forked from pytorch/pytorch.github.io
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathhub.html
910 lines (714 loc) · 50.4 KB
/
hub.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
<!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 name="robots" content="noindex">
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>torch.hub — PyTorch master documentation</title>
<link rel="canonical" href="https://pytorch.org/docs/stable/hub.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="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="index" title="Index" href="genindex.html" />
<link rel="search" title="Search" href="search.html" />
<link rel="next" title="TorchScript" href="jit.html" />
<link rel="prev" title="Probability distributions - torch.distributions" href="distributions.html" />
<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">
</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>
<div class="ecosystem-dropdown">
<a id="dropdownMenuButton" data-toggle="ecosystem-dropdown">
Ecosystem
</a>
<div class="ecosystem-dropdown-menu">
<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>
<a class="nav-dropdown-item" href="https://pytorch.org/ecosystem">
<span class=dropdown-title>Tools & Libraries</span>
<p>Explore the ecosystem of tools and libraries</p>
</a>
</div>
</div>
</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">
<a href="https://pytorch.org/docs/stable/index.html">Docs</a>
</li>
<li>
<div class="resources-dropdown">
<a id="resourcesDropdownButton" data-toggle="resources-dropdown">
Resources
</a>
<div class="resources-dropdown-menu">
<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/features">
<span class=dropdown-title>About</span>
<p>Learn about PyTorch’s features and capabilities</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='http://pytorch.org/docs/versions.html'>1.4.0 ▼</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"><span class="caption-text">Notes</span></p>
<ul>
<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/distributed_autograd.html">Distributed Autograd Design</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/faq.html">Frequently Asked Questions</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/multiprocessing.html">Multiprocessing best practices</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/rref.html">Remote Reference Protocol</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>
<li class="toctree-l1"><a class="reference external" href="http://pytorch.org/xla/">PyTorch on XLA Devices</a></li>
</ul>
<p class="caption"><span class="caption-text">Language Bindings</span></p>
<ul>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/cppdocs/">C++ API</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/javadoc/">Javadoc</a></li>
</ul>
<p class="caption"><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"><a class="reference internal" href="tensor_attributes.html">Tensor Attributes</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="cuda.html">torch.cuda</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="distributions.html">torch.distributions</a></li>
<li class="toctree-l1 current"><a class="current reference internal" href="#">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="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="optim.html">torch.optim</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="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="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="dlpack.html">torch.utils.dlpack</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__.html">torch.__config__</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/index.html">torchvision</a></li>
</ul>
<p class="caption"><span class="caption-text">torchaudio Reference</span></p>
<ul>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/audio">torchaudio</a></li>
</ul>
<p class="caption"><span class="caption-text">torchtext Reference</span></p>
<ul>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/text">torchtext</a></li>
</ul>
<p class="caption"><span class="caption-text">Community</span></p>
<ul>
<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/governance.html">PyTorch Governance</a></li>
<li class="toctree-l1"><a class="reference internal" href="community/persons_of_interest.html">PyTorch Governance | Persons of Interest</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>torch.hub</li>
<li class="pytorch-breadcrumbs-aside">
<a href="_sources/hub.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">
<div class="section" id="torch-hub">
<h1>torch.hub<a class="headerlink" href="#torch-hub" title="Permalink to this headline">¶</a></h1>
<p>Pytorch Hub is a pre-trained model repository designed to facilitate research reproducibility.</p>
<div class="section" id="publishing-models">
<h2>Publishing models<a class="headerlink" href="#publishing-models" title="Permalink to this headline">¶</a></h2>
<p>Pytorch Hub supports publishing pre-trained models(model definitions and pre-trained weights)
to a github repository by adding a simple <code class="docutils literal notranslate"><span class="pre">hubconf.py</span></code> file;</p>
<p><code class="docutils literal notranslate"><span class="pre">hubconf.py</span></code> can have multiple entrypoints. Each entrypoint is defined as a python function
(example: a pre-trained model you want to publish).</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">entrypoint_name</span><span class="p">(</span><span class="o">*</span><span class="n">args</span><span class="p">,</span> <span class="o">**</span><span class="n">kwargs</span><span class="p">):</span>
<span class="c1"># args & kwargs are optional, for models which take positional/keyword arguments.</span>
<span class="o">...</span>
</pre></div>
</div>
<div class="section" id="how-to-implement-an-entrypoint">
<h3>How to implement an entrypoint?<a class="headerlink" href="#how-to-implement-an-entrypoint" title="Permalink to this headline">¶</a></h3>
<p>Here is a code snippet specifies an entrypoint for <code class="docutils literal notranslate"><span class="pre">resnet18</span></code> model if we expand
the implementation in <code class="docutils literal notranslate"><span class="pre">pytorch/vision/hubconf.py</span></code>.
