Register
Login
Resources
Docs Blog Datasets Glossary Case Studies Tutorials & Webinars
Product
Data Engine LLMs Platform Enterprise
Pricing Explore
Connect to our Discord channel
Resources  
Docs Blog Datasets Glossary Case Studies Tutorials & Webinars
Explore Pricing
Product  
Data Engine LLMs Platform Enterprise
Connect to Dagshub Discord channel
Register
Login

Deci-AI
/
super-gradients
connected to https://github.com/Deci-AI/super-gradients.git

Watch 4
Issues 62 Pull Requests 44 Discussions 0

super_gradients.training.sg_model.html 58 KB
History Raw

You have to be logged in to leave a comment. Sign In
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 
  1. <!DOCTYPE html>
    2. 

  2. <html class="writer-html5" lang="en" >
    3. 

  3. <head>
    4. 

  4.   <meta charset="utf-8" /><meta name="generator" content="Docutils 0.17.1: http://docutils.sourceforge.net/" />
    5. 

  5. 

  6.   <meta name="viewport" content="width=device-width, initial-scale=1.0" />
    7. 

  7.   <title>super_gradients.training.sg_model package &mdash; SuperGradients 1.0 documentation</title>
    8. 

  8.       <link rel="stylesheet" href="_static/pygments.css" type="text/css" />
    9. 

  9.       <link rel="stylesheet" href="_static/css/theme.css" type="text/css" />
    10. 

  10.   <!--[if lt IE 9]>
    11. 

  11.     <script src="_static/js/html5shiv.min.js"></script>
    12. 

  12.   <![endif]-->
    13. 

  13.   
    14. 

  14.         <script data-url_root="./" id="documentation_options" src="_static/documentation_options.js"></script>
    15. 

  15.         <script src="_static/jquery.js"></script>
    16. 

  16.         <script src="_static/underscore.js"></script>
    17. 

  17.         <script src="_static/doctools.js"></script>
    18. 

  18.     <script src="_static/js/theme.js"></script>
    19. 

  19.     <link rel="index" title="Index" href="genindex.html" />
    20. 

  20.     <link rel="search" title="Search" href="search.html" />
    21. 

  21.     <link rel="next" title="super_gradients.training.utils package" href="super_gradients.training.utils.html" />
    22. 

  22.     <link rel="prev" title="super_gradients.training.models package" href="super_gradients.training.models.html" /> 
    23. 

  23. </head>
    24. 

  24. 

  25. <body class="wy-body-for-nav"> 
    26. 

  26.   <div class="wy-grid-for-nav">
    27. 

  27.     <nav data-toggle="wy-nav-shift" class="wy-nav-side">
    28. 

  28.       <div class="wy-side-scroll">
    29. 

  29.         <div class="wy-side-nav-search" >
    30. 

  30.             <a href="index.html" class="icon icon-home"> SuperGradients
    31. 

  31.           </a>
    32. 

  32. <div role="search">
    33. 

  33.   <form id="rtd-search-form" class="wy-form" action="search.html" method="get">
    34. 

  34.     <input type="text" name="q" placeholder="Search docs" />
    35. 

  35.     <input type="hidden" name="check_keywords" value="yes" />
    36. 

  36.     <input type="hidden" name="area" value="default" />
    37. 

  37.   </form>
    38. 

  38. </div>
    39. 

  39.         </div><div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="Navigation menu">
    40. 

  40.               <p class="caption"><span class="caption-text">Contents:</span></p>
    41. 

  41. <ul class="current">
    42. 

  42. <li class="toctree-l1"><a class="reference internal" href="intro.html">Introduction</a></li>
    43. 

  43. <li class="toctree-l1"><a class="reference internal" href="intro.html#installation">Installation</a></li>
    44. 

  44. <li class="toctree-l1 current"><a class="reference internal" href="super_gradients.training.html">super_gradients.training package</a><ul class="current">
    45. 

  45. <li class="toctree-l2 current"><a class="reference internal" href="super_gradients.training.html#subpackages">Subpackages</a><ul class="current">
    46. 

  46. <li class="toctree-l3"><a class="reference internal" href="super_gradients.training.datasets.html">super_gradients.training.datasets package</a></li>
    47. 

  47. <li class="toctree-l3"><a class="reference internal" href="super_gradients.training.exceptions.html">super_gradients.training.exceptions package</a></li>
    48. 

  48. <li class="toctree-l3"><a class="reference internal" href="super_gradients.training.legacy.html">super_gradients.training.legacy package</a></li>
    49. 

  49. <li class="toctree-l3"><a class="reference internal" href="super_gradients.training.losses.html">super_gradients.training.losses package</a></li>
    50. 

  50. <li class="toctree-l3"><a class="reference internal" href="super_gradients.training.metrics.html">super_gradients.training.metrics package</a></li>
    51. 

  51. <li class="toctree-l3"><a class="reference internal" href="super_gradients.training.models.html">super_gradients.training.models package</a></li>
    52. 

  52. <li class="toctree-l3 current"><a class="current reference internal" href="#">super_gradients.training.sg_model package</a><ul>
    53. 

  53. <li class="toctree-l4"><a class="reference internal" href="#submodules">Submodules</a></li>
    54. 

  54. <li class="toctree-l4"><a class="reference internal" href="#module-super_gradients.training.sg_model.sg_model">super_gradients.training.sg_model.sg_model module</a></li>
    55. 

  55. <li class="toctree-l4"><a class="reference internal" href="#module-super_gradients.training.sg_model">Module contents</a></li>
    56. 

  56. </ul>
    57. 

  57. </li>
    58. 

  58. <li class="toctree-l3"><a class="reference internal" href="super_gradients.training.utils.html">super_gradients.training.utils package</a></li>
    59. 

  59. </ul>
    60. 

  60. </li>
    61. 

  61. <li class="toctree-l2"><a class="reference internal" href="super_gradients.training.html#submodules">Submodules</a></li>
    62. 

  62. <li class="toctree-l2"><a class="reference internal" href="super_gradients.training.html#module-super_gradients.training.params">super_gradients.training.params module</a></li>
    63. 

  63. <li class="toctree-l2"><a class="reference internal" href="super_gradients.training.html#module-super_gradients.training.pretrained_models">super_gradients.training.pretrained_models module</a></li>
    64. 

  64. <li class="toctree-l2"><a class="reference internal" href="super_gradients.training.html#module-super_gradients.training">Module contents</a></li>
    65. 

  65. </ul>
    66. 

  66. </li>
    67. 

  67. </ul>
    68. 

  68. 

  69.         </div>
    70. 

  70.       </div>
    71. 

  71.     </nav>
    72. 

  72. 

  73.     <section data-toggle="wy-nav-shift" class="wy-nav-content-wrap"><nav class="wy-nav-top" aria-label="Mobile navigation menu" >
    74. 

  74.           <i data-toggle="wy-nav-top" class="fa fa-bars"></i>
    75. 

  75.           <a href="index.html">SuperGradients</a>
    76. 

  76.       </nav>
    77. 

  77. 

  78.       <div class="wy-nav-content">
    79. 

  79.         <div class="rst-content">
    80. 

  80.           <div role="navigation" aria-label="Page navigation">
    81. 

  81.   <ul class="wy-breadcrumbs">
    82. 

  82.       <li><a href="index.html" class="icon icon-home"></a> &raquo;</li>
    83. 

  83.           <li><a href="super_gradients.training.html">super_gradients.training package</a> &raquo;</li>
    84. 

  84.       <li>super_gradients.training.sg_model package</li>
    85. 

  85.       <li class="wy-breadcrumbs-aside">
    86. 

