-
Notifications
You must be signed in to change notification settings - Fork 258
/
Copy pathchar_rnn_classification_tutorial.html
1330 lines (1100 loc) ยท 106 KB
/
char_rnn_classification_tutorial.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="ko" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="ko" > <!--<![endif]-->
<head>
<meta charset="utf-8">
<meta property="og:title" content="๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: ๋ฌธ์-๋จ์ RNN์ผ๋ก ์ด๋ฆ ๋ถ๋ฅํ๊ธฐ" />
<meta property="og:type" content="article" />
<meta property="og:url" content="https://tutorials.pytorch.kr/intermediate/char_rnn_classification_tutorial.html" />
<meta property="og:site_name" content="PyTorch Tutorials KR" />
<meta property="og:description" content="Author: Sean Robertson, ๋ฒ์ญ: ํฉ์ฑ์, ๊น์ ํ,. ์ฌ๊ธฐ์์๋ ๋จ์ด๋ฅผ ๋ถ๋ฅํ๊ธฐ ์ํด ๊ธฐ์ด์ ์ธ ๋ฌธ์-๋จ์์ ์ํ ์ ๊ฒฝ๋ง(RNN, Recurrent Nueral Network)์ ๊ตฌ์ถํ๊ณ ํ์ตํ ์์ ์
๋๋ค. ์ด ํํ ๋ฆฌ์ผ ๋ฐ ์ดํ 2๊ฐ ํํ ๋ฆฌ์ผ์ธ ๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: ๋ฌธ์-๋จ์ RNN์ผ๋ก ์ด๋ฆ ์์ฑํ๊ธฐ ๋ฐ ๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: Sequence to Sequence ๋คํธ์ํฌ์ Attention์ ์ด์ฉํ ๋ฒ์ญ ์์๋ ์์ฐ์ด ์ฒ๋ฆฌ(NLP, Natural Language Processing) ๋ถ์ผ์์ ์ด๋ป๊ฒ ๋ฐ..." />
<meta property="og:image" content="https://tutorials.pytorch.kr/_static/logos/logo-kr-sm-dark.png" />
<meta property="og:image:alt" content="PyTorch Tutorials KR" />
<meta name="description" content="Author: Sean Robertson, ๋ฒ์ญ: ํฉ์ฑ์, ๊น์ ํ,. ์ฌ๊ธฐ์์๋ ๋จ์ด๋ฅผ ๋ถ๋ฅํ๊ธฐ ์ํด ๊ธฐ์ด์ ์ธ ๋ฌธ์-๋จ์์ ์ํ ์ ๊ฒฝ๋ง(RNN, Recurrent Nueral Network)์ ๊ตฌ์ถํ๊ณ ํ์ตํ ์์ ์
๋๋ค. ์ด ํํ ๋ฆฌ์ผ ๋ฐ ์ดํ 2๊ฐ ํํ ๋ฆฌ์ผ์ธ ๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: ๋ฌธ์-๋จ์ RNN์ผ๋ก ์ด๋ฆ ์์ฑํ๊ธฐ ๋ฐ ๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: Sequence to Sequence ๋คํธ์ํฌ์ Attention์ ์ด์ฉํ ๋ฒ์ญ ์์๋ ์์ฐ์ด ์ฒ๋ฆฌ(NLP, Natural Language Processing) ๋ถ์ผ์์ ์ด๋ป๊ฒ ๋ฐ..." />
<meta property="og:ignore_canonical" content="true" />
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: ๋ฌธ์-๋จ์ RNN์ผ๋ก ์ด๋ฆ ๋ถ๋ฅํ๊ธฐ — ํ์ดํ ์น ํ๊ตญ์ด ํํ ๋ฆฌ์ผ (PyTorch tutorials in Korean)</title>
<link rel="shortcut icon" href="../_static/favicon.ico"/>
<link rel="stylesheet" href="../_static/css/theme.css" type="text/css" />
<!-- <link rel="stylesheet" href="../_static/pygments.css" type="text/css" /> -->
<link rel="stylesheet" href="../_static/pygments.css" type="text/css" />
<link rel="stylesheet" href="../_static/css/theme.css" type="text/css" />
<link rel="stylesheet" href="../_static/copybutton.css" type="text/css" />
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/katex@0.16.10/dist/katex.min.css" type="text/css" />
<link rel="stylesheet" href="../_static/katex-math.css" type="text/css" />
<link rel="stylesheet" href="../_static/sg_gallery.css" type="text/css" />
<link rel="stylesheet" href="../_static/sg_gallery-binder.css" type="text/css" />
<link rel="stylesheet" href="../_static/sg_gallery-dataframe.css" type="text/css" />
<link rel="stylesheet" href="../_static/sg_gallery-rendered-html.css" type="text/css" />
<link rel="stylesheet" href="../_static/sphinx-design.5ea377869091fd0449014c60fc090103.min.css" type="text/css" />
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/katex@0.10.0-beta/dist/katex.min.css" type="text/css" />
<link rel="stylesheet" href="../_static/css/custom.css" type="text/css" />
<link rel="stylesheet" href="../_static/css/custom2.css" type="text/css" />
<link rel="index" title="์์ธ" href="../genindex.html" />
<link rel="search" title="๊ฒ์" href="../search.html" />
<link rel="next" title="๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: ๋ฌธ์-๋จ์ RNN์ผ๋ก ์ด๋ฆ ์์ฑํ๊ธฐ" href="char_rnn_generation_tutorial.html" />
<link rel="prev" title="Fast Transformer Inference with Better Transformer" href="../beginner/bettertransformer_tutorial.html" />
<script src="../_static/js/modernizr.min.js"></script>
<!-- Preload the theme fonts -->
<link rel="preload" href="../_static/fonts/FreightSans/freight-sans-book.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="../_static/fonts/FreightSans/freight-sans-medium.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="../_static/fonts/IBMPlexMono/IBMPlexMono-Medium.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="../_static/fonts/FreightSans/freight-sans-bold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="../_static/fonts/FreightSans/freight-sans-medium-italic.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="../_static/fonts/IBMPlexMono/IBMPlexMono-SemiBold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<!-- Preload the katex fonts -->
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Math-Italic.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Main-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Main-Bold.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Size1-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Size4-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Size2-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Size3-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="preload" href="https://cdn.jsdelivr.net/npm/katex@0.10.0/dist/fonts/KaTeX_Caligraphic-Regular.woff2" as="font" type="font/woff2" crossorigin="anonymous">
<link rel="stylesheet" href="https://use.fontawesome.com/releases/v5.15.2/css/all.css" integrity="sha384-vSIIfh2YWi9wW0r9iZe7RJPrKwp6bG+s9QZMoITbCckVJqGCCRhc+ccxNcdpHuYu" crossorigin="anonymous">
</head>
<div class="container-fluid header-holder tutorials-header" id="header-holder">
<div class="container">
<div class="header-container">
<a class="header-logo" href="https://pytorch.kr/" aria-label="PyTorch"></a>
<div class="main-menu">
<ul>
<li>
<a href="https://pytorch.kr/get-started">์์ํ๊ธฐ</a>
</li>
<li class="active">
<a href="https://tutorials.pytorch.kr/">ํํ ๋ฆฌ์ผ</a>
</li>
<li>
<a href="https://pytorch.kr/hub">ํ๋ธ</a>
</li>
<li>
<a href="https://discuss.pytorch.kr/">์ปค๋ฎค๋ํฐ</a>
</li>
</ul>
</div>
<a class="main-menu-open-button" href="#" data-behavior="open-mobile-menu"></a>
</div>
</div>
</div>
<body class="pytorch-body">
<div class="table-of-contents-link-wrapper">
<span>Table of Contents</span>
<a href="#" class="toggle-table-of-contents" data-behavior="toggle-table-of-contents"></a>
</div>
<nav data-toggle="wy-nav-shift" class="pytorch-left-menu" id="pytorch-left-menu">
<div class="pytorch-side-scroll">
<div class="pytorch-menu pytorch-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
<div class="pytorch-left-menu-search">
<div class="version">
2.3.1+cu121
</div>
<div role="search">
<form id="rtd-search-form" class="wy-form" action="../search.html" method="get">
<input type="text" name="q" placeholder="Search Tutorials" />
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
</div>
</div>
<p class="caption" role="heading"><span class="caption-text">ํ์ดํ ์น(PyTorch) ๋ ์ํผ</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../recipes/recipes_index.html">๋ชจ๋ ๋ ์ํผ ๋ณด๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../prototype/prototype_index.html">๋ชจ๋ ํ๋กํ ํ์
๋ ์ํผ ๋ณด๊ธฐ</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">ํ์ดํ ์น(PyTorch) ์์ํ๊ธฐ</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../beginner/basics/intro.html">ํ์ดํ ์น(PyTorch) ๊ธฐ๋ณธ ์ตํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/basics/quickstart_tutorial.html">๋น ๋ฅธ ์์(Quickstart)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/basics/tensorqs_tutorial.html">ํ
์(Tensor)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/basics/data_tutorial.html">Dataset๊ณผ DataLoader</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/basics/transforms_tutorial.html">๋ณํ(Transform)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/basics/buildmodel_tutorial.html">์ ๊ฒฝ๋ง ๋ชจ๋ธ ๊ตฌ์ฑํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/basics/autogradqs_tutorial.html"><code class="docutils literal notranslate"><span class="pre">torch.autograd</span></code>๋ฅผ ์ฌ์ฉํ ์๋ ๋ฏธ๋ถ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/basics/optimization_tutorial.html">๋ชจ๋ธ ๋งค๊ฐ๋ณ์ ์ต์ ํํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/basics/saveloadrun_tutorial.html">๋ชจ๋ธ ์ ์ฅํ๊ณ ๋ถ๋ฌ์ค๊ธฐ</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Introduction to PyTorch on YouTube</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../beginner/introyt.html">PyTorch ์๊ฐ - YouTube ์๋ฆฌ์ฆ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/introyt/introyt1_tutorial.html">PyTorch ์๊ฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/introyt/tensors_deeper_tutorial.html">Pytorch Tensor ์๊ฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/introyt/autogradyt_tutorial.html">The Fundamentals of Autograd</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/introyt/modelsyt_tutorial.html">Building Models with PyTorch</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/introyt/tensorboardyt_tutorial.html">PyTorch TensorBoard Support</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/introyt/trainingyt.html">Training with PyTorch</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/introyt/captumyt.html">Model Understanding with Captum</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">ํ์ดํ ์น(PyTorch) ๋ฐฐ์ฐ๊ธฐ</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../beginner/deep_learning_60min_blitz.html">PyTorch๋ก ๋ฅ๋ฌ๋ํ๊ธฐ: 60๋ถ๋ง์ ๋์ฅ๋ด๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/pytorch_with_examples.html">์์ ๋ก ๋ฐฐ์ฐ๋ ํ์ดํ ์น(PyTorch)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/nn_tutorial.html"><cite>torch.nn</cite> ์ด <em>์ค์ ๋ก</em> ๋ฌด์์ธ๊ฐ์?</a></li>
<li class="toctree-l1"><a class="reference internal" href="tensorboard_tutorial.html">TensorBoard๋ก ๋ชจ๋ธ, ๋ฐ์ดํฐ, ํ์ต ์๊ฐํํ๊ธฐ</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">์ด๋ฏธ์ง/๋น๋์ค</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="torchvision_tutorial.html">TorchVision Object Detection Finetuning Tutorial</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/transfer_learning_tutorial.html">์ปดํจํฐ ๋น์ (Vision)์ ์ํ ์ ์ดํ์ต(Transfer Learning)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/fgsm_tutorial.html">์ ๋์ ์์ ์์ฑ(Adversarial Example Generation)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/dcgan_faces_tutorial.html">DCGAN ํํ ๋ฆฌ์ผ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/vt_tutorial.html">๋ฐฐํฌ๋ฅผ ์ํด ๋น์ ํธ๋์คํฌ๋จธ(Vision Transformer) ๋ชจ๋ธ ์ต์ ํํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="tiatoolbox_tutorial.html">Whole Slide Image Classification Using PyTorch and TIAToolbox</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">์ค๋์ค</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../beginner/audio_io_tutorial.html">Audio I/O</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/audio_resampling_tutorial.html">Audio Resampling</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/audio_data_augmentation_tutorial.html">Audio Data Augmentation</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/audio_feature_extractions_tutorial.html">Audio Feature Extractions</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/audio_feature_augmentation_tutorial.html">Audio Feature Augmentation</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/audio_datasets_tutorial.html">Audio Datasets</a></li>