In most case importing the right function in <code class="docutils literal notranslate"><span class="pre">hubconf.py</span></code> is sufficient. Here we
just want to use the expanded version as an example to show how it works.
You can see the full script in
<a class="reference external" href="https://github.com/pytorch/vision/blob/master/hubconf.py">pytorch/vision repo</a></p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">dependencies</span> <span class="o">=</span> <span class="p">[</span><span class="s1">'torch'</span><span class="p">]</span>
<span class="kn">from</span> <span class="nn">torchvision.models.resnet</span> <span class="kn">import</span> <span class="n">resnet18</span> <span class="k">as</span> <span class="n">_resnet18</span>
<span class="c1"># resnet18 is the name of entrypoint</span>
<span class="k">def</span> <span class="nf">resnet18</span><span class="p">(</span><span class="n">pretrained</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span> <span class="o">**</span><span class="n">kwargs</span><span class="p">):</span>
<span class="sd">""" # This docstring shows up in hub.help()</span>
<span class="sd"> Resnet18 model</span>
<span class="sd"> pretrained (bool): kwargs, load pretrained weights into the model</span>
<span class="sd"> """</span>
<span class="c1"># Call the model, load pretrained weights</span>
<span class="n">model</span> <span class="o">=</span> <span class="n">_resnet18</span><span class="p">(</span><span class="n">pretrained</span><span class="o">=</span><span class="n">pretrained</span><span class="p">,</span> <span class="o">**</span><span class="n">kwargs</span><span class="p">)</span>
<span class="k">return</span> <span class="n">model</span>
</pre></div>
</div>
<ul class="simple">
<li><p><code class="docutils literal notranslate"><span class="pre">dependencies</span></code> variable is a <strong>list</strong> of package names required to <strong>load</strong> the model. Note this might
be slightly different from dependencies required for training a model.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">args</span></code> and <code class="docutils literal notranslate"><span class="pre">kwargs</span></code> are passed along to the real callable function.</p></li>
<li><p>Docstring of the function works as a help message. It explains what does the model do and what
are the allowed positional/keyword arguments. It’s highly recommended to add a few examples here.</p></li>
<li><p>Entrypoint function can either return a model(nn.module), or auxiliary tools to make the user workflow smoother, e.g. tokenizers.</p></li>
<li><p>Callables prefixed with underscore are considered as helper functions which won’t show up in <code class="docutils literal notranslate"><span class="pre">torch.hub.list()</span></code>.</p></li>
<li><p>Pretrained weights can either be stored locally in the github repo, or loadable by
<code class="docutils literal notranslate"><span class="pre">torch.hub.load_state_dict_from_url()</span></code>. If less than 2GB, it’s recommended to attach it to a <a class="reference external" href="https://help.github.com/en/articles/distributing-large-binaries">project release</a>
and use the url from the release.
