-
Notifications
You must be signed in to change notification settings - Fork 336
/
pytorch-examples.html
1350 lines (1168 loc) · 253 KB
/
pytorch-examples.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta property="og:title" content="Writing better code with pytorch and einops">
<meta property="og:description" content="Learning by example: rewriting and fixing popular code fragments">
<meta property="og:image" content="http://arogozhnikov.github.io/images/einops/einops_video.gif">
<meta property="og:video" content="http://arogozhnikov.github.io/images/einops/einops_video.mp4" />
<meta property="og:url" content="https://arogozhnikov.github.io/einops/pytorch-examples.html">
<meta name="twitter:card" content="summary_large_image">
<!-- Non-Essential, But Recommended -->
<meta property="og:site_name" content="Writing better code with pytorch and einops">
<meta name="twitter:image:alt" content="Learning by example: rewriting and fixing popular code fragments">
<title>Writing better code with pytorch+einops</title>
<style>.highlight .hll { background-color: #ffffcc }
.highlight { background: #f8f8f8; }
.highlight .c { color: #408080; font-style: italic } /* Comment */
.highlight .err { border: 1px solid #FF0000 } /* Error */
.highlight .k { color: #008000; font-weight: bold } /* Keyword */
.highlight .o { color: #666666 } /* Operator */
.highlight .ch { color: #408080; font-style: italic } /* Comment.Hashbang */
.highlight .cm { color: #408080; font-style: italic } /* Comment.Multiline */
.highlight .cp { color: #BC7A00 } /* Comment.Preproc */
.highlight .cpf { color: #408080; font-style: italic } /* Comment.PreprocFile */
.highlight .c1 { color: #408080; font-style: italic } /* Comment.Single */
.highlight .cs { color: #408080; font-style: italic } /* Comment.Special */
.highlight .gd { color: #A00000 } /* Generic.Deleted */
.highlight .ge { font-style: italic } /* Generic.Emph */
.highlight .gr { color: #FF0000 } /* Generic.Error */
.highlight .gh { color: #000080; font-weight: bold } /* Generic.Heading */
.highlight .gi { color: #00A000 } /* Generic.Inserted */
.highlight .go { color: #888888 } /* Generic.Output */
.highlight .gp { color: #000080; font-weight: bold } /* Generic.Prompt */
.highlight .gs { font-weight: bold } /* Generic.Strong */
.highlight .gu { color: #800080; font-weight: bold } /* Generic.Subheading */
.highlight .gt { color: #0044DD } /* Generic.Traceback */
.highlight .kc { color: #008000; font-weight: bold } /* Keyword.Constant */
.highlight .kd { color: #008000; font-weight: bold } /* Keyword.Declaration */
.highlight .kn { color: #008000; font-weight: bold } /* Keyword.Namespace */
.highlight .kp { color: #008000 } /* Keyword.Pseudo */
.highlight .kr { color: #008000; font-weight: bold } /* Keyword.Reserved */
.highlight .kt { color: #B00040 } /* Keyword.Type */
.highlight .m { color: #666666 } /* Literal.Number */
.highlight .s { color: #BA2121 } /* Literal.String */
.highlight .na { color: #7D9029 } /* Name.Attribute */
.highlight .nb { color: #008000 } /* Name.Builtin */
.highlight .nc { color: #0000FF; font-weight: bold } /* Name.Class */
.highlight .no { color: #880000 } /* Name.Constant */
.highlight .nd { color: #AA22FF } /* Name.Decorator */
.highlight .ni { color: #999999; font-weight: bold } /* Name.Entity */
.highlight .ne { color: #D2413A; font-weight: bold } /* Name.Exception */
.highlight .nf { color: #0000FF } /* Name.Function */
.highlight .nl { color: #A0A000 } /* Name.Label */
.highlight .nn { color: #0000FF; font-weight: bold } /* Name.Namespace */
.highlight .nt { color: #008000; font-weight: bold } /* Name.Tag */
.highlight .nv { color: #19177C } /* Name.Variable */
.highlight .ow { color: #AA22FF; font-weight: bold } /* Operator.Word */
.highlight .w { color: #bbbbbb } /* Text.Whitespace */
.highlight .mb { color: #666666 } /* Literal.Number.Bin */
.highlight .mf { color: #666666 } /* Literal.Number.Float */
.highlight .mh { color: #666666 } /* Literal.Number.Hex */
.highlight .mi { color: #666666 } /* Literal.Number.Integer */
.highlight .mo { color: #666666 } /* Literal.Number.Oct */
.highlight .sa { color: #BA2121 } /* Literal.String.Affix */
.highlight .sb { color: #BA2121 } /* Literal.String.Backtick */
.highlight .sc { color: #BA2121 } /* Literal.String.Char */
.highlight .dl { color: #BA2121 } /* Literal.String.Delimiter */
.highlight .sd { color: #BA2121; font-style: italic } /* Literal.String.Doc */
.highlight .s2 { color: #BA2121 } /* Literal.String.Double */
.highlight .se { color: #BB6622; font-weight: bold } /* Literal.String.Escape */
.highlight .sh { color: #BA2121 } /* Literal.String.Heredoc */
.highlight .si { color: #BB6688; font-weight: bold } /* Literal.String.Interpol */
.highlight .sx { color: #008000 } /* Literal.String.Other */
.highlight .sr { color: #BB6688 } /* Literal.String.Regex */
.highlight .s1 { color: #BA2121 } /* Literal.String.Single */
.highlight .ss { color: #19177C } /* Literal.String.Symbol */
.highlight .bp { color: #008000 } /* Name.Builtin.Pseudo */
.highlight .fm { color: #0000FF } /* Name.Function.Magic */
.highlight .vc { color: #19177C } /* Name.Variable.Class */
.highlight .vg { color: #19177C } /* Name.Variable.Global */
.highlight .vi { color: #19177C } /* Name.Variable.Instance */
.highlight .vm { color: #19177C } /* Name.Variable.Magic */
.highlight .il { color: #666666 } /* Literal.Number.Integer.Long */
body {
padding: 50px 10px;
}
.leftright-wrapper {
text-align: center;
overflow-x: auto;
}
.leftright-cells {
display: inline-flex;
text-align: left;
}
.leftright-cells > div {
padding: 0px 10px;
min-width: 350px;
}
.markdown-cell{
max-width: 700px;
margin: 0px auto;
}
h1 {
text-align: center;
padding: 10px 0px 0px;
}
</style>
</head>
<body>
<a href="https://github.com/arogozhnikov/einops" class="github-corner" aria-label="View source on GitHub">
<svg width="80" height="80" viewBox="0 0 250 250" style="fill:#151513; color:#fff; position: absolute; top: 0; border: 0; right: 0;" aria-hidden="true">
<path d="M0,0 L115,115 L130,115 L142,142 L250,250 L250,0 Z"></path><path d="M128.3,109.0 C113.8,99.7 119.0,89.6 119.0,89.6 C122.0,82.7 120.5,78.6 120.5,78.6 C119.2,72.0 123.4,76.3 123.4,76.3 C127.3,80.9 125.5,87.3 125.5,87.3 C122.9,97.6 130.6,101.9 134.4,103.2" fill="currentColor" style="transform-origin: 130px 106px;" class="octo-arm"></path>
<path d="M115.0,115.0 C114.9,115.1 118.7,116.5 119.8,115.4 L133.7,101.6 C136.9,99.2 139.9,98.4 142.2,98.6 C133.8,88.0 127.5,74.4 143.8,58.0 C148.5,53.4 154.0,51.2 159.7,51.0 C160.3,49.4 163.2,43.6 171.4,40.1 C171.4,40.1 176.1,42.5 178.8,56.2 C183.1,58.6 187.2,61.8 190.9,65.4 C194.5,69.0 197.7,73.2 200.1,77.6 C213.8,80.2 216.3,84.9 216.3,84.9 C212.7,93.1 206.9,96.0 205.4,96.6 C205.1,102.4 203.0,107.8 198.3,112.5 C181.9,128.9 168.3,122.5 157.7,114.1 C157.9,116.9 156.7,120.9 152.7,124.9 L141.0,136.5 C139.8,137.7 141.6,141.9 141.8,141.8 Z" fill="currentColor" class="octo-body"></path>
</svg></a>
<style>.github-corner:hover .octo-arm{animation:octocat-wave 560ms ease-in-out}@keyframes octocat-wave{0%,100%{transform:rotate(0)}20%,60%{transform:rotate(-25deg)}40%,80%{transform:rotate(10deg)}}@media (max-width:500px){.github-corner:hover .octo-arm{animation:none}.github-corner .octo-arm{animation:octocat-wave 560ms ease-in-out}}</style>
<div class='markdown-cell'><div align="center">
<a href="https://github.com/arogozhnikov/einops">
<img src="http://arogozhnikov.github.io/images/einops/einops_logo_350x350.png" alt="einops package logo" width="150" height="150" style='padding: 50px 50px 25px;' />
</a>
<div>
<a href="https://github.com/arogozhnikov/einops">[github]</a>,
tutorials
<a href="https://github.com/arogozhnikov/einops/blob/master/docs/1-einops-basics.ipynb">[1]</a> and
<a href="https://github.com/arogozhnikov/einops/blob/master/docs/2-einops-for-deep-learning.ipynb">[2]</a>
<br />
<br />
</div>
</div>
<h1>Writing a better code with pytorch and einops</h1>
<p><br /><br /></p>
<h2>Rewriting building blocks of deep learning</h2>
<p>Now let's get to examples from real world.
