forked from cvg/glue-factory
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy path2048-Homog-11-08-distributedTraining-newScenes-32-grayscale.log
4526 lines (3953 loc) · 342 KB
/
2048-Homog-11-08-distributedTraining-newScenes-32-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 Sun 11 Aug 14:47:27 BST 2024
Command running in screen session distributed_homog-newScenes_0_20240811-144727 on ray18
Command running in screen session distributed_homog-newScenes_1_20240811-144727 on ray06
[08/11/2024 14:47:34 gluefactory INFO] Training loader has 1948 batches
[08/11/2024 14:47:34 gluefactory INFO] Validation loader has 2594 batches
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory_nonfree/superpoint.py:426: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
scaler = GradScaler()
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/lightglue.py:646: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
getting last checkpoint pretrain_lightglue
getting best checkpoint pretrain_lightglue
Checkpoint path used is /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/pretrain_lightglue/checkpoint_best.tar
model conf after setting, loaded from checkpoint!!! {'name': 'two_view_pipeline', 'extractor': {'name': 'gluefactory_nonfree.superpoint', 'max_num_keypoints': 2048, 'force_num_keypoints': True, 'detection_threshold': 0.0, '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': True, 'checkpointed': True}}
the rank is 2 and world size is 4
cuda:0
23 initpy data sets path gluefactory.datasets.homographiesTartan
205 tools.py utils the classes are [('HomographyTartanTreeDataset', <class 'gluefactory.datasets.homographiesTartan.HomographyTartanTreeDataset'>)]
homog taratan init image_dir: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/data/syntheticForestData/imageData
shuffling images
Loaded 7792 training, 2594 validation, and 1262 test images.
Loaded training dataset: homographiesTartan
calling get_data_loader on dataset object - dataset.get_data_loader('train', distributed=args.distributed, shuffle=True)
205 tools.py utils the classes are [('TwoViewPipeline', <class 'gluefactory.models.two_view_pipeline.TwoViewPipeline'>)]
205 tools.py utils the classes are [('SuperPoint', <class 'gluefactory_nonfree.superpoint.SuperPoint'>)]
None
ModuleNotFoundError: matchers.lightglue
205 tools.py utils the classes are []
400 lightglue conf flash is True
ModuleNotFoundError: matchers.homography_matcher
205 tools.py utils the classes are [('HomographyMatcher', <class 'gluefactory.models.matchers.homography_matcher.HomographyMatcher'>)]
93 base_model.py HomographyMatcher is trainable
93 base_model.py TwoViewPipeline is trainable
conf.model.name two_view_pipeline
replaced!! the weights using two_view_pipeline
args distibuted
device ids are [device(type='cuda', index=0)]
[rank2]:[W811 14:47:35.044577119 ProcessGroupNCCL.cpp:1168] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator())
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
nts/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/multiprocessing/spawn.py", line 282, in spawn
return start_processes(fn, args, nprocs, join, daemon, start_method="spawn")
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/multiprocessing/spawn.py", line 238, in start_processes
while not context.join():
^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/multiprocessing/spawn.py", line 189, 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/redoGlue/lib/python3.12/site-packages/torch/multiprocessing/spawn.py", line 76, in _wrap
fn(i, *args)
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py", line 2486, 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 1859, in training
model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[device])
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/parallel/distributed.py", line 822, in __init__
_verify_param_shape_across_processes(self.process_group, parameters)
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/distributed/utils.py", line 286, in _verify_param_shape_across_processes
return dist._verify_params_across_processes(process_group, tensors, logger)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
torch.distributed.DistBackendError: [2] is setting up NCCL communicator and retrieving ncclUniqueId from [0] via c10d key-value store by key '0', but store->get('0') got error: Connection reset by peer
Exception raised from recvBytes at /opt/conda/conda-bld/pytorch_1720538439675/work/torch/csrc/distributed/c10d/Utils.hpp:672 (most recent call first):
frame #0: c10::Error::Error(c10::SourceLocation, std::string) + 0x96 (0x741ce81ebf86 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libc10.so)
frame #1: <unknown function> + 0x59976fe (0x741d3029f6fe in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #2: c10d::TCPStore::doWait(c10::ArrayRef<std::string>, std::chrono::duration<long, std::ratio<1l, 1000l> >) + 0x2c7 (0x741d30299f97 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #3: c10d::TCPStore::doGet(std::string const&) + 0x32 (0x741d3029a2c2 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #4: c10d::TCPStore::get(std::string const&) + 0xa1 (0x741d3029b4b1 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #5: c10d::PrefixStore::get(std::string const&) + 0x31 (0x741d3024ff01 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #6: c10d::PrefixStore::get(std::string const&) + 0x31 (0x741d3024ff01 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #7: c10d::PrefixStore::get(std::string const&) + 0x31 (0x741d3024ff01 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #8: c10d::ProcessGroupNCCL::broadcastUniqueNCCLID(ncclUniqueId*, bool, std::string const&, int) + 0xaf (0x741ce942a0df in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cuda.so)
frame #9: c10d::ProcessGroupNCCL::getNCCLComm(std::string const&, c10::Device&, c10d::OpType, int, bool) + 0x114c (0x741ce9435ebc in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cuda.so)
frame #10: c10d::ProcessGroupNCCL::allgather(std::vector<std::vector<at::Tensor, std::allocator<at::Tensor> >, std::allocator<std::vector<at::Tensor, std::allocator<at::Tensor> > > >&, std::vector<at::Tensor, std::allocator<at::Tensor> >&, c10d::AllgatherOptions const&) + 0x8db (0x741ce944672b in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cuda.so)
frame #11: <unknown function> + 0x593c62d (0x741d3024462d in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #12: <unknown function> + 0x5945807 (0x741d3024d807 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #13: <unknown function> + 0x4f676bb (0x741d2f86f6bb in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #14: <unknown function> + 0x4f64f34 (0x741d2f86cf34 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #15: <unknown function> + 0x1766f68 (0x741d2c06ef68 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #16: <unknown function> + 0x594dad3 (0x741d30255ad3 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #17: <unknown function> + 0x5958e5f (0x741d30260e5f in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #18: c10d::verify_params_across_processes(c10::intrusive_ptr<c10d::ProcessGroup, c10::detail::intrusive_target_default_null_type<c10d::ProcessGroup> > const&, std::vector<at::Tensor, std::allocator<at::Tensor> > const&, std::optional<std::weak_ptr<c10d::Logger> > const&) + 0x26d (0x741d302c70bd in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_cpu.so)
frame #19: <unknown function> + 0xd5cd8c (0x741d388a6d8c in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_python.so)
frame #20: <unknown function> + 0x4b10b3 (0x741d37ffb0b3 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/lib/libtorch_python.so)
frame #21: <unknown function> + 0x22d111 (0x58a0fc87b111 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #22: _PyObject_MakeTpCall + 0x2eb (0x58a0fc85ad8b in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #23: <unknown function> + 0x113f20 (0x58a0fc761f20 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #24: _PyObject_FastCallDictTstate + 0x291 (0x58a0fc85d7b1 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #25: <unknown function> + 0x23ac53 (0x58a0fc888c53 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #26: _PyObject_MakeTpCall + 0x2bc (0x58a0fc85ad5c in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #27: <unknown function> + 0x113f20 (0x58a0fc761f20 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #28: PyEval_EvalCode + 0xae (0x58a0fc9156fe in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #29: <unknown function> + 0x2eb7ba (0x58a0fc9397ba in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #30: <unknown function> + 0x2e6a9b (0x58a0fc934a9b in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #31: PyRun_StringFlags + 0x62 (0x58a0fc926862 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #32: PyRun_SimpleStringFlags + 0x3c (0x58a0fc92677c in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #33: Py_RunMain + 0x450 (0x58a0fc945340 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #34: Py_BytesMain + 0x37 (0x58a0fc8fdfa7 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
frame #35: <unknown function> + 0x29d90 (0x741d41229d90 in /lib/x86_64-linux-gnu/libc.so.6)
frame #36: __libc_start_main + 0x80 (0x741d41229e40 in /lib/x86_64-linux-gnu/libc.so.6)
frame #37: <unknown function> + 0x2afe21 (0x58a0fc8fde21 in /homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/bin/python)
. This may indicate a possible application crash on rank 0 or a network set up issue.