<li class="toctree-l1"><a class="reference internal" href="speech_recognition_pipeline_tutorial.html">Speech Recognition with Wav2Vec2</a></li>
<li class="toctree-l1"><a class="reference internal" href="text_to_speech_with_torchaudio.html">Text-to-speech with Tacotron2</a></li>
<li class="toctree-l1"><a class="reference internal" href="forced_alignment_with_torchaudio_tutorial.html">wav2vec2์ ์ด์ฉํ ๊ฐ์ ์ ๋ ฌ</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">ํ
์คํธ</span></p>
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="../beginner/bettertransformer_tutorial.html">Fast Transformer Inference with Better Transformer</a></li>
<li class="toctree-l1 current"><a class="current reference internal" href="#">๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: ๋ฌธ์-๋จ์ RNN์ผ๋ก ์ด๋ฆ ๋ถ๋ฅํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="char_rnn_generation_tutorial.html">๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: ๋ฌธ์-๋จ์ RNN์ผ๋ก ์ด๋ฆ ์์ฑํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="seq2seq_translation_tutorial.html">๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: Sequence to Sequence ๋คํธ์ํฌ์ Attention์ ์ด์ฉํ ๋ฒ์ญ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/text_sentiment_ngrams_tutorial.html">torchtext ๋ผ์ด๋ธ๋ฌ๋ฆฌ๋ก ํ
์คํธ ๋ถ๋ฅํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/translation_transformer.html"><code class="docutils literal notranslate"><span class="pre">nn.Transformer</span></code> ์ torchtext๋ก ์ธ์ด ๋ฒ์ญํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/torchtext_custom_dataset_tutorial.html">Preprocess custom text dataset using Torchtext</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">๋ฐฑ์๋</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../beginner/onnx/intro_onnx.html">Introduction to ONNX</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">๊ฐํํ์ต</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="reinforcement_q_learning.html">๊ฐํ ํ์ต (DQN) ํํ ๋ฆฌ์ผ</a></li>
<li class="toctree-l1"><a class="reference internal" href="reinforcement_ppo.html">Reinforcement Learning (PPO) with TorchRL Tutorial</a></li>
<li class="toctree-l1"><a class="reference internal" href="mario_rl_tutorial.html">๋ง๋ฆฌ์ค ๊ฒ์ RL ์์ด์ ํธ๋ก ํ์ตํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/pendulum.html">Pendulum: Writing your environment and transforms with TorchRL</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">PyTorch ๋ชจ๋ธ์ ํ๋ก๋์
ํ๊ฒฝ์ ๋ฐฐํฌํ๊ธฐ</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../beginner/onnx/intro_onnx.html">Introduction to ONNX</a></li>
<li class="toctree-l1"><a class="reference internal" href="flask_rest_api_tutorial.html">Flask๋ฅผ ์ฌ์ฉํ์ฌ Python์์ PyTorch๋ฅผ REST API๋ก ๋ฐฐํฌํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/Intro_to_TorchScript_tutorial.html">TorchScript ์๊ฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/cpp_export.html">C++์์ TorchScript ๋ชจ๋ธ ๋ก๋ฉํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/super_resolution_with_onnxruntime.html">(์ ํ) PyTorch ๋ชจ๋ธ์ ONNX์ผ๋ก ๋ณํํ๊ณ ONNX ๋ฐํ์์์ ์คํํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="realtime_rpi.html">Raspberry Pi 4 ์์ ์ค์๊ฐ ์ถ๋ก (Inference) (30fps!)</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">PyTorch ํ๋กํ์ผ๋ง</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../beginner/profiler.html">PyTorch ๋ชจ๋ ํ๋กํ์ผ๋งํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/hta_intro_tutorial.html">Introduction to Holistic Trace Analysis</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/hta_trace_diff_tutorial.html">Trace Diff using Holistic Trace Analysis</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Code Transforms with FX</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="fx_conv_bn_fuser.html">(๋ฒ ํ) FX์์ ํฉ์ฑ๊ณฑ/๋ฐฐ์น ์ ๊ทํ(Convolution/Batch Norm) ๊ฒฐํฉ๊ธฐ(Fuser) ๋ง๋ค๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="fx_profiling_tutorial.html">(beta) Building a Simple CPU Performance Profiler with FX</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">ํ๋ก ํธ์๋ API</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="memory_format_tutorial.html">(๋ฒ ํ) PyTorch๋ฅผ ์ฌ์ฉํ Channels Last ๋ฉ๋ชจ๋ฆฌ ํ์</a></li>
<li class="toctree-l1"><a class="reference internal" href="forward_ad_usage.html">Forward-mode Automatic Differentiation (Beta)</a></li>
<li class="toctree-l1"><a class="reference internal" href="jacobians_hessians.html">Jacobians, Hessians, hvp, vhp, and more: composing function transforms</a></li>
<li class="toctree-l1"><a class="reference internal" href="ensembling.html">๋ชจ๋ธ ์์๋ธ</a></li>
<li class="toctree-l1"><a class="reference internal" href="per_sample_grads.html">Per-sample-gradients</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/cpp_frontend.html">PyTorch C++ ํ๋ก ํธ์๋ ์ฌ์ฉํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/torch-script-parallelism.html">TorchScript์ ๋์ ๋ณ๋ ฌ ์ฒ๋ฆฌ(Dynamic Parallelism)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/cpp_autograd.html">C++ ํ๋ก ํธ์๋์ ์๋ ๋ฏธ๋ถ (autograd)</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">PyTorch ํ์ฅํ๊ธฐ</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="custom_function_double_backward_tutorial.html">Double Backward with Custom Functions</a></li>
<li class="toctree-l1"><a class="reference internal" href="custom_function_conv_bn_tutorial.html">Fusing Convolution and Batch Norm using Custom Function</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/cpp_extension.html">Custom C++ and CUDA Extensions</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/torch_script_custom_ops.html">Extending TorchScript with Custom C++ Operators</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/torch_script_custom_classes.html">์ปค์คํ
C++ ํด๋์ค๋ก TorchScript ํ์ฅํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/dispatcher.html">Registering a Dispatched Operator in C++</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/extend_dispatcher.html">Extending dispatcher for a new backend in C++</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/privateuseone.html">Facilitating New Backend Integration by PrivateUse1</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">๋ชจ๋ธ ์ต์ ํ</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../beginner/profiler.html">PyTorch ๋ชจ๋ ํ๋กํ์ผ๋งํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="tensorboard_profiler_tutorial.html">ํ
์๋ณด๋๋ฅผ ์ด์ฉํ ํ์ดํ ์น ํ๋กํ์ผ๋ฌ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/hyperparameter_tuning_tutorial.html">Ray Tune์ ์ฌ์ฉํ ํ์ดํผํ๋ผ๋ฏธํฐ ํ๋</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/vt_tutorial.html">๋ฐฐํฌ๋ฅผ ์ํด ๋น์ ํธ๋์คํฌ๋จธ(Vision Transformer) ๋ชจ๋ธ ์ต์ ํํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="parametrizations.html">Parametrizations Tutorial</a></li>
<li class="toctree-l1"><a class="reference internal" href="pruning_tutorial.html">๊ฐ์ง์น๊ธฐ ๊ธฐ๋ฒ(Pruning) ํํ ๋ฆฌ์ผ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/dynamic_quantization_tutorial.html">(๋ฒ ํ) LSTM ๊ธฐ๋ฐ ๋จ์ด ๋จ์ ์ธ์ด ๋ชจ๋ธ์ ๋์ ์์ํ</a></li>
<li class="toctree-l1"><a class="reference internal" href="dynamic_quantization_bert_tutorial.html">(๋ฒ ํ) BERT ๋ชจ๋ธ ๋์ ์์ํํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="quantized_transfer_learning_tutorial.html">(๋ฒ ํ) ์ปดํจํฐ ๋น์ ํํ ๋ฆฌ์ผ์ ์ํ ์์ํ๋ ์ ์ดํ์ต(Quantized Transfer Learning)</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/static_quantization_tutorial.html">(๋ฒ ํ) PyTorch์์ Eager Mode๋ฅผ ์ด์ฉํ ์ ์ ์์ํ</a></li>
<li class="toctree-l1"><a class="reference internal" href="torchserve_with_ipex.html">Grokking PyTorch Intel CPU performance from first principles</a></li>
<li class="toctree-l1"><a class="reference internal" href="torchserve_with_ipex_2.html">Grokking PyTorch Intel CPU performance from first principles (Part 2)</a></li>
<li class="toctree-l1"><a class="reference internal" href="nvfuser_intro_tutorial.html">Getting Started - Accelerate Your Scripts with nvFuser</a></li>
<li class="toctree-l1"><a class="reference internal" href="ax_multiobjective_nas_tutorial.html">Multi-Objective NAS with Ax</a></li>
<li class="toctree-l1"><a class="reference internal" href="torch_compile_tutorial.html">Introduction to <code class="docutils literal notranslate"><span class="pre">torch.compile</span></code></a></li>
<li class="toctree-l1"><a class="reference internal" href="inductor_debug_cpu.html">Inductor CPU backend debugging and profiling</a></li>
<li class="toctree-l1"><a class="reference internal" href="scaled_dot_product_attention_tutorial.html">(Beta) Scaled Dot Product Attention (SDPA)๋ก ๊ณ ์ฑ๋ฅ ํธ๋์คํฌ๋จธ(Transformers) ๊ตฌํํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="scaled_dot_product_attention_tutorial.html#torch-compile-sdpa"><code class="docutils literal notranslate"><span class="pre">torch.compile</span></code> ๊ณผ ํจ๊ป SDPA ์ฌ์ฉํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="scaled_dot_product_attention_tutorial.html#sdpa-atteition-bias">SDPA๋ฅผ <code class="docutils literal notranslate"><span class="pre">atteition.bias</span></code> ํ์ ํด๋์ค์ ์ฌ์ฉํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="scaled_dot_product_attention_tutorial.html#id8">๊ฒฐ๋ก </a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/knowledge_distillation_tutorial.html">Knowledge Distillation Tutorial</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">๋ณ๋ ฌ ๋ฐ ๋ถ์ฐ ํ์ต</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../distributed/home.html">Distributed and Parallel Training Tutorials</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/dist_overview.html">PyTorch Distributed Overview</a></li>
