forked from cvg/glue-factory
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtartan-18-07-distributedTraining-128-grayscale.log
1091 lines (1072 loc) · 228 KB
/
tartan-18-07-distributedTraining-128-grayscale.log
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
Starting the distributed training at Thu 18 Jul 23:20:45 BST 2024
Command running in screen session distributed_train_0_20240718-232045 on gpu12
Command running in screen session distributed_train_1_20240718-232045 on gpu07
Command running in screen session distributed_train_2_20240718-232045 on gpu08
Command running in screen session distributed_train_3_20240718-232045 on gpu20
Command running in screen session distributed_train_4_20240718-232045 on gpu10
Command running in screen session distributed_train_5_20240718-232045 on gpu14
Command running in screen session distributed_train_6_20240718-232045 on gpu16
Command running in screen session distributed_train_7_20240718-232045 on gpu24
Distributed training completed at Thu 18 Jul 23:21:00 BST 2024
[07/18/2024 23:21:13 gluefactory INFO] Starting experiment sp+lg_18-07-tartan-homog-gray-128
[07/18/2024 23:21:26 gluefactory INFO] Will fine-tune from weights of pretrain_lightglue
[07/18/2024 23:21:26 gluefactory INFO] Training in distributed mode with 8 GPUs
[07/18/2024 23:21:40 gluefactory INFO] Starting experiment sp+lg_18-07-tartan-homog-gray-128
[07/18/2024 23:22:03 gluefactory INFO] Will fine-tune from weights of pretrain_lightglue
[07/18/2024 23:22:06 gluefactory INFO] Training in distributed mode with 8 GPUs
[07/18/2024 23:22:42 gluefactory INFO] Using device cuda:0
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/18/2024 23:22:44 gluefactory.datasets.base_dataset INFO] Creating dataset HomographyTartanTreeDataset
NFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/18/2024 23:22:45 gluefactory.datasets.base_dataset INFO] Creating dataset HomographyTartanTreeDataset
[07/18/2024 23:23:42 gluefactory INFO] Training loader has 303 batches
[07/18/2024 23:23:42 gluefactory INFO] Validation loader has 608 batches
[07/18/2024 23:23:57 gluefactory INFO] Parameters with scaled learning rate:
{}
[07/18/2024 23:23:57 gluefactory INFO] Training with mixed_precision=None
[07/18/2024 23:23:57 gluefactory INFO] Starting training with configuration:
data:
name: homographiesTartan
data_dir: syntheticForestData/imageData
train_size: 40000
val_size: 3000
batch_size: 128
num_workers: 4
homography:
difficulty: 0.7
max_angle: 45
photometric:
name: lg
model:
name: two_view_pipeline
extractor:
name: extractors.superpoint_open
max_num_keypoints: 512
force_num_keypoints: true
detection_threshold: 0
nms_radius: 3
trainable: false
ground_truth:
name: matchers.homography_matcher
th_positive: 3
th_negative: 3
matcher:
name: matchers.lightglue
filter_threshold: 0.1
flash: false
checkpointed: true
train:
seed: 0
epochs: 40
optimizer: adam
opt_regexp: null
optimizer_options: {}
lr: 0.0001
lr_schedule:
type: exp
start: 20
exp_div_10: 10
on_epoch: true
factor: 1.0
options: {}
lr_scaling:
- - 100
- - dampingnet.const
eval_every_iter: 500
save_every_iter: 5000
log_every_iter: 100
log_grad_every_iter: null
test_every_epoch: 1
keep_last_checkpoints: 10
load_experiment: sp+lg_densehomography
median_metrics: []
recall_metrics: {}
pr_metrics: {}
best_key: loss/total
dataset_callback_fn: null
dataset_callback_on_val: false
clip_grad: null
pr_curves: {}
plot:
- 5
- gluefactory.visualization.visualize_batch.make_match_figures
submodules: []
benchmarks:
hpatches:
eval:
estimator: opencv
ransac_th: 0.5
image_mode: null
[07/18/2024 23:23:57 gluefactory INFO] Starting epoch 0
loading pretrain lightglue, replacing model
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_densehomography/checkpoint_39_61799.tar
loading pretrain lightglue, replacing model
Configuration fields in conf.model:
renaming old state dictionary keys
lines 410 of train.py with renaming keys may not work?
{'name': 'two_view_pipeline', 'extractor': {'name': 'extractors.superpoint_open', 'max_num_keypoints': 512, 'force_num_keypoints': True, 'detection_threshold': 0, 'nms_radius': 3, 'trainable': False}, 'ground_truth': {'name': 'matchers.homography_matcher', 'th_positive': 3, 'th_negative': 3}, 'matcher': {'name': 'matchers.lightglue', 'filter_threshold': 0.1, 'flash': False, 'checkpointed': True}}
the rank is 3 and world size is 8
cuda:0
23 initpy data sets path gluefactory.datasets.homographiesTartan
205 tools.py utils the classes are [('HomographyTartanTreeDataset', <class 'gluefactory.datasets.homographiesTartan.HomographyTartanTreeDataset'>)]
Loaded training dataset: homographiesTartan
Using the same dataset for validation as for training.
205 tools.py utils the classes are [('TwoViewPipeline', <class 'gluefactory.models.two_view_pipeline.TwoViewPipeline'>)]
ModuleNotFoundError: extractors.superpoint_open
205 tools.py utils the classes are [('SuperPoint', <class 'gluefactory.models.extractors.superpoint_open.SuperPoint'>)]
None
Superpoint grayscale mode selected
ModuleNotFoundError: matchers.lightglue
205 tools.py utils the classes are []
368 lightglue loading weights: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/pretrain_lightglue/superpoint_lightglue.pth
ModuleNotFoundError: matchers.homography_matcher
205 tools.py utils the classes are [('HomographyMatcher', <class 'gluefactory.models.matchers.homography_matcher.HomographyMatcher'>)]
initcp not none
args distibuted
device ids are [device(type='cuda', index=0)]
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/utils/checkpoint.py:464: UserWarning: torch.utils.checkpoint: the use_reentrant parameter should be passed explicitly. In version 2.4 we will raise an exception if use_reentrant is not passed. use_reentrant=False is recommended, but if you need to preserve the current default behavior, you can pass use_reentrant=True. Refer to docs for more details on the differences between the two variants.
warnings.warn(
[07/18/2024 23:24:35 gluefactory INFO] [E 0 | it 0] loss {total 3.217E+00, last 2.886E+00, assignment_nll 2.886E+00, nll_pos 5.302E+00, nll_neg 4.703E-01, num_matchable 7.799E+01, num_unmatchable 2.376E+02, confidence 1.278E-01, row_norm 3.400E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 0, 'iteration': 0, 'total': 3.2168147563934326, 'last': 2.8858933448791504, 'assignment_nll': 2.8858933448791504, 'nll_pos': 5.301520824432373, 'nll_neg': 0.47026604413986206, 'num_matchable': 77.9921875, 'num_unmatchable': 237.56640625, 'confidence': 0.12775671482086182, 'row_norm': 0.3400484621524811}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py:814: FutureWarning: The behavior of DataFrame concatenation with empty or all-NA entries is deprecated. In a future version, this will no longer exclude empty or all-NA columns when determining the result dtypes. To retain the old behavior, exclude the relevant entries before the concat operation.
results_df = pd.concat([results_df, new_entry], ignore_index=True)
[07/18/2024 23:32:37 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240718_233237.pkl
[07/18/2024 23:32:37 gluefactory INFO] [Validation] {match_recall 1.997E-01, match_precision 4.684E-01, accuracy 7.482E-01, average_precision 1.523E-01, loss/total 4.326E+00, loss/last 4.326E+00, loss/assignment_nll 4.326E+00, loss/nll_pos 7.212E+00, loss/nll_neg 1.439E+00, loss/num_matchable 1.644E+02, loss/num_unmatchable 3.400E+02, loss/row_norm 3.451E-01}
[07/18/2024 23:32:37 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/18/2024 23:32:38 gluefactory INFO] New best val: loss/total=4.325604209787983
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py:814: FutureWarning: The behavior of DataFrame concatenation with empty or all-NA entries is deprecated. In a future version, this will no longer exclude empty or all-NA columns when determining the result dtypes. To retain the old behavior, exclude the relevant entries before the concat operation.
results_df = pd.concat([results_df, new_entry], ignore_index=True)
[07/18/2024 23:43:32 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240718_234332.pkl
[07/18/2024 23:43:32 gluefactory INFO] [Validation] {match_recall 1.997E-01, match_precision 4.684E-01, accuracy 7.482E-01, average_precision 1.523E-01, loss/total 4.326E+00, loss/last 4.326E+00, loss/assignment_nll 4.326E+00, loss/nll_pos 7.212E+00, loss/nll_neg 1.439E+00, loss/num_matchable 1.644E+02, loss/num_unmatchable 3.400E+02, loss/row_norm 3.451E-01}
[07/18/2024 23:43:32 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/18/2024 23:43:33 gluefactory INFO] New best val: loss/total=4.325601578593279
/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/utils/checkpoint.py:464: UserWarning: torch.utils.checkpoint: the use_reentrant parameter should be passed explicitly. In version 2.4 we will raise an exception if use_reentrant is not passed. use_reentrant=False is recommended, but if you need to preserve the current default behavior, you can pass use_reentrant=True. Refer to docs for more details on the differences between the two variants.
warnings.warn(
[07/18/2024 23:48:11 gluefactory INFO] [E 0 | it 100] loss {total 1.567E+00, last 4.068E-01, assignment_nll 4.068E-01, nll_pos 6.610E-01, nll_neg 1.527E-01, num_matchable 1.462E+02, num_unmatchable 3.586E+02, confidence 2.973E-01, row_norm 9.363E-01}
[07/18/2024 23:52:43 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_0_200_20240718_235243.pkl
[07/18/2024 23:52:43 gluefactory INFO] [E 0 | it 200] loss {total 1.052E+00, last 1.545E-01, assignment_nll 1.545E-01, nll_pos 1.759E-01, nll_neg 1.331E-01, num_matchable 1.381E+02, num_unmatchable 3.658E+02, confidence 2.589E-01, row_norm 9.669E-01}
[07/18/2024 23:58:47 gluefactory INFO] [E 0 | it 300] loss {total 1.050E+00, last 3.016E-01, assignment_nll 3.016E-01, nll_pos 4.537E-01, nll_neg 1.495E-01, num_matchable 1.342E+02, num_unmatchable 3.709E+02, confidence 2.275E-01, row_norm 9.508E-01}
Data to be logged: {'epoch': 0, 'iteration': 0, 'match_recall': 0.19966769337865803, 'match_precision': 0.4684297389667025, 'accuracy': 0.7481534674699252, 'average_precision': 0.1522978252524367, 'loss/total': 4.325601578593279, 'loss/last': 4.325601578593279, 'loss/assignment_nll': 4.325601578593279, 'loss/nll_pos': 7.212359272110774, 'loss/nll_neg': 1.4388438750619086, 'loss/num_matchable': 164.43802758904673, 'loss/num_unmatchable': 340.0155445748404, 'loss/row_norm': 0.345067403230349}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 0, 'iteration': 100, 'total': 1.5672080516815186, 'last': 0.40682947635650635, 'assignment_nll': 0.40682947635650635, 'nll_pos': 0.6609779000282288, 'nll_neg': 0.15268096327781677, 'num_matchable': 146.15625, 'num_unmatchable': 358.625, 'confidence': 0.2972630262374878, 'row_norm': 0.9362682104110718}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 0, 'iteration': 200, 'total': 1.0521240234375, 'last': 0.15446671843528748, 'assignment_nll': 0.15446671843528748, 'nll_pos': 0.1758536547422409, 'nll_neg': 0.13307979702949524, 'num_matchable': 138.0625, 'num_unmatchable': 365.76171875, 'confidence': 0.2589189410209656, 'row_norm': 0.9668808579444885}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 0, 'iteration': 300, 'total': 1.0502104759216309, 'last': 0.3015606105327606, 'assignment_nll': 0.3015606105327606, 'nll410675, 'nll_pos': 0.16916073858737946, 'nll_neg': 0.055637046694755554, 'num_matchable': 50.4296875, 'num_unmatchable': 139.078125, 'confidence': 0.08571251481771469, 'row_norm': 0.3576112985610962}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 00:14:57 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_001457.pkl
[07/19/2024 00:14:57 gluefactory INFO] [Validation] {match_recall 9.421E-01, match_precision 8.256E-01, accuracy 9.585E-01, average_precision 8.017E-01, loss/total 2.479E-01, loss/last 2.479E-01, loss/assignment_nll 2.479E-01, loss/nll_pos 3.648E-01, loss/nll_neg 1.310E-01, loss/num_matchable 1.448E+02, loss/num_unmatchable 3.610E+02, loss/row_norm 9.582E-01}
[07/19/2024 00:14:57 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 00:14:59 gluefactory INFO] New best val: loss/total=0.247899299627448
[07/19/2024 00:15:02 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_0_2416.tar
[07/19/2024 00:15:04 gluefactory INFO] Starting epoch 1
[07/19/2024 00:15:04 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 0, 'iteration': 302, 'match_recall': 0.9420979920178744, 'match_precision': 0.8255841600715759, 'accuracy': 0.9584852505152378, 'average_precision': 0.8016511524364196, 'loss/total': 0.247899299627448, 'loss/last': 0.247899299627448, 'loss/assignment_nll': 0.247899299627448, 'loss/nll_pos': 0.36476548151215676, 'loss/nll_neg': 0.1310331202891596, 'loss/num_matchable': 144.83323038912908, 'loss/num_unmatchable': 361.0229050854437, 'loss/row_norm': 0.9582096908424822}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 00:15:43 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_001543.pkl
[07/19/2024 00:15:43 gluefactory INFO] [Validation] {match_recall 9.421E-01, match_precision 8.256E-01, accuracy 9.585E-01, average_precision 8.017E-01, loss/total 2.479E-01, loss/last 2.479E-01, loss/assignment_nll 2.479E-01, loss/nll_pos 3.648E-01, loss/nll_neg 1.310E-01, loss/num_matchable 1.448E+02, loss/num_unmatchable 3.610E+02, loss/row_norm 9.582E-01}
[07/19/2024 00:15:44 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 00:15:46 gluefactory INFO] New best val: loss/total=0.2478993002901309
[07/19/2024 00:15:49 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_0_2416.tar
[07/19/2024 00:15:51 gluefactory INFO] Starting epoch 1
[07/19/2024 00:15:51 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 0, 'iteration': 302, 'match_recall': 0.9420979920178744, 'match_precision': 0.8255841600715759, 'accuracy': 0.9584852505152378, 'average_precision': 0.8016511524364196, 'loss/total': 0.2478993002901309, 'loss/last': 0.2478993002901309, 'loss/assignment_nll': 0.2478993002901309, 'loss/nll_pos': 0.3647654808740177, 'loss/nll_neg': 0.1310331202891596, 'loss/num_matchable': 144.83323038912908, 'loss/num_unmatchable': 361.0229050854437, 'loss/row_norm': 0.9582096908424822}
INFO:INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 00:17:23 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_001723.pkl
[07/19/2024 00:17:23 gluefactory INFO] [Validation] {match_recall 9.421E-01, match_precision 8.256E-01, accuracy 9.585E-01, average_precision 8.017E-01, loss/total 2.479E-01, loss/last 2.479E-01, loss/assignment_nll 2.479E-01, loss/nll_pos 3.648E-01, loss/nll_neg 1.310E-01, loss/num_matchable 1.448E+02, loss/num_unmatchable 3.610E+02, loss/row_norm 9.582E-01}
[07/19/2024 00:17:23 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 00:17:25 gluefactory INFO] New best val: loss/total=0.247899299756303
[07/19/2024 00:17:29 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_0_2416.tar
[07/19/2024 00:17:30 gluefactory INFO] Starting epoch 1
[07/19/2024 00:17:30 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 0, 'iteration': 302, 'match_recall': 0.9420979920178744, 'match_precision': 0.8255841600715759, 'accuracy': 0.9584852505152378, 'average_precision': 0.8016511524364196, 'loss/total': 0.247899299756303, 'loss/last': 0.247899299756303, 'loss/assignment_nll': 0.247899299756303, 'loss/nll_pos': 0.36476548069914305, 'loss/nll_neg': 0.1310331202891596, 'loss/num_matchable': 144.83323038912908, 'loss/num_unmatchable': 361.0229050854437, 'loss/row_norm': 0.9582096908424822}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 00:18:00 gluefactory INFO] [E 1 | it 0] loss {total 9.506E-01, last 1.607E-01, assignment_nll 1.607E-01, nll_pos 1.716E-01, nll_neg 1.499E-01, num_matchable 1.316E+02, num_unmatchable 3.726E+02, confidence 2.447E-01, row_norm 9.680E-01}
[07/19/2024 00:22:01 gluefactory INFO] [E 1 | it 100] loss {total 1.040E+00, last 1.822E-01, assignment_nll 1.822E-01, nll_pos 2.645E-01, nll_neg 9.991E-02, num_matchable 1.508E+02, num_unmatchable 3.529E+02, confidence 2.402E-01, row_norm 9.646E-01}
[07/19/2024 00:26:06 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_1_200_20240719_002606.pkl
[07/19/2024 00:26:06 gluefactory INFO] [E 1 | it 200] loss {total 8.287E-01, last 1.356E-01, assignment_nll 1.356E-01, nll_pos 1.728E-01, nll_neg 9.838E-02, num_matchable 1.417E+02, num_unmatchable 3.626E+02, confidence 2.131E-01, row_norm 9.758E-01}
