segformer-b0-finetuned_orthophoto_tile_crack_0902_350

This model is a fine-tuned version of nvidia/mit-b0 on the alphaca/orthophoto_tile_crack_0902 dataset. It achieves the following results on the evaluation set:

• Loss: 0.1030
• Mean Iou: 0.1814
• Mean Accuracy: 0.3629
• Overall Accuracy: 0.3629
• Accuracy Unlabeled: nan
• Accuracy Crack: 0.3629
• Iou Unlabeled: 0.0
• Iou Crack: 0.3629

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:

• learning_rate: 0.0001
• train_batch_size: 20
• eval_batch_size: 20
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 300

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabeled	Accuracy Crack	Iou Unlabeled	Iou Crack
0.0488	1.8987	300	0.0646	0.0250	0.0499	0.0499	nan	0.0499	0.0	0.0499
0.04	3.7975	600	0.0562	0.1269	0.2537	0.2537	nan	0.2537	0.0	0.2537
0.0325	5.6962	900	0.0518	0.1278	0.2556	0.2556	nan	0.2556	0.0	0.2556
0.0407	7.5949	1200	0.0495	0.1704	0.3408	0.3408	nan	0.3408	0.0	0.3408
0.036	9.4937	1500	0.0476	0.1390	0.2780	0.2780	nan	0.2780	0.0	0.2780
0.0379	11.3924	1800	0.0469	0.1657	0.3314	0.3314	nan	0.3314	0.0	0.3314
0.0298	13.2911	2100	0.0464	0.1730	0.3459	0.3459	nan	0.3459	0.0	0.3459
0.0332	15.1899	2400	0.0475	0.1716	0.3431	0.3431	nan	0.3431	0.0	0.3431
0.0443	17.0886	2700	0.0472	0.1750	0.3501	0.3501	nan	0.3501	0.0	0.3501
0.0359	18.9873	3000	0.0481	0.1444	0.2888	0.2888	nan	0.2888	0.0	0.2888
0.0268	20.8861	3300	0.0473	0.1593	0.3185	0.3185	nan	0.3185	0.0	0.3185
0.0241	22.7848	3600	0.0496	0.1675	0.3349	0.3349	nan	0.3349	0.0	0.3349
0.0223	24.6835	3900	0.0484	0.2083	0.4166	0.4166	nan	0.4166	0.0	0.4166
0.0418	26.5823	4200	0.0500	0.1953	0.3905	0.3905	nan	0.3905	0.0	0.3905
0.0362	28.4810	4500	0.0495	0.1907	0.3814	0.3814	nan	0.3814	0.0	0.3814
0.0237	30.3797	4800	0.0512	0.1869	0.3737	0.3737	nan	0.3737	0.0	0.3737
0.0215	32.2785	5100	0.0511	0.1842	0.3683	0.3683	nan	0.3683	0.0	0.3683
0.0282	34.1772	5400	0.0526	0.1735	0.3470	0.3470	nan	0.3470	0.0	0.3470
0.0222	36.0759	5700	0.0522	0.1854	0.3708	0.3708	nan	0.3708	0.0	0.3708
0.022	37.9747	6000	0.0518	0.1817	0.3633	0.3633	nan	0.3633	0.0	0.3633
0.0245	39.8734	6300	0.0534	0.1733	0.3465	0.3465	nan	0.3465	0.0	0.3465
0.0262	41.7722	6600	0.0539	0.1714	0.3427	0.3427	nan	0.3427	0.0	0.3427
0.0254	43.6709	6900	0.0542	0.1857	0.3713	0.3713	nan	0.3713	0.0	0.3713
0.0222	45.5696	7200	0.0547	0.1815	0.3631	0.3631	nan	0.3631	0.0	0.3631
0.0204	47.4684	7500	0.0543	0.1728	0.3457	0.3457	nan	0.3457	0.0	0.3457