These code fragments taken from official tutorials and popular repositories.</p>
<p>Learn how to improve code and how <code>einops</code> can help you.</p>
<p><strong>Left</strong>: as it was, <strong>Right</strong>: improved version</p></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="c1"># start from importing some stuff</span>
<span class="kn">import</span> <span class="nn">torch</span>
<span class="kn">import</span> <span class="nn">torch.nn</span> <span class="k">as</span> <span class="nn">nn</span>
<span class="kn">import</span> <span class="nn">torch.nn.functional</span> <span class="k">as</span> <span class="nn">F</span>
<span class="kn">import</span> <span class="nn">numpy</span> <span class="k">as</span> <span class="nn">np</span>
<span class="kn">import</span> <span class="nn">math</span>
<span class="kn">from</span> <span class="nn">einops</span> <span class="kn">import</span> <span class="n">rearrange</span><span class="p">,</span> <span class="n">reduce</span><span class="p">,</span> <span class="n">asnumpy</span><span class="p">,</span> <span class="n">parse_shape</span>
<span class="kn">from</span> <span class="nn">einops.layers.torch</span> <span class="kn">import</span> <span class="n">Rearrange</span><span class="p">,</span> <span class="n">Reduce</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><h1>Simple ConvNet</h1></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">Net</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="nb">super</span><span class="p">(</span><span class="n">Net</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">10</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">5</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv2</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">10</span><span class="p">,</span> <span class="mi">20</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">5</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv2_drop</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Dropout2d</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">fc1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">320</span><span class="p">,</span> <span class="mi">50</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">fc2</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">50</span><span class="p">,</span> <span class="mi">10</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="n">F</span><span class="o">.</span><span class="n">max_pool2d</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv1</span><span class="p">(</span><span class="n">x</span><span class="p">),</span> <span class="mi">2</span><span class="p">))</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="n">F</span><span class="o">.</span><span class="n">max_pool2d</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv2_drop</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv2</span><span class="p">(</span><span class="n">x</span><span class="p">)),</span> <span class="mi">2</span><span class="p">))</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="mi">320</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">fc1</span><span class="p">(</span><span class="n">x</span><span class="p">))</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">dropout</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">training</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">training</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">fc2</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="k">return</span> <span class="n">F</span><span class="o">.</span><span class="n">log_softmax</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">dim</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
<span class="n">conv_net_old</span> <span class="o">=</span> <span class="n">Net</span><span class="p">()</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="n">conv_net_new</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">10</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">5</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">MaxPool2d</span><span class="p">(</span><span class="n">kernel_size</span><span class="o">=</span><span class="mi">2</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">(),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">10</span><span class="p">,</span> <span class="mi">20</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">5</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">MaxPool2d</span><span class="p">(</span><span class="n">kernel_size</span><span class="o">=</span><span class="mi">2</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">(),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Dropout2d</span><span class="p">(),</span>
<span class="n">Rearrange</span><span class="p">(</span><span class="s1">'b c h w -> b (c h w)'</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">320</span><span class="p">,</span> <span class="mi">50</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">(),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Dropout</span><span class="p">(),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">50</span><span class="p">,</span> <span class="mi">10</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">LogSoftmax</span><span class="p">(</span><span class="n">dim</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
<span class="p">)</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><p>Reasons to prefer new implementation:</p>
<ul>
<li>in the original code (to the left) if input size is changed and batch size is divisible by 16 (that's usually so), we'll get something senseless after reshaping<ul>
<li>new code will explicitly raise an error in this case</li>
</ul>
</li>
<li>we won't forget to use dropout with flag self.training with new version</li>
<li>code is straightforward to read and analyze</li>
<li>sequential makes printing / saving / passing trivial. And there is no need in your code to load a model (which also has a number of benefits)</li>
<li>don't need logsoftmax? Now you can use <code>conv_net_new[:-1]</code>. One more reason to prefer <code>nn.Sequential</code></li>
<li>... and we could also add inplace for ReLU</li>
</ul></div><div class='markdown-cell'><h1>Super-resolution</h1>
<!-- minified https://github.com/pytorch/examples/tree/master/super_resolution, withour initialization --></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">SuperResolutionNetOld</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">upscale_factor</span><span class="p">):</span>
<span class="nb">super</span><span class="p">(</span><span class="n">SuperResolutionNetOld</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">relu</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">64</span><span class="p">,</span> <span class="p">(</span><span class="mi">5</span><span class="p">,</span> <span class="mi">5</span><span class="p">),</span> <span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">),</span> <span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">2</span><span class="p">))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv2</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">64</span><span class="p">,</span> <span class="mi">64</span><span class="p">,</span> <span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">3</span><span class="p">),</span> <span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">),</span> <span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv3</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">64</span><span class="p">,</span> <span class="mi">32</span><span class="p">,</span> <span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">3</span><span class="p">),</span> <span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">),</span> <span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv4</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">32</span><span class="p">,</span> <span class="n">upscale_factor</span> <span class="o">**</span> <span class="mi">2</span><span class="p">,</span> <span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">3</span><span class="p">),</span> <span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">),</span> <span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">pixel_shuffle</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">PixelShuffle</span><span class="p">(</span><span class="n">upscale_factor</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv1</span><span class="p">(</span><span class="n">x</span><span class="p">))</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv2</span><span class="p">(</span><span class="n">x</span><span class="p">))</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv3</span><span class="p">(</span><span class="n">x</span><span class="p">))</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">pixel_shuffle</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv4</span><span class="p">(</span><span class="n">x</span><span class="p">))</span>
<span class="k">return</span> <span class="n">x</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">SuperResolutionNetNew</span><span class="p">(</span><span class="n">upscale_factor</span><span class="p">):</span>
<span class="k">return</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">64</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">5</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">2</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">(</span><span class="n">inplace</span><span class="o">=</span><span class="kc">True</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">64</span><span class="p">,</span> <span class="mi">64</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">(</span><span class="n">inplace</span><span class="o">=</span><span class="kc">True</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">64</span><span class="p">,</span> <span class="mi">32</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">(</span><span class="n">inplace</span><span class="o">=</span><span class="kc">True</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">32</span><span class="p">,</span> <span class="n">upscale_factor</span> <span class="o">**</span> <span class="mi">2</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">),</span>
<span class="n">Rearrange</span><span class="p">(</span><span class="s1">'b (h2 w2) h w -> b (h h2) (w w2)'</span><span class="p">,</span> <span class="n">h2</span><span class="o">=</span><span class="n">upscale_factor</span><span class="p">,</span> <span class="n">w2</span><span class="o">=</span><span class="n">upscale_factor</span><span class="p">),</span>
<span class="p">)</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><p>Here is the difference:</p>
<ul>
<li>no need in special instruction pixel_shuffle (and result is transferrable between frameworks)</li>
<li>output doesn't contain a fake axis (and we could do the same for the input)</li>
<li>inplace ReLU used now, for high resolution pictures that becomes critical and saves us much memory</li>
<li>and all the benefits of nn.Sequential again</li>
</ul></div><div class='markdown-cell'><h1>Restyling Gram matrix for style transfer</h1></div><div class='markdown-cell'><!-- from https://github.com/pytorch/examples/blob/29c2ed8ca6dc36fc78a3e74a5908615619987863/fast_neural_style/neural_style/utils.py#L21-L26 -->
<p>Original code is already good - first line shows what kind of input is expected</p>
<ul>
<li>einsum operation should be read like:</li>
<li>for each batch and for each pair of channels, we sum over h and w.</li>
<li>I've also changed normalization, because that's how Gram matrix is defined, otherwise we should call it normalized Gram matrix or alike</li>
</ul></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">gram_matrix_old</span><span class="p">(</span><span class="n">y</span><span class="p">):</span>
<span class="p">(</span><span class="n">b</span><span class="p">,</span> <span class="n">ch</span><span class="p">,</span> <span class="n">h</span><span class="p">,</span> <span class="n">w</span><span class="p">)</span> <span class="o">=</span> <span class="n">y</span><span class="o">.</span><span class="n">size</span><span class="p">()</span>
<span class="n">features</span> <span class="o">=</span> <span class="n">y</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">b</span><span class="p">,</span> <span class="n">ch</span><span class="p">,</span> <span class="n">w</span> <span class="o">*</span> <span class="n">h</span><span class="p">)</span>
<span class="n">features_t</span> <span class="o">=</span> <span class="n">features</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="n">gram</span> <span class="o">=</span> <span class="n">features</span><span class="o">.</span><span class="n">bmm</span><span class="p">(</span><span class="n">features_t</span><span class="p">)</span> <span class="o">/</span> <span class="p">(</span><span class="n">ch</span> <span class="o">*</span> <span class="n">h</span> <span class="o">*</span> <span class="n">w</span><span class="p">)</span>
<span class="k">return</span> <span class="n">gram</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">gram_matrix_new</span><span class="p">(</span><span class="n">y</span><span class="p">):</span>
<span class="n">b</span><span class="p">,</span> <span class="n">ch</span><span class="p">,</span> <span class="n">h</span><span class="p">,</span> <span class="n">w</span> <span class="o">=</span> <span class="n">y</span><span class="o">.</span><span class="n">shape</span>
<span class="k">return</span> <span class="n">torch</span><span class="o">.</span><span class="n">einsum</span><span class="p">(</span><span class="s1">'bchw,bdhw->bcd'</span><span class="p">,</span> <span class="p">[</span><span class="n">y</span><span class="p">,</span> <span class="n">y</span><span class="p">])</span> <span class="o">/</span> <span class="p">(</span><span class="n">h</span> <span class="o">*</span> <span class="n">w</span><span class="p">)</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><p>It would be great to use just <code>'b c1 h w,b c2 h w->b c1 c2'</code>, but einsum supports only one-letter axes</p></div><div class='markdown-cell'><h1>Recurrent model</h1>
<p>All we did here is just made information about shapes explicit to skip deciphering</p>
<!-- simplified version of https://github.com/pytorch/examples/blob/master/word_language_model/model.py --></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">RNNModelOld</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="sd">"""Container module with an encoder, a recurrent module, and a decoder."""</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">ntoken</span><span class="p">,</span> <span class="n">ninp</span><span class="p">,</span> <span class="n">nhid</span><span class="p">,</span> <span class="n">nlayers</span><span class="p">,</span> <span class="n">dropout</span><span class="o">=</span><span class="mf">0.5</span><span class="p">):</span>
<span class="nb">super</span><span class="p">(</span><span class="n">RNNModel</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">drop</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Dropout</span><span class="p">(</span><span class="n">dropout</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">encoder</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Embedding</span><span class="p">(</span><span class="n">ntoken</span><span class="p">,</span> <span class="n">ninp</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">rnn</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">LSTM</span><span class="p">(</span><span class="n">ninp</span><span class="p">,</span> <span class="n">nhid</span><span class="p">,</span> <span class="n">nlayers</span><span class="p">,</span> <span class="n">dropout</span><span class="o">=</span><span class="n">dropout</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">decoder</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">nhid</span><span class="p">,</span> <span class="n">ntoken</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="nb">input</span><span class="p">,</span> <span class="n">hidden</span><span class="p">):</span>
<span class="n">emb</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">drop</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">encoder</span><span class="p">(</span><span class="nb">input</span><span class="p">))</span>
<span class="n">output</span><span class="p">,</span> <span class="n">hidden</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">rnn</span><span class="p">(</span><span class="n">emb</span><span class="p">,</span> <span class="n">hidden</span><span class="p">)</span>
<span class="n">output</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">drop</span><span class="p">(</span><span class="n">output</span><span class="p">)</span>
