-
Notifications
You must be signed in to change notification settings - Fork 47
/
Copy pathadapt.parameter_based.FineTuning.html
968 lines (933 loc) · 76.1 KB
/
adapt.parameter_based.FineTuning.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
<!DOCTYPE html>
<html class="writer-html5" lang="en" >
<head>
<meta charset="utf-8" /><meta name="generator" content="Docutils 0.17.1: http://docutils.sourceforge.net/" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title>adapt.parameter_based.FineTuning — adapt 0.1.0 documentation</title>
<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/sg_gallery.css" type="text/css" />
<link rel="stylesheet" href="../_static/css/custom.css" type="text/css" />
<!--[if lt IE 9]>
<script src="../_static/js/html5shiv.min.js"></script>
<![endif]-->
<script data-url_root="../" id="documentation_options" src="../_static/documentation_options.js"></script>
<script src="../_static/jquery.js"></script>
<script src="../_static/underscore.js"></script>
<script src="../_static/doctools.js"></script>
<script crossorigin="anonymous" integrity="sha256-Ae2Vz/4ePdIu6ZyI/5ZGsYnb+m0JlOmKPjt6XZ9JJkA=" src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.4/require.min.js"></script>
<script src="../_static/js/custom.js"></script>
<script src="../_static/js/theme.js"></script>
<link rel="index" title="Index" href="../genindex.html" />
<link rel="search" title="Search" href="../search.html" />
</head>
<body class="wy-body-for-nav">
<div class="wy-grid-for-nav">
<nav data-toggle="wy-nav-shift" class="wy-nav-side">
<div class="wy-side-scroll">
<div class="wy-side-nav-search" >
<a href="../index.html" class="icon icon-home"> adapt
</a>
<div role="search">
<form id="rtd-search-form" class="wy-form" action="../search.html" method="get">
<input type="text" name="q" placeholder="Search docs" />
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
</div>
</div><div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="Navigation menu">
<p class="caption" role="heading"><span class="caption-text">Installation</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../install.html">Installation</a></li>
<li class="toctree-l1"><a class="reference internal" href="https://github.com/adapt-python/adapt">Github</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Tutorials</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../map.html">Choosing the right algorithm</a></li>
<li class="toctree-l1"><a class="reference internal" href="../examples/Quick_start.html">Quick-Start</a></li>
<li class="toctree-l1"><a class="reference internal" href="../examples/Developer_Guide.html">Developer Guide</a></li>
</ul>
<p class="caption" role="heading"><span class="caption-text">API Documentation</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../contents.html#adapt-feature-based">Feature-based</a><ul>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.PRED.html">PRED</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.FA.html">FA</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.CORAL.html">CORAL</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.SA.html">SA</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.TCA.html">TCA</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.fMMD.html">fMMD</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.DeepCORAL.html">DeepCORAL</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.DANN.html">DANN</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.ADDA.html">ADDA</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.WDGRL.html">WDGRL</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.CDAN.html">CDAN</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.MCD.html">MCD</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.MDD.html">MDD</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.feature_based.CCSA.html">CCSA</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../contents.html#adapt-instance-based">Instance-based</a><ul>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.LDM.html">LDM</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.KLIEP.html">KLIEP</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.KMM.html">KMM</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.ULSIF.html">ULSIF</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.RULSIF.html">RULSIF</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.NearestNeighborsWeighting.html">NearestNeighborsWeighting</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.IWC.html">IWC</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.IWN.html">IWN</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.BalancedWeighting.html">BalancedWeighting</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.TrAdaBoost.html">TrAdaBoost</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.TrAdaBoostR2.html">TrAdaBoostR2</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.TwoStageTrAdaBoostR2.html">TwoStageTrAdaBoostR2</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.instance_based.WANN.html">WANN</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../contents.html#adapt-parameter-based">Parameter-based</a><ul>
<li class="toctree-l2"><a class="reference internal" href="adapt.parameter_based.LinInt.html">LinInt</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.parameter_based.RegularTransferLR.html">RegularTransferLR</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.parameter_based.RegularTransferLC.html">RegularTransferLC</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.parameter_based.RegularTransferNN.html">RegularTransferNN</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.parameter_based.RegularTransferGP.html">RegularTransferGP</a></li>
<li class="toctree-l2"><a class="reference internal" href="#">FineTuning</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.parameter_based.TransferTreeClassifier.html">TransferTreeClassifier</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.parameter_based.TransferForestClassifier.html">TransferForestClassifier</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.parameter_based.TransferTreeSelector.html">TransferTreeSelector</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.parameter_based.TransferForestSelector.html">TransferForestSelector</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../contents.html#adapt-metrics">Metrics</a><ul>
<li class="toctree-l2"><a class="reference internal" href="adapt.metrics.make_uda_scorer.html">make_uda_scorer</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.metrics.cov_distance.html">cov_distance</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.metrics.neg_j_score.html">neg_j_score</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.metrics.linear_discrepancy.html">linear_discrepancy</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.metrics.normalized_linear_discrepancy.html">normalized_linear_discrepancy</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.metrics.frechet_distance.html">frechet_distance</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.metrics.normalized_frechet_distance.html">normalized_frechet_distance</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.metrics.domain_classifier.html">domain_classifier</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.metrics.reverse_validation.html">reverse_validation</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../contents.html#adapt-utils">Utility Functions</a><ul>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.UpdateLambda.html">UpdateLambda</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.accuracy.html">accuracy</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.check_arrays.html">check_arrays</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.check_estimator.html">check_estimator</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.check_network.html">check_network</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.get_default_encoder.html">get_default_encoder</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.get_default_task.html">get_default_task</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.get_default_discriminator.html">get_default_discriminator</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.GradientHandler.html">GradientHandler</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.make_classification_da.html">make_classification_da</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.make_regression_da.html">make_regression_da</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.check_sample_weight.html">check_sample_weight</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.set_random_seed.html">set_random_seed</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.check_fitted_estimator.html">check_fitted_estimator</a></li>
<li class="toctree-l2"><a class="reference internal" href="adapt.utils.check_fitted_network.html">check_fitted_network</a></li>
</ul>