<li class="toctree-l1"><a class="reference internal" href="../beginner/ddp_series_intro.html">Distributed Data Parallel in PyTorch - Video Tutorials</a></li>
<li class="toctree-l1"><a class="reference internal" href="model_parallel_tutorial.html">๋จ์ผ ๋จธ์ ์ ์ฌ์ฉํ ๋ชจ๋ธ ๋ณ๋ ฌํ ๋ชจ๋ฒ ์ฌ๋ก</a></li>
<li class="toctree-l1"><a class="reference internal" href="ddp_tutorial.html">๋ถ์ฐ ๋ฐ์ดํฐ ๋ณ๋ ฌ ์ฒ๋ฆฌ ์์ํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="dist_tuto.html">PyTorch๋ก ๋ถ์ฐ ์ดํ๋ฆฌ์ผ์ด์
๊ฐ๋ฐํ๊ธฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="FSDP_tutorial.html">Getting Started with Fully Sharded Data Parallel(FSDP)</a></li>
<li class="toctree-l1"><a class="reference internal" href="FSDP_adavnced_tutorial.html">Advanced Model Training with Fully Sharded Data Parallel (FSDP)</a></li>
<li class="toctree-l1"><a class="reference internal" href="TP_tutorial.html">Large Scale Transformer model training with Tensor Parallel (TP)</a></li>
<li class="toctree-l1"><a class="reference internal" href="process_group_cpp_extension_tutorial.html">Cpp ํ์ฅ์ ์ฌ์ฉํ ํ๋ก์ธ์ค ๊ทธ๋ฃน ๋ฐฑ์๋ ์ฌ์ฉ์ ์ ์</a></li>
<li class="toctree-l1"><a class="reference internal" href="rpc_tutorial.html">Getting Started with Distributed RPC Framework</a></li>
<li class="toctree-l1"><a class="reference internal" href="rpc_param_server_tutorial.html">Implementing a Parameter Server Using Distributed RPC Framework</a></li>
<li class="toctree-l1"><a class="reference internal" href="dist_pipeline_parallel_tutorial.html">Distributed Pipeline Parallelism Using RPC</a></li>
<li class="toctree-l1"><a class="reference internal" href="rpc_async_execution.html">Implementing Batch RPC Processing Using Asynchronous Executions</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/rpc_ddp_tutorial.html">๋ถ์ฐ ๋ฐ์ดํฐ ๋ณ๋ ฌ(DDP)๊ณผ ๋ถ์ฐ RPC ํ๋ ์์ํฌ ๊ฒฐํฉ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/ddp_pipeline.html">๋ถ์ฐ ๋ฐ์ดํฐ ๋ณ๋ ฌ ์ฒ๋ฆฌ์ ๋ณ๋ ฌ ์ฒ๋ฆฌ ํ์ดํ๋ผ์ธ์ ์ฌ์ฉํ ํธ๋์คํฌ๋จธ ๋ชจ๋ธ ํ์ต</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/generic_join.html">Distributed Training with Uneven Inputs Using the Join Context Manager</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Edge with ExecuTorch</span></p>
<ul>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/executorch/stable/tutorials/export-to-executorch-tutorial.html">Exporting to ExecuTorch Tutorial</a></li>
<li class="toctree-l1"><a class="reference external" href=" https://pytorch.org/executorch/stable/running-a-model-cpp-tutorial.html">Running an ExecuTorch Model in C++ Tutorial</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/executorch/stable/tutorials/sdk-integration-tutorial.html">Using the ExecuTorch SDK to Profile a Model</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/executorch/stable/demo-apps-ios.html">Building an ExecuTorch iOS Demo App</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/executorch/stable/demo-apps-android.html">Building an ExecuTorch Android Demo App</a></li>
<li class="toctree-l1"><a class="reference external" href="https://pytorch.org/executorch/stable/examples-end-to-end-to-lower-model-to-delegate.html">Lowering a Model as a Delegate</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">์ถ์ฒ ์์คํ
</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="torchrec_tutorial.html">TorchRec ์๊ฐ</a></li>
<li class="toctree-l1"><a class="reference internal" href="../advanced/sharding.html">Exploring TorchRec sharding</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Multimodality</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../beginner/flava_finetuning_tutorial.html">TorchMultimodal ํํ ๋ฆฌ์ผ: FLAVA ๋ฏธ์ธ์กฐ์ </a></li>
</ul>
</div>
</div>
</nav>
<div class="pytorch-container">
<div class="pytorch-page-level-bar" id="pytorch-page-level-bar">
<div class="pytorch-breadcrumbs-wrapper">
<div role="navigation" aria-label="breadcrumbs navigation">
<ul class="pytorch-breadcrumbs">
<li>
<a href="../index.html">
Tutorials
</a> >
</li>
<li>๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: ๋ฌธ์-๋จ์ RNN์ผ๋ก ์ด๋ฆ ๋ถ๋ฅํ๊ธฐ</li>
<li class="pytorch-breadcrumbs-aside">
<a href="../_sources/intermediate/char_rnn_classification_tutorial.rst.txt" rel="nofollow"><img src="../_static/images/view-page-source-icon.svg"></a>
</li>
</ul>
</div>
</div>
<div class="pytorch-shortcuts-wrapper" id="pytorch-shortcuts-wrapper">
Shortcuts
</div>
</div>
<section data-toggle="wy-nav-shift" id="pytorch-content-wrap" class="pytorch-content-wrap">
<div class="pytorch-content-left">
<div class="pytorch-call-to-action-links">
<div id="tutorial-type">intermediate/char_rnn_classification_tutorial</div>
<div id="google-colab-link">
<img class="call-to-action-img" src="../_static/images/pytorch-colab.svg"/>
<div class="call-to-action-desktop-view">Run in Google Colab</div>
<div class="call-to-action-mobile-view">Colab</div>
</div>
<div id="download-notebook-link">
<img class="call-to-action-notebook-img" src="../_static/images/pytorch-download.svg"/>
<div class="call-to-action-desktop-view">Download Notebook</div>
<div class="call-to-action-mobile-view">Notebook</div>
</div>
<div id="github-view-link">
<img class="call-to-action-img" src="../_static/images/pytorch-github.svg"/>
<div class="call-to-action-desktop-view">View on GitHub</div>
<div class="call-to-action-mobile-view">GitHub</div>
</div>
</div>
<div class="rst-content">
<div role="main" class="main-content" itemscope="itemscope" itemtype="http://schema.org/Article">
<article itemprop="articleBody" id="pytorch-article" class="pytorch-article">
<div class="sphx-glr-download-link-note admonition note">
<p class="admonition-title">์ฐธ๊ณ </p>
<p>Click <a class="reference internal" href="#sphx-glr-download-intermediate-char-rnn-classification-tutorial-py"><span class="std std-ref">here</span></a>
to download the full example code</p>
</div>
<div class="sphx-glr-example-title section" id="nlp-rnn">
<span id="sphx-glr-intermediate-char-rnn-classification-tutorial-py"></span><h1>๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: ๋ฌธ์-๋จ์ RNN์ผ๋ก ์ด๋ฆ ๋ถ๋ฅํ๊ธฐ<a class="headerlink" href="#nlp-rnn" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h1>
<dl class="simple">
<dt><strong>Author</strong>: <a class="reference external" href="https://github.com/spro">Sean Robertson</a></dt><dd><p><strong>๋ฒ์ญ</strong>: <a class="reference external" href="https://github.com/adonisues">ํฉ์ฑ์</a>, <a class="reference external" href="https://github.com/garlicvread">๊น์ ํ</a></p>
</dd>
</dl>
<p>์ฌ๊ธฐ์์๋ ๋จ์ด๋ฅผ ๋ถ๋ฅํ๊ธฐ ์ํด ๊ธฐ์ด์ ์ธ ๋ฌธ์-๋จ์์ ์ํ ์ ๊ฒฝ๋ง(RNN, Recurrent Nueral Network)์
๊ตฌ์ถํ๊ณ ํ์ตํ ์์ ์
๋๋ค. ์ด ํํ ๋ฆฌ์ผ ๋ฐ ์ดํ 2๊ฐ ํํ ๋ฆฌ์ผ์ธ <a class="reference internal" href="char_rnn_generation_tutorial.html"><span class="doc">๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: ๋ฌธ์-๋จ์ RNN์ผ๋ก ์ด๋ฆ ์์ฑํ๊ธฐ</span></a>
๋ฐ <a class="reference internal" href="seq2seq_translation_tutorial.html"><span class="doc">๊ธฐ์ด๋ถํฐ ์์ํ๋ NLP: Sequence to Sequence ๋คํธ์ํฌ์ Attention์ ์ด์ฉํ ๋ฒ์ญ</span></a> ์์๋ ์์ฐ์ด ์ฒ๋ฆฌ(NLP, Natural Language Processing)
๋ถ์ผ์์ ์ด๋ป๊ฒ ๋ฐ์ดํฐ๋ฅผ ์ ์ฒ๋ฆฌํ๊ณ NLP ๋ชจ๋ธ์ ๊ตฌ์ถํ๋์ง๋ฅผ ๋ฐ๋ฐ๋ฅ๋ถํฐ(from scratch) ์ค๋ช
ํฉ๋๋ค.
์ด๋ฅผ ์ํด ์ด ํํ ๋ฆฌ์ผ ์๋ฆฌ์ฆ์์๋ <cite>torchtext</cite> ์ ์๋ง์ ํธ๋ฆฌํ ๊ธฐ๋ฅ๋ค์ ์ฌ์ฉํ์ง ์๊ณ
NLP ๋ชจ๋ธ๋ง์ ์ํ ๋ฐ์ดํฐ ์ ์ฒ๋ฆฌ๊ฐ ๋ฐ๋ฐ๋ฅ(low-level)์์ ์ด๋ป๊ฒ ์งํ๋๋์ง ์ ์ ์์ต๋๋ค.</p>
<p>๋ฌธ์-๋จ์ RNN์ ๋จ์ด๋ฅผ ๋ฌธ์์ ์ฐ์์ผ๋ก ์ฝ์ด ๋ค์ฌ์ ๊ฐ ๋จ๊ณ์ ์์ธก๊ณผ
ใ์๋ ์ํ(Hidden State)ใ๋ฅผ ์ถ๋ ฅํ๊ณ , ๋ค์ ๋จ๊ณ์ ์ด์ ๋จ๊ณ์ ์๋ ์ํ๋ฅผ ์ ๋ฌํฉ๋๋ค.
๋จ์ด๊ฐ ์ํ ํด๋์ค๋ก ์ถ๋ ฅ๋๋๋ก ์ต์ข
์์ธก์ผ๋ก ์ ํํฉ๋๋ค.</p>
<p>๊ตฌ์ฒด์ ์ผ๋ก, 18๊ฐ ์ธ์ด๋ก ๋ ์์ฒ ๊ฐ์ ์ฑ(ๅง)์ ํ๋ จ์ํค๊ณ ,
์ฒ ์์ ๋ฐ๋ผ ์ด๋ฆ์ด ์ด๋ค ์ธ์ด์ธ์ง ์์ธกํฉ๋๋ค:</p>
<div class="highlight-sh notranslate"><div class="highlight"><pre><span></span>$<span class="w"> </span>python<span class="w"> </span>predict.py<span class="w"> </span>Hinton
<span class="o">(</span>-0.47<span class="o">)</span><span class="w"> </span>Scottish
<span class="o">(</span>-1.52<span class="o">)</span><span class="w"> </span>English
<span class="o">(</span>-3.57<span class="o">)</span><span class="w"> </span>Irish
$<span class="w"> </span>python<span class="w"> </span>predict.py<span class="w"> </span>Schmidhuber
<span class="o">(</span>-0.19<span class="o">)</span><span class="w"> </span>German
<span class="o">(</span>-2.48<span class="o">)</span><span class="w"> </span>Czech
<span class="o">(</span>-2.68<span class="o">)</span><span class="w"> </span>Dutch
</pre></div>
</div>
<div class="section" id="id3">
<h2>์ค๋น ๊ณผ์ <a class="headerlink" href="#id3" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h2>
<p>์ด ํํ ๋ฆฌ์ผ์ ์์ํ๊ธฐ ์ , PyTorch๋ฅผ ์ด๋ฏธ ์ค์นํ์ผ๋ฉฐ, Python ํ๋ก๊ทธ๋๋ฐ ์ธ์ด์
Tensor์ ๋ํ ๊ธฐ๋ณธ์ ์ธ ์ดํด๋ฅผ ํ๊ณ ๊ณ์
์ผ ํฉ๋๋ค:</p>
<ul class="simple">
<li><p><a class="reference external" href="https://pytorch.kr/">https://pytorch.kr/</a> ์์ ์ค์น ์๋ด๋ฅผ ์ฐพ์ ์ ์์ผ๋ฉฐ,</p></li>
<li><p><a class="reference internal" href="../beginner/deep_learning_60min_blitz.html"><span class="doc">PyTorch๋ก ๋ฅ๋ฌ๋ํ๊ธฐ: 60๋ถ๋ง์ ๋์ฅ๋ด๊ธฐ</span></a> ๋ฅผ ํตํด PyTorch์ ์ผ๋ฐ์ ์ธ ๋ด์ฉ๊ณผ Tensor์ ๋ํด ์ตํ ์ ์์ต๋๋ค.</p></li>
<li><p><a class="reference internal" href="../beginner/pytorch_with_examples.html"><span class="doc">์์ ๋ก ๋ฐฐ์ฐ๋ ํ์ดํ ์น(PyTorch)</span></a> ๋ ๋๊ณ ๊น์ ๊ฐ์(overview)๋ฅผ ์ ๊ณตํ๋ฉฐ,</p></li>
<li><p><a class="reference internal" href="../beginner/former_torchies_tutorial.html"><span class="doc">Torch ์ฌ์ฉ์๋ฅผ ์ํ PyTorch</span></a> ์ด์ Lua Torch ์ฌ์ฉ์๋ฅผ ์ํ ์๋ฃ์
๋๋ค.</p></li>
</ul>
<p>RNN๊ณผ ๊ทธ ์๋ ๋ฐฉ์์ ์๋ ๊ฒ ๋ํ ์ ์ฉํฉ๋๋ค:</p>
<ul class="simple">
<li><p><a class="reference external" href="https://karpathy.github.io/2015/05/21/rnn-effectiveness/">The Unreasonable Effectiveness of Recurrent Neural
Networks</a>
์ค์ํ ์์ ๋ฅผ ๋ณด์ฌ ์ค๋๋ค.</p></li>
<li><p><a class="reference external" href="https://colah.github.io/posts/2015-08-Understanding-LSTMs/">Understanding LSTM
Networks</a>
LSTM์ ๊ดํ ๊ฒ์ด์ง๋ง RNN์ ๊ดํด์๋ ์ ์ตํฉ๋๋ค.</p></li>
</ul>
</div>
<div class="section" id="id4">
<h2>๋ฐ์ดํฐ ์ค๋น<a class="headerlink" href="#id4" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h2>
<div class="admonition note">
<p class="admonition-title">์ฐธ๊ณ </p>
<p><a class="reference external" href="https://download.pytorch.org/tutorial/data.zip">์ฌ๊ธฐ</a> ์์ ๋ฐ์ดํฐ๋ฅผ ๋ค์ด๋ก๋ ๋ฐ๊ณ
ํ์ฌ ๋๋ ํ ๋ฆฌ์ ์์ถ์ ํธ์ญ์์ค.</p>
</div>
<p><code class="docutils literal notranslate"><span class="pre">data/names</span></code> ๋๋ ํ ๋ฆฌ์๋ <code class="docutils literal notranslate"><span class="pre">[Language].txt</span></code> ๋ผ๋ 18 ๊ฐ์ ํ
์คํธ ํ์ผ์ด ์์ต๋๋ค.