<span class="n">decoded</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">decoder</span><span class="p">(</span><span class="n">output</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">output</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="mi">0</span><span class="p">)</span><span class="o">*</span><span class="n">output</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="mi">1</span><span class="p">),</span> <span class="n">output</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="mi">2</span><span class="p">)))</span>
<span class="k">return</span> <span class="n">decoded</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">output</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="mi">0</span><span class="p">),</span> <span class="n">output</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="mi">1</span><span class="p">),</span> <span class="n">decoded</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="mi">1</span><span class="p">)),</span> <span class="n">hidden</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">RNNModelNew</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="sd">"""Container module with an encoder, a recurrent module, and a decoder."""</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">ntoken</span><span class="p">,</span> <span class="n">ninp</span><span class="p">,</span> <span class="n">nhid</span><span class="p">,</span> <span class="n">nlayers</span><span class="p">,</span> <span class="n">dropout</span><span class="o">=</span><span class="mf">0.5</span><span class="p">):</span>
<span class="nb">super</span><span class="p">(</span><span class="n">RNNModel</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">drop</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Dropout</span><span class="p">(</span><span class="n">p</span><span class="o">=</span><span class="n">dropout</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">encoder</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Embedding</span><span class="p">(</span><span class="n">ntoken</span><span class="p">,</span> <span class="n">ninp</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">rnn</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">LSTM</span><span class="p">(</span><span class="n">ninp</span><span class="p">,</span> <span class="n">nhid</span><span class="p">,</span> <span class="n">nlayers</span><span class="p">,</span> <span class="n">dropout</span><span class="o">=</span><span class="n">dropout</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">decoder</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">nhid</span><span class="p">,</span> <span class="n">ntoken</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="nb">input</span><span class="p">,</span> <span class="n">hidden</span><span class="p">):</span>
<span class="n">t</span><span class="p">,</span> <span class="n">b</span> <span class="o">=</span> <span class="nb">input</span><span class="o">.</span><span class="n">shape</span>
<span class="n">emb</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">drop</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">encoder</span><span class="p">(</span><span class="nb">input</span><span class="p">))</span>
<span class="n">output</span><span class="p">,</span> <span class="n">hidden</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">rnn</span><span class="p">(</span><span class="n">emb</span><span class="p">,</span> <span class="n">hidden</span><span class="p">)</span>
<span class="n">output</span> <span class="o">=</span> <span class="n">rearrange</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">drop</span><span class="p">(</span><span class="n">output</span><span class="p">),</span> <span class="s1">'t b nhid -> (t b) nhid'</span><span class="p">)</span>
<span class="n">decoded</span> <span class="o">=</span> <span class="n">rearrange</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">decoder</span><span class="p">(</span><span class="n">output</span><span class="p">),</span> <span class="s1">'(t b) token -> t b token'</span><span class="p">,</span> <span class="n">t</span><span class="o">=</span><span class="n">t</span><span class="p">,</span> <span class="n">b</span><span class="o">=</span><span class="n">b</span><span class="p">)</span>
<span class="k">return</span> <span class="n">decoded</span><span class="p">,</span> <span class="n">hidden</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><h1>Channel shuffle (from shufflenet)</h1>
<!-- from https://github.com/jaxony/ShuffleNet/blob/master/model.py --></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">channel_shuffle_old</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">groups</span><span class="p">):</span>
<span class="n">batchsize</span><span class="p">,</span> <span class="n">num_channels</span><span class="p">,</span> <span class="n">height</span><span class="p">,</span> <span class="n">width</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">size</span><span class="p">()</span>
<span class="n">channels_per_group</span> <span class="o">=</span> <span class="n">num_channels</span> <span class="o">//</span> <span class="n">groups</span>
<span class="c1"># reshape</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">batchsize</span><span class="p">,</span> <span class="n">groups</span><span class="p">,</span>
<span class="n">channels_per_group</span><span class="p">,</span> <span class="n">height</span><span class="p">,</span> <span class="n">width</span><span class="p">)</span>
<span class="c1"># transpose</span>
<span class="c1"># - contiguous() required if transpose() is used before view().</span>
<span class="c1"># See https://github.com/pytorch/pytorch/issues/764</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span><span class="o">.</span><span class="n">contiguous</span><span class="p">()</span>
<span class="c1"># flatten</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">batchsize</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="n">height</span><span class="p">,</span> <span class="n">width</span><span class="p">)</span>
<span class="k">return</span> <span class="n">x</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">channel_shuffle_new</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">groups</span><span class="p">):</span>
<span class="k">return</span> <span class="n">rearrange</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="s1">'b (c1 c2) h w -> b (c2 c1) h w'</span><span class="p">,</span> <span class="n">c1</span><span class="o">=</span><span class="n">groups</span><span class="p">)</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><p>While progress is obvious, this is not the limit. As you'll see below, we don't even need to write these couple of lines.</p></div><div class='markdown-cell'><h1>Shufflenet</h1></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="kn">from</span> <span class="nn">collections</span> <span class="kn">import</span> <span class="n">OrderedDict</span>
<span class="k">def</span> <span class="nf">channel_shuffle</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">groups</span><span class="p">):</span>
<span class="n">batchsize</span><span class="p">,</span> <span class="n">num_channels</span><span class="p">,</span> <span class="n">height</span><span class="p">,</span> <span class="n">width</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">size</span><span class="p">()</span>
<span class="n">channels_per_group</span> <span class="o">=</span> <span class="n">num_channels</span> <span class="o">//</span> <span class="n">groups</span>
<span class="c1"># reshape</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">batchsize</span><span class="p">,</span> <span class="n">groups</span><span class="p">,</span>
<span class="n">channels_per_group</span><span class="p">,</span> <span class="n">height</span><span class="p">,</span> <span class="n">width</span><span class="p">)</span>
<span class="c1"># transpose</span>
<span class="c1"># - contiguous() required if transpose() is used before view().</span>
<span class="c1"># See https://github.com/pytorch/pytorch/issues/764</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span><span class="o">.</span><span class="n">contiguous</span><span class="p">()</span>
<span class="c1"># flatten</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">batchsize</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="n">height</span><span class="p">,</span> <span class="n">width</span><span class="p">)</span>
<span class="k">return</span> <span class="n">x</span>
<span class="k">class</span> <span class="nc">ShuffleUnitOld</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">in_channels</span><span class="p">,</span> <span class="n">out_channels</span><span class="p">,</span> <span class="n">groups</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span>
<span class="n">grouped_conv</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span> <span class="n">combine</span><span class="o">=</span><span class="s1">'add'</span><span class="p">):</span>
<span class="nb">super</span><span class="p">(</span><span class="n">ShuffleUnitOld</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">in_channels</span> <span class="o">=</span> <span class="n">in_channels</span>
<span class="bp">self</span><span class="o">.</span><span class="n">out_channels</span> <span class="o">=</span> <span class="n">out_channels</span>
<span class="bp">self</span><span class="o">.</span><span class="n">grouped_conv</span> <span class="o">=</span> <span class="n">grouped_conv</span>
<span class="bp">self</span><span class="o">.</span><span class="n">combine</span> <span class="o">=</span> <span class="n">combine</span>
<span class="bp">self</span><span class="o">.</span><span class="n">groups</span> <span class="o">=</span> <span class="n">groups</span>
<span class="bp">self</span><span class="o">.</span><span class="n">bottleneck_channels</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">out_channels</span> <span class="o">//</span> <span class="mi">4</span>
<span class="c1"># define the type of ShuffleUnit</span>
<span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">combine</span> <span class="o">==</span> <span class="s1">'add'</span><span class="p">:</span>
<span class="c1"># ShuffleUnit Figure 2b</span>
<span class="bp">self</span><span class="o">.</span><span class="n">depthwise_stride</span> <span class="o">=</span> <span class="mi">1</span>
<span class="bp">self</span><span class="o">.</span><span class="n">_combine_func</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_add</span>
<span class="k">elif</span> <span class="bp">self</span><span class="o">.</span><span class="n">combine</span> <span class="o">==</span> <span class="s1">'concat'</span><span class="p">:</span>
<span class="c1"># ShuffleUnit Figure 2c</span>
<span class="bp">self</span><span class="o">.</span><span class="n">depthwise_stride</span> <span class="o">=</span> <span class="mi">2</span>
<span class="bp">self</span><span class="o">.</span><span class="n">_combine_func</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_concat</span>
<span class="c1"># ensure output of concat has the same channels as </span>
<span class="c1"># original output channels.</span>
<span class="bp">self</span><span class="o">.</span><span class="n">out_channels</span> <span class="o">-=</span> <span class="bp">self</span><span class="o">.</span><span class="n">in_channels</span>
<span class="k">else</span><span class="p">:</span>
<span class="k">raise</span> <span class="ne">ValueError</span><span class="p">(</span><span class="s2">"Cannot combine tensors with </span><span class="se">\"</span><span class="si">{}</span><span class="se">\"</span><span class="s2">"</span> \
<span class="s2">"Only </span><span class="se">\"</span><span class="s2">add</span><span class="se">\"</span><span class="s2"> and </span><span class="se">\"</span><span class="s2">concat</span><span class="se">\"</span><span class="s2"> are"</span> \
<span class="s2">"supported"</span><span class="o">.</span><span class="n">format</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">combine</span><span class="p">))</span>
<span class="c1"># Use a 1x1 grouped or non-grouped convolution to reduce input channels</span>
<span class="c1"># to bottleneck channels, as in a ResNet bottleneck module.</span>
<span class="c1"># NOTE: Do not use group convolution for the first conv1x1 in Stage 2.</span>
<span class="bp">self</span><span class="o">.</span><span class="n">first_1x1_groups</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">groups</span> <span class="k">if</span> <span class="n">grouped_conv</span> <span class="k">else</span> <span class="mi">1</span>
<span class="bp">self</span><span class="o">.</span><span class="n">g_conv_1x1_compress</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_make_grouped_conv1x1</span><span class="p">(</span>
<span class="bp">self</span><span class="o">.</span><span class="n">in_channels</span><span class="p">,</span>
<span class="bp">self</span><span class="o">.</span><span class="n">bottleneck_channels</span><span class="p">,</span>
<span class="bp">self</span><span class="o">.</span><span class="n">first_1x1_groups</span><span class="p">,</span>
<span class="n">batch_norm</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span>
<span class="n">relu</span><span class="o">=</span><span class="kc">True</span>
<span class="p">)</span>
<span class="c1"># 3x3 depthwise convolution followed by batch normalization</span>
<span class="bp">self</span><span class="o">.</span><span class="n">depthwise_conv3x3</span> <span class="o">=</span> <span class="n">conv3x3</span><span class="p">(</span>
<span class="bp">self</span><span class="o">.</span><span class="n">bottleneck_channels</span><span class="p">,</span> <span class="bp">self</span><span class="o">.</span><span class="n">bottleneck_channels</span><span class="p">,</span>
<span class="n">stride</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">depthwise_stride</span><span class="p">,</span> <span class="n">groups</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">bottleneck_channels</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">bn_after_depthwise</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">bottleneck_channels</span><span class="p">)</span>
<span class="c1"># Use 1x1 grouped convolution to expand from </span>
<span class="c1"># bottleneck_channels to out_channels</span>
<span class="bp">self</span><span class="o">.</span><span class="n">g_conv_1x1_expand</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_make_grouped_conv1x1</span><span class="p">(</span>
<span class="bp">self</span><span class="o">.</span><span class="n">bottleneck_channels</span><span class="p">,</span>
<span class="bp">self</span><span class="o">.</span><span class="n">out_channels</span><span class="p">,</span>
<span class="bp">self</span><span class="o">.</span><span class="n">groups</span><span class="p">,</span>
<span class="n">batch_norm</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span>
<span class="n">relu</span><span class="o">=</span><span class="kc">False</span>
<span class="p">)</span>
<span class="nd">@staticmethod</span>
<span class="k">def</span> <span class="nf">_add</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">out</span><span class="p">):</span>
<span class="c1"># residual connection</span>
<span class="k">return</span> <span class="n">x</span> <span class="o">+</span> <span class="n">out</span>
<span class="nd">@staticmethod</span>
<span class="k">def</span> <span class="nf">_concat</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="n">out</span><span class="p">):</span>
<span class="c1"># concatenate along channel axis</span>
<span class="k">return</span> <span class="n">torch</span><span class="o">.</span><span class="n">cat</span><span class="p">((</span><span class="n">x</span><span class="p">,</span> <span class="n">out</span><span class="p">),</span> <span class="mi">1</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">_make_grouped_conv1x1</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">in_channels</span><span class="p">,</span> <span class="n">out_channels</span><span class="p">,</span> <span class="n">groups</span><span class="p">,</span>
<span class="n">batch_norm</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span> <span class="n">relu</span><span class="o">=</span><span class="kc">False</span><span class="p">):</span>
<span class="n">modules</span> <span class="o">=</span> <span class="n">OrderedDict</span><span class="p">()</span>
<span class="n">conv</span> <span class="o">=</span> <span class="n">conv1x1</span><span class="p">(</span><span class="n">in_channels</span><span class="p">,</span> <span class="n">out_channels</span><span class="p">,</span> <span class="n">groups</span><span class="o">=</span><span class="n">groups</span><span class="p">)</span>
<span class="n">modules</span><span class="p">[</span><span class="s1">'conv1x1'</span><span class="p">]</span> <span class="o">=</span> <span class="n">conv</span>
<span class="k">if</span> <span class="n">batch_norm</span><span class="p">:</span>
