generated_from_trainer

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

segformer-b5-finetuned-segments-crop_crack_early-lr6-8

This model is a fine-tuned version of nvidia/mit-b5 on the None dataset. It achieves the following results on the evaluation set:

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

Training results

Training Loss Epoch Step Validation Loss Mean Iou Mean Accuracy Overall Accuracy Accuracy Unlabeled Accuracy Crack Accuracy Potholes Iou Unlabeled Iou Crack Iou Potholes
0.3187 0.07 20 0.3997 0.0630 0.0962 0.0920 nan 0.0720 0.1203 0.0 0.0707 0.1182
0.1572 0.14 40 0.1967 0.1256 0.1920 0.1806 nan 0.1259 0.2582 0.0 0.1242 0.2524
0.1316 0.21 60 0.1744 0.1255 0.2012 0.2238 nan 0.3322 0.0701 0.0 0.3065 0.0701
0.1051 0.28 80 0.1538 0.1497 0.2309 0.2448 nan 0.3111 0.1507 0.0 0.2985 0.1505
0.2733 0.35 100 0.1587 0.1738 0.2751 0.2985 nan 0.4104 0.1397 0.0 0.3816 0.1397
0.1544 0.42 120 0.1306 0.2941 0.4527 0.4424 nan 0.3930 0.5124 0.0 0.3865 0.4957
0.2418 0.49 140 0.1322 0.2712 0.4150 0.3977 nan 0.3150 0.5150 0.0 0.3115 0.5020
0.1182 0.56 160 0.1189 0.2774 0.4224 0.4091 nan 0.3453 0.4996 0.0 0.3403 0.4920
0.1287 0.63 180 0.1263 0.2728 0.4170 0.4111 nan 0.3828 0.4512 0.0 0.3675 0.4508
0.0895 0.7 200 0.1285 0.2252 0.3426 0.3489 nan 0.3787 0.3065 0.0 0.3738 0.3017
0.1097 0.77 220 0.1254 0.2902 0.4440 0.4396 nan 0.4185 0.4696 0.0 0.4015 0.4691
0.1234 0.84 240 0.1180 0.2718 0.4132 0.4029 nan 0.3535 0.4730 0.0 0.3439 0.4717
0.1774 0.91 260 0.1170 0.3964 0.6114 0.5946 nan 0.5145 0.7084 0.0 0.5040 0.6851
0.0754 0.98 280 0.1170 0.2904 0.4468 0.4031 nan 0.1942 0.6994 0.0 0.1931 0.6781
0.1382 1.05 300 0.1054 0.3532 0.5403 0.5220 nan 0.4344 0.6462 0.0 0.4284 0.6312
0.0812 1.12 320 0.1029 0.3404 0.5203 0.4872 nan 0.3294 0.7111 0.0 0.3273 0.6938
0.0754 1.19 340 0.1113 0.2714 0.4123 0.4066 nan 0.3794 0.4451 0.0 0.3698 0.4443
0.1153 1.26 360 0.1053 0.3478 0.5274 0.5136 nan 0.4478 0.6070 0.0 0.4407 0.6028
0.1036 1.33 380 0.1052 0.3907 0.5977 0.5784 nan 0.4859 0.7095 0.0 0.4810 0.6912
0.1002 1.4 400 0.1050 0.3212 0.4887 0.4820 nan 0.4498 0.5276 0.0 0.4373 0.5261
0.0667 1.47 420 0.1145 0.2930 0.4444 0.4272 nan 0.3453 0.5434 0.0 0.3374 0.5416