</li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Synthetic Examples</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../examples/Classification.html">Classification</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/Classification.html#Experimental-Setup">Experimental Setup</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Classification.html#Src-Only">Src Only</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Classification.html#DANN">DANN</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Classification.html#Instance-Based">Instance Based</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../examples/Two_moons.html">Two Moons</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#Setup">Setup</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#Network">Network</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#Source-Only">Source Only</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#DANN">DANN</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#ADDA">ADDA</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#DeepCORAL">DeepCORAL</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#CORAL">CORAL</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#MCD">MCD</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#MDD">MDD</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#WDGRL">WDGRL</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Two_moons.html#CDAN">CDAN</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../examples/Rotation.html">Rotation</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/Rotation.html#Setup">Setup</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Rotation.html#Source-Only">Source Only</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Rotation.html#CORAL">CORAL</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Rotation.html#DANN">DANN</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Rotation.html#ADDA">ADDA</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../examples/Regression.html">Toy Regression</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/Regression.html#Experimental-Setup">Experimental Setup</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Regression.html#TGT-Only">TGT Only</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Regression.html#Src-Only">Src Only</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Regression.html#All">All</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Regression.html#CORAL">CORAL</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Regression.html#TrAdaBoostR2">TrAdaBoostR2</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Regression.html#RegularTransferNN">RegularTransferNN</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../examples/sample_bias.html">Sample Bias 1D</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/sample_bias.html#Setup">Setup</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/sample_bias.html#KMM">KMM</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/sample_bias.html#KLIEP">KLIEP</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../examples/sample_bias_2d.html">Sample Bias 2D</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/sample_bias_2d.html#Setup">Setup</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/sample_bias_2d.html#Estimator">Estimator</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/sample_bias_2d.html#Source-Only">Source Only</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/sample_bias_2d.html#KMM">KMM</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/sample_bias_2d.html#KLIEP">KLIEP</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../examples/Multi_fidelity.html">Multi-Fidelity</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/Multi_fidelity.html#Setup">Setup</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Multi_fidelity.html#Network">Network</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Multi_fidelity.html#Low-fidelity-only">Low fidelity only</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Multi_fidelity.html#High-fidelity-only">High fidelity only</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Multi_fidelity.html#RegularTransferNN">RegularTransferNN</a></li>
</ul>
</li>
</ul>
<p class="caption" role="heading"><span class="caption-text">Real Examples</span></p>
<ul>
<li class="toctree-l1"><a class="reference internal" href="../examples/Sample_bias_example.html">Sample Bias</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/Sample_bias_example.html#Sample-Bias-on-the-diabetes-dataset">Diabetes Dataset</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Sample_bias_example.html#Applying-Transfer-Learning-for-correcting-sample-bias">Sample bias correction</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Sample_bias_example.html#Hidden-bias-and-features-importance-estimation">Features importance</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../examples/Flowers_example.html">Fine-Tuning</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/Flowers_example.html#Training-a-model-from-scratch">Train from Scratch</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Flowers_example.html#Model-based-Transfer">Model-based Transfer</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../examples/Office_example.html">Deep Domain Adaptation</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/Office_example.html#Dataset-Preprocessing">Dataset Preprocessing</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Office_example.html#Fit-without-adaptation">Fit without adaptation</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Office_example.html#Fit-with-adaptation">Fit with adaptation</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../examples/tradaboost_experiments.html">TrAdaBoost Experiments</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/tradaboost_experiments.html#Mushrooms">Mushrooms</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/tradaboost_experiments.html#20-NewsGroup">20-NewsGroup</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../examples/Heart_Failure.html">Heart Failure</a><ul>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#Setup">Setup</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#Network">Network</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#Source-Only">Source Only</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#DANN">DANN</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#ADDA">ADDA</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#DeepCORAL">DeepCORAL</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#CORAL">CORAL</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#MCD">MCD</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#MDD">MDD</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#WDGRL">WDGRL</a></li>
<li class="toctree-l2"><a class="reference internal" href="../examples/Heart_Failure.html#CDAN">CDAN</a></li>
</ul>
</li>
</ul>
</div>
</div>
</nav>
<section data-toggle="wy-nav-shift" class="wy-nav-content-wrap"><nav class="wy-nav-top" aria-label="Mobile navigation menu" >
<i data-toggle="wy-nav-top" class="fa fa-bars"></i>
<a href="../index.html">adapt</a>
</nav>
<div class="wy-nav-content">
<div class="rst-content">
<div role="navigation" aria-label="Page navigation">
<ul class="wy-breadcrumbs">
<li><a href="../index.html" class="icon icon-home"></a> »</li>
<li><span class="xref std std-ref">adapt.parameter_based</span>.FineTuning</li>
<li class="wy-breadcrumbs-aside">
</li>
</ul>
<hr/>
</div>
<div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
<div itemprop="articleBody">
<style>
/* CSS overrides for sphinx_rtd_theme */
/* 24px margin */
.nbinput.nblast.container,
.nboutput.nblast.container {
margin-bottom: 19px; /* padding has already 5px */
}
/* ... except between code cells! */
.nblast.container + .nbinput.container {
margin-top: -19px;
}
.admonition > p:before {
margin-right: 4px; /* make room for the exclamation icon */
}
/* Fix math alignment, see https://github.com/rtfd/sphinx_rtd_theme/pull/686 */
.math {
text-align: unset;
}
</style>
<section id="adapt-parameter-based-finetuning">
<h1><a class="reference internal" href="../contents.html#adapt-parameter-based"><span class="std std-ref">adapt.parameter_based</span></a>.FineTuning<a class="headerlink" href="#adapt-parameter-based-finetuning" title="Permalink to this headline"></a></h1>
<dl class="py class">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning">
<em class="property"><span class="pre">class</span><span class="w"> </span></em><span class="sig-prename descclassname"><span class="pre">adapt.parameter_based.</span></span><span class="sig-name descname"><span class="pre">FineTuning</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">encoder</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">task</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">Xt</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">yt</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">training</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pretrain</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">verbose</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">1</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">copy</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">random_state</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">params</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning" title="Permalink to this definition"></a></dt>
<dd><p>FineTuning : finetunes pretrained networks on target data.</p>
<p>A pretrained source encoder should be given. A task network,
pretrained or not, should be given too.</p>
<p>Finetuning train both networks. The task network can be
fitted first using the <code class="docutils literal notranslate"><span class="pre">pretrain</span></code> parameter. The layers
to train in the encoder can be set via the parameter <code class="docutils literal notranslate"><span class="pre">training</span></code>.</p>
<figure class="align-center" id="id3">
<img alt="../_images/regulartransfer.png" src="../_images/regulartransfer.png" />
<figcaption>
<p><span class="caption-text">Transferring parameters of a CNN pretrained on Imagenet
(source: [1])</span><a class="headerlink" href="#id3" title="Permalink to this image"></a></p>
</figcaption>
</figure>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>encoder</strong><span class="classifier">tensorflow Model (default=None)</span></dt><dd><p>Encoder netwok. If <code class="docutils literal notranslate"><span class="pre">None</span></code>, a shallow network with 10
neurons and ReLU activation is used as encoder network.</p>
</dd>
<dt><strong>task</strong><span class="classifier">tensorflow Model (default=None)</span></dt><dd><p>Task netwok. If <code class="docutils literal notranslate"><span class="pre">None</span></code>, a two layers network with 10
neurons per layer and ReLU activation is used as task network.</p>
</dd>
<dt><strong>Xt</strong><span class="classifier">numpy array (default=None)</span></dt><dd><p>Target input data.</p>
</dd>
<dt><strong>yt</strong><span class="classifier">numpy array (default=None)</span></dt><dd><p>Target output data.</p>
</dd>
<dt><strong>training</strong><span class="classifier">bool or list of bool, optional (default=True)</span></dt><dd><p>Weither to train the <code class="docutils literal notranslate"><span class="pre">encoder</span></code> or not.