<span class="n">modules</span><span class="p">[</span><span class="s1">'batch_norm'</span><span class="p">]</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="n">out_channels</span><span class="p">)</span>
<span class="k">if</span> <span class="n">relu</span><span class="p">:</span>
<span class="n">modules</span><span class="p">[</span><span class="s1">'relu'</span><span class="p">]</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">()</span>
<span class="k">if</span> <span class="nb">len</span><span class="p">(</span><span class="n">modules</span><span class="p">)</span> <span class="o">></span> <span class="mi">1</span><span class="p">:</span>
<span class="k">return</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span><span class="n">modules</span><span class="p">)</span>
<span class="k">else</span><span class="p">:</span>
<span class="k">return</span> <span class="n">conv</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="c1"># save for combining later with output</span>
<span class="n">residual</span> <span class="o">=</span> <span class="n">x</span>
<span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">combine</span> <span class="o">==</span> <span class="s1">'concat'</span><span class="p">:</span>
<span class="n">residual</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">avg_pool2d</span><span class="p">(</span><span class="n">residual</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span>
<span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
<span class="n">out</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">g_conv_1x1_compress</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">out</span> <span class="o">=</span> <span class="n">channel_shuffle</span><span class="p">(</span><span class="n">out</span><span class="p">,</span> <span class="bp">self</span><span class="o">.</span><span class="n">groups</span><span class="p">)</span>
<span class="n">out</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">depthwise_conv3x3</span><span class="p">(</span><span class="n">out</span><span class="p">)</span>
<span class="n">out</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">bn_after_depthwise</span><span class="p">(</span><span class="n">out</span><span class="p">)</span>
<span class="n">out</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">g_conv_1x1_expand</span><span class="p">(</span><span class="n">out</span><span class="p">)</span>
<span class="n">out</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_combine_func</span><span class="p">(</span><span class="n">residual</span><span class="p">,</span> <span class="n">out</span><span class="p">)</span>
<span class="k">return</span> <span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="n">out</span><span class="p">)</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">ShuffleUnitNew</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">in_channels</span><span class="p">,</span> <span class="n">out_channels</span><span class="p">,</span> <span class="n">groups</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span>
<span class="n">grouped_conv</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span> <span class="n">combine</span><span class="o">=</span><span class="s1">'add'</span><span class="p">):</span>
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="n">first_1x1_groups</span> <span class="o">=</span> <span class="n">groups</span> <span class="k">if</span> <span class="n">grouped_conv</span> <span class="k">else</span> <span class="mi">1</span>
<span class="n">bottleneck_channels</span> <span class="o">=</span> <span class="n">out_channels</span> <span class="o">//</span> <span class="mi">4</span>
<span class="bp">self</span><span class="o">.</span><span class="n">combine</span> <span class="o">=</span> <span class="n">combine</span>
<span class="k">if</span> <span class="n">combine</span> <span class="o">==</span> <span class="s1">'add'</span><span class="p">:</span>
<span class="c1"># ShuffleUnit Figure 2b</span>
<span class="bp">self</span><span class="o">.</span><span class="n">left</span> <span class="o">=</span> <span class="n">Rearrange</span><span class="p">(</span><span class="s1">'...->...'</span><span class="p">)</span> <span class="c1"># identity</span>
<span class="n">depthwise_stride</span> <span class="o">=</span> <span class="mi">1</span>
<span class="k">else</span><span class="p">:</span>
<span class="c1"># ShuffleUnit Figure 2c</span>
<span class="bp">self</span><span class="o">.</span><span class="n">left</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">AvgPool2d</span><span class="p">(</span><span class="n">kernel_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
<span class="n">depthwise_stride</span> <span class="o">=</span> <span class="mi">2</span>
<span class="c1"># ensure output of concat has the same channels as original output channels.</span>
<span class="n">out_channels</span> <span class="o">-=</span> <span class="n">in_channels</span>
<span class="k">assert</span> <span class="n">out_channels</span> <span class="o">></span> <span class="mi">0</span>
<span class="bp">self</span><span class="o">.</span><span class="n">right</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span>
<span class="c1"># Use a 1x1 grouped or non-grouped convolution to reduce input channels</span>
<span class="c1"># to bottleneck channels, as in a ResNet bottleneck module.</span>
<span class="n">conv1x1</span><span class="p">(</span><span class="n">in_channels</span><span class="p">,</span> <span class="n">bottleneck_channels</span><span class="p">,</span> <span class="n">groups</span><span class="o">=</span><span class="n">first_1x1_groups</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="n">bottleneck_channels</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">(</span><span class="n">inplace</span><span class="o">=</span><span class="kc">True</span><span class="p">),</span>
<span class="c1"># channel shuffle</span>
<span class="n">Rearrange</span><span class="p">(</span><span class="s1">'b (c1 c2) h w -> b (c2 c1) h w'</span><span class="p">,</span> <span class="n">c1</span><span class="o">=</span><span class="n">groups</span><span class="p">),</span>
<span class="c1"># 3x3 depthwise convolution followed by batch </span>
<span class="n">conv3x3</span><span class="p">(</span><span class="n">bottleneck_channels</span><span class="p">,</span> <span class="n">bottleneck_channels</span><span class="p">,</span>
<span class="n">stride</span><span class="o">=</span><span class="n">depthwise_stride</span><span class="p">,</span> <span class="n">groups</span><span class="o">=</span><span class="n">bottleneck_channels</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="n">bottleneck_channels</span><span class="p">),</span>
<span class="c1"># Use 1x1 grouped convolution to expand from </span>
<span class="c1"># bottleneck_channels to out_channels</span>
<span class="n">conv1x1</span><span class="p">(</span><span class="n">bottleneck_channels</span><span class="p">,</span> <span class="n">out_channels</span><span class="p">,</span> <span class="n">groups</span><span class="o">=</span><span class="n">groups</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="n">out_channels</span><span class="p">),</span>
<span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="k">if</span> <span class="bp">self</span><span class="o">.</span><span class="n">combine</span> <span class="o">==</span> <span class="s1">'add'</span><span class="p">:</span>
<span class="n">combined</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">left</span><span class="p">(</span><span class="n">x</span><span class="p">)</span> <span class="o">+</span> <span class="bp">self</span><span class="o">.</span><span class="n">right</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="k">else</span><span class="p">:</span>
<span class="n">combined</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">cat</span><span class="p">([</span><span class="bp">self</span><span class="o">.</span><span class="n">left</span><span class="p">(</span><span class="n">x</span><span class="p">),</span> <span class="bp">self</span><span class="o">.</span><span class="n">right</span><span class="p">(</span><span class="n">x</span><span class="p">)],</span> <span class="n">dim</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
<span class="k">return</span> <span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="n">combined</span><span class="p">,</span> <span class="n">inplace</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><p>Rewriting the code helped to identify:</p>
<ul>
<li>There is no sense in doing reshuffling and not using groups in the first convolution
(indeed, in the paper it is not so). However, result is an equivalent model.</li>
<li>It is also strange that the first convolution may be not grouped, while the last convolution is always grouped
(and that is different from the paper)</li>
</ul>
<p>Other comments:</p>
<ul>
<li>There is an identity layer for pytorch introduced here</li>
<li>The last thing left is get rid of conv1x1 and conv3x3 in the code - those are not better than standard</li>
</ul></div><div class='markdown-cell'><h1>Simplifying ResNet</h1></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">ResNetOld</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">block</span><span class="p">,</span> <span class="n">layers</span><span class="p">,</span> <span class="n">num_classes</span><span class="o">=</span><span class="mi">1000</span><span class="p">):</span>
<span class="bp">self</span><span class="o">.</span><span class="n">inplanes</span> <span class="o">=</span> <span class="mi">64</span>
<span class="nb">super</span><span class="p">(</span><span class="n">ResNetOld</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">64</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">7</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span>
<span class="n">bias</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">bn1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="mi">64</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">relu</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">(</span><span class="n">inplace</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">maxpool</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">MaxPool2d</span><span class="p">(</span><span class="n">kernel_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">layer1</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_make_layer</span><span class="p">(</span><span class="n">block</span><span class="p">,</span> <span class="mi">64</span><span class="p">,</span> <span class="n">layers</span><span class="p">[</span><span class="mi">0</span><span class="p">])</span>
<span class="bp">self</span><span class="o">.</span><span class="n">layer2</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_make_layer</span><span class="p">(</span><span class="n">block</span><span class="p">,</span> <span class="mi">128</span><span class="p">,</span> <span class="n">layers</span><span class="p">[</span><span class="mi">1</span><span class="p">],</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">layer3</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_make_layer</span><span class="p">(</span><span class="n">block</span><span class="p">,</span> <span class="mi">256</span><span class="p">,</span> <span class="n">layers</span><span class="p">[</span><span class="mi">2</span><span class="p">],</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">layer4</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">_make_layer</span><span class="p">(</span><span class="n">block</span><span class="p">,</span> <span class="mi">512</span><span class="p">,</span> <span class="n">layers</span><span class="p">[</span><span class="mi">3</span><span class="p">],</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">avgpool</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">AvgPool2d</span><span class="p">(</span><span class="mi">7</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">fc</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">512</span> <span class="o">*</span> <span class="n">block</span><span class="o">.</span><span class="n">expansion</span><span class="p">,</span> <span class="n">num_classes</span><span class="p">)</span>
<span class="k">for</span> <span class="n">m</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">modules</span><span class="p">():</span>
<span class="k">if</span> <span class="nb">isinstance</span><span class="p">(</span><span class="n">m</span><span class="p">,</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">):</span>
<span class="n">n</span> <span class="o">=</span> <span class="n">m</span><span class="o">.</span><span class="n">kernel_size</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span> <span class="o">*</span> <span class="n">m</span><span class="o">.</span><span class="n">kernel_size</span><span class="p">[</span><span class="mi">1</span><span class="p">]</span> <span class="o">*</span> <span class="n">m</span><span class="o">.</span><span class="n">out_channels</span>
<span class="n">m</span><span class="o">.</span><span class="n">weight</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">normal_</span><span class="p">(</span><span class="mi">0</span><span class="p">,</span> <span class="n">math</span><span class="o">.</span><span class="n">sqrt</span><span class="p">(</span><span class="mf">2.</span> <span class="o">/</span> <span class="n">n</span><span class="p">))</span>
<span class="k">elif</span> <span class="nb">isinstance</span><span class="p">(</span><span class="n">m</span><span class="p">,</span> <span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">):</span>
<span class="n">m</span><span class="o">.</span><span class="n">weight</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">fill_</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span>
<span class="n">m</span><span class="o">.</span><span class="n">bias</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">zero_</span><span class="p">()</span>
<span class="k">def</span> <span class="nf">_make_layer</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">block</span><span class="p">,</span> <span class="n">planes</span><span class="p">,</span> <span class="n">blocks</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">):</span>
<span class="n">downsample</span> <span class="o">=</span> <span class="kc">None</span>
<span class="k">if</span> <span class="n">stride</span> <span class="o">!=</span> <span class="mi">1</span> <span class="ow">or</span> <span class="bp">self</span><span class="o">.</span><span class="n">inplanes</span> <span class="o">!=</span> <span class="n">planes</span> <span class="o">*</span> <span class="n">block</span><span class="o">.</span><span class="n">expansion</span><span class="p">:</span>
<span class="n">downsample</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">inplanes</span><span class="p">,</span> <span class="n">planes</span> <span class="o">*</span> <span class="n">block</span><span class="o">.</span><span class="n">expansion</span><span class="p">,</span>
<span class="n">kernel_size</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="n">stride</span><span class="p">,</span> <span class="n">bias</span><span class="o">=</span><span class="kc">False</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="n">planes</span> <span class="o">*</span> <span class="n">block</span><span class="o">.</span><span class="n">expansion</span><span class="p">),</span>
<span class="p">)</span>
<span class="n">layers</span> <span class="o">=</span> <span class="p">[]</span>
<span class="n">layers</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">block</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">inplanes</span><span class="p">,</span> <span class="n">planes</span><span class="p">,</span> <span class="n">stride</span><span class="p">,</span> <span class="n">downsample</span><span class="p">))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">inplanes</span> <span class="o">=</span> <span class="n">planes</span> <span class="o">*</span> <span class="n">block</span><span class="o">.</span><span class="n">expansion</span>
<span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">blocks</span><span class="p">):</span>
<span class="n">layers</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">block</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">inplanes</span><span class="p">,</span> <span class="n">planes</span><span class="p">))</span>