0.022	49.3671	7800	0.0559	0.1718	0.3435	0.3435	nan	0.3435	0.0	0.3435
0.0318	51.2658	8100	0.0579	0.1710	0.3420	0.3420	nan	0.3420	0.0	0.3420
0.0277	53.1646	8400	0.0586	0.1796	0.3593	0.3593	nan	0.3593	0.0	0.3593
0.0187	55.0633	8700	0.0596	0.1793	0.3585	0.3585	nan	0.3585	0.0	0.3585
0.0263	56.9620	9000	0.0580	0.1664	0.3328	0.3328	nan	0.3328	0.0	0.3328
0.0292	58.8608	9300	0.0595	0.1837	0.3673	0.3673	nan	0.3673	0.0	0.3673
0.0288	60.7595	9600	0.0583	0.1947	0.3894	0.3894	nan	0.3894	0.0	0.3894
0.0276	62.6582	9900	0.0584	0.2077	0.4153	0.4153	nan	0.4153	0.0	0.4153
0.0285	64.5570	10200	0.0610	0.1685	0.3369	0.3369	nan	0.3369	0.0	0.3369
0.0249	66.4557	10500	0.0614	0.1888	0.3777	0.3777	nan	0.3777	0.0	0.3777
0.0368	68.3544	10800	0.0590	0.1888	0.3775	0.3775	nan	0.3775	0.0	0.3775
0.0231	70.2532	11100	0.0646	0.1816	0.3632	0.3632	nan	0.3632	0.0	0.3632
0.0211	72.1519	11400	0.0620	0.1815	0.3630	0.3630	nan	0.3630	0.0	0.3630
0.0176	74.0506	11700	0.0647	0.2006	0.4012	0.4012	nan	0.4012	0.0	0.4012
0.0169	75.9494	12000	0.0653	0.1826	0.3651	0.3651	nan	0.3651	0.0	0.3651
0.0299	77.8481	12300	0.0646	0.1945	0.3890	0.3890	nan	0.3890	0.0	0.3890
0.0271	79.7468	12600	0.0640	0.1873	0.3745	0.3745	nan	0.3745	0.0	0.3745
0.0198	81.6456	12900	0.0681	0.1770	0.3539	0.3539	nan	0.3539	0.0	0.3539
0.0226	83.5443	13200	0.0675	0.1849	0.3699	0.3699	nan	0.3699	0.0	0.3699
0.0261	85.4430	13500	0.0686	0.1720	0.3441	0.3441	nan	0.3441	0.0	0.3441
0.0194	87.3418	13800	0.0670	0.1936	0.3871	0.3871	nan	0.3871	0.0	0.3871
0.0231	89.2405	14100	0.0651	0.1987	0.3974	0.3974	nan	0.3974	0.0	0.3974
0.0168	91.1392	14400	0.0727	0.1759	0.3518	0.3518	nan	0.3518	0.0	0.3518
0.0221	93.0380	14700	0.0708	0.1973	0.3946	0.3946	nan	0.3946	0.0	0.3946
0.0296	94.9367	15000	0.0722	0.1787	0.3575	0.3575	nan	0.3575	0.0	0.3575
0.0202	96.8354	15300	0.0723	0.1892	0.3783	0.3783	nan	0.3783	0.0	0.3783
0.0177	98.7342	15600	0.0686	0.1943	0.3885	0.3885	nan	0.3885	0.0	0.3885
0.0139	100.6329	15900	0.0694	0.1881	0.3762	0.3762	nan	0.3762	0.0	0.3762
0.0165	102.5316	16200	0.0713	0.1737	0.3473	0.3473	nan	0.3473	0.0	0.3473
0.0187	104.4304	16500	0.0739	0.1862	0.3725	0.3725	nan	0.3725	0.0	0.3725
0.0228	106.3291	16800	0.0706	0.1681	0.3361	0.3361	nan	0.3361	0.0	0.3361
0.0149	108.2278	17100	0.0710	0.1840	0.3680	0.3680	nan	0.3680	0.0	0.3680
0.0212	110.1266	17400	0.0742	0.1958	0.3916	0.3916	nan	0.3916	0.0	0.3916
0.0224	112.0253	17700	0.0775	0.1735	0.3469	0.3469	nan	0.3469	0.0	0.3469