๊ฐ ํ์ผ์๋ ํ ์ค์ ํ๋์ ์ด๋ฆ์ด ํฌํจ๋์ด ์์ผ๋ฉฐ ๋๋ถ๋ถ ๋ก๋ง์๋ก ๋์ด ์์ต๋๋ค.
(๊ทธ๋ฌ๋, ์ ๋์ฝ๋์์ ASCII๋ก ๋ณํํด์ผ ํจ).</p>
<p>๊ฐ ์ธ์ด ๋ณ๋ก ์ด๋ฆ ๋ชฉ๋ก ์ฌ์ <code class="docutils literal notranslate"><span class="pre">{language:</span> <span class="pre">[names</span> <span class="pre">...]}</span></code> ์ ๋ง๋ญ๋๋ค.
์ผ๋ฐ ๋ณ์ ใcategoryใ ์ ใlineใ (์ฐ๋ฆฌ์ ๊ฒฝ์ฐ ์ธ์ด์ ์ด๋ฆ)์ ์ดํ์ ํ์ฅ์ฑ์ ์ํด ์ฌ์ฉ๋ฉ๋๋ค.</p>
<div class="admonition note">
<p class="admonition-title">์ฐธ๊ณ </p>
<p>์ญ์ ์ฃผ: ใlineใ ์ ์
๋ ฅ์ ใcategoryใ์ ํด๋์ค๋ฅผ ์ ์ฉํ์ฌ ๋ค๋ฅธ ๋ฌธ์ ์๋ ํ์ฉํ ์ ์์ต๋๋ค.
์ฌ๊ธฐ์๋ ใlineใ์ ์ด๋ฆ(ex. Robert)์ ์
๋ ฅ์ผ๋ก, ใcategoryใ์ ํด๋์ค(ex. english)๋ก ์ฌ์ฉํฉ๋๋ค.</p>
</div>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="kn">from</span> <span class="nn">__future__</span> <span class="kn">import</span> <span class="n">unicode_literals</span><span class="p">,</span> <span class="n">print_function</span><span class="p">,</span> <span class="n">division</span>
<span class="kn">from</span> <span class="nn">io</span> <span class="kn">import</span> <span class="nb">open</span>
<span class="kn">import</span> <span class="nn">glob</span>
<span class="kn">import</span> <span class="nn">os</span>
<span class="k">def</span> <span class="nf">findFiles</span><span class="p">(</span><span class="n">path</span><span class="p">):</span> <span class="k">return</span> <span class="n">glob</span><span class="o">.</span><span class="n">glob</span><span class="p">(</span><span class="n">path</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">findFiles</span><span class="p">(</span><span class="s1">'data/names/*.txt'</span><span class="p">))</span>
<span class="kn">import</span> <span class="nn">unicodedata</span>
<span class="kn">import</span> <span class="nn">string</span>
<span class="n">all_letters</span> <span class="o">=</span> <span class="n">string</span><span class="o">.</span><span class="n">ascii_letters</span> <span class="o">+</span> <span class="s2">" .,;'"</span>
<span class="n">n_letters</span> <span class="o">=</span> <span class="nb">len</span><span class="p">(</span><span class="n">all_letters</span><span class="p">)</span>
<span class="c1"># ์ ๋์ฝ๋ ๋ฌธ์์ด์ ASCII๋ก ๋ณํ, https://stackoverflow.com/a/518232/2809427</span>
<span class="k">def</span> <span class="nf">unicodeToAscii</span><span class="p">(</span><span class="n">s</span><span class="p">):</span>
<span class="k">return</span> <span class="s1">''</span><span class="o">.</span><span class="n">join</span><span class="p">(</span>
<span class="n">c</span> <span class="k">for</span> <span class="n">c</span> <span class="ow">in</span> <span class="n">unicodedata</span><span class="o">.</span><span class="n">normalize</span><span class="p">(</span><span class="s1">'NFD'</span><span class="p">,</span> <span class="n">s</span><span class="p">)</span>
<span class="k">if</span> <span class="n">unicodedata</span><span class="o">.</span><span class="n">category</span><span class="p">(</span><span class="n">c</span><span class="p">)</span> <span class="o">!=</span> <span class="s1">'Mn'</span>
<span class="ow">and</span> <span class="n">c</span> <span class="ow">in</span> <span class="n">all_letters</span>
<span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">unicodeToAscii</span><span class="p">(</span><span class="s1">'ลlusร rski'</span><span class="p">))</span>
<span class="c1"># ๊ฐ ์ธ์ด์ ์ด๋ฆ ๋ชฉ๋ก์ธ category_lines ์ฌ์ ์์ฑ</span>
<span class="n">category_lines</span> <span class="o">=</span> <span class="p">{}</span>
<span class="n">all_categories</span> <span class="o">=</span> <span class="p">[]</span>
<span class="c1"># ํ์ผ์ ์ฝ๊ณ ์ค ๋จ์๋ก ๋ถ๋ฆฌ</span>
<span class="k">def</span> <span class="nf">readLines</span><span class="p">(</span><span class="n">filename</span><span class="p">):</span>
<span class="n">lines</span> <span class="o">=</span> <span class="nb">open</span><span class="p">(</span><span class="n">filename</span><span class="p">,</span> <span class="n">encoding</span><span class="o">=</span><span class="s1">'utf-8'</span><span class="p">)</span><span class="o">.</span><span class="n">read</span><span class="p">()</span><span class="o">.</span><span class="n">strip</span><span class="p">()</span><span class="o">.</span><span class="n">split</span><span class="p">(</span><span class="s1">'</span><span class="se">\n</span><span class="s1">'</span><span class="p">)</span>
<span class="k">return</span> <span class="p">[</span><span class="n">unicodeToAscii</span><span class="p">(</span><span class="n">line</span><span class="p">)</span> <span class="k">for</span> <span class="n">line</span> <span class="ow">in</span> <span class="n">lines</span><span class="p">]</span>
<span class="k">for</span> <span class="n">filename</span> <span class="ow">in</span> <span class="n">findFiles</span><span class="p">(</span><span class="s1">'data/names/*.txt'</span><span class="p">):</span>
<span class="n">category</span> <span class="o">=</span> <span class="n">os</span><span class="o">.</span><span class="n">path</span><span class="o">.</span><span class="n">splitext</span><span class="p">(</span><span class="n">os</span><span class="o">.</span><span class="n">path</span><span class="o">.</span><span class="n">basename</span><span class="p">(</span><span class="n">filename</span><span class="p">))[</span><span class="mi">0</span><span class="p">]</span>
<span class="n">all_categories</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">category</span><span class="p">)</span>
<span class="n">lines</span> <span class="o">=</span> <span class="n">readLines</span><span class="p">(</span><span class="n">filename</span><span class="p">)</span>
<span class="n">category_lines</span><span class="p">[</span><span class="n">category</span><span class="p">]</span> <span class="o">=</span> <span class="n">lines</span>
<span class="n">n_categories</span> <span class="o">=</span> <span class="nb">len</span><span class="p">(</span><span class="n">all_categories</span><span class="p">)</span>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>['data/names/German.txt', 'data/names/Spanish.txt', 'data/names/Chinese.txt', 'data/names/Japanese.txt', 'data/names/Scottish.txt', 'data/names/Arabic.txt', 'data/names/English.txt', 'data/names/Vietnamese.txt', 'data/names/Dutch.txt', 'data/names/Irish.txt', 'data/names/Czech.txt', 'data/names/Polish.txt', 'data/names/Russian.txt', 'data/names/Korean.txt', 'data/names/Italian.txt', 'data/names/Greek.txt', 'data/names/Portuguese.txt', 'data/names/French.txt']
Slusarski
</pre></div>
</div>
<p>์ด์ ๊ฐ <code class="docutils literal notranslate"><span class="pre">category</span></code> (์ธ์ด)๋ฅผ <code class="docutils literal notranslate"><span class="pre">line</span></code> (์ด๋ฆ)์ ๋งคํํ๋ ์ฌ์ ์ธ
<code class="docutils literal notranslate"><span class="pre">category_lines</span></code> ๋ฅผ ๋ง๋ค์์ต๋๋ค. ๋์ค์ ์ฐธ์กฐํ ์ ์๋๋ก
<code class="docutils literal notranslate"><span class="pre">all_categories</span></code> (์ธ์ด ๋ชฉ๋ก)์ <code class="docutils literal notranslate"><span class="pre">n_categories</span></code> ๋ ์ถ์ ํฉ๋๋ค.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="nb">print</span><span class="p">(</span><span class="n">category_lines</span><span class="p">[</span><span class="s1">'Italian'</span><span class="p">][:</span><span class="mi">5</span><span class="p">])</span>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>['Abandonato', 'Abatangelo', 'Abatantuono', 'Abate', 'Abategiovanni']
</pre></div>
</div>
<div class="section" id="tensor">
<h3>์ด๋ฆ์ Tensor๋ก ๋ณ๊ฒฝ<a class="headerlink" href="#tensor" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h3>
<p>์ด์ ๋ชจ๋ ์ด๋ฆ์ ์ฒด๊ณํํ์ผ๋ฏ๋ก, ์ด๋ฅผ ํ์ฉํ๊ธฐ ์ํด Tensor๋ก
๋ณํํด์ผ ํฉ๋๋ค.</p>
<p>ํ๋์ ๋ฌธ์๋ฅผ ํํํ๊ธฐ ์ํด ํฌ๊ธฐ๊ฐ <code class="docutils literal notranslate"><span class="pre"><1</span> <span class="pre">x</span> <span class="pre">n_letters></span></code> ์ธ
ใOne-Hot ๋ฒกํฐใ๋ฅผ ์ฌ์ฉํฉ๋๋ค. One-Hot ๋ฒกํฐ๋ ํ์ฌ ๋ฌธ์์
์ฃผ์์๋ 1์ด, ๊ทธ ์ธ ๋๋จธ์ง ์ฃผ์์๋ 0์ด ์ฑ์์ง ๋ฒกํฐ์
๋๋ค.