In the example above <code class="docutils literal notranslate"><span class="pre">torchvision.models.resnet.resnet18</span></code> handles <code class="docutils literal notranslate"><span class="pre">pretrained</span></code>, alternatively you can put the following logic in the entrypoint definition.</p></li>
</ul>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">if</span> <span class="n">pretrained</span><span class="p">:</span>
<span class="c1"># For checkpoint saved in local github repo, e.g. <RELATIVE_PATH_TO_CHECKPOINT>=weights/save.pth</span>
<span class="n">dirname</span> <span class="o">=</span> <span class="n">os</span><span class="o">.</span><span class="n">path</span><span class="o">.</span><span class="n">dirname</span><span class="p">(</span><span class="vm">__file__</span><span class="p">)</span>
<span class="n">checkpoint</span> <span class="o">=</span> <span class="n">os</span><span class="o">.</span><span class="n">path</span><span class="o">.</span><span class="n">join</span><span class="p">(</span><span class="n">dirname</span><span class="p">,</span> <span class="o"><</span><span class="n">RELATIVE_PATH_TO_CHECKPOINT</span><span class="o">></span><span class="p">)</span>
<span class="n">state_dict</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">load</span><span class="p">(</span><span class="n">checkpoint</span><span class="p">)</span>
<span class="n">model</span><span class="o">.</span><span class="n">load_state_dict</span><span class="p">(</span><span class="n">state_dict</span><span class="p">)</span>
<span class="c1"># For checkpoint saved elsewhere</span>
<span class="n">checkpoint</span> <span class="o">=</span> <span class="s1">'https://download.pytorch.org/models/resnet18-5c106cde.pth'</span>
<span class="n">model</span><span class="o">.</span><span class="n">load_state_dict</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">hub</span><span class="o">.</span><span class="n">load_state_dict_from_url</span><span class="p">(</span><span class="n">checkpoint</span><span class="p">,</span> <span class="n">progress</span><span class="o">=</span><span class="kc">False</span><span class="p">))</span>
</pre></div>
</div>
</div>
<div class="section" id="important-notice">
<h3>Important Notice<a class="headerlink" href="#important-notice" title="Permalink to this headline">¶</a></h3>
<ul class="simple">
<li><p>The published models should be at least in a branch/tag. It can’t be a random commit.</p></li>
</ul>
</div>
</div>
<div class="section" id="loading-models-from-hub">
<h2>Loading models from Hub<a class="headerlink" href="#loading-models-from-hub" title="Permalink to this headline">¶</a></h2>
<p>Pytorch Hub provides convenient APIs to explore all available models in hub through <code class="docutils literal notranslate"><span class="pre">torch.hub.list()</span></code>,
show docstring and examples through <code class="docutils literal notranslate"><span class="pre">torch.hub.help()</span></code> and load the pre-trained models using <code class="docutils literal notranslate"><span class="pre">torch.hub.load()</span></code></p>
<span class="target" id="module-torch.hub"></span><dl class="function">
<dt id="torch.hub.list">
<code class="sig-prename descclassname">torch.hub.</code><code class="sig-name descname">list</code><span class="sig-paren">(</span><em class="sig-param">github</em>, <em class="sig-param">force_reload=False</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/hub.html#list"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.hub.list" title="Permalink to this definition">¶</a></dt>
<dd><p>List all entrypoints available in <cite>github</cite> hubconf.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><ul class="simple">
<li><p><strong>github</strong> (<em>string</em>) – a string with format “repo_owner/repo_name[:tag_name]” with an optional
tag/branch. The default branch is <cite>master</cite> if not specified.
Example: ‘pytorch/vision[:hub]’</p></li>
<li><p><strong>force_reload</strong> (<em>bool</em><em>, </em><em>optional</em>) – whether to discard the existing cache and force a fresh download.
Default is <cite>False</cite>.</p></li>
</ul>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><p>a list of available entrypoint names</p>
</dd>
<dt class="field-odd">Return type</dt>
<dd class="field-odd"><p>entrypoints</p>
</dd>
</dl>
<p class="rubric">Example</p>
<div class="doctest highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">entrypoints</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">hub</span><span class="o">.</span><span class="n">list</span><span class="p">(</span><span class="s1">'pytorch/vision'</span><span class="p">,</span> <span class="n">force_reload</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.hub.help">
<code class="sig-prename descclassname">torch.hub.</code><code class="sig-name descname">help</code><span class="sig-paren">(</span><em class="sig-param">github</em>, <em class="sig-param">model</em>, <em class="sig-param">force_reload=False</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/hub.html#help"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.hub.help" title="Permalink to this definition">¶</a></dt>
<dd><p>Show the docstring of entrypoint <cite>model</cite>.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><ul class="simple">
<li><p><strong>github</strong> (<em>string</em>) – a string with format <repo_owner/repo_name[:tag_name]> with an optional
tag/branch. The default branch is <cite>master</cite> if not specified.