0.0295	113.9241	18000	0.0730	0.1828	0.3656	0.3656	nan	0.3656	0.0	0.3656
0.015	115.8228	18300	0.0756	0.1883	0.3766	0.3766	nan	0.3766	0.0	0.3766
0.0153	117.7215	18600	0.0777	0.1736	0.3471	0.3471	nan	0.3471	0.0	0.3471
0.0177	119.6203	18900	0.0757	0.1798	0.3595	0.3595	nan	0.3595	0.0	0.3595
0.021	121.5190	19200	0.0753	0.1815	0.3630	0.3630	nan	0.3630	0.0	0.3630
0.0163	123.4177	19500	0.0786	0.1930	0.3860	0.3860	nan	0.3860	0.0	0.3860
0.0203	125.3165	19800	0.0773	0.1892	0.3784	0.3784	nan	0.3784	0.0	0.3784
0.0218	127.2152	20100	0.0774	0.1906	0.3813	0.3813	nan	0.3813	0.0	0.3813
0.0158	129.1139	20400	0.0800	0.1794	0.3589	0.3589	nan	0.3589	0.0	0.3589
0.0214	131.0127	20700	0.0810	0.1808	0.3616	0.3616	nan	0.3616	0.0	0.3616
0.0234	132.9114	21000	0.0800	0.1837	0.3675	0.3675	nan	0.3675	0.0	0.3675
0.0171	134.8101	21300	0.0806	0.1747	0.3495	0.3495	nan	0.3495	0.0	0.3495
0.022	136.7089	21600	0.0791	0.1714	0.3428	0.3428	nan	0.3428	0.0	0.3428
0.0158	138.6076	21900	0.0790	0.1844	0.3688	0.3688	nan	0.3688	0.0	0.3688
0.0209	140.5063	22200	0.0784	0.1830	0.3661	0.3661	nan	0.3661	0.0	0.3661
0.0269	142.4051	22500	0.0824	0.1765	0.3529	0.3529	nan	0.3529	0.0	0.3529
0.0149	144.3038	22800	0.0801	0.1923	0.3846	0.3846	nan	0.3846	0.0	0.3846
0.0169	146.2025	23100	0.0826	0.1815	0.3630	0.3630	nan	0.3630	0.0	0.3630
0.0198	148.1013	23400	0.0820	0.1876	0.3753	0.3753	nan	0.3753	0.0	0.3753
0.0081	150.0	23700	0.0824	0.1749	0.3498	0.3498	nan	0.3498	0.0	0.3498
0.0162	151.8987	24000	0.0837	0.1837	0.3673	0.3673	nan	0.3673	0.0	0.3673
0.0186	153.7975	24300	0.0814	0.1844	0.3688	0.3688	nan	0.3688	0.0	0.3688
0.0226	155.6962	24600	0.0803	0.1938	0.3875	0.3875	nan	0.3875	0.0	0.3875
0.016	157.5949	24900	0.0831	0.1748	0.3497	0.3497	nan	0.3497	0.0	0.3497
0.0155	159.4937	25200	0.0839	0.1902	0.3804	0.3804	nan	0.3804	0.0	0.3804
0.0241	161.3924	25500	0.0826	0.1887	0.3775	0.3775	nan	0.3775	0.0	0.3775
0.0163	163.2911	25800	0.0848	0.1823	0.3647	0.3647	nan	0.3647	0.0	0.3647
0.0139	165.1899	26100	0.0835	0.1858	0.3717	0.3717	nan	0.3717	0.0	0.3717
0.0177	167.0886	26400	0.0851	0.1918	0.3836	0.3836	nan	0.3836	0.0	0.3836
0.0205	168.9873	26700	0.0862	0.1819	0.3638	0.3638	nan	0.3638	0.0	0.3638
0.0273	170.8861	27000	0.0865	0.1822	0.3643	0.3643	nan	0.3643	0.0	0.3643
0.0186	172.7848	27300	0.0877	0.1912	0.3824	0.3824	nan	0.3824	0.0	0.3824
0.0133	174.6835	27600	0.0867	0.1859	0.3717	0.3717	nan	0.3717	0.0	0.3717