[07/19/2024 00:30:14 gluefactory INFO] [E 1 | it 300] loss {total 2.051E-01, last 4.411E-02, assignment_nll 4.411E-02, nll_pos 5.769E-02, nll_neg 3.053E-02, num_matchable 3.537E+01, num_unmatchable 9.087E+01, confidence 5.092E-02, row_norm 2.423E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 1, 'iteration': 0, 'total': 0.23177975416183472, 'last': 0.040704838931560516, 'assignment_nll': 0.040704838931560516, 'nll_pos': 0.04578513279557228, 'nll_neg': 0.03562454879283905, 'num_matchable': 33.7734375, 'num_unmatchable': 92.5, 'confidence': 0.06064105033874512, 'row_norm': 0.24239961802959442}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 1, 'iteration': 100, 'total': 0.2960001230239868, 'last': 0.08000849187374115, 'assignment_nll': 0.08000849187374115, 'nll_pos': 0.1252911239862442, 'nll_neg': 0.03472583740949631, 'num_matchable': 35.296875, 'num_unmatchable': 90.8828125, 'confidence': 0.059240035712718964, 'row_norm': 0.2350689172744751}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 1, 'iteration': 200, 'total': 0.20700475573539734, 'last': 0.027203775942325592, 'assignment_nll': 0.027203775942325592, 'nll_pos': 0.030648862943053246, 'nll_neg': 0.02375868707895279, 'num_matchable': 35.0390625, 'num_unmatchable': 91.3515625, 'confidence': 0.05279494449496269, 'row_norm': 0.24554228782653809}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 1, 'iteration': 300, 'total': 0.20508405566215515, 'last': 0.044106967747211456, 'assignment_nll': 0.044106967747211456, 'nll_pos': 0.05768510699272156, 'nll_neg': 0.030528824776411057, 'num_matchable': 35.3671875, 'num_unmatchable': 90.87109375, 'confidence': 0.050916045904159546, 'row_norm': 0.24231597781181335}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
tions
[07/19/2024 00:38:28 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_003828.pkl
[07/19/2024 00:38:28 gluefactory INFO] [Validation] {match_recall 9.583E-01, match_precision 8.414E-01, accuracy 9.641E-01, average_precision 8.222E-01, loss/total 1.796E-01, loss/last 1.796E-01, loss/assignment_nll 1.796E-01, loss/nll_pos 2.503E-01, loss/nll_neg 1.090E-01, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.693E-01}
[07/19/2024 00:38:28 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 00:38:29 gluefactory INFO] New best val: loss/total=0.17960983444912604
[07/19/2024 00:38:32 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_1_4840.tar
[07/19/2024 00:38:34 gluefactory INFO] Starting epoch 2
[07/19/2024 00:38:34 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 1, 'iteration': 302, 'match_recall': 0.9583239338882474, 'match_precision': 0.8413573994125043, 'accuracy': 0.9641142070403874, 'average_precision': 0.8221950139532076, 'loss/total': 0.17960983444912604, 'loss/last': 0.17960983444912604, 'loss/assignment_nll': 0.17960983444912604, 'loss/nll_pos': 0.25026229050624804, 'loss/nll_neg': 0.10895737873791839, 'loss/num_matchable': 145.12476837554047, 'loss/num_unmatchable': 360.71819024088944, 'loss/row_norm': 0.9692710533735068}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 00:41:17 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_004117.pkl
[07/19/2024 00:41:17 gluefactory INFO] [Validation] {match_recall 9.583E-01, match_precision 8.414E-01, accuracy 9.641E-01, average_precision 8.222E-01, loss/total 1.796E-01, loss/last 1.796E-01, loss/assignment_nll 1.796E-01, loss/nll_pos 2.503E-01, loss/nll_neg 1.090E-01, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.693E-01}
[07/19/2024 00:41:17 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 00:41:18 gluefactory INFO] New best val: loss/total=0.17960890884449238
[07/19/2024 00:41:23 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_1_4840.tar
[07/19/2024 00:41:24 gluefactory INFO] Starting epoch 2
[07/19/2024 00:41:24 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 1, 'iteration': 302, 'match_recall': 0.9583198171057572, 'match_precision': 0.8413531918179137, 'accuracy': 0.9641125955919387, 'average_precision': 0.8221904545967191, 'loss/total': 0.17960890884449238, 'loss/last': 0.17960890884449238, 'loss/assignment_nll': 0.17960890884449238, 'loss/nll_pos': 0.2502614926602036, 'loss/nll_neg': 0.1089563257163914, 'loss/num_matchable': 145.1246654313362, 'loss/num_unmatchable': 360.71829318509367, 'loss/row_norm': 0.9692719277222264}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 00:42:40 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_004240.pkl
[07/19/2024 00:42:40 gluefactory INFO] [Validation] {match_recall 9.583E-01, match_precision 8.414E-01, accuracy 9.641E-01, average_precision 8.222E-01, loss/total 1.796E-01, loss/last 1.796E-01, loss/assignment_nll 1.796E-01, loss/nll_pos 2.503E-01, loss/nll_neg 1.090E-01, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.693E-01}
[07/19/2024 00:42:40 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 00:42:42 gluefactory INFO] New best val: loss/total=0.17960890884449238
[07/19/2024 00:42:46 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_1_4840.tar
[07/19/2024 00:42:47 gluefactory INFO] Starting epoch 2
[07/19/2024 00:42:47 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 1, 'iteration': 302, 'match_recall': 0.9583198171057572, 'match_precision': 0.8413531918179137, 'accuracy': 0.9641125955919387, 'average_precision': 0.8221904545967191, 'loss/total': 0.17960890884449238, 'loss/last': 0.17960890884449238, 'loss/assignment_nll': 0.17960890884449238, 'loss/nll_pos': 0.2502614926602036, 'loss/nll_neg': 0.1089563257163914, 'loss/num_matchable': 145.1246654313362, 'loss/num_unmatchable': 360.71829318509367, 'loss/row_norm': 0.9692719277222264}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 00:49:15 gluefactory INFO] [E 2 | it 0] loss {total 4.153E-01, last 6.810E-02, assignment_nll 6.810E-02, nll_pos 7.040E-02, nll_neg 6.579E-02, num_matchable 6.756E+01, num_unmatchable 1.845E+02, confidence 1.092E-01, row_norm 4.860E-01}
[07/19/2024 00:50:21 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_2_5000.tar
[07/19/2024 00:53:33 gluefactory INFO] [E 2 | it 100] loss {total 5.052E-01, last 4.445E-02, assignment_nll 4.445E-02, nll_pos 5.066E-02, nll_neg 3.823E-02, num_matchable 8.961E+01, num_unmatchable 2.263E+02, confidence 1.311E-01, row_norm 6.151E-01}
[07/19/2024 00:57:49 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_2_200_20240719_005749.pkl
[07/19/2024 00:57:49 gluefactory INFO] [E 2 | it 200] loss {total 4.544E-01, last 5.115E-02, assignment_nll 5.115E-02, nll_pos 5.010E-02, nll_neg 5.219E-02, num_matchable 8.951E+01, num_unmatchable 2.256E+02, confidence 1.262E-01, row_norm 6.149E-01}
[07/19/2024 01:02:03 gluefactory INFO] [E 2 | it 300] loss {total 5.252E-01, last 1.180E-01, assignment_nll 1.180E-01, nll_pos 1.594E-01, nll_neg 7.657E-02, num_matchable 8.391E+01, num_unmatchable 2.313E+02, confidence 1.273E-01, row_norm 6.061E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 2, 'iteration': 0, 'total': 0.5167489051818848, 'last': 0.08434426784515381, 'assignment_nll': 0.08434426784515381, 'nll_pos': 0.08543968200683594, 'nll_neg': 0.08324885368347168, 'num_matchable': 83.8203125, 'num_unmatchable': 231.3125, 'confidence': 0.1373283416032791, 'row_norm': 0.6072310209274292}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 2, 'iteration': 100, 'total': 0.5052091479301453, 'last': 0.04444748908281326, 'assignment_nll': 0.04444748908281326, 'nll_pos': 0.05066191777586937, 'nll_neg': 0.038233064115047455, 'num_matchable': 89.609375, 'num_unmatchable': 226.2734375, 'confidence': 0.131139874458313, 'row_norm': 0.6151103377342224}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 2, 'iteration': 200, 'total': 0.4543853998184204, 'last': 0.05114690959453583, 'assignment_nll': 0.05114690959453583, 'nll_pos': 0.05009973421692848, 'nll_neg': 0.052194081246852875, 'num_matchable': 89.5078125, 'num_unmatchable': 225.6328125, 'confidence': 0.12620893120765686, 'row_norm': 0.614868700504303}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 2, 'iteration': 300, 'total': 0.5252363681793213, 'last': 0.11799365282058716, 'assignment_nll': 0.11799365282058716, 'nll_pos': 0.159419447183609, 'nll_neg': 0.0765678733587265, 'num_matchable': 83.9140625, 'num_unmatchable': 231.29296875, 'confidence': 0.12733669579029083, 'row_norm': 0.606062114238739}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 01:10:15 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_011015.pkl
[07/19/2024 01:10:15 gluefactory INFO] [Validation] {match_recall 9.627E-01, match_precision 8.496E-01, accuracy 9.676E-01, average_precision 8.313E-01, loss/total 1.602E-01, loss/last 1.602E-01, loss/assignment_nll 1.602E-01, loss/nll_pos 2.300E-01, loss/nll_neg 9.032E-02, loss/num_matchable 1.453E+02, loss/num_unmatchable 3.605E+02, loss/row_norm 9.721E-01}
[07/19/2024 01:10:15 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 01:10:16 gluefactory INFO] New best val: loss/total=0.16016533432134525
[07/19/2024 01:10:19 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_2_7264.tar
[07/19/2024 01:10:21 gluefactory INFO] Starting epoch 3
[07/19/2024 01:10:21 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 2, 'iteration': 302, 'match_recall': 0.962720435466299, 'match_precision': 0.84960603435417, 'accuracy': 0.9676469040427503, 'average_precision': 0.8312980272727222, 'loss/total': 0.16016533432134525, 'loss/last': 0.16016533432134525, 'loss/assignment_nll': 0.16016533432134525, 'loss/nll_pos': 0.23001542033994635, 'loss/nll_neg': 0.09031524774724867, 'loss/num_matchable': 145.29256742845376, 'loss/num_unmatchable': 360.53129503808935, 'loss/row_norm': 0.9721340546152438}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 01:13:04 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_011304.pkl
[07/19/2024 01:13:04 gluefactory INFO] [Validation] {match_recall 9.627E-01, match_precision 8.496E-01, accuracy 9.676E-01, average_precision 8.313E-01, loss/total 1.602E-01, loss/last 1.602E-01, loss/assignment_nll 1.602E-01, loss/nll_pos 2.301E-01, loss/nll_neg 9.032E-02, loss/num_matchable 1.453E+02, loss/num_unmatchable 3.605E+02, loss/row_norm 9.721E-01}
[07/19/2024 01:13:04 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 01:13:06 gluefactory INFO] New best val: loss/total=0.16019922610397616
[07/19/2024 01:13:10 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_2_7264.tar
[07/19/2024 01:13:11 gluefactory INFO] Starting epoch 3
[07/19/2024 01:13:11 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 2, 'iteration': 302, 'match_recall': 0.9627200122328239, 'match_precision': 0.8496026270841653, 'accuracy': 0.9676452791442934, 'average_precision': 0.8312944491777936, 'loss/total': 0.16019922610397616, 'loss/last': 0.16019922610397616, 'loss/assignment_nll': 0.16019922610397616, 'loss/nll_pos': 0.23008187422373547, 'loss/nll_neg': 0.0903165768747599, 'loss/num_matchable': 145.29267037265802, 'loss/num_unmatchable': 360.5311920938851, 'loss/row_norm': 0.9721336526489808}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 01:15:40 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_011540.pkl
[07/19/2024 01:15:40 gluefactory INFO] [Validation] {match_recall 9.627E-01, match_precision 8.496E-01, accuracy 9.676E-01, average_precision 8.313E-01, loss/total 1.602E-01, loss/last 1.602E-01, loss/assignment_nll 1.602E-01, loss/nll_pos 2.301E-01, loss/nll_neg 9.032E-02, loss/num_matchable 1.453E+02, loss/num_unmatchable 3.605E+02, loss/row_norm 9.721E-01}
[07/19/2024 01:15:41 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 01:15:42 gluefactory INFO] New best val: loss/total=0.16019922610397616
[07/19/2024 01:15:46 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_2_7264.tar
[07/19/2024 01:15:47 gluefactory INFO] Starting epoch 3
[07/19/2024 01:15:48 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 2, 'iteration': 302, 'match_recall': 0.9627200122328239, 'match_precision': 0.8496026270841653, 'accuracy': 0.9676452791442934, 'average_precision': 0.8312944491777936, 'loss/total': 0.16019922610397616, 'loss/last': 0.16019922610397616, 'loss/assignment_nll': 0.16019922610397616, 'loss/nll_pos': 0.23008187422373547, 'loss/nll_neg': 0.0903165768747599, 'loss/num_matchable': 145.29267037265802, 'loss/num_unmatchable': 360.5311920938851, 'loss/row_norm': 0.9721336526489808}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 01:18:24 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_011824.pkl
[07/19/2024 01:18:24 gluefactory INFO] [Validation] {match_recall 9.627E-01, match_precision 8.496E-01, accuracy 9.676E-01, average_precision 8.313E-01, loss/total 1.602E-01, loss/last 1.602E-01, loss/assignment_nll 1.602E-01, loss/nll_pos 2.300E-01, loss/nll_neg 9.031E-02, loss/num_matchable 1.453E+02, loss/num_unmatchable 3.605E+02, loss/row_norm 9.721E-01}
[07/19/2024 01:18:10 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 01:18:11 gluefactory INFO] New best val: loss/total=0.16019922606409245
[07/19/2024 01:18:15 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_2_7264.tar
[07/19/2024 01:18:17 gluefactory INFO] Starting epoch 3
[07/19/2024 01:18:17 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 2, 'iteration': 302, 'match_recall': 0.9627200122328239, 'match_precision': 0.8496026270841653, 'accuracy': 0.9676452791442934, 'average_precision': 0.8312944491777936, 'loss/total': 0.16019922606409245, 'loss/last': 0.16019922606409245, 'loss/assignment_nll': 0.16019922606409245, 'loss/nll_pos': 0.23008187458575668, 'loss/nll_neg': 0.0903165768747599, 'loss/num_matchable': 145.29267037265802, 'loss/num_unmatchable': 360.5311920938851, 'loss/row_norm': 0.9721336526489808}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 01:21:02 gluefactory INFO] [E 3 | it 0] loss {total 2.840E-01, last 4.416E-02, assignment_nll 4.416E-02, nll_pos 4.821E-02, nll_neg 4.012E-02, num_matchable 4.997E+01, num_unmatchable 1.391E+02, confidence 8.054E-02, row_norm 3.660E-01}
[07/19/2024 01:27:04 gluefactory INFO] [E 3 | it 100] loss {total 3.226E-01, last 7.293E-02, assignment_nll 7.293E-02, nll_pos 1.133E-01, nll_neg 3.254E-02, num_matchable 5.391E+01, num_unmatchable 1.356E+02, confidence 7.357E-02, row_norm 3.637E-01}
[07/19/2024 01:32:48 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_3_200_20240719_013248.pkl
[07/19/2024 01:32:48 gluefactory INFO] [E 3 | it 200] loss {total 2.559E-01, last 2.805E-02, assignment_nll 2.805E-02, nll_pos 2.944E-02, nll_neg 2.666E-02, num_matchable 5.204E+01, num_unmatchable 1.372E+02, confidence 7.015E-02, row_norm 3.696E-01}
[07/19/2024 01:37:09 gluefactory INFO] [E 3 | it 300] loss {total 5.103E-01, last 8.020E-02, assignment_nll 8.020E-02, nll_pos 8.207E-02, nll_neg 7.834E-02, num_matchable 1.051E+02, num_unmatchable 2.739E+02, confidence 1.424E-01, row_norm 7.340E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 3, 'iteration': 0, 'total': 0.6637953519821167, 'last': 0.11581026762723923, 'assignment_nll': 0.11581026762723923, 'nll_pos': 0.13696502149105072, 'nll_neg': 0.09465551376342773, 'num_matchable': 113.0, 'num_unmatchable': 328.4296875, 'confidence': 0.18234649300575256, 'row_norm': 0.8537601828575134}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 3, 'iteration': 100, 'total': 0.7588695883750916, 'last': 0.1619521677494049, 'assignment_nll': 0.1619521677494049, 'nll_pos': 0.2574416697025299, 'nll_neg': 0.06646265834569931, 'num_matchable': 127.3125, 'num_unmatchable': 313.7578125, 'confidence': 0.1744966059923172, 'row_norm': 0.8513035774230957}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 3, 'iteration': 200, 'total': 0.623030424118042, 'last': 0.08553212881088257, 'assignment_nll': 0.08553212881088257, 'nll_pos': 0.10135137289762497, 'nll_neg': 0.06971287727355957, 'num_matchable': 119.1328125, 'num_unmatchable': 321.40625, 'confidence': 0.16623465716838837, 'row_norm': 0.8601997494697571}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 3, 'iteration': 300, 'total': 0.590624213218689, 'last': 0.09368140250444412, 'assignment_nll': 0.09368140250444412, 'nll_pos': 0.09749023616313934, 'nll_neg': 0.0898725762963295, 'num_matchable': 122.421875, 'num_unmatchable': 319.90234375, 'confidence': 0.16499102115631104, 'row_norm': 0.8569871187210083}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 01:48:10 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_014810.pkl
[07/19/2024 01:48:10 gluefactory INFO] [Validation] {match_recall 9.675E-01, match_precision 8.529E-01, accuracy 9.690E-01, average_precision 8.362E-01, loss/total 1.409E-01, loss/last 1.409E-01, loss/assignment_nll 1.409E-01, loss/nll_pos 1.953E-01, loss/nll_neg 8.661E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.761E-01}
[07/19/2024 01:48:10 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 01:48:12 gluefactory INFO] New best val: loss/total=0.1409485853736001
[07/19/2024 01:48:15 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_3_9688.tar
[07/19/2024 01:48:17 gluefactory INFO] Starting epoch 4
[07/19/2024 01:48:17 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 3, 'iteration': 302, 'match_recall': 0.9674982360932521, 'match_precision': 0.8528633220759787, 'accuracy': 0.969011164285873, 'average_precision': 0.836202205074209, 'loss/total': 0.1409485853736001, 'loss/last': 0.1409485853736001, 'loss/assignment_nll': 0.1409485853736001, 'loss/nll_pos': 0.19528853224422832, 'loss/nll_neg': 0.08660863976199268, 'loss/num_matchable': 145.02429483220095, 'loss/num_unmatchable': 360.81264154828085, 'loss/row_norm': 0.9760518551290047}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
ions
[07/19/2024 01:49:26 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_014910.pkl
[07/19/2024 01:49:26 gluefactory INFO] [Validation] {match_recall 9.675E-01, match_precision 8.529E-01, accuracy 9.690E-01, average_precision 8.362E-01, loss/total 1.409E-01, loss/last 1.409E-01, loss/assignment_nll 1.409E-01, loss/nll_pos 1.953E-01, loss/nll_neg 8.661E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.761E-01}
