segformer-b0-finetuned-wrinkle

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

• Loss: 0.0170
• Mean Iou: 0.1948
• Mean Accuracy: 0.3896
• Overall Accuracy: 0.3896
• Accuracy Unlabeled: nan
• Accuracy Wrinkle: 0.3896
• Iou Unlabeled: 0.0
• Iou Wrinkle: 0.3896

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: 6e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 10
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Accuracy Unlabeled	Accuracy Wrinkle	Iou Unlabeled	Iou Wrinkle	Validation Loss	Mean Accuracy	Mean Iou	Overall Accuracy
0.0207	0.0617	20	nan	0.3445	0.0	0.3445	0.0224	0.3445	0.1723	0.3445
0.0235	0.1235	40	nan	0.3026	0.0	0.3026	0.0214	0.3026	0.1513	0.3026
0.0159	0.1852	60	nan	0.3031	0.0	0.3031	0.0206	0.3031	0.1515	0.3031
0.0214	0.2469	80	nan	0.2827	0.0	0.2827	0.0204	0.2827	0.1413	0.2827
0.0204	0.3086	100	nan	0.2768	0.0	0.2768	0.0202	0.2768	0.1384	0.2768
0.0231	0.3704	120	nan	0.2846	0.0	0.2846	0.0200	0.2846	0.1423	0.2846
0.0151	0.4321	140	nan	0.2569	0.0	0.2569	0.0199	0.2569	0.1285	0.2569
0.0245	0.4938	160	nan	0.2752	0.0	0.2752	0.0197	0.2752	0.1376	0.2752
0.023	0.5556	180	nan	0.3322	0.0	0.3322	0.0200	0.3322	0.1661	0.3322
0.0321	0.6173	200	nan	0.3031	0.0	0.3031	0.0198	0.3031	0.1516	0.3031
0.0197	0.6790	220	nan	0.3157	0.0	0.3157	0.0196	0.3157	0.1578	0.3157
0.0192	0.7407	240	nan	0.3430	0.0	0.3430	0.0198	0.3430	0.1715	0.3430
0.0228	0.8025	260	nan	0.3144	0.0	0.3144	0.0194	0.3144	0.1572	0.3144
0.0216	0.8642	280	nan	0.3338	0.0	0.3338	0.0194	0.3338	0.1669	0.3338
0.0186	0.9259	300	nan	0.3411	0.0	0.3411	0.0193	0.3411	0.1705	0.3411
0.0202	0.9877	320	nan	0.3215	0.0	0.3215	0.0192	0.3215	0.1608	0.3215
0.0194	1.0494	340	nan	0.3286	0.0	0.3286	0.0194	0.3286	0.1643	0.3286
0.0166	1.1111	360	nan	0.2658	0.0	0.2658	0.0193	0.2658	0.1329	0.2658
0.0187	1.1728	380	nan	0.3133	0.0	0.3133	0.0191	0.3133	0.1567	0.3133
0.0208	1.2346	400	nan	0.3254	0.0	0.3254	0.0190	0.3254	0.1627	0.3254
0.0237	1.2963	420	nan	0.3093	0.0	0.3093	0.0190	0.3093	0.1546	0.3093
0.0244	1.3580	440	nan	0.3167	0.0	0.3167	0.0189	0.3167	0.1583	0.3167
0.0156	1.4198	460	nan	0.3408	0.0	0.3408	0.0191	0.3408	0.1704	0.3408
0.0231	1.4815	480	nan	0.2716	0.0	0.2716	0.0190	0.2716	0.1358	0.2716
0.0197	1.5432	500	nan	0.3540	0.0	0.3540	0.0190	0.3540	0.1770	0.3540
0.013	1.6049	520	nan	0.3358	0.0	0.3358	0.0188	0.3358	0.1679	0.3358
0.0158	1.6667	540	nan	0.3619	0.0	0.3619	0.0188	0.3619	0.1810	0.3619