  86.             <a href="_sources/super_gradients.training.sg_model.rst.txt" rel="nofollow"> View page source</a>
    87. 

  87.       </li>
    88. 

  88.   </ul>
    89. 

  89.   <hr/>
    90. 

  90. </div>
    91. 

  91.           <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
    92. 

  92.            <div itemprop="articleBody">
    93. 

  93.              
    94. 

  94.   <section id="super-gradients-training-sg-model-package">
    95. 

  95. <h1>super_gradients.training.sg_model package<a class="headerlink" href="#super-gradients-training-sg-model-package" title="Permalink to this headline"></a></h1>
    96. 

  96. <section id="submodules">
    97. 

  97. <h2>Submodules<a class="headerlink" href="#submodules" title="Permalink to this headline"></a></h2>
    98. 

  98. </section>
    99. 

  99. <section id="module-super_gradients.training.sg_model.sg_model">
    100. 

  100. <span id="super-gradients-training-sg-model-sg-model-module"></span><h2>super_gradients.training.sg_model.sg_model module<a class="headerlink" href="#module-super_gradients.training.sg_model.sg_model" title="Permalink to this headline"></a></h2>
    101. 

  101. <dl class="py class">
    102. 

  102. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.EvaluationType">
    103. 

  103. <em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">super_gradients.training.sg_model.sg_model.</span></span><span class="sig-name descname"><span class="pre">EvaluationType</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">value</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#EvaluationType"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.EvaluationType" title="Permalink to this definition"></a></dt>
    104. 

  104. <dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">str</span></code>, <code class="xref py py-class docutils literal notranslate"><span class="pre">enum.Enum</span></code></p>
    105. 

  105. <p>Passed to SgModel.evaluate(..), and controls which phase callbacks should be triggered (if at all).</p>
    106. 

  106. <blockquote>
    107. 

  107. <div><dl class="simple">
    108. 

  108. <dt>Attributes:</dt><dd><p>TEST
    109. 

  109. VALIDATION</p>
    110. 

  110. </dd>
    111. 

  111. </dl>
    112. 

  112. </div></blockquote>
    113. 

  113. <dl class="py attribute">
    114. 

  114. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.EvaluationType.TEST">
    115. 

  115. <span class="sig-name descname"><span class="pre">TEST</span></span><em class="property"> <span class="pre">=</span> <span class="pre">'TEST'</span></em><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.EvaluationType.TEST" title="Permalink to this definition"></a></dt>
    116. 

  116. <dd></dd></dl>
    117. 

  117. 

  118. <dl class="py attribute">
    119. 

  119. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.EvaluationType.VALIDATION">
    120. 

  120. <span class="sig-name descname"><span class="pre">VALIDATION</span></span><em class="property"> <span class="pre">=</span> <span class="pre">'VALIDATION'</span></em><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.EvaluationType.VALIDATION" title="Permalink to this definition"></a></dt>
    121. 

  121. <dd></dd></dl>
    122. 

  122. 

  123. </dd></dl>
    124. 

  124. 

  125. <dl class="py class">
    126. 

  126. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.MultiGPUMode">
    127. 

  127. <em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">super_gradients.training.sg_model.sg_model.</span></span><span class="sig-name descname"><span class="pre">MultiGPUMode</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">value</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#MultiGPUMode"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.MultiGPUMode" title="Permalink to this definition"></a></dt>
    128. 

  128. <dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">str</span></code>, <code class="xref py py-class docutils literal notranslate"><span class="pre">enum.Enum</span></code></p>
    129. 

  129. <dl class="simple">
    130. 

  130. <dt>Attributes:</dt><dd><p>OFF                       - Single GPU Mode / CPU Mode
    131. 

  131. DATA_PARALLEL             - Multiple GPUs, Synchronous
    132. 

  132. DISTRIBUTED_DATA_PARALLEL - Multiple GPUs, Asynchronous</p>
    133. 

  133. </dd>
    134. 

  134. </dl>
    135. 

  135. <dl class="py attribute">
    136. 

  136. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.MultiGPUMode.DATA_PARALLEL">
    137. 

  137. <span class="sig-name descname"><span class="pre">DATA_PARALLEL</span></span><em class="property"> <span class="pre">=</span> <span class="pre">'DP'</span></em><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.MultiGPUMode.DATA_PARALLEL" title="Permalink to this definition"></a></dt>
    138. 

  138. <dd></dd></dl>
    139. 

  139. 

  140. <dl class="py attribute">
    141. 

  141. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.MultiGPUMode.DISTRIBUTED_DATA_PARALLEL">
    142. 

  142. <span class="sig-name descname"><span class="pre">DISTRIBUTED_DATA_PARALLEL</span></span><em class="property"> <span class="pre">=</span> <span class="pre">'DDP'</span></em><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.MultiGPUMode.DISTRIBUTED_DATA_PARALLEL" title="Permalink to this definition"></a></dt>
    143. 

  143. <dd></dd></dl>
    144. 

  144. 

  145. <dl class="py attribute">
    146. 

  146. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.MultiGPUMode.OFF">
    147. 

  147. <span class="sig-name descname"><span class="pre">OFF</span></span><em class="property"> <span class="pre">=</span> <span class="pre">'Off'</span></em><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.MultiGPUMode.OFF" title="Permalink to this definition"></a></dt>
    148. 

  148. <dd></dd></dl>
    149. 

  149. 

  150. </dd></dl>
    151. 

  151. 

  152. <dl class="py class">
    153. 

  153. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel">
    154. 

  154. <em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">super_gradients.training.sg_model.sg_model.</span></span><span class="sig-name descname"><span class="pre">SgModel</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="pre">experiment_name:</span> <span class="pre">str</span></em>, <em class="sig-param"><span class="pre">device:</span> <span class="pre">Optional[str]</span> <span class="pre">=</span> <span class="pre">None</span></em>, <em class="sig-param"><span class="pre">multi_gpu:</span> <span class="pre">super_gradients.training.sg_model.sg_model.MultiGPUMode</span> <span class="pre">=</span> <span class="pre">&lt;MultiGPUMode.OFF:</span> <span class="pre">'Off'&gt;</span></em>, <em class="sig-param"><span class="pre">model_checkpoints_location:</span> <span class="pre">str</span> <span class="pre">=</span> <span class="pre">'local'</span></em>, <em class="sig-param"><span class="pre">overwrite_local_checkpoint:</span> <span class="pre">bool</span> <span class="pre">=</span> <span class="pre">True</span></em>, <em class="sig-param"><span class="pre">ckpt_name:</span> <span class="pre">str</span> <span class="pre">=</span> <span class="pre">'ckpt_latest.pth'</span></em>, <em class="sig-param"><span class="pre">post_prediction_callback:</span> <span class="pre">Optional[super_gradients.training.utils.detection_utils.DetectionPostPredictionCallback]</span> <span class="pre">=</span> <span class="pre">None</span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel" title="Permalink to this definition"></a></dt>
    155. 

  155. <dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">object</span></code></p>
    156. 

  156. <p>SuperGradient Model - Base Class for Sg Models</p>
    157. 

  157. <dl class="simple">
    158. 

  158. <dt>train(max_epochs<span class="classifier">int, initial_epoch</span><span class="classifier">int, save_model</span><span class="classifier">bool)</span></dt><dd><p>the main function used for the training, h.p. updating, logging etc.</p>
    159. 

  159. </dd>
    160. 

  160. <dt>predict(idx<span class="classifier">int)</span></dt><dd><p>returns the predictions and label of the current inputs</p>
    161. 

  161. </dd>
    162. 