0.0789 1.54 440 0.1021 0.3482 0.5335 0.5037 nan 0.3611 0.7060 0.0 0.3584 0.6861
0.1067 1.61 460 0.1052 0.3891 0.6020 0.5718 nan 0.4277 0.7762 0.0 0.4243 0.7430
0.0931 1.68 480 0.1172 0.2708 0.4163 0.4196 nan 0.4354 0.3971 0.0 0.4155 0.3970
0.0963 1.75 500 0.0984 0.3751 0.5701 0.5526 nan 0.4688 0.6713 0.0 0.4637 0.6615
0.0787 1.82 520 0.0973 0.3590 0.5448 0.5282 nan 0.4490 0.6406 0.0 0.4402 0.6367
0.0699 1.89 540 0.1156 0.3087 0.4707 0.4711 nan 0.4731 0.4683 0.0 0.4581 0.4680
0.0949 1.96 560 0.0978 0.3474 0.5313 0.5081 nan 0.3974 0.6653 0.0 0.3936 0.6486
0.1397 2.03 580 0.0969 0.3561 0.5429 0.5201 nan 0.4112 0.6746 0.0 0.4067 0.6616
0.0749 2.1 600 0.0992 0.3904 0.5960 0.5901 nan 0.5620 0.6299 0.0 0.5493 0.6220
0.0916 2.17 620 0.1031 0.4125 0.6294 0.6244 nan 0.6005 0.6583 0.0 0.5889 0.6486
0.1168 2.24 640 0.0958 0.3648 0.5568 0.5369 nan 0.4420 0.6716 0.0 0.4366 0.6577
0.0879 2.31 660 0.0963 0.3617 0.5539 0.5220 nan 0.3698 0.7379 0.0 0.3657 0.7195
0.074 2.38 680 0.1027 0.3534 0.5387 0.5240 nan 0.4541 0.6233 0.0 0.4458 0.6144
0.0448 2.45 700 0.0968 0.3591 0.5437 0.5241 nan 0.4303 0.6571 0.0 0.4233 0.6540
0.1011 2.52 720 0.0925 0.3884 0.5908 0.5692 nan 0.4660 0.7156 0.0 0.4580 0.7071
0.0898 2.59 740 0.0929 0.3687 0.5610 0.5424 nan 0.4539 0.6680 0.0 0.4469 0.6592
0.0645 2.66 760 0.0920 0.4107 0.6235 0.6133 nan 0.5648 0.6821 0.0 0.5560 0.6761
0.0657 2.73 780 0.0905 0.3793 0.5822 0.5561 nan 0.4309 0.7336 0.0 0.4293 0.7087
0.0792 2.8 800 0.1014 0.3526 0.5358 0.5216 nan 0.4539 0.6177 0.0 0.4409 0.6168
0.1359 2.87 820 0.0988 0.3957 0.6123 0.5772 nan 0.4095 0.8152 0.0 0.4075 0.7794
0.1265 2.94 840 0.1329 0.2427 0.3705 0.3669 nan 0.3497 0.3912 0.0 0.3372 0.3911
0.1051 3.01 860 0.0971 0.3935 0.5987 0.5920 nan 0.5603 0.6371 0.0 0.5495 0.6311
0.1066 3.08 880 0.0906 0.3882 0.5888 0.5667 nan 0.4614 0.7161 0.0 0.4576 0.7070
0.1001 3.15 900 0.0940 0.3714 0.5661 0.5318 nan 0.3684 0.7637 0.0 0.3657 0.7485
0.114 3.22 920 0.1004 0.3721 0.5649 0.5571 nan 0.5198 0.6100 0.0 0.5086 0.6078
0.0969 3.29 940 0.0960 0.3933 0.5965 0.5751 nan 0.4728 0.7201 0.0 0.4664 0.7136
0.1018 3.36 960 0.0973 0.3314 0.5007 0.4885 nan 0.4302 0.5712 0.0 0.4255 0.5686
0.1173 3.43 980 0.0890 0.4049 0.6205 0.5990 nan 0.4962 0.7448 0.0 0.4928 0.7219