0.0212	1.7284	560	nan	0.3649	0.0	0.3649	0.0188	0.3649	0.1824	0.3649
0.0231	1.7901	580	nan	0.2906	0.0	0.2906	0.0189	0.2906	0.1453	0.2906
0.0158	1.8519	600	nan	0.3210	0.0	0.3210	0.0187	0.3210	0.1605	0.3210
0.024	1.9136	620	nan	0.3117	0.0	0.3117	0.0187	0.3117	0.1558	0.3117
0.0193	1.9753	640	nan	0.3596	0.0	0.3596	0.0187	0.3596	0.1798	0.3596
0.0201	2.0370	660	nan	0.3628	0.0	0.3628	0.0187	0.3628	0.1814	0.3628
0.0162	2.0988	680	nan	0.3238	0.0	0.3238	0.0185	0.3238	0.1619	0.3238
0.0219	2.1605	700	nan	0.3622	0.0	0.3622	0.0186	0.3622	0.1811	0.3622
0.0207	2.2222	720	nan	0.3919	0.0	0.3919	0.0189	0.3919	0.1959	0.3919
0.0176	2.2840	740	nan	0.3116	0.0	0.3116	0.0185	0.3116	0.1558	0.3116
0.0149	2.3457	760	nan	0.3917	0.0	0.3917	0.0186	0.3917	0.1959	0.3917
0.0192	2.4074	780	nan	0.3097	0.0	0.3097	0.0184	0.3097	0.1548	0.3097
0.0202	2.4691	800	nan	0.3745	0.0	0.3745	0.0183	0.3745	0.1872	0.3745
0.0181	2.5309	820	nan	0.3687	0.0	0.3687	0.0185	0.3687	0.1844	0.3687
0.0168	2.5926	840	nan	0.3629	0.0	0.3629	0.0183	0.3629	0.1815	0.3629
0.0171	2.6543	860	nan	0.3148	0.0	0.3148	0.0183	0.3148	0.1574	0.3148
0.0166	2.7160	880	nan	0.3426	0.0	0.3426	0.0182	0.3426	0.1713	0.3426
0.0123	2.7778	900	nan	0.3670	0.0	0.3670	0.0183	0.3670	0.1835	0.3670
0.0149	2.8395	920	nan	0.3535	0.0	0.3535	0.0182	0.3535	0.1767	0.3535
0.016	2.9012	940	nan	0.3856	0.0	0.3856	0.0183	0.3856	0.1928	0.3856
0.0192	2.9630	960	nan	0.3647	0.0	0.3647	0.0182	0.3647	0.1823	0.3647
0.0188	3.0247	980	nan	0.3313	0.0	0.3313	0.0182	0.3313	0.1657	0.3313
0.0203	3.0864	1000	nan	0.3233	0.0	0.3233	0.0181	0.3233	0.1617	0.3233
0.0152	3.1481	1020	nan	0.3047	0.0	0.3047	0.0182	0.3047	0.1523	0.3047
0.0175	3.2099	1040	nan	0.3545	0.0	0.3545	0.0181	0.3545	0.1772	0.3545
0.0241	3.2716	1060	nan	0.3898	0.0	0.3898	0.0182	0.3898	0.1949	0.3898
0.0215	3.3333	1080	nan	0.3102	0.0	0.3102	0.0181	0.3102	0.1551	0.3102
0.0205	3.3951	1100	nan	0.3581	0.0	0.3581	0.0180	0.3581	0.1791	0.3581
0.0208	3.4568	1120	nan	0.3611	0.0	0.3611	0.0179	0.3611	0.1805	0.3611
0.0184	3.5185	1140	nan	0.3151	0.0	0.3151	0.0180	0.3151	0.1576	0.3151
0.0159	3.5802	1160	nan	0.4039	0.0	0.4039	0.0182	0.4039	0.2019	0.4039
0.0219	3.6420	1180	nan	0.3444	0.0	0.3444	0.0181	0.3444	0.1722	0.3444
0.023	3.7037	1200	nan	0.3265	0.0	0.3265	0.0179	0.3265	0.1633	0.3265
0.0148	3.7654	1220	nan	0.3677	0.0	0.3677	0.0179	0.3677	0.1838	0.3677