<span class="k">return</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span><span class="o">*</span><span class="n">layers</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">conv1</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">bn1</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">maxpool</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">layer1</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">layer2</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">layer3</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">layer4</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">avgpool</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">x</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="mi">0</span><span class="p">),</span> <span class="o">-</span><span class="mi">1</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">fc</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="k">return</span> <span class="n">x</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">make_layer</span><span class="p">(</span><span class="n">inplanes</span><span class="p">,</span> <span class="n">planes</span><span class="p">,</span> <span class="n">block</span><span class="p">,</span> <span class="n">n_blocks</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">):</span>
<span class="n">downsample</span> <span class="o">=</span> <span class="kc">None</span>
<span class="k">if</span> <span class="n">stride</span> <span class="o">!=</span> <span class="mi">1</span> <span class="ow">or</span> <span class="n">inplanes</span> <span class="o">!=</span> <span class="n">planes</span> <span class="o">*</span> <span class="n">block</span><span class="o">.</span><span class="n">expansion</span><span class="p">:</span>
<span class="c1"># output size won't match input, so adjust residual</span>
<span class="n">downsample</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="n">inplanes</span><span class="p">,</span> <span class="n">planes</span> <span class="o">*</span> <span class="n">block</span><span class="o">.</span><span class="n">expansion</span><span class="p">,</span>
<span class="n">kernel_size</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="n">stride</span><span class="p">,</span> <span class="n">bias</span><span class="o">=</span><span class="kc">False</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="n">planes</span> <span class="o">*</span> <span class="n">block</span><span class="o">.</span><span class="n">expansion</span><span class="p">),</span>
<span class="p">)</span>
<span class="k">return</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span>
<span class="n">block</span><span class="p">(</span><span class="n">inplanes</span><span class="p">,</span> <span class="n">planes</span><span class="p">,</span> <span class="n">stride</span><span class="p">,</span> <span class="n">downsample</span><span class="p">),</span>
<span class="o">*</span><span class="p">[</span><span class="n">block</span><span class="p">(</span><span class="n">planes</span> <span class="o">*</span> <span class="n">block</span><span class="o">.</span><span class="n">expansion</span><span class="p">,</span> <span class="n">planes</span><span class="p">)</span> <span class="k">for</span> <span class="n">_</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">n_blocks</span><span class="p">)]</span>
<span class="p">)</span>
<span class="k">def</span> <span class="nf">ResNetNew</span><span class="p">(</span><span class="n">block</span><span class="p">,</span> <span class="n">layers</span><span class="p">,</span> <span class="n">num_classes</span><span class="o">=</span><span class="mi">1000</span><span class="p">):</span>
<span class="n">e</span> <span class="o">=</span> <span class="n">block</span><span class="o">.</span><span class="n">expansion</span>
<span class="n">resnet</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span>
<span class="n">Rearrange</span><span class="p">(</span><span class="s1">'b c h w -> b c h w'</span><span class="p">,</span> <span class="n">c</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">h</span><span class="o">=</span><span class="mi">224</span><span class="p">,</span> <span class="n">w</span><span class="o">=</span><span class="mi">224</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="mi">3</span><span class="p">,</span> <span class="mi">64</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">7</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">bias</span><span class="o">=</span><span class="kc">False</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">(</span><span class="mi">64</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">(</span><span class="n">inplace</span><span class="o">=</span><span class="kc">True</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">MaxPool2d</span><span class="p">(</span><span class="n">kernel_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">),</span>
<span class="n">make_layer</span><span class="p">(</span><span class="mi">64</span><span class="p">,</span> <span class="mi">64</span><span class="p">,</span> <span class="n">block</span><span class="p">,</span> <span class="n">layers</span><span class="p">[</span><span class="mi">0</span><span class="p">],</span> <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">),</span>
<span class="n">make_layer</span><span class="p">(</span><span class="mi">64</span> <span class="o">*</span> <span class="n">e</span><span class="p">,</span> <span class="mi">128</span><span class="p">,</span> <span class="n">block</span><span class="p">,</span> <span class="n">layers</span><span class="p">[</span><span class="mi">1</span><span class="p">],</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">),</span>
<span class="n">make_layer</span><span class="p">(</span><span class="mi">128</span> <span class="o">*</span> <span class="n">e</span><span class="p">,</span> <span class="mi">256</span><span class="p">,</span> <span class="n">block</span><span class="p">,</span> <span class="n">layers</span><span class="p">[</span><span class="mi">2</span><span class="p">],</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">),</span>
<span class="n">make_layer</span><span class="p">(</span><span class="mi">256</span> <span class="o">*</span> <span class="n">e</span><span class="p">,</span> <span class="mi">512</span><span class="p">,</span> <span class="n">block</span><span class="p">,</span> <span class="n">layers</span><span class="p">[</span><span class="mi">3</span><span class="p">],</span> <span class="n">stride</span><span class="o">=</span><span class="mi">2</span><span class="p">),</span>
<span class="c1"># combined AvgPool and view in one averaging operation</span>
<span class="n">Reduce</span><span class="p">(</span><span class="s1">'b c h w -> b c'</span><span class="p">,</span> <span class="s1">'mean'</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="mi">512</span> <span class="o">*</span> <span class="n">e</span><span class="p">,</span> <span class="n">num_classes</span><span class="p">),</span>
<span class="p">)</span>
<span class="c1"># initialization</span>
<span class="k">for</span> <span class="n">m</span> <span class="ow">in</span> <span class="n">resnet</span><span class="o">.</span><span class="n">modules</span><span class="p">():</span>
<span class="k">if</span> <span class="nb">isinstance</span><span class="p">(</span><span class="n">m</span><span class="p">,</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">):</span>
<span class="n">n</span> <span class="o">=</span> <span class="n">m</span><span class="o">.</span><span class="n">kernel_size</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span> <span class="o">*</span> <span class="n">m</span><span class="o">.</span><span class="n">kernel_size</span><span class="p">[</span><span class="mi">1</span><span class="p">]</span> <span class="o">*</span> <span class="n">m</span><span class="o">.</span><span class="n">out_channels</span>
<span class="n">m</span><span class="o">.</span><span class="n">weight</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">normal_</span><span class="p">(</span><span class="mi">0</span><span class="p">,</span> <span class="n">math</span><span class="o">.</span><span class="n">sqrt</span><span class="p">(</span><span class="mf">2.</span> <span class="o">/</span> <span class="n">n</span><span class="p">))</span>
<span class="k">elif</span> <span class="nb">isinstance</span><span class="p">(</span><span class="n">m</span><span class="p">,</span> <span class="n">nn</span><span class="o">.</span><span class="n">BatchNorm2d</span><span class="p">):</span>
<span class="n">m</span><span class="o">.</span><span class="n">weight</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">fill_</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span>
<span class="n">m</span><span class="o">.</span><span class="n">bias</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">zero_</span><span class="p">()</span>
<span class="k">return</span> <span class="n">resnet</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><p>Changes:</p>
<ul>
<li>explicit check for input shape</li>
<li>no views and simple sequential structure, output is just nn.Sequential, so can always be saved/passed/etc</li>
<li>no need in AvgPool and additional views, this place is much clearer now</li>
<li><code>make_layer</code> doesn't use internal state (that's quite faulty place)</li>
</ul></div><div class='markdown-cell'><h1>Improving RNN language modelling</h1></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">RNNOld</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">,</span> <span class="n">hidden_dim</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">,</span> <span class="n">n_layers</span><span class="p">,</span> <span class="n">bidirectional</span><span class="p">,</span> <span class="n">dropout</span><span class="p">):</span>
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">embedding</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Embedding</span><span class="p">(</span><span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">rnn</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">LSTM</span><span class="p">(</span><span class="n">embedding_dim</span><span class="p">,</span> <span class="n">hidden_dim</span><span class="p">,</span> <span class="n">num_layers</span><span class="o">=</span><span class="n">n_layers</span><span class="p">,</span>
<span class="n">bidirectional</span><span class="o">=</span><span class="n">bidirectional</span><span class="p">,</span> <span class="n">dropout</span><span class="o">=</span><span class="n">dropout</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">fc</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">hidden_dim</span><span class="o">*</span><span class="mi">2</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">dropout</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Dropout</span><span class="p">(</span><span class="n">dropout</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="c1">#x = [sent len, batch size]</span>
<span class="n">embedded</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">dropout</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">embedding</span><span class="p">(</span><span class="n">x</span><span class="p">))</span>
<span class="c1">#embedded = [sent len, batch size, emb dim]</span>
<span class="n">output</span><span class="p">,</span> <span class="p">(</span><span class="n">hidden</span><span class="p">,</span> <span class="n">cell</span><span class="p">)</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">rnn</span><span class="p">(</span><span class="n">embedded</span><span class="p">)</span>
<span class="c1">#output = [sent len, batch size, hid dim * num directions]</span>
<span class="c1">#hidden = [num layers * num directions, batch size, hid dim]</span>
<span class="c1">#cell = [num layers * num directions, batch size, hid dim]</span>
<span class="c1">#concat the final forward (hidden[-2,:,:]) and backward (hidden[-1,:,:]) hidden layers</span>
<span class="c1">#and apply dropout</span>
<span class="n">hidden</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">dropout</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">cat</span><span class="p">((</span><span class="n">hidden</span><span class="p">[</span><span class="o">-</span><span class="mi">2</span><span class="p">,:,:],</span> <span class="n">hidden</span><span class="p">[</span><span class="o">-</span><span class="mi">1</span><span class="p">,:,:]),</span> <span class="n">dim</span><span class="o">=</span><span class="mi">1</span><span class="p">))</span>
<span class="c1">#hidden = [batch size, hid dim * num directions]</span>
<span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">fc</span><span class="p">(</span><span class="n">hidden</span><span class="o">.</span><span class="n">squeeze</span><span class="p">(</span><span class="mi">0</span><span class="p">))</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">RNNNew</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">,</span> <span class="n">hidden_dim</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">,</span> <span class="n">n_layers</span><span class="p">,</span> <span class="n">bidirectional</span><span class="p">,</span> <span class="n">dropout</span><span class="p">):</span>
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">embedding</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Embedding</span><span class="p">(</span><span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">rnn</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">LSTM</span><span class="p">(</span><span class="n">embedding_dim</span><span class="p">,</span> <span class="n">hidden_dim</span><span class="p">,</span> <span class="n">num_layers</span><span class="o">=</span><span class="n">n_layers</span><span class="p">,</span>
<span class="n">bidirectional</span><span class="o">=</span><span class="n">bidirectional</span><span class="p">,</span> <span class="n">dropout</span><span class="o">=</span><span class="n">dropout</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">dropout</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Dropout</span><span class="p">(</span><span class="n">dropout</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">directions</span> <span class="o">=</span> <span class="mi">2</span> <span class="k">if</span> <span class="n">bidirectional</span> <span class="k">else</span> <span class="mi">1</span>
<span class="bp">self</span><span class="o">.</span><span class="n">fc</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">hidden_dim</span> <span class="o">*</span> <span class="bp">self</span><span class="o">.</span><span class="n">directions</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="c1">#x = [sent len, batch size] </span>
<span class="n">embedded</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">dropout</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">embedding</span><span class="p">(</span><span class="n">x</span><span class="p">))</span>
<span class="c1">#embedded = [sent len, batch size, emb dim]</span>
<span class="n">output</span><span class="p">,</span> <span class="p">(</span><span class="n">hidden</span><span class="p">,</span> <span class="n">cell</span><span class="p">)</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">rnn</span><span class="p">(</span><span class="n">embedded</span><span class="p">)</span>
<span class="n">hidden</span> <span class="o">=</span> <span class="n">rearrange</span><span class="p">(</span><span class="n">hidden</span><span class="p">,</span> <span class="s1">'(layer dir) b c -> layer b (dir c)'</span><span class="p">,</span>
<span class="nb">dir</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">directions</span><span class="p">)</span>
<span class="c1"># take the final layer's hidden</span>
<span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">fc</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">dropout</span><span class="p">(</span><span class="n">hidden</span><span class="p">[</span><span class="o">-</span><span class="mi">1</span><span class="p">]))</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><ul>
<li>original code misbehaves for non-bidirectional models</li>
<li>... and fails when bidirectional = False, and there is only one layer</li>
<li>modification of the code shows both how hidden is structured and how it is modified</li>
</ul></div><div class='markdown-cell'><h1>Writing FastText faster</h1>
<!-- from # https://github.com/bentrevett/pytorch-sentiment-analysis/blob/master/3%20-%20Faster%20Sentiment%20Analysis.ipynb --></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">FastTextOld</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">):</span>
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">embedding</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Embedding</span><span class="p">(</span><span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">fc</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">embedding_dim</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="c1">#x = [sent len, batch size]</span>
<span class="n">embedded</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">embedding</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="c1">#embedded = [sent len, batch size, emb dim]</span>