  162. <dt>test(epoch<span class="classifier">int, idx</span><span class="classifier">int, save</span><span class="classifier">bool):</span></dt><dd><p>returns the test loss, accuracy and runtime</p>
    163. 

  163. </dd>
    164. 

  164. </dl>
    165. 

  165. <dl class="py method">
    166. 

  166. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.backward_step">
    167. 

  167. <span class="sig-name descname"><span class="pre">backward_step</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">loss</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">torch.Tensor</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">epoch</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">int</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">batch_idx</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">int</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">context</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><a class="reference internal" href="super_gradients.training.utils.html#super_gradients.training.utils.callbacks.PhaseContext" title="super_gradients.training.utils.callbacks.PhaseContext"><span class="pre">super_gradients.training.utils.callbacks.PhaseContext</span></a></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.backward_step"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.backward_step" title="Permalink to this definition"></a></dt>
    168. 

  168. <dd><p>Run backprop on the loss and perform a step
    169. 

  169. :param loss: The value computed by the loss function
    170. 

  170. :param optimizer: An object that can perform a gradient step and zeroize model gradient
    171. 

  171. :param epoch: number of epoch the training is on
    172. 

  172. :param batch_idx: number of iteration inside the current epoch
    173. 

  173. :param context: current phase context
    174. 

  174. :return:</p>
    175. 

  175. </dd></dl>
    176. 

  176. 

  177. <dl class="py method">
    178. 

  178. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.build_model">
    179. 

  179. <span class="sig-name descname"><span class="pre">build_model</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="pre">architecture:</span> <span class="pre">Union[str,</span> <span class="pre">torch.nn.modules.module.Module],</span> <span class="pre">arch_params={},</span> <span class="pre">load_checkpoint:</span> <span class="pre">bool</span> <span class="pre">=</span> <span class="pre">False,</span> <span class="pre">strict_load:</span> <span class="pre">super_gradients.training.sg_model.sg_model.StrictLoad</span> <span class="pre">=</span> <span class="pre">&lt;StrictLoad.ON:</span> <span class="pre">True&gt;,</span> <span class="pre">source_ckpt_folder_name:</span> <span class="pre">Optional[str]</span> <span class="pre">=</span> <span class="pre">None,</span> <span class="pre">load_weights_only:</span> <span class="pre">bool</span> <span class="pre">=</span> <span class="pre">False,</span> <span class="pre">load_backbone:</span> <span class="pre">bool</span> <span class="pre">=</span> <span class="pre">False,</span> <span class="pre">external_checkpoint_path:</span> <span class="pre">Optional[str]</span> <span class="pre">=</span> <span class="pre">None</span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.build_model"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.build_model" title="Permalink to this definition"></a></dt>
    180. 

  180. <dd><dl class="field-list simple">
    181. 

  181. <dt class="field-odd">Parameters</dt>
    182. 

  182. <dd class="field-odd"><ul class="simple">
    183. 

  183. <li><p><strong>architecture</strong> – Defines the network’s architecture from models/ALL_ARCHITECTURES</p></li>
    184. 

  184. <li><p><strong>arch_params</strong> – Architecture H.P. e.g.: block, num_blocks, num_classes, etc.</p></li>
    185. 

  185. <li><p><strong>load_checkpoint</strong> – Load a pre-trained checkpoint</p></li>
    186. 

  186. <li><p><strong>strict_load</strong> – See StrictLoad class documentation for details.</p></li>
    187. 

  187. <li><p><strong>source_ckpt_folder_name</strong> – folder name to load the checkpoint from (self.experiment_name if none is given)</p></li>
    188. 

  188. <li><p><strong>load_weights_only</strong> – loads only the weight from the checkpoint and zeroize the training params</p></li>
    189. 

  189. <li><p><strong>load_backbone</strong> – loads the provided checkpoint to self.net.backbone instead of self.net</p></li>
    190. 

  190. <li><p><strong>external_checkpoint_path</strong> – The path to the external checkpoint to be loaded. Can be absolute or relative
    191. 

  191. (ie: path/to/checkpoint.pth). If provided, will automatically attempt to
    192. 

  192. load the checkpoint even if the load_checkpoint flag is not provided.</p></li>
    193. 

  193. </ul>
    194. 

  194. </dd>
    195. 

  195. </dl>
    196. 

  196. </dd></dl>
    197. 

  197. 

  198. <dl class="py method">
    199. 

  199. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.compute_model_runtime">
    200. 

  200. <span class="sig-name descname"><span class="pre">compute_model_runtime</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">input_dims</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">tuple</span><span class="p"><span class="pre">]</span></span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">batch_sizes</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">Union</span><span class="p"><span class="pre">[</span></span><span class="pre">tuple</span><span class="p"><span class="pre">,</span> </span><span class="pre">list</span><span class="p"><span class="pre">,</span> </span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">(1,</span> <span class="pre">8,</span> <span class="pre">16,</span> <span class="pre">32,</span> <span class="pre">64)</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">verbose</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">bool</span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">True</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.compute_model_runtime"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.compute_model_runtime" title="Permalink to this definition"></a></dt>
    201. 

  201. <dd><p>Compute the “atomic” inference time and throughput.
    202. 

  202. Atomic refers to calculating the forward pass independently, discarding effects such as data augmentation,
    203. 

  203. data upload to device, multi-gpu distribution etc.
    204. 

  204. :param input_dims: tuple</p>
    205. 

  205. <blockquote>
    206. 

  206. <div><p>shape of a basic input to the network (without the first index) e.g. (3, 224, 224)
    207. 

  207. if None uses an input from the test loader</p>
    208. 

  208. </div></blockquote>
    209. 

  209. <dl class="field-list simple">
    210. 

  210. <dt class="field-odd">Parameters</dt>
    211. 

  211. <dd class="field-odd"><ul class="simple">
    212. 

  212. <li><p><strong>batch_sizes</strong> – int or list
    213. 

  213. Batch sizes for latency calculation</p></li>
    214. 

  214. <li><p><strong>verbose</strong> – bool
    215. 

  215. Prints results to screen</p></li>
    216. 

  216. </ul>
    217. 

  217. </dd>
    218. 

  218. <dt class="field-even">Returns</dt>
    219. 

  219. <dd class="field-even"><p>log: dict
    220. 

  220. Latency and throughput for each tested batch size</p>
    221. 

  221. </dd>
    222. 

  222. </dl>
    223. 

  223. </dd></dl>
    224. 

  224. 

  225. <dl class="py method">
    226. 

  226. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.connect_dataset_interface">
    227. 

  227. <span class="sig-name descname"><span class="pre">connect_dataset_interface</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">dataset_interface</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><a class="reference internal" href="super_gradients.training.datasets.dataset_interfaces.html#super_gradients.training.datasets.dataset_interfaces.dataset_interface.DatasetInterface" title="super_gradients.training.datasets.dataset_interfaces.dataset_interface.DatasetInterface"><span class="pre">super_gradients.training.datasets.dataset_interfaces.dataset_interface.DatasetInterface</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">data_loader_num_workers</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">int</span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">8</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.connect_dataset_interface"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.connect_dataset_interface" title="Permalink to this definition"></a></dt>
    228. 

  228. <dd><dl class="field-list simple">
    229. 

  229. <dt class="field-odd">Parameters</dt>
    230. 

  230. <dd class="field-odd"><ul class="simple">
    231. 

  231. <li><p><strong>dataset_interface</strong> – DatasetInterface object</p></li>
    232. 

  232. <li><p><strong>data_loader_num_workers</strong> – The number of threads to initialize the Data Loaders with
    233. 