0.019	3.8272	1240	nan	0.3422	0.0	0.3422	0.0178	0.3422	0.1711	0.3422
0.0154	3.8889	1260	nan	0.3611	0.0	0.3611	0.0179	0.3611	0.1806	0.3611
0.0212	3.9506	1280	nan	0.3274	0.0	0.3274	0.0178	0.3274	0.1637	0.3274
0.0192	4.0123	1300	nan	0.3829	0.0	0.3829	0.0178	0.3829	0.1914	0.3829
0.0183	4.0741	1320	nan	0.4120	0.0	0.4120	0.0180	0.4120	0.2060	0.4120
0.0141	4.1358	1340	nan	0.3613	0.0	0.3613	0.0177	0.3613	0.1806	0.3613
0.0165	4.1975	1360	nan	0.3552	0.0	0.3552	0.0180	0.3552	0.1776	0.3552
0.018	4.2593	1380	nan	0.3646	0.0	0.3646	0.0177	0.3646	0.1823	0.3646
0.0166	4.3210	1400	nan	0.3557	0.0	0.3557	0.0177	0.3557	0.1778	0.3557
0.0193	4.3827	1420	nan	0.3518	0.0	0.3518	0.0177	0.3518	0.1759	0.3518
0.0173	4.4444	1440	nan	0.3640	0.0	0.3640	0.0176	0.3640	0.1820	0.3640
0.0159	4.5062	1460	nan	0.4083	0.0	0.4083	0.0178	0.4083	0.2042	0.4083
0.0186	4.5679	1480	nan	0.3679	0.0	0.3679	0.0177	0.3679	0.1839	0.3679
0.0214	4.6296	1500	nan	0.3567	0.0	0.3567	0.0176	0.3567	0.1783	0.3567
0.0192	4.6914	1520	nan	0.3595	0.0	0.3595	0.0176	0.3595	0.1797	0.3595
0.0173	4.7531	1540	nan	0.3567	0.0	0.3567	0.0176	0.3567	0.1784	0.3567
0.0185	4.8148	1560	nan	0.3830	0.0	0.3830	0.0176	0.3830	0.1915	0.3830
0.0179	4.8765	1580	nan	0.3851	0.0	0.3851	0.0176	0.3851	0.1926	0.3851
0.0152	4.9383	1600	nan	0.4017	0.0	0.4017	0.0177	0.4017	0.2009	0.4017
0.0281	5.0	1620	nan	0.3503	0.0	0.3503	0.0176	0.3503	0.1752	0.3503
0.0185	5.0617	1640	nan	0.4148	0.0	0.4148	0.0178	0.4148	0.2074	0.4148
0.0209	5.1235	1660	nan	0.3782	0.0	0.3782	0.0176	0.3782	0.1891	0.3782
0.0191	5.1852	1680	nan	0.3441	0.0	0.3441	0.0175	0.3441	0.1720	0.3441
0.0134	5.2469	1700	nan	0.3189	0.0	0.3189	0.0176	0.3189	0.1594	0.3189
0.0178	5.3086	1720	nan	0.3452	0.0	0.3452	0.0176	0.3452	0.1726	0.3452
0.0214	5.3704	1740	nan	0.4143	0.0	0.4143	0.0177	0.4143	0.2072	0.4143
0.0179	5.4321	1760	nan	0.3732	0.0	0.3732	0.0175	0.3732	0.1866	0.3732
0.0187	5.4938	1780	nan	0.3843	0.0	0.3843	0.0176	0.3843	0.1921	0.3843
0.0218	5.5556	1800	nan	0.3807	0.0	0.3807	0.0175	0.3807	0.1904	0.3807
0.0154	5.6173	1820	nan	0.3602	0.0	0.3602	0.0175	0.3602	0.1801	0.3602
0.021	5.6790	1840	nan	0.3445	0.0	0.3445	0.0175	0.3445	0.1723	0.3445
0.0191	5.7407	1860	nan	0.3290	0.0	0.3290	0.0175	0.3290	0.1645	0.3290
0.0194	5.8025	1880	nan	0.3564	0.0	0.3564	0.0174	0.3564	0.1782	0.3564
0.0121	5.8642	1900	nan	0.3860	0.0	0.3860	0.0175	0.3860	0.1930	0.3860