<span class="n">embedded</span> <span class="o">=</span> <span class="n">embedded</span><span class="o">.</span><span class="n">permute</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="c1">#embedded = [batch size, sent len, emb dim]</span>
<span class="n">pooled</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">avg_pool2d</span><span class="p">(</span><span class="n">embedded</span><span class="p">,</span> <span class="p">(</span><span class="n">embedded</span><span class="o">.</span><span class="n">shape</span><span class="p">[</span><span class="mi">1</span><span class="p">],</span> <span class="mi">1</span><span class="p">))</span><span class="o">.</span><span class="n">squeeze</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span>
<span class="c1">#pooled = [batch size, embedding_dim]</span>
<span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">fc</span><span class="p">(</span><span class="n">pooled</span><span class="p">)</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">FastTextNew</span><span class="p">(</span><span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">):</span>
<span class="k">return</span> <span class="n">nn</span><span class="o">.</span><span class="n">Sequential</span><span class="p">(</span>
<span class="n">Rearrange</span><span class="p">(</span><span class="s1">'t b -> t b'</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Embedding</span><span class="p">(</span><span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">),</span>
<span class="n">Reduce</span><span class="p">(</span><span class="s1">'t b c -> b c'</span><span class="p">,</span> <span class="s1">'mean'</span><span class="p">),</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">embedding_dim</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">),</span>
<span class="n">Rearrange</span><span class="p">(</span><span class="s1">'b c -> b c'</span><span class="p">),</span>
<span class="p">)</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><p>Some comments on new code:</p>
<ul>
<li>first and last operations do nothing and can be removed<ul>
<li>but were added to explicitly show expected input and output</li>
</ul>
</li>
<li>this also gives you a flexibility of changing interface by editing a single line. Should you need to accept inputs as (batch, time),
you just change first line to <code>Rearrange('b t -> t b'),</code></li>
</ul></div><div class='markdown-cell'><h1>CNNs for text classification</h1></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">CNNOld</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">,</span> <span class="n">n_filters</span><span class="p">,</span> <span class="n">filter_sizes</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">,</span> <span class="n">dropout</span><span class="p">):</span>
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">embedding</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Embedding</span><span class="p">(</span><span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv_0</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="n">in_channels</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">out_channels</span><span class="o">=</span><span class="n">n_filters</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="p">(</span><span class="n">filter_sizes</span><span class="p">[</span><span class="mi">0</span><span class="p">],</span><span class="n">embedding_dim</span><span class="p">))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv_1</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="n">in_channels</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">out_channels</span><span class="o">=</span><span class="n">n_filters</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="p">(</span><span class="n">filter_sizes</span><span class="p">[</span><span class="mi">1</span><span class="p">],</span><span class="n">embedding_dim</span><span class="p">))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv_2</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Conv2d</span><span class="p">(</span><span class="n">in_channels</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">out_channels</span><span class="o">=</span><span class="n">n_filters</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="p">(</span><span class="n">filter_sizes</span><span class="p">[</span><span class="mi">2</span><span class="p">],</span><span class="n">embedding_dim</span><span class="p">))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">fc</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="nb">len</span><span class="p">(</span><span class="n">filter_sizes</span><span class="p">)</span><span class="o">*</span><span class="n">n_filters</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">dropout</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Dropout</span><span class="p">(</span><span class="n">dropout</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="c1">#x = [sent len, batch size]</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">permute</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">0</span><span class="p">)</span>
<span class="c1">#x = [batch size, sent len]</span>
<span class="n">embedded</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">embedding</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="c1">#embedded = [batch size, sent len, emb dim]</span>
<span class="n">embedded</span> <span class="o">=</span> <span class="n">embedded</span><span class="o">.</span><span class="n">unsqueeze</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span>
<span class="c1">#embedded = [batch size, 1, sent len, emb dim]</span>
<span class="n">conved_0</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv_0</span><span class="p">(</span><span class="n">embedded</span><span class="p">)</span><span class="o">.</span><span class="n">squeeze</span><span class="p">(</span><span class="mi">3</span><span class="p">))</span>
<span class="n">conved_1</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv_1</span><span class="p">(</span><span class="n">embedded</span><span class="p">)</span><span class="o">.</span><span class="n">squeeze</span><span class="p">(</span><span class="mi">3</span><span class="p">))</span>
<span class="n">conved_2</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv_2</span><span class="p">(</span><span class="n">embedded</span><span class="p">)</span><span class="o">.</span><span class="n">squeeze</span><span class="p">(</span><span class="mi">3</span><span class="p">))</span>
<span class="c1">#conv_n = [batch size, n_filters, sent len - filter_sizes[n]]</span>
<span class="n">pooled_0</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">max_pool1d</span><span class="p">(</span><span class="n">conved_0</span><span class="p">,</span> <span class="n">conved_0</span><span class="o">.</span><span class="n">shape</span><span class="p">[</span><span class="mi">2</span><span class="p">])</span><span class="o">.</span><span class="n">squeeze</span><span class="p">(</span><span class="mi">2</span><span class="p">)</span>
<span class="n">pooled_1</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">max_pool1d</span><span class="p">(</span><span class="n">conved_1</span><span class="p">,</span> <span class="n">conved_1</span><span class="o">.</span><span class="n">shape</span><span class="p">[</span><span class="mi">2</span><span class="p">])</span><span class="o">.</span><span class="n">squeeze</span><span class="p">(</span><span class="mi">2</span><span class="p">)</span>
<span class="n">pooled_2</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">max_pool1d</span><span class="p">(</span><span class="n">conved_2</span><span class="p">,</span> <span class="n">conved_2</span><span class="o">.</span><span class="n">shape</span><span class="p">[</span><span class="mi">2</span><span class="p">])</span><span class="o">.</span><span class="n">squeeze</span><span class="p">(</span><span class="mi">2</span><span class="p">)</span>
<span class="c1">#pooled_n = [batch size, n_filters]</span>
<span class="n">cat</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">dropout</span><span class="p">(</span><span class="n">torch</span><span class="o">.</span><span class="n">cat</span><span class="p">((</span><span class="n">pooled_0</span><span class="p">,</span> <span class="n">pooled_1</span><span class="p">,</span> <span class="n">pooled_2</span><span class="p">),</span> <span class="n">dim</span><span class="o">=</span><span class="mi">1</span><span class="p">))</span>
<span class="c1">#cat = [batch size, n_filters * len(filter_sizes)]</span>
<span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">fc</span><span class="p">(</span><span class="n">cat</span><span class="p">)</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">CNNNew</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">,</span> <span class="n">n_filters</span><span class="p">,</span> <span class="n">filter_sizes</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">,</span> <span class="n">dropout</span><span class="p">):</span>
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">embedding</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Embedding</span><span class="p">(</span><span class="n">vocab_size</span><span class="p">,</span> <span class="n">embedding_dim</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">convs</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">ModuleList</span><span class="p">([</span>
<span class="n">nn</span><span class="o">.</span><span class="n">Conv1d</span><span class="p">(</span><span class="n">embedding_dim</span><span class="p">,</span> <span class="n">n_filters</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="n">size</span><span class="p">)</span> <span class="k">for</span> <span class="n">size</span> <span class="ow">in</span> <span class="n">filter_sizes</span>
<span class="p">])</span>
<span class="bp">self</span><span class="o">.</span><span class="n">fc</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="nb">len</span><span class="p">(</span><span class="n">filter_sizes</span><span class="p">)</span> <span class="o">*</span> <span class="n">n_filters</span><span class="p">,</span> <span class="n">output_dim</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">dropout</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Dropout</span><span class="p">(</span><span class="n">dropout</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">x</span><span class="p">):</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">rearrange</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="s1">'t b -> t b'</span><span class="p">)</span>
<span class="n">emb</span> <span class="o">=</span> <span class="n">rearrange</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">embedding</span><span class="p">(</span><span class="n">x</span><span class="p">),</span> <span class="s1">'t b c -> b c t'</span><span class="p">)</span>
<span class="n">pooled</span> <span class="o">=</span> <span class="p">[</span><span class="n">reduce</span><span class="p">(</span><span class="n">conv</span><span class="p">(</span><span class="n">emb</span><span class="p">),</span> <span class="s1">'b c t -> b c'</span><span class="p">,</span> <span class="s1">'max'</span><span class="p">)</span> <span class="k">for</span> <span class="n">conv</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">convs</span><span class="p">]</span>
<span class="n">concatenated</span> <span class="o">=</span> <span class="n">rearrange</span><span class="p">(</span><span class="n">pooled</span><span class="p">,</span> <span class="s1">'filter b c -> b (filter c)'</span><span class="p">)</span>
<span class="k">return</span> <span class="bp">self</span><span class="o">.</span><span class="n">fc</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">dropout</span><span class="p">(</span><span class="n">F</span><span class="o">.</span><span class="n">relu</span><span class="p">(</span><span class="n">concatenated</span><span class="p">)))</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><ul>
<li>Original code misuses Conv2d, while Conv1d is the right choice</li>
<li>Fixed code can work with any number of filter_sizes (and won't fail)</li>
<li>First line in new code does nothing, but was added for simplicity</li>
</ul></div><div class='markdown-cell'><h1>Highway convolutions</h1>
<ul>
<li>Highway convolutions are common in TTS systems. Code below makes splitting a bit more explicit.</li>
<li>Splitting policy may eventually turn out to be important if input had previously groups over channel axes (group convolutions or bidirectional LSTMs/GRUs)</li>
<li>Same applies to GLU and gated units in general</li>
</ul></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">HighwayConv1dOld</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Conv1d</span><span class="p">):</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">inputs</span><span class="p">):</span>
<span class="n">L</span> <span class="o">=</span> <span class="nb">super</span><span class="p">(</span><span class="n">HighwayConv1dOld</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="n">forward</span><span class="p">(</span><span class="n">inputs</span><span class="p">)</span>
<span class="n">H1</span><span class="p">,</span> <span class="n">H2</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">chunk</span><span class="p">(</span><span class="n">L</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">1</span><span class="p">)</span> <span class="c1"># chunk at the feature dim</span>
<span class="n">torch</span><span class="o">.</span><span class="n">sigmoid_</span><span class="p">(</span><span class="n">H1</span><span class="p">)</span>
<span class="k">return</span> <span class="n">H1</span> <span class="o">*</span> <span class="n">H2</span> <span class="o">+</span> <span class="p">(</span><span class="mf">1.0</span> <span class="o">-</span> <span class="n">H1</span><span class="p">)</span> <span class="o">*</span> <span class="n">inputs</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">HighwayConv1dNew</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Conv1d</span><span class="p">):</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">inputs</span><span class="p">):</span>
<span class="n">L</span> <span class="o">=</span> <span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="n">forward</span><span class="p">(</span><span class="n">inputs</span><span class="p">)</span>
<span class="n">H1</span><span class="p">,</span> <span class="n">H2</span> <span class="o">=</span> <span class="n">rearrange</span><span class="p">(</span><span class="n">L</span><span class="p">,</span> <span class="s1">'b (split c) t -> split b c t'</span><span class="p">,</span> <span class="n">split</span><span class="o">=</span><span class="mi">2</span><span class="p">)</span>
<span class="n">torch</span><span class="o">.</span><span class="n">sigmoid_</span><span class="p">(</span><span class="n">H1</span><span class="p">)</span>
<span class="k">return</span> <span class="n">H1</span> <span class="o">*</span> <span class="n">H2</span> <span class="o">+</span> <span class="p">(</span><span class="mf">1.0</span> <span class="o">-</span> <span class="n">H1</span><span class="p">)</span> <span class="o">*</span> <span class="n">inputs</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><h1>Tacotron's CBHG module</h1>
<!-- https://github.com/r9y9/tacotron_pytorch/blob/master/tacotron_pytorch/tacotron.py --></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">CBHG_Old</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="sd">"""CBHG module: a recurrent neural network composed of:</span>
<span class="sd"> - 1-d convolution banks</span>
<span class="sd"> - Highway networks + residual connections</span>
<span class="sd"> - Bidirectional gated recurrent units</span>
<span class="sd"> """</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">in_dim</span><span class="p">,</span> <span class="n">K</span><span class="o">=</span><span class="mi">16</span><span class="p">,</span> <span class="n">projections</span><span class="o">=</span><span class="p">[</span><span class="mi">128</span><span class="p">,</span> <span class="mi">128</span><span class="p">]):</span>
<span class="nb">super</span><span class="p">(</span><span class="n">CBHG</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">in_dim</span> <span class="o">=</span> <span class="n">in_dim</span>
<span class="bp">self</span><span class="o">.</span><span class="n">relu</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">ReLU</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv1d_banks</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">ModuleList</span><span class="p">(</span>
<span class="p">[</span><span class="n">BatchNormConv1d</span><span class="p">(</span><span class="n">in_dim</span><span class="p">,</span> <span class="n">in_dim</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="n">k</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