[07/19/2024 01:49:26 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 01:49:27 gluefactory INFO] New best val: loss/total=0.1409485853736001
[07/19/2024 01:49:31 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_3_9688.tar
[07/19/2024 01:49:33 gluefactory INFO] Starting epoch 4
[07/19/2024 01:49:33 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 3, 'iteration': 302, 'match_recall': 0.9674982360932521, 'match_precision': 0.8528633220759787, 'accuracy': 0.969011164285873, 'average_precision': 0.836202205074209, 'loss/total': 0.1409485853736001, 'loss/last': 0.1409485853736001, 'loss/assignment_nll': 0.1409485853736001, 'loss/nll_pos': 0.19528853224422832, 'loss/nll_neg': 0.08660863976199268, 'loss/num_matchable': 145.02429483220095, 'loss/num_unmatchable': 360.81264154828085, 'loss/row_norm': 0.9760518551290047}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 01:53:46 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_015346.pkl
[07/19/2024 01:53:46 gluefactory INFO] [Validation] {match_recall 9.675E-01, match_precision 8.529E-01, accuracy 9.690E-01, average_precision 8.362E-01, loss/total 1.409E-01, loss/last 1.409E-01, loss/assignment_nll 1.409E-01, loss/nll_pos 1.953E-01, loss/nll_neg 8.661E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.761E-01}
[07/19/2024 01:53:46 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 01:53:47 gluefactory INFO] New best val: loss/total=0.14094087506272626
[07/19/2024 01:53:52 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_3_9688.tar
[07/19/2024 01:53:53 gluefactory INFO] Starting epoch 4
[07/19/2024 01:53:53 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 3, 'iteration': 302, 'match_recall': 0.9674945664008213, 'match_precision': 0.8528629979994996, 'accuracy': 0.9690113356998483, 'average_precision': 0.8362048934069047, 'loss/total': 0.14094087506272626, 'loss/last': 0.14094087506272626, 'loss/assignment_nll': 0.14094087506272626, 'loss/nll_pos': 0.19527164449777615, 'loss/nll_neg': 0.08661010530975735, 'loss/num_matchable': 145.02511838583487, 'loss/num_unmatchable': 360.81161210623844, 'loss/row_norm': 0.9760539299646032}
INFO:amentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 01:56:23 gluefactory INFO] [E 4 | it 0] loss {total 1.735E-01, last 2.761E-02, assignment_nll 2.761E-02, nll_pos 2.884E-02, nll_neg 2.638E-02, num_matchable 3.293E+01, num_unmatchable 9.335E+01, confidence 4.874E-02, row_norm 2.443E-01}
[07/19/2024 01:57:52 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_4_10000.tar
[07/19/2024 02:00:28 gluefactory INFO] [E 4 | it 100] loss {total 1.626E-01, last 1.065E-02, assignment_nll 1.065E-02, nll_pos 1.102E-02, nll_neg 1.029E-02, num_matchable 3.839E+01, num_unmatchable 8.780E+01, confidence 4.668E-02, row_norm 2.475E-01}
[07/19/2024 02:04:40 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_4_200_20240719_020440.pkl
[07/19/2024 02:04:40 gluefactory INFO] [E 4 | it 200] loss {total 1.420E-01, last 1.742E-02, assignment_nll 1.742E-02, nll_pos 2.007E-02, nll_neg 1.477E-02, num_matchable 3.709E+01, num_unmatchable 8.922E+01, confidence 4.257E-02, row_norm 2.466E-01}
[07/19/2024 02:08:46 gluefactory INFO] [E 4 | it 300] loss {total 1.667E-01, last 2.475E-02, assignment_nll 2.475E-02, nll_pos 2.767E-02, nll_neg 2.184E-02, num_matchable 3.623E+01, num_unmatchable 8.960E+01, confidence 4.740E-02, row_norm 2.448E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 4, 'iteration': 0, 'total': 0.1735338270664215, 'last': 0.027609027922153473, 'assignment_nll': 0.027609027922153473, 'nll_pos': 0.028841078281402588, 'nll_neg': 0.026376977562904358, 'num_matchable': 32.9296875, 'num_unmatchable': 93.3515625, 'confidence': 0.04873637109994888, 'row_norm': 0.24426883459091187}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 4, 'iteration': 100, 'total': 0.1625748574733734, 'last': 0.01065435353666544, 'assignment_nll': 0.01065435353666544, 'nll_pos': 0.011020440608263016, 'nll_neg': 0.010288266465067863, 'num_matchable': 38.390625, 'num_unmatchable': 87.796875, 'confidence': 0.04668007045984268, 'row_norm': 0.24746033549308777}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 4, 'iteration': 200, 'total': 0.1420159488173, 'last': 0.06328193098306656, 'assignment_nll': 0.06328193098306656, 'nll_pos': 0.07042115181684494, 'nll_neg': 0.056142695248126984, 'num_matchable': 124.703125, 'num_unmatchable': 316.58984375, 'confidence': 0.1534072607755661, 'row_norm': 0.8625954389572144}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 4, 'iteration': 300, 'total': 0.593636155128479, 'last': 0.1004217192530632, 'assignment_nll': 0.1004217192530632, 'nll_pos': 0.12093313038349152, 'nll_neg': 0.07991030812263489, 'num_matchable': 121.0625, 'num_unmatchable': 320.19140625, 'confidence': 0.1640627384185791, 'row_norm': 0.8569554090499878}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 02:19:48 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_021948.pkl
[07/19/2024 02:19:48 gluefactory INFO] [Validation] {match_recall 9.686E-01, match_precision 8.543E-01, accuracy 9.697E-01, average_precision 8.379E-01, loss/total 1.385E-01, loss/last 1.385E-01, loss/assignment_nll 1.385E-01, loss/nll_pos 1.872E-01, loss/nll_neg 8.978E-02, loss/num_matchable 1.449E+02, loss/num_unmatchable 3.610E+02, loss/row_norm 9.778E-01}
[07/19/2024 02:19:48 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 02:19:49 gluefactory INFO] New best val: loss/total=0.1385118592785898
[07/19/2024 02:19:53 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_4_12112.tar
[07/19/2024 02:19:55 gluefactory INFO] Starting epoch 5
[07/19/2024 02:19:55 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 4, 'iteration': 302, 'match_recall': 0.9685835592879762, 'match_precision': 0.8542679985143384, 'accuracy': 0.9696909221253917, 'average_precision': 0.8379315853069809, 'loss/total': 0.1385118592785898, 'loss/last': 0.1385118592785898, 'loss/assignment_nll': 0.1385118592785898, 'loss/nll_pos': 0.18723886527407924, 'loss/nll_neg': 0.0897848550705801, 'loss/num_matchable': 144.89324686020177, 'loss/num_unmatchable': 360.9564031295038, 'loss/row_norm': 0.9778242948640883}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 02:27:47 gluefactory INFO] [E 5 | it 0] loss {total 8.019E-02, last 1.363E-02, assignment_nll 1.363E-02, nll_pos 1.286E-02, nll_neg 1.441E-02, num_matchable 1.520E+01, num_unmatchable 4.791E+01, confidence 2.190E-02, row_norm 1.220E-01}
efactory INFO] [Validation] {match_recall 9.686E-01, match_precision 8.543E-01, accuracy 9.697E-01, average_precision 8.379E-01, loss/total 1.385E-01, loss/last 1.385E-01, loss/assignment_nll 1.385E-01, loss/nll_pos 1.872E-01, loss/nll_neg 8.978E-02, loss/num_matchable 1.449E+02, loss/num_unmatchable 3.610E+02, loss/row_norm 9.778E-01}
[07/19/2024 02:27:19 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 02:27:21 gluefactory INFO] New best val: loss/total=0.13851185943198863
[07/19/2024 02:27:25 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_4_12112.tar
[07/19/2024 02:27:26 gluefactory INFO] Starting epoch 5
[07/19/2024 02:27:26 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 4, 'iteration': 302, 'match_recall': 0.9685835592879762, 'match_precision': 0.8542679985143384, 'accuracy': 0.9696909221253917, 'average_precision': 0.8379315853069809, 'loss/total': 0.13851185943198863, 'loss/last': 0.13851185943198863, 'loss/assignment_nll': 0.13851185943198863, 'loss/nll_pos': 0.18723886441657983, 'loss/nll_neg': 0.0897848550705801, 'loss/num_matchable': 144.89324686020177, 'loss/num_unmatchable': 360.9564031295038, 'loss/row_norm': 0.9778242948640883}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 02:27:47 gluefactory INFO] [E 5 | it 0] loss {total 7.450E-01, last 1.958E-01, assignment_nll 1.958E-01, nll_pos 2.693E-01, nll_neg 1.222E-01, num_matchable 1.271E+02, num_unmatchable 3.777E+02, confidence 1.814E-01, row_norm 9.731E-01}
[07/19/2024 02:32:09 gluefactory INFO] [E 5 | it 100] loss {total 7.831E-01, last 1.896E-01, assignment_nll 1.896E-01, nll_pos 3.122E-01, nll_neg 6.693E-02, num_matchable 1.457E+02, num_unmatchable 3.595E+02, confidence 1.837E-01, row_norm 9.751E-01}
[07/19/2024 02:36:26 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_5_200_20240719_023626.pkl
[07/19/2024 02:36:26 gluefactory INFO] [E 5 | it 200] loss {total 6.146E-01, last 8.965E-02, assignment_nll 8.965E-02, nll_pos 9.858E-02, nll_neg 8.073E-02, num_matchable 1.394E+02, num_unmatchable 3.644E+02, confidence 1.731E-01, row_norm 9.830E-01}
[07/19/2024 02:40:47 gluefactory INFO] [E 5 | it 300] loss {total 6.618E-01, last 1.033E-01, assignment_nll 1.033E-01, nll_pos 1.170E-01, nll_neg 8.969E-02, num_matchable 1.354E+02, num_unmatchable 3.687E+02, confidence 1.800E-01, row_norm 9.778E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 5, 'iteration': 0, 'total': 0.7450042963027954, 'last': 0.19577321410179138, 'assignment_nll': 0.19577321410179138, 'nll_pos': 0.26932159066200256, 'nll_neg': 0.1222248524427414, 'num_matchable': 127.1171875, 'num_unmatchable': 377.6640625, 'confidence': 0.18138301372528076, 'row_norm': 0.9731352925300598}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 5, 'iteration': 100, 'total': 0.7830953001976013, 'last': 0.1895786076784134, 'assignment_nll': 0.1895786076784134, 'nll_pos': 0.3122245669364929, 'nll_neg': 0.06693264842033386, 'num_matchable': 145.6953125, 'num_unmatchable': 359.49609375, 'confidence': 0.18370141088962555, 'row_norm': 0.9751060605049133}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 5, 'iteration': 200, 'total': 0.6145608425140381, 'last': 0.08965399861335754, 'assignment_nll': 0.08965399861335754, 'nll_pos': 0.0985802635550499, 'nll_neg': 0.080727718770504, 'num_matchable': 139.421875, 'num_unmatchable': 364.35546875, 'confidence': 0.17310713231563568, 'row_norm': 0.9829509854316711}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 5, 'iteration': 300, 'total': 0.6618357300758362, 'last': 0.10332278162240982, 'assignment_nll': 0.10332278162240982, 'nll_pos': 0.11695057898759842, 'nll_neg': 0.08969499170780182, 'num_matchable': 135.3515625, 'num_unmatchable': 368.70703125, 'confidence': 0.17996379733085632, 'row_norm': 0.9778371453285217}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 02:51:48 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_025148.pkl
[07/19/2024 02:51:48 gluefactory INFO] [Validation] {match_recall 9.724E-01, match_precision 8.621E-01, accuracy 9.727E-01, average_precision 8.465E-01, loss/total 1.196E-01, loss/last 1.196E-01, loss/assignment_nll 1.196E-01, loss/nll_pos 1.661E-01, loss/nll_neg 7.309E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.803E-01}
[07/19/2024 02:51:48 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 02:51:50 gluefactory INFO] New best val: loss/total=0.11959211713952159
[07/19/2024 02:51:54 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_5_14536.tar
[07/19/2024 02:51:55 gluefactory INFO] Starting epoch 6
[07/19/2024 02:51:55 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 5, 'iteration': 302, 'match_recall': 0.9723794042560969, 'match_precision': 0.8621442129466179, 'accuracy': 0.9726916598345926, 'average_precision': 0.8464724658839811, 'loss/total': 0.11959211713952159, 'loss/last': 0.11959211713952159, 'loss/assignment_nll': 0.11959211713952159, 'loss/nll_pos': 0.16609171318399954, 'loss/nll_neg': 0.07309252175791828, 'loss/num_matchable': 145.14453366275478, 'loss/num_unmatchable': 360.6845274861025, 'loss/row_norm': 0.9802962993431759}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 02:52:50 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_025250.pkl
[07/19/2024 02:52:50 gluefactory INFO] [Validation] {match_recall 9.724E-01, match_precision 8.621E-01, accuracy 9.727E-01, average_precision 8.465E-01, loss/total 1.196E-01, loss/last 1.196E-01, loss/assignment_nll 1.196E-01, loss/nll_pos 1.661E-01, loss/nll_neg 7.309E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.803E-01}
[07/19/2024 02:52:50 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 02:52:52 gluefactory INFO] New best val: loss/total=0.11959211713952159
[07/19/2024 02:52:55 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_5_14536.tar
[07/19/2024 02:52:57 gluefactory INFO] Starting epoch 6
[07/19/2024 02:52:57 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 5, 'iteration': 302, 'match_recall': 0.9723794042560969, 'match_precision': 0.8621442129466179, 'accuracy': 0.9726916598345926, 'average_precision': 0.8464724658839811, 'loss/total': 0.11959211713952159, 'loss/last': 0.11959211713952159, 'loss/assignment_nll': 0.11959211713952159, 'loss/nll_pos': 0.16609171318399954, 'loss/nll_neg': 0.07309252175791828, 'loss/num_matchable': 145.14453366275478, 'loss/num_unmatchable': 360.6845274861025, 'loss/row_norm': 0.9802962993431759}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 02:59:47 gluefactory INFO] [E 6 | it 0] loss {total 7.359E-02, last 1.284E-02, assignment_nll 1.284E-02, nll_pos 1.456E-02, nll_neg 1.112E-02, num_matchable 1.672E+01, num_unmatchable 4.640E+01, confidence 2.114E-02, row_norm 1.229E-01}
efactory INFO] [Validation] {match_recall 9.724E-01, match_precision 8.621E-01, accuracy 9.727E-01, average_precision 8.465E-01, loss/total 1.196E-01, loss/last 1.196E-01, loss/assignment_nll 1.196E-01, loss/nll_pos 1.661E-01, loss/nll_neg 7.309E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.803E-01}
[07/19/2024 02:59:19 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 02:59:20 gluefactory INFO] New best val: loss/total=0.11959211724536677
[07/19/2024 02:59:24 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_5_14536.tar
[07/19/2024 02:59:26 gluefactory INFO] Starting epoch 6
[07/19/2024 02:59:26 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 5, 'iteration': 302, 'match_recall': 0.9723794042560969, 'match_precision': 0.8621442129466179, 'accuracy': 0.9726916598345926, 'average_precision': 0.8464724658839811, 'loss/total': 0.11959211724536677, 'loss/last': 0.11959211724536677, 'loss/assignment_nll': 0.11959211724536677, 'loss/nll_pos': 0.16609171313184395, 'loss/nll_neg': 0.07309252175791828, 'loss/num_matchable': 145.14453366275478, 'loss/num_unmatchable': 360.6845274861025, 'loss/row_norm': 0.9802962993431759}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 02:59:47 gluefactory INFO] [E 6 | it 0] loss {total 6.361E-01, last 1.315E-01, assignment_nll 1.315E-01, nll_pos 1.683E-01, nll_neg 9.472E-02, num_matchable 1.252E+02, num_unmatchable 3.800E+02, confidence 1.711E-01, row_norm 9.789E-01}
[07/19/2024 03:02:01 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_6_15000.tar
[07/19/2024 03:03:56 gluefactory INFO] [E 6 | it 100] loss {total 6.346E-01, last 6.307E-02, assignment_nll 6.307E-02, nll_pos 8.166E-02, nll_neg 4.449E-02, num_matchable 1.469E+02, num_unmatchable 3.579E+02, confidence 1.749E-01, row_norm 9.874E-01}
[07/19/2024 03:08:09 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_6_200_20240719_030809.pkl
[07/19/2024 03:08:09 gluefactory INFO] [E 6 | it 200] loss {total 5.731E-01, last 7.172E-02, assignment_nll 7.172E-02, nll_pos 8.369E-02, nll_neg 5.976E-02, num_matchable 1.392E+02, num_unmatchable 3.649E+02, confidence 1.641E-01, row_norm 9.877E-01}
[07/19/2024 03:13:09 gluefactory INFO] [E 6 | it 300] loss {total 6.343E-01, last 9.498E-02, assignment_nll 9.498E-02, nll_pos 1.029E-01, nll_neg 8.702E-02, num_matchable 1.400E+02, num_unmatchable 3.646E+02, confidence 1.794E-01, row_norm 9.802E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 6, 'iteration': 0, 'total': 0.636076033115387, 'last': 0.13152532279491425, 'assignment_nll': 0.13152532279491425, 'nll_pos': 0.16833198070526123, 'nll_neg': 0.09471868723630905, 'num_matchable': 125.2265625, 'num_unmatchable': 379.9921875, 'confidence': 0.17107342183589935, 'row_norm': 0.978909969329834}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 6, 'iteration': 100, 'total': 0.6345576047897339, 'last': 0.06307317316532135, 'assignment_nll': 0.06307317316532135, 'nll_pos': 0.08165948837995529, 'nll_neg': 0.04448685795068741, 'num_matchable': 146.8984375, 'num_unmatchable': 357.91015625, 'confidence': 0.17494012415409088, 'row_norm': 0.9874324202537537}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 6, 'iteration': 200, 'total': 0.5730946063995361, 'last': 0.07172160595655441, 'assignment_nll': 0.07172160595655441, 'nll_pos': 0.08368706703186035, 'nll_neg': 0.05975613743066788, 'num_matchable': 139.2109375, 'num_unmatchable': 364.9296875, 'confidence': 0.1640731394290924, 'row_norm': 0.9876804947853088}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 6, 'iteration': 300, 'total': 0.634344756603241, 'last': 0.09497731178998947, 'assignment_nll': 0.09497731178998947, 'nll_pos': 0.1029299721121788, 'nll_neg': 0.08702463656663895, 'num_matchable': 140.0078125, 'num_unmatchable': 364.6484375, 'confidence': 0.179414764046669, 'row_norm': 0.9801960587501526}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 03:21:20 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_032120.pkl