์์ <code class="docutils literal notranslate"><span class="pre">"b"</span> <span class="pre">=</span> <span class="pre"><0</span> <span class="pre">1</span> <span class="pre">0</span> <span class="pre">0</span> <span class="pre">0</span> <span class="pre">...></span></code> .</p>
<p>๋จ์ด๋ฅผ ๋ง๋ค๊ธฐ ์ํด One-Hot ๋ฒกํฐ๋ค์ 2์ฐจ์ ํ๋ ฌ
<code class="docutils literal notranslate"><span class="pre"><line_length</span> <span class="pre">x</span> <span class="pre">1</span> <span class="pre">x</span> <span class="pre">n_letters></span></code> ์ ๊ฒฐํฉ์ํต๋๋ค.</p>
<p>์์์ ๋ณด์ด๋ ์ถ๊ฐ์ ์ธ 1์ฐจ์์ PyTorch์์ ๋ชจ๋ ๊ฒ์ด ๋ฐฐ์น(batch)์ ์๋ค๊ณ ๊ฐ์ ํ๊ธฐ
๋๋ฌธ์ ๋ฐ์ํฉ๋๋ค. ์ฌ๊ธฐ์๋ ๋ฐฐ์น ํฌ๊ธฐ 1์ ์ฌ์ฉํ๊ณ ์์ต๋๋ค.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># .. note::</span>
<span class="c1"># ์ญ์ ์ฃผ: One-Hot ๋ฒกํฐ๋ ์ธ์ด ๋ฐ ๋ฒ์ฃผํ ๋ฐ์ดํฐ๋ฅผ ๋ค๋ฃฐ ๋ ์ฃผ๋ก ์ฌ์ฉํ๋ฉฐ,</span>
<span class="c1"># ๋จ์ด, ๊ธ์ ๋ฑ์ ๋ฒกํฐ๋ก ํํํ ๋ ๋จ์ด, ๊ธ์ ์ฌ์ด์ ์๊ด ๊ด๊ณ๋ฅผ ๋ฏธ๋ฆฌ ์ ์ ์์ ๊ฒฝ์ฐ,</span>
<span class="c1"># One-Hot์ผ๋ก ํํํ์ฌ ์๋ก ์ง๊ตํ๋ค๊ณ ๊ฐ์ ํ๊ณ ํ์ต์ ์์ํฉ๋๋ค.</span>
<span class="c1"># ์ด์ ๋์ผํ๊ฒ, ์๊ด ๊ด๊ณ๋ฅผ ์ ์ ์๋ ๋ค๋ฅธ ๋ฐ์ดํฐ์ ๊ฒฝ์ฐ์๋ One-Hot ๋ฒกํฐ๋ฅผ ํ์ฉํ ์ ์์ต๋๋ค.</span>
<span class="c1">#</span>
<span class="kn">import</span> <span class="nn">torch</span>
<span class="c1"># all_letters ๋ก ๋ฌธ์์ ์ฃผ์ ์ฐพ๊ธฐ, ์์ "a" = 0</span>
<span class="k">def</span> <span class="nf">letterToIndex</span><span class="p">(</span><span class="n">letter</span><span class="p">):</span>
<span class="k">return</span> <span class="n">all_letters</span><span class="o">.</span><span class="n">find</span><span class="p">(</span><span class="n">letter</span><span class="p">)</span>
<span class="c1"># ๊ฒ์ฆ์ ์ํด์ ํ ๊ฐ์ ๋ฌธ์๋ฅผ <1 x n_letters> Tensor๋ก ๋ณํ</span>
<span class="k">def</span> <span class="nf">letterToTensor</span><span class="p">(</span><span class="n">letter</span><span class="p">):</span>
<span class="n">tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">zeros</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">n_letters</span><span class="p">)</span>
<span class="n">tensor</span><span class="p">[</span><span class="mi">0</span><span class="p">][</span><span class="n">letterToIndex</span><span class="p">(</span><span class="n">letter</span><span class="p">)]</span> <span class="o">=</span> <span class="mi">1</span>
<span class="k">return</span> <span class="n">tensor</span>
<span class="c1"># ํ ์ค(์ด๋ฆ)์ <line_length x 1 x n_letters>,</span>
<span class="c1"># ๋๋ One-Hot ๋ฌธ์ ๋ฒกํฐ์ Array๋ก ๋ณ๊ฒฝ</span>
<span class="k">def</span> <span class="nf">lineToTensor</span><span class="p">(</span><span class="n">line</span><span class="p">):</span>
<span class="n">tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">zeros</span><span class="p">(</span><span class="nb">len</span><span class="p">(</span><span class="n">line</span><span class="p">),</span> <span class="mi">1</span><span class="p">,</span> <span class="n">n_letters</span><span class="p">)</span>
<span class="k">for</span> <span class="n">li</span><span class="p">,</span> <span class="n">letter</span> <span class="ow">in</span> <span class="nb">enumerate</span><span class="p">(</span><span class="n">line</span><span class="p">):</span>
<span class="n">tensor</span><span class="p">[</span><span class="n">li</span><span class="p">][</span><span class="mi">0</span><span class="p">][</span><span class="n">letterToIndex</span><span class="p">(</span><span class="n">letter</span><span class="p">)]</span> <span class="o">=</span> <span class="mi">1</span>
<span class="k">return</span> <span class="n">tensor</span>
<span class="nb">print</span><span class="p">(</span><span class="n">letterToTensor</span><span class="p">(</span><span class="s1">'J'</span><span class="p">))</span>
<span class="nb">print</span><span class="p">(</span><span class="n">lineToTensor</span><span class="p">(</span><span class="s1">'Jones'</span><span class="p">)</span><span class="o">.</span><span class="n">size</span><span class="p">())</span>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>tensor([[0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1.,
0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
0., 0., 0.]])
torch.Size([5, 1, 57])
</pre></div>
</div>
</div>
</div>
<div class="section" id="id5">
<h2>๋คํธ์ํฌ ์์ฑ<a class="headerlink" href="#id5" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h2>
<p>Autograd ์ ์, Torch์์ RNN(recurrent neural network) ์์ฑ์
์ฌ๋ฌ ์๊ฐ ๋จ๊ณ ๊ฑธ์ณ์ ๊ณ์ธต์ ๋งค๊ฐ๋ณ์๋ฅผ ๋ณต์ ํ๋ ์์
์ ํฌํจํฉ๋๋ค.
๊ณ์ธต์ ์๋ ์ํ์ ๋ณํ๋(Gradient)๋ฅผ ๊ฐ์ง๋ฉฐ, ์ด์ ์ด๊ฒ๋ค์ ๊ทธ๋ํ ์์ฒด์์
์์ ํ ์ฒ๋ฆฌ๋ฉ๋๋ค. ์ด๋ feed-forward ๊ณ์ธต๊ณผ
๊ฐ์ ๋งค์ฐ ใ์์ํใ ๋ฐฉ๋ฒ์ผ๋ก RNN์ ๊ตฌํํ ์ ์์์ ์๋ฏธํฉ๋๋ค.</p>
<div class="admonition note">
<p class="admonition-title">์ฐธ๊ณ </p>
<p>์ญ์ ์ฃผ: ์ฌ๊ธฐ์๋ ํ์ต ๋ชฉ์ ์ผ๋ก nn.RNN ๋์ ์ง์ RNN์ ์ฌ์ฉํฉ๋๋ค.</p>
</div>
<p>์ด RNN ๋ชจ๋์ ใ๊ธฐ๋ณธ(vanilla)์ ์ธ RNNใ์ ๊ตฌํํ๋ฉฐ, ์
๋ ฅ๊ณผ ์๋ ์ํ(hidden state),
๊ทธ๋ฆฌ๊ณ ์ถ๋ ฅ ๋ค ๋์ํ๋ <code class="docutils literal notranslate"><span class="pre">LogSoftmax</span></code> ๊ณ์ธต์ด ์๋ 3๊ฐ์ ์ ํ ๊ณ์ธต๋ง์ ๊ฐ์ง๋๋ค.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">torch.nn</span> <span class="k">as</span> <span class="nn">nn</span>
<span class="kn">import</span> <span class="nn">torch.nn.functional</span> <span class="k">as</span> <span class="nn">F</span>
<span class="k">class</span> <span class="nc">RNN</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">input_size</span><span class="p">,</span> <span class="n">hidden_size</span><span class="p">,</span> <span class="n">output_size</span><span class="p">):</span>
<span class="nb">super</span><span class="p">(</span><span class="n">RNN</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">hidden_size</span> <span class="o">=</span> <span class="n">hidden_size</span>
<span class="bp">self</span><span class="o">.</span><span class="n">i2h</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">input_size</span><span class="p">,</span> <span class="n">hidden_size</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">h2h</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_size</span><span class="p">,</span> <span class="n">hidden_size</span><span class="p">)</span>
<span class="bp">self</span><span class="o">.</span><span class="n">h2o</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_size</span><span class="p">,</span> <span class="n">output_size</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">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="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">hidden</span> <span class="o">=</span> <span class="n">F</span><span class="o">.</span><span class="n">tanh</span><span class="p">(</span><span class="bp">self</span><span class="o">.</span><span class="n">i2h</span><span class="p">(</span><span class="nb">input</span><span class="p">)</span> <span class="o">+</span> <span class="bp">self</span><span class="o">.</span><span class="n">h2h</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">h2o</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">softmax</span><span class="p">(</span><span class="n">output</span><span class="p">)</span>
<span class="k">return</span> <span class="n">output</span><span class="p">,</span> <span class="n">hidden</span>
<span class="k">def</span> <span class="nf">initHidden</span><span class="p">(</span><span class="bp">self</span><span class="p">):</span>
<span class="k">return</span> <span class="n">torch</span><span class="o">.</span><span class="n">zeros</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">hidden_size</span><span class="p">)</span>
<span class="n">n_hidden</span> <span class="o">=</span> <span class="mi">128</span>
<span class="n">rnn</span> <span class="o">=</span> <span class="n">RNN</span><span class="p">(</span><span class="n">n_letters</span><span class="p">,</span> <span class="n">n_hidden</span><span class="p">,</span> <span class="n">n_categories</span><span class="p">)</span>
</pre></div>
</div>
<p>์ด ๋คํธ์ํฌ์ ํ ๋จ๊ณ๋ฅผ ์คํํ๋ ค๋ฉด ์
๋ ฅ(ํ์ฌ ๋ฌธ์ Tensor)๊ณผ
์ด์ ์ ์๋ ์ํ(์ฒ์์๋ 0์ผ๋ก ์ด๊ธฐํ)๋ฅผ ์ ๋ฌํด์ผ ํฉ๋๋ค.
์ถ๋ ฅ(๊ฐ ์ธ์ด์ ํ๋ฅ )๊ณผ ๋ค์ ์๋ ์ํ(๋ค์ ๋จ๊ณ๋ฅผ ์ํด ์ ์ง)๋ฅผ
๋๋ ค๋ฐ์ต๋๋ค.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="nb">input</span> <span class="o">=</span> <span class="n">letterToTensor</span><span class="p">(</span><span class="s1">'A'</span><span class="p">)</span>
<span class="n">hidden</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">zeros</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">n_hidden</span><span class="p">)</span>
<span class="n">output</span><span class="p">,</span> <span class="n">next_hidden</span> <span class="o">=</span> <span class="n">rnn</span><span class="p">(</span><span class="nb">input</span><span class="p">,</span> <span class="n">hidden</span><span class="p">)</span>
</pre></div>
</div>
<p>ํจ์จ์ฑ์ ์ํด์ ๋งค ๋จ๊ณ๋ง๋ค ์๋ก์ด Tensor๋ฅผ ๋ง๋ค๊ณ ์ถ์ง ์๊ธฐ ๋๋ฌธ์
<code class="docutils literal notranslate"><span class="pre">letterToTensor</span></code> ๋์ <code class="docutils literal notranslate"><span class="pre">lineToTensor</span></code> ๋ฅผ ์๋ผ์ ์ฌ์ฉํ
๊ฒ์
๋๋ค. ์ด๊ฒ์ Tensor์ ์ฌ์ ์ฐ์ฐ(pre-computing) ๋ฐฐ์น์ ์ํด
๋์ฑ ์ต์ ํ๋ ์ ์์ต๋๋ค.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="nb">input</span> <span class="o">=</span> <span class="n">lineToTensor</span><span class="p">(</span><span class="s1">'Albert'</span><span class="p">)</span>
<span class="n">hidden</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">zeros</span><span class="p">(</span><span class="mi">1</span><span class="p">,</span> <span class="n">n_hidden</span><span class="p">)</span>
<span class="n">output</span><span class="p">,</span> <span class="n">next_hidden</span> <span class="o">=</span> <span class="n">rnn</span><span class="p">(</span><span class="nb">input</span><span class="p">[</span><span class="mi">0</span><span class="p">],</span> <span class="n">hidden</span><span class="p">)</span>
<span class="nb">print</span><span class="p">(</span><span class="n">output</span><span class="p">)</span>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>tensor([[-3.0014, -2.8677, -2.9758, -2.9196, -3.1387, -2.8728, -2.8886, -2.8754,
-2.5694, -2.8957, -2.8363, -2.9602, -2.9206, -2.8656, -2.8350, -2.7372,
-3.0470, -2.9479]], grad_fn=<LogSoftmaxBackward0>)
</pre></div>
</div>
<p>๋ณด์๋ค์ํผ ์ถ๋ ฅ์ <code class="docutils literal notranslate"><span class="pre"><1</span> <span class="pre">x</span> <span class="pre">n_categories></span></code> Tensor์ด๊ณ , ๋ชจ๋ ํญ๋ชฉ์
ํด๋น ์นดํ
๊ณ ๋ฆฌ์ ์ฐ๋(likelihood)์
๋๋ค(๋ ๋์ ๊ฒ์ด ๋ ํ๋ฅ ๋์).</p>
</div>
<div class="section" id="id6">
<h2>ํ์ต<a class="headerlink" href="#id6" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h2>
<div class="section" id="id7">
<h3>ํ์ต ์ค๋น<a class="headerlink" href="#id7" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h3>
<p>ํ์ต์ ๋ค์ด๊ฐ๊ธฐ ์ , ๋ช๋ช ๋์์ด ๋๋ ํจ์(helper function)๋ฅผ ๋ง๋ค์ด์ผ ํฉ๋๋ค.