0.0957 3.5 1000 0.0941 0.3724 0.5672 0.5452 nan 0.4404 0.6939 0.0 0.4346 0.6825
0.0976 3.57 1020 0.0994 0.3447 0.5219 0.5082 nan 0.4430 0.6007 0.0 0.4360 0.5980
0.145 3.64 1040 0.0916 0.3668 0.5572 0.5300 nan 0.4000 0.7144 0.0 0.3951 0.7053
0.0494 3.71 1060 0.0867 0.4152 0.6317 0.6105 nan 0.5093 0.7540 0.0 0.5023 0.7432
0.0851 3.78 1080 0.0873 0.3918 0.5949 0.5719 nan 0.4618 0.7281 0.0 0.4547 0.7208
0.0463 3.85 1100 0.0907 0.3686 0.5600 0.5335 nan 0.4068 0.7131 0.0 0.4009 0.7048
0.0825 3.92 1120 0.0930 0.3694 0.5606 0.5369 nan 0.4237 0.6976 0.0 0.4175 0.6909
0.0622 3.99 1140 0.0934 0.3848 0.5835 0.5638 nan 0.4697 0.6974 0.0 0.4623 0.6919
0.0633 4.06 1160 0.0932 0.4324 0.6573 0.6438 nan 0.5791 0.7356 0.0 0.5700 0.7272
0.0975 4.13 1180 0.0925 0.4221 0.6430 0.6230 nan 0.5277 0.7584 0.0 0.5188 0.7474
0.0961 4.2 1200 0.0899 0.3870 0.5889 0.5640 nan 0.4448 0.7330 0.0 0.4404 0.7206
0.0528 4.27 1220 0.0914 0.4008 0.6107 0.5804 nan 0.4354 0.7860 0.0 0.4282 0.7742
0.0877 4.34 1240 0.0933 0.3777 0.5732 0.5536 nan 0.4599 0.6866 0.0 0.4525 0.6808
0.0579 4.41 1260 0.0906 0.3926 0.5972 0.5783 nan 0.4881 0.7064 0.0 0.4801 0.6977
0.1101 4.48 1280 0.0946 0.4184 0.6364 0.6313 nan 0.6072 0.6655 0.0 0.5986 0.6565
0.0898 4.55 1300 0.0930 0.4115 0.6248 0.6156 nan 0.5715 0.6781 0.0 0.5619 0.6726
0.0636 4.62 1320 0.0920 0.3985 0.6036 0.5852 nan 0.4970 0.7103 0.0 0.4892 0.7062
0.0714 4.69 1340 0.0900 0.3977 0.6032 0.5884 nan 0.5179 0.6885 0.0 0.5093 0.6837
0.0862 4.76 1360 0.0880 0.4039 0.6131 0.5891 nan 0.4749 0.7513 0.0 0.4702 0.7416
0.0801 4.83 1380 0.0878 0.4142 0.6297 0.6114 nan 0.5239 0.7355 0.0 0.5173 0.7254
0.0652 4.9 1400 0.0933 0.3704 0.5614 0.5336 nan 0.4010 0.7217 0.0 0.3966 0.7145
0.0633 4.97 1420 0.0914 0.3790 0.5754 0.5630 nan 0.5039 0.6469 0.0 0.4944 0.6427
0.0905 5.03 1440 0.0920 0.3680 0.5620 0.5331 nan 0.3949 0.7291 0.0 0.3911 0.7127
0.0461 5.1 1460 0.0851 0.4118 0.6274 0.5987 nan 0.4614 0.7934 0.0 0.4560 0.7795
0.0598 5.17 1480 0.0907 0.4047 0.6128 0.5942 nan 0.5054 0.7202 0.0 0.4959 0.7181
0.1264 5.24 1500 0.1093 0.3601 0.5495 0.5494 nan 0.5490 0.5501 0.0 0.5315 0.5487
0.065 5.31 1520 0.0904 0.4181 0.6369 0.6184 nan 0.5299 0.7439 0.0 0.5210 0.7332