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
[08/11/2024 14:47:56 gluefactory INFO] Starting experiment 11-08-fullHomog-newScenes
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 14:48:19 gluefactory.datasets.base_dataset INFO] Creating dataset HomographyTartanTreeDataset
install --upgrade albumentations
[08/11/2024 14:48:23 gluefactory.datasets.base_dataset INFO] Creating dataset HomographyTartanTreeDataset
training/11-08-fullHomog-newScenes/log.txt
[08/11/2024 14:48:24 gluefactory INFO] Will fine-tune from weights of pretrain_lightglue
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py:1724: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
init_cp = torch.load(str(init_cp), map_location="cpu")
[08/11/2024 14:48:24 gluefactory INFO] Training in distributed mode with 4 GPUs
[08/11/2024 14:48:24 gluefactory INFO] Using device cuda:0
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 14:48:27 gluefactory.datasets.base_dataset INFO] Creating dataset HomographyTartanTreeDataset
[08/11/2024 14:49:50 gluefactory INFO] Parameters with scaled learning rate:
{}
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py:1889: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
scaler = GradScaler(enabled=args.mixed_precision is not None)
[08/11/2024 14:49:50 gluefactory INFO] Training with mixed_precision=None
[08/11/2024 14:49:50 gluefactory INFO] Starting training with configuration:
data:
name: homographiesTartan
data_dir: syntheticForestData/imageData/
train_size: 40000
val_size: 5000
batch_size: 32
num_workers: 4
homography:
difficulty: 0.7
max_angle: 45
photometric:
name: lg
model:
name: two_view_pipeline
extractor:
name: gluefactory_nonfree.superpoint
max_num_keypoints: 2048
force_num_keypoints: true
detection_threshold: 0.0
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: true
ch[08/11/2024 14:49:50 gluefactory INFO] Parameters with scaled learning rate:
{}
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py:1889: FutureWarning: `torch.cuda.amp.GradScaler(args...)` is deprecated. Please use `torch.amp.GradScaler('cuda', args...)` instead.
scaler = GradScaler(enabled=args.mixed_precision is not None)
[08/11/2024 14:49:50 gluefactory INFO] Training with mixed_precision=None
[08/11/2024 14:49:50 gluefactory INFO] Starting training with configuration:
data:
name: homographiesTartan
data_dir: syntheticForestData/imageData/
train_size: 40000
val_size: 5000
batch_size: 32
num_workers: 4
homography:
difficulty: 0.7
max_angle: 45
photometric:
name: lg
model:
name: two_view_pipeline
extractor:
name: gluefactory_nonfree.superpoint
max_num_keypoints: 2048
force_num_keypoints: true
detection_threshold: 0.0
nms_radius: 3
trainable: false
ground_truth:
name: matchers.homography_ma = GradScaler(enabled=args.mixed_precision is not None)
[08/11/2024 14:49:52 gluefactory INFO] Training with mixed_precision=None
[08/11/2024 14:49:52 gluefactory INFO] Starting training with configuration:
data:
name: homographiesTartan
data_dir: syntheticForestData/imageData/
train_size: 40000
val_size: 5000
batch_size: 32
num_workers: 4
homography:
difficulty: 0.7
max_angle: 45
photometric:
name: lg
model:
name: two_view_pipeline
extractor:
name: gluefactory_nonfree.superpoint
max_num_keypoints: 2048
force_num_keypoints: true
detection_threshold: 0.0
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: true
checkpointed: true
train:
seed: 42
epochs: 40
optimizer: adam
opt_regexp: null
optimizer_options: {}
lr: 0.0001
lr_schedule:
type: exp
start: 30
exp_div_10: 45
on_epoch: true
factor: 1.0
options: {}
lr_scaling:
- - 100
- - dampingnet.const
eval_every_iter: 500
save_every_iter: 5000
log_every_iter: 250
log_grad_every_iter: null
test_every_epoch: 1
keep_last_checkpoints: 50
load_experiment: pretrain_lightglue
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
[08/11/2024 14:49:52 gluefactory INFO] Starting epoch 0
getting last checkpoint pretrain_lightglue
getting best checkpoint pretrain_lightglue
Checkpoint path used is /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/outputs/training/pretrain_lightglue/checkpoint_best.tar
model conf after setting, loaded from checkpoint!!! {'name': 'two_view_pipeline', 'extractor': {'name': 'gluefactory_nonfree.superpoint', 'max_num_keypoints': 2048, 'force_num_keypoints': True, 'detection_threshold': 0.0, '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': True, 'checkpointed': True}}
the rank is 2 and world size is 4
cuda:0
23 initpy data sets path gluefactory.datasets.homographiesTartan
205 tools.py utils the classes are [('HomographyTartanTreeDataset', <class 'gluefactory.datasets.homographiesTartan.HomographyTartanTreeDataset'>)]
homog taratan init image_dir: /vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/data/syntheticForestData/imageData
shuffling images
Loaded 7792 training, 2594 validation, and 1262 test images.
Loaded training dataset: homographiesTartan
calling get_data_loader on dataset object - dataset.get_data_loader('train', distributed=args.distributed, shuffle=True)
205 tools.py utils the classes are [('TwoViewPipeline', <class 'gluefactory.models.two_view_pipeline.TwoViewPipeline'>)]
205 tools.py utils the classes are [('SuperPoint', <class 'gluefactory_nonfree.superpoint.SuperPoint'>)]
None
ModuleNotFoundError: matchers.lightglue
205 tools.py utils the classes are []
400 lightglue conf flash is True
ModuleNotFoundError: matchers.homography_matcher
205 tools.py utils the classes are [('HomographyMatcher', <class 'gluefactory.models.matchers.homography_matcher.HomographyMatcher'>)]
93 base_model.py HomographyMatcher is trainable
93 base_model.py TwoViewPipeline is trainable
conf.model.name two_view_pipeline
replaced!! the weights using two_view_pipeline
args distibuted
device ids are [device(type='cuda', index=0)]
after mpodel weight sub in
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py:2098: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
with autocast(enabled=args.mixed_precision is not None, dtype=mp_dtype):
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory_nonfree/superpoint.py:647: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
with autocast():
ray22:1995324:1995324 [0] NCCL INFO cudaDriverVersion 12020
ray22:1995324:1995324 [0] NCCL INFO NCCL_SOCKET_IFNAME set by environment to eno1
ray22:1995324:1995324 [0] NCCL INFO Bootstrap : Using eno1:146.169.52.22<0>
ray22:1995324:1995324 [0] NCCL INFO NET/Plugin : dlerror=libnccl-net.so: cannot open shared object file: No such file or directory No plugin found (libnccl-net.so), using internal implementation
ray22:1995324:1996424 [0] NCCL INFO NCCL_SOCKET_IFNAME set by environment to eno1
ray22:1995324:1996424 [0] NCCL INFO NET/IB : No device found.