0.0194	5.9259	1920	nan	0.4023	0.0	0.4023	0.0175	0.4023	0.2011	0.4023
0.0159	5.9877	1940	nan	0.3953	0.0	0.3953	0.0173	0.3953	0.1976	0.3953
0.0142	6.0494	1960	nan	0.4036	0.0	0.4036	0.0173	0.4036	0.2018	0.4036
0.0165	6.1111	1980	nan	0.3593	0.0	0.3593	0.0173	0.3593	0.1796	0.3593
0.0166	6.1728	2000	nan	0.4104	0.0	0.4104	0.0174	0.4104	0.2052	0.4104
0.0164	6.2346	2020	nan	0.3653	0.0	0.3653	0.0174	0.3653	0.1826	0.3653
0.0267	6.2963	2040	nan	0.3582	0.0	0.3582	0.0173	0.3582	0.1791	0.3582
0.0202	6.3580	2060	nan	0.3613	0.0	0.3613	0.0173	0.3613	0.1806	0.3613
0.0134	6.4198	2080	nan	0.4004	0.0	0.4004	0.0174	0.4004	0.2002	0.4004
0.014	6.4815	2100	nan	0.3781	0.0	0.3781	0.0173	0.3781	0.1890	0.3781
0.0197	6.5432	2120	nan	0.3724	0.0	0.3724	0.0173	0.3724	0.1862	0.3724
0.0185	6.6049	2140	nan	0.3306	0.0	0.3306	0.0174	0.3306	0.1653	0.3306
0.0205	6.6667	2160	nan	0.3523	0.0	0.3523	0.0173	0.3523	0.1761	0.3523
0.0152	6.7284	2180	nan	0.3690	0.0	0.3690	0.0172	0.3690	0.1845	0.3690
0.0158	6.7901	2200	0.0172	0.1793	0.3586	0.3586	nan	0.3586	0.0	0.3586
0.0199	6.8519	2220	0.0174	0.2063	0.4126	0.4126	nan	0.4126	0.0	0.4126
0.0186	6.9136	2240	0.0173	0.1691	0.3382	0.3382	nan	0.3382	0.0	0.3382
0.0228	6.9753	2260	0.0173	0.1804	0.3609	0.3609	nan	0.3609	0.0	0.3609
0.0193	7.0370	2280	0.0172	0.1887	0.3775	0.3775	nan	0.3775	0.0	0.3775
0.0161	7.0988	2300	0.0172	0.1900	0.3799	0.3799	nan	0.3799	0.0	0.3799
0.0182	7.1605	2320	0.0172	0.1888	0.3775	0.3775	nan	0.3775	0.0	0.3775
0.0165	7.2222	2340	0.0172	0.2037	0.4075	0.4075	nan	0.4075	0.0	0.4075
0.0197	7.2840	2360	0.0172	0.1900	0.3800	0.3800	nan	0.3800	0.0	0.3800
0.0246	7.3457	2380	0.0172	0.1861	0.3722	0.3722	nan	0.3722	0.0	0.3722
0.0191	7.4074	2400	0.0172	0.1935	0.3870	0.3870	nan	0.3870	0.0	0.3870
0.0157	7.4691	2420	0.0172	0.1913	0.3826	0.3826	nan	0.3826	0.0	0.3826
0.0187	7.5309	2440	0.0172	0.1863	0.3727	0.3727	nan	0.3727	0.0	0.3727
0.0199	7.5926	2460	0.0172	0.1817	0.3634	0.3634	nan	0.3634	0.0	0.3634
0.0226	7.6543	2480	0.0172	0.1944	0.3889	0.3889	nan	0.3889	0.0	0.3889
0.0167	7.7160	2500	0.0172	0.1833	0.3666	0.3666	nan	0.3666	0.0	0.3666
0.017	7.7778	2520	0.0171	0.1924	0.3848	0.3848	nan	0.3848	0.0	0.3848
0.0219	7.8395	2540	0.0171	0.1885	0.3769	0.3769	nan	0.3769	0.0	0.3769
0.0149	7.9012	2560	0.0172	0.2088	0.4177	0.4177	nan	0.4177	0.0	0.4177
0.0137	7.9630	2580	0.0171	0.1932	0.3865	0.3865	nan	0.3865	0.0	0.3865