If a list is given, values from <code class="docutils literal notranslate"><span class="pre">training</span></code> are assigned
successively to the <code class="docutils literal notranslate"><span class="pre">trainable</span></code> attribute of the
<code class="docutils literal notranslate"><span class="pre">encoder</span></code> layers going from the last layer to the first one.
If the length of <code class="docutils literal notranslate"><span class="pre">training</span></code> is smaller than the length of
the <code class="docutils literal notranslate"><span class="pre">encoder</span></code> layers list, the last value of <code class="docutils literal notranslate"><span class="pre">training</span></code>
will be asigned to the remaining layers.</p>
</dd>
<dt><strong>pretrain</strong><span class="classifier">bool (default=False)</span></dt><dd><p>If True, the task network is first trained alone on the outputs
of the encoder.</p>
</dd>
<dt><strong>copy</strong><span class="classifier">boolean (default=True)</span></dt><dd><p>Whether to make a copy of <code class="docutils literal notranslate"><span class="pre">estimator</span></code> or not.</p>
</dd>
<dt><strong>verbose</strong><span class="classifier">int (default=1)</span></dt><dd><p>Verbosity level.</p>
</dd>
<dt><strong>random_state</strong><span class="classifier">int (default=None)</span></dt><dd><p>Seed of random generator.</p>
</dd>
<dt><strong>params</strong><span class="classifier">key, value arguments</span></dt><dd><p>Arguments given at the different level of the adapt object.
It can be, for instance, compile or fit parameters of the
estimator or kernel parameters etc…
Accepted parameters can be found by calling the method
<code class="docutils literal notranslate"><span class="pre">_get_legal_params(params)</span></code>.</p>
</dd>
</dl>
</dd>
<dt class="field-even">Yields</dt>
<dd class="field-even"><dl class="simple">
<dt><strong>optimizer</strong><span class="classifier">str or instance of tf.keras.optimizers (default=”rmsprop”)</span></dt><dd><p>Optimizer for the task. It should be an
instance of tf.keras.optimizers as:
<code class="docutils literal notranslate"><span class="pre">tf.keras.optimizers.SGD(0.001)</span></code> or
<code class="docutils literal notranslate"><span class="pre">tf.keras.optimizers.Adam(lr=0.001,</span> <span class="pre">beta_1=0.5)</span></code>.
A string can also be given as <code class="docutils literal notranslate"><span class="pre">"adam"</span></code>.
Default optimizer is <code class="docutils literal notranslate"><span class="pre">rmsprop</span></code>.</p>
</dd>
<dt><strong>loss</strong><span class="classifier">str or instance of tf.keras.losses (default=”mse”)</span></dt><dd><p>Loss for the task. It should be an
instance of tf.keras.losses as:
<code class="docutils literal notranslate"><span class="pre">tf.keras.losses.MeanSquaredError()</span></code> or
<code class="docutils literal notranslate"><span class="pre">tf.keras.losses.CategoricalCrossentropy()</span></code>.
A string can also be given as <code class="docutils literal notranslate"><span class="pre">"mse"</span></code> or
<code class="docutils literal notranslate"><span class="pre">categorical_crossentropy</span></code>.
Default loss is <code class="docutils literal notranslate"><span class="pre">mse</span></code>.</p>
</dd>
<dt><strong>metrics</strong><span class="classifier">list of str or list of tf.keras.metrics.Metric instance</span></dt><dd><p>List of metrics to be evaluated by the model during training
and testing. Typically you will use <code class="docutils literal notranslate"><span class="pre">metrics=['accuracy']</span></code>.</p>
</dd>
<dt><strong>optimizer_enc</strong><span class="classifier">str or instance of tf.keras.optimizers</span></dt><dd><p>If the Adapt Model has an <code class="docutils literal notranslate"><span class="pre">encoder</span></code> attribute,
a specific optimizer for the <code class="docutils literal notranslate"><span class="pre">encoder</span></code> network can
be given. Typically, this parameter can be used to
give a smaller learning rate to the encoder.
If not specified, <code class="docutils literal notranslate"><span class="pre">optimizer_enc=optimizer</span></code>.</p>
</dd>
<dt><strong>optimizer_disc</strong><span class="classifier">str or instance of tf.keras.optimizers</span></dt><dd><p>If the Adapt Model has a <code class="docutils literal notranslate"><span class="pre">discriminator</span></code> attribute,
a specific optimizer for the <code class="docutils literal notranslate"><span class="pre">discriminator</span></code> network can
be given. If not specified, <code class="docutils literal notranslate"><span class="pre">optimizer_disc=optimizer</span></code>.</p>
</dd>
<dt><strong>kwargs</strong><span class="classifier">key, value arguments</span></dt><dd><p>Any arguments of the <code class="docutils literal notranslate"><span class="pre">fit</span></code> method from the Tensorflow
Model can be given, as <code class="docutils literal notranslate"><span class="pre">epochs</span></code> and <code class="docutils literal notranslate"><span class="pre">batch_size</span></code>.