ray22:1995324:1996424 [0] NCCL INFO NCCL_SOCKET_IFNAME set by environment to eno1
ray22:1995324:1996424 [0] NCCL INFO NET/Socket : Using [0]eno1:146.169.52.22<0>
ray22:1995324:1996424 [0] NCCL INFO Using non-device net plugin version 0
ray22:1995324:1996424 [0] NCCL INFO Using network Socket
ray22:1995324:1996424 [0] NCCL INFO comm 0x61cefbee1eb0 rank 3 nranks 4 cudaDev 0 nvmlDev 0 busId 1000 commId 0xeee99a9961e68c57 - Init START
ray22:1995324:1996424 [0] NCCL INFO comm 0x61cefbee1eb0 rank 3 nRanks 4 nNodes 4 localRanks 1 localRank 0 MNNVL 0
ray22:1995324:1996424 [0] NCCL INFO Trees [0] -1/-1/-1->3->2 [1] 1/-1/-1->3->-1
ray22:1995324:1996424 [0] NCCL INFO P2P Chunksize set to 131072
ray22:1995324:1996424 [0] NCCL INFO Channel 00/0 : 2[0] -> 3[0] [receive] via NET/Socket/0
ray22:1995324:1996424 [0] NCCL INFO Channel 01/0 : 2[0] -> 3[0] [receive] via NET/Socket/0
ray22:1995324:1996424 [0] NCCL INFO Channel 00/0 : 3[0] -> 0[0] [send] via NET/Socket/0
ray22:1995324:1996424 [0] NCCL INFO Channel 01/0 : 3[0] -> 0[0] [send] via NET/Socket/0
ray22:1995324:1996424 [0] NCCL INFO Connected all rings
ray22:1995324:1996424 [0] NCCL INFO Channel 01/0 : 1[0] -> 3[0] [receive] via NET/Socket/0
ray22:1995324:1996424 [0] NCCL INFO Channel 01/0 : 3[0] -> 1[0] [send] via NET/Socket/0
ray22:1995324:1996424 [0] NCCL INFO Channel 00/0 : 3[0] -> 2[0] [send] via NET/Socket/0
ray22:1995324:1996424 [0] NCCL INFO Connected all trees
ray22:1995324:1996424 [0] NCCL INFO threadThresholds 8/8/64 | 32/8/64 | 512 | 512
ray22:1995324:1996424 [0] NCCL INFO 2 coll channels, 0 collnet channels, 0 nvls channels, 2 p2p channels, 2 p2p channels per peer
ray22:1995324:1996424 [0] NCCL INFO comm 0x61cefbee1eb0 rank 3 nranks 4 cudaDev 0 nvmlDev 0 busId 1000 commId 0xeee99a9961e68c57 - Init COMPLETE
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/_dynamo/eval_frame.py:600: 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.
return fn(*args, **kwargs)
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/utils/checkpoint.py:295: Fut between the two variants.
return fn(*args, **kwargs)
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/utils/checkpoint.py:295: FutureWarning: `torch.cpu.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cpu', args...)` instead.
with torch.enable_grad(), device_autocast_ctx, torch.cpu.amp.autocast(**ctx.cpu_autocast_kwargs): # type: ignore[attr-defined]
[08/11/2024 14:50:49 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_0_0_20240811_145049.pkl
[08/11/2024 14:50:49 gluefactory INFO] [E 0 | it 0] loss {total 7.050E-01, last 5.342E-01, assignment_nll 5.342E-01, nll_pos 9.965E-01, nll_neg 7.199E-02, num_matchable 1.550E+02, num_unmatchable 3.246E+02, confidence 7.130E-02, row_norm 1.991E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 0, 'iteration': 0, 'total': 0.7049579620361328, 'last': 0.5342415571212769, 'assignment_nll': 0.5342415571212769, 'nll_pos': 0.9964897632598877, 'nll_neg': 0.07199341803789139, 'num_matchable': 154.96875, 'num_unmatchable': 324.640625, 'confidence': 0.07130053639411926, 'row_norm': 0.19914929568767548}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
876648, 'last': 0.3985535502433777, 'assignment_nll': 0.3985535502433777, 'nll_pos': 0.7280431389808655, 'nll_neg': 0.06906405836343765, 'num_matchable': 157.53125, 'num_unmatchable': 317.0, 'confidence': 0.08137056231498718, 'row_norm': 0.2062034159898758}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory_nonfree/superpoint.py:647: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
with autocast():
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py:2295: 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)
[08/11/2024 15:03:50 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_150350.pkl
[08/11/2024 15:03:50 gluefactory INFO] [Validation] {match_recall 8.309E-01, match_precision 7.675E-01, accuracy 9.109E-01, average_precision 6.871E-01, loss/total 8.304E-01, loss/last 8.304E-01, loss/assignment_nll 8.304E-01, loss/nll_pos 1.352E+00, loss/nll_neg 3.092E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 8.620E-01}
[08/11/2024 15:03:50 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_best.tar
[08/11/2024 15:03:52 gluefactory INFO] New best val: loss/total=0.8304440510806067
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py:2098: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
with autocast(enabled=args.mixed_precision is not None, dtype=mp_dtype):
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory_nonfree/superpoint.py:647: FutureWarning: `torch.cuda.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cuda', args...)` instead.
with autocast():
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/_dynamo/eval_frame.py:600: 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.
return fn(*args, **kwargs)
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/utils/checkpoint.py:295: FutureWarning: `torch.cpu.amp.autocast(args...)` is deprecated. Please use `torch.amp.autocast('cpu', args...)` instead.
