">torch.layout</span></code> class is in beta and subject to change.</p>
</div>
<p>A <a class="reference internal" href="#torch.layout" title="torch.layout"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.layout</span></code></a> is an object that represents the memory layout of a
<a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.Tensor</span></code></a>. Currently, we support <code class="docutils literal notranslate"><span class="pre">torch.strided</span></code> (dense Tensors)
and have beta support for <code class="docutils literal notranslate"><span class="pre">torch.sparse_coo</span></code> (sparse COO Tensors).</p>
<p><code class="docutils literal notranslate"><span class="pre">torch.strided</span></code> represents dense Tensors and is the memory layout that
is most commonly used. Each strided tensor has an associated
<code class="xref py py-class docutils literal notranslate"><span class="pre">torch.Storage</span></code>, which holds its data. These tensors provide
multi-dimensional, <a class="reference external" href="https://en.wikipedia.org/wiki/Stride_of_an_array">strided</a>
view of a storage. Strides are a list of integers: the k-th stride
represents the jump in the memory necessary to go from one element to the
next one in the k-th dimension of the Tensor. This concept makes it possible
to perform many tensor operations efficiently.</p>
<p>Example:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">x</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">([[</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">5</span><span class="p">],</span> <span class="p">[</span><span class="mi">6</span><span class="p">,</span> <span class="mi">7</span><span class="p">,</span> <span class="mi">8</span><span class="p">,</span> <span class="mi">9</span><span class="p">,</span> <span class="mi">10</span><span class="p">]])</span>
<span class="gp">>>> </span><span class="n">x</span><span class="o">.</span><span class="n">stride</span><span class="p">()</span>
<span class="go">(5, 1)</span>
<span class="gp">>>> </span><span class="n">x</span><span class="o">.</span><span class="n">t</span><span class="p">()</span><span class="o">.</span><span class="n">stride</span><span class="p">()</span>
<span class="go">(1, 5)</span>
</pre></div>
</div>
<p>For more information on <code class="docutils literal notranslate"><span class="pre">torch.sparse_coo</span></code> tensors, see <a class="reference internal" href="sparse.html#sparse-docs"><span class="std std-ref">torch.sparse</span></a>.</p>
</section>
<section id="torch-memory-format">
<h2>torch.memory_format<a class="headerlink" href="#torch-memory-format" title="Permalink to this heading">¶</a></h2>
<dl class="py class">
<dt class="sig sig-object py" id="torch.memory_format">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">torch.</span></span><span class="sig-name descname"><span class="pre">memory_format</span></span><a class="headerlink" href="#torch.memory_format" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<p>A <a class="reference internal" href="#torch.memory_format" title="torch.memory_format"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.memory_format</span></code></a> is an object representing the memory format on which a <a class="reference internal" href="tensors.html#torch.Tensor" title="torch.Tensor"><code class="xref py py-class docutils literal notranslate"><span class="pre">torch.Tensor</span></code></a> is
or will be allocated.</p>
<p>Possible values are:</p>
<ul class="simple">
<li><p><code class="docutils literal notranslate"><span class="pre">torch.contiguous_format</span></code>:
Tensor is or will be allocated in dense non-overlapping memory. Strides represented by values in decreasing order.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">torch.channels_last</span></code>:
Tensor is or will be allocated in dense non-overlapping memory. Strides represented by values in
<code class="docutils literal notranslate"><span class="pre">strides[0]</span> <span class="pre">></span> <span class="pre">strides[2]</span> <span class="pre">></span> <span class="pre">strides[3]</span> <span class="pre">></span> <span class="pre">strides[1]</span> <span class="pre">==</span> <span class="pre">1</span></code> aka NHWC order.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">torch.channels_last_3d</span></code>:
Tensor is or will be allocated in dense non-overlapping memory. Strides represented by values in
<code class="docutils literal notranslate"><span class="pre">strides[0]</span> <span class="pre">></span> <span class="pre">strides[2]</span> <span class="pre">></span> <span class="pre">strides[3]</span> <span class="pre">></span> <span class="pre">strides[4]</span> <span class="pre">></span> <span class="pre">strides[1]</span> <span class="pre">==</span> <span class="pre">1</span></code> aka NDHWC order.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">torch.preserve_format</span></code>:
Used in functions like <cite>clone</cite> to preserve the memory format of the input tensor. If input tensor is
allocated in dense non-overlapping memory, the output tensor strides will be copied from the input.