Specific arguments from <code class="docutils literal notranslate"><span class="pre">optimizer</span></code> can also be given
as <code class="docutils literal notranslate"><span class="pre">learning_rate</span></code> or <code class="docutils literal notranslate"><span class="pre">beta_1</span></code> for <code class="docutils literal notranslate"><span class="pre">Adam</span></code>.
This allows to perform <code class="docutils literal notranslate"><span class="pre">GridSearchCV</span></code> from scikit-learn
on these arguments.</p>
</dd>
</dl>
</dd>
</dl>
<p class="rubric">References</p>
<dl class="citation">
<dt class="label" id="r8f631e03c5d8-1"><span class="brackets">1</span></dt>
<dd><p><a class="reference external" href="https://hal.inria.fr/hal-00911179v1/document">[1]</a> Oquab M., Bottou L., Laptev I., Sivic J. “Learning and transferring mid-level image representations using convolutional neural networks”. In CVPR, 2014.</p>
</dd>
</dl>
<p class="rubric">Examples</p>
<div class="doctest highlight-default notranslate"><div class="highlight"><pre><span></span><span class="gp">>>> </span><span class="kn">from</span> <span class="nn">adapt.utils</span> <span class="kn">import</span> <span class="n">make_regression_da</span>
<span class="gp">>>> </span><span class="kn">from</span> <span class="nn">adapt.parameter_based</span> <span class="kn">import</span> <span class="n">FineTuning</span>
<span class="gp">>>> </span><span class="n">Xs</span><span class="p">,</span> <span class="n">ys</span><span class="p">,</span> <span class="n">Xt</span><span class="p">,</span> <span class="n">yt</span> <span class="o">=</span> <span class="n">make_regression_da</span><span class="p">()</span>
<span class="gp">>>> </span><span class="n">src_model</span> <span class="o">=</span> <span class="n">FineTuning</span><span class="p">(</span><span class="n">loss</span><span class="o">=</span><span class="s2">"mse"</span><span class="p">,</span> <span class="n">pretrain</span><span class="o">=</span><span class="kc">False</span><span class="p">,</span> <span class="n">random_state</span><span class="o">=</span><span class="mi">0</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">src_model</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">Xs</span><span class="p">,</span> <span class="n">ys</span><span class="p">,</span> <span class="n">epochs</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span> <span class="n">verbose</span><span class="o">=</span><span class="mi">0</span><span class="p">)</span>
<span class="gp">>>> </span><span class="nb">print</span><span class="p">(</span><span class="n">src_model</span><span class="o">.</span><span class="n">score</span><span class="p">(</span><span class="n">Xt</span><span class="p">,</span> <span class="n">yt</span><span class="p">))</span>
<span class="go">1/1 [==============================] - 0s 82ms/step - loss: 0.2645</span>
<span class="go">0.26447802782058716</span>
<span class="gp">>>> </span><span class="n">tgt_model</span> <span class="o">=</span> <span class="n">FineTuning</span><span class="p">(</span><span class="n">encoder</span><span class="o">=</span><span class="n">src_model</span><span class="o">.</span><span class="n">encoder_</span><span class="p">,</span> <span class="n">loss</span><span class="o">=</span><span class="s2">"mse"</span><span class="p">,</span> <span class="n">pretrain</span><span class="o">=</span><span class="kc">True</span><span class="p">,</span>
<span class="gp">... </span> <span class="n">pretrain__epochs</span><span class="o">=</span><span class="mi">30</span><span class="p">,</span> <span class="n">random_state</span><span class="o">=</span><span class="mi">0</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">tgt_model</span><span class="o">.</span><span class="n">fit</span><span class="p">(</span><span class="n">Xt</span><span class="p">[:</span><span class="mi">3</span><span class="p">],</span> <span class="n">yt</span><span class="p">[:</span><span class="mi">3</span><span class="p">],</span> <span class="n">epochs</span><span class="o">=</span><span class="mi">100</span><span class="p">,</span> <span class="n">verbose</span><span class="o">=</span><span class="mi">0</span><span class="p">)</span>
<span class="gp">>>> </span><span class="n">tgt_model</span><span class="o">.</span><span class="n">score</span><span class="p">(</span><span class="n">Xt</span><span class="p">,</span> <span class="n">yt</span><span class="p">)</span>
<span class="go">1/1 [==============================] - 0s 78ms/step - loss: 0.1114</span>
<span class="go">0.11144775152206421</span>
</pre></div>
</div>
<dl class="field-list simple">
<dt class="field-odd">Attributes</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>encoder_</strong><span class="classifier">tensorflow Model</span></dt><dd><p>encoder network.</p>
</dd>
<dt><strong>task_</strong><span class="classifier">tensorflow Model</span></dt><dd><p>Network.</p>
</dd>
<dt><strong>history_</strong><span class="classifier">dict</span></dt><dd><p>history of the losses and metrics across the epochs
of the network training.</p>
</dd>
</dl>
</dd>
</dl>
<p class="rubric">Methods</p>
<table class="longtable docutils align-default">
<colgroup>
<col style="width: 10%" />
<col style="width: 90%" />
</colgroup>
<tbody>
<tr class="row-odd"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.__init__" title="adapt.parameter_based.FineTuning.__init__"><code class="xref py py-obj docutils literal notranslate"><span class="pre">__init__</span></code></a>([encoder, task, Xt, yt, training, ...])</p></td>
<td><p></p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.compile" title="adapt.parameter_based.FineTuning.compile"><code class="xref py py-obj docutils literal notranslate"><span class="pre">compile</span></code></a>([optimizer, loss, metrics, ...])</p></td>
<td><p>Configures the model for training.</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.fit" title="adapt.parameter_based.FineTuning.fit"><code class="xref py py-obj docutils literal notranslate"><span class="pre">fit</span></code></a>([Xt, yt])</p></td>
<td><p>Fit FineTuning.</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.get_params" title="adapt.parameter_based.FineTuning.get_params"><code class="xref py py-obj docutils literal notranslate"><span class="pre">get_params</span></code></a>([deep])</p></td>
<td><p>Get parameters for this estimator.</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.load_weights" title="adapt.parameter_based.FineTuning.load_weights"><code class="xref py py-obj docutils literal notranslate"><span class="pre">load_weights</span></code></a>(filepath[, skip_mismatch, ...])</p></td>
<td><p>Loads all layer weights from a saved files.</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.predict" title="adapt.parameter_based.FineTuning.predict"><code class="xref py py-obj docutils literal notranslate"><span class="pre">predict</span></code></a>(x[, batch_size, verbose, steps, ...])</p></td>