[07/19/2024 03:21:20 gluefactory INFO] [Validation] {match_recall 9.732E-01, match_precision 8.681E-01, accuracy 9.749E-01, average_precision 8.520E-01, loss/total 1.172E-01, loss/last 1.172E-01, loss/assignment_nll 1.172E-01, loss/nll_pos 1.710E-01, loss/nll_neg 6.345E-02, loss/num_matchable 1.457E+02, loss/num_unmatchable 3.601E+02, loss/row_norm 9.819E-01}
[07/19/2024 03:21:20 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 03:21:22 gluefactory INFO] New best val: loss/total=0.11724805030860551
[07/19/2024 03:21:25 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_6_16960.tar
[07/19/2024 03:21:27 gluefactory INFO] Starting epoch 7
[07/19/2024 03:21:27 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 6, 'iteration': 302, 'match_recall': 0.9731721038898863, 'match_precision': 0.8680962954918396, 'accuracy': 0.974899054015397, 'average_precision': 0.8520204075722383, 'loss/total': 0.11724805030860551, 'loss/last': 0.11724805030860551, 'loss/assignment_nll': 0.11724805030860551, 'loss/nll_pos': 0.17104925938711468, 'loss/nll_neg': 0.063446840690516, 'loss/num_matchable': 145.65009264978383, 'loss/num_unmatchable': 360.13794523368335, 'loss/row_norm': 0.9818554343523399}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 03:24:10 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_032410.pkl
[07/19/2024 03:24:10 gluefactory INFO] [Validation] {match_recall 9.732E-01, match_precision 8.681E-01, accuracy 9.749E-01, average_precision 8.520E-01, loss/total 1.172E-01, loss/last 1.172E-01, loss/assignment_nll 1.172E-01, loss/nll_pos 1.710E-01, loss/nll_neg 6.345E-02, loss/num_matchable 1.456E+02, loss/num_unmatchable 3.601E+02, loss/row_norm 9.819E-01}
[07/19/2024 03:24:10 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 03:24:12 gluefactory INFO] New best val: loss/total=0.11724639334020814
[07/19/2024 03:24:16 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_6_16960.tar
[07/19/2024 03:24:17 gluefactory INFO] Starting epoch 7
[07/19/2024 03:24:17 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 6, 'iteration': 302, 'match_recall': 0.9731779166497099, 'match_precision': 0.8680967531848254, 'accuracy': 0.9748998882104419, 'average_precision': 0.8520247299336886, 'loss/total': 0.11724639334020814, 'loss/last': 0.11724639334020814, 'loss/assignment_nll': 0.11724639334020814, 'loss/nll_pos': 0.17104746732996562, 'loss/nll_neg': 0.0634453195161214, 'loss/num_matchable': 145.64988676137534, 'loss/num_unmatchable': 360.13825406629604, 'loss/row_norm': 0.9818553513942591}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 03:25:11 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_032511.pkl
[07/19/2024 03:25:11 gluefactory INFO] [Validation] {match_recall 9.732E-01, match_precision 8.681E-01, accuracy 9.749E-01, average_precision 8.520E-01, loss/total 1.172E-01, loss/last 1.172E-01, loss/assignment_nll 1.172E-01, loss/nll_pos 1.710E-01, loss/nll_neg 6.345E-02, loss/num_matchable 1.456E+02, loss/num_unmatchable 3.601E+02, loss/row_norm 9.819E-01}
[07/19/2024 03:25:11 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 03:25:13 gluefactory INFO] New best val: loss/total=0.11724639334020814
[07/19/2024 03:25:16 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_6_16960.tar
[07/19/2024 03:25:18 gluefactory INFO] Starting epoch 7
[07/19/2024 03:25:18 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 6, 'iteration': 302, 'match_recall': 0.9731779166497099, 'match_precision': 0.8680967531848254, 'accuracy': 0.9748998882104419, 'average_precision': 0.8520247299336886, 'loss/total': 0.11724639334020814, 'loss/last': 0.11724639334020814, 'loss/assignment_nll': 0.11724639334020814, 'loss/nll_pos': 0.17104746732996562, 'loss/nll_neg': 0.0634453195161214, 'loss/num_matchable': 145.64988676137534, 'loss/num_unmatchable': 360.13825406629604, 'loss/row_norm': 0.9818553513942591}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 03:32:11 gluefactory INFO] [E 7 | it 0] loss {total 3.697E-01, last 1.321E-01, assignment_nll 1.321E-01, nll_pos 2.174E-01, nll_neg 4.685E-02, num_matchable 6.371E+01, num_unmatchable 1.886E+02, confidence 8.452E-02, row_norm 4.854E-01}
[07/19/2024 03:36:12 gluefactory INFO] [E 7 | it 100] loss {total 3.189E-01, last 2.991E-02, assignment_nll 2.991E-02, nll_pos 3.379E-02, nll_neg 2.604E-02, num_matchable 7.148E+01, num_unmatchable 1.809E+02, confidence 8.700E-02, row_norm 4.927E-01}
[07/19/2024 03:40:20 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_7_200_20240719_034020.pkl
[07/19/2024 03:40:20 gluefactory INFO] [E 7 | it 200] loss {total 4.776E-01, last 5.591E-02, assignment_nll 5.591E-02, nll_pos 6.009E-02, nll_neg 5.173E-02, num_matchable 1.234E+02, num_unmatchable 3.176E+02, confidence 1.426E-01, row_norm 8.645E-01}
[07/19/2024 03:44:31 gluefactory INFO] [E 7 | it 300] loss {total 4.971E-01, last 6.854E-02, assignment_nll 6.854E-02, nll_pos 6.503E-02, nll_neg 7.206E-02, num_matchable 1.220E+02, num_unmatchable 3.201E+02, confidence 1.482E-01, row_norm 8.614E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 7, 'iteration': 0, 'total': 0.6084822416305542, 'last': 0.17892150580883026, 'assignment_nll': 0.17892150580883026, 'nll_pos': 0.27831730246543884, 'nll_neg': 0.0795256644487381, 'num_matchable': 111.1953125, 'num_unmatchable': 330.0625, 'confidence': 0.14901837706565857, 'row_norm': 0.8514452576637268}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 7, 'iteration': 100, 'total': 0.5712657570838928, 'last': 0.050963111221790314, 'assignment_nll': 0.050963111221790314, 'nll_pos': 0.05998575687408447, 'nll_neg': 0.041940465569496155, 'num_matchable': 122.078125, 'num_unmatchable': 319.11328125, 'confidence': 0.15380656719207764, 'row_norm': 0.8628181219100952}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 7, 'iteration': 200, 'total': 0.47764021158218384, 'last': 0.05591220408678055, 'assignment_nll': 0.05591220408678055, 'nll_pos': 0.06009034067392349, 'nll_neg': 0.051734067499637604, 'num_matchable': 123.4296875, 'num_unmatchable': 317.62109375, 'confidence': 0.1425517499446869, 'row_norm': 0.8645414113998413}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 7, 'iteration': 300, 'total': 0.497113972902298, 'last': 0.0685424655675888, 'assignment_nll': 0.0685424655675888, 'nll_pos': 0.065025195479393, 'nll_neg': 0.0720597505569458, 'num_matchable': 122.0, 'num_unmatchable': 320.09375, 'confidence': 0.14816951751708984, 'row_norm': 0.8614315986633301}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 03:55:33 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_035533.pkl
[07/19/2024 03:55:33 gluefactory INFO] [Validation] {match_recall 9.722E-01, match_precision 8.665E-01, accuracy 9.743E-01, average_precision 8.505E-01, loss/total 1.208E-01, loss/last 1.208E-01, loss/assignment_nll 1.208E-01, loss/nll_pos 1.679E-01, loss/nll_neg 7.364E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.609E+02, loss/row_norm 9.798E-01}
[07/19/2024 03:55:37 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_7_19384.tar
[07/19/2024 03:55:39 gluefactory INFO] Starting epoch 8
[07/19/2024 03:55:39 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 7, 'iteration': 302, 'match_recall': 0.9721524162206087, 'match_precision': 0.8665092095735757, 'accuracy': 0.9743267585152776, 'average_precision': 0.850481017003561, 'loss/total': 0.12078977341266854, 'loss/last': 0.12078977341266854, 'loss/assignment_nll': 0.12078977341266854, 'loss/nll_pos': 0.16793483582478486, 'loss/nll_neg': 0.07364471028590548, 'loss/num_matchable': 144.95274861025325, 'loss/num_unmatchable': 360.8875849289685, 'loss/row_norm': 0.9797700174998079}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 03:56:34 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_035634.pkl
[07/19/2024 03:56:34 gluefactory INFO] [Validation] {match_recall 9.722E-01, match_precision 8.665E-01, accuracy 9.743E-01, average_precision 8.505E-01, loss/total 1.208E-01, loss/last 1.208E-01, loss/assignment_nll 1.208E-01, loss/nll_pos 1.679E-01, loss/nll_neg 7.364E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.609E+02, loss/row_norm 9.798E-01}
[07/19/2024 03:56:38 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_7_19384.tar
[07/19/2024 03:56:40 gluefactory INFO] Starting epoch 8
[07/19/2024 03:56:40 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 7, 'iteration': 302, 'match_recall': 0.9721524162206087, 'match_precision': 0.8665092095735757, 'accuracy': 0.9743267585152776, 'average_precision': 0.850481017003561, 'loss/total': 0.12078977341266854, 'loss/last': 0.12078977341266854, 'loss/assignment_nll': 0.12078977341266854, 'loss/nll_pos': 0.16793483582478486, 'loss/nll_neg': 0.07364471028590548, 'loss/num_matchable': 144.95274861025325, 'loss/num_unmatchable': 360.8875849289685, 'loss/row_norm': 0.9797700174998079}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 04:00:50 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_040050.pkl
[07/19/2024 04:00:50 gluefactory INFO] [Validation] {match_recall 9.722E-01, match_precision 8.665E-01, accuracy 9.743E-01, average_precision 8.505E-01, loss/total 1.208E-01, loss/last 1.208E-01, loss/assignment_nll 1.208E-01, loss/nll_pos 1.679E-01, loss/nll_neg 7.364E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.609E+02, loss/row_norm 9.798E-01}
[07/19/2024 04:00:54 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_7_19384.tar
[07/19/2024 04:00:56 gluefactory INFO] Starting epoch 8
[07/19/2024 04:00:56 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 7, 'iteration': 302, 'match_recall': 0.9721581484767324, 'match_precision': 0.8665099622831685, 'accuracy': 0.9743262987606118, 'average_precision': 0.8504824763785573, 'loss/total': 0.12078709726290697, 'loss/last': 0.12078709726290697, 'loss/assignment_nll': 0.12078709726290697, 'loss/nll_pos': 0.16793193740320378, 'loss/nll_neg': 0.07364225660124592, 'loss/num_matchable': 144.9543957175211, 'loss/num_unmatchable': 360.88604076590485, 'loss/row_norm': 0.9797705726808105}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you hve 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 04:03:30 gluefactory INFO] [E 8 | it 0] loss {total 1.905E-01, last 7.798E-02, assignment_nll 7.798E-02, nll_pos 1.295E-01, nll_neg 2.643E-02, num_matchable 3.202E+01, num_unmatchable 9.448E+01, confidence 3.919E-02, row_norm 2.415E-01}
[07/19/2024 04:06:40 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_8_20000.tar
[07/19/2024 04:07:37 gluefactory INFO] [E 8 | it 100] loss {total 1.472E-01, last 1.098E-02, assignment_nll 1.098E-02, nll_pos 1.037E-02, nll_neg 1.159E-02, num_matchable 3.584E+01, num_unmatchable 9.039E+01, confidence 4.310E-02, row_norm 2.473E-01}
[07/19/2024 04:11:35 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_8_200_20240719_041135.pkl
[07/19/2024 04:11:35 gluefactory INFO] [E 8 | it 200] loss {total 1.373E-01, last 1.589E-02, assignment_nll 1.589E-02, nll_pos 1.535E-02, nll_neg 1.644E-02, num_matchable 3.416E+01, num_unmatchable 9.161E+01, confidence 4.131E-02, row_norm 2.464E-01}
[07/19/2024 04:17:12 gluefactory INFO] [E 8 | it 300] loss {total 3.917E-01, last 6.693E-02, assignment_nll 6.693E-02, nll_pos 7.606E-02, nll_neg 5.779E-02, num_matchable 8.573E+01, num_unmatchable 2.300E+02, confidence 1.063E-01, row_norm 6.102E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 8, 'iteration': 0, 'total': 0.41500815749168396, 'last': 0.11600033193826675, 'assignment_nll': 0.11600033193826675, 'nll_pos': 0.17257782816886902, 'nll_neg': 0.0594228059053421, 'num_matchable': 82.2421875, 'num_unmatchable': 233.46875, 'confidence': 0.10287880897521973, 'row_norm': 0.6095202565193176}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 8, 'iteration': 100, 'total': 0.4004122316837311, 'last': 0.059255313128232956, 'assignment_nll': 0.059255313128232956, 'nll_pos': 0.0837908387184143, 'nll_neg': 0.03471977263689041, 'num_matchable': 88.6484375, 'num_unmatchable': 226.890625, 'confidence': 0.10600505024194717, 'row_norm': 0.6122204065322876}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 8, 'iteration': 200, 'total': 0.3372008204460144, 'last': 0.04080798849463463, 'assignment_nll': 0.04080798849463463, 'nll_pos': 0.04206263646483421, 'nll_neg': 0.03955334052443504, 'num_matchable': 85.8125, 'num_u5, 'num_unmatchable': 184.41796875, 'confidence': 0.07970865815877914, 'row_norm': 0.49375027418136597}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 8, 'iteration': 300, 'total': 0.31817853450775146, 'last': 0.05586061626672745, 'assignment_nll': 0.05586061626672745, 'nll_pos': 0.06423848867416382, 'nll_neg': 0.04748274385929108, 'num_matchable': 69.5859375, 'num_unmatchable': 182.9140625, 'confidence': 0.08585064113140106, 'row_norm': 0.487515926361084}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 04:29:15 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_042915.pkl
[07/19/2024 04:29:15 gluefactory INFO] [Validation] {match_recall 9.736E-01, match_precision 8.661E-01, accuracy 9.739E-01, average_precision 8.503E-01, loss/total 1.121E-01, loss/last 1.121E-01, loss/assignment_nll 1.121E-01, loss/nll_pos 1.543E-01, loss/nll_neg 6.985E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.797E-01}
[07/19/2024 04:29:15 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 04:29:17 gluefactory INFO] New best val: loss/total=0.11206387497932609
[07/19/2024 04:29:20 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_8_21808.tar
[07/19/2024 04:29:22 gluefactory INFO] Starting epoch 9
[07/19/2024 04:29:22 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 8, 'iteration': 302, 'match_recall': 0.9735796649391719, 'match_precision': 0.8660547769631725, 'accuracy': 0.9738971758875455, 'average_precision': 0.8502697742400112, 'loss/total': 0.11206387497932609, 'loss/last': 0.11206387497932609, 'loss/assignment_nll': 0.11206387497932609, 'loss/nll_pos': 0.15428253064497335, 'loss/nll_neg': 0.0698452196433904, 'loss/num_matchable': 145.12147416100473, 'loss/num_unmatchable': 360.7172122709491, 'loss/row_norm': 0.979717480711714}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 04:33:35 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_043335.pkl
[07/19/2024 04:33:35 gluefactory INFO] [Validation] {match_recall 9.736E-01, match_precision 8.661E-01, accuracy 9.739E-01, average_precision 8.503E-01, loss/total 1.121E-01, loss/last 1.121E-01, loss/assignment_nll 1.121E-01, loss/nll_pos 1.543E-01, loss/nll_neg 6.984E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.797E-01}
[07/19/2024 04:33:35 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 04:33:37 gluefactory INFO] New best val: loss/total=0.11205934311804186
[07/19/2024 04:33:43 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_8_21808.tar
[07/19/2024 04:33:44 gluefactory INFO] Starting epoch 9
[07/19/2024 04:33:44 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 8, 'iteration': 302, 'match_recall': 0.9735814146183636, 'match_precision': 0.8660691322454681, 'accuracy': 0.9739014406201278, 'average_precision': 0.8502787717065621, 'loss/total': 0.11205934311804186, 'loss/last': 0.11205934311804186, 'loss/assignment_nll': 0.11205934311804186, 'loss/nll_pos': 0.15427432931215262, 'loss/nll_neg': 0.06984435750214878, 'loss/num_matchable': 145.12301832406837, 'loss/num_unmatchable': 360.7157710520898, 'loss/row_norm': 0.9797178051808942}
FO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 04:35:45 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_043544.pkl
[07/19/2024 04:35:45 gluefactory INFO] [Validation] {match_recall 9.736E-01, match_precision 8.661E-01, accuracy 9.739E-01, average_precision 8.503E-01, loss/total 1.121E-01, loss/last 1.121E-01, loss/assignment_nll 1.121E-01, loss/nll_pos 1.543E-01, loss/nll_neg 6.985E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.797E-01}
[07/19/2024 04:35:45 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 04:36:05 gluefactory INFO] New best val: loss/total=0.11206387486427698
[07/19/2024 04:36:09 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_8_21808.tar
[07/19/2024 04:36:11 gluefactory INFO] Starting epoch 9
[07/19/2024 04:36:11 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 8, 'iteration': 302, 'match_recall': 0.9735796649391719, 'match_precision': 0.8660547769631725, 'accuracy': 0.9738971758875455, 'average_precision': 0.8502697742400112, 'loss/total': 0.11206387486427698, 'loss/last': 0.11206387486427698, 'loss/assignment_nll': 0.11206387486427698, 'loss/nll_pos': 0.15428252998382444, 'loss/nll_neg': 0.0698452196433904, 'loss/num_matchable': 145.12147416100473, 'loss/num_unmatchable': 360.7172122709491, 'loss/row_norm': 0.979717480711714}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 04:36:32 gluefactory INFO] [E 9 | it 0] loss {total 5.632E-01, last 9.199E-02, assignment_nll 9.199E-02, nll_pos 9.863E-02, nll_neg 8.535E-02, num_matchable 1.244E+02, num_unmatchable 3.802E+02, confidence 1.664E-01, row_norm 9.809E-01}