Otherwise output strides will follow <code class="docutils literal notranslate"><span class="pre">torch.contiguous_format</span></code></p></li>
</ul>
</section>
</section>
</article>
</div>
<footer>
<div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
<a href="tensor_view.html" class="btn btn-neutral float-right" title="Tensor Views" accesskey="n" rel="next">Next <img src="_static/images/chevron-right-orange.svg" class="next-page"></a>
<a href="generated/torch.Tensor.zero_.html" class="btn btn-neutral" title="torch.Tensor.zero_" accesskey="p" rel="prev"><img src="_static/images/chevron-right-orange.svg" class="previous-page"> Previous</a>
</div>
<hr>
<div role="contentinfo">
<p>
© Copyright 2023, PyTorch Contributors.
</p>
</div>
<div>
Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/rtfd/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>.
</div>
</footer>
</div>
<script>
var match = window.location.href.match(/\/_[a-zA-Z0-9_]*.html|_dynamo/gi);
var url = window.location.href.lastIndexOf(match[match.length-1]);
if (url)
{
var div = '<div class="admonition note"><p class="admonition-title">Note</p><p><i class="fa fa-exclamation-circle" aria-hidden="true"> </i> This page describes an internal API which is not intended to be used outside of the PyTorch codebase and can be modified or removed without notice.</p></div>'
document.getElementById("pytorch-article").insertAdjacentHTML('afterBegin', div)
}
</script>
</div>
<div class="pytorch-content-right" id="pytorch-content-right">
<div class="pytorch-right-menu" id="pytorch-right-menu">
<div class="pytorch-side-scroll" id="pytorch-side-scroll-right">
<ul>
<li><a class="reference internal" href="#">Tensor Attributes</a><ul>
<li><a class="reference internal" href="#torch-dtype">torch.dtype</a></li>
<li><a class="reference internal" href="#torch-device">torch.device</a></li>
<li><a class="reference internal" href="#torch-layout">torch.layout</a></li>
<li><a class="reference internal" href="#torch-memory-format">torch.memory_format</a></li>
</ul>
</li>
</ul>
</div>
</div>
</div>
</section>
</div>
<script type="text/javascript" id="documentation_options" data-url_root="./" src="_static/documentation_options.js"></script>
<script data-url_root="./" id="documentation_options" src="_static/documentation_options.js"></script>
<script src="_static/jquery.js"></script>
<script src="_static/underscore.js"></script>
<script src="_static/_sphinx_javascript_frameworks_compat.js"></script>
<script src="_static/doctools.js"></script>
<script src="_static/clipboard.min.js"></script>
<script src="_static/copybutton.js"></script>
<script type="text/javascript" src="_static/js/vendor/popper.min.js"></script>
<script type="text/javascript" src="_static/js/vendor/bootstrap.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/list.js/1.5.0/list.min.js"></script>
<script type="text/javascript" src="_static/js/theme.js"></script>
<script type="text/javascript">
jQuery(function () {
SphinxRtdTheme.Navigation.enable(true);
});
</script>
<script script type="text/javascript">
var collapsedSections = ['Developer Notes', 'Language Bindings', 'Libraries', 'Community'];
</script>
<img height="1" width="1" style="border-style:none;" alt="" src="https://www.googleadservices.com/pagead/conversion/795629140/?label=txkmCPmdtosBENSssfsC&guid=ON&script=0"/>
<!-- Begin Footer -->
<div class="container-fluid docs-tutorials-resources" id="docs-tutorials-resources">
<div class="container">
<div class="row">
<div class="col-md-4 text-center">
<h2>Docs</h2>
<p>Access comprehensive developer documentation for PyTorch</p>
<a class="with-right-arrow" href="https://pytorch.org/docs/stable/index.html">View Docs</a>
</div>
<div class="col-md-4 text-center">
<h2>Tutorials</h2>
<p>Get in-depth tutorials for beginners and advanced developers</p>
<a class="with-right-arrow" href="https://pytorch.org/tutorials">View Tutorials</a>
</div>
<div class="col-md-4 text-center">
<h2>Resources</h2>
<p>Find development resources and get your questions answered</p>
<a class="with-right-arrow" href="https://pytorch.org/resources">View Resources</a>
</div>
</div>
</div>
</div>