<td><p>Generates output predictions for the input samples.</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.predict_disc" title="adapt.parameter_based.FineTuning.predict_disc"><code class="xref py py-obj docutils literal notranslate"><span class="pre">predict_disc</span></code></a>(X)</p></td>
<td><p>Not used.</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.predict_task" title="adapt.parameter_based.FineTuning.predict_task"><code class="xref py py-obj docutils literal notranslate"><span class="pre">predict_task</span></code></a>(X)</p></td>
<td><p>Return predictions of the task on the encoded features.</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.pretrain_step" title="adapt.parameter_based.FineTuning.pretrain_step"><code class="xref py py-obj docutils literal notranslate"><span class="pre">pretrain_step</span></code></a>(data)</p></td>
<td><p></p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.save_weights" title="adapt.parameter_based.FineTuning.save_weights"><code class="xref py py-obj docutils literal notranslate"><span class="pre">save_weights</span></code></a>(filepath[, overwrite, ...])</p></td>
<td><p>Saves all layer weights.</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.score" title="adapt.parameter_based.FineTuning.score"><code class="xref py py-obj docutils literal notranslate"><span class="pre">score</span></code></a>(X, y[, sample_weight])</p></td>
<td><p>Return the evaluation of the model on X, y.</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.set_params" title="adapt.parameter_based.FineTuning.set_params"><code class="xref py py-obj docutils literal notranslate"><span class="pre">set_params</span></code></a>(**params)</p></td>
<td><p>Set the parameters of this estimator.</p></td>
</tr>
<tr class="row-odd"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.transform" title="adapt.parameter_based.FineTuning.transform"><code class="xref py py-obj docutils literal notranslate"><span class="pre">transform</span></code></a>(X)</p></td>
<td><p>Return the encoded features of X.</p></td>
</tr>
<tr class="row-even"><td><p><a class="reference internal" href="#adapt.parameter_based.FineTuning.unsupervised_score" title="adapt.parameter_based.FineTuning.unsupervised_score"><code class="xref py py-obj docutils literal notranslate"><span class="pre">unsupervised_score</span></code></a>(Xs, Xt)</p></td>
<td><p>Return unsupervised score.</p></td>
</tr>
</tbody>
</table>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.__init__">
<span class="sig-name descname"><span class="pre">__init__</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">encoder</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">task</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">Xt</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">yt</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">training</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">pretrain</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">verbose</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">1</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">copy</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">random_state</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">params</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.__init__.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.__init__" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.compile">
<span class="sig-name descname"><span class="pre">compile</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">optimizer</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">loss</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">metrics</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">loss_weights</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">weighted_metrics</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">run_eagerly</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">steps_per_execution</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">kwargs</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.compile.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.compile" title="Permalink to this definition"></a></dt>
<dd><p>Configures the model for training.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>optimizer: str or `tf.keras.optimizer` instance</strong></dt><dd><p>Optimizer</p>
</dd>
<dt><strong>loss: str or `tf.keras.losses.Loss` instance</strong></dt><dd><p>Loss function. A loss function is any callable
with the signature <cite>loss = fn(y_true, y_pred)</cite>,
where <cite>y_true</cite> are the ground truth values, and
<cite>y_pred</cite> are the model’s predictions.
<cite>y_true</cite> should have shape
<cite>(batch_size, d0, .. dN)</cite> (except in the case of
sparse loss functions such as
sparse categorical crossentropy which expects integer arrays of shape
<cite>(batch_size, d0, .. dN-1)</cite>).
<cite>y_pred</cite> should have shape <cite>(batch_size, d0, .. dN)</cite>.
The loss function should return a float tensor.
If a custom <cite>Loss</cite> instance is
used and reduction is set to <cite>None</cite>, return value has shape
<cite>(batch_size, d0, .. dN-1)</cite> i.e. per-sample or per-timestep loss
values; otherwise, it is a scalar. If the model has multiple outputs,
you can use a different loss on each output by passing a dictionary
or a list of losses. The loss value that will be minimized by the
model will then be the sum of all individual losses, unless
<cite>loss_weights</cite> is specified.</p>
</dd>
<dt><strong>metrics: list of str or list of `tf.keras.metrics.Metric` instance</strong></dt><dd><p>List of metrics to be evaluated by the model during training
and testing. Typically you will use <cite>metrics=[‘accuracy’]</cite>. A
function is any callable with the signature <cite>result = fn(y_true,
y_pred)</cite>. To specify different metrics for different outputs of a
multi-output model, you could also pass a dictionary, such as
<cite>metrics={‘output_a’: ‘accuracy’, ‘output_b’: [‘accuracy’, ‘mse’]}</cite>.
You can also pass a list to specify a metric or a list of metrics
for each output, such as <cite>metrics=[[‘accuracy’], [‘accuracy’, ‘mse’]]</cite>
or <cite>metrics=[‘accuracy’, [‘accuracy’, ‘mse’]]</cite>. When you pass the
strings ‘accuracy’ or ‘acc’, we convert this to one of
<cite>tf.keras.metrics.BinaryAccuracy</cite>,
<cite>tf.keras.metrics.CategoricalAccuracy</cite>,
<cite>tf.keras.metrics.SparseCategoricalAccuracy</cite> based on the loss
function used and the model output shape. We do a similar
conversion for the strings ‘crossentropy’ and ‘ce’ as well.</p>
</dd>
<dt><strong>loss_weights: List or dict of floats</strong></dt><dd><p>Scalars to weight the loss contributions of different model
outputs. The loss value that will be minimized by the model will then
be the <em>weighted sum</em> of all individual losses, weighted by the
<cite>loss_weights</cite> coefficients.