with torch.enable_grad(), device_autocast_ctx, torch.cpu.amp.autocast(**ctx.cpu_autocast_kwargs): # type: ignore[attr-defined]
Data logged to validation: {'epoch': 0, 'iteration': 0, 'match_recall': 0.8308602879391143, 'match_precision': 0.7675385015600906, 'accuracy': 0.9109162504891753, 'average_precision': 0.6871112929736088, 'loss/total': 0.8304440510806067, 'loss/last': 0.8304440510806067, 'loss/assignment_nll': 0.8304440510806067, 'loss/nll_pos': 1.351721740977619, 'loss/nll_neg': 0.309166366468509, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.8619861938260387}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 0.8304440510806067
best_eval inf
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 15:30:40 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_153040.pkl
[08/11/2024 15:30:40 gluefactory INFO] [Validation] {match_recall 6.730E-01, match_precision 6.924E-01, accuracy 8.445E-01, average_precision 5.086E-01, loss/total 1.404E+00, loss/last 1.404E+00, loss/assignment_nll 1.404E+00, loss/nll_pos 2.403E+00, loss/nll_neg 4.053E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 7.305E-01}
[08/11/2024 15:30:27 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_0_968.tar
[08/11/2024 15:30:29 gluefactory INFO] Starting epoch 1
[08/11/2024 15:30:29 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 0, 'iteration': 242, 'match_recall': 0.6729549974685646, 'match_precision': 0.692415002271039, 'accuracy': 0.8445438080414865, 'average_precision': 0.5086134102772821, 'loss/total': 1.40427974636222, 'loss/last': 1.40427974636222, 'loss/assignment_nll': 1.40427974636222, 'loss/nll_pos': 2.4032740457846553, 'loss/nll_neg': 0.40528544477986295, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.7304731832620449}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.40427974636222
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 15:31:29 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_1_0_20240811_153129.pkl
[08/11/2024 15:31:29 gluefactory INFO] [E 1 | it 0] loss {total 7.860E-01, last 6.889E-01, assignment_nll 6.889E-01, nll_pos 1.263E+00, nll_neg 1.150E-01, num_matchable 1.582E+02, num_unmatchable 3.164E+02, confidence 6.462E-02, row_norm 1.791E-01}
neg 4.053E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 7.305E-01}
[08/11/2024 15:30:59 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_0_968.tar
[08/11/2024 15:31:01 gluefactory INFO] Starting epoch 1
[08/11/2024 15:31:01 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 0, 'iteration': 242, 'match_recall': 0.6729549974685646, 'match_precision': 0.692415002271039, 'accuracy': 0.8445438080414865, 'average_precision': 0.5086134102772821, 'loss/total': 1.40427974636222, 'loss/last': 1.40427974636222, 'loss/assignment_nll': 1.40427974636222, 'loss/nll_pos': 2.4032740457846553, 'loss/nll_neg': 0.40528544477986295, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.7304731832620449}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.40427974636222
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 15:31:29 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_1_0_20240811_153129.pkl
[08/11/2024 15:31:29 gluefactory INFO] [E 1 | it 0] loss {total 7.023E-01, last 5.700E-01, assignment_nll 5.700E-01, nll_pos 1.019E+00, nll_neg 1.209E-01, num_matchable 1.710E+02, num_unmatchable 3.055E+02, confidence 6.278E-02, row_norm 1.791E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 1, 'iteration': 0, 'total': 0.7023179531097412, 'last': 0.5700079202651978, 'assignment_nll': 0.5700079202651978, 'nll_pos': 1.0191385746002197, 'nll_neg': 0.12087714672088623, 'num_matchable': 170.96875, 'num_unmatchable': 305.546875, 'confidence': 0.06278319656848907, 'row_norm': 0.17907094955444336}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 15:58:30 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_155830.pkl
[08/11/2024 15:58:30 gluefactory INFO] [Validation] {match_recall 6.736E-01, match_precision 6.844E-01, accuracy 8.409E-01, average_precision 5.026E-01, loss/total 1.382E+00, loss/last 1.382E+00, loss/assignment_nll 1.382E+00, loss/nll_pos 2.369E+00, loss/nll_neg 3.946E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 7.308E-01}
[08/11/2024 15:58:34 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_1_1940.tar
[08/11/2024 15:58:36 gluefactory INFO] Starting epoch 2
[08/11/2024 15:58:36 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 1, 'iteration': 242, 'match_recall': 0.6736333122047536, 'match_precision': 0.6843835626460997, 'accuracy': 0.8409169483111285, 'average_precision': 0.5026261796837324, 'loss/total': 1.3817455689541267, 'loss/last': 1.3817455689541267, 'loss/assignment_nll': 1.3817455689541267, 'loss/nll_pos': 2.3689242723637025, 'loss/nll_neg': 0.3945668727825859, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.7307591401988393}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.3817455689541267
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 15:59:08 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_2_0_20240811_155908.pkl
[08/11/2024 15:59:08 gluefactory INFO] [E 2 | it 0] loss {total 6.801E-01, last 5.983E-01, assignment_nll 5.983E-01, nll_pos 1.090E+00, nll_neg 1.068E-01, num_matchable 1.571E+02, num_unmatchable 3.164E+02, confidence 6.797E-02, row_norm 1.749E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 2, 'iteration': 0, 'total': 0.6800676584243774, 'last': 0.5982542037963867, 'assignment_nll': 0.5982542037963867, 'nll_pos': 1.0897454023361206, 'nll_neg': 0.1067628562450409, 'num_matchable': 157.09375, 'num_unmatchable': 316.4375, 'confidence': 0.06796541064977646, 'row_norm': 0.17486032843589783}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 16:25:55 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_162555.pkl
[08/11/2024 16:25:55 gluefactory INFO] [Validation] {match_recall 4.824E-01, match_precision 7.422E-01, accuracy 8.031E-01, average_precision 3.951E-01, loss/total 1.927E+00, loss/last 1.927E+00, loss/assignment_nll 1.927E+00, loss/nll_pos 3.622E+00, loss/nll_neg 2.322E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 7.698E-01}
[08/11/2024 16:25:58 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_2_2912.tar
[08/11/2024 16:25:59 gluefactory INFO] Starting epoch 3
[08/11/2024 16:25:59 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 2, 'iteration': 242, 'match_recall': 0.482441188605271, 'match_precision': 0.7422227251042562, 'accuracy': 0.803105602929844, 'average_precision': 0.3951348732688745, 'loss/total': 1.9273439976648818, 'loss/last': 1.9273439976648818, 'loss/assignment_nll': 1.9273439976648818, 'loss/nll_pos': 3.622446606135313, 'loss/nll_neg': 0.2322413797505378, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.7698341881484367}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.9273439976648818
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 16:26:26 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_3_0_20240811_162626.pkl
[08/11/2024 16:26:26 gluefactory INFO] [E 3 | it 0] loss {total 3.138E+00, last 2.835E+00, assignment_nll 2.835E+00, nll_pos 5.440E+00, nll_neg 2.293E-01, num_matchable 6.487E+02, num_unmatchable 1.257E+03, confidence 3.947E-01, row_norm 7.686E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 3, 'iteration': 0, 'total': 3.137537956237793, 'last': 2.8346574306488037, 'assignment_nll': 2.8346574306488037, 'nll_pos': 5.440009117126465, 'nll_neg': 0.22930589318275452, 'num_matchable': 648.71875, 'num_unmatchable': 1256.953125, 'confidence': 0.3947061002254486, 'row_norm': 0.768608808517456}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 16:53:14 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_165314.pkl
[08/11/2024 16:53:14 gluefactory INFO] [Validation] {match_recall 5.918E-01, match_precision 6.767E-01, accuracy 8.209E-01, average_precision 4.430E-01, loss/total 1.779E+00, loss/last 1.779E+00, loss/assignment_nll 1.779E+00, loss/nll_pos 2.879E+00, loss/nll_neg 6.781E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 6.220E-01}