  233. The dataset to be connected</p></li>
    234. 

  234. </ul>
    235. 

  235. </dd>
    236. 

  236. </dl>
    237. 

  237. </dd></dl>
    238. 

  238. 

  239. <dl class="py method">
    240. 

  240. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.evaluate">
    241. 

  241. <span class="sig-name descname"><span class="pre">evaluate</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">data_loader</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">torch.utils.data.dataloader.DataLoader</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">metrics</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">torchmetrics.collections.MetricCollection</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">evaluation_type</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><a class="reference internal" href="#super_gradients.training.sg_model.sg_model.EvaluationType" title="super_gradients.training.sg_model.sg_model.EvaluationType"><span class="pre">super_gradients.training.sg_model.sg_model.EvaluationType</span></a></span></em>, <em class="sig-param"><span class="n"><span class="pre">epoch</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">silent_mode</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">bool</span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">False</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">metrics_progress_verbose</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">bool</span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">False</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.evaluate"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.evaluate" title="Permalink to this definition"></a></dt>
    242. 

  242. <dd><p>Evaluates the model on given dataloader and metrics.</p>
    243. 

  243. <dl class="field-list simple">
    244. 

  244. <dt class="field-odd">Parameters</dt>
    245. 

  245. <dd class="field-odd"><ul class="simple">
    246. 

  246. <li><p><strong>data_loader</strong> – dataloader to perform evaluataion on</p></li>
    247. 

  247. <li><p><strong>metrics</strong> – (MetricCollection) metrics for evaluation</p></li>
    248. 

  248. <li><p><strong>evaluation_type</strong> – (EvaluationType) controls which phase callbacks will be used (for example, on batch end,
    249. 

  249. when evaluation_type=EvaluationType.VALIDATION the Phase.VALIDATION_BATCH_END callbacks will be triggered)</p></li>
    250. 

  250. <li><p><strong>epoch</strong> – (int) epoch idx</p></li>
    251. 

  251. <li><p><strong>silent_mode</strong> – (bool) controls verbosity</p></li>
    252. 

  252. <li><p><strong>metrics_progress_verbose</strong> – (bool) controls the verbosity of metrics progress (default=False).
    253. 

  253. Slows down the program significantly.</p></li>
    254. 

  254. </ul>
    255. 

  255. </dd>
    256. 

  256. <dt class="field-even">Returns</dt>
    257. 

  257. <dd class="field-even"><p>results tuple (tuple) containing the loss items and metric values.</p>
    258. 

  258. </dd>
    259. 

  259. </dl>
    260. 

  260. </dd></dl>
    261. 

  261. 

  262. <dl class="py method">
    263. 

  263. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.get_arch_params">
    264. 

  264. <span class="sig-name descname"><span class="pre">get_arch_params</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.get_arch_params"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.get_arch_params" title="Permalink to this definition"></a></dt>
    265. 

  265. <dd></dd></dl>
    266. 

  266. 

  267. <dl class="py method">
    268. 

  268. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.get_architecture">
    269. 

  269. <span class="sig-name descname"><span class="pre">get_architecture</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.get_architecture"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.get_architecture" title="Permalink to this definition"></a></dt>
    270. 

  270. <dd></dd></dl>
    271. 

  271. 

  272. <dl class="py method">
    273. 

  273. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.get_module">
    274. 

  274. <span class="sig-name descname"><span class="pre">get_module</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.get_module"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.get_module" title="Permalink to this definition"></a></dt>
    275. 

  275. <dd></dd></dl>
    276. 

  276. 

  277. <dl class="py method">
    278. 

  278. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.get_structure">
    279. 

  279. <span class="sig-name descname"><span class="pre">get_structure</span></span><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.get_structure"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.get_structure" title="Permalink to this definition"></a></dt>
    280. 

  280. <dd></dd></dl>
    281. 

  281. 

  282. <dl class="py method">
    283. 

  283. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.predict">
    284. 

  284. <span class="sig-name descname"><span class="pre">predict</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">inputs</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">targets</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">half</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">normalize</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">verbose</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">move_outputs_to_cpu</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.predict"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.predict" title="Permalink to this definition"></a></dt>
    285. 

  285. <dd><p>A fast predictor for a batch of inputs
    286. 

  286. :param inputs: torch.tensor or numpy.array</p>
    287. 

  287. <blockquote>
    288. 

  288. <div><p>a batch of inputs</p>
    289. 

  289. </div></blockquote>
    290. 

  290. <dl class="field-list simple">
    291. 

  291. <dt class="field-odd">Parameters</dt>
    292. 

  292. <dd class="field-odd"><ul class="simple">
    293. 

  293. <li><p><strong>targets</strong> – torch.tensor()
    294. 

  294. corresponding labels - if non are given - accuracy will not be computed</p></li>
    295. 

  295. <li><p><strong>verbose</strong> – bool
    296. 

  296. print the results to screen</p></li>
    297. 

  297. <li><p><strong>normalize</strong> – bool
    298. 

  298. If true, normalizes the tensor according to the dataloader’s normalization values</p></li>
    299. 

  299. <li><p><strong>half</strong> – Performs half precision evaluation</p></li>
    300. 

  300. <li><p><strong>move_outputs_to_cpu</strong> – Moves the results from the GPU to the CPU</p></li>
    301. 

  301. </ul>
    302. 

  302. </dd>
    303. 

  303. <dt class="field-even">Returns</dt>
    304. 

  304. <dd class="field-even"><p>outputs, acc, net_time, gross_time
    305. 

  305. networks predictions, accuracy calculation, forward pass net time, function gross time</p>
    306. 

  306. </dd>
    307. 

  307. </dl>
    308. 

  308. </dd></dl>
    309. 

  309. 

  310. <dl class="py method">
    311. 

  311. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.re_build_model">
    312. 

  312. <span class="sig-name descname"><span class="pre">re_build_model</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">arch_params</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">{}</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.re_build_model"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.re_build_model" title="Permalink to this definition"></a></dt>
    313. 

  313. <dd><dl class="simple">
    314. 

  314. <dt>arch_params<span class="classifier">dict</span></dt><dd><p>Architecture H.P. e.g.: block, num_blocks, num_classes, etc.</p>
    315. 

  315. </dd>
    316. 

  316. </dl>
    317. 

  317. <dl class="field-list simple">
    318. 

  318. <dt class="field-odd">Returns</dt>
    319. 

  319. <dd class="field-odd"><p></p>
    320. 

  320. </dd>
    321. 

  321. </dl>
    322. 

  322. </dd></dl>
    323. 

  323. 

  324. <dl class="py method">
    325. 

  325. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.save_checkpoint">
    326. 

  326. <span class="sig-name descname"><span class="pre">save_checkpoint</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">optimizer</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">epoch</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">int</span><span class="p"><span class="pre">]</span></span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">validation_results_tuple</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">tuple</span><span class="p"><span class="pre">]</span></span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.save_checkpoint"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.save_checkpoint" title="Permalink to this definition"></a></dt>
    327. 

  327. <dd><p>Save the current state dict as latest (always), best (if metric was improved), epoch# (if determined in training
    328. 

  328. params)</p>
    329. 

  329. </dd></dl>
    330. 

  330. 

  331. <dl class="py method">
    332. 

  332. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.set_experiment_name">
    333. 

  333. <span class="sig-name descname"><span class="pre">set_experiment_name</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">experiment_name</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.set_experiment_name"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.set_experiment_name" title="Permalink to this definition"></a></dt>
    334. 

  334. <dd></dd></dl>
    335. 

  335. 

  336. <dl class="py method">
    337. 

  337. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.set_module">
    338. 

