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-lr8-350-8

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
0.4007 0.07 20 0.3652 0.0624 0.0949 0.0850 nan 0.0374 0.1524 0.0 0.0372 0.1501
0.1665 0.14 40 0.2121 0.1691 0.2604 0.2421 nan 0.1550 0.3657 0.0 0.1517 0.3555
0.1145 0.21 60 0.1806 0.1182 0.1856 0.2049 nan 0.2972 0.0740 0.0 0.2807 0.0739
0.0835 0.28 80 0.1521 0.2059 0.3151 0.3161 nan 0.3207 0.3095 0.0 0.3104 0.3073
0.2789 0.35 100 0.1636 0.1249 0.1951 0.2099 nan 0.2803 0.1100 0.0 0.2648 0.1099
0.1435 0.42 120 0.1490 0.2611 0.4122 0.4119 nan 0.4104 0.4139 0.0 0.4062 0.3772
0.2562 0.49 140 0.1370 0.2293 0.3530 0.3499 nan 0.3350 0.3711 0.0 0.3265 0.3613
0.1211 0.56 160 0.1310 0.2644 0.4043 0.3979 nan 0.3674 0.4411 0.0 0.3551 0.4381
0.1379 0.63 180 0.1331 0.3149 0.4811 0.4793 nan 0.4704 0.4918 0.0 0.4542 0.4907
0.1017 0.7 200 0.1477 0.1476 0.2284 0.2249 nan 0.2079 0.2489 0.0 0.2062 0.2367
0.1201 0.77 220 0.1340 0.3514 0.5395 0.5410 nan 0.5482 0.5307 0.0 0.5278 0.5262
0.1098 0.84 240 0.1263 0.3347 0.5155 0.4942 nan 0.3926 0.6384 0.0 0.3861 0.6181
0.1992 0.91 260 0.1235 0.3532 0.5454 0.5258 nan 0.4320 0.6588 0.0 0.4261 0.6334
0.0827 0.98 280 0.1295 0.2970 0.4775 0.4428 nan 0.2768 0.6782 0.0 0.2762 0.6147
0.1425 1.05 300 0.1229 0.3292 0.5049 0.5057 nan 0.5095 0.5003 0.0 0.4962 0.4915
0.0923 1.12 320 0.1092 0.3527 0.5398 0.5058 nan 0.3433 0.7363 0.0 0.3395 0.7186
0.0869 1.19 340 0.1298 0.2323 0.3519 0.3558 nan 0.3746 0.3292 0.0 0.3695 0.3275
0.1178 1.26 360 0.1221 0.2744 0.4174 0.4113 nan 0.3819 0.4530 0.0 0.3727 0.4506
0.1186 1.33 380 0.1162 0.2850 0.4392 0.4270 nan 0.3688 0.5096 0.0 0.3651 0.4899
0.1096 1.4 400 0.1152 0.3531 0.5435 0.5225 nan 0.4224 0.6645 0.0 0.4160 0.6435
0.0721 1.47 420 0.1093 0.3263 0.4962 0.4784 nan 0.3936 0.5987 0.0 0.3852 0.5937
0.0771 1.54 440 0.1220 0.2754 0.4189 0.4082 nan 0.3571 0.4808 0.0 0.3485 0.4779
0.1211 1.61 460 0.1183 0.3253 0.5003 0.4915 nan 0.4495 0.5511 0.0 0.4350 0.5411
0.1116 1.68 480 0.1151 0.2657 0.4045 0.3948 nan 0.3487 0.4602 0.0 0.3376 0.4595
0.0958 1.75 500 0.1086 0.3547 0.5430 0.5300 nan 0.4682 0.6177 0.0 0.4621 0.6021
0.0861 1.82 520 0.1115 0.3653 0.5563 0.5405 nan 0.4651 0.6476 0.0 0.4543 0.6415
0.0792 1.89 540 0.1236 0.3061 0.4694 0.4720 nan 0.4843 0.4545 0.0 0.4668 0.4516