<span class="n">padding</span><span class="o">=</span><span class="n">k</span> <span class="o">//</span> <span class="mi">2</span><span class="p">,</span> <span class="n">activation</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">relu</span><span class="p">)</span>
<span class="k">for</span> <span class="n">k</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">K</span> <span class="o">+</span> <span class="mi">1</span><span class="p">)])</span>
<span class="bp">self</span><span class="o">.</span><span class="n">max_pool1d</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">MaxPool1d</span><span class="p">(</span><span class="n">kernel_size</span><span class="o">=</span><span class="mi">2</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
<span class="n">in_sizes</span> <span class="o">=</span> <span class="p">[</span><span class="n">K</span> <span class="o">*</span> <span class="n">in_dim</span><span class="p">]</span> <span class="o">+</span> <span class="n">projections</span><span class="p">[:</span><span class="o">-</span><span class="mi">1</span><span class="p">]</span>
<span class="n">activations</span> <span class="o">=</span> <span class="p">[</span><span class="bp">self</span><span class="o">.</span><span class="n">relu</span><span class="p">]</span> <span class="o">*</span> <span class="p">(</span><span class="nb">len</span><span class="p">(</span><span class="n">projections</span><span class="p">)</span> <span class="o">-</span> <span class="mi">1</span><span class="p">)</span> <span class="o">+</span> <span class="p">[</span><span class="kc">None</span><span class="p">]</span>
<span class="bp">self</span><span class="o">.</span><span class="n">conv1d_projections</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">ModuleList</span><span class="p">(</span>
<span class="p">[</span><span class="n">BatchNormConv1d</span><span class="p">(</span><span class="n">in_size</span><span class="p">,</span> <span class="n">out_size</span><span class="p">,</span> <span class="n">kernel_size</span><span class="o">=</span><span class="mi">3</span><span class="p">,</span> <span class="n">stride</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span>
<span class="n">padding</span><span class="o">=</span><span class="mi">1</span><span class="p">,</span> <span class="n">activation</span><span class="o">=</span><span class="n">ac</span><span class="p">)</span>
<span class="k">for</span> <span class="p">(</span><span class="n">in_size</span><span class="p">,</span> <span class="n">out_size</span><span class="p">,</span> <span class="n">ac</span><span class="p">)</span> <span class="ow">in</span> <span class="nb">zip</span><span class="p">(</span>
<span class="n">in_sizes</span><span class="p">,</span> <span class="n">projections</span><span class="p">,</span> <span class="n">activations</span><span class="p">)])</span>
<span class="bp">self</span><span class="o">.</span><span class="n">pre_highway</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">projections</span><span class="p">[</span><span class="o">-</span><span class="mi">1</span><span class="p">],</span> <span class="n">in_dim</span><span class="p">,</span> <span class="n">bias</span><span class="o">=</span><span class="kc">False</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">highways</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">ModuleList</span><span class="p">(</span>
<span class="p">[</span><span class="n">Highway</span><span class="p">(</span><span class="n">in_dim</span><span class="p">,</span> <span class="n">in_dim</span><span class="p">)</span> <span class="k">for</span> <span class="n">_</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">4</span><span class="p">)])</span>
<span class="bp">self</span><span class="o">.</span><span class="n">gru</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">GRU</span><span class="p">(</span>
<span class="n">in_dim</span><span class="p">,</span> <span class="n">in_dim</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="n">batch_first</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span> <span class="n">bidirectional</span><span class="o">=</span><span class="kc">True</span><span class="p">)</span>
</pre></div>
</div></div></div> <div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">forward_old</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">inputs</span><span class="p">):</span>
<span class="c1"># (B, T_in, in_dim)</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">inputs</span>
<span class="c1"># Needed to perform conv1d on time-axis</span>
<span class="c1"># (B, in_dim, T_in)</span>
<span class="k">if</span> <span class="n">x</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="o">-</span><span class="mi">1</span><span class="p">)</span> <span class="o">==</span> <span class="bp">self</span><span class="o">.</span><span class="n">in_dim</span><span class="p">:</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="n">T</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="o">-</span><span class="mi">1</span><span class="p">)</span>
<span class="c1"># (B, in_dim*K, T_in)</span>
<span class="c1"># Concat conv1d bank outputs</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">cat</span><span class="p">([</span><span class="n">conv1d</span><span class="p">(</span><span class="n">x</span><span class="p">)[:,</span> <span class="p">:,</span> <span class="p">:</span><span class="n">T</span><span class="p">]</span> <span class="k">for</span> <span class="n">conv1d</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">conv1d_banks</span><span class="p">],</span> <span class="n">dim</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
<span class="k">assert</span> <span class="n">x</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span> <span class="o">==</span> <span class="bp">self</span><span class="o">.</span><span class="n">in_dim</span> <span class="o">*</span> <span class="nb">len</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">conv1d_banks</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">max_pool1d</span><span class="p">(</span><span class="n">x</span><span class="p">)[:,</span> <span class="p">:,</span> <span class="p">:</span><span class="n">T</span><span class="p">]</span>
<span class="k">for</span> <span class="n">conv1d</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">conv1d_projections</span><span class="p">:</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">conv1d</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="c1"># (B, T_in, in_dim)</span>
<span class="c1"># Back to the original shape</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">)</span>
<span class="k">if</span> <span class="n">x</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="o">-</span><span class="mi">1</span><span class="p">)</span> <span class="o">!=</span> <span class="bp">self</span><span class="o">.</span><span class="n">in_dim</span><span class="p">:</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">pre_highway</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="c1"># Residual connection</span>
<span class="n">x</span> <span class="o">+=</span> <span class="n">inputs</span>
<span class="k">for</span> <span class="n">highway</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">highways</span><span class="p">:</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">highway</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="c1"># (B, T_in, in_dim*2)</span>
<span class="n">outputs</span><span class="p">,</span> <span class="n">_</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">gru</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="k">return</span> <span class="n">outputs</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">forward_new</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">inputs</span><span class="p">,</span> <span class="n">input_lengths</span><span class="o">=</span><span class="kc">None</span><span class="p">):</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">rearrange</span><span class="p">(</span><span class="n">inputs</span><span class="p">,</span> <span class="s1">'b t c -> b c t'</span><span class="p">)</span>
<span class="n">_</span><span class="p">,</span> <span class="n">_</span><span class="p">,</span> <span class="n">T</span> <span class="o">=</span> <span class="n">x</span><span class="o">.</span><span class="n">shape</span>
<span class="c1"># Concat conv1d bank outputs</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">rearrange</span><span class="p">([</span><span class="n">conv1d</span><span class="p">(</span><span class="n">x</span><span class="p">)[:,</span> <span class="p">:,</span> <span class="p">:</span><span class="n">T</span><span class="p">]</span> <span class="k">for</span> <span class="n">conv1d</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">conv1d_banks</span><span class="p">],</span>
<span class="s1">'bank b c t -> b (bank c) t'</span><span class="p">,</span> <span class="n">c</span><span class="o">=</span><span class="bp">self</span><span class="o">.</span><span class="n">in_dim</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">max_pool1d</span><span class="p">(</span><span class="n">x</span><span class="p">)[:,</span> <span class="p">:,</span> <span class="p">:</span><span class="n">T</span><span class="p">]</span>
<span class="k">for</span> <span class="n">conv1d</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">conv1d_projections</span><span class="p">:</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">conv1d</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">rearrange</span><span class="p">(</span><span class="n">x</span><span class="p">,</span> <span class="s1">'b c t -> b t c'</span><span class="p">)</span>
<span class="k">if</span> <span class="n">x</span><span class="o">.</span><span class="n">size</span><span class="p">(</span><span class="o">-</span><span class="mi">1</span><span class="p">)</span> <span class="o">!=</span> <span class="bp">self</span><span class="o">.</span><span class="n">in_dim</span><span class="p">:</span>
<span class="n">x</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">pre_highway</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="c1"># Residual connection</span>
<span class="n">x</span> <span class="o">+=</span> <span class="n">inputs</span>
<span class="k">for</span> <span class="n">highway</span> <span class="ow">in</span> <span class="bp">self</span><span class="o">.</span><span class="n">highways</span><span class="p">:</span>
<span class="n">x</span> <span class="o">=</span> <span class="n">highway</span><span class="p">(</span><span class="n">x</span><span class="p">)</span>
<span class="c1"># (B, T_in, in_dim*2)</span>
<span class="n">outputs</span><span class="p">,</span> <span class="n">_</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">gru</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">highways</span><span class="p">(</span><span class="n">x</span><span class="p">))</span>
<span class="k">return</span> <span class="n">outputs</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><p>There is still a large room for improvements, but in this example only forward function was changed</p></div><div class='markdown-cell'><h1>Simple attention</h1>
<p>Good news: there is no more need to guess order of dimensions. Neither for inputs nor for outputs</p></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">Attention</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="nb">super</span><span class="p">(</span><span class="n">Attention</span><span class="p">,</span> <span class="bp">self</span><span class="p">)</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">K</span><span class="p">,</span> <span class="n">V</span><span class="p">,</span> <span class="n">Q</span><span class="p">):</span>
<span class="n">A</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">bmm</span><span class="p">(</span><span class="n">K</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span><span class="mi">2</span><span class="p">),</span> <span class="n">Q</span><span class="p">)</span> <span class="o">/</span> <span class="n">np</span><span class="o">.</span><span class="n">sqrt</span><span class="p">(</span><span class="n">Q</span><span class="o">.</span><span class="n">shape</span><span class="p">[</span><span class="mi">1</span><span class="p">])</span>
<span class="n">A</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">softmax</span><span class="p">(</span><span class="n">A</span><span class="p">,</span> <span class="mi">1</span><span class="p">)</span>
<span class="n">R</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">bmm</span><span class="p">(</span><span class="n">V</span><span class="p">,</span> <span class="n">A</span><span class="p">)</span>
<span class="k">return</span> <span class="n">torch</span><span class="o">.</span><span class="n">cat</span><span class="p">((</span><span class="n">R</span><span class="p">,</span> <span class="n">Q</span><span class="p">),</span> <span class="n">dim</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">attention</span><span class="p">(</span><span class="n">K</span><span class="p">,</span> <span class="n">V</span><span class="p">,</span> <span class="n">Q</span><span class="p">):</span>
<span class="n">_</span><span class="p">,</span> <span class="n">n_channels</span><span class="p">,</span> <span class="n">_</span> <span class="o">=</span> <span class="n">K</span><span class="o">.</span><span class="n">shape</span>
<span class="n">A</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">einsum</span><span class="p">(</span><span class="s1">'bct,bcl->btl'</span><span class="p">,</span> <span class="p">[</span><span class="n">K</span><span class="p">,</span> <span class="n">Q</span><span class="p">])</span>
<span class="n">A</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">softmax</span><span class="p">(</span><span class="n">A</span> <span class="o">*</span> <span class="n">n_channels</span> <span class="o">**</span> <span class="p">(</span><span class="o">-</span><span class="mf">0.5</span><span class="p">),</span> <span class="mi">1</span><span class="p">)</span>
<span class="n">R</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">einsum</span><span class="p">(</span><span class="s1">'bct,btl->bcl'</span><span class="p">,</span> <span class="p">[</span><span class="n">V</span><span class="p">,</span> <span class="n">A</span><span class="p">])</span>
<span class="k">return</span> <span class="n">torch</span><span class="o">.</span><span class="n">cat</span><span class="p">((</span><span class="n">R</span><span class="p">,</span> <span class="n">Q</span><span class="p">),</span> <span class="n">dim</span><span class="o">=</span><span class="mi">1</span><span class="p">)</span>
</pre></div>
</div></div></div> <div class='markdown-cell'><h1>Transformer's attention needs more attention</h1></div><div class='leftright-wrapper'><div class='leftright-cells'><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">ScaledDotProductAttention</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="sd">''' Scaled Dot-Product Attention '''</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">temperature</span><span class="p">,</span> <span class="n">attn_dropout</span><span class="o">=</span><span class="mf">0.1</span><span class="p">):</span>
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">temperature</span> <span class="o">=</span> <span class="n">temperature</span>
<span class="bp">self</span><span class="o">.</span><span class="n">dropout</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Dropout</span><span class="p">(</span><span class="n">attn_dropout</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">softmax</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Softmax</span><span class="p">(</span><span class="n">dim</span><span class="o">=</span><span class="mi">2</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">q</span><span class="p">,</span> <span class="n">k</span><span class="p">,</span> <span class="n">v</span><span class="p">,</span> <span class="n">mask</span><span class="o">=</span><span class="kc">None</span><span class="p">):</span>
<span class="n">attn</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">bmm</span><span class="p">(</span><span class="n">q</span><span class="p">,</span> <span class="n">k</span><span class="o">.</span><span class="n">transpose</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">))</span>
<span class="n">attn</span> <span class="o">=</span> <span class="n">attn</span> <span class="o">/</span> <span class="bp">self</span><span class="o">.</span><span class="n">temperature</span>
<span class="k">if</span> <span class="n">mask</span> <span class="ow">is</span> <span class="ow">not</span> <span class="kc">None</span><span class="p">:</span>