  338. <span class="sig-name descname"><span class="pre">set_module</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">module</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.set_module"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.set_module" title="Permalink to this definition"></a></dt>
    339. 

  339. <dd></dd></dl>
    340. 

  340. 

  341. <dl class="py method">
    342. 

  342. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.test">
    343. 

  343. <span class="sig-name descname"><span class="pre">test</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">test_loader</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">torch.utils.data.dataloader.DataLoader</span><span class="p"><span class="pre">]</span></span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">loss</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">Optional</span><span class="p"><span class="pre">[</span></span><span class="pre">torch.nn.modules.loss._Loss</span><span class="p"><span class="pre">]</span></span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">silent_mode</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">bool</span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">False</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">test_metrics_list</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">loss_logging_items_names</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">metrics_progress_verbose</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">test_phase_callbacks</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span> &#x2192; <span class="pre">tuple</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.test"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.test" title="Permalink to this definition"></a></dt>
    344. 

  344. <dd><p>Evaluates the model on given dataloader and metrics.</p>
    345. 

  345. <dl class="field-list simple">
    346. 

  346. <dt class="field-odd">Parameters</dt>
    347. 

  347. <dd class="field-odd"><ul class="simple">
    348. 

  348. <li><p><strong>test_loader</strong> – dataloader to perform test on.</p></li>
    349. 

  349. <li><p><strong>test_metrics_list</strong> – (list(torchmetrics.Metric)) metrics list for evaluation.</p></li>
    350. 

  350. <li><p><strong>silent_mode</strong> – (bool) controls verbosity</p></li>
    351. 

  351. <li><p><strong>metrics_progress_verbose</strong> – (bool) controls the verbosity of metrics progress (default=False). Slows down the program.</p></li>
    352. 

  352. </ul>
    353. 

  353. </dd>
    354. 

  354. <dt class="field-even">Returns</dt>
    355. 

  355. <dd class="field-even"><p>results tuple (tuple) containing the loss items and metric values.</p>
    356. 

  356. </dd>
    357. 

  357. </dl>
    358. 

  358. <dl class="simple">
    359. 

  359. <dt>All of the above args will override SgModel’s corresponding attribute when not equal to None. Then evaluation</dt><dd><p>is ran on self.test_loader with self.test_metrics.</p>
    360. 

  360. </dd>
    361. 

  361. </dl>
    362. 

  362. </dd></dl>
    363. 

  363. 

  364. <dl class="py method">
    365. 

  365. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.train">
    366. 

  366. <span class="sig-name descname"><span class="pre">train</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">training_params</span></span><span class="p"><span class="pre">:</span></span> <span class="n"><span class="pre">dict</span></span> <span class="o"><span class="pre">=</span></span> <span class="default_value"><span class="pre">{}</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.train"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.train" title="Permalink to this definition"></a></dt>
    367. 

  367. <dd><p>train - Trains the Model</p>
    368. 

  368. <dl>
    369. 

  369. <dt>IMPORTANT NOTE: Additional batch parameters can be added as a third item (optional) if a tuple is returned by</dt><dd><p>the data loaders, as dictionary. The phase context will hold the additional items, under an attribute with
    370. 

  370. the same name as the key in this dictionary. Then such items can be accessed through phase callbacks.</p>
    371. 

  371. <blockquote>
    372. 

  372. <div><dl class="field-list">
    373. 

  373. <dt class="field-odd">param training_params</dt>
    374. 

  374. <dd class="field-odd"><ul>
    375. 

  375. <li><p><cite>max_epochs</cite> : int</p>
    376. 

  376. <blockquote>
    377. 

  377. <div><p>Number of epochs to run training.</p>
    378. 

  378. </div></blockquote>
    379. 

  379. </li>
    380. 

  380. <li><p><cite>lr_updates</cite> : list(int)</p>
    381. 

  381. <blockquote>
    382. 

  382. <div><p>List of fixed epoch numbers to perform learning rate updates when <cite>lr_mode=’step’</cite>.</p>
    383. 

  383. </div></blockquote>
    384. 

  384. </li>
    385. 

  385. <li><p><cite>lr_decay_factor</cite> : float</p>
    386. 

  386. <blockquote>
    387. 

  387. <div><p>Decay factor to apply to the learning rate at each update when <cite>lr_mode=’step’</cite>.</p>
    388. 

  388. </div></blockquote>
    389. 

  389. </li>
    390. 

  390. <li><p><cite>lr_mode</cite> : str</p>
    391. 

  391. <blockquote>
    392. 

  392. <div><p>Learning rate scheduling policy, one of [‘step’,’poly’,’cosine’,’function’]. ‘step’ refers to
    393. 

  393. constant updates at epoch numbers passed through <cite>lr_updates</cite>. ‘cosine’ refers to Cosine Anealing
    394. 

  394. policy as mentioned in <a class="reference external" href="https://arxiv.org/abs/1608.03983">https://arxiv.org/abs/1608.03983</a>. ‘poly’ refers to polynomial decrease i.e
    395. 

  395. in each epoch iteration <cite>self.lr = self.initial_lr * pow((1.0 - (current_iter / max_iter)),
    396. 

  396. 0.9)</cite> ‘function’ refers to user defined learning rate scheduling function, that is passed through
    397. 

  397. <cite>lr_schedule_function</cite>.</p>
    398. 

  398. </div></blockquote>
    399. 

  399. </li>
    400. 

  400. <li><p><cite>lr_schedule_function</cite> : Union[callable,None]</p>
    401. 

  401. <blockquote>
    402. 

  402. <div><p>Learning rate scheduling function to be used when <cite>lr_mode</cite> is ‘function’.</p>
    403. 

  403. </div></blockquote>
    404. 

  404. </li>
    405. 

  405. <li><p><cite>lr_warmup_epochs</cite> : int (default=0)</p>
    406. 

  406. <blockquote>
    407. 

  407. <div><p>Number of epochs for learning rate warm up - see <a class="reference external" href="https://arxiv.org/pdf/1706.02677.pdf">https://arxiv.org/pdf/1706.02677.pdf</a> (Section 2.2).</p>
    408. 

  408. </div></blockquote>
    409. 

  409. </li>
    410. 

  410. <li><dl class="simple">
    411. 

  411. <dt><cite>cosine_final_lr_ratio</cite><span class="classifier">float (default=0.01)</span></dt><dd><dl class="simple">
    412. 

  412. <dt>Final learning rate ratio (only relevant when <a href="#id1"><span class="problematic" id="id2">`</span></a>lr_mode`=’cosine’). The cosine starts from initial_lr and reaches</dt><dd><p>initial_lr * cosine_final_lr_ratio in last epoch</p>
    413. 

  413. </dd>
    414. 

  414. </dl>
    415. 

  415. </dd>
    416. 

  416. </dl>
    417. 

  417. </li>
    418. 

  418. <li><p><cite>inital_lr</cite> : float</p>
    419. 

  419. <blockquote>
    420. 

  420. <div><p>Initial learning rate.</p>
    421. 

  421. </div></blockquote>
    422. 

  422. </li>
    423. 

  423. <li><p><cite>loss</cite> : Union[nn.module, str]</p>
    424. 

  424. <blockquote>
    425. 

  425. <div><p>Loss function for training.
    426. 

  426. One of SuperGradient’s built in options:</p>
    427. 

  427. <blockquote>
    428. 

  428. <div><p>“cross_entropy”: LabelSmoothingCrossEntropyLoss,
    429. 

  429. “mse”: MSELoss,
    430. 

  430. “r_squared_loss”: RSquaredLoss,
    431. 

  431. “detection_loss”: YoLoV3DetectionLoss,
    432. 