Example: ‘pytorch/vision[:hub]’</p></li>
<li><p><strong>model</strong> (<em>string</em>) – a string of entrypoint name defined in repo’s hubconf.py</p></li>
<li><p><strong>force_reload</strong> (<em>bool</em><em>, </em><em>optional</em>) – whether to discard the existing cache and force a fresh download.
Default is <cite>False</cite>.</p></li>
</ul>
</dd>
</dl>
<p class="rubric">Example</p>
<div class="doctest highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="nb">print</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">hub</span><span class="o">.</span><span class="n">help</span><span class="p">(</span><span class="s1">'pytorch/vision'</span><span class="p">,</span> <span class="s1">'resnet18'</span><span class="p">,</span> <span class="n">force_reload</span><span class="o">=</span><span class="kc">True</span><span class="p">))</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.hub.load">
<code class="sig-prename descclassname">torch.hub.</code><code class="sig-name descname">load</code><span class="sig-paren">(</span><em class="sig-param">github</em>, <em class="sig-param">model</em>, <em class="sig-param">*args</em>, <em class="sig-param">**kwargs</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/hub.html#load"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.hub.load" title="Permalink to this definition">¶</a></dt>
<dd><p>Load a model from a github repo, with pretrained weights.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><ul class="simple">
<li><p><strong>github</strong> (<em>string</em>) – a string with format “repo_owner/repo_name[:tag_name]” with an optional
tag/branch. The default branch is <cite>master</cite> if not specified.
Example: ‘pytorch/vision[:hub]’</p></li>
<li><p><strong>model</strong> (<em>string</em>) – a string of entrypoint name defined in repo’s hubconf.py</p></li>
<li><p><strong>*args</strong> (<em>optional</em>) – the corresponding args for callable <cite>model</cite>.</p></li>
<li><p><strong>force_reload</strong> (<em>bool</em><em>, </em><em>optional</em>) – whether to force a fresh download of github repo unconditionally.
Default is <cite>False</cite>.</p></li>
<li><p><strong>verbose</strong> (<em>bool</em><em>, </em><em>optional</em>) – If False, mute messages about hitting local caches. Note that the message
about first download is cannot be muted.
Default is <cite>True</cite>.</p></li>
<li><p><strong>**kwargs</strong> (<em>optional</em>) – the corresponding kwargs for callable <cite>model</cite>.</p></li>
</ul>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><p>a single model with corresponding pretrained weights.</p>
</dd>
</dl>
<p class="rubric">Example</p>
<div class="doctest highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">model</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">hub</span><span class="o">.</span><span class="n">load</span><span class="p">(</span><span class="s1">'pytorch/vision'</span><span class="p">,</span> <span class="s1">'resnet50'</span><span class="p">,</span> <span class="n">pretrained</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.hub.download_url_to_file">
<code class="sig-prename descclassname">torch.hub.</code><code class="sig-name descname">download_url_to_file</code><span class="sig-paren">(</span><em class="sig-param">url</em>, <em class="sig-param">dst</em>, <em class="sig-param">hash_prefix=None</em>, <em class="sig-param">progress=True</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/hub.html#download_url_to_file"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.hub.download_url_to_file" title="Permalink to this definition">¶</a></dt>
<dd><p>Download object at the given URL to a local path.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><ul class="simple">
<li><p><strong>url</strong> (<em>string</em>) – URL of the object to download</p></li>
<li><p><strong>dst</strong> (<em>string</em>) – Full path where object will be saved, e.g. <cite>/tmp/temporary_file</cite></p></li>
<li><p><strong>hash_prefix</strong> (<em>string</em><em>, </em><em>optional</em>) – If not None, the SHA256 downloaded file should start with <cite>hash_prefix</cite>.
