Modelpublic

jhaberbe/segformer-b0-finetuned-lipid-droplets-v4

sourceHugging Faceotherupdated 2y agoView on Hugging Face

0likes4downloads

Model Card

segformer-b0-finetuned-lipid-droplets-v4

This model is a fine-tuned version of nvidia/mit-b0 on the jhaberbe/lipid-droplets-v4 dataset. It achieves the following results on the evaluation set:

—Loss: 0.0579
—Mean Iou: 0.2161
—Mean Accuracy: 0.4321
—Overall Accuracy: 0.4321
—Accuracy Unlabeled: nan
—Accuracy Lipid: 0.4321
—Iou Unlabeled: 0.0
—Iou Lipid: 0.4321

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
—trainbatchsize: 2
—evalbatchsize: 2
—seed: 42
—optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
—lrschedulertype: linear
—num_epochs: 1000

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy	Accuracy Unlabeled	Accuracy Lipid	Iou Lipid
0.0683	5.0	20	0.4807	0.006	0.012	0.012	nan	0.012	0.012
0.0727	10.0	40	0.2153	0.0	0.0	0.0	nan	0.0	0.0
0.0324	15.0	60	0.4132	0.0	0.0	0.0	nan	0.0	0.0
0.0229	20.0	80	0.1490	0.0	0.0	0.0	nan	0.0	0.0
0.0314	25.0	100	0.4783	0.0097	0.0193	0.0193	nan	0.0193	0.0193
0.0187	30.0	120	0.0686	0.0	0.0	0.0	nan	0.0	0.0
0.0326	35.0	140	0.0456	0.0	0.0	0.0	nan	0.0	0.0
0.024	40.0	160	0.2099	0.0	0.0	0.0	nan	0.0	0.0
0.0162	45.0	180	0.1570	0.0009	0.0018	0.0018	nan	0.0018	0.0018
0.0129	50.0	200	0.1300	0.0101	0.0201	0.0201	nan	0.0201	0.0201
0.013	55.0	220	0.1372	0.0044	0.0088	0.0088	nan	0.0088	0.0088
0.0147	60.0	240	0.1529	0.0484	0.0968	0.0968	nan	0.0968	0.0968
0.0192	65.0	260	0.0777	0.0094	0.0188	0.0188	nan	0.0188	0.0188
0.0105	70.0	280	0.1133	0.0073	0.0146	0.0146	nan	0.0146	0.0146
0.0086	75.0	300	0.1237	0.0039	0.0077	0.0077	nan	0.0077	0.0077
0.0151	80.0	320	0.0472	0.0001	0.0001	0.0001	nan	0.0001	0.0001
0.013	85.0	340	0.0373	0.0014	0.0028	0.0028	nan	0.0028	0.0028
0.053	90.0	360	0.0745	0.0036	0.0072	0.0072	nan	0.0072	0.0072
0.0117	95.0	380	0.0097	0.0	0.0	0.0	nan	0.0	0.0
0.018	100.0	400	0.1363	0.1399	0.2797	0.2797	nan	0.2797	0.2797
0.0099	105.0	420	0.0888	0.0053	0.0106	0.0106	nan	0.0106	0.0106
0.0063	110.0	440	0.0889	0.0134	0.0269	0.0269	nan	0.0269	0.0269
0.0127	115.0	460	0.0389	0.0059	0.0119	0.0119	nan	0.0119	0.0119
0.0153	120.0	480	0.0579	0.0048	0.0095	0.0095	nan	0.0095	0.0095
0.0064	125.0	500	0.0589	0.0415	0.0830	0.0830	nan	0.0830	0.0830
0.0074	130.0	520	0.1361	0.0917	0.1833	0.1833	nan	0.1833	0.1833
0.0062	135.0	540	0.1212	0.1098	0.2196	0.2196	nan	0.2196	0.2196
0.0053	140.0	560	0.0814	0.0954	0.1908	0.1908	nan	0.1908	0.1908
0.0065	145.0	580	0.1418	0.1704	0.3408	0.3408	nan	0.3408	0.3408
0.0118	150.0	600	0.0541	0.0210	0.0419	0.0419	nan	0.0419	0.0419
0.0111	155.0	620	0.0801	0.0419	0.0839	0.0839	nan	0.0839	0.0839