์ฒซ์งธ๋ ์ฐ๋ฆฌ๊ฐ ์์๋ธ ๊ฐ ์นดํ
๊ณ ๋ฆฌ์ ์ฐ๋(likelihood)์ธ ๋คํธ์ํฌ ์ถ๋ ฅ์ ํด์ํ๋ ํจ์์
๋๋ค.
๊ฐ์ฅ ํฐ ๊ฐ์ ์ฃผ์๋ฅผ ์๊ธฐ ์ํด์ <code class="docutils literal notranslate"><span class="pre">Tensor.topk</span></code> ๋ฅผ ์ฌ์ฉํ ์ ์์ต๋๋ค.</p>
<div class="admonition note">
<p class="admonition-title">์ฐธ๊ณ </p>
<p>์ญ์ ์ฃผ: ๋คํธ์ํฌ ์ถ๋ ฅ(๊ฐ ์นดํ
๊ณ ๋ฆฌ์ ์ฐ๋, likelihood)์ผ๋ก ๊ฐ์ฅ ํ๋ฅ ์ด ๋์
์นดํ
๊ณ ๋ฆฌ์ ์ด๋ฆ(์ธ์ด)๊ณผ ์นดํ
๊ณ ๋ฆฌ ๋ฒํธ๋ฅผ ๋ฐํํฉ๋๋ค.</p>
</div>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">categoryFromOutput</span><span class="p">(</span><span class="n">output</span><span class="p">):</span>
<span class="n">top_n</span><span class="p">,</span> <span class="n">top_i</span> <span class="o">=</span> <span class="n">output</span><span class="o">.</span><span class="n">topk</span><span class="p">(</span><span class="mi">1</span><span class="p">)</span> <span class="c1"># ํ
์์ ๊ฐ์ฅ ํฐ ๊ฐ ๋ฐ ์ฃผ์</span>
<span class="n">category_i</span> <span class="o">=</span> <span class="n">top_i</span><span class="p">[</span><span class="mi">0</span><span class="p">]</span><span class="o">.</span><span class="n">item</span><span class="p">()</span> <span class="c1"># ํ
์์์ ์ ์ ๊ฐ์ผ๋ก ๋ณ๊ฒฝ</span>
<span class="k">return</span> <span class="n">all_categories</span><span class="p">[</span><span class="n">category_i</span><span class="p">],</span> <span class="n">category_i</span>
<span class="nb">print</span><span class="p">(</span><span class="n">categoryFromOutput</span><span class="p">(</span><span class="n">output</span><span class="p">))</span>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>('Dutch', 8)
</pre></div>
</div>
<p>ํ์ต ์์(ํ๋์ ์ด๋ฆ๊ณผ ๊ทธ ์ธ์ด)๋ฅผ ์ป๋ ๋น ๋ฅธ ๋ฐฉ๋ฒ๋ ํ์ํฉ๋๋ค.:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">random</span>
<span class="k">def</span> <span class="nf">randomChoice</span><span class="p">(</span><span class="n">l</span><span class="p">):</span>
<span class="k">return</span> <span class="n">l</span><span class="p">[</span><span class="n">random</span><span class="o">.</span><span class="n">randint</span><span class="p">(</span><span class="mi">0</span><span class="p">,</span> <span class="nb">len</span><span class="p">(</span><span class="n">l</span><span class="p">)</span> <span class="o">-</span> <span class="mi">1</span><span class="p">)]</span>
<span class="k">def</span> <span class="nf">randomTrainingExample</span><span class="p">():</span>
<span class="n">category</span> <span class="o">=</span> <span class="n">randomChoice</span><span class="p">(</span><span class="n">all_categories</span><span class="p">)</span>
<span class="n">line</span> <span class="o">=</span> <span class="n">randomChoice</span><span class="p">(</span><span class="n">category_lines</span><span class="p">[</span><span class="n">category</span><span class="p">])</span>
<span class="n">category_tensor</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">tensor</span><span class="p">([</span><span class="n">all_categories</span><span class="o">.</span><span class="n">index</span><span class="p">(</span><span class="n">category</span><span class="p">)],</span> <span class="n">dtype</span><span class="o">=</span><span class="n">torch</span><span class="o">.</span><span class="n">long</span><span class="p">)</span>
<span class="n">line_tensor</span> <span class="o">=</span> <span class="n">lineToTensor</span><span class="p">(</span><span class="n">line</span><span class="p">)</span>
<span class="k">return</span> <span class="n">category</span><span class="p">,</span> <span class="n">line</span><span class="p">,</span> <span class="n">category_tensor</span><span class="p">,</span> <span class="n">line_tensor</span>
<span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="mi">10</span><span class="p">):</span>
<span class="n">category</span><span class="p">,</span> <span class="n">line</span><span class="p">,</span> <span class="n">category_tensor</span><span class="p">,</span> <span class="n">line_tensor</span> <span class="o">=</span> <span class="n">randomTrainingExample</span><span class="p">()</span>
<span class="nb">print</span><span class="p">(</span><span class="s1">'category ='</span><span class="p">,</span> <span class="n">category</span><span class="p">,</span> <span class="s1">'/ line ='</span><span class="p">,</span> <span class="n">line</span><span class="p">)</span>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>category = Spanish / line = Robledo
category = Greek / line = Paraskevopoulos
category = German / line = Gros
category = Italian / line = Parri
category = Dutch / line = Kan
category = Spanish / line = Ubina
category = Greek / line = Kaglantge
category = Chinese / line = Seto
category = Polish / line = Bukoski
category = Korean / line = Gil
</pre></div>
</div>
</div>
<div class="section" id="id8">
<h3>๋คํธ์ํฌ ํ์ต<a class="headerlink" href="#id8" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h3>
<p>์ด์ ์ด ๋คํธ์ํฌ๋ฅผ ํ์ตํ๋ ๋ฐ ํ์ํ ์์(ํ์ต ๋ฐ์ดํฐ)๋ฅผ ๋ณด์ฌ์ฃผ๊ณ ์ถ์ ํฉ๋๋ค.
๋ง์ผ ํ๋ ธ๋ค๋ฉด ์๋ ค ์ค๋๋ค.</p>
<p>RNN์ ๋ง์ง๋ง ๊ณ์ธต์ด <code class="docutils literal notranslate"><span class="pre">nn.LogSoftmax</span></code> ์ด๋ฏ๋ก ์์ค ํจ์๋ก
<code class="docutils literal notranslate"><span class="pre">nn.NLLLoss</span></code> ๊ฐ ์ ํฉํฉ๋๋ค.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">criterion</span> <span class="o">=</span> <span class="n">nn</span><span class="o">.</span><span class="n">NLLLoss</span><span class="p">()</span>
</pre></div>
</div>
<p>๊ฐ ํ์ต ๋ฃจํ๋ ๋ค์๊ณผ ๊ฐ์ต๋๋ค:</p>
<ul class="simple">
<li><p>์
๋ ฅ๊ณผ ๋ชฉํ Tensor ์์ฑ</p></li>
<li><p>0 ๋ก ์ด๊ธฐํ ๋ ์๋ ์ํ ์์ฑ</p></li>
<li><p>๊ฐ ๋ฌธ์ ์ฝ๊ธฐ</p></li>
<li><p>๋ค์ ๋ฌธ์๋ฅผ ์ํ ์๋ ์ํ ์ ์ง</p></li>
<li><p>๋ชฉํ์ ์ต์ข
์ถ๋ ฅ ๋น๊ต</p></li>
<li><p>์ญ์ ํ</p></li>
<li><p>์ถ๋ ฅ๊ณผ ์์ค ๋ฐํ</p></li>
</ul>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">learning_rate</span> <span class="o">=</span> <span class="mf">0.005</span> <span class="c1"># ํ์ต๋ฅ ์ ๋๋ฌด ๋๊ฒ ์ค์ ํ๋ฉด ๋ฐ์ฐํ ์ ์๊ณ , ๋๋ฌด ๋ฎ์ผ๋ฉด ํ์ต์ด ๋์ง ์์ ์ ์์ต๋๋ค.</span>
<span class="k">def</span> <span class="nf">train</span><span class="p">(</span><span class="n">category_tensor</span><span class="p">,</span> <span class="n">line_tensor</span><span class="p">):</span>
<span class="n">hidden</span> <span class="o">=</span> <span class="n">rnn</span><span class="o">.</span><span class="n">initHidden</span><span class="p">()</span>
<span class="n">rnn</span><span class="o">.</span><span class="n">zero_grad</span><span class="p">()</span>
<span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="n">line_tensor</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="p">,</span> <span class="n">hidden</span> <span class="o">=</span> <span class="n">rnn</span><span class="p">(</span><span class="n">line_tensor</span><span class="p">[</span><span class="n">i</span><span class="p">],</span> <span class="n">hidden</span><span class="p">)</span>
<span class="n">loss</span> <span class="o">=</span> <span class="n">criterion</span><span class="p">(</span><span class="n">output</span><span class="p">,</span> <span class="n">category_tensor</span><span class="p">)</span>
<span class="n">loss</span><span class="o">.</span><span class="n">backward</span><span class="p">()</span>
<span class="c1"># ๋งค๊ฐ๋ณ์์ ๊ฒฝ์ฌ๋์ ํ์ต๋ฅ ์ ๊ณฑํด์ ๊ทธ ๋งค๊ฐ๋ณ์์ ๊ฐ์ ๋ํฉ๋๋ค.</span>
<span class="k">for</span> <span class="n">p</span> <span class="ow">in</span> <span class="n">rnn</span><span class="o">.</span><span class="n">parameters</span><span class="p">():</span>
<span class="n">p</span><span class="o">.</span><span class="n">data</span><span class="o">.</span><span class="n">add_</span><span class="p">(</span><span class="n">p</span><span class="o">.</span><span class="n">grad</span><span class="o">.</span><span class="n">data</span><span class="p">,</span> <span class="n">alpha</span><span class="o">=-</span><span class="n">learning_rate</span><span class="p">)</span>
<span class="k">return</span> <span class="n">output</span><span class="p">,</span> <span class="n">loss</span><span class="o">.</span><span class="n">item</span><span class="p">()</span>
</pre></div>
</div>
<p>์ด์ ์์ ๋ฐ์ดํฐ๋ฅผ ์ฌ์ฉํ์ฌ ์คํํด์ผ ํฉ๋๋ค. <code class="docutils literal notranslate"><span class="pre">train</span></code> ํจ์๊ฐ ์ถ๋ ฅ๊ณผ ์์ค์
๋ฐํํ๊ธฐ ๋๋ฌธ์ ์ถ์ธก์ ํ๋ฉด์ ์ถ๋ ฅํ๊ณ ๋์ํ๋ฅผ ์ํ ์์ค์ ์ถ์ ํ ์
์์ต๋๋ค. 1000๊ฐ์ ์์ ๋ฐ์ดํฐ๊ฐ ์๊ธฐ ๋๋ฌธ์ <code class="docutils literal notranslate"><span class="pre">print_every</span></code> ์์ ๋ง
์ถ๋ ฅํ๊ณ , ์์ค์ ํ๊ท ์ ์ป์ต๋๋ค.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">time</span>
<span class="kn">import</span> <span class="nn">math</span>
<span class="n">n_iters</span> <span class="o">=</span> <span class="mi">100000</span>
<span class="n">print_every</span> <span class="o">=</span> <span class="mi">5000</span>
<span class="n">plot_every</span> <span class="o">=</span> <span class="mi">1000</span>
<span class="c1"># ๋์ํ๋ฅผ ์ํ ์์ค ์ถ์ </span>
<span class="n">current_loss</span> <span class="o">=</span> <span class="mi">0</span>
<span class="n">all_losses</span> <span class="o">=</span> <span class="p">[]</span>
<span class="k">def</span> <span class="nf">timeSince</span><span class="p">(</span><span class="n">since</span><span class="p">):</span>
<span class="n">now</span> <span class="o">=</span> <span class="n">time</span><span class="o">.</span><span class="n">time</span><span class="p">()</span>
<span class="n">s</span> <span class="o">=</span> <span class="n">now</span> <span class="o">-</span> <span class="n">since</span>
<span class="n">m</span> <span class="o">=</span> <span class="n">math</span><span class="o">.</span><span class="n">floor</span><span class="p">(</span><span class="n">s</span> <span class="o">/</span> <span class="mi">60</span><span class="p">)</span>
<span class="n">s</span> <span class="o">-=</span> <span class="n">m</span> <span class="o">*</span> <span class="mi">60</span>
<span class="k">return</span> <span class="s1">'</span><span class="si">%d</span><span class="s1">m </span><span class="si">%d</span><span class="s1">s'</span> <span class="o">%</span> <span class="p">(</span><span class="n">m</span><span class="p">,</span> <span class="n">s</span><span class="p">)</span>
<span class="n">start</span> <span class="o">=</span> <span class="n">time</span><span class="o">.</span><span class="n">time</span><span class="p">()</span>
<span class="k">for</span> <span class="nb">iter</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_iters</span> <span class="o">+</span> <span class="mi">1</span><span class="p">):</span>
<span class="n">category</span><span class="p">,</span> <span class="n">line</span><span class="p">,</span> <span class="n">category_tensor</span><span class="p">,</span> <span class="n">line_tensor</span> <span class="o">=</span> <span class="n">randomTrainingExample</span><span class="p">()</span>
<span class="n">output</span><span class="p">,</span> <span class="n">loss</span> <span class="o">=</span> <span class="n">train</span><span class="p">(</span><span class="n">category_tensor</span><span class="p">,</span> <span class="n">line_tensor</span><span class="p">)</span>
<span class="n">current_loss</span> <span class="o">+=</span> <span class="n">loss</span>