[07/19/2024 04:40:31 gluefactory INFO] [E 9 | it 100] loss {total 6.251E-01, last 1.101E-01, assignment_nll 1.101E-01, nll_pos 1.647E-01, nll_neg 5.546E-02, num_matchable 1.426E+02, num_unmatchable 3.633E+02, confidence 1.608E-01, row_norm 9.825E-01}
[07/19/2024 04:47:13 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_9_200_20240719_044713.pkl
[07/19/2024 04:47:13 gluefactory INFO] [E 9 | it 200] loss {total 5.458E-01, last 9.777E-02, assignment_nll 9.777E-02, nll_pos 1.371E-01, nll_neg 5.839E-02, num_matchable 1.373E+02, num_unmatchable 3.663E+02, confidence 1.519E-01, row_norm 9.869E-01}
[07/19/2024 04:53:31 gluefactory INFO] [E 9 | it 300] loss {total 6.307E-01, last 1.748E-01, assignment_nll 1.748E-01, nll_pos 2.656E-01, nll_neg 8.391E-02, num_matchable 1.302E+02, num_unmatchable 3.744E+02, confidence 1.566E-01, row_norm 9.780E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 9, 'iteration': 0, 'total': 0.5632457733154297, 'last': 0.09199404716491699, 'assignment_nll': 0.09199404716491699, 'nll_pos': 0.09863408654928207, 'nll_neg': 0.08535400032997131, 'num_matchable': 124.3984375, 'num_unmatchable': 380.1640625, 'confidence': 0.16637730598449707, 'row_norm': 0.9809427261352539}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 9, 'iteration': 100, 'total': 0.625067949295044, 'last': 0.11010412871837616, 'assignment_nll': 0.11010412871837616, 'nll_pos': 0.16474467515945435, 'nll_neg': 0.05546358600258827, 'num_matchable': 142.5546875, 'num_unmatchable': 363.296875, 'confidence': 0.16080012917518616, 'row_norm': 0.9824689030647278}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 9, 'iteration': 200, 'total': 0.5457751750946045, 'last': 0.09776579588651657, 'assignment_nll': 0.09776579588651657, 'nll_pos': 0.13713832199573517, 'nll_neg': 0.058393266052007675, 'num_matchable': 137.2734375, 'num_unmatchable': 366.26953125, 'confidence': 0.15192100405693054, 'row_norm': 0.9869339466094971}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 9, 'iteration': 300, 'total': 0.6307042837142944, 'last': 0.17476817965507507, 'assignment_nll': 0.17476817965507507, 'nll_pos': 0.2656254768371582, 'nll_neg': 0.08391088247299194, 'num_matchable': 130.2109375, 'num_unmatchable': 374.41015625, 'confidence': 0.1565827876329422, 'row_norm': 0.9780216217041016}
IINFO:alINFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 05:01:44 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_050144.pkl
[07/19/2024 05:01:44 gluefactory INFO] [Validation] {match_recall 9.769E-01, match_precision 8.707E-01, accuracy 9.759E-01, average_precision 8.558E-01, loss/total 9.961E-02, loss/last 9.961E-02, loss/assignment_nll 9.961E-02, loss/nll_pos 1.346E-01, loss/nll_neg 6.461E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.831E-01}
[07/19/2024 05:01:44 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 05:01:46 gluefactory INFO] New best val: loss/total=0.09961142170110467
[07/19/2024 05:01:49 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_9_24232.tar
[07/19/2024 05:01:50 gluefactory.utils.experiments INFO] Deleting checkpoint checkpoint_2_7264.tar
[07/19/2024 05:01:50 gluefactory INFO] Starting epoch 10
[07/19/2024 05:01:50 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 9, 'iteration': 302, 'match_recall': 0.9769118979852436, 'match_precision': 0.8706550979432633, 'accuracy': 0.9759010396574122, 'average_precision': 0.8557971866135247, 'loss/total': 0.09961142170110467, 'loss/last': 0.09961142170110467, 'loss/assignment_nll': 0.09961142170110467, 'loss/nll_pos': 0.13460823425335516, 'loss/nll_neg': 0.06461460954730762, 'loss/num_matchable': 145.04992793905703, 'loss/num_unmatchable': 360.7808832612724, 'loss/row_norm': 0.9831353461445421}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 05:04:33 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_050433.pkl
[07/19/2024 05:04:33 gluefactory INFO] [Validation] {match_recall 9.769E-01, match_precision 8.707E-01, accuracy 9.759E-01, average_precision 8.558E-01, loss/total 9.960E-02, loss/last 9.960E-02, loss/assignment_nll 9.960E-02, loss/nll_pos 1.346E-01, loss/nll_neg 6.462E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.831E-01}
[07/19/2024 05:04:33 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 05:04:34 gluefactory INFO] New best val: loss/total=0.09959801907839232
[07/19/2024 05:04:38 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_9_24232.tar
[07/19/2024 05:04:40 gluefactory INFO] Starting epoch 10
[07/19/2024 05:04:40 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 9, 'iteration': 302, 'match_recall': 0.9769166218834402, 'match_precision': 0.8706534492863974, 'accuracy': 0.9759002128255104, 'average_precision': 0.8557958559462728, 'loss/total': 0.09959801907839232, 'loss/last': 0.09959801907839232, 'loss/assignment_nll': 0.09959801907839232, 'loss/nll_pos': 0.13457999224372827, 'loss/nll_neg': 0.06461604582427681, 'loss/num_matchable': 145.0496191064443, 'loss/num_unmatchable': 360.7812435659872, 'loss/row_norm': 0.9831353139307903}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 05:05:33 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_050533.pkl
[07/19/2024 05:05:33 gluefactory INFO] [Validation] {match_recall 9.769E-01, match_precision 8.707E-01, accuracy 9.759E-01, average_precision 8.558E-01, loss/total 9.960E-02, loss/last 9.960E-02, loss/assignment_nll 9.960E-02, loss/nll_pos 1.346E-01, loss/nll_neg 6.462E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.831E-01}
[07/19/2024 05:05:33 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 05:05:35 gluefactory INFO] New best val: loss/total=0.09959801907839232
[07/19/2024 05:05:39 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_9_24232.tar
[07/19/2024 05:05:40 gluefactory INFO] Starting epoch 10
[07/19/2024 05:05:40 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 9, 'iteration': 302, 'match_recall': 0.9769166218834402, 'match_precision': 0.8706534492863974, 'accuracy': 0.9759002128255104, 'average_precision': 0.8557958559462728, 'loss/total': 0.09959801907839232, 'loss/last': 0.09959801907839232, 'loss/assignment_nll': 0.09959801907839232, 'loss/nll_pos': 0.13457999224372827, 'loss/nll_neg': 0.06461604582427681, 'loss/num_matchable': 145.0496191064443, 'loss/num_unmatchable': 360.7812435659872, 'loss/row_norm': 0.9831353139307903}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 05:12:30 gluefactory INFO] [E 10 | it 0] loss {total 3.061E-01, last 5.947E-02, assignment_nll 5.947E-02, nll_pos 7.317E-02, nll_neg 4.576E-02, num_matchable 6.300E+01, num_unmatchable 1.891E+02, confidence 8.298E-02, row_norm 4.889E-01}
factory INFO] [Validation] {match_recall 9.769E-01, match_precision 8.707E-01, accuracy 9.759E-01, average_precision 8.558E-01, loss/total 9.960E-02, loss/last 9.960E-02, loss/assignment_nll 9.960E-02, loss/nll_pos 1.346E-01, loss/nll_neg 6.462E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.831E-01}
[07/19/2024 05:12:02 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 05:12:04 gluefactory INFO] New best val: loss/total=0.09959801869949723
[07/19/2024 05:12:07 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_9_24232.tar
[07/19/2024 05:12:09 gluefactory INFO] Starting epoch 10
[07/19/2024 05:12:09 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 9, 'iteration': 302, 'match_recall': 0.9769166218834402, 'match_precision': 0.8706534492863974, 'accuracy': 0.9759002128255104, 'average_precision': 0.8557958559462728, 'loss/total': 0.09959801869949723, 'loss/last': 0.09959801869949723, 'loss/assignment_nll': 0.09959801869949723, 'loss/nll_pos': 0.1345799907066721, 'loss/nll_neg': 0.06461604582427681, 'loss/num_matchable': 145.0496191064443, 'loss/num_unmatchable': 360.7812435659872, 'loss/row_norm': 0.9831353139307903}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 05:12:30 gluefactory INFO] [E 10 | it 0] loss {total 5.752E-01, last 9.591E-02, assignment_nll 9.591E-02, nll_pos 1.057E-01, nll_neg 8.617E-02, num_matchable 1.251E+02, num_unmatchable 3.794E+02, confidence 1.630E-01, row_norm 9.799E-01}
[07/19/2024 05:16:08 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_10_25000.tar
[07/19/2024 05:16:28 gluefactory INFO] [E 10 | it 100] loss {total 5.539E-01, last 5.388E-02, assignment_nll 5.388E-02, nll_pos 5.684E-02, nll_neg 5.092E-02, num_matchable 1.469E+02, num_unmatchable 3.580E+02, confidence 1.611E-01, row_norm 9.844E-01}
[07/19/2024 05:21:09 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_10_200_20240719_052109.pkl
[07/19/2024 05:21:09 gluefactory INFO] [E 10 | it 200] loss {total 5.091E-01, last 5.480E-02, assignment_nll 5.480E-02, nll_pos 5.858E-02, nll_neg 5.102E-02, num_matchable 1.368E+02, num_unmatchable 3.675E+02, confidence 1.544E-01, row_norm 9.887E-01}
[07/19/2024 05:25:38 gluefactory INFO] [E 10 | it 300] loss {total 5.791E-01, last 1.179E-01, assignment_nll 1.179E-01, nll_pos 1.568E-01, nll_neg 7.902E-02, num_matchable 1.383E+02, num_unmatchable 3.665E+02, confidence 1.606E-01, row_norm 9.789E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 10, 'iteration': 0, 'total': 0.5752401947975159, 'last': 0.09591487050056458, 'assignment_nll': 0.09591487050056458, 'nll_pos': 0.10565511137247086, 'nll_neg': 0.08617464452981949, 'num_matchable': 125.1015625, 'num_unmatchable': 379.3671875, 'confidence': 0.16295066475868225, 'row_norm': 0.9799225330352783}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 10, 'iteration': 100, 'total': 0.5538526177406311, 'last': 0.05387643724679947, 'assignment_nll': 0.05387643724679947, 'nll_pos': 0.05683784931898117, 'nll_neg': 0.05091501772403717, 'num_matchable': 146.875, 'num_unmatchable': 357.98046875, 'confidence': 0.16110444068908691, 'row_norm': 0.9844377636909485}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 10, 'iteration': 200, 'total': 0.5090692043304443, 'last': 0.05479682981967926, 'assignment_nll': 0.05479682981967926, 'nll_pos': 0.058576058596372604, 'nll_neg': 0.05101760849356651, 'num_matchable': 136.8046875, 'num_unmatchable': 367.5234375, 'confidence': 0.15438911318778992, 'row_norm': 0.9886764883995056}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 10, 'iteration': 300, 'total': 0.5791491270065308, 'last': 0.1178896576166153, 'assignment_nll': 0.1178896576166153, 'nll_pos': 0.1567545235157013, 'nll_neg': 0.07902482151985168, 'num_matchable': 138.34375, 'num_unmatchable': 366.5078125, 'confidence': 0.16062481701374054, 'row_norm': 0.978922963142395}
IINFO:aINFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 05:33:55 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_053355.pkl
[07/19/2024 05:33:55 gluefactory INFO] [Validation] {match_recall 9.775E-01, match_precision 8.745E-01, accuracy 9.773E-01, average_precision 8.596E-01, loss/total 9.582E-02, loss/last 9.582E-02, loss/assignment_nll 9.582E-02, loss/nll_pos 1.315E-01, loss/nll_neg 6.016E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.842E-01}
[07/19/2024 05:33:55 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 05:33:57 gluefactory INFO] New best val: loss/total=0.09581809049859161
[07/19/2024 05:34:00 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_10_26656.tar
[07/19/2024 05:34:01 gluefactory.utils.experiments INFO] Deleting checkpoint checkpoint_4_10000.tar
[07/19/2024 05:34:01 gluefactory INFO] Starting epoch 11
[07/19/2024 05:34:01 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 10, 'iteration': 302, 'match_recall': 0.977508303901652, 'match_precision': 0.8744883592147897, 'accuracy': 0.9772859782084826, 'average_precision': 0.8596158305730304, 'loss/total': 0.09581809049859161, 'loss/last': 0.09581809049859161, 'loss/assignment_nll': 0.09581809049859161, 'loss/nll_pos': 0.1314766597760535, 'loss/nll_neg': 0.060159520072172595, 'loss/num_matchable': 145.13176858142887, 'loss/num_unmatchable': 360.6988882025942, 'loss/row_norm': 0.9841530644307049}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 05:36:41 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_053641.pkl
[07/19/2024 05:36:41 gluefactory INFO] [Validation] {match_recall 9.775E-01, match_precision 8.745E-01, accuracy 9.773E-01, average_precision 8.596E-01, loss/total 9.583E-02, loss/last 9.583E-02, loss/assignment_nll 9.583E-02, loss/nll_pos 1.315E-01, loss/nll_neg 6.016E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.842E-01}
[07/19/2024 05:37:41 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 05:36:43 gluefactory INFO] New best val: loss/total=0.09582898173376148
[07/19/2024 05:36:47 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_10_26656.tar
[07/19/2024 05:38:11 gluefactory INFO] New best val: loss/total=0.09582199530884246
[07/19/2024 05:38:15 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_10_26656.tar
[07/19/2024 05:38:17 gluefactory INFO] Starting epoch 11
[07/19/2024 05:38:17 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 10, 'iteration': 302, 'match_recall': 0.9775097827153464, 'match_precision': 0.8744866915119266, 'accuracy': 0.9772864307963556, 'average_precision': 0.8596173773239933, 'loss/total': 0.09582199530884246, 'loss/last': 0.09582199530884246, 'loss/assignment_nll': 0.09582199530884246, 'loss/nll_pos': 0.13148279870874277, 'loss/nll_neg': 0.060161191281349245, 'loss/num_matchable': 145.1313568046119, 'loss/num_unmatchable': 360.6993514515133, 'loss/row_norm': 0.9841526971571186}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
0.0001 to 0.0001
Data to be logged: {'epoch': 10, 'iteration': 302, 'match_recall': 0.9775097827153464, 'match_precision': 0.8744866915119266, 'accuracy': 0.9772864307963556, 'average_precision': 0.8596173773239933, 'loss/total': 0.09582199530884246, 'loss/last': 0.09582199530884246, 'loss/assignment_nll': 0.09582199530884246, 'loss/nll_pos': 0.13148279870874277, 'loss/nll_neg': 0.060161191281349245, 'loss/num_matchable': 145.1313568046119, 'loss/num_unmatchable': 360.6993514515133, 'loss/row_norm': 0.9841526971571186}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 05:41:59 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_054159.pkl
[07/19/2024 05:41:59 gluefactory INFO] [Validation] {match_recall 9.775E-01, match_precision 8.745E-01, accuracy 9.773E-01, average_precision 8.596E-01, loss/total 9.582E-02, loss/last 9.582E-02, loss/assignment_nll 9.582E-02, loss/nll_pos 1.315E-01, loss/nll_neg 6.016E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.607E+02, loss/row_norm 9.842E-01}
[07/19/2024 05:41:51 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 05:41:52 gluefactory INFO] New best val: loss/total=0.09582199501278274
[07/19/2024 05:41:56 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_10_26656.tar
[07/19/2024 05:41:58 gluefactory INFO] Starting epoch 11
[07/19/2024 05:41:58 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 10, 'iteration': 302, 'match_recall': 0.9775097827153464, 'match_precision': 0.8744866915119266, 'accuracy': 0.9772864307963556, 'average_precision': 0.8596173773239933, 'loss/total': 0.09582199501278274, 'loss/last': 0.09582199501278274, 'loss/assignment_nll': 0.09582199501278274, 'loss/nll_pos': 0.13148279869953883, 'loss/nll_neg': 0.060161191281349245, 'loss/num_matchable': 145.1313568046119, 'loss/num_unmatchable': 360.6993514515133, 'loss/row_norm': 0.9841526971571186}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 05:44:40 gluefactory INFO] [E 11 | it 0] loss {total 1.371E-01, last 2.225E-02, assignment_nll 2.225E-02, nll_pos 2.679E-02, nll_neg 1.771E-02, num_matchable 3.202E+01, num_unmatchable 9.411E+01, confidence 4.021E-02, row_norm 2.463E-01}
[07/19/2024 05:49:47 gluefactory INFO] [E 11 | it 100] loss {total 1.430E-01, last 1.392E-02, assignment_nll 1.392E-02, nll_pos 1.597E-02, nll_neg 1.187E-02, num_matchable 3.820E+01, num_unmatchable 8.796E+01, confidence 4.157E-02, row_norm 2.474E-01}
[07/19/2024 05:54:45 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_11_200_20240719_055445.pkl
[07/19/2024 05:54:45 gluefactory INFO] [E 11 | it 200] loss {total 1.301E-01, last 1.647E-02, assignment_nll 1.647E-02, nll_pos 1.783E-02, nll_neg 1.511E-02, num_matchable 3.640E+01, num_unmatchable 8.971E+01, confidence 3.779E-02, row_norm 2.471E-01}
[07/19/2024 06:01:39 gluefactory INFO] [E 11 | it 300] loss {total 4.780E-01, last 6.701E-02, assignment_nll 6.701E-02, nll_pos 6.776E-02, nll_neg 6.627E-02, num_matchable 1.214E+02, num_unmatchable 3.205E+02, confidence 1.383E-01, row_norm 8.590E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 11, 'iteration': 0, 'total': 0.4701029062271118, 'last': 0.07014874368906021, 'assignment_nll': 0.07014874368906021, 'nll_pos': 0.07649897783994675, 'nll_neg': 0.06379850953817368, 'num_matchable': 112.640625, 'num_unmatchable': 328.8046875, 'confidence': 0.14011701941490173, 'row_norm': 0.8607200980186462}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 11, 'iteration': 100, 'total': 0.5339773893356323, 'last': 0.08466776460409164, 'assignment_nll': 0.08466776460409164, 'nll_pos': 0.12593907117843628, 'nll_neg': 0.04339646175503731, 'num_matchable': 130.4765625, 'num_unmatchable': 311.3671875, 'confidence': 0.14325541257858276, 'row_norm': 0.8624038100242615}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 11, 'iteration': 200, 'total': 0.4530918598175049, 'last': 0.061728812754154205, 'assignment_nll': 0.061728812754154205, 'nll_pos': 0.07126956433057785, 'nll_neg': 0.05218805372714996, 'num_matchable': 121.4453125, 'num_unmatchable': 320.3671875, 'confidence': 0.1310410350561142, 'row_norm': 0.8636348843574524}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 11, 'iteration': 300, 'total': 0.4779666066169739, 'last': 0.0670144259929657, 'assignment_nll': 0.0670144259929657, 'nll_pos': 0.06775568425655365, 'nll_neg': 0.06627317517995834, 'num_matchable': 121.4140625, 'num_unmatchable': 320.4609375, 'confidence': 0.13831964135169983, 'row_norm': 0.8589667081832886}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 06:12:40 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_061240.pkl