0.1062 1.96 560 0.1381 0.2473 0.3812 0.3945 nan 0.4581 0.3042 0.0 0.4380 0.3039
0.1673 2.03 580 0.1208 0.2577 0.4016 0.3795 nan 0.2737 0.5296 0.0 0.2708 0.5024
0.0792 2.1 600 0.1131 0.3264 0.4979 0.4946 nan 0.4790 0.5168 0.0 0.4685 0.5108
0.0746 2.17 620 0.1110 0.3276 0.4999 0.5010 nan 0.5063 0.4935 0.0 0.4986 0.4844
0.1398 2.24 640 0.1144 0.2616 0.3995 0.3822 nan 0.2998 0.4992 0.0 0.2970 0.4878
0.0998 2.31 660 0.1018 0.3241 0.4946 0.4600 nan 0.2951 0.6940 0.0 0.2921 0.6803
0.0804 2.38 680 0.1070 0.3108 0.4730 0.4570 nan 0.3806 0.5654 0.0 0.3730 0.5594
0.0467 2.45 700 0.1154 0.2576 0.3901 0.3715 nan 0.2828 0.4974 0.0 0.2782 0.4946
0.1093 2.52 720 0.1035 0.3819 0.5911 0.5641 nan 0.4353 0.7469 0.0 0.4326 0.7132
0.0975 2.59 740 0.1105 0.3186 0.4839 0.4634 nan 0.3658 0.6019 0.0 0.3570 0.5990
0.073 2.66 760 0.0997 0.3789 0.5831 0.5574 nan 0.4348 0.7314 0.0 0.4308 0.7058
0.0707 2.73 780 0.1029 0.3297 0.5169 0.4728 nan 0.2621 0.7716 0.0 0.2619 0.7273
0.0773 2.8 800 0.1080 0.3066 0.4655 0.4518 nan 0.3860 0.5451 0.0 0.3773 0.5424
0.1331 2.87 820 0.0986 0.3536 0.5407 0.5157 nan 0.3962 0.6852 0.0 0.3915 0.6693
0.1179 2.94 840 0.1160 0.2542 0.3856 0.3731 nan 0.3131 0.4581 0.0 0.3051 0.4576
0.107 3.01 860 0.1082 0.3336 0.5072 0.5030 nan 0.4831 0.5313 0.0 0.4726 0.5282
0.1113 3.08 880 0.1055 0.3116 0.4874 0.4569 nan 0.3110 0.6639 0.0 0.3102 0.6247
0.1252 3.15 900 0.1177 0.2671 0.4057 0.3930 nan 0.3322 0.4792 0.0 0.3285 0.4727
0.1175 3.22 920 0.1033 0.3595 0.5479 0.5346 nan 0.4709 0.6249 0.0 0.4618 0.6166
0.1227 3.29 940 0.1054 0.3316 0.5022 0.4793 nan 0.3696 0.6349 0.0 0.3640 0.6307
0.112 3.36 960 0.1165 0.3649 0.5604 0.5391 nan 0.4373 0.6836 0.0 0.4256 0.6692
0.1261 3.43 980 0.1025 0.3636 0.5579 0.5369 nan 0.4365 0.6793 0.0 0.4294 0.6616
0.0946 3.5 1000 0.1065 0.3146 0.4856 0.4603 nan 0.3395 0.6317 0.0 0.3364 0.6075
0.1034 3.57 1020 0.1142 0.2993 0.4536 0.4347 nan 0.3444 0.5628 0.0 0.3388 0.5589
0.148 3.64 1040 0.1038 0.3243 0.4949 0.4704 nan 0.3534 0.6363 0.0 0.3493 0.6234
0.0485 3.71 1060 0.1084 0.3227 0.4911 0.4750 nan 0.3979 0.5842 0.0 0.3895 0.5786
0.0977 3.78 1080 0.1158 0.2886 0.4387 0.4213 nan 0.3378 0.5396 0.0 0.3305 0.5352
0.0483 3.85 1100 0.1095 0.3233 0.4906 0.4659 nan 0.3480 0.6331 0.0 0.3406 0.6295
0.0904 3.92 1120 0.1191 0.2858 0.4341 0.4219 nan 0.3637 0.5046 0.0 0.3539 0.5034