0.0175	176.5823	27900	0.0864	0.1846	0.3692	0.3692	nan	0.3692	0.0	0.3692
0.0259	178.4810	28200	0.0890	0.1865	0.3730	0.3730	nan	0.3730	0.0	0.3730
0.018	180.3797	28500	0.0872	0.1891	0.3782	0.3782	nan	0.3782	0.0	0.3782
0.0152	182.2785	28800	0.0861	0.1869	0.3737	0.3737	nan	0.3737	0.0	0.3737
0.0161	184.1772	29100	0.0910	0.1822	0.3644	0.3644	nan	0.3644	0.0	0.3644
0.0204	186.0759	29400	0.0888	0.1914	0.3828	0.3828	nan	0.3828	0.0	0.3828
0.0167	187.9747	29700	0.0897	0.1817	0.3634	0.3634	nan	0.3634	0.0	0.3634
0.0209	189.8734	30000	0.0900	0.1870	0.3739	0.3739	nan	0.3739	0.0	0.3739
0.02	191.7722	30300	0.0912	0.1929	0.3858	0.3858	nan	0.3858	0.0	0.3858
0.0167	193.6709	30600	0.0911	0.1849	0.3698	0.3698	nan	0.3698	0.0	0.3698
0.0184	195.5696	30900	0.0925	0.1845	0.3690	0.3690	nan	0.3690	0.0	0.3690
0.0202	197.4684	31200	0.0901	0.1869	0.3738	0.3738	nan	0.3738	0.0	0.3738
0.0143	199.3671	31500	0.0920	0.1844	0.3688	0.3688	nan	0.3688	0.0	0.3688
0.0151	201.2658	31800	0.0957	0.1846	0.3692	0.3692	nan	0.3692	0.0	0.3692
0.014	203.1646	32100	0.0925	0.1842	0.3684	0.3684	nan	0.3684	0.0	0.3684
0.0155	205.0633	32400	0.0971	0.1793	0.3587	0.3587	nan	0.3587	0.0	0.3587
0.0182	206.9620	32700	0.0929	0.1845	0.3689	0.3689	nan	0.3689	0.0	0.3689
0.0156	208.8608	33000	0.0936	0.1894	0.3788	0.3788	nan	0.3788	0.0	0.3788
0.0117	210.7595	33300	0.0925	0.1883	0.3766	0.3766	nan	0.3766	0.0	0.3766
0.0198	212.6582	33600	0.0942	0.1840	0.3679	0.3679	nan	0.3679	0.0	0.3679
0.0149	214.5570	33900	0.0942	0.1883	0.3766	0.3766	nan	0.3766	0.0	0.3766
0.0128	216.4557	34200	0.0956	0.1852	0.3704	0.3704	nan	0.3704	0.0	0.3704
0.0282	218.3544	34500	0.0960	0.1843	0.3687	0.3687	nan	0.3687	0.0	0.3687
0.0172	220.2532	34800	0.0941	0.1805	0.3610	0.3610	nan	0.3610	0.0	0.3610
0.0175	222.1519	35100	0.0957	0.1890	0.3781	0.3781	nan	0.3781	0.0	0.3781
0.0178	224.0506	35400	0.0960	0.1869	0.3738	0.3738	nan	0.3738	0.0	0.3738
0.0196	225.9494	35700	0.0939	0.1888	0.3777	0.3777	nan	0.3777	0.0	0.3777
0.0135	227.8481	36000	0.0962	0.1884	0.3768	0.3768	nan	0.3768	0.0	0.3768
0.0157	229.7468	36300	0.0968	0.1849	0.3697	0.3697	nan	0.3697	0.0	0.3697
0.0158	231.6456	36600	0.0991	0.1803	0.3606	0.3606	nan	0.3606	0.0	0.3606
0.0195	233.5443	36900	0.0985	0.1862	0.3723	0.3723	nan	0.3723	0.0	0.3723
0.0117	235.4430	37200	0.0974	0.1840	0.3679	0.3679	nan	0.3679	0.0	0.3679
0.0146	237.3418	37500	0.0987	0.1879	0.3757	0.3757	nan	0.3757	0.0	0.3757
0.0213	239.2405	37800	0.0999	0.1788	0.3577	0.3577	nan	0.3577	0.0	0.3577