Default: None</p></li>
<li><p><strong>progress</strong> (<em>bool</em><em>, </em><em>optional</em>) – whether or not to display a progress bar to stderr
Default: True</p></li>
</ul>
</dd>
</dl>
<p class="rubric">Example</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">hub</span><span class="o">.</span><span class="n">download_url_to_file</span><span class="p">(</span><span class="s1">'https://s3.amazonaws.com/pytorch/models/resnet18-5c106cde.pth'</span><span class="p">,</span> <span class="s1">'/tmp/temporary_file'</span><span class="p">)</span>
</pre></div>
</div>
</dd></dl>
<dl class="function">
<dt id="torch.hub.load_state_dict_from_url">
<code class="sig-prename descclassname">torch.hub.</code><code class="sig-name descname">load_state_dict_from_url</code><span class="sig-paren">(</span><em class="sig-param">url</em>, <em class="sig-param">model_dir=None</em>, <em class="sig-param">map_location=None</em>, <em class="sig-param">progress=True</em>, <em class="sig-param">check_hash=False</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/hub.html#load_state_dict_from_url"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.hub.load_state_dict_from_url" title="Permalink to this definition">¶</a></dt>
<dd><p>Loads the Torch serialized object at the given URL.</p>
<p>If downloaded file is a zip file, it will be automatically
decompressed.</p>
<p>If the object is already present in <cite>model_dir</cite>, it’s deserialized and
returned.
The default value of <cite>model_dir</cite> is <code class="docutils literal notranslate"><span class="pre">$TORCH_HOME/checkpoints</span></code> where
environment variable <code class="docutils literal notranslate"><span class="pre">$TORCH_HOME</span></code> defaults to <code class="docutils literal notranslate"><span class="pre">$XDG_CACHE_HOME/torch</span></code>.
<code class="docutils literal notranslate"><span class="pre">$XDG_CACHE_HOME</span></code> follows the X Design Group specification of the Linux
filesytem layout, with a default value <code class="docutils literal notranslate"><span class="pre">~/.cache</span></code> if not set.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><ul class="simple">
<li><p><strong>url</strong> (<em>string</em>) – URL of the object to download</p></li>
<li><p><strong>model_dir</strong> (<em>string</em><em>, </em><em>optional</em>) – directory in which to save the object</p></li>
<li><p><strong>map_location</strong> (<em>optional</em>) – a function or a dict specifying how to remap storage locations (see torch.load)</p></li>
<li><p><strong>progress</strong> (<em>bool</em><em>, </em><em>optional</em>) – whether or not to display a progress bar to stderr.
Default: True</p></li>
<li><p><strong>check_hash</strong> (<em>bool</em><em>, </em><em>optional</em>) – If True, the filename part of the URL should follow the naming convention
<code class="docutils literal notranslate"><span class="pre">filename-<sha256>.ext</span></code> where <code class="docutils literal notranslate"><span class="pre"><sha256></span></code> is the first eight or more
digits of the SHA256 hash of the contents of the file. The hash is used to
ensure unique names and to verify the contents of the file.
Default: False</p></li>
</ul>
</dd>
</dl>
<p class="rubric">Example</p>
<div class="doctest highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="n">state_dict</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">hub</span><span class="o">.</span><span class="n">load_state_dict_from_url</span><span class="p">(</span><span class="s1">'https://s3.amazonaws.com/pytorch/models/resnet18-5c106cde.pth'</span><span class="p">)</span>
</pre></div>
</div>
</dd></dl>
<div class="section" id="running-a-loaded-model">
<h3>Running a loaded model:<a class="headerlink" href="#running-a-loaded-model" title="Permalink to this headline">¶</a></h3>
<p>Note that <code class="docutils literal notranslate"><span class="pre">*args,</span> <span class="pre">**kwargs</span></code> in <code class="docutils literal notranslate"><span class="pre">torch.load()</span></code> are used to <strong>instantiate</strong> a model.
After you loaded a model, how can you find out what you can do with the model?