0.0652 3.99 1140 0.1077 0.3449 0.5275 0.5087 nan 0.4189 0.6361 0.0 0.4110 0.6237
0.0647 4.06 1160 0.1184 0.3120 0.4741 0.4731 nan 0.4681 0.4802 0.0 0.4572 0.4788
0.1083 4.13 1180 0.1026 0.3622 0.5591 0.5305 nan 0.3940 0.7243 0.0 0.3900 0.6964
0.097 4.2 1200 0.1150 0.3055 0.4638 0.4407 nan 0.3302 0.5974 0.0 0.3252 0.5913
0.0577 4.27 1220 0.0989 0.3769 0.5744 0.5452 nan 0.4059 0.7429 0.0 0.3988 0.7319
0.0905 4.34 1240 0.1048 0.3180 0.4861 0.4615 nan 0.3442 0.6280 0.0 0.3425 0.6115
0.0622 4.41 1260 0.0958 0.3764 0.5772 0.5470 nan 0.4029 0.7515 0.0 0.3989 0.7304
0.1154 4.48 1280 0.1102 0.3525 0.5393 0.5386 nan 0.5352 0.5435 0.0 0.5186 0.5390
0.1005 4.55 1300 0.1081 0.3593 0.5496 0.5437 nan 0.5154 0.5838 0.0 0.5040 0.5739
0.0698 4.62 1320 0.1156 0.3279 0.4992 0.4926 nan 0.4615 0.5368 0.0 0.4487 0.5349
0.0758 4.69 1340 0.1124 0.3384 0.5161 0.5110 nan 0.4866 0.5457 0.0 0.4734 0.5417
0.0997 4.76 1360 0.1065 0.3157 0.4800 0.4559 nan 0.3405 0.6196 0.0 0.3362 0.6109
0.0878 4.83 1380 0.1015 0.3561 0.5424 0.5249 nan 0.4412 0.6437 0.0 0.4330 0.6354
0.0707 4.9 1400 0.1002 0.3398 0.5182 0.4862 nan 0.3331 0.7033 0.0 0.3297 0.6897
0.0649 4.97 1420 0.1267 0.3123 0.4773 0.4715 nan 0.4438 0.5107 0.0 0.4270 0.5100
0.0972 5.03 1440 0.1053 0.3397 0.5248 0.4902 nan 0.3251 0.7245 0.0 0.3242 0.6950
0.0572 5.1 1460 0.0991 0.3752 0.5773 0.5486 nan 0.4115 0.7431 0.0 0.4087 0.7170
0.0708 5.17 1480 0.0969 0.3593 0.5495 0.5186 nan 0.3712 0.7277 0.0 0.3667 0.7111
0.1332 5.24 1500 0.0984 0.3562 0.5433 0.5243 nan 0.4340 0.6525 0.0 0.4277 0.6409
0.0694 5.31 1520 0.0972 0.3703 0.5649 0.5440 nan 0.4438 0.6860 0.0 0.4370 0.6738
0.08 5.38 1540 0.0982 0.3846 0.5857 0.5671 nan 0.4783 0.6930 0.0 0.4711 0.6826
0.0813 5.45 1560 0.0966 0.3769 0.5737 0.5484 nan 0.4276 0.7198 0.0 0.4237 0.7071
0.1169 5.52 1580 0.1064 0.3271 0.4972 0.4849 nan 0.4260 0.5683 0.0 0.4149 0.5663
0.0689 5.59 1600 0.1038 0.3670 0.5610 0.5462 nan 0.4756 0.6463 0.0 0.4672 0.6339
0.0999 5.66 1620 0.1013 0.3647 0.5576 0.5412 nan 0.4630 0.6522 0.0 0.4536 0.6406
0.0787 5.73 1640 0.1090 0.3247 0.4924 0.4720 nan 0.3749 0.6098 0.0 0.3668 0.6073
0.0715 5.8 1660 0.1099 0.3695 0.5636 0.5551 nan 0.5147 0.6124 0.0 0.4976 0.6110
0.0926 5.87 1680 0.0988 0.3825 0.5837 0.5581 nan 0.4357 0.7316 0.0 0.4296 0.7178
0.0884 5.94 1700 0.1006 0.3566 0.5411 0.5199 nan 0.4188 0.6634 0.0 0.4138 0.6561