[08/11/2024 16:53:18 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_3_3884.tar
[08/11/2024 16:53:19 gluefactory INFO] Starting epoch 4
[08/11/2024 16:53:19 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 3, 'iteration': 242, 'match_recall': 0.5917641625738916, 'match_precision': 0.6766829316029662, 'accuracy': 0.8209027551464972, 'average_precision': 0.4429852530739357, 'loss/total': 1.7786290462695367, 'loss/last': 1.7786290462695367, 'loss/assignment_nll': 1.7786290462695367, 'loss/nll_pos': 2.8791489955913865, 'loss/nll_neg': 0.6781090961204829, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.6219951710337019}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.7786290462695367
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 16:53:45 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_4_0_20240811_165345.pkl
[08/11/2024 16:53:45 gluefactory INFO] [E 4 | it 0] loss {total 2.841E+00, last 2.582E+00, assignment_nll 2.582E+00, nll_pos 4.472E+00, nll_neg 6.913E-01, num_matchable 6.531E+02, num_unmatchable 1.243E+03, confidence 2.442E-01, row_norm 6.181E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 4, 'iteration': 0, 'total': 2.8412179946899414, 'last': 2.5819091796875, 'assignment_nll': 2.5819091796875, 'nll_pos': 4.472494602203369, 'nll_neg': 0.6913242340087891, 'num_matchable': 653.0625, 'num_unmatchable': 1243.40625, 'confidence': 0.24423885345458984, 'row_norm': 0.6181074976921082}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 17:20:32 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_172032.pkl
[08/11/2024 17:20:32 gluefactory INFO] [Validation] {match_recall 5.123E-01, match_precision 6.727E-01, accuracy 8.012E-01, average_precision 3.865E-01, loss/total 2.408E+00, loss/last 2.408E+00, loss/assignment_nll 2.408E+00, loss/nll_pos 3.312E+00, loss/nll_neg 1.505E+00, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 3.723E-01}
[08/11/2024 17:20:35 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_4_4856.tar
[08/11/2024 17:20:36 gluefactory INFO] Starting epoch 5
[08/11/2024 17:20:36 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 4, 'iteration': 242, 'match_recall': 0.5122976548964001, 'match_precision': 0.6726861511089293, 'accuracy': 0.8011842052478835, 'average_precision': 0.3865104264126618, 'loss/total': 2.4083744926276167, 'loss/last': 2.4083744926276167, 'loss/assignment_nll': 2.4083744926276167, 'loss/nll_pos': 3.3119657118686274, 'loss/nll_neg': 1.5047832355190445, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.3723331404678989}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 2.4083744926276167
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 17:21:02 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_5_0_20240811_172102.pkl
[08/11/2024 17:21:02 gluefactory INFO] [E 5 | it 0] loss {total 3.285E+00, last 3.155E+00, assignment_nll 3.155E+00, nll_pos 4.834E+00, nll_neg 1.477E+00, num_matchable 6.393E+02, num_unmatchable 1.257E+03, confidence 2.717E-01, row_norm 3.611E-01}
[08/11/2024 17:24:00 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_5_5000.tar
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 5, 'iteration': 0, 'total': 3.285405158996582, 'last': 3.155405044555664, 'assignment_nll': 3.155405044555664, 'nll_pos': 4.833714485168457, 'nll_neg': 1.4770957231521606, 'num_matchable': 639.28125, 'num_unmatchable': 1257.0, 'confidence': 0.27170389890670776, 'row_norm': 0.3610788583755493}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 17:47:51 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_174751.pkl
[08/11/2024 17:47:51 gluefactory INFO] [Validation] {match_recall 5.177E-01, match_precision 6.777E-01, accuracy 8.038E-01, average_precision 3.925E-01, loss/total 2.204E+00, loss/last 2.204E+00, loss/assignment_nll 2.204E+00, loss/nll_pos 3.313E+00, loss/nll_neg 1.094E+00, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 4.616E-01}
[08/11/2024 17:47:54 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_5_5828.tar
[08/11/2024 17:47:56 gluefactory INFO] Starting epoch 6
[08/11/2024 17:47:56 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 5, 'iteration': 242, 'match_recall': 0.5176509756166933, 'match_precision': 0.6777031810447264, 'accuracy': 0.8038038232736433, 'average_precision': 0.3924968448297373, 'loss/total': 2.2037678456067487, 'loss/last': 2.2037678456067487, 'loss/assignment_nll': 2.2037678456067487, 'loss/nll_pos': 3.313408838755917, 'loss/nll_neg': 1.0941268378436428, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.46164000937363325}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 2.2037678456067487
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 17:48:22 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_6_0_20240811_174822.pkl
[08/11/2024 17:48:22 gluefactory INFO] [E 6 | it 0] loss {total 3.159E+00, last 2.955E+00, assignment_nll 2.955E+00, nll_pos 4.805E+00, nll_neg 1.105E+00, num_matchable 6.482E+02, num_unmatchable 1.255E+03, confidence 2.479E-01, row_norm 4.650E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 6, 'iteration': 0, 'total': 3.1586520671844482, 'last': 2.954873561859131, 'assignment_nll': 2.954873561859131, 'nll_pos': 4.805235862731934, 'nll_neg': 1.1045114994049072, 'num_matchable': 648.21875, 'num_unmatchable': 1254.765625, 'confidence': 0.2478969842195511, 'row_norm': 0.4649752974510193}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 18:15:08 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_181508.pkl
[08/11/2024 18:15:08 gluefactory INFO] [Validation] {match_recall 5.223E-01, match_precision 6.915E-01, accuracy 8.081E-01, average_precision 4.021E-01, loss/total 2.028E+00, loss/last 2.028E+00, loss/assignment_nll 2.028E+00, loss/nll_pos 3.275E+00, loss/nll_neg 7.805E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 5.533E-01}
[08/11/2024 18:15:11 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_6_6800.tar
[08/11/2024 18:15:13 gluefactory INFO] Starting epoch 7
[08/11/2024 18:15:13 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 6, 'iteration': 242, 'match_recall': 0.5223093145198793, 'match_precision': 0.691459924878758, 'accuracy': 0.8080559000200552, 'average_precision': 0.40211703249740893, 'loss/total': 2.027754293925595, 'loss/last': 2.027754293925595, 'loss/assignment_nll': 2.027754293925595, 'loss/nll_pos': 3.2750004071314334, 'loss/nll_neg': 0.7805081740102128, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.5533190706002317}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 2.027754293925595
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 18:15:38 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_7_0_20240811_181538.pkl
[08/11/2024 18:15:38 gluefactory INFO] [E 7 | it 0] loss {total 3.052E+00, last 2.808E+00, assignment_nll 2.808E+00, nll_pos 4.853E+00, nll_neg 7.623E-01, num_matchable 6.426E+02, num_unmatchable 1.259E+03, confidence 2.605E-01, row_norm 5.490E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 7, 'iteration': 0, 'total': 3.051649332046509, 'last': 2.8078601360321045, 'assignment_nll': 2.8078601360321045, 'nll_pos': 4.853404998779297, 'nll_neg': 0.7623153924942017, 'num_matchable': 642.59375, 'num_unmatchable': 1259.453125, 'confidence': 0.26048970222473145, 'row_norm': 0.5489634871482849}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 18:42:24 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_184224.pkl
[08/11/2024 18:42:24 gluefactory INFO] [Validation] {match_recall 5.468E-01, match_precision 6.839E-01, accuracy 8.118E-01, average_precision 4.148E-01, loss/total 1.929E+00, loss/last 1.929E+00, loss/assignment_nll 1.929E+00, loss/nll_pos 3.135E+00, loss/nll_neg 7.235E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 5.841E-01}
[08/11/2024 18:42:27 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_7_7772.tar
[08/11/2024 18:42:29 gluefactory INFO] Starting epoch 8