0.0711 5.38 1540 0.1021 0.4315 0.6585 0.6563 nan 0.6460 0.6709 0.0 0.6259 0.6685
0.0726 5.45 1560 0.0993 0.3713 0.5609 0.5420 nan 0.4516 0.6703 0.0 0.4445 0.6694
0.1181 5.52 1580 0.0957 0.3985 0.6066 0.5941 nan 0.5346 0.6786 0.0 0.5300 0.6654
0.1049 5.59 1600 0.0982 0.4586 0.6967 0.6912 nan 0.6649 0.7284 0.0 0.6517 0.7242
0.1034 5.66 1620 0.0877 0.4091 0.6254 0.5941 nan 0.4448 0.8060 0.0 0.4409 0.7863
0.0745 5.73 1640 0.0912 0.3913 0.5938 0.5731 nan 0.4745 0.7131 0.0 0.4692 0.7047
0.0665 5.8 1660 0.0861 0.4349 0.6610 0.6365 nan 0.5196 0.8024 0.0 0.5142 0.7906
0.0836 5.87 1680 0.0953 0.3982 0.6031 0.5838 nan 0.4912 0.7151 0.0 0.4818 0.7128
0.0765 5.94 1700 0.0899 0.3915 0.5976 0.5774 nan 0.4811 0.7141 0.0 0.4755 0.6990
0.0456 6.01 1720 0.1060 0.3429 0.5206 0.4982 nan 0.3908 0.6504 0.0 0.3796 0.6493
0.0939 6.08 1740 0.0892 0.4107 0.6285 0.5976 nan 0.4499 0.8072 0.0 0.4459 0.7861
0.0756 6.15 1760 0.0867 0.4091 0.6207 0.5969 nan 0.4831 0.7582 0.0 0.4765 0.7507
0.0435 6.22 1780 0.0898 0.3919 0.5929 0.5660 nan 0.4374 0.7484 0.0 0.4308 0.7449
0.0673 6.29 1800 0.0861 0.4474 0.6795 0.6569 nan 0.5489 0.8100 0.0 0.5424 0.7998
0.0499 6.36 1820 0.0877 0.3879 0.5923 0.5593 nan 0.4014 0.7832 0.0 0.3981 0.7655
0.0532 6.43 1840 0.0889 0.4420 0.6742 0.6438 nan 0.4987 0.8496 0.0 0.4933 0.8328
0.0483 6.5 1860 0.0907 0.4078 0.6178 0.5928 nan 0.4737 0.7619 0.0 0.4677 0.7558
0.0886 6.57 1880 0.0938 0.4683 0.7105 0.6990 nan 0.6441 0.7769 0.0 0.6309 0.7739
0.0603 6.64 1900 0.0883 0.4295 0.6553 0.6282 nan 0.4989 0.8117 0.0 0.4951 0.7935
0.0844 6.71 1920 0.0917 0.4115 0.6241 0.6039 nan 0.5076 0.7406 0.0 0.5017 0.7327
0.0688 6.78 1940 0.0910 0.4314 0.6549 0.6359 nan 0.5451 0.7647 0.0 0.5372 0.7570
0.0863 6.85 1960 0.0886 0.3809 0.5779 0.5517 nan 0.4265 0.7293 0.0 0.4220 0.7208
0.0744 6.92 1980 0.0904 0.3983 0.6038 0.5781 nan 0.4554 0.7521 0.0 0.4483 0.7465
0.0836 6.99 2000 0.0878 0.4239 0.6440 0.6216 nan 0.5149 0.7731 0.0 0.5091 0.7625
0.0677 7.06 2020 0.0892 0.4554 0.6921 0.6757 nan 0.5973 0.7868 0.0 0.5898 0.7763
0.0601 7.13 2040 0.0862 0.4311 0.6546 0.6354 nan 0.5435 0.7657 0.0 0.5382 0.7551
0.0682 7.2 2060 0.1049 0.3975 0.6057 0.5953 nan 0.5456 0.6658 0.0 0.5283 0.6641

Framework versions