0.0552 6.01 1720 0.1058 0.3475 0.5279 0.4992 nan 0.3620 0.6937 0.0 0.3529 0.6896
0.0878 6.08 1740 0.1020 0.3471 0.5284 0.5012 nan 0.3711 0.6857 0.0 0.3651 0.6763
0.0887 6.15 1760 0.1062 0.3315 0.5034 0.4784 nan 0.3591 0.6477 0.0 0.3513 0.6431
0.0475 6.22 1780 0.1019 0.3594 0.5464 0.5227 nan 0.4094 0.6834 0.0 0.4024 0.6756
0.0659 6.29 1800 0.0989 0.4000 0.6156 0.5872 nan 0.4512 0.7800 0.0 0.4483 0.7517
0.0565 6.36 1820 0.1119 0.3080 0.4690 0.4436 nan 0.3221 0.6159 0.0 0.3187 0.6053
0.0598 6.43 1840 0.1119 0.3122 0.4755 0.4506 nan 0.3320 0.6190 0.0 0.3276 0.6089
0.0498 6.5 1860 0.1142 0.2978 0.4520 0.4271 nan 0.3079 0.5962 0.0 0.3051 0.5883
0.0968 6.57 1880 0.1041 0.3703 0.5638 0.5481 nan 0.4731 0.6545 0.0 0.4669 0.6440
0.0648 6.64 1900 0.1073 0.3697 0.5627 0.5423 nan 0.4449 0.6804 0.0 0.4356 0.6735
0.0892 6.71 1920 0.1095 0.3276 0.4978 0.4720 nan 0.3490 0.6465 0.0 0.3442 0.6386
0.066 6.78 1940 0.1111 0.3513 0.5344 0.5296 nan 0.5066 0.5623 0.0 0.4972 0.5566
0.0921 6.85 1960 0.1050 0.3576 0.5453 0.5285 nan 0.4481 0.6425 0.0 0.4406 0.6322
0.0762 6.92 1980 0.1066 0.3585 0.5476 0.5281 nan 0.4354 0.6597 0.0 0.4248 0.6507
0.0874 6.99 2000 0.1056 0.3390 0.5168 0.4934 nan 0.3816 0.6519 0.0 0.3753 0.6418
0.0775 7.06 2020 0.1010 0.3632 0.5550 0.5372 nan 0.4521 0.6579 0.0 0.4459 0.6438
0.0578 7.13 2040 0.1033 0.3673 0.5603 0.5442 nan 0.4675 0.6531 0.0 0.4588 0.6432
0.0722 7.2 2060 0.1048 0.3518 0.5356 0.5161 nan 0.4228 0.6485 0.0 0.4167 0.6387
0.0599 7.27 2080 0.1097 0.3453 0.5261 0.5100 nan 0.4328 0.6195 0.0 0.4256 0.6102
0.0798 7.34 2100 0.1064 0.3494 0.5371 0.5139 nan 0.4030 0.6711 0.0 0.3982 0.6499
0.0671 7.41 2120 0.1288 0.2880 0.4355 0.4143 nan 0.3132 0.5578 0.0 0.3084 0.5555
0.0688 7.48 2140 0.1080 0.3426 0.5207 0.4964 nan 0.3800 0.6614 0.0 0.3740 0.6538
0.0772 7.55 2160 0.1409 0.2690 0.4090 0.4058 nan 0.3906 0.4275 0.0 0.3805 0.4266
0.0542 7.62 2180 0.1142 0.3239 0.4905 0.4707 nan 0.3763 0.6047 0.0 0.3697 0.6020
0.0699 7.69 2200 0.1046 0.3775 0.5771 0.5432 nan 0.3808 0.7734 0.0 0.3758 0.7567
0.0484 7.76 2220 0.1163 0.3437 0.5225 0.5123 nan 0.4634 0.5816 0.0 0.4525 0.5786
0.0581 7.83 2240 0.1020 0.3700 0.5653 0.5367 nan 0.4002 0.7304 0.0 0.3966 0.7135
0.0565 7.9 2260 0.1041 0.3648 0.5529 0.5298 nan 0.4190 0.6869 0.0 0.4137 0.6809

Framework versions