<footer class="site-footer">
<div class="container footer-container">
<div class="footer-logo-wrapper">
<a href="https://pytorch.org/" class="footer-logo"></a>
</div>
<div class="footer-links-wrapper">
<div class="footer-links-col">
<ul>
<li class="list-title"><a href="https://pytorch.org/">PyTorch</a></li>
<li><a href="https://pytorch.org/get-started">Get Started</a></li>
<li><a href="https://pytorch.org/features">Features</a></li>
<li><a href="https://pytorch.org/ecosystem">Ecosystem</a></li>
<li><a href="https://pytorch.org/blog/">Blog</a></li>
<li><a href="https://github.com/pytorch/pytorch/blob/master/CONTRIBUTING.md">Contributing</a></li>
</ul>
</div>
<div class="footer-links-col">
<ul>
<li class="list-title"><a href="https://pytorch.org/resources">Resources</a></li>
<li><a href="https://pytorch.org/tutorials">Tutorials</a></li>
<li><a href="https://pytorch.org/docs/stable/index.html">Docs</a></li>
<li><a href="https://discuss.pytorch.org" target="_blank">Discuss</a></li>
<li><a href="https://github.com/pytorch/pytorch/issues" target="_blank">Github Issues</a></li>
<li><a href="https://pytorch.org/assets/brand-guidelines/PyTorch-Brand-Guidelines.pdf" target="_blank">Brand Guidelines</a></li>
</ul>
</div>
<div class="footer-links-col">
<ul>
<li class="list-title">Stay up to date</li>
<li><a href="https://www.facebook.com/pytorch" target="_blank">Facebook</a></li>
<li><a href="https://twitter.com/pytorch" target="_blank">Twitter</a></li>
<li><a href="https://www.youtube.com/pytorch" target="_blank">YouTube</a></li>
<li><a href="https://www.linkedin.com/company/pytorch" target="_blank">LinkedIn</a></li>
</ul>
</div>
<div class="footer-links-col">
<ul>
<li class="list-title">PyTorch Podcasts</li>
<li><a href="https://open.spotify.com/show/6UzHKeiy368jKfQMKKvJY5" target="_blank">Spotify</a></li>
<li><a href="https://podcasts.apple.com/us/podcast/pytorch-developer-podcast/id1566080008" target="_blank">Apple</a></li>
<li><a href="https://www.google.com/podcasts?feed=aHR0cHM6Ly9mZWVkcy5zaW1wbGVjYXN0LmNvbS9PQjVGa0lsOA%3D%3D" target="_blank">Google</a></li>
<li><a href="https://music.amazon.com/podcasts/7a4e6f0e-26c2-49e9-a478-41bd244197d0/PyTorch-Developer-Podcast?" target="_blank">Amazon</a></li>
</ul>
</div>
</div>
<div class="privacy-policy">
<ul>
<li class="privacy-policy-links"><a href="https://www.linuxfoundation.org/terms/" target="_blank">Terms</a></li>
<li class="privacy-policy-links">|</li>
<li class="privacy-policy-links"><a href="https://www.linuxfoundation.org/privacy-policy/" target="_blank">Privacy</a></li>
</ul>
</div>
<div class="copyright">
<p>© Copyright The Linux Foundation. The PyTorch Foundation is a project of The Linux Foundation.
For web site terms of use, trademark policy and other policies applicable to The PyTorch Foundation please see
<a href="www.linuxfoundation.org/policies/">www.linuxfoundation.org/policies/</a>. The PyTorch Foundation supports the PyTorch open source
project, which has been established as PyTorch Project a Series of LF Projects, LLC. For policies applicable to the PyTorch Project a Series of LF Projects, LLC,
please see <a href="www.lfprojects.org/policies/">www.lfprojects.org/policies/</a>.</p>
</div>
</div>
</footer>
<div class="cookie-banner-wrapper">
<div class="container">
<p class="gdpr-notice">To analyze traffic and optimize your experience, we serve cookies on this site. By clicking or navigating, you agree to allow our usage of cookies. As the current maintainers of this site, Facebook’s Cookies Policy applies. Learn more, including about available controls: <a href="https://www.facebook.com/policies/cookies/">Cookies Policy</a>.</p>
<img class="close-button" src="_static/images/pytorch-x.svg">
</div>
</div>
<!-- End Footer -->
<!-- Begin Mobile Menu -->
<div class="mobile-main-menu">
<div class="container-fluid">
<div class="container">
<div class="mobile-main-menu-header-container">
<a class="header-logo" href="https://pytorch.org/" aria-label="PyTorch"></a>
<a class="main-menu-close-button" href="#" data-behavior="close-mobile-menu"></a>
</div>
</div>
</div>
<div class="mobile-main-menu-links-container">
<div class="main-menu">
<ul>