A suggested workflow is</p>
<ul class="simple">
<li><p><code class="docutils literal notranslate"><span class="pre">dir(model)</span></code> to see all avaialble methods of the model.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">help(model.foo)</span></code> to check what arguments <code class="docutils literal notranslate"><span class="pre">model.foo</span></code> takes to run</p></li>
</ul>
<p>To help users explore without refering to documentation back and forth, we strongly
recommend repo owners make function help messages clear and succinct. It’s also helpful
to include a minimal working example.</p>
</div>
<div class="section" id="where-are-my-downloaded-models-saved">
<h3>Where are my downloaded models saved?<a class="headerlink" href="#where-are-my-downloaded-models-saved" title="Permalink to this headline">¶</a></h3>
<p>The locations are used in the order of</p>
<ul class="simple">
<li><p>Calling <code class="docutils literal notranslate"><span class="pre">hub.set_dir(<PATH_TO_HUB_DIR>)</span></code></p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">$TORCH_HOME/hub</span></code>, if environment variable <code class="docutils literal notranslate"><span class="pre">TORCH_HOME</span></code> is set.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">$XDG_CACHE_HOME/torch/hub</span></code>, if environment variable <code class="docutils literal notranslate"><span class="pre">XDG_CACHE_HOME</span></code> is set.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">~/.cache/torch/hub</span></code></p></li>
</ul>
<dl class="function">
<dt id="torch.hub.set_dir">
<code class="sig-prename descclassname">torch.hub.</code><code class="sig-name descname">set_dir</code><span class="sig-paren">(</span><em class="sig-param">d</em><span class="sig-paren">)</span><a class="reference internal" href="_modules/torch/hub.html#set_dir"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#torch.hub.set_dir" title="Permalink to this definition">¶</a></dt>
<dd><p>Optionally set hub_dir to a local dir to save downloaded models & weights.</p>
<p>If <code class="docutils literal notranslate"><span class="pre">set_dir</span></code> is not called, default path is <code class="docutils literal notranslate"><span class="pre">$TORCH_HOME/hub</span></code> where
environment variable <code class="docutils literal notranslate"><span class="pre">$TORCH_HOME</span></code> defaults to <code class="docutils literal notranslate"><span class="pre">$XDG_CACHE_HOME/torch</span></code>.
<code class="docutils literal notranslate"><span class="pre">$XDG_CACHE_HOME</span></code> follows the X Design Group specification of the Linux
filesytem layout, with a default value <code class="docutils literal notranslate"><span class="pre">~/.cache</span></code> if the environment
variable is not set.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><p><strong>d</strong> (<em>string</em>) – path to a local folder to save downloaded models & weights.</p>
</dd>
</dl>
</dd></dl>
</div>
<div class="section" id="caching-logic">
<h3>Caching logic<a class="headerlink" href="#caching-logic" title="Permalink to this headline">¶</a></h3>
<p>By default, we don’t clean up files after loading it. Hub uses the cache by default if it already exists in <code class="docutils literal notranslate"><span class="pre">hub_dir</span></code>.</p>
<p>Users can force a reload by calling <code class="docutils literal notranslate"><span class="pre">hub.load(...,</span> <span class="pre">force_reload=True)</span></code>. This will delete
the existing github folder and downloaded weights, reinitialize a fresh download. This is useful
when updates are published to the same branch, users can keep up with the latest release.</p>
</div>
<div class="section" id="known-limitations">
<h3>Known limitations:<a class="headerlink" href="#known-limitations" title="Permalink to this headline">¶</a></h3>
<p>Torch hub works by importing the package as if it was installed. There’re some side effects
introduced by importing in Python. For example, you can see new items in Python caches
<code class="docutils literal notranslate"><span class="pre">sys.modules</span></code> and <code class="docutils literal notranslate"><span class="pre">sys.path_importer_cache</span></code> which is normal Python behavior.</p>
<p>A known limitation that worth mentioning here is user <strong>CANNOT</strong> load two different branches of
the same repo in the <strong>same python process</strong>. It’s just like installing two packages with the
same name in Python, which is not good. Cache might join the party and give you surprises if you
actually try that. Of course it’s totally fine to load them in separate processes.</p>
</div>
</div>
</div>
</article>
</div>
<footer>
<div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
<a href="jit.html" class="btn btn-neutral float-right" title="TorchScript" accesskey="n" rel="next">Next <img src="_static/images/chevron-right-orange.svg" class="next-page"></a>