0.0235	8.0247	2600	0.0171	0.1837	0.3674	0.3674	nan	0.3674	0.0	0.3674
0.0169	8.0864	2620	0.0171	0.2023	0.4046	0.4046	nan	0.4046	0.0	0.4046
0.0183	8.1481	2640	0.0171	0.1876	0.3753	0.3753	nan	0.3753	0.0	0.3753
0.0171	8.2099	2660	0.0171	0.1876	0.3753	0.3753	nan	0.3753	0.0	0.3753
0.0202	8.2716	2680	0.0172	0.2087	0.4173	0.4173	nan	0.4173	0.0	0.4173
0.0229	8.3333	2700	0.0171	0.1799	0.3599	0.3599	nan	0.3599	0.0	0.3599
0.0138	8.3951	2720	0.0171	0.1941	0.3882	0.3882	nan	0.3882	0.0	0.3882
0.0155	8.4568	2740	0.0171	0.2031	0.4063	0.4063	nan	0.4063	0.0	0.4063
0.0171	8.5185	2760	0.0171	0.2038	0.4075	0.4075	nan	0.4075	0.0	0.4075
0.0211	8.5802	2780	0.0170	0.1946	0.3893	0.3893	nan	0.3893	0.0	0.3893
0.0161	8.6420	2800	0.0170	0.1950	0.3900	0.3900	nan	0.3900	0.0	0.3900
0.0175	8.7037	2820	0.0171	0.1946	0.3893	0.3893	nan	0.3893	0.0	0.3893
0.0199	8.7654	2840	0.0171	0.1948	0.3895	0.3895	nan	0.3895	0.0	0.3895
0.0169	8.8272	2860	0.0171	0.2013	0.4027	0.4027	nan	0.4027	0.0	0.4027
0.0213	8.8889	2880	0.0170	0.1964	0.3927	0.3927	nan	0.3927	0.0	0.3927
0.018	8.9506	2900	0.0171	0.1850	0.3700	0.3700	nan	0.3700	0.0	0.3700
0.0195	9.0123	2920	0.0171	0.2041	0.4081	0.4081	nan	0.4081	0.0	0.4081
0.0156	9.0741	2940	0.0171	0.1922	0.3844	0.3844	nan	0.3844	0.0	0.3844
0.0201	9.1358	2960	0.0171	0.1985	0.3971	0.3971	nan	0.3971	0.0	0.3971
0.0223	9.1975	2980	0.0171	0.1970	0.3940	0.3940	nan	0.3940	0.0	0.3940
0.0193	9.2593	3000	0.0170	0.1918	0.3836	0.3836	nan	0.3836	0.0	0.3836
0.0121	9.3210	3020	0.0170	0.1978	0.3956	0.3956	nan	0.3956	0.0	0.3956
0.0183	9.3827	3040	0.0170	0.1922	0.3845	0.3845	nan	0.3845	0.0	0.3845
0.0208	9.4444	3060	0.0170	0.1966	0.3932	0.3932	nan	0.3932	0.0	0.3932
0.0207	9.5062	3080	0.0170	0.1967	0.3933	0.3933	nan	0.3933	0.0	0.3933
0.0194	9.5679	3100	0.0170	0.1922	0.3843	0.3843	nan	0.3843	0.0	0.3843
0.0158	9.6296	3120	0.0170	0.1965	0.3929	0.3929	nan	0.3929	0.0	0.3929
0.0158	9.6914	3140	0.0170	0.1898	0.3796	0.3796	nan	0.3796	0.0	0.3796
0.0215	9.7531	3160	0.0170	0.1922	0.3845	0.3845	nan	0.3845	0.0	0.3845
0.0143	9.8148	3180	0.0170	0.1989	0.3979	0.3979	nan	0.3979	0.0	0.3979
0.0177	9.8765	3200	0.0170	0.1935	0.3870	0.3870	nan	0.3870	0.0	0.3870
0.0159	9.9383	3220	0.0170	0.1911	0.3822	0.3822	nan	0.3822	0.0	0.3822
0.0189	10.0	3240	0.0170	0.1948	0.3896	0.3896	nan	0.3896	0.0	0.3896

Framework versions

• Transformers 4.46.3
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.20.3