[08/11/2024 18:42:29 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 7, 'iteration': 242, 'match_recall': 0.5468018908590745, 'match_precision': 0.6838796224241176, 'accuracy': 0.811783653594203, 'average_precision': 0.4147706303397968, 'loss/total': 1.9291095623348709, 'loss/last': 1.9291095623348709, 'loss/assignment_nll': 1.9291095623348709, 'loss/nll_pos': 3.134731969928962, 'loss/nll_neg': 0.7234871538216643, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.5840673057409829}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.9291095623348709
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 18:42:55 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_8_0_20240811_184255.pkl
[08/11/2024 18:42:55 gluefactory INFO] [E 8 | it 0] loss {total 3.010E+00, last 2.741E+00, assignment_nll 2.741E+00, nll_pos 4.810E+00, nll_neg 6.728E-01, num_matchable 6.373E+02, num_unmatchable 1.268E+03, confidence 2.103E-01, row_norm 5.920E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 8, 'iteration': 0, 'total': 3.0099689960479736, 'last': 2.7412166595458984, 'assignment_nll': 2.7412166595458984, 'nll_pos': 4.809628009796143, 'nll_neg': 0.6728047728538513, 'num_matchable': 637.3125, 'num_unmatchable': 1268.375, 'confidence': 0.2103021889925003, 'row_norm': 0.5920091867446899}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 19:09:41 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_190941.pkl
[08/11/2024 19:09:41 gluefactory INFO] [Validation] {match_recall 5.574E-01, match_precision 6.818E-01, accuracy 8.144E-01, average_precision 4.211E-01, loss/total 1.927E+00, loss/last 1.927E+00, loss/assignment_nll 1.927E+00, loss/nll_pos 3.139E+00, loss/nll_neg 7.146E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 6.029E-01}
[08/11/2024 19:09:44 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_8_8744.tar
[08/11/2024 19:09:46 gluefactory INFO] Starting epoch 9
[08/11/2024 19:09:46 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 8, 'iteration': 242, 'match_recall': 0.5573808733279096, 'match_precision': 0.681840143161456, 'accuracy': 0.8144385245274652, 'average_precision': 0.4211431207708331, 'loss/total': 1.9266676528690958, 'loss/last': 1.9266676528690958, 'loss/assignment_nll': 1.9266676528690958, 'loss/nll_pos': 3.138705268306923, 'loss/nll_neg': 0.7146300189570454, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.6028603087412732}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.9266676528690958
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 19:10:12 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_9_0_20240811_191012.pkl
[08/11/2024 19:10:12 gluefactory INFO] [E 9 | it 0] loss {total 2.978E+00, last 2.720E+00, assignment_nll 2.720E+00, nll_pos 4.721E+00, nll_neg 7.202E-01, num_matchable 6.553E+02, num_unmatchable 1.244E+03, confidence 2.282E-01, row_norm 5.951E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 9, 'iteration': 0, 'total': 2.977609634399414, 'last': 2.7204647064208984, 'assignment_nll': 2.7204647064208984, 'nll_pos': 4.720734119415283, 'nll_neg': 0.720195472240448, 'num_matchable': 655.28125, 'num_unmatchable': 1244.34375, 'confidence': 0.22823242843151093, 'row_norm': 0.5950886607170105}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 19:36:58 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_193658.pkl
[08/11/2024 19:36:58 gluefactory INFO] [Validation] {match_recall 6.030E-01, match_precision 6.602E-01, accuracy 8.191E-01, average_precision 4.413E-01, loss/total 1.889E+00, loss/last 1.889E+00, loss/assignment_nll 1.889E+00, loss/nll_pos 2.679E+00, loss/nll_neg 1.099E+00, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 5.050E-01}
[08/11/2024 19:37:01 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_9_9716.tar
[08/11/2024 19:37:03 gluefactory INFO] Starting epoch 10
[08/11/2024 19:37:03 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 9, 'iteration': 242, 'match_recall': 0.6030274196689461, 'match_precision': 0.6602322403154837, 'accuracy': 0.8190635866446043, 'average_precision': 0.44127454209437994, 'loss/total': 1.8892739333827033, 'loss/last': 1.8892739333827033, 'loss/assignment_nll': 1.8892739333827033, 'loss/nll_pos': 2.6794716219949835, 'loss/nll_neg': 1.0990762427483693, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.5050227826801555}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.8892739333827033
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 19:37:29 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_10_0_20240811_193729.pkl
[08/11/2024 19:37:29 gluefactory INFO] [E 10 | it 0] loss {total 3.009E+00, last 2.796E+00, assignment_nll 2.796E+00, nll_pos 4.502E+00, nll_neg 1.090E+00, num_matchable 6.240E+02, num_unmatchable 1.280E+03, confidence 2.331E-01, row_norm 4.909E-01}
[08/11/2024 19:43:28 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_10_10000.tar
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 10, 'iteration': 0, 'total': 3.0094141960144043, 'last': 2.7962589263916016, 'assignment_nll': 2.7962589263916016, 'nll_pos': 4.502206802368164, 'nll_neg': 1.0903114080429077, 'num_matchable': 624.0, 'num_unmatchable': 1280.15625, 'confidence': 0.23306190967559814, 'row_norm': 0.4909394383430481}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 20:04:17 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_200417.pkl
[08/11/2024 20:04:17 gluefactory INFO] [Validation] {match_recall 5.848E-01, match_precision 6.680E-01, accuracy 8.175E-01, average_precision 4.333E-01, loss/total 1.919E+00, loss/last 1.919E+00, loss/assignment_nll 1.919E+00, loss/nll_pos 2.812E+00, loss/nll_neg 1.026E+00, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 5.162E-01}
[08/11/2024 20:04:20 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_10_10688.tar
[08/11/2024 20:04:22 gluefactory INFO] Starting epoch 11
[08/11/2024 20:04:22 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 10, 'iteration': 242, 'match_recall': 0.5848016487973051, 'match_precision': 0.668015190336276, 'accuracy': 0.8174944984609565, 'average_precision': 0.4332785748847191, 'loss/total': 1.919047249494007, 'loss/last': 1.919047249494007, 'loss/assignment_nll': 1.919047249494007, 'loss/nll_pos': 2.812128987418934, 'loss/nll_neg': 1.0259654942437513, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.516213732217952}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.919047249494007
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 20:04:49 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_11_0_20240811_200449.pkl
[08/11/2024 20:04:49 gluefactory INFO] [E 11 | it 0] loss {total 3.006E+00, last 2.757E+00, assignment_nll 2.757E+00, nll_pos 4.482E+00, nll_neg 1.033E+00, num_matchable 6.429E+02, num_unmatchable 1.260E+03, confidence 2.067E-01, row_norm 5.175E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 11, 'iteration': 0, 'total': 3.0056416988372803, 'last': 2.7573800086975098, 'assignment_nll': 2.7573800086975098, 'nll_pos': 4.481653213500977, 'nll_neg': 1.033106803894043, 'num_matchable': 642.9375, 'num_unmatchable': 1259.5, 'confidence': 0.20673595368862152, 'row_norm': 0.517504870891571}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 20:31:34 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_203134.pkl
[08/11/2024 20:31:34 gluefactory INFO] [Validation] {match_recall 5.484E-01, match_precision 6.917E-01, accuracy 8.148E-01, average_precision 4.203E-01, loss/total 1.877E+00, loss/last 1.877E+00, loss/assignment_nll 1.877E+00, loss/nll_pos 3.183E+00, loss/nll_neg 5.714E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 6.472E-01}
[08/11/2024 20:31:37 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_11_11660.tar
[08/11/2024 20:31:39 gluefactory INFO] Starting epoch 12
[08/11/2024 20:31:39 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 11, 'iteration': 242, 'match_recall': 0.5484137541529024, 'match_precision': 0.6917001681780025, 'accuracy': 0.8148485914780346, 'average_precision': 0.4203421145149075, 'loss/total': 1.8769562010959928, 'loss/last': 1.8769562010959928, 'loss/assignment_nll': 1.8769562010959928, 'loss/nll_pos': 3.182557123297439, 'loss/nll_neg': 0.5713552657971496, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.6472467787466145}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.8769562010959928