  432. “shelfnet_ohem_loss”: ShelfNetOHEMLoss,
    433. 

  433. “shelfnet_se_loss”: ShelfNetSemanticEncodingLoss,
    434. 

  434. “yolo_v5_loss”: YoLoV5DetectionLoss,
    435. 

  435. “ssd_loss”: SSDLoss,</p>
    436. 

  436. </div></blockquote>
    437. 

  437. <p>or user defined nn.module loss function.</p>
    438. 

  438. <p>IMPORTANT: forward(…) should return a (loss, loss_items) tuple where loss is the tensor used
    439. 

  439. for backprop (i.e what your original loss function returns), and loss_items should be a tensor of
    440. 

  440. shape (n_items), of values computed during the forward pass which we desire to log over the
    441. 

  441. entire epoch. For example- the loss itself should always be logged. Another example is a scenario
    442. 

  442. where the computed loss is the sum of a few components we would like to log- these entries in
    443. 

  443. loss_items).</p>
    444. 

  444. <p>When training, set the loss_logging_items_names parameter in train_params to be a list of
    445. 

  445. strings, of length n_items who’s ith element is the name of the ith entry in loss_items. Then
    446. 

  446. each item will be logged, rendered on tensorboard and “watched” (i.e saving model checkpoints
    447. 

  447. according to it).</p>
    448. 

  448. <p>Since running logs will save the loss_items in some internal state, it is recommended that
    449. 

  449. loss_items are detached from their computational graph for memory efficiency.</p>
    450. 

  450. </div></blockquote>
    451. 

  451. </li>
    452. 

  452. <li><p><cite>optimizer</cite> : Union[str, torch.optim.Optimizer]</p>
    453. 

  453. <blockquote>
    454. 

  454. <div><p>Optimization algorithm. One of [‘Adam’,’SGD’,’RMSProp’] corresponding to the torch.optim
    455. 

  455. optimzers implementations, or any object that implements torch.optim.Optimizer.</p>
    456. 

  456. </div></blockquote>
    457. 

  457. </li>
    458. 

  458. <li><p><cite>criterion_params</cite> : dict</p>
    459. 

  459. <blockquote>
    460. 

  460. <div><p>Loss function parameters.</p>
    461. 

  461. </div></blockquote>
    462. 

  462. </li>
    463. 

  463. <li><dl>
    464. 

  464. <dt><cite>optimizer_params</cite><span class="classifier">dict</span></dt><dd><p>When <cite>optimizer</cite> is one of [‘Adam’,’SGD’,’RMSProp’], it will be initialized with optimizer_params.</p>
    465. 

  465. <p>(see <a class="reference external" href="https://pytorch.org/docs/stable/optim.html">https://pytorch.org/docs/stable/optim.html</a> for the full list of
    466. 

  466. parameters for each optimizer).</p>
    467. 

  467. </dd>
    468. 

  468. </dl>
    469. 

  469. </li>
    470. 

  470. <li><p><cite>train_metrics_list</cite> : list(torchmetrics.Metric)</p>
    471. 

  471. <blockquote>
    472. 

  472. <div><p>Metrics to log during training. For more information on torchmetrics see
    473. 

  473. <a class="reference external" href="https://torchmetrics.rtfd.io/en/latest/">https://torchmetrics.rtfd.io/en/latest/</a>.</p>
    474. 

  474. </div></blockquote>
    475. 

  475. </li>
    476. 

  476. <li><p><cite>valid_metrics_list</cite> : list(torchmetrics.Metric)</p>
    477. 

  477. <blockquote>
    478. 

  478. <div><p>Metrics to log during validation/testing. For more information on torchmetrics see
    479. 

  479. <a class="reference external" href="https://torchmetrics.rtfd.io/en/latest/">https://torchmetrics.rtfd.io/en/latest/</a>.</p>
    480. 

  480. </div></blockquote>
    481. 

  481. </li>
    482. 

  482. <li><p><cite>loss_logging_items_names</cite> : list(str)</p>
    483. 

  483. <blockquote>
    484. 

  484. <div><p>The list of names/titles for the outputs returned from the loss functions forward pass (reminder-
    485. 

  485. the loss function should return the tuple (loss, loss_items)). These names will be used for
    486. 

  486. logging their values.</p>
    487. 

  487. </div></blockquote>
    488. 

  488. </li>
    489. 

  489. <li><p><cite>metric_to_watch</cite> : str (default=”Accuracy”)</p>
    490. 

  490. <blockquote>
    491. 

  491. <div><p>will be the metric which the model checkpoint will be saved according to, and can be set to any
    492. 

  492. of the following:</p>
    493. 

  493. <blockquote>
    494. 

  494. <div><p>a metric name (str) of one of the metric objects from the valid_metrics_list</p>
    495. 

  495. <p>a “metric_name” if some metric in valid_metrics_list has an attribute component_names which
    496. 

  496. is a list referring to the names of each entry in the output metric (torch tensor of size n)</p>
    497. 

  497. <p>one of “loss_logging_items_names” i.e which will correspond to an item returned during the
    498. 

  498. loss function’s forward pass.</p>
    499. 

  499. </div></blockquote>
    500. 

  500. <p>At the end of each epoch, if a new best metric_to_watch value is achieved, the models checkpoint
    501. 

  501. is saved in YOUR_PYTHON_PATH/checkpoints/ckpt_best.pth</p>
    502. 

  502. </div></blockquote>
    503. 

  503. </li>
    504. 

  504. <li><p><cite>greater_metric_to_watch_is_better</cite> : bool</p>
    505. 

  505. <blockquote>
    506. 

  506. <div><dl class="simple">
    507. 

  507. <dt>When choosing a model’s checkpoint to be saved, the best achieved model is the one that maximizes the</dt><dd><p>metric_to_watch when this parameter is set to True, and a one that minimizes it otherwise.</p>
    508. 

  508. </dd>
    509. 

  509. </dl>
    510. 

  510. </div></blockquote>
    511. 

  511. </li>
    512. 

  512. <li><p><cite>ema</cite> : bool (default=False)</p>
    513. 

  513. <blockquote>
    514. 

  514. <div><p>Whether to use Model Exponential Moving Average (see
    515. 

  515. <a class="reference external" href="https://github.com/rwightman/pytorch-image-models">https://github.com/rwightman/pytorch-image-models</a> ema implementation)</p>
    516. 

  516. </div></blockquote>
    517. 

  517. </li>
    518. 

  518. <li><p><cite>batch_accumulate</cite> : int (default=1)</p>
    519. 

  519. <blockquote>
    520. 

  520. <div><p>Number of batches to accumulate before every backward pass.</p>
    521. 

  521. </div></blockquote>
    522. 

  522. </li>
    523. 

  523. <li><p><cite>ema_params</cite> : dict</p>
    524. 

  524. <blockquote>
    525. 

  525. <div><p>Parameters for the ema model.</p>
    526. 

  526. </div></blockquote>
    527. 

  527. </li>
    528. 

  528. <li><p><cite>zero_weight_decay_on_bias_and_bn</cite> : bool (default=False)</p>
    529. 

  529. <blockquote>
    530. 

  530. <div><p>Whether to apply weight decay on batch normalization parameters or not (ignored when the passed
    531. 

  531. optimizer has already been initialized).</p>
    532. 

  532. </div></blockquote>
    533. 

  533. </li>
    534. 

  534. <li><p><cite>load_opt_params</cite> : bool (default=True)</p>
    535. 

  535. <blockquote>
    536. 

  536. <div><p>Whether to load the optimizers parameters as well when loading a model’s checkpoint.</p>
    537. 