<span class="c1"># ``iter`` ์ซ์, ์์ค, ์ด๋ฆ, ์ถ์ธก ํ๋ฉด ์ถ๋ ฅ</span>
<span class="k">if</span> <span class="nb">iter</span> <span class="o">%</span> <span class="n">print_every</span> <span class="o">==</span> <span class="mi">0</span><span class="p">:</span>
<span class="n">guess</span><span class="p">,</span> <span class="n">guess_i</span> <span class="o">=</span> <span class="n">categoryFromOutput</span><span class="p">(</span><span class="n">output</span><span class="p">)</span>
<span class="n">correct</span> <span class="o">=</span> <span class="s1">'โ'</span> <span class="k">if</span> <span class="n">guess</span> <span class="o">==</span> <span class="n">category</span> <span class="k">else</span> <span class="s1">'โ (</span><span class="si">%s</span><span class="s1">)'</span> <span class="o">%</span> <span class="n">category</span>
<span class="nb">print</span><span class="p">(</span><span class="s1">'</span><span class="si">%d</span><span class="s1"> </span><span class="si">%d%%</span><span class="s1"> (</span><span class="si">%s</span><span class="s1">) </span><span class="si">%.4f</span><span class="s1"> </span><span class="si">%s</span><span class="s1"> / </span><span class="si">%s</span><span class="s1"> </span><span class="si">%s</span><span class="s1">'</span> <span class="o">%</span> <span class="p">(</span><span class="nb">iter</span><span class="p">,</span> <span class="nb">iter</span> <span class="o">/</span> <span class="n">n_iters</span> <span class="o">*</span> <span class="mi">100</span><span class="p">,</span> <span class="n">timeSince</span><span class="p">(</span><span class="n">start</span><span class="p">),</span> <span class="n">loss</span><span class="p">,</span> <span class="n">line</span><span class="p">,</span> <span class="n">guess</span><span class="p">,</span> <span class="n">correct</span><span class="p">))</span>
<span class="c1"># ํ์ฌ ํ๊ท ์์ค์ ์ ์ฒด ์์ค ๋ฆฌ์คํธ์ ์ถ๊ฐ</span>
<span class="k">if</span> <span class="nb">iter</span> <span class="o">%</span> <span class="n">plot_every</span> <span class="o">==</span> <span class="mi">0</span><span class="p">:</span>
<span class="n">all_losses</span><span class="o">.</span><span class="n">append</span><span class="p">(</span><span class="n">current_loss</span> <span class="o">/</span> <span class="n">plot_every</span><span class="p">)</span>
<span class="n">current_loss</span> <span class="o">=</span> <span class="mi">0</span>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>5000 5% (0m 3s) 2.4839 Kasimor / Russian โ (Czech)
10000 10% (0m 7s) 3.1097 Kumiega / Spanish โ (Polish)
15000 15% (0m 11s) 1.7582 O'Halloran / Dutch โ (Irish)
20000 20% (0m 15s) 0.5089 Gimondi / Italian โ
25000 25% (0m 19s) 1.4201 Reuter / Dutch โ (German)
30000 30% (0m 23s) 1.8297 Seger / German โ (Czech)
35000 35% (0m 27s) 2.7120 Matos / Arabic โ (Portuguese)
40000 40% (0m 31s) 0.3220 Tokudome / Japanese โ
45000 45% (0m 35s) 0.4470 Luc / Vietnamese โ
50000 50% (0m 39s) 2.0357 Smith / English โ (Scottish)
55000 55% (0m 43s) 3.9332 Steer / German โ (English)
60000 60% (0m 47s) 1.1131 Hamilton / Russian โ (Scottish)
65000 65% (0m 51s) 0.5333 Teoh / Chinese โ
70000 70% (0m 55s) 2.9210 Onoe / Irish โ (Japanese)
75000 75% (0m 59s) 2.7158 Maier / French โ (German)
80000 80% (1m 3s) 4.4062 Mas / Greek โ (Dutch)
85000 85% (1m 7s) 0.0838 Mai / Chinese โ
90000 90% (1m 11s) 0.4596 Dubhshlaine / Irish โ
95000 95% (1m 15s) 0.0770 Nosek / Polish โ
100000 100% (1m 19s) 0.1322 Elansky / Russian โ
</pre></div>
</div>
</div>
<div class="section" id="id9">
<h3>๊ฒฐ๊ณผ ๋์ํ<a class="headerlink" href="#id9" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h3>
<p><code class="docutils literal notranslate"><span class="pre">all_losses</span></code> ๋ฅผ ์ด์ฉํ ์์ค ๋์ํ๋
๋คํธ์ํฌ์ ํ์ต์ ๋ณด์ฌ์ค๋๋ค:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="kn">import</span> <span class="nn">matplotlib.pyplot</span> <span class="k">as</span> <span class="nn">plt</span>
<span class="kn">import</span> <span class="nn">matplotlib.ticker</span> <span class="k">as</span> <span class="nn">ticker</span>
<span class="n">plt</span><span class="o">.</span><span class="n">figure</span><span class="p">()</span>
<span class="n">plt</span><span class="o">.</span><span class="n">plot</span><span class="p">(</span><span class="n">all_losses</span><span class="p">)</span>
</pre></div>
</div>
<img src="../_images/sphx_glr_char_rnn_classification_tutorial_001.png" srcset="../_images/sphx_glr_char_rnn_classification_tutorial_001.png" alt="char rnn classification tutorial" class = "sphx-glr-single-img"/><div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>[<matplotlib.lines.Line2D object at 0x7775c83aa0b0>]
</pre></div>
</div>
</div>
</div>
<div class="section" id="id10">
<h2>๊ฒฐ๊ณผ ํ๊ฐ<a class="headerlink" href="#id10" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h2>
<p>๋คํธ์ํฌ๊ฐ ๋ค๋ฅธ ์นดํ
๊ณ ๋ฆฌ์์ ์ผ๋ง๋ ์ ์๋ํ๋์ง ๋ณด๊ธฐ ์ํด
๋ชจ๋ ์ค์ ์ธ์ด(ํ)๊ฐ ๋คํธ์ํฌ์์ ์ด๋ค ์ธ์ด๋ก ์ถ์ธก(์ด)๋๋์ง ๋ํ๋ด๋
ํผ๋ ํ๋ ฌ(confusion matrix)์ ๋ง๋ญ๋๋ค. ํผ๋ ํ๋ ฌ์ ๊ณ์ฐํ๊ธฐ ์ํด
<code class="docutils literal notranslate"><span class="pre">evaluate()</span></code> ๋ก ๋ง์ ์์ ์ํ์ ๋คํธ์ํฌ์ ์คํํฉ๋๋ค.
<code class="docutils literal notranslate"><span class="pre">evaluate()</span></code> ์ <code class="docutils literal notranslate"><span class="pre">train</span> <span class="pre">()</span></code> ๊ณผ ์ญ์ ํ๋ฅผ ๋นผ๋ฉด ๋์ผํฉ๋๋ค.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="c1"># ํผ๋ ํ๋ ฌ์์ ์ ํํ ์ถ์ธก์ ์ถ์ </span>
<span class="n">confusion</span> <span class="o">=</span> <span class="n">torch</span><span class="o">.</span><span class="n">zeros</span><span class="p">(</span><span class="n">n_categories</span><span class="p">,</span> <span class="n">n_categories</span><span class="p">)</span>
<span class="n">n_confusion</span> <span class="o">=</span> <span class="mi">10000</span>
<span class="c1"># ์ฃผ์ด์ง ๋ผ์ธ์ ์ถ๋ ฅ ๋ฐํ</span>
<span class="k">def</span> <span class="nf">evaluate</span><span class="p">(</span><span class="n">line_tensor</span><span class="p">):</span>
<span class="n">hidden</span> <span class="o">=</span> <span class="n">rnn</span><span class="o">.</span><span class="n">initHidden</span><span class="p">()</span>
<span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="n">line_tensor</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="p">,</span> <span class="n">hidden</span> <span class="o">=</span> <span class="n">rnn</span><span class="p">(</span><span class="n">line_tensor</span><span class="p">[</span><span class="n">i</span><span class="p">],</span> <span class="n">hidden</span><span class="p">)</span>
<span class="k">return</span> <span class="n">output</span>
<span class="c1"># ์์ ์ค ์ด๋ค ๊ฒ์ด ์ ํํ ์์ธก๋์๋์ง ๊ธฐ๋ก</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="n">n_confusion</span><span class="p">):</span>
<span class="n">category</span><span class="p">,</span> <span class="n">line</span><span class="p">,</span> <span class="n">category_tensor</span><span class="p">,</span> <span class="n">line_tensor</span> <span class="o">=</span> <span class="n">randomTrainingExample</span><span class="p">()</span>
<span class="n">output</span> <span class="o">=</span> <span class="n">evaluate</span><span class="p">(</span><span class="n">line_tensor</span><span class="p">)</span>
<span class="n">guess</span><span class="p">,</span> <span class="n">guess_i</span> <span class="o">=</span> <span class="n">categoryFromOutput</span><span class="p">(</span><span class="n">output</span><span class="p">)</span>
<span class="n">category_i</span> <span class="o">=</span> <span class="n">all_categories</span><span class="o">.</span><span class="n">index</span><span class="p">(</span><span class="n">category</span><span class="p">)</span>
<span class="n">confusion</span><span class="p">[</span><span class="n">category_i</span><span class="p">][</span><span class="n">guess_i</span><span class="p">]</span> <span class="o">+=</span> <span class="mi">1</span>
<span class="c1"># ๋ชจ๋ ํ์ ํฉ๊ณ๋ก ๋๋์ด ์ ๊ทํ</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="n">n_categories</span><span class="p">):</span>
<span class="n">confusion</span><span class="p">[</span><span class="n">i</span><span class="p">]</span> <span class="o">=</span> <span class="n">confusion</span><span class="p">[</span><span class="n">i</span><span class="p">]</span> <span class="o">/</span> <span class="n">confusion</span><span class="p">[</span><span class="n">i</span><span class="p">]</span><span class="o">.</span><span class="n">sum</span><span class="p">()</span>
<span class="c1"># ๋์ ์ค์ </span>
<span class="n">fig</span> <span class="o">=</span> <span class="n">plt</span><span class="o">.</span><span class="n">figure</span><span class="p">()</span>
<span class="n">ax</span> <span class="o">=</span> <span class="n">fig</span><span class="o">.</span><span class="n">add_subplot</span><span class="p">(</span><span class="mi">111</span><span class="p">)</span>
<span class="n">cax</span> <span class="o">=</span> <span class="n">ax</span><span class="o">.</span><span class="n">matshow</span><span class="p">(</span><span class="n">confusion</span><span class="o">.</span><span class="n">numpy</span><span class="p">())</span>
<span class="n">fig</span><span class="o">.</span><span class="n">colorbar</span><span class="p">(</span><span class="n">cax</span><span class="p">)</span>
<span class="c1"># ์ถ ์ค์ </span>
<span class="n">ax</span><span class="o">.</span><span class="n">set_xticklabels</span><span class="p">([</span><span class="s1">''</span><span class="p">]</span> <span class="o">+</span> <span class="n">all_categories</span><span class="p">,</span> <span class="n">rotation</span><span class="o">=</span><span class="mi">90</span><span class="p">)</span>
<span class="n">ax</span><span class="o">.</span><span class="n">set_yticklabels</span><span class="p">([</span><span class="s1">''</span><span class="p">]</span> <span class="o">+</span> <span class="n">all_categories</span><span class="p">)</span>
<span class="c1"># ๋ชจ๋ tick์์ ๋ ์ด๋ธ ์ง์ </span>
<span class="n">ax</span><span class="o">.</span><span class="n">xaxis</span><span class="o">.</span><span class="n">set_major_locator</span><span class="p">(</span><span class="n">ticker</span><span class="o">.</span><span class="n">MultipleLocator</span><span class="p">(</span><span class="mi">1</span><span class="p">))</span>
<span class="n">ax</span><span class="o">.</span><span class="n">yaxis</span><span class="o">.</span><span class="n">set_major_locator</span><span class="p">(</span><span class="n">ticker</span><span class="o">.</span><span class="n">MultipleLocator</span><span class="p">(</span><span class="mi">1</span><span class="p">))</span>
<span class="c1"># sphinx_gallery_thumbnail_number = 2</span>
<span class="n">plt</span><span class="o">.</span><span class="n">show</span><span class="p">()</span>
</pre></div>
</div>
<img src="../_images/sphx_glr_char_rnn_classification_tutorial_002.png" srcset="../_images/sphx_glr_char_rnn_classification_tutorial_002.png" alt="char rnn classification tutorial" class = "sphx-glr-single-img"/><p>์ฃผ์ถ์์ ๋ฒ์ด๋ ๋ฐ์ ์ ์ ์ ํํ์ฌ ์๋ชป ์ถ์ธกํ ์ธ์ด๋ฅผ ํ์ํ ์ ์์ต๋๋ค.