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 20:32:05 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_12_0_20240811_203205.pkl
[08/11/2024 20:32:05 gluefactory INFO] [E 12 | it 0] loss {total 2.986E+00, last 2.763E+00, assignment_nll 2.763E+00, nll_pos 4.972E+00, nll_neg 5.542E-01, num_matchable 6.482E+02, num_unmatchable 1.254E+03, confidence 2.183E-01, row_norm 6.397E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 12, 'iteration': 0, 'total': 2.9862289428710938, 'last': 2.7632009983062744, 'assignment_nll': 2.7632009983062744, 'nll_pos': 4.972247123718262, 'nll_neg': 0.5541548132896423, 'num_matchable': 648.25, 'num_unmatchable': 1254.09375, 'confidence': 0.2182750403881073, 'row_norm': 0.6397279500961304}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 20:58:51 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_205851.pkl
[08/11/2024 20:58:51 gluefactory INFO] [Validation] {match_recall 5.698E-01, match_precision 6.816E-01, accuracy 8.180E-01, average_precision 4.309E-01, loss/total 1.838E+00, loss/last 1.838E+00, loss/assignment_nll 1.838E+00, loss/nll_pos 2.994E+00, loss/nll_neg 6.823E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 6.084E-01}
[08/11/2024 20:58:54 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_12_12632.tar
[08/11/2024 20:58:55 gluefactory INFO] Starting epoch 13
[08/11/2024 20:58:55 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 12, 'iteration': 242, 'match_recall': 0.5698440130986704, 'match_precision': 0.6815818423202797, 'accuracy': 0.8180440344439163, 'average_precision': 0.43093461370412994, 'loss/total': 1.838014819296673, 'loss/last': 1.838014819296673, 'loss/assignment_nll': 1.838014819296673, 'loss/nll_pos': 2.99373458400541, 'loss/nll_neg': 0.6822950640548259, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.6083923716129664}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.838014819296673
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 20:59:22 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_13_0_20240811_205921.pkl
[08/11/2024 20:59:22 gluefactory INFO] [E 13 | it 0] loss {total 3.043E+00, last 2.755E+00, assignment_nll 2.755E+00, nll_pos 4.840E+00, nll_neg 6.709E-01, num_matchable 6.321E+02, num_unmatchable 1.271E+03, confidence 2.243E-01, row_norm 6.079E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 13, 'iteration': 0, 'total': 3.04274845123291, 'last': 2.75531005859375, 'assignment_nll': 2.75531005859375, 'nll_pos': 4.839730262756348, 'nll_neg': 0.6708897352218628, 'num_matchable': 632.0625, 'num_unmatchable': 1270.875, 'confidence': 0.22434628009796143, 'row_norm': 0.6078522205352783}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 21:26:08 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_212608.pkl
[08/11/2024 21:26:08 gluefactory INFO] [Validation] {match_recall 5.661E-01, match_precision 6.852E-01, accuracy 8.176E-01, average_precision 4.296E-01, loss/total 1.830E+00, loss/last 1.830E+00, loss/assignment_nll 1.830E+00, loss/nll_pos 3.096E+00, loss/nll_neg 5.644E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 6.527E-01}
[08/11/2024 21:26:11 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_13_13604.tar
[08/11/2024 21:26:13 gluefactory INFO] Starting epoch 14
[08/11/2024 21:26:13 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 13, 'iteration': 242, 'match_recall': 0.5660545758862815, 'match_precision': 0.6852277393238123, 'accuracy': 0.8175550596293065, 'average_precision': 0.4296104955783512, 'loss/total': 1.8301748888439275, 'loss/last': 1.8301748888439275, 'loss/assignment_nll': 1.8301748888439275, 'loss/nll_pos': 3.095911073794986, 'loss/nll_neg': 0.5644387100049874, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.6527019129777377}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.8301748888439275
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 21:26:39 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_14_0_20240811_212639.pkl
[08/11/2024 21:26:39 gluefactory INFO] [E 14 | it 0] loss {total 3.010E+00, last 2.764E+00, assignment_nll 2.764E+00, nll_pos 4.992E+00, nll_neg 5.370E-01, num_matchable 6.368E+02, num_unmatchable 1.264E+03, confidence 1.852E-01, row_norm 6.640E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 14, 'iteration': 0, 'total': 3.0104494094848633, 'last': 2.7643980979919434, 'assignment_nll': 2.7643980979919434, 'nll_pos': 4.991819381713867, 'nll_neg': 0.5369763374328613, 'num_matchable': 636.8125, 'num_unmatchable': 1264.015625, 'confidence': 0.18516969680786133, 'row_norm': 0.6639700531959534}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 21:53:23 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_215323.pkl
[08/11/2024 21:53:23 gluefactory INFO] [Validation] {match_recall 6.184E-01, match_precision 6.502E-01, accuracy 8.187E-01, average_precision 4.453E-01, loss/total 1.837E+00, loss/last 1.837E+00, loss/assignment_nll 1.837E+00, loss/nll_pos 2.570E+00, loss/nll_neg 1.104E+00, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 5.378E-01}
[08/11/2024 21:53:27 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_14_14576.tar
[08/11/2024 21:53:28 gluefactory INFO] Starting epoch 15
[08/11/2024 21:53:28 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 14, 'iteration': 242, 'match_recall': 0.6184103965943101, 'match_precision': 0.650172815657434, 'accuracy': 0.8186817461008646, 'average_precision': 0.4452680399211995, 'loss/total': 1.8368443418854277, 'loss/last': 1.8368443418854277, 'loss/assignment_nll': 1.8368443418854277, 'loss/nll_pos': 2.569672258431487, 'loss/nll_neg': 1.104016403923976, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.5378319451509298}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.8368443418854277
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 21:53:55 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_15_0_20240811_215355.pkl
[08/11/2024 21:53:55 gluefactory INFO] [E 15 | it 0] loss {total 2.840E+00, last 2.601E+00, assignment_nll 2.601E+00, nll_pos 4.129E+00, nll_neg 1.073E+00, num_matchable 6.447E+02, num_unmatchable 1.256E+03, confidence 1.968E-01, row_norm 5.214E-01}
[08/11/2024 22:02:53 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_15_15000.tar
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 15, 'iteration': 0, 'total': 2.839789867401123, 'last': 2.6012961864471436, 'assignment_nll': 2.6012961864471436, 'nll_pos': 4.129098892211914, 'nll_neg': 1.073493480682373, 'num_matchable': 644.71875, 'num_unmatchable': 1256.125, 'confidence': 0.19677306711673737, 'row_norm': 0.5214076042175293}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 22:20:38 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_222038.pkl
[08/11/2024 22:20:38 gluefactory INFO] [Validation] {match_recall 6.022E-01, match_precision 6.628E-01, accuracy 8.202E-01, average_precision 4.424E-01, loss/total 1.839E+00, loss/last 1.839E+00, loss/assignment_nll 1.839E+00, loss/nll_pos 2.741E+00, loss/nll_neg 9.374E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 5.524E-01}
[08/11/2024 22:20:41 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_15_15548.tar
[08/11/2024 22:20:43 gluefactory INFO] Starting epoch 16
[08/11/2024 22:20:43 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 15, 'iteration': 242, 'match_recall': 0.602188015223101, 'match_precision': 0.6628373125101293, 'accuracy': 0.820192689149308, 'average_precision': 0.4423708698339617, 'loss/total': 1.8392750131412938, 'loss/last': 1.8392750131412938, 'loss/assignment_nll': 1.8392750131412938, 'loss/nll_pos': 2.7411731489824165, 'loss/nll_neg': 0.937376880057518, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.5523987803536373}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.8392750131412938