0.0077	160.0	640	0.0628	0.2041	0.4081	0.4081	nan	0.4081	0.4081
0.0118	165.0	660	0.1001	0.1372	0.2743	0.2743	nan	0.2743	0.2743
0.0078	170.0	680	0.1254	0.1439	0.2879	0.2879	nan	0.2879	0.2879
0.0058	175.0	700	0.1119	0.2726	0.5452	0.5452	nan	0.5452	0.5452
0.0097	180.0	720	0.1608	0.3098	0.6196	0.6196	nan	0.6196	0.6196
0.0063	185.0	740	0.1617	0.2622	0.5244	0.5244	nan	0.5244	0.5244
0.0095	190.0	760	0.0526	0.0804	0.1608	0.1608	nan	0.1608	0.1608
0.0119	195.0	780	0.1438	0.2352	0.4703	0.4703	nan	0.4703	0.4703
0.0264	200.0	800	0.1563	0.3272	0.6545	0.6545	nan	0.6545	0.6545
0.0043	205.0	820	0.0605	0.2524	0.5048	0.5048	nan	0.5048	0.5048
0.0055	210.0	840	0.1585	0.3650	0.7299	0.7299	nan	0.7299	0.7299
0.009	215.0	860	0.1218	0.2845	0.5690	0.5690	nan	0.5690	0.5690
0.0295	220.0	880	0.5198	0.4981	0.9963	0.9963	nan	0.9963	0.9963
0.004	225.0	900	0.1908	0.2370	0.4741	0.4741	nan	0.4741	0.4741
0.0108	230.0	920	0.0077	0.0	0.0	0.0	nan	0.0	0.0
0.0034	235.0	940	0.1812	0.1838	0.3676	0.3676	nan	0.3676	0.3676
0.0058	240.0	960	0.1195	0.1750	0.3501	0.3501	nan	0.3501	0.3501
0.0076	245.0	980	0.1365	0.0957	0.1914	0.1914	nan	0.1914	0.1914
0.0056	250.0	1000	0.0098	0.0	0.0	0.0	nan	0.0	0.0
0.0088	255.0	1020	0.0514	0.1783	0.3567	0.3567	nan	0.3567	0.3567
0.0143	260.0	1040	0.0106	0.0002	0.0004	0.0004	nan	0.0004	0.0004
0.0052	265.0	1060	0.0749	0.0476	0.0952	0.0952	nan	0.0952	0.0952
0.0084	270.0	1080	0.1056	0.2186	0.4372	0.4372	nan	0.4372	0.4372
0.0084	275.0	1100	0.1106	0.0734	0.1469	0.1469	nan	0.1469	0.1469
0.0043	280.0	1120	0.1078	0.3212	0.6425	0.6425	nan	0.6425	0.6425
0.0079	285.0	1140	0.0610	0.2426	0.4851	0.4851	nan	0.4851	0.4851
0.0037	290.0	1160	0.0624	0.2492	0.4985	0.4985	nan	0.4985	0.4985
0.0084	295.0	1180	0.1833	0.2811	0.5622	0.5622	nan	0.5622	0.5622
0.0049	300.0	1200	0.0843	0.2817	0.5634	0.5634	nan	0.5634	0.5634
0.0064	305.0	1220	0.0625	0.1790	0.3581	0.3581	nan	0.3581	0.3581
0.011	310.0	1240	0.0653	0.1121	0.2241	0.2241	nan	0.2241	0.2241
0.0041	315.0	1260	0.0620	0.198	0.396	0.396	nan	0.396	0.396
0.0079	320.0	1280	0.0733	0.11	0.22	0.22	nan	0.22	0.22
0.0039	325.0	1300	0.2672	0.2041	0.4081	0.4081	nan	0.4081	0.4081
0.0081	330.0	1320	0.0231	0.0708	0.1415	0.1415	nan	0.1415	0.1415
0.0038	335.0	1340	0.0244	0.1270	0.2541	0.2541	nan	0.2541	0.2541
0.0078	340.0	1360	0.0095	0.0198	0.0396	0.0396	nan	0.0396	0.0396
0.0061	345.0	1380	0.1181	0.1386	0.2772	0.2772	nan	0.2772	0.2772
0.0099	350.0	1400	0.0350	0.0970	0.1939	0.1939	nan	0.1939	0.1939
0.0066	355.0	1420	0.0303	0.1477	0.2954	0.2954	nan	0.2954	0.2954
0.0083	360.0	1440	0.0919	0.1237	0.2474	0.2474	nan	0.2474	0.2474
0.0082	365.0	1460	0.0087	0.0	0.0	0.0	nan	0.0	0.0
0.0078	370.0	1480	0.1443	0.3666	0.7331	0.7331	nan	0.7331	0.7331