If a list, it is expected to have a 1:1 mapping to the model’s
outputs. If a dict, it is expected to map output names (strings)
to scalar coefficients.</p>
</dd>
<dt><strong>weighted_metrics: list of metrics</strong></dt><dd><p>List of metrics to be evaluated and weighted by
<cite>sample_weight</cite> or <cite>class_weight</cite> during training and testing.</p>
</dd>
<dt><strong>run_eagerly: bool (default=False)</strong></dt><dd><p>If <cite>True</cite>, this <cite>Model</cite>’s logic will not be wrapped
in a <cite>tf.function</cite>. Recommended to leave
this as <cite>None</cite> unless your <cite>Model</cite> cannot be run inside a
<cite>tf.function</cite>. <cite>run_eagerly=True</cite> is not supported when using
<cite>tf.distribute.experimental.ParameterServerStrategy</cite>.</p>
</dd>
<dt><strong>steps_per_execution: int (default=1)</strong></dt><dd><p>The number of batches to run during each
<cite>tf.function</cite> call. Running multiple batches
inside a single <cite>tf.function</cite> call can greatly improve performance
on TPUs or small models with a large Python overhead.
At most, one full epoch will be run each
execution. If a number larger than the size of the epoch is passed,
the execution will be truncated to the size of the epoch.
Note that if <cite>steps_per_execution</cite> is set to <cite>N</cite>,
<cite>Callback.on_batch_begin</cite> and <cite>Callback.on_batch_end</cite> methods
will only be called every <cite>N</cite> batches
(i.e. before/after each <cite>tf.function</cite> execution).</p>
</dd>
<dt><strong>**kwargs: key, value arguments</strong></dt><dd><p>Arguments supported for backwards compatibility only.</p>
</dd>
</dl>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><dl class="simple">
<dt>None: None</dt><dd></dd>
</dl>
</dd>
</dl>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.fit">
<span class="sig-name descname"><span class="pre">fit</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">Xt</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">yt</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">fit_params</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.fit.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.fit" title="Permalink to this definition"></a></dt>
<dd><p>Fit FineTuning.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>Xt</strong><span class="classifier">numpy array (default=None)</span></dt><dd><p>Target input data.</p>
</dd>
<dt><strong>yt</strong><span class="classifier">numpy array (default=None)</span></dt><dd><p>Target output data.</p>
</dd>
<dt><strong>fit_params</strong><span class="classifier">key, value arguments</span></dt><dd><p>Arguments given to the fit method of the model
(epochs, batch_size, callbacks…).</p>
</dd>
</dl>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><dl class="simple">
<dt><strong>self</strong><span class="classifier">returns an instance of self</span></dt><dd></dd>
</dl>
</dd>
</dl>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.get_params">
<span class="sig-name descname"><span class="pre">get_params</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">deep</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.get_params.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.get_params" title="Permalink to this definition"></a></dt>
<dd><p>Get parameters for this estimator.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>deep</strong><span class="classifier">bool, default=True</span></dt><dd><p>Not used, here for scikit-learn compatibility.</p>
</dd>
</dl>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><dl class="simple">
<dt><strong>params</strong><span class="classifier">dict</span></dt><dd><p>Parameter names mapped to their values.</p>
</dd>
</dl>
</dd>
</dl>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.load_weights">
<span class="sig-name descname"><span class="pre">load_weights</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">filepath</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">skip_mismatch</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">by_name</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">options</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.load_weights.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.load_weights" title="Permalink to this definition"></a></dt>
<dd><p>Loads all layer weights from a saved files.</p>
<p>The saved file could be a SavedModel file, a <cite>.keras</cite> file (v3 saving
format), or a file created via <cite>model.save_weights()</cite>.</p>
<p>By default, weights are loaded based on the network’s
topology. This means the architecture should be the same as when the
weights were saved. Note that layers that don’t have weights are not
taken into account in the topological ordering, so adding or removing
layers is fine as long as they don’t have weights.</p>
<p><strong>Partial weight loading</strong></p>
<p>If you have modified your model, for instance by adding a new layer
(with weights) or by changing the shape of the weights of a layer,
you can choose to ignore errors and continue loading
by setting <cite>skip_mismatch=True</cite>. In this case any layer with
mismatching weights will be skipped. A warning will be displayed
for each skipped layer.</p>
<p><strong>Weight loading by name</strong></p>
<p>If your weights are saved as a <cite>.h5</cite> file created
via <cite>model.save_weights()</cite>, you can use the argument <cite>by_name=True</cite>.</p>
<p>In this case, weights are loaded into layers only if they share
the same name. This is useful for fine-tuning or transfer-learning
models where some of the layers have changed.</p>
<p>Note that only topological loading (<cite>by_name=False</cite>) is supported when
loading weights from the <cite>.keras</cite> v3 format or from the TensorFlow
SavedModel format.</p>
<dl class="simple">
<dt>Args:</dt><dd><dl class="simple">
<dt>filepath: String, path to the weights file to load. For weight files</dt><dd><p>in TensorFlow format, this is the file prefix (the same as was
passed to <cite>save_weights()</cite>). This can also be a path to a
SavedModel or a <cite>.keras</cite> file (v3 saving format) saved
via <cite>model.save()</cite>.</p>
</dd>
<dt>skip_mismatch: Boolean, whether to skip loading of layers where</dt><dd><p>there is a mismatch in the number of weights, or a mismatch in
the shape of the weights.</p>
</dd>
<dt>by_name: Boolean, whether to load weights by name or by topological</dt><dd><p>order. Only topological loading is supported for weight files in
the <cite>.keras</cite> v3 format or in the TensorFlow SavedModel format.</p>
</dd>
<dt>options: Optional <cite>tf.train.CheckpointOptions</cite> object that specifies</dt><dd><p>options for loading weights (only valid for a SavedModel file).</p>
</dd>
</dl>
</dd>
</dl>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.predict">
<span class="sig-name descname"><span class="pre">predict</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">x</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">batch_size</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">verbose</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">0</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">steps</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">callbacks</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">max_queue_size</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">10</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">workers</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">1</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">use_multiprocessing</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">False</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.predict.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.predict" title="Permalink to this definition"></a></dt>
<dd><p>Generates output predictions for the input samples.</p>
<p>Computation is done in batches. This method is designed for performance in
large scale inputs. For small amount of inputs that fit in one batch,
directly using <cite>__call__()</cite> is recommended for faster execution, e.g.,
<cite>model(x)</cite>, or <cite>model(x, training=False)</cite> if you have layers such as
<cite>tf.keras.layers.BatchNormalization</cite> that behaves differently during
inference. Also, note the fact that test loss is not affected by
regularization layers like noise and dropout.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>x: array</strong></dt><dd><p>Input samples.</p>
</dd>
<dt><strong>batch_size: int (default=`None`)</strong></dt><dd><p>Number of samples per batch.
If unspecified, <cite>batch_size</cite> will default to 32.