<span class="n">attn</span> <span class="o">=</span> <span class="n">attn</span><span class="o">.</span><span class="n">masked_fill</span><span class="p">(</span><span class="n">mask</span><span class="p">,</span> <span class="o">-</span><span class="n">np</span><span class="o">.</span><span class="n">inf</span><span class="p">)</span>
<span class="n">attn</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">softmax</span><span class="p">(</span><span class="n">attn</span><span class="p">)</span>
<span class="n">attn</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">dropout</span><span class="p">(</span><span class="n">attn</span><span class="p">)</span>
<span class="n">output</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">bmm</span><span class="p">(</span><span class="n">attn</span><span class="p">,</span> <span class="n">v</span><span class="p">)</span>
<span class="k">return</span> <span class="n">output</span><span class="p">,</span> <span class="n">attn</span>
<span class="k">class</span> <span class="nc">MultiHeadAttentionOld</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="sd">''' Multi-Head Attention module '''</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">n_head</span><span class="p">,</span> <span class="n">d_model</span><span class="p">,</span> <span class="n">d_k</span><span class="p">,</span> <span class="n">d_v</span><span class="p">,</span> <span class="n">dropout</span><span class="o">=</span><span class="mf">0.1</span><span class="p">):</span>
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">n_head</span> <span class="o">=</span> <span class="n">n_head</span>
<span class="bp">self</span><span class="o">.</span><span class="n">d_k</span> <span class="o">=</span> <span class="n">d_k</span>
<span class="bp">self</span><span class="o">.</span><span class="n">d_v</span> <span class="o">=</span> <span class="n">d_v</span>
<span class="bp">self</span><span class="o">.</span><span class="n">w_qs</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">d_model</span><span class="p">,</span> <span class="n">n_head</span> <span class="o">*</span> <span class="n">d_k</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">w_ks</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">d_model</span><span class="p">,</span> <span class="n">n_head</span> <span class="o">*</span> <span class="n">d_k</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">w_vs</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">d_model</span><span class="p">,</span> <span class="n">n_head</span> <span class="o">*</span> <span class="n">d_v</span><span class="p">)</span>
<span class="n">nn</span><span class="o">.</span><span class="n">init</span><span class="o">.</span><span class="n">normal_</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">w_qs</span><span class="o">.</span><span class="n">weight</span><span class="p">,</span> <span class="n">mean</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="n">std</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">sqrt</span><span class="p">(</span><span class="mf">2.0</span> <span class="o">/</span> <span class="p">(</span><span class="n">d_model</span> <span class="o">+</span> <span class="n">d_k</span><span class="p">)))</span>
<span class="n">nn</span><span class="o">.</span><span class="n">init</span><span class="o">.</span><span class="n">normal_</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">w_ks</span><span class="o">.</span><span class="n">weight</span><span class="p">,</span> <span class="n">mean</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="n">std</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">sqrt</span><span class="p">(</span><span class="mf">2.0</span> <span class="o">/</span> <span class="p">(</span><span class="n">d_model</span> <span class="o">+</span> <span class="n">d_k</span><span class="p">)))</span>
<span class="n">nn</span><span class="o">.</span><span class="n">init</span><span class="o">.</span><span class="n">normal_</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">w_vs</span><span class="o">.</span><span class="n">weight</span><span class="p">,</span> <span class="n">mean</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="n">std</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">sqrt</span><span class="p">(</span><span class="mf">2.0</span> <span class="o">/</span> <span class="p">(</span><span class="n">d_model</span> <span class="o">+</span> <span class="n">d_v</span><span class="p">)))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">attention</span> <span class="o">=</span> <span class="n">ScaledDotProductAttention</span><span class="p">(</span><span class="n">temperature</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">power</span><span class="p">(</span><span class="n">d_k</span><span class="p">,</span> <span class="mf">0.5</span><span class="p">))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">layer_norm</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">LayerNorm</span><span class="p">(</span><span class="n">d_model</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">fc</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">n_head</span> <span class="o">*</span> <span class="n">d_v</span><span class="p">,</span> <span class="n">d_model</span><span class="p">)</span>
<span class="n">nn</span><span class="o">.</span><span class="n">init</span><span class="o">.</span><span class="n">xavier_normal_</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">fc</span><span class="o">.</span><span class="n">weight</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">dropout</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Dropout</span><span class="p">(</span><span class="n">dropout</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">q</span><span class="p">,</span> <span class="n">k</span><span class="p">,</span> <span class="n">v</span><span class="p">,</span> <span class="n">mask</span><span class="o">=</span><span class="kc">None</span><span class="p">):</span>
<span class="n">d_k</span><span class="p">,</span> <span class="n">d_v</span><span class="p">,</span> <span class="n">n_head</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">d_k</span><span class="p">,</span> <span class="bp">self</span><span class="o">.</span><span class="n">d_v</span><span class="p">,</span> <span class="bp">self</span><span class="o">.</span><span class="n">n_head</span>
<span class="n">sz_b</span><span class="p">,</span> <span class="n">len_q</span><span class="p">,</span> <span class="n">_</span> <span class="o">=</span> <span class="n">q</span><span class="o">.</span><span class="n">size</span><span class="p">()</span>
<span class="n">sz_b</span><span class="p">,</span> <span class="n">len_k</span><span class="p">,</span> <span class="n">_</span> <span class="o">=</span> <span class="n">k</span><span class="o">.</span><span class="n">size</span><span class="p">()</span>
<span class="n">sz_b</span><span class="p">,</span> <span class="n">len_v</span><span class="p">,</span> <span class="n">_</span> <span class="o">=</span> <span class="n">v</span><span class="o">.</span><span class="n">size</span><span class="p">()</span>
<span class="n">residual</span> <span class="o">=</span> <span class="n">q</span>
<span class="n">q</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">w_qs</span><span class="p">(</span><span class="n">q</span><span class="p">)</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">sz_b</span><span class="p">,</span> <span class="n">len_q</span><span class="p">,</span> <span class="n">n_head</span><span class="p">,</span> <span class="n">d_k</span><span class="p">)</span>
<span class="n">k</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">w_ks</span><span class="p">(</span><span class="n">k</span><span class="p">)</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">sz_b</span><span class="p">,</span> <span class="n">len_k</span><span class="p">,</span> <span class="n">n_head</span><span class="p">,</span> <span class="n">d_k</span><span class="p">)</span>
<span class="n">v</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">w_vs</span><span class="p">(</span><span class="n">v</span><span class="p">)</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">sz_b</span><span class="p">,</span> <span class="n">len_v</span><span class="p">,</span> <span class="n">n_head</span><span class="p">,</span> <span class="n">d_v</span><span class="p">)</span>
<span class="n">q</span> <span class="o">=</span> <span class="n">q</span><span class="o">.</span><span class="n">permute</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span><span class="o">.</span><span class="n">contiguous</span><span class="p">()</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="n">len_q</span><span class="p">,</span> <span class="n">d_k</span><span class="p">)</span> <span class="c1"># (n*b) x lq x dk</span>
<span class="n">k</span> <span class="o">=</span> <span class="n">k</span><span class="o">.</span><span class="n">permute</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span><span class="o">.</span><span class="n">contiguous</span><span class="p">()</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="n">len_k</span><span class="p">,</span> <span class="n">d_k</span><span class="p">)</span> <span class="c1"># (n*b) x lk x dk</span>
<span class="n">v</span> <span class="o">=</span> <span class="n">v</span><span class="o">.</span><span class="n">permute</span><span class="p">(</span><span class="mi">2</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span><span class="o">.</span><span class="n">contiguous</span><span class="p">()</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="o">-</span><span class="mi">1</span><span class="p">,</span> <span class="n">len_v</span><span class="p">,</span> <span class="n">d_v</span><span class="p">)</span> <span class="c1"># (n*b) x lv x dv</span>
<span class="n">mask</span> <span class="o">=</span> <span class="n">mask</span><span class="o">.</span><span class="n">repeat</span><span class="p">(</span><span class="n">n_head</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">1</span><span class="p">)</span> <span class="c1"># (n*b) x .. x ..</span>
<span class="n">output</span><span class="p">,</span> <span class="n">attn</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">attention</span><span class="p">(</span><span class="n">q</span><span class="p">,</span> <span class="n">k</span><span class="p">,</span> <span class="n">v</span><span class="p">,</span> <span class="n">mask</span><span class="o">=</span><span class="n">mask</span><span class="p">)</span>
<span class="n">output</span> <span class="o">=</span> <span class="n">output</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">n_head</span><span class="p">,</span> <span class="n">sz_b</span><span class="p">,</span> <span class="n">len_q</span><span class="p">,</span> <span class="n">d_v</span><span class="p">)</span>
<span class="n">output</span> <span class="o">=</span> <span class="n">output</span><span class="o">.</span><span class="n">permute</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">3</span><span class="p">)</span><span class="o">.</span><span class="n">contiguous</span><span class="p">()</span><span class="o">.</span><span class="n">view</span><span class="p">(</span><span class="n">sz_b</span><span class="p">,</span> <span class="n">len_q</span><span class="p">,</span> <span class="o">-</span><span class="mi">1</span><span class="p">)</span> <span class="c1"># b x lq x (n*dv)</span>
<span class="n">output</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">dropout</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">fc</span><span class="p">(</span><span class="n">output</span><span class="p">))</span>
<span class="n">output</span> <span class="o">=</span> <span class="bp">self</span><span class="o">.</span><span class="n">layer_norm</span><span class="p">(</span><span class="n">output</span> <span class="o">+</span> <span class="n">residual</span><span class="p">)</span>
<span class="k">return</span> <span class="n">output</span><span class="p">,</span> <span class="n">attn</span>
</pre></div>
</div><div><div class="highlight"><pre><span></span><span class="k">class</span> <span class="nc">MultiHeadAttentionNew</span><span class="p">(</span><span class="n">nn</span><span class="o">.</span><span class="n">Module</span><span class="p">):</span>
<span class="k">def</span> <span class="fm">__init__</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">n_head</span><span class="p">,</span> <span class="n">d_model</span><span class="p">,</span> <span class="n">d_k</span><span class="p">,</span> <span class="n">d_v</span><span class="p">,</span> <span class="n">dropout</span><span class="o">=</span><span class="mf">0.1</span><span class="p">):</span>
<span class="nb">super</span><span class="p">()</span><span class="o">.</span><span class="fm">__init__</span><span class="p">()</span>
<span class="bp">self</span><span class="o">.</span><span class="n">n_head</span> <span class="o">=</span> <span class="n">n_head</span>
<span class="bp">self</span><span class="o">.</span><span class="n">w_qs</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">d_model</span><span class="p">,</span> <span class="n">n_head</span> <span class="o">*</span> <span class="n">d_k</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">w_ks</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">d_model</span><span class="p">,</span> <span class="n">n_head</span> <span class="o">*</span> <span class="n">d_k</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">w_vs</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">d_model</span><span class="p">,</span> <span class="n">n_head</span> <span class="o">*</span> <span class="n">d_v</span><span class="p">)</span>
<span class="n">nn</span><span class="o">.</span><span class="n">init</span><span class="o">.</span><span class="n">normal_</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">w_qs</span><span class="o">.</span><span class="n">weight</span><span class="p">,</span> <span class="n">mean</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="n">std</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">sqrt</span><span class="p">(</span><span class="mf">2.0</span> <span class="o">/</span> <span class="p">(</span><span class="n">d_model</span> <span class="o">+</span> <span class="n">d_k</span><span class="p">)))</span>
<span class="n">nn</span><span class="o">.</span><span class="n">init</span><span class="o">.</span><span class="n">normal_</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">w_ks</span><span class="o">.</span><span class="n">weight</span><span class="p">,</span> <span class="n">mean</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="n">std</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">sqrt</span><span class="p">(</span><span class="mf">2.0</span> <span class="o">/</span> <span class="p">(</span><span class="n">d_model</span> <span class="o">+</span> <span class="n">d_k</span><span class="p">)))</span>
<span class="n">nn</span><span class="o">.</span><span class="n">init</span><span class="o">.</span><span class="n">normal_</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">w_vs</span><span class="o">.</span><span class="n">weight</span><span class="p">,</span> <span class="n">mean</span><span class="o">=</span><span class="mi">0</span><span class="p">,</span> <span class="n">std</span><span class="o">=</span><span class="n">np</span><span class="o">.</span><span class="n">sqrt</span><span class="p">(</span><span class="mf">2.0</span> <span class="o">/</span> <span class="p">(</span><span class="n">d_model</span> <span class="o">+</span> <span class="n">d_v</span><span class="p">)))</span>
<span class="bp">self</span><span class="o">.</span><span class="n">fc</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Linear</span><span class="p">(</span><span class="n">n_head</span> <span class="o">*</span> <span class="n">d_v</span><span class="p">,</span> <span class="n">d_model</span><span class="p">)</span>
<span class="n">nn</span><span class="o">.</span><span class="n">init</span><span class="o">.</span><span class="n">xavier_normal_</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">fc</span><span class="o">.</span><span class="n">weight</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">dropout</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">Dropout</span><span class="p">(</span><span class="n">p</span><span class="o">=</span><span class="n">dropout</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">layer_norm</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">LayerNorm</span><span class="p">(</span><span class="n">d_model</span><span class="p">)</span>
<span class="k">def</span> <span class="nf">forward</span><span class="p">(</span><span class="bp">self</span><span class="p">,</span> <span class="n">q</span><span class="p">,</span> <span class="n">k</span><span class="p">,</span> <span class="n">v</span><span class="p">,</span> <span class="n">mask</span><span class="o">=</span><span class="kc">None</span><span class="p">):</span>