์๋ฅผ ๋ค์ด ํ๊ตญ์ด๋ ์ค๊ตญ์ด๋ก ์ดํ๋ฆฌ์์ด๋ก ์คํ์ธ์ด๋ก.
๊ทธ๋ฆฌ์ค์ด๋ ๋งค์ฐ ์๋๋ ๊ฒ์ผ๋ก ์์ด๋ ๋งค์ฐ ๋์ ๊ฒ์ผ๋ก ๋ณด์
๋๋ค.
(๋ค๋ฅธ ์ธ์ด๋ค๊ณผ์ ์ค์ฒฉ ๋๋ฌธ์ผ๋ก ์ถ์ )</p>
<div class="section" id="id11">
<h3>์ฌ์ฉ์ ์
๋ ฅ์ผ๋ก ์คํ<a class="headerlink" href="#id11" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h3>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="k">def</span> <span class="nf">predict</span><span class="p">(</span><span class="n">input_line</span><span class="p">,</span> <span class="n">n_predictions</span><span class="o">=</span><span class="mi">3</span><span class="p">):</span>
<span class="nb">print</span><span class="p">(</span><span class="s1">'</span><span class="se">\n</span><span class="s1">> </span><span class="si">%s</span><span class="s1">'</span> <span class="o">%</span> <span class="n">input_line</span><span class="p">)</span>
<span class="k">with</span> <span class="n">torch</span><span class="o">.</span><span class="n">no_grad</span><span class="p">():</span>
<span class="n">output</span> <span class="o">=</span> <span class="n">evaluate</span><span class="p">(</span><span class="n">lineToTensor</span><span class="p">(</span><span class="n">input_line</span><span class="p">))</span>
<span class="c1"># Get top N categories</span>
<span class="n">topv</span><span class="p">,</span> <span class="n">topi</span> <span class="o">=</span> <span class="n">output</span><span class="o">.</span><span class="n">topk</span><span class="p">(</span><span class="n">n_predictions</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="kc">True</span><span class="p">)</span>
<span class="n">predictions</span> <span class="o">=</span> <span class="p">[]</span>
<span class="k">for</span> <span class="n">i</span> <span class="ow">in</span> <span class="nb">range</span><span class="p">(</span><span class="n">n_predictions</span><span class="p">):</span>
<span class="n">value</span> <span class="o">=</span> <span class="n">topv</span><span class="p">[</span><span class="mi">0</span><span class="p">][</span><span class="n">i</span><span class="p">]</span><span class="o">.</span><span class="n">item</span><span class="p">()</span>
<span class="n">category_index</span> <span class="o">=</span> <span class="n">topi</span><span class="p">[</span><span class="mi">0</span><span class="p">][</span><span class="n">i</span><span class="p">]</span><span class="o">.</span><span class="n">item</span><span class="p">()</span>
<span class="nb">print</span><span class="p">(</span><span class="s1">'(</span><span class="si">%.2f</span><span class="s1">) </span><span class="si">%s</span><span class="s1">'</span> <span class="o">%</span> <span class="p">(</span><span class="n">value</span><span class="p">,</span> <span class="n">all_categories</span><span class="p">[</span><span class="n">category_index</span><span class="p">]))</span>
<span class="n">predictions</span><span class="o">.</span><span class="n">append</span><span class="p">([</span><span class="n">value</span><span class="p">,</span> <span class="n">all_categories</span><span class="p">[</span><span class="n">category_index</span><span class="p">]])</span>
<span class="n">predict</span><span class="p">(</span><span class="s1">'Dovesky'</span><span class="p">)</span>
<span class="n">predict</span><span class="p">(</span><span class="s1">'Jackson'</span><span class="p">)</span>
<span class="n">predict</span><span class="p">(</span><span class="s1">'Satoshi'</span><span class="p">)</span>
</pre></div>
</div>
<div class="sphx-glr-script-out highlight-none notranslate"><div class="highlight"><pre><span></span>> Dovesky
(-0.67) Czech
(-1.02) Russian
(-2.66) English
> Jackson
(-0.08) Scottish
(-2.99) English
(-4.35) French
> Satoshi
(-1.11) Italian
(-1.11) Polish
(-1.47) Japanese
</pre></div>
</div>
<p><a class="reference external" href="https://github.com/spro/practical-pytorch/tree/master/char-rnn-classification">์ค์ฉ PyTorch ์ ์ฅ์</a>
์ ์ต์ข
๋ฒ์ ์คํฌ๋ฆฝํธ๋ ์ ์ฝ๋๋ฅผ ๋ช ๊ฐ์ ํ์ผ๋ก ๋ถํ ํ์ต๋๋ค.:</p>
<ul class="simple">
<li><p><code class="docutils literal notranslate"><span class="pre">data.py</span></code> (ํ์ผ ์ฝ๊ธฐ)</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">model.py</span></code> (RNN ์ ์)</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">train.py</span></code> (ํ์ต ์คํ)</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">predict.py</span></code> (์ปค๋งจ๋ ๋ผ์ธ ์ธ์๋ก <code class="docutils literal notranslate"><span class="pre">predict()</span></code> ์คํ)</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">server.py</span></code> (<code class="docutils literal notranslate"><span class="pre">bottle.py</span></code> ๋ฅผ ์ฌ์ฉํ์ฌ JSON API๋ก ์์ธก ์ ๊ณต)</p></li>
</ul>
<p><code class="docutils literal notranslate"><span class="pre">train.py</span></code> ๋ฅผ ์คํํ๋ฉด ํ์ต ๋ฐ ์ ๊ฒฝ๋ง์ ์ ์ฅํฉ๋๋ค.</p>
<p><code class="docutils literal notranslate"><span class="pre">predict.py</span></code> ๋ฅผ ์คํํ๋ฉด ์ด๋ฆ์ผ๋ก๋ถํฐ ์์ธก์ ์คํํฉ๋๋ค:</p>
<div class="highlight-sh notranslate"><div class="highlight"><pre><span></span>$<span class="w"> </span>python<span class="w"> </span>predict.py<span class="w"> </span>Hazaki
<span class="o">(</span>-0.42<span class="o">)</span><span class="w"> </span>Japanese
<span class="o">(</span>-1.39<span class="o">)</span><span class="w"> </span>Polish
<span class="o">(</span>-3.51<span class="o">)</span><span class="w"> </span>Czech
</pre></div>
</div>
<p><code class="docutils literal notranslate"><span class="pre">server.py</span></code> ๋ฅผ ์คํํ๊ณ <a class="reference external" href="http://localhost:5533/Yourname">http://localhost:5533/Yourname</a> ์ ๋ฐฉ๋ฌธํ๋ฉด
์์ธก ๊ฐ์ ๋ํ JSON ์ถ๋ ฅ์ ํ์ธํ ์ ์์ต๋๋ค.</p>
</div>
</div>
<div class="section" id="id12">
<h2>์ฐ์ต<a class="headerlink" href="#id12" title="์ด ์ ๋ชฉ์ ๋ํ ํผ๋จธ๋งํฌ">ยถ</a></h2>
<ul class="simple">
<li><p>ใline -> categoryใ ์ ๋ค๋ฅธ ๋ฐ์ดํฐ ์งํฉ์ผ๋ก ์๋ํด ๋ณด์ญ์์ค, ์๋ฅผ ๋ค์ด:</p>
<ul>
<li><p>๋จ์ด -> ์ธ์ด</p></li>
<li><p>์ด๋ฆ -> ์ฑ๋ณ</p></li>
<li><p>์บ๋ฆญํฐ ์ด๋ฆ -> ์๊ฐ</p></li>
<li><p>ํ์ด์ง ์ ๋ชฉ -> ๋ธ๋ก๊ทธ ๋๋ ์๋ธ๋ ๋ง</p></li>
</ul>
</li>
<li><p>๋ ํฌ๊ณ ๋ ๋์ ๋ชจ์์ ๋คํธ์ํฌ๋ก ๋ ๋์ ๊ฒฐ๊ณผ๋ฅผ ์ป์ผ์ญ์์ค.</p>
<ul>
<li><p>๋ ๋ง์ ์ ํ ๊ณ์ธต์ ์ถ๊ฐํด ๋ณด์ญ์์ค.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">nn.LSTM</span></code> ๊ณผ <code class="docutils literal notranslate"><span class="pre">nn.GRU</span></code> ๊ณ์ธต์ ์ถ๊ฐํด ๋ณด์ญ์์ค.</p></li>
<li><p>์์ ๊ฐ์ RNN ์ฌ๋ฌ ๊ฐ๋ฅผ ์์ ์์ค ๋คํธ์ํฌ๋ก ๊ฒฐํฉํด ๋ณด์ญ์์ค.</p></li>
</ul>
</li>
</ul>
<p class="sphx-glr-timing"><strong>Total running time of the script:</strong> ( 1 minutes 23.059 seconds)</p>
<div class="sphx-glr-footer sphx-glr-footer-example docutils container" id="sphx-glr-download-intermediate-char-rnn-classification-tutorial-py">
<div class="sphx-glr-download sphx-glr-download-python docutils container">
<p><a class="reference download internal" download="" href="../_downloads/37c8905519d3fd3f437b783a48d06eac/char_rnn_classification_tutorial.py"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Python</span> <span class="pre">source</span> <span class="pre">code:</span> <span class="pre">char_rnn_classification_tutorial.py</span></code></a></p>
</div>
<div class="sphx-glr-download sphx-glr-download-jupyter docutils container">
<p><a class="reference download internal" download="" href="../_downloads/13b143c2380f4768d9432d808ad50799/char_rnn_classification_tutorial.ipynb"><code class="xref download docutils literal notranslate"><span class="pre">Download</span> <span class="pre">Jupyter</span> <span class="pre">notebook:</span> <span class="pre">char_rnn_classification_tutorial.ipynb</span></code></a></p>
</div>