Do not specify the <cite>batch_size</cite> if your data is in the
form of dataset, generators, or <cite>keras.utils.Sequence</cite> instances
(since they generate batches).</p>
</dd>
<dt><strong>verbose: int (default=0)</strong></dt><dd><p>Verbosity mode, 0 or 1.</p>
</dd>
<dt><strong>steps: int (default=None)</strong></dt><dd><p>Total number of steps (batches of samples)
before declaring the prediction round finished.
Ignored with the default value of <cite>None</cite>. If x is a <cite>tf.data</cite>
dataset and <cite>steps</cite> is None, <cite>predict()</cite> will
run until the input dataset is exhausted.</p>
</dd>
<dt><strong>callbacks: List of `keras.callbacks.Callback` instances.</strong></dt><dd><p>List of callbacks to apply during prediction.
See [callbacks](/api_docs/python/tf/keras/callbacks).</p>
</dd>
<dt><strong>max_queue_size: int (default=10)</strong></dt><dd><p>Used for generator or <cite>keras.utils.Sequence</cite>
input only. Maximum size for the generator queue.
If unspecified, <cite>max_queue_size</cite> will default to 10.</p>
</dd>
<dt><strong>workers: int (default=1)</strong></dt><dd><p>Used for generator or <cite>keras.utils.Sequence</cite> input
only. Maximum number of processes to spin up when using
process-based threading. If unspecified, <cite>workers</cite> will default
to 1.</p>
</dd>
<dt><strong>use_multiprocessing: bool (default=False)</strong></dt><dd><p>Used for generator or <cite>keras.utils.Sequence</cite> input only.
If <cite>True</cite>, use process-based
threading. If unspecified, <cite>use_multiprocessing</cite> will default to
<cite>False</cite>. Note that because this implementation relies on
multiprocessing, you should not pass non-picklable arguments to
the generator as they can’t be passed easily to children processes.</p>
</dd>
</dl>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><dl class="simple">
<dt><strong>y_pred</strong><span class="classifier">array</span></dt><dd><p>Numpy array(s) of predictions.</p>
</dd>
</dl>
</dd>
</dl>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.predict_disc">
<span class="sig-name descname"><span class="pre">predict_disc</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">X</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.predict_disc.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.predict_disc" title="Permalink to this definition"></a></dt>
<dd><p>Not used.</p>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.predict_task">
<span class="sig-name descname"><span class="pre">predict_task</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">X</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.predict_task.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.predict_task" title="Permalink to this definition"></a></dt>
<dd><p>Return predictions of the task on the encoded features.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>X</strong><span class="classifier">array</span></dt><dd><p>input data</p>
</dd>
</dl>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><dl class="simple">
<dt><strong>y_task</strong><span class="classifier">array</span></dt><dd><p>predictions of task network</p>
</dd>
</dl>
</dd>
</dl>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.pretrain_step">
<span class="sig-name descname"><span class="pre">pretrain_step</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">data</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.pretrain_step.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.pretrain_step" title="Permalink to this definition"></a></dt>
<dd></dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.save_weights">
<span class="sig-name descname"><span class="pre">save_weights</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">filepath</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">overwrite</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">True</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">save_format</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">options</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.save_weights.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.save_weights" title="Permalink to this definition"></a></dt>
<dd><p>Saves all layer weights.</p>
<p>Either saves in HDF5 or in TensorFlow format based on the <cite>save_format</cite>
argument.</p>
<dl class="simple">
<dt>When saving in HDF5 format, the weight file has:</dt><dd><ul class="simple">
<li><dl class="simple">
<dt><cite>layer_names</cite> (attribute), a list of strings</dt><dd><p>(ordered names of model layers).</p>
</dd>
</dl>
</li>
<li><dl class="simple">
<dt>For every layer, a <cite>group</cite> named <cite>layer.name</cite></dt><dd><ul>
<li><dl class="simple">
<dt>For every such layer group, a group attribute <cite>weight_names</cite>,</dt><dd><p>a list of strings
(ordered names of weights tensor of the layer).</p>
</dd>
</dl>
</li>
<li><dl class="simple">
<dt>For every weight in the layer, a dataset</dt><dd><p>storing the weight value, named after the weight tensor.</p>
</dd>
</dl>
</li>
</ul>
</dd>
</dl>
</li>
</ul>
</dd>
</dl>
<p>When saving in TensorFlow format, all objects referenced by the network
are saved in the same format as <cite>tf.train.Checkpoint</cite>, including any
<cite>Layer</cite> instances or <cite>Optimizer</cite> instances assigned to object
attributes. For networks constructed from inputs and outputs using
<cite>tf.keras.Model(inputs, outputs)</cite>, <cite>Layer</cite> instances used by the network
are tracked/saved automatically. For user-defined classes which inherit
from <cite>tf.keras.Model</cite>, <cite>Layer</cite> instances must be assigned to object
attributes, typically in the constructor. See the documentation of
<cite>tf.train.Checkpoint</cite> and <cite>tf.keras.Model</cite> for details.</p>
<p>While the formats are the same, do not mix <cite>save_weights</cite> and
<cite>tf.train.Checkpoint</cite>. Checkpoints saved by <cite>Model.save_weights</cite> should
be loaded using <cite>Model.load_weights</cite>. Checkpoints saved using
<cite>tf.train.Checkpoint.save</cite> should be restored using the corresponding
<cite>tf.train.Checkpoint.restore</cite>. Prefer <cite>tf.train.Checkpoint</cite> over
<cite>save_weights</cite> for training checkpoints.</p>
<p>The TensorFlow format matches objects and variables by starting at a
root object, <cite>self</cite> for <cite>save_weights</cite>, and greedily matching attribute
names. For <cite>Model.save</cite> this is the <cite>Model</cite>, and for <cite>Checkpoint.save</cite>
this is the <cite>Checkpoint</cite> even if the <cite>Checkpoint</cite> has a model attached.