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[08/11/2024 22:21:10 gluefactory INFO] \Training intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/train_11-08-fullHomog-newScenes_intermediate_16_0_20240811_222110.pkl
[08/11/2024 22:21:10 gluefactory INFO] [E 16 | it 0] loss {total 7.146E-01, last 6.614E-01, assignment_nll 6.614E-01, nll_pos 1.079E+00, nll_neg 2.433E-01, num_matchable 1.700E+02, num_unmatchable 3.039E+02, confidence 5.065E-02, row_norm 1.344E-01}
Current iteraiton: 0
Data logged to training DataFrame: {'epoch': 16, 'iteration': 0, 'total': 0.7145894169807434, 'last': 0.6613680124282837, 'assignment_nll': 0.6613680124282837, 'nll_pos': 1.0793884992599487, 'nll_neg': 0.24334745109081268, 'num_matchable': 170.0, 'num_unmatchable': 303.9375, 'confidence': 0.05064709484577179, 'row_norm': 0.1343717873096466}
line 813in train.py doing validation
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
[rank3]:[W811 22:42:15.484063462 ProcessGroupNCCL.cpp:1168] Warning: WARNING: process group has NOT been destroyed before we destruct ProcessGroupNCCL. On normal program exit, the application should call destroy_process_group to ensure that any pending NCCL operations have finished in this process. In rare cases this process can exit before this point and block the progress of another member of the process group. This constraint has always been present, but this warning has only been added since PyTorch 2.4 (function operator())
ray22:1995324:1996526 [0] NCCL INFO [Service thread] Connection closed by localRank 0
ray22:1995324:2358496 [0] NCCL INFO comm 0x61cefbee1eb0 rank 3 nranks 4 cudaDev 0 busId 1000 - Abort COMPLETE
Traceback (most recent call last):
File "<frozen runpy>", line 198, in _run_module_as_main
File "<frozen runpy>", line 88, in _run_code
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py", line 2577, in <module>
torch.multiprocessing.spawn(
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/multiprocessing/spawn.py", line 282, in spawn
return start_processes(fn, args, nprocs, join, daemon, start_method="spawn")
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/multiprocessing/spawn.py", line 238, in start_processes
while not context.join():
^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/multiprocessing/spawn.py", line 189, 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/redoGlue/lib/python3.12/site-packages/torch/multiprocessing/spawn.py", line 76, in _wrap
fn(i, *args)
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py", line 2486, 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 2263, in training
results, pr_metrics, figures = do_evaluation(
^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/utils/_contextlib.py", line 116, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/train.py", line 1551, in do_evaluation
pred = model(data)
^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/parallel/distributed.py", line 1636, in forward
else self._run_ddp_forward(*inputs, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/parallel/distributed.py", line 1454, in _run_ddp_forward
return self.module(*inputs, **kwargs) # type: ignore[index]
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/base_model.py", line 129, in forward
return self._forward(data)
^^^^^^^^^^^^^^^^^^^
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/two_view_pipeline.py", line 102, in _forward
pred0 = self.extract_view(data, "0")
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/two_view_pipeline.py", line 83, in extract_view
pred_i = {**pred_i, **self.extractor(data_i)}
^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/base_model.py", line 129, in forward
return self._forward(data)
^^^^^^^^^^^^^^^^^^^
File "/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory_nonfree/superpoint.py", line 748, in _forward
x = self.relu(self.conv1b(x))
^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1553, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/modules/module.py", line 1562, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/modules/conv.py", line 458, in forward
return self._conv_forward(input, self.weight, self.bias)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/torch/nn/modules/conv.py", line 454, in _conv_forward
return F.conv2d(input, weight, bias, self.stride,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
torch.OutOfMemoryError: CUDA out of memory. Tried to allocate 300.00 MiB. GPU 0 has a total capacity of 7.92 GiB of which 73.50 MiB is free. Including non-PyTorch memory, this process has 862.00 MiB memory in use. Process 2357904 has 6.92 GiB memory in use. Of the allocated memory 613.86 MiB is allocated by PyTorch, and 54.14 MiB is reserved by PyTorch but unallocated. If reserved but unallocated memory is large try setting PYTORCH_CUDA_ALLOC_CONF=expandable_segments:True to avoid fragmentation. See documentation for Memory Management (https://pytorch.org/docs/stable/notes/cuda.html#environment-variables)
[08/11/2024 22:47:50 gluefactory INFO] VALIDATION Intermediate DataFrame saved to: /homes/tp4618/Documents/bitbucket/SuperGlueThesis/external/glue-factory/outputs/training/11-08-fullHomog-newScenes/validation_11-08-fullHomog-newScenes_20240811_224750.pkl
[08/11/2024 22:47:50 gluefactory INFO] [Validation] {match_recall 5.865E-01, match_precision 6.699E-01, accuracy 8.188E-01, average_precision 4.360E-01, loss/total 1.822E+00, loss/last 1.822E+00, loss/assignment_nll 1.822E+00, loss/nll_pos 2.874E+00, loss/nll_neg 7.702E-01, loss/num_matchable 7.042E+02, loss/num_unmatchable 1.194E+03, loss/row_norm 5.975E-01}
[08/11/2024 22:47:54 gluefactory.utils.experiments INFO] Saving checkpoint checkpoint_16_16520.tar
[08/11/2024 22:47:55 gluefactory INFO] Starting epoch 17
[08/11/2024 22:47:55 gluefactory INFO] lr changed from 0.0001 to 0.0001
Data logged to validation: {'epoch': 16, 'iteration': 242, 'match_recall': 0.5865109902027486, 'match_precision': 0.6698667999865371, 'accuracy': 0.8187862771607034, 'average_precision': 0.43595921457870795, 'loss/total': 1.821845043413255, 'loss/last': 1.821845043413255, 'loss/assignment_nll': 1.821845043413255, 'loss/nll_pos': 2.873509849689516, 'loss/nll_neg': 0.7701802341958239, 'loss/num_matchable': 704.1873554356207, 'loss/num_unmatchable': 1193.9313801079413, 'loss/row_norm': 0.5975280397657072}
comparing to see if it hte best experiment 875 train.py
key used was loss/total
FROM RESULTS [conf.train.best_key] 1.821845043413255
best_eval 0.8304440510806067
591 train.py, conf.train.lr_schedule.on_epoch is: True
/homes/tp4618/Documents/bitbucket/miniconda3/envs/redoGlue/lib/python3.12/site-packages/kornia/feature/lightglue.py:44: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
/vol/bitbucket/tp4618/SuperGlueThesis/external/glue-factory/gluefactory/models/matchers/loadedLightGlue.py:25: FutureWarning: `torch.cuda.amp.custom_fwd(args...)` is deprecated. Please use `torch.amp.custom_fwd(args..., device_type='cuda')` instead.
@torch.cuda.amp.custom_fwd(cast_inputs=torch.float32)
INFO:albumentations.check_version:A new version of Albumentations is available: 1.4.13 (you have 1.4.10). Upgrade using: pip install --upgrade albumentations
ray06:1273847:1275036 [0] misc/socket.cc:50 NCCL WARN socketProgress: Connection closed by remote peer ray18.doc.ic.ac.uk<45220>
ray06:1273847:1275036 [0] NCCL INFO misc/socket.cc:752 -> 6
ray06:1273847:1275036 [0] NCCL INFO transport/net_socket.cc:474 -> 6
ray06:1273847:1275036 [0] NCCL INFO transport/net.cc:1298 -> 6
ray06:1273847:1275036 [0] NCCL INFO proxy.cc:694 -> 6
ray06:1273847:1275036 [0] NCCL INFO proxy.cc:874 -> 6 [Progress Thread]
ray06:1273847:1275036 [0] misc/socket.cc:50 NCCL WARN socketProgress: Connection closed by remote peer ray18.doc.ic.ac.uk<45220>
ray06:1273847:1275036 [0] NCCL INFO misc/socket.cc:752 -> 6
ray06:1273847:1275036 [0] NCCL INFO transport/net_socket.cc:474 -> 6
ray06:1273847:1275036 [0] NCCL INFO transport/net.cc:1298 -> 6