  537. </div></blockquote>
    538. 

  538. </li>
    539. 

  539. <li><p><cite>run_validation_freq</cite> : int (default=1)</p>
    540. 

  540. <blockquote>
    541. 

  541. <div><dl class="simple">
    542. 

  542. <dt>The frequency in which validation is performed during training (i.e the validation is ran every</dt><dd><p><cite>run_validation_freq</cite> epochs.</p>
    543. 

  543. </dd>
    544. 

  544. </dl>
    545. 

  545. </div></blockquote>
    546. 

  546. </li>
    547. 

  547. <li><p><cite>save_model</cite> : bool (default=True)</p>
    548. 

  548. <blockquote>
    549. 

  549. <div><p>Whether to save the model checkpoints.</p>
    550. 

  550. </div></blockquote>
    551. 

  551. </li>
    552. 

  552. <li><p><cite>launch_tensorboard</cite> : bool (default=False)</p>
    553. 

  553. <blockquote>
    554. 

  554. <div><p>Whether to launch a TensorBoard process.</p>
    555. 

  555. </div></blockquote>
    556. 

  556. </li>
    557. 

  557. <li><p><cite>tb_files_user_prompt</cite> : bool</p>
    558. 

  558. <blockquote>
    559. 

  559. <div><p>Asks user for Tensorboard deletion prompt.</p>
    560. 

  560. </div></blockquote>
    561. 

  561. </li>
    562. 

  562. <li><p><cite>silent_mode</cite> : bool</p>
    563. 

  563. <blockquote>
    564. 

  564. <div><p>Silents the print outs.</p>
    565. 

  565. </div></blockquote>
    566. 

  566. </li>
    567. 

  567. <li><p><cite>mixed_precision</cite> : bool</p>
    568. 

  568. <blockquote>
    569. 

  569. <div><p>Whether to use mixed precision or not.</p>
    570. 

  570. </div></blockquote>
    571. 

  571. </li>
    572. 

  572. <li><p><cite>tensorboard_port</cite> : int, None (default=None)</p>
    573. 

  573. <blockquote>
    574. 

  574. <div><p>Specific port number for the tensorboard to use when launched (when set to None, some free port
    575. 

  575. number will be used).</p>
    576. 

  576. </div></blockquote>
    577. 

  577. </li>
    578. 

  578. <li><p><cite>save_ckpt_epoch_list</cite> : list(int) (default=[])</p>
    579. 

  579. <blockquote>
    580. 

  580. <div><p>List of fixed epoch indices the user wishes to save checkpoints in.</p>
    581. 

  581. </div></blockquote>
    582. 

  582. </li>
    583. 

  583. <li><p><cite>average_best_models</cite> : bool (default=False)</p>
    584. 

  584. <blockquote>
    585. 

  585. <div><p>If set, a snapshot dictionary file and the average model will be saved / updated at every epoch
    586. 

  586. and evaluated only when training is completed. The snapshot file will only be deleted upon
    587. 

  587. completing the training. The snapshot dict will be managed on cpu.</p>
    588. 

  588. </div></blockquote>
    589. 

  589. </li>
    590. 

  590. <li><p><cite>save_tensorboard_to_s3</cite> : bool (default=False)</p>
    591. 

  591. <blockquote>
    592. 

  592. <div><p>Saves tensorboard in s3.</p>
    593. 

  593. </div></blockquote>
    594. 

  594. </li>
    595. 

  595. <li><p><cite>precise_bn</cite> : bool (default=False)</p>
    596. 

  596. <blockquote>
    597. 

  597. <div><p>Whether to use precise_bn calculation during the training.</p>
    598. 

  598. </div></blockquote>
    599. 

  599. </li>
    600. 

  600. <li><p><cite>precise_bn_batch_size</cite> : int (default=None)</p>
    601. 

  601. <blockquote>
    602. 

  602. <div><p>The effective batch size we want to calculate the batchnorm on. For example, if we are training a model
    603. 

  603. on 8 gpus, with a batch of 128 on each gpu, a good rule of thumb would be to give it 8192
    604. 

  604. (ie: effective_batch_size * num_gpus = batch_per_gpu * num_gpus * num_gpus).
    605. 

  605. If precise_bn_batch_size is not provided in the training_params, the latter heuristic will be taken.</p>
    606. 

  606. </div></blockquote>
    607. 

  607. </li>
    608. 

  608. <li><p><cite>seed</cite> : int (default=42)</p>
    609. 

  609. <blockquote>
    610. 

  610. <div><p>Random seed to be set for torch, numpy, and random. When using DDP each process will have it’s seed
    611. 

  611. set to seed + rank.</p>
    612. 

  612. </div></blockquote>
    613. 

  613. </li>
    614. 

  614. <li><p><cite>log_installed_packages</cite> : bool (default=False)</p>
    615. 

  615. <blockquote>
    616. 

  616. <div><dl class="simple">
    617. 

  617. <dt>When set, the list of all installed packages (and their versions) will be written to the tensorboard</dt><dd><p>and logfile (useful when trying to reproduce results).</p>
    618. 

  618. </dd>
    619. 

  619. </dl>
    620. 

  620. </div></blockquote>
    621. 

  621. </li>
    622. 

  622. <li><p><cite>dataset_statistics</cite> : bool (default=False)</p>
    623. 

  623. <blockquote>
    624. 

  624. <div><p>Enable a statistic analysis of the dataset. If set to True the dataset will be analyzed and a report
    625. 

  625. will be added to the tensorboard along with some sample images from the dataset. Currently only
    626. 

  626. detection datasets are supported for analysis.</p>
    627. 

  627. </div></blockquote>
    628. 

  628. </li>
    629. 

  629. <li><p><cite>save_full_train_log</cite> : bool (default=False)</p>
    630. 

  630. <blockquote>
    631. 

  631. <div><dl class="simple">
    632. 

  632. <dt>When set, a full log (of all super_gradients modules, including uncaught exceptions from any other</dt><dd><p>module) of the training will be saved in the checkpoint directory under full_train_log.log</p>
    633. 

  633. </dd>
    634. 

  634. </dl>
    635. 

  635. </div></blockquote>
    636. 

  636. </li>
    637. 

  637. </ul>
    638. 

  638. </dd>
    639. 

  639. </dl>
    640. 

  640. </div></blockquote>
    641. 

  641. </dd>
    642. 

  642. </dl>
    643. 

  643. <dl class="field-list simple">
    644. 

  644. <dt class="field-odd">Returns</dt>
    645. 

  645. <dd class="field-odd"><p></p>
    646. 

  646. </dd>
    647. 

  647. </dl>
    648. 

  648. </dd></dl>
    649. 

  649. 

  650. <dl class="py method">
    651. 

  651. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.SgModel.update_architecture">
    652. 

  652. <span class="sig-name descname"><span class="pre">update_architecture</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">structure</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#SgModel.update_architecture"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.SgModel.update_architecture" title="Permalink to this definition"></a></dt>
    653. 

  653. <dd><dl class="simple">
    654. 

  654. <dt>architecture<span class="classifier">str</span></dt><dd><p>Defines the network’s architecture according to the options in models/all_architectures</p>
    655. 

  655. </dd>
    656. 

  656. <dt>load_checkpoint<span class="classifier">bool</span></dt><dd><p>Loads a checkpoint according to experiment_name</p>
    657. 

  657. </dd>
    658. 

  658. <dt>arch_params<span class="classifier">dict</span></dt><dd><p>Architecture H.P. e.g.: block, num_blocks, num_classes, etc.</p>
    659. 

  659. </dd>
    660. 