This means saving a <cite>tf.keras.Model</cite> using <cite>save_weights</cite> and loading
into a <cite>tf.train.Checkpoint</cite> with a <cite>Model</cite> attached (or vice versa)
will not match the <cite>Model</cite>’s variables. See the
[guide to training checkpoints](
<a class="reference external" href="https://www.tensorflow.org/guide/checkpoint">https://www.tensorflow.org/guide/checkpoint</a>) for details on
the TensorFlow format.</p>
<dl class="simple">
<dt>Args:</dt><dd><dl class="simple">
<dt>filepath: String or PathLike, path to the file to save the weights</dt><dd><p>to. When saving in TensorFlow format, this is the prefix used
for checkpoint files (multiple files are generated). Note that
the ‘.h5’ suffix causes weights to be saved in HDF5 format.</p>
</dd>
<dt>overwrite: Whether to silently overwrite any existing file at the</dt><dd><p>target location, or provide the user with a manual prompt.</p>
</dd>
<dt>save_format: Either ‘tf’ or ‘h5’. A <cite>filepath</cite> ending in ‘.h5’ or</dt><dd><p>‘.keras’ will default to HDF5 if <cite>save_format</cite> is <cite>None</cite>.
Otherwise, <cite>None</cite> becomes ‘tf’. Defaults to <cite>None</cite>.</p>
</dd>
<dt>options: Optional <cite>tf.train.CheckpointOptions</cite> object that specifies</dt><dd><p>options for saving weights.</p>
</dd>
</dl>
</dd>
<dt>Raises:</dt><dd><dl class="simple">
<dt>ImportError: If <cite>h5py</cite> is not available when attempting to save in</dt><dd><p>HDF5 format.</p>
</dd>
</dl>
</dd>
</dl>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.score">
<span class="sig-name descname"><span class="pre">score</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">X</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">y</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">sample_weight</span></span><span class="o"><span class="pre">=</span></span><span class="default_value"><span class="pre">None</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.score.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.score" title="Permalink to this definition"></a></dt>
<dd><p>Return the evaluation of the model on X, y.</p>
<p>Call <cite>evaluate</cite> on tensorflow Model.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>X</strong><span class="classifier">array</span></dt><dd><p>input data</p>
</dd>
<dt><strong>y</strong><span class="classifier">array</span></dt><dd><p>output data</p>
</dd>
<dt><strong>sample_weight</strong><span class="classifier">array (default=None)</span></dt><dd><p>Sample weights</p>
</dd>
</dl>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><dl class="simple">
<dt><strong>score</strong><span class="classifier">float</span></dt><dd><p>Score.</p>
</dd>
</dl>
</dd>
</dl>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.set_params">
<span class="sig-name descname"><span class="pre">set_params</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="o"><span class="pre">**</span></span><span class="n"><span class="pre">params</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.set_params.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.set_params" title="Permalink to this definition"></a></dt>
<dd><p>Set the parameters of this estimator.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>**params</strong><span class="classifier">dict</span></dt><dd><p>Estimator parameters.</p>
</dd>
</dl>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><dl class="simple">
<dt><strong>self</strong><span class="classifier">estimator instance</span></dt><dd><p>Estimator instance.</p>
</dd>
</dl>
</dd>
</dl>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.transform">
<span class="sig-name descname"><span class="pre">transform</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">X</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.transform.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.transform" title="Permalink to this definition"></a></dt>
<dd><p>Return the encoded features of X.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>X</strong><span class="classifier">array</span></dt><dd><p>input data</p>
</dd>
</dl>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><dl class="simple">
<dt><strong>X_enc</strong><span class="classifier">array</span></dt><dd><p>predictions of encoder network</p>
</dd>
</dl>
</dd>
</dl>
</dd></dl>
<dl class="py method">
<dt class="sig sig-object py" id="adapt.parameter_based.FineTuning.unsupervised_score">
<span class="sig-name descname"><span class="pre">unsupervised_score</span></span><span class="sig-paren">(</span><em class="sig-param"><span class="n"><span class="pre">Xs</span></span></em>, <em class="sig-param"><span class="n"><span class="pre">Xt</span></span></em><span class="sig-paren">)</span><a class="reference external" href="https://github.com/antoinedemathelin/adapt/tree/master/adapt/parameter_based/_finetuning.unsupervised_score.py"><span class="viewcode-link"><span class="pre">[source]</span></span></a><a class="headerlink" href="#adapt.parameter_based.FineTuning.unsupervised_score" title="Permalink to this definition"></a></dt>
<dd><p>Return unsupervised score.</p>
<p>The normalized discrepancy distance is computed
between the reweighted/transformed source input
data and the target input data.</p>
<dl class="field-list simple">
<dt class="field-odd">Parameters</dt>
<dd class="field-odd"><dl class="simple">
<dt><strong>Xs</strong><span class="classifier">array</span></dt><dd><p>Source input data.</p>
</dd>
<dt><strong>Xt</strong><span class="classifier">array</span></dt><dd><p>Source input data.</p>
</dd>
</dl>
</dd>
<dt class="field-even">Returns</dt>
<dd class="field-even"><dl class="simple">
<dt><strong>score</strong><span class="classifier">float</span></dt><dd><p>Unsupervised score.</p>
</dd>
</dl>
</dd>
</dl>
</dd></dl>
</dd></dl>
<h2> Examples </h2><div class="toctree-wrapper compound">
</div>
<div class="sphx-glr-thumbcontainer">
<div class="figure align-center">
<img alt="thumbnail" src="../_static/images/finetuned.png" />
<p class="caption">
<span class="caption-text">
<a class="reference internal" href="../examples/Flowers_example.html">
<span class="std std-ref">Model-Based Transfer Learning</span>
</a>
</span>
</p>
</div>
</div>
<div class="sphx-glr-thumbcontainer">
<div class="figure align-center">
<img alt="thumbnail" src="../_static/images/office_item.png" />
<p class="caption">
<span class="caption-text">
<a class="reference internal" href="../examples/Office_example.html">
<span class="std std-ref">Deep Domain Adaptation: Transfer one domain to another</span>
</a>
</span>
</p>
</div>
</div>
<div class="sphx-glr-clear"></div></section>
</div>
</div>
<footer>
<hr/>
<div role="contentinfo">
<p>© Copyright 2020, ADAPT team, Michelin and Centre Borelli, ENS Paris-Saclay.</p>
</div>
Built with <a href="https://www.sphinx-doc.org/">Sphinx</a> using a
<a href="https://github.com/readthedocs/sphinx_rtd_theme">theme</a>
provided by <a href="https://readthedocs.org">Read the Docs</a>.
</footer>
</div>
</div>
</section>
</div>
<script>
jQuery(function () {
SphinxRtdTheme.Navigation.enable(true);
});
</script>
</body>
</html>