[07/19/2024 06:12:40 gluefactory INFO] [Validation] {match_recall 9.781E-01, match_precision 8.687E-01, accuracy 9.751E-01, average_precision 8.545E-01, loss/total 9.959E-02, loss/last 9.959E-02, loss/assignment_nll 9.959E-02, loss/nll_pos 1.313E-01, loss/nll_neg 6.787E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.827E-01}
[07/19/2024 06:12:44 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_11_29080.tar
[07/19/2024 06:12:45 gluefactory INFO] Starting epoch 12
[07/19/2024 06:12:45 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 11, 'iteration': 302, 'match_recall': 0.9781288125766804, 'match_precision': 0.8687282800674438, 'accuracy': 0.9750706944810232, 'average_precision': 0.8545315088203764, 'loss/total': 0.09958807810518215, 'loss/last': 0.09958807810518215, 'loss/assignment_nll': 0.09958807810518215, 'loss/nll_pos': 0.1313036004145029, 'loss/nll_neg': 0.06787255640236198, 'loss/num_matchable': 145.0705167799053, 'loss/num_unmatchable': 360.75782375952235, 'loss/row_norm': 0.9827291153452635}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 06:13:43 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_061343.pkl
[07/19/2024 06:13:43 gluefactory INFO] [Validation] {match_recall 9.781E-01, match_precision 8.687E-01, accuracy 9.751E-01, average_precision 8.545E-01, loss/total 9.959E-02, loss/last 9.959E-02, loss/assignment_nll 9.959E-02, loss/nll_pos 1.313E-01, loss/nll_neg 6.787E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.827E-01}
[07/19/2024 06:13:47 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_11_29080.tar
[07/19/2024 06:13:48 gluefactory INFO] Starting epoch 12
[07/19/2024 06:13:48 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 11, 'iteration': 302, 'match_recall': 0.9781288125766804, 'match_precision': 0.8687282800674438, 'accuracy': 0.9750706944810232, 'average_precision': 0.8545315088203764, 'loss/total': 0.09958807810518215, 'loss/last': 0.09958807810518215, 'loss/assignment_nll': 0.09958807810518215, 'loss/nll_pos': 0.1313036004145029, 'loss/nll_neg': 0.06787255640236198, 'loss/num_matchable': 145.0705167799053, 'loss/num_unmatchable': 360.75782375952235, 'loss/row_norm': 0.9827291153452635}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 06:17:56 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_061756.pkl
[07/19/2024 06:17:56 gluefactory INFO] [Validation] {match_recall 9.781E-01, match_precision 8.687E-01, accuracy 9.751E-01, average_precision 8.545E-01, loss/total 9.959E-02, loss/last 9.959E-02, loss/assignment_nll 9.959E-02, loss/nll_pos 1.313E-01, loss/nll_neg 6.787E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.827E-01}
[07/19/2024 06:18:00 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_11_29080.tar
[07/19/2024 06:18:02 gluefactory INFO] Starting epoch 12
[07/19/2024 06:18:02 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 11, 'iteration': 302, 'match_recall': 0.9781263479853648, 'match_precision': 0.8687280714450512, 'accuracy': 0.97507044128707, 'average_precision': 0.8545342929721099, 'loss/total': 0.09959237468029114, 'loss/last': 0.09959237468029114, 'loss/assignment_nll': 0.09959237468029114, 'loss/nll_pos': 0.1313113797816897, 'loss/nll_neg': 0.06787336886999831, 'loss/num_matchable': 145.07216388717316, 'loss/num_unmatchable': 360.75612518015237, 'loss/row_norm': 0.9827280728223516}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 06:20:12 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_062012.pkl
[07/19/2024 06:20:12 gluefactory INFO] [Validation] {match_recall 9.781E-01, match_precision 8.687E-01, accuracy 9.751E-01, average_precision 8.545E-01, loss/total 9.959E-02, loss/last 9.959E-02, loss/assignment_nll 9.959E-02, loss/nll_pos 1.313E-01, loss/nll_neg 6.787E-02, loss/num_matchable 1.451E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.827E-01}
[07/19/2024 06:20:16 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_11_29080.tar
[07/19/2024 06:20:17 gluefactory INFO] Starting epoch 12
[07/19/2024 06:20:17 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 11, 'iteration': 302, 'match_recall': 0.9781288125766804, 'match_precision': 0.8687282800674438, 'accuracy': 0.9750706944810232, 'average_precision': 0.8545315088203764, 'loss/total': 0.09958807804075465, 'loss/last': 0.09958807804075465, 'loss/assignment_nll': 0.09958807804075465, 'loss/nll_pos': 0.13130360016522982, 'loss/nll_neg': 0.06787255640236198, 'loss/num_matchable': 145.0705167799053, 'loss/num_unmatchable': 360.75782375952235, 'loss/row_norm': 0.9827291153452635}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 06:20:43 gluefactory INFO] [E 12 | it 0] loss {total 5.111E-01, last 7.886E-02, assignment_nll 7.886E-02, nll_pos 7.323E-02, nll_neg 8.449E-02, num_matchable 1.240E+02, num_unmatchable 3.806E+02, confidence 1.557E-01, row_norm 9.839E-01}
[07/19/2024 06:24:43 gluefactory INFO] [E 12 | it 100] loss {total 5.895E-01, last 1.143E-01, assignment_nll 1.143E-01, nll_pos 1.804E-01, nll_neg 4.823E-02, num_matchable 1.474E+02, num_unmatchable 3.579E+02, confidence 1.545E-01, row_norm 9.822E-01}
[07/19/2024 06:26:12 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_12_30000.tar
[07/19/2024 06:29:32 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_12_200_20240719_062932.pkl
[07/19/2024 06:29:32 gluefactory INFO] [E 12 | it 200] loss {total 4.975E-01, last 7.070E-02, assignment_nll 7.070E-02, nll_pos 8.656E-02, nll_neg 5.484E-02, num_matchable 1.384E+02, num_unmatchable 3.658E+02, confidence 1.459E-01, row_norm 9.876E-01}
[07/19/2024 06:33:20 gluefactory INFO] [E 12 | it 300] loss {total 3.669E-01, last 8.418E-02, assignment_nll 8.418E-02, nll_pos 1.174E-01, nll_neg 5.094E-02, num_matchable 8.726E+01, num_unmatchable 2.278E+02, confidence 9.969E-02, row_norm 6.113E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 12, 'iteration': 0, 'total': 0.31643933057785034, 'last': 0.04607146233320236, 'assignment_nll': 0.04607146233320236, 'nll_pos': 0.041402630507946014, 'nll_neg': 0.05074029415845871, 'num_matchable': 78.1875, 'num_unmatchable': 237.046875, 'confidence': 0.09758885204792023, 'row_norm': 0.6152355670928955}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 12, 'iteration': 100, 'total': 0.36347052454948425, 'last': 0.0688772052526474, 'assignment_nll': 0.0688772052526474, 'nll_pos': 0.11010728031396866, 'nll_neg': 0.027647122740745544, 'num_matchable': 93.984375, 'num_unmatchable': 221.8671875, 'confidence': 0.096393883228302, 'row_norm': 0.6155098676681519}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 12, 'iteration': 200, 'total': 0.32513731718063354, 'last': 0.050260357558727264, 'assignment_nll': 0.050260357558727264, 'nll_pos': 0.06305480003356934, 'nll_neg': 0.03746592253446579, 'num_matchable': 84.2734375, 'num_unmatchable': 230.6171875, 'confidence': 0.09249962866306305, 'row_norm': 0.6162296533584595}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 12, 'iteran': 300, 'total': 0.2976231873035431, 'last': 0.07457821071147919, 'assignment_nll': 0.07457821071147919, 'nll_pos': 0.10804644972085953, 'nll_neg': 0.041109975427389145, 'num_matchable': 67.8671875, 'num_unmatchable': 184.2109375, 'confidence': 0.07805308699607849, 'row_norm': 0.48835986852645874}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 06:45:22 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_064522.pkl
[07/19/2024 06:45:22 gluefactory INFO] [Validation] {match_recall 9.785E-01, match_precision 8.736E-01, accuracy 9.770E-01, average_precision 8.591E-01, loss/total 9.522E-02, loss/last 9.522E-02, loss/assignment_nll 9.522E-02, loss/nll_pos 1.290E-01, loss/nll_neg 6.144E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.834E-01}
[07/19/2024 06:45:22 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 06:45:24 gluefactory INFO] New best val: loss/total=0.09521832219935725
[07/19/2024 06:45:27 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_12_31504.tar
[07/19/2024 06:45:29 gluefactory INFO] Starting epoch 13
[07/19/2024 06:45:29 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 12, 'iteration': 302, 'match_recall': 0.978499435644619, 'match_precision': 0.8736458606918285, 'accuracy': 0.9770432769307291, 'average_precision': 0.8591328315438373, 'loss/total': 0.09521832219935725, 'loss/last': 0.09521832219935725, 'loss/assignment_nll': 0.09521832219935725, 'loss/nll_pos': 0.12900161114017986, 'loss/nll_neg': 0.061435031363484, 'loss/num_matchable': 145.02779493514515, 'loss/num_unmatchable': 360.80811200329424, 'loss/row_norm': 0.9833825477631907}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 06:52:20 gluefactory INFO] [E 13 | it 0] loss {total 6.492E-02, last 7.370E-03, assignment_nll 7.370E-03, nll_pos 6.540E-03, nll_neg 8.199E-03, num_matchable 1.564E+01, num_unmatchable 4.720E+01, confidence 2.084E-02, row_norm 1.229E-01}
factory INFO] [Validation] {match_recall 9.785E-01, match_precision 8.736E-01, accuracy 9.770E-01, average_precision 8.591E-01, loss/total 9.522E-02, loss/last 9.522E-02, loss/assignment_nll 9.522E-02, loss/nll_pos 1.290E-01, loss/nll_neg 6.144E-02, loss/num_matchable 1.450E+02, loss/num_unmatchable 3.608E+02, loss/row_norm 9.834E-01}
[07/19/2024 06:51:52 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 06:51:53 gluefactory INFO] New best val: loss/total=0.09521832167166532
[07/19/2024 06:51:57 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_12_31504.tar
[07/19/2024 06:51:59 gluefactory INFO] Starting epoch 13
[07/19/2024 06:51:59 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 12, 'iteration': 302, 'match_recall': 0.978499435644619, 'match_precision': 0.8736458606918285, 'accuracy': 0.9770432769307291, 'average_precision': 0.8591328315438373, 'loss/total': 0.09521832167166532, 'loss/last': 0.09521832167166532, 'loss/assignment_nll': 0.09521832167166532, 'loss/nll_pos': 0.12900161096377122, 'loss/nll_neg': 0.061435031363484, 'loss/num_matchable': 145.02779493514515, 'loss/num_unmatchable': 360.80811200329424, 'loss/row_norm': 0.9833825477631907}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 06:52:20 gluefactory INFO] [E 13 | it 0] loss {total 5.229E-01, last 7.394E-02, assignment_nll 7.394E-02, nll_pos 7.727E-02, nll_neg 7.062E-02, num_matchable 1.259E+02, num_unmatchable 3.781E+02, confidence 1.596E-01, row_norm 9.854E-01}
[07/19/2024 06:56:08 gluefactory INFO] [E 13 | it 100] loss {total 5.700E-01, last 4.629E-02, assignment_nll 4.629E-02, nll_pos 5.322E-02, nll_neg 3.937E-02, num_matchable 1.477E+02, num_unmatchable 3.572E+02, confidence 1.620E-01, row_norm 9.896E-01}
[07/19/2024 06:59:58 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_13_200_20240719_065958.pkl
[07/19/2024 06:59:58 gluefactory INFO] [E 13 | it 200] loss {total 4.772E-01, last 5.273E-02, assignment_nll 5.273E-02, nll_pos 5.082E-02, nll_neg 5.464E-02, num_matchable 1.346E+02, num_unmatchable 3.699E+02, confidence 1.427E-01, row_norm 9.900E-01}
[07/19/2024 07:03:47 gluefactory INFO] [E 13 | it 300] loss {total 5.374E-01, last 9.557E-02, assignment_nll 9.557E-02, nll_pos 1.108E-01, nll_neg 8.032E-02, num_matchable 1.397E+02, num_unmatchable 3.644E+02, confidence 1.559E-01, row_norm 9.822E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 13, 'iteration': 0, 'total': 0.5229314565658569, 'last': 0.07394435256719589, 'assignment_nll': 0.07394435256719589, 'nll_pos': 0.07727114856243134, 'nll_neg': 0.07061754912137985, 'num_matchable': 125.90625, 'num_unmatchable': 378.09375, 'confidence': 0.15964552760124207, 'row_norm': 0.9854207634925842}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 13, 'iteration': 100, 'total': 0.5700095891952515, 'last': 0.046294569969177246, 'assignment_nll': 0.046294569969177246, 'nll_pos': 0.053221143782138824, 'nll_neg': 0.039367999881505966, 'num_matchable': 147.734375, 'num_unmatchable': 357.22265625, 'confidence': 0.1620023250579834, 'row_norm': 0.9895974397659302}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 13, 'iteration': 200, 'total': 0.4772421419620514, 'last': 0.052730660885572433, 'assignment_nll': 0.052730660885572433, 'nll_pos': 0.05082148313522339, 'nll_neg': 0.05463983491063118, 'num_matchable': 134.6171875, 'num_unmatchable': 369.93359375, 'confidence': 0.142684206366539, 'row_norm': 0.9899759292602539}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 13, 'iteration': 300, 'total': 0.537379264831543, 'last': 0.09556866437196732, 'assignment_nll': 0.09556866437196732, 'nll_pos': 0.1108156219124794, 'nll_neg': 0.08032169938087463, 'num_matchable': 139.6953125, 'num_unmatchable': 364.37890625, 'confidence': 0.1559063047170639, 'row_norm': 0.982157826423645}
INFO:aININFO:alINFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 07:11:56 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_071156.pkl
[07/19/2024 07:11:56 gluefactory INFO] [Validation] {match_recall 9.797E-01, match_precision 8.733E-01, accuracy 9.766E-01, average_precision 8.591E-01, loss/total 8.997E-02, loss/last 8.997E-02, loss/assignment_nll 8.997E-02, loss/nll_pos 1.215E-01, loss/nll_neg 5.839E-02, loss/num_matchable 1.456E+02, loss/num_unmatchable 3.602E+02, loss/row_norm 9.838E-01}
[07/19/2024 07:11:57 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 07:11:58 gluefactory INFO] New best val: loss/total=0.08996732140890679
[07/19/2024 07:12:01 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_13_33928.tar
[07/19/2024 07:12:03 gluefactory.utils.experiments INFO] Deleting checkpoint checkpoint_6_16960.tar
[07/19/2024 07:12:03 gluefactory INFO] Starting epoch 14
[07/19/2024 07:12:03 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 13, 'iteration': 302, 'match_recall': 0.9796985180161484, 'match_precision': 0.8733189972629511, 'accuracy': 0.9766342749388492, 'average_precision': 0.8591296663494397, 'loss/total': 0.08996732140890679, 'loss/last': 0.08996732140890679, 'loss/assignment_nll': 0.08996732140890679, 'loss/nll_pos': 0.12154089043786281, 'loss/nll_neg': 0.058393752471990044, 'loss/num_matchable': 145.61045913115092, 'loss/num_unmatchable': 360.18782170063827, 'loss/row_norm': 0.9837754447690396}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 07:14:46 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_071446.pkl
[07/19/2024 07:14:46 gluefactory INFO] [Validation] {match_recall 9.797E-01, match_precision 8.733E-01, accuracy 9.766E-01, average_precision 8.591E-01, loss/total 8.996E-02, loss/last 8.996E-02, loss/assignment_nll 8.996E-02, loss/nll_pos 1.215E-01, loss/nll_neg 5.839E-02, loss/num_matchable 1.456E+02, loss/num_unmatchable 3.602E+02, loss/row_norm 9.838E-01}
[07/19/2024 07:14:46 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 07:14:48 gluefactory INFO] New best val: loss/total=0.08995647156438127
[07/19/2024 07:14:52 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_13_33928.tar
[07/19/2024 07:14:53 gluefactory INFO] Starting epoch 14
[07/19/2024 07:14:53 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 13, 'iteration': 302, 'match_recall': 0.9796917939937136, 'match_precision': 0.8733170710646716, 'accuracy': 0.9766328914533163, 'average_precision': 0.8591263925977664, 'loss/total': 0.08995647156438127, 'loss/last': 0.08995647156438127, 'loss/assignment_nll': 0.08995647156438127, 'loss/nll_pos': 0.12152245217834158, 'loss/nll_neg': 0.058390491094615866, 'loss/num_matchable': 145.61045913115092, 'loss/num_unmatchable': 360.18782170063827, 'loss/row_norm': 0.9837767239679199}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 07:15:47 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_071547.pkl
[07/19/2024 07:15:47 gluefactory INFO] [Validation] {match_recall 9.797E-01, match_precision 8.733E-01, accuracy 9.766E-01, average_precision 8.591E-01, loss/total 8.996E-02, loss/last 8.996E-02, loss/assignment_nll 8.996E-02, loss/nll_pos 1.215E-01, loss/nll_neg 5.839E-02, loss/num_matchable 1.456E+02, loss/num_unmatchable 3.602E+02, loss/row_norm 9.838E-01}
[07/19/2024 07:15:47 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 07:15:49 gluefactory INFO] New best val: loss/total=0.08995647156438127
[07/19/2024 07:15:53 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_13_33928.tar
[07/19/2024 07:15:54 gluefactory INFO] Starting epoch 14