0.0208	241.1392	38100	0.0982	0.1798	0.3597	0.3597	nan	0.3597	0.0	0.3597
0.0181	243.0380	38400	0.0971	0.1834	0.3667	0.3667	nan	0.3667	0.0	0.3667
0.0159	244.9367	38700	0.0986	0.1820	0.3640	0.3640	nan	0.3640	0.0	0.3640
0.0281	246.8354	39000	0.0987	0.1846	0.3692	0.3692	nan	0.3692	0.0	0.3692
0.0202	248.7342	39300	0.0987	0.1866	0.3732	0.3732	nan	0.3732	0.0	0.3732
0.0217	250.6329	39600	0.0991	0.1815	0.3631	0.3631	nan	0.3631	0.0	0.3631
0.0127	252.5316	39900	0.1008	0.1840	0.3680	0.3680	nan	0.3680	0.0	0.3680
0.0157	254.4304	40200	0.0998	0.1861	0.3722	0.3722	nan	0.3722	0.0	0.3722
0.0169	256.3291	40500	0.0981	0.1859	0.3718	0.3718	nan	0.3718	0.0	0.3718
0.0252	258.2278	40800	0.1000	0.1855	0.3710	0.3710	nan	0.3710	0.0	0.3710
0.0132	260.1266	41100	0.1016	0.1771	0.3542	0.3542	nan	0.3542	0.0	0.3542
0.0131	262.0253	41400	0.0992	0.1884	0.3768	0.3768	nan	0.3768	0.0	0.3768
0.017	263.9241	41700	0.1022	0.1848	0.3695	0.3695	nan	0.3695	0.0	0.3695
0.0139	265.8228	42000	0.1016	0.1832	0.3664	0.3664	nan	0.3664	0.0	0.3664
0.0204	267.7215	42300	0.1019	0.1831	0.3663	0.3663	nan	0.3663	0.0	0.3663
0.0134	269.6203	42600	0.1027	0.1858	0.3715	0.3715	nan	0.3715	0.0	0.3715
0.0203	271.5190	42900	0.1011	0.1846	0.3692	0.3692	nan	0.3692	0.0	0.3692
0.0173	273.4177	43200	0.1028	0.1821	0.3642	0.3642	nan	0.3642	0.0	0.3642
0.0204	275.3165	43500	0.1018	0.1866	0.3731	0.3731	nan	0.3731	0.0	0.3731
0.0167	277.2152	43800	0.1017	0.1857	0.3714	0.3714	nan	0.3714	0.0	0.3714
0.0219	279.1139	44100	0.1022	0.1853	0.3705	0.3705	nan	0.3705	0.0	0.3705
0.0238	281.0127	44400	0.1015	0.1876	0.3753	0.3753	nan	0.3753	0.0	0.3753
0.0191	282.9114	44700	0.1037	0.1836	0.3672	0.3672	nan	0.3672	0.0	0.3672
0.0164	284.8101	45000	0.1023	0.1845	0.3689	0.3689	nan	0.3689	0.0	0.3689
0.0148	286.7089	45300	0.1033	0.1844	0.3687	0.3687	nan	0.3687	0.0	0.3687
0.0177	288.6076	45600	0.1028	0.1812	0.3624	0.3624	nan	0.3624	0.0	0.3624
0.0158	290.5063	45900	0.1020	0.1854	0.3709	0.3709	nan	0.3709	0.0	0.3709
0.0169	292.4051	46200	0.1036	0.1828	0.3656	0.3656	nan	0.3656	0.0	0.3656
0.0168	294.3038	46500	0.1037	0.1837	0.3674	0.3674	nan	0.3674	0.0	0.3674
0.017	296.2025	46800	0.1031	0.1842	0.3685	0.3685	nan	0.3685	0.0	0.3685
0.015	298.1013	47100	0.1030	0.1838	0.3676	0.3676	nan	0.3676	0.0	0.3676
0.0287	300.0	47400	0.1030	0.1814	0.3629	0.3629	nan	0.3629	0.0	0.3629

Framework versions

• Transformers 4.42.3
• Pytorch 2.3.1+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1