  660. </dl>
    661. 

  661. <dl class="field-list simple">
    662. 

  662. <dt class="field-odd">Returns</dt>
    663. 

  663. <dd class="field-odd"><p></p>
    664. 

  664. </dd>
    665. 

  665. </dl>
    666. 

  666. </dd></dl>
    667. 

  667. 

  668. </dd></dl>
    669. 

  669. 

  670. <dl class="py class">
    671. 

  671. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.StrictLoad">
    672. 

  672. <em class="property"><span class="pre">class</span> </em><span class="sig-prename descclassname"><span class="pre">super_gradients.training.sg_model.sg_model.</span></span><span class="sig-name descname"><span class="pre">StrictLoad</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">value</span></span></em><span class="sig-paren">)</span><a class="reference internal" href="_modules/super_gradients/training/sg_model/sg_model.html#StrictLoad"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.StrictLoad" title="Permalink to this definition"></a></dt>
    673. 

  673. <dd><p>Bases: <code class="xref py py-class docutils literal notranslate"><span class="pre">enum.Enum</span></code></p>
    674. 

  674. <dl>
    675. 

  675. <dt>Wrapper for adding more functionality to torch’s strict_load parameter in load_state_dict().</dt><dd><dl>
    676. 

  676. <dt>Attributes:</dt><dd><p>OFF              - Native torch “strict_load = off” behaviour. See nn.Module.load_state_dict() documentation for more details.
    677. 

  677. ON               - Native torch “strict_load = on” behaviour. See nn.Module.load_state_dict() documentation for more details.
    678. 

  678. NO_KEY_MATCHING  - Allows the usage of SuperGradient’s adapt_checkpoint function, which loads a checkpoint by matching each</p>
    679. 

  679. <blockquote>
    680. 

  680. <div><p>layer’s shapes (and bypasses the strict matching of the names of each layer (ie: disregards the state_dict key matching)).</p>
    681. 

  681. </div></blockquote>
    682. 

  682. </dd>
    683. 

  683. </dl>
    684. 

  684. </dd>
    685. 

  685. </dl>
    686. 

  686. <dl class="py attribute">
    687. 

  687. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.StrictLoad.NO_KEY_MATCHING">
    688. 

  688. <span class="sig-name descname"><span class="pre">NO_KEY_MATCHING</span></span><em class="property"> <span class="pre">=</span> <span class="pre">'no_key_matching'</span></em><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.StrictLoad.NO_KEY_MATCHING" title="Permalink to this definition"></a></dt>
    689. 

  689. <dd></dd></dl>
    690. 

  690. 

  691. <dl class="py attribute">
    692. 

  692. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.StrictLoad.OFF">
    693. 

  693. <span class="sig-name descname"><span class="pre">OFF</span></span><em class="property"> <span class="pre">=</span> <span class="pre">False</span></em><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.StrictLoad.OFF" title="Permalink to this definition"></a></dt>
    694. 

  694. <dd></dd></dl>
    695. 

  695. 

  696. <dl class="py attribute">
    697. 

  697. <dt class="sig sig-object py" id="super_gradients.training.sg_model.sg_model.StrictLoad.ON">
    698. 

  698. <span class="sig-name descname"><span class="pre">ON</span></span><em class="property"> <span class="pre">=</span> <span class="pre">True</span></em><a class="headerlink" href="#super_gradients.training.sg_model.sg_model.StrictLoad.ON" title="Permalink to this definition"></a></dt>
    699. 

  699. <dd></dd></dl>
    700. 

  700. 

  701. </dd></dl>
    702. 

  702. 

  703. </section>
    704. 

  704. <section id="module-super_gradients.training.sg_model">
    705. 

  705. <span id="module-contents"></span><h2>Module contents<a class="headerlink" href="#module-super_gradients.training.sg_model" title="Permalink to this headline"></a></h2>
    706. 

  706. </section>
    707. 

  707. </section>
    708. 

  708. 

  709. 

  710.            </div>
    711. 

  711.           </div>
    712. 

  712.           <footer><div class="rst-footer-buttons" role="navigation" aria-label="Footer">
    713. 

  713.         <a href="super_gradients.training.models.html" class="btn btn-neutral float-left" title="super_gradients.training.models package" accesskey="p" rel="prev"><span class="fa fa-arrow-circle-left" aria-hidden="true"></span> Previous</a>
    714. 

  714.         <a href="super_gradients.training.utils.html" class="btn btn-neutral float-right" title="super_gradients.training.utils package" accesskey="n" rel="next">Next <span class="fa fa-arrow-circle-right" aria-hidden="true"></span></a>
    715. 

  715.     </div>
    716. 

  716. 

  717.   <hr/>
    718. 

  718. 

  719.   <div role="contentinfo">
    720. 

  720.     <p>&#169; Copyright 2021, SuperGradients team.</p>
    721. 

  721.   </div>
    722. 

  722. 

  723.   Built with <a href="https://www.sphinx-doc.org/">Sphinx</a> using a
    724. 

  724.     <a href="https://github.com/readthedocs/sphinx_rtd_theme">theme</a>
    725. 

  725.     provided by <a href="https://readthedocs.org">Read the Docs</a>.
    726. 

  726.    
    727. 

  727. 

  728. </footer>
    729. 

  729.         </div>
    730. 

  730.       </div>
    731. 

  731.     </section>
    732. 

  732.   </div>
    733. 

  733.   <script>
    734. 

  734.       jQuery(function () {
    735. 

  735.           SphinxRtdTheme.Navigation.enable(true);
    736. 

  736.       });
    737. 

  737.   </script> 
    738. 

  738. 

  739. </body>
    740. 

  740. </html>
    741.
Prev
Next
Tip!

Press p or to see the previous file or, n or to see the next file

Comments

Write Preview
Loading...
Integrate AWS S3
Use S3 remote
Select bucket
Access key
Finish

Use AWS S3 as storage!

Browsing data directories saved to S3 is possible with DAGsHub. Let's configure your repository to easily display your data in the context of any commit!

Specify your S3 bucket

Access key (If needed)

Congratulations!

super-gradients is now integrated with AWS S3!

Delete Storage Key

Are you sure you want to delete this access key?

No
Yes
Integrate Google Cloud Storage
Use Google Storage
Select bucket
Upload key
Finish

Use Google Cloud Storage!

Browsing data directories saved to Google Cloud Storage is possible with DAGsHub. Let's configure your repository to easily display your data in the context of any commit!

Specify your Google Storage bucket

Service Account Key

Congratulations!

super-gradients is now integrated with Google Cloud Storage!

Delete Storage Key

Are you sure you want to delete this access key?

No
Yes
Integrate Azure Cloud Storage
Use Azure Storage
Select bucket
Set key
Finish

Use Azure Cloud Storage!

Browsing data directories saved to Azure Cloud Storage is possible with DAGsHub. Let's configure your repository to easily display your data in the context of any commit!

Specify your Azure Storage bucket

Access key (If needed)

Congratulations!

super-gradients is now integrated with Azure Cloud Storage!

Delete Storage Key

Are you sure you want to delete this access key?

No
Yes
Integrate S3 compatible storage
Use S3 like remote
Select bucket
Access key
Finish

Use any S3 compatible storage!

Browsing data directories saved to S3 compatible storage is possible with DAGsHub. Let's configure your repository to easily display your data in the context of any commit!

Specify your S3 bucket

Bucket name cannot be the same as the repository name. Please change one of them.

Access key (If needed)

Congratulations!

super-gradients is now integrated with your S3 compatible storage!

Delete Storage Key

Are you sure you want to delete this access key?

No
Yes