[07/19/2024 07:15:54 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 13, 'iteration': 302, 'match_recall': 0.9796917939937136, 'match_precision': 0.8733170710646716, 'accuracy': 0.9766328914533163, 'average_precision': 0.8591263925977664, 'loss/total': 0.08995647156438127, 'loss/last': 0.08995647156438127, 'loss/assignment_nll': 0.08995647156438127, 'loss/nll_pos': 0.12152245217834158, 'loss/nll_neg': 0.058390491094615866, 'loss/num_matchable': 145.61045913115092, 'loss/num_unmatchable': 360.18782170063827, 'loss/row_norm': 0.9837767239679199}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 07:22:45 gluefactory INFO] [E 14 | it 0] loss {total 1.842E-01, last 2.679E-02, assignment_nll 2.679E-02, nll_pos 2.534E-02, nll_neg 2.824E-02, num_matchable 4.888E+01, num_unmatchable 1.405E+02, confidence 5.647E-02, row_norm 3.693E-01}
factory INFO] [Validation] {match_recall 9.797E-01, match_precision 8.733E-01, accuracy 9.766E-01, average_precision 8.591E-01, loss/total 8.996E-02, loss/last 8.996E-02, loss/assignment_nll 8.996E-02, loss/nll_pos 1.215E-01, loss/nll_neg 5.839E-02, loss/num_matchable 1.456E+02, loss/num_unmatchable 3.602E+02, loss/row_norm 9.838E-01}
[07/19/2024 07:22:18 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 07:22:19 gluefactory INFO] New best val: loss/total=0.08995647169323628
[07/19/2024 07:22:23 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_13_33928.tar
[07/19/2024 07:22:25 gluefactory INFO] Starting epoch 14
[07/19/2024 07:22:25 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 13, 'iteration': 302, 'match_recall': 0.9796917939937136, 'match_precision': 0.8733170710646716, 'accuracy': 0.9766328914533163, 'average_precision': 0.8591263925977664, 'loss/total': 0.08995647169323628, 'loss/last': 0.08995647169323628, 'loss/assignment_nll': 0.08995647169323628, 'loss/nll_pos': 0.12152245206175848, 'loss/nll_neg': 0.058390491094615866, 'loss/num_matchable': 145.61045913115092, 'loss/num_unmatchable': 360.18782170063827, 'loss/row_norm': 0.9837767239679199}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 07:22:45 gluefactory INFO] [E 14 | it 0] loss {total 5.176E-01, last 8.581E-02, assignment_nll 8.581E-02, nll_pos 9.852E-02, nll_neg 7.310E-02, num_matchable 1.249E+02, num_unmatchable 3.801E+02, confidence 1.510E-01, row_norm 9.840E-01}
[07/19/2024 07:26:34 gluefactory INFO] [E 14 | it 100] loss {total 5.159E-01, last 4.667E-02, assignment_nll 4.667E-02, nll_pos 5.404E-02, nll_neg 3.930E-02, num_matchable 1.514E+02, num_unmatchable 3.539E+02, confidence 1.522E-01, row_norm 9.886E-01}
[07/19/2024 07:27:50 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_14_35000.tar
[07/19/2024 07:30:24 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_14_200_20240719_073024.pkl
[07/19/2024 07:30:24 gluefactory INFO] [E 14 | it 200] loss {total 4.569E-01, last 5.034E-02, assignment_nll 5.034E-02, nll_pos 5.268E-02, nll_neg 4.799E-02, num_matchable 1.418E+02, num_unmatchable 3.631E+02, confidence 1.414E-01, row_norm 9.905E-01}
[07/19/2024 07:34:13 gluefactory INFO] [E 14 | it 300] loss {total 1.663E-01, last 6.254E-02, assignment_nll 6.254E-02, nll_pos 1.033E-01, nll_neg 2.181E-02, num_matchable 3.291E+01, num_unmatchable 9.346E+01, confidence 3.551E-02, row_norm 2.420E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 14, 'iteration': 0, 'total': 0.11734546720981598, 'last': 0.01612767204642296, 'assignment_nll': 0.01612767204642296, 'nll_pos': 0.015072465874254704, 'nll_neg': 0.017182881012558937, 'num_matchable': 32.3125, 'num_unmatchable': 94.13671875, 'confidence': 0.03631329536437988, 'row_norm': 0.24662044644355774}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 14, 'iteration': 100, 'total': 0.1303878277540207, 'last': 0.00763398502022028, 'assignment_nll': 0.00763398502022028, 'nll_pos': 0.007122581824660301, 'nll_neg': 0.008145388215780258, 'num_matchable': 38.4453125, 'num_unmatchable': 87.72265625, 'confidence': 0.03995055332779884, 'row_norm': 0.2476438283920288}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 14, 'iteration': 200, 'total': 0.1030401960015297, 'last': 0.009629645384848118, 'assignment_nll': 0.009629645384848118, 'nll_pos': 0.008826246485114098, 'nll_neg': 0.010433044284582138, 'num_matchable': 35.5546875, 'num_unmatchable': 90.765625, 'confidence': 0.03299621492624283, 'row_norm': 0.2481810450553894}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 14, 'iteration': 300, 'total': 0.16632197797298431, 'last': 0.062543585896492, 'assignment_nll': 0.062543585896492, 'nll_pos': 0.10327466577291489, 'nll_neg': 0.02181249111890793, 'num_matchable': 32.90625, 'num_unmatchable': 93.4609375, 'confidence': 0.03550752252340317, 'row_norm': 0.2419990599155426}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
tations
umentations
[07/19/2024 07:42:23 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_074223.pkl
[07/19/2024 07:42:23 gluefactory INFO] [Validation] {match_recall 9.798E-01, match_precision 8.795E-01, accuracy 9.788E-01, average_precision 8.649E-01, loss/total 8.826E-02, loss/last 8.826E-02, loss/assignment_nll 8.826E-02, loss/nll_pos 1.213E-01, loss/nll_neg 5.517E-02, loss/num_matchable 1.456E+02, loss/num_unmatchable 3.602E+02, loss/row_norm 9.852E-01}
[07/19/2024 07:42:23 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 07:42:25 gluefactory INFO] New best val: loss/total=0.08825756831024442
[07/19/2024 07:42:28 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_14_36352.tar
[07/19/2024 07:42:29 gluefactory.utils.experiments INFO] Deleting checkpoint checkpoint_8_20000.tar
[07/19/2024 07:42:29 gluefactory INFO] Starting epoch 15
[07/19/2024 07:42:29 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 14, 'iteration': 302, 'match_recall': 0.9798099195022064, 'match_precision': 0.8794660609470907, 'accuracy': 0.9788280125137512, 'average_precision': 0.8648875176059824, 'loss/total': 0.08825756831024442, 'loss/last': 0.08825756831024442, 'loss/assignment_nll': 0.08825756831024442, 'loss/nll_pos': 0.12134171604100805, 'loss/nll_neg': 0.055173420284188084, 'loss/num_matchable': 145.5717521103562, 'loss/num_unmatchable': 360.22040354128063, 'loss/row_norm': 0.9851923967944218}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 07:45:12 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_074512.pkl
[07/19/2024 07:45:12 gluefactory INFO] [Validation] {match_recall 9.798E-01, match_precision 8.795E-01, accuracy 9.788E-01, average_precision 8.649E-01, loss/total 8.827E-02, loss/last 8.827E-02, loss/assignment_nll 8.827E-02, loss/nll_pos 1.214E-01, loss/nll_neg 5.518E-02, loss/num_matchable 1.456E+02, loss/num_unmatchable 3.602E+02, loss/row_norm 9.852E-01}
[07/19/2024 07:45:12 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 07:45:14 gluefactory INFO] New best val: loss/total=0.08827067092034198
[07/19/2024 07:45:18 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_14_36352.tar
[07/19/2024 07:45:19 gluefactory INFO] Starting epoch 15
[07/19/2024 07:45:19 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 14, 'iteration': 302, 'match_recall': 0.9798099143970938, 'match_precision': 0.8794666976135455, 'accuracy': 0.9788274019619111, 'average_precision': 0.8648883478740176, 'loss/total': 0.08827067092034198, 'loss/last': 0.08827067092034198, 'loss/assignment_nll': 0.08827067092034198, 'loss/nll_pos': 0.12136586556718865, 'loss/nll_neg': 0.05517547658038883, 'loss/num_matchable': 145.5717521103562, 'loss/num_unmatchable': 360.2203005970764, 'loss/row_norm': 0.9851911551966598}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 07:46:15 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_074615.pkl
[07/19/2024 07:46:15 gluefactory INFO] [Validation] {match_recall 9.798E-01, match_precision 8.795E-01, accuracy 9.788E-01, average_precision 8.649E-01, loss/total 8.827E-02, loss/last 8.827E-02, loss/assignment_nll 8.827E-02, loss/nll_pos 1.214E-01, loss/nll_neg 5.518E-02, loss/num_matchable 1.456E+02, loss/num_unmatchable 3.602E+02, loss/row_norm 9.852E-01}
[07/19/2024 07:46:15 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 07:46:16 gluefactory INFO] New best val: loss/total=0.08827067092034198
[07/19/2024 07:46:20 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_14_36352.tar
[07/19/2024 07:46:21 gluefactory INFO] Starting epoch 15
[07/19/2024 07:46:22 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 14, 'iteration': 302, 'match_recall': 0.9798099143970938, 'match_precision': 0.8794666976135455, 'accuracy': 0.9788274019619111, 'average_precision': 0.8648883478740176, 'loss/total': 0.08827067092034198, 'loss/last': 0.08827067092034198, 'loss/assignment_nll': 0.08827067092034198, 'loss/nll_pos': 0.12136586556718865, 'loss/nll_neg': 0.05517547658038883, 'loss/num_matchable': 145.5717521103562, 'loss/num_unmatchable': 360.2203005970764, 'loss/row_norm': 0.9851911551966598}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 07:50:31 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_075022.pkl
[07/19/2024 07:50:22 gluefactory INFO] [Validation] {match_recall 9.798E-01, match_precision 8.795E-01, accuracy 9.788E-01, average_precision 8.649E-01, loss/total 8.827E-02, loss/last 8.827E-02, loss/assignment_nll 8.827E-02, loss/nll_pos 1.214E-01, loss/nll_neg 5.518E-02, loss/num_matchable 1.456E+02, loss/num_unmatchable 3.602E+02, loss/row_norm 9.852E-01}
[07/19/2024 07:50:22 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 07:50:24 gluefactory INFO] New best val: loss/total=0.08827067079762292
[07/19/2024 07:50:28 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_14_36352.tar
[07/19/2024 07:50:29 gluefactory INFO] Starting epoch 15
[07/19/2024 07:50:29 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 14, 'iteration': 302, 'match_recall': 0.9798099143970938, 'match_precision': 0.8794666976135455, 'accuracy': 0.9788274019619111, 'average_precision': 0.8648883478740176, 'loss/total': 0.08827067079762292, 'loss/last': 0.08827067079762292, 'loss/assignment_nll': 0.08827067079762292, 'loss/nll_pos': 0.12136586487075801, 'loss/nll_neg': 0.05517547658038883, 'loss/num_matchable': 145.5717521103562, 'loss/num_unmatchable': 360.2203005970764, 'loss/row_norm': 0.9851911551966598}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 07:53:13 gluefactory INFO] [E 15 | it 0] loss {total 1.554E-01, last 3.179E-02, assignment_nll 3.179E-02, nll_pos 4.497E-02, nll_neg 1.861E-02, num_matchable 3.015E+01, num_unmatchable 9.554E+01, confidence 3.899E-02, row_norm 2.459E-01}
[07/19/2024 07:57:17 gluefactory INFO] [E 15 | it 100] loss {total 1.140E-01, last 6.115E-03, assignment_nll 6.115E-03, nll_pos 6.666E-03, nll_neg 5.564E-03, num_matchable 3.636E+01, num_unmatchable 9.022E+01, confidence 3.574E-02, row_norm 2.484E-01}
[07/19/2024 08:01:06 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_15_200_20240719_080106.pkl
[07/19/2024 08:01:06 gluefactory INFO] [E 15 | it 200] loss {total 4.129E-01, last 6.814E-02, assignment_nll 6.814E-02, nll_pos 8.534E-02, nll_neg 5.093E-02, num_matchable 1.197E+02, num_unmatchable 3.216E+02, confidence 1.220E-01, row_norm 8.630E-01}
[07/19/2024 08:04:55 gluefactory INFO] [E 15 | it 300] loss {total 4.680E-01, last 9.659E-02, assignment_nll 9.659E-02, nll_pos 1.382E-01, nll_neg 5.497E-02, num_matchable 1.179E+02, num_unmatchable 3.241E+02, confidence 1.308E-01, row_norm 8.606E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 15, 'iteration': 0, 'total': 0.5086313486099243, 'last': 0.10998239368200302, 'assignment_nll': 0.10998239368200302, 'nll_pos': 0.15391433238983154, 'nll_neg': 0.0660504624247551, 'num_matchable': 108.7109375, 'num_unmatchable': 332.16015625, 'confidence': 0.13704577088356018, 'row_norm': 0.8592292070388794}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 15, 'iteration': 100, 'total': 0.4349021315574646, 'last': 0.035576287657022476, 'assignment_nll': 0.035576287657022476, 'nll_pos': 0.045966167002916336, 'nll_neg': 0.025186408311128616, 'num_matchable': 125.6171875, 'num_unmatchable': 317.015625, 'confidence': 0.12864699959754944, 'row_norm': 0.8673216104507446}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 15, 'iteration': 200, 'total': 0.4128551781177521, 'last': 0.06813547015190125, 'assignment_nll': 0.06813547015190125, 'nll_pos': 0.08533987402915955, 'nll_neg': 0.05093105509877205, 'num_matchable': 119.6796875, 'num_unmatchable': 321.609375, 'confidence': 0.12202440947294235, 'row_norm': 0.8629527688026428}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 15, 'iteration': 300, 'total': 0.467989444732666, 'last': 0.09659280627965927, 'assignment_nll': 0.09659280627965927, 'nll_pos': 0.13821280002593994, 'nll_neg': 0.0549728125333786, 'num_matchable': 117.890625, 'num_unmatchable': 324.0859375, 'confidence': 0.13082429766654968, 'row_norm': 0.8606218099594116}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 08:15:54 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_081554.pkl
[07/19/2024 08:15:54 gluefactory INFO] [Validation] {match_recall 9.790E-01, match_precision 8.756E-01, accuracy 9.777E-01, average_precision 8.611E-01, loss/total 9.071E-02, loss/last 9.071E-02, loss/assignment_nll 9.071E-02, loss/nll_pos 1.206E-01, loss/nll_neg 6.079E-02, loss/num_matchable 1.448E+02, loss/num_unmatchable 3.611E+02, loss/row_norm 9.845E-01}
[07/19/2024 08:15:58 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_15_38776.tar
[07/19/2024 08:15:59 gluefactory INFO] Starting epoch 16
[07/19/2024 08:15:59 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 15, 'iteration': 302, 'match_recall': 0.9790222564793085, 'match_precision': 0.8756479908586996, 'accuracy': 0.9776779573248914, 'average_precision': 0.8611122997596918, 'loss/total': 0.09071156381391064, 'loss/last': 0.09071156381391064, 'loss/assignment_nll': 0.09071156381391064, 'loss/nll_pos': 0.12063222040976605, 'loss/nll_neg': 0.06079090737260453, 'loss/num_matchable': 144.80420012353304, 'loss/num_unmatchable': 361.0538912909203, 'loss/row_norm': 0.9845047211705938}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 08:16:56 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_081656.pkl
[07/19/2024 08:16:56 gluefactory INFO] [Validation] {match_recall 9.790E-01, match_precision 8.756E-01, accuracy 9.777E-01, average_precision 8.611E-01, loss/total 9.071E-02, loss/last 9.071E-02, loss/assignment_nll 9.071E-02, loss/nll_pos 1.206E-01, loss/nll_neg 6.079E-02, loss/num_matchable 1.448E+02, loss/num_unmatchable 3.611E+02, loss/row_norm 9.845E-01}
[07/19/2024 08:17:00 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_15_38776.tar
[07/19/2024 08:17:02 gluefactory INFO] Starting epoch 16
[07/19/2024 08:17:02 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 15, 'iteration': 302, 'match_recall': 0.9790222564793085, 'match_precision': 0.8756479908586996, 'accuracy': 0.9776779573248914, 'average_precision': 0.8611122997596918, 'loss/total': 0.09071156381391064, 'loss/last': 0.09071156381391064, 'loss/assignment_nll': 0.09071156381391064, 'loss/nll_pos': 0.12063222040976605, 'loss/nll_neg': 0.06079090737260453, 'loss/num_matchable': 144.80420012353304, 'loss/num_unmatchable': 361.0538912909203, 'loss/row_norm': 0.9845047211705938}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 08:21:14 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_082114.pkl
[07/19/2024 08:21:14 gluefactory INFO] [Validation] {match_recall 9.790E-01, match_precision 8.756E-01, accuracy 9.777E-01, average_precision 8.611E-01, loss/total 9.069E-02, loss/last 9.069E-02, loss/assignment_nll 9.069E-02, loss/nll_pos 1.206E-01, loss/nll_neg 6.079E-02, loss/num_matchable 1.448E+02, loss/num_unmatchable 3.611E+02, loss/row_norm 9.845E-01}
[07/19/2024 08:21:18 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_15_38776.tar
[07/19/2024 08:21:20 gluefactory INFO] Starting epoch 16
[07/19/2024 08:21:20 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 15, 'iteration': 302, 'match_recall': 0.9790296901124039, 'match_precision': 0.8756473911061364, 'accuracy': 0.9776755683285134, 'average_precision': 0.8611122599987182, 'loss/total': 0.0906851195036331, 'loss/last': 0.0906851195036331, 'loss/assignment_nll': 0.0906851195036331, 'loss/nll_pos': 0.12057912350889768, 'loss/nll_neg': 0.06079111489666166, 'loss/num_matchable': 144.80512662137122, 'loss/num_unmatchable': 361.05306773728637, 'loss/row_norm': 0.9845048420243188}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade usingg: pip install --upgrade albumentations
[07/19/2024 08:23:52 gluefactory INFO] [E 16 | it 0] loss {total 1.857E-01, last 2.878E-02, assignment_nll 2.878E-02, nll_pos 3.065E-02, nll_neg 2.691E-02, num_matchable 4.754E+01, num_unmatchable 1.414E+02, confidence 5.801E-02, row_norm 3.689E-01}
[07/19/2024 08:27:41 gluefactory INFO] [E 16 | it 100] loss {total 2.023E-01, last 2.992E-02, assignment_nll 2.992E-02, nll_pos 4.281E-02, nll_neg 1.703E-02, num_matchable 5.430E+01, num_unmatchable 1.354E+02, confidence 5.651E-02, row_norm 3.699E-01}
[07/19/2024 08:29:40 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_16_40000.tar
[07/19/2024 08:31:31 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_16_200_20240719_083131.pkl