<a href="distributions.html" class="btn btn-neutral" title="Probability distributions - torch.distributions" accesskey="p" rel="prev"><img src="_static/images/chevron-right-orange.svg" class="previous-page"> Previous</a>
</div>
<hr>
<div role="contentinfo">
<p>
© Copyright 2019, Torch 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>
</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="#">torch.hub</a><ul>
<li><a class="reference internal" href="#publishing-models">Publishing models</a><ul>
<li><a class="reference internal" href="#how-to-implement-an-entrypoint">How to implement an entrypoint?</a></li>
<li><a class="reference internal" href="#important-notice">Important Notice</a></li>
</ul>
</li>
<li><a class="reference internal" href="#loading-models-from-hub">Loading models from Hub</a><ul>
<li><a class="reference internal" href="#running-a-loaded-model">Running a loaded model:</a></li>
<li><a class="reference internal" href="#where-are-my-downloaded-models-saved">Where are my downloaded models saved?</a></li>
<li><a class="reference internal" href="#caching-logic">Caching logic</a></li>
<li><a class="reference internal" href="#known-limitations">Known limitations:</a></li>
</ul>
</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 src="_static/jquery.js"></script>
<script src="_static/underscore.js"></script>
<script src="_static/doctools.js"></script>
<script src="_static/language_data.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 type="text/javascript" src="_static/js/theme.js"></script>
<script type="text/javascript">
jQuery(function () {
SphinxRtdTheme.Navigation.enable(true);
});
</script>
<script>
(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
(i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),
m=s.getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,m)
})(window,document,'script','https://www.google-analytics.com/analytics.js','ga');
ga('create', 'UA-90545585-1', 'auto');
ga('send', 'pageview');
</script>
<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>
<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://pytorch.org/resources">Resources</a></li>
</ul>
</div>
<div class="footer-links-col">
<ul>
<li class="list-title"><a href="https://pytorch.org/support">Support</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.slack.com" target="_blank">Slack</a></li>
<li><a href="https://github.com/pytorch/pytorch/blob/master/CONTRIBUTING.md" target="_blank">Contributing</a></li>
</ul>
</div>
<div class="footer-links-col follow-us-col">
<ul>
<li class="list-title">Follow Us</li>
<li>
<div id="mc_embed_signup">
<form
action="https://twitter.us14.list-manage.com/subscribe/post?u=75419c71fe0a935e53dfa4a3f&id=91d0dccd39"
method="post"
id="mc-embedded-subscribe-form"
name="mc-embedded-subscribe-form"
class="email-subscribe-form validate"
target="_blank"
novalidate>
<div id="mc_embed_signup_scroll" class="email-subscribe-form-fields-wrapper">
<div class="mc-field-group">
<label for="mce-EMAIL" style="display:none;">Email Address</label>
<input type="email" value="" name="EMAIL" class="required email" id="mce-EMAIL" placeholder="Email Address">
</div>
<div id="mce-responses" class="clear">
<div class="response" id="mce-error-response" style="display:none"></div>
<div class="response" id="mce-success-response" style="display:none"></div>
</div> <!-- real people should not fill this in and expect good things - do not remove this or risk form bot signups-->
<div style="position: absolute; left: -5000px;" aria-hidden="true"><input type="text" name="b_75419c71fe0a935e53dfa4a3f_91d0dccd39" tabindex="-1" value=""></div>
<div class="clear">
<input type="submit" value="" name="subscribe" id="mc-embedded-subscribe" class="button email-subscribe-button">
</div>
</div>
</form>
</div>
</li>
</ul>
<div class="footer-social-icons">
<a href="https://www.facebook.com/pytorch" target="_blank" class="facebook"></a>
<a href="https://twitter.com/pytorch" target="_blank" class="twitter"></a>
</div>
</div>
</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>
<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/mobile">Mobile</a>
</li>
<li>
<a href="https://pytorch.org/hub">PyTorch Hub</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">
<a href="https://pytorch.org/docs/stable/index.html">Docs</a>
</li>
<li>
<a href="https://pytorch.org/resources">Resources</a>
</li>
<li>
<a href="https://github.com/pytorch/pytorch">Github</a>
</li>
</ul>
</div>
</div>
</div>
<!-- End Mobile Menu -->
<script type="text/javascript" src="_static/js/vendor/anchor.min.js"></script>
<script type="text/javascript">
$(document).ready(function() {
mobileMenu.bind();
mobileTOC.bind();
pytorchAnchors.bind();
sideMenus.bind();
scrollToAnchor.bind();
highlightNavigation.bind();
mainMenuDropdown.bind();
// Add class to links that have code blocks, since we cannot create links in code blocks
$("article.pytorch-article a span.pre").each(function(e) {
$(this).closest("a").addClass("has-code");
});
})
</script>
</body>
</html>