[07/19/2024 08:31:31 gluefactory INFO] [E 16 | it 200] loss {total 1.655E-01, last 1.812E-02, assignment_nll 1.812E-02, nll_pos 1.922E-02, nll_neg 1.702E-02, num_matchable 5.252E+01, num_unmatchable 1.369E+02, confidence 5.077E-02, row_norm 3.715E-01}
[07/19/2024 08:35:19 gluefactory INFO] [E 16 | it 300] loss {total 2.303E-01, last 2.900E-02, assignment_nll 2.900E-02, nll_pos 2.667E-02, nll_neg 3.133E-02, num_matchable 7.059E+01, num_unmatchable 1.821E+02, confidence 7.240E-02, row_norm 4.931E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 16, 'iteration': 0, 'total': 0.24941813945770264, 'last': 0.04023239016532898, 'assignment_nll': 0.04023239016532898, 'nll_pos': 0.04293233901262283, 'nll_neg': 0.03753243759274483, 'num_matchable': 63.2578125, 'num_unmatchable': 188.95703125, 'confidence': 0.07650088518857956, 'row_norm': 0.49162644147872925}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 16, 'iteration': 100, 'total': 0.26189619302749634, 'last': 0.03363392874598503, 'assignment_nll': 0.03363392874598503, 'nll_pos': 0.04587893933057785, 'nll_neg': 0.021388914436101913, 'num_matchable': 71.2421875, 'num_unmatchable': 181.578125, 'confidence': 0.07506972551345825, 'row_norm': 0.4939190149307251}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 16, 'iteration': 200, 'total': 0.22907127439975739, 'last': 0.031154338270425797, 'assignment_nll': 0.031154338270425797, 'nll_pos': 0.03851472586393356, 'nll_neg': 0.02379394695162773, 'num_matchable': 67.1328125, 'num_unmatchable': 185.421875, 'confidence': 0.06703512370586395, 'row_norm': 0.4947540760040283}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 16, 'iteration': 300, 'total': 0.23033611476421356, 'last': 0.029002539813518524, 'assignment_nll': 0.029002539813518524, 'nll_pos': 0.0266729723662138, 'nll_neg': 0.0313321091234684, 'num_matchable': 70.59375, 'num_unmatchable': 182.140625, 'confidence': 0.07240433990955353, 'row_norm': 0.493096262216568}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 08:47:21 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_084721.pkl
[07/19/2024 08:47:21 gluefactory INFO] [Validation] {match_recall 9.797E-01, match_precision 8.778E-01, accuracy 9.782E-01, average_precision 8.635E-01, loss/total 8.730E-02, loss/last 8.730E-02, loss/assignment_nll 8.730E-02, loss/nll_pos 1.193E-01, loss/nll_neg 5.532E-02, loss/num_matchable 1.455E+02, loss/num_unmatchable 3.603E+02, loss/row_norm 9.846E-01}
[07/19/2024 08:47:21 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 08:47:22 gluefactory INFO] New best val: loss/total=0.08730234281501825
[07/19/2024 08:47:26 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_16_41200.tar
[07/19/2024 08:47:27 gluefactory INFO] Starting epoch 17
[07/19/2024 08:47:27 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 16, 'iteration': 302, 'match_recall': 0.9796886111277721, 'match_precision': 0.877844539068107, 'accuracy': 0.97821476513745, 'average_precision': 0.8634746568743427, 'loss/total': 0.08730234281501825, 'loss/last': 0.08730234281501825, 'loss/assignment_nll': 0.08730234281501825, 'loss/nll_pos': 0.11928585357168282, 'loss/nll_neg': 0.055318832415868796, 'loss/num_matchable': 145.53376569899115, 'loss/num_unmatchable': 360.2679637636401, 'loss/row_norm': 0.9846045256200975}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 08:54:18 gluefactory INFO] [E 17 | it 0] loss {total 1.976E-01, last 3.315E-02, assignment_nll 3.315E-02, nll_pos 3.624E-02, nll_neg 3.005E-02, num_matchable 4.741E+01, num_unmatchable 1.414E+02, confidence 5.865E-02, row_norm 3.672E-01}
factory INFO] [Validation] {match_recall 9.797E-01, match_precision 8.778E-01, accuracy 9.782E-01, average_precision 8.635E-01, loss/total 8.730E-02, loss/last 8.730E-02, loss/assignment_nll 8.730E-02, loss/nll_pos 1.193E-01, loss/nll_neg 5.532E-02, loss/num_matchable 1.455E+02, loss/num_unmatchable 3.603E+02, loss/row_norm 9.846E-01}
[07/19/2024 08:53:51 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[07/19/2024 08:53:52 gluefactory INFO] New best val: loss/total=0.0873023428595039
[07/19/2024 08:53:56 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_16_41200.tar
[07/19/2024 08:53:57 gluefactory INFO] Starting epoch 17
[07/19/2024 08:53:57 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 16, 'iteration': 302, 'match_recall': 0.9796886111277721, 'match_precision': 0.877844539068107, 'accuracy': 0.97821476513745, 'average_precision': 0.8634746568743427, 'loss/total': 0.0873023428595039, 'loss/last': 0.0873023428595039, 'loss/assignment_nll': 0.0873023428595039, 'loss/nll_pos': 0.11928585281235868, 'loss/nll_neg': 0.055318832415868796, 'loss/num_matchable': 145.53376569899115, 'loss/num_unmatchable': 360.2679637636401, 'loss/row_norm': 0.9846045256200975}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 08:54:18 gluefactory INFO] [E 17 | it 0] loss {total 5.184E-01, last 9.836E-02, assignment_nll 9.836E-02, nll_pos 1.213E-01, nll_neg 7.540E-02, num_matchable 1.304E+02, num_unmatchable 3.740E+02, confidence 1.517E-01, row_norm 9.815E-01}
[07/19/2024 08:58:06 gluefactory INFO] [E 17 | it 100] loss {total 5.735E-01, last 1.197E-01, assignment_nll 1.197E-01, nll_pos 2.018E-01, nll_neg 3.771E-02, num_matchable 1.416E+02, num_unmatchable 3.633E+02, confidence 1.465E-01, row_norm 9.835E-01}
[07/19/2024 09:02:03 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_17_200_20240719_090203.pkl
[07/19/2024 09:02:03 gluefactory INFO] [E 17 | it 200] loss {total 4.467E-01, last 5.200E-02, assignment_nll 5.200E-02, nll_pos 5.560E-02, nll_neg 4.841E-02, num_matchable 1.440E+02, num_unmatchable 3.603E+02, confidence 1.390E-01, row_norm 9.899E-01}
[07/19/2024 09:05:52 gluefactory INFO] [E 17 | it 300] loss {total 4.755E-01, last 6.924E-02, assignment_nll 6.924E-02, nll_pos 7.328E-02, nll_neg 6.520E-02, num_matchable 1.391E+02, num_unmatchable 3.658E+02, confidence 1.466E-01, row_norm 9.855E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 17, 'iteration': 0, 'total': 0.5183765888214111, 'last': 0.09835910052061081, 'assignment_nll': 0.09835910052061081, 'nll_pos': 0.12131451815366745, 'nll_neg': 0.07540367543697357, 'num_matchable': 130.3515625, 'num_unmatchable': 374.01171875, 'confidence': 0.15169863402843475, 'row_norm': 0.9814704656600952}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 17, 'iteration': 100, 'total': 0.5734535455703735, 'last': 0.11973454058170319, 'assignment_nll': 0.11973454058170319, 'nll_pos': 0.2017599642276764, 'nll_neg': 0.03770909830927849, 'num_matchable': 141.578125, 'num_unmatchable': 363.34375, 'confidence': 0.1464703381061554, 'row_norm': 0.9834812879562378}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 17, 'iteration': 200, 'total': 0.44670283794403076, 'last': 0.05200238898396492, 'assignment_nll': 0.05200238898396492, 'nll_pos': 0.055597662925720215, 'nll_neg': 0.04840711131691933, 'num_matchable': 143.953125, 'num_unmatchable': 360.34375, 'confidence': 0.13899841904640198, 'row_norm': 0.9899464845657349}
debugging in data_to_log, iteraiton: 300
Data to be logged to training DataFrame: {'epoch': 17, 'iteration': 300, 'total': 0.47547540068626404, 'last': 0.06923820078372955, 'assignment_nll': 0.06923820078372955, 'nll_pos': 0.07327544689178467, 'nll_neg': 0.06520095467567444, 'num_matchable': 139.1484375, 'num_unmatchable': 365.77734375, 'confidence': 0.14655102789402008, 'row_norm': 0.9855203628540039}
INFO:alINFOINFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 09:14:13 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_091413.pkl
[07/19/2024 09:14:13 gluefactory INFO] [Validation] {match_recall 9.802E-01, match_precision 8.781E-01, accuracy 9.783E-01, average_precision 8.640E-01, loss/total 8.901E-02, loss/last 8.901E-02, loss/assignment_nll 8.901E-02, loss/nll_pos 1.221E-01, loss/nll_neg 5.593E-02, loss/num_matchable 1.455E+02, loss/num_unmatchable 3.603E+02, loss/row_norm 9.843E-01}
[07/19/2024 09:14:16 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_17_43624.tar
[07/19/2024 09:14:17 gluefactory.utils.experiments INFO] Deleting checkpoint checkpoint_10_26656.tar
[07/19/2024 09:14:17 gluefactory INFO] Starting epoch 18
[07/19/2024 09:14:17 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 17, 'iteration': 302, 'match_recall': 0.9802281888443819, 'match_precision': 0.8780821425840643, 'accuracy': 0.9783438069721738, 'average_precision': 0.864032695144691, 'loss/total': 0.0890108070274731, 'loss/last': 0.0890108070274731, 'loss/assignment_nll': 0.0890108070274731, 'loss/nll_pos': 0.12208673276412915, 'loss/nll_neg': 0.05593488060359034, 'loss/num_matchable': 145.502985381923, 'loss/num_unmatchable': 360.30564134239245, 'loss/row_norm': 0.9843045582073275}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 09:17:01 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_091701.pkl
[07/19/2024 09:17:01 gluefactory INFO] [Validation] {match_recall 9.802E-01, match_precision 8.781E-01, accuracy 9.783E-01, average_precision 8.640E-01, loss/total 8.901E-02, loss/last 8.901E-02, loss/assignment_nll 8.901E-02, loss/nll_pos 1.221E-01, loss/nll_neg 5.594E-02, loss/num_matchable 1.455E+02, loss/num_unmatchable 3.603E+02, loss/row_norm 9.843E-01}
[07/19/2024 09:17:05 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_17_43624.tar
[07/19/2024 09:17:06 gluefactory INFO] Starting epoch 18
[07/19/2024 09:17:06 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 17, 'iteration': 302, 'match_recall': 0.9802241804473607, 'match_precision': 0.8780788688323908, 'accuracy': 0.9783438007871335, 'average_precision': 0.8640324868168241, 'loss/total': 0.08900740218596348, 'loss/last': 0.08900740218596348, 'loss/assignment_nll': 0.08900740218596348, 'loss/nll_pos': 0.12207863331305095, 'loss/nll_neg': 0.055936172454613506, 'loss/num_matchable': 145.50319127033148, 'loss/num_unmatchable': 360.30543545398393, 'loss/row_norm': 0.9843027444319737}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 09:18:01 gluefactory INFO] Intermediate DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_20240719_091801.pkl
[07/19/2024 09:18:01 gluefactory INFO] [Validation] {match_recall 9.802E-01, match_precision 8.781E-01, accuracy 9.783E-01, average_precision 8.640E-01, loss/total 8.901E-02, loss/last 8.901E-02, loss/assignment_nll 8.901E-02, loss/nll_pos 1.221E-01, loss/nll_neg 5.594E-02, loss/num_matchable 1.455E+02, loss/num_unmatchable 3.603E+02, loss/row_norm 9.843E-01}
[07/19/2024 09:18:05 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_17_43624.tar
[07/19/2024 09:18:06 gluefactory INFO] Starting epoch 18
[07/19/2024 09:18:06 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data to be logged: {'epoch': 17, 'iteration': 302, 'match_recall': 0.9802241804473607, 'match_precision': 0.8780788688323908, 'accuracy': 0.9783438007871335, 'average_precision': 0.8640324868168241, 'loss/total': 0.08900740218596348, 'loss/last': 0.08900740218596348, 'loss/assignment_nll': 0.08900740218596348, 'loss/nll_pos': 0.12207863331305095, 'loss/nll_neg': 0.055936172454613506, 'loss/num_matchable': 145.50319127033148, 'loss/num_unmatchable': 360.30543545398393, 'loss/row_norm': 0.9843027444319737}
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.11 (you have 1.4.7). Upgrade using: pip install --upgrade albumentations
[07/19/2024 09:24:55 gluefactory INFO] [E 18 | it 0] loss {total 2.413E-01, last 3.652E-02, assignment_nll 3.652E-02, nll_pos 3.946E-02, nll_neg 3.358E-02, num_matchable 6.391E+01, num_unmatchable 1.886E+02, confidence 7.384E-02, row_norm 4.926E-01}
[07/19/2024 09:29:54 gluefactory INFO] [E 18 | it 100] loss {total 2.552E-01, last 2.305E-02, assignment_nll 2.305E-02, nll_pos 2.436E-02, nll_neg 2.173E-02, num_matchable 7.190E+01, num_unmatchable 1.801E+02, confidence 7.469E-02, row_norm 4.947E-01}
[07/19/2024 09:33:38 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_18_45000.tar
[07/19/2024 09:34:47 gluefactory INFO] Intermediate training DataFrame saved to: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/sp+lg_18-07-tartan-homog-gray-128/sp+lg_densehomography_intermediate_18_200_20240719_093447.pkl
[07/19/2024 09:34:47 gluefactory INFO] [E 18 | it 200] loss {total 6.201E-02, last 9.117E-03, assignment_nll 9.117E-03, nll_pos 1.206E-02, nll_neg 6.175E-03, num_matchable 1.727E+01, num_unmatchable 4.546E+01, confidence 1.737E-02, row_norm 1.237E-01}
debugging in data_to_log, iteraiton: 0
Data to be logged to training DataFrame: {'epoch': 18, 'iteration': 0, 'total': 0.06177719309926033, 'last': 0.009034791961312294, 'assignment_nll': 0.009034791961312294, 'nll_pos': 0.009477515704929829, 'nll_neg': 0.008592067286372185, 'num_matchable': 16.2734375, 'num_unmatchable': 46.76171875, 'confidence': 0.019191492348909378, 'row_norm': 0.12317077815532684}
debugging in data_to_log, iteraiton: 100
Data to be logged to training DataFrame: {'epoch': 18, 'iteration': 100, 'total': 0.05603495612740517, 'last': 0.0036895303055644035, 'assignment_nll': 0.0036895303055644035, 'nll_pos': 0.0032039224170148373, 'nll_neg': 0.00417513819411397, 'num_matchable': 19.125, 'num_unmatchable': 43.828125, 'confidence': 0.01832251250743866, 'row_norm': 0.1240987777709961}
debugging in data_to_log, iteraiton: 200
Data to be logged to training DataFrame: {'epoch': 18, 'iteration': 200, 'total': 0.06201095134019852, 'last': 0.0091172456741333, 'assignment_nll': 0.0091172456741333, 'nll_pos': 0.012059901840984821, 'nll_neg': 0.00617458950728178, 'num_matchable': 17.2734375, 'num_unmatchable': 45.4609375, 'confidence': 0.017367806285619736, 'row_norm': 0.123725526034832}
Traceback (most recent call last):
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/runpy.py", line 86, in _run_code
exec(code, run_globals)
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py", line 1076, in <module>
torch.multiprocessing.spawn(
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 281, in spawn
return start_processes(fn, args, nprocs, join, daemon, start_method="spawn")
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 237, in start_processes
while not context.join():
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 188, in join
raise ProcessRaisedException(msg, error_index, failed_process.pid)
torch.multiprocessing.spawn.ProcessRaisedException:
-- Process 0 terminated with the following error:
Traceback (most recent call last):
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/multiprocessing/spawn.py", line 75, in _wrap
fn(i, *args)
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py", line 992, in main_worker
training(rank, conf, output_dir, args, results_df, train_df)
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py", line 640, in training
pred = model(data)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl
return forward_call(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/parallel/distributed.py", line 1593, in forward
else self._run_ddp_forward(*inputs, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/parallel/distributed.py", line 1411, in _run_ddp_forward
return self.module(*inputs, **kwargs) # type: ignore[index]
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl
return forward_call(*args, **kwargs)
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/base_model.py", line 120, in forward
return self._forward(data)
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/two_view_pipeline.py", line 79, in _forward
pred0 = self.extract_view(data, "0")
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/two_view_pipeline.py", line 69, in extract_view
pred_i = {**pred_i, **self.extractor(data_i)}
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl
return forward_call(*args, **kwargs)
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/base_model.py", line 120, in forward
return self._forward(data)
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/extractors/superpoint_open.py", line 370, in _forward
features = self.backbone(image)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl
return forward_call(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/container.py", line 217, in forward
input = module(input)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl
return forward_call(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/container.py", line 217, in forward
input = module(input)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl
return forward_call(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/container.py", line 217, in forward
input = module(input)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1532, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1541, in _call_impl
return forward_call(*args, **kwargs)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/batchnorm.py", line 776, in forward
return sync_batch_norm.apply(
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/autograd/function.py", line 598, in apply
return super().apply(*args, **kwargs) # type: ignore[misc]
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/GlueFactory/lib/python3.10/site-packages/torch/nn/modules/_functions.py", line 69, in forward
dist.all_gather(combined_list, combined, process_group, async_op=False)