0.0075	375.0	1500	0.0493	0.038	0.076	0.076	nan	0.076	0.076
0.0058	380.0	1520	0.0941	0.2978	0.5956	0.5956	nan	0.5956	0.5956
0.0045	385.0	1540	0.0353	0.1239	0.2479	0.2479	nan	0.2479	0.2479
0.0077	390.0	1560	0.0469	0.1066	0.2131	0.2131	nan	0.2131	0.2131
0.0057	395.0	1580	0.0533	0.1192	0.2383	0.2383	nan	0.2383	0.2383
0.0073	400.0	1600	0.1126	0.1331	0.2662	0.2662	nan	0.2662	0.2662
0.0042	405.0	1620	0.1222	0.2205	0.4410	0.4410	nan	0.4410	0.4410
0.0078	410.0	1640	0.1111	0.2519	0.5039	0.5039	nan	0.5039	0.5039
0.0056	415.0	1660	0.0988	0.2311	0.4622	0.4622	nan	0.4622	0.4622
0.0093	420.0	1680	0.0464	0.0693	0.1386	0.1386	nan	0.1386	0.1386
0.0072	425.0	1700	0.0751	0.228	0.456	0.456	nan	0.456	0.456
0.0026	430.0	1720	0.0728	0.2397	0.4793	0.4793	nan	0.4793	0.4793
0.0051	435.0	1740	0.0443	0.0879	0.1757	0.1757	nan	0.1757	0.1757
0.0092	440.0	1760	0.0694	0.1514	0.3029	0.3029	nan	0.3029	0.3029
0.0079	445.0	1780	0.0287	0.1381	0.2761	0.2761	nan	0.2761	0.2761
0.0048	450.0	1800	0.0597	0.2146	0.4292	0.4292	nan	0.4292	0.4292
0.007	455.0	1820	0.1145	0.232	0.464	0.464	nan	0.464	0.464
0.0062	460.0	1840	0.0854	0.2715	0.5430	0.5430	nan	0.5430	0.5430
0.0031	465.0	1860	0.1299	0.2943	0.5886	0.5886	nan	0.5886	0.5886
0.0056	470.0	1880	0.0560	0.1650	0.3299	0.3299	nan	0.3299	0.3299
0.0031	475.0	1900	0.0706	0.1521	0.3043	0.3043	nan	0.3043	0.3043
0.0024	480.0	1920	0.0478	0.1862	0.3724	0.3724	nan	0.3724	0.3724
0.007	485.0	1940	0.0357	0.1796	0.3592	0.3592	nan	0.3592	0.3592
0.0075	490.0	1960	0.0598	0.1544	0.3088	0.3088	nan	0.3088	0.3088
0.0031	495.0	1980	0.0447	0.3236	0.6472	0.6472	nan	0.6472	0.6472
0.007	500.0	2000	0.1198	0.3270	0.6541	0.6541	nan	0.6541	0.6541
0.009	505.0	2020	0.0550	0.2481	0.4963	0.4963	nan	0.4963	0.4963
0.0042	510.0	2040	0.0649	0.2150	0.4299	0.4299	nan	0.4299	0.4299
0.0024	515.0	2060	0.0731	0.2039	0.4079	0.4079	nan	0.4079	0.4079
0.0034	520.0	2080	0.0812	0.2652	0.5303	0.5303	nan	0.5303	0.5303
0.0039	525.0	2100	0.0719	0.1894	0.3788	0.3788	nan	0.3788	0.3788
0.0061	530.0	2120	0.0943	0.2135	0.4270	0.4270	nan	0.4270	0.4270
0.0054	535.0	2140	0.0578	0.2337	0.4674	0.4674	nan	0.4674	0.4674
0.0074	540.0	2160	0.1498	0.4530	0.9059	0.9059	nan	0.9059	0.9059
0.0052	545.0	2180	0.1614	0.4556	0.9112	0.9112	nan	0.9112	0.9112
0.0068	550.0	2200	0.0747	0.3214	0.6428	0.6428	nan	0.6428	0.6428
0.0068	555.0	2220	0.0462	0.1623	0.3247	0.3247	nan	0.3247	0.3247
0.0097	560.0	2240	0.0748	0.3294	0.6588	0.6588	nan	0.6588	0.6588
0.0066	565.0	2260	0.0913	0.2101	0.4203	0.4203	nan	0.4203	0.4203
0.0054	570.0	2280	0.0472	0.1623	0.3247	0.3247	nan	0.3247	0.3247
0.0039	575.0	2300	0.0361	0.0861	0.1721	0.1721	nan	0.1721	0.1721
0.0023	580.0	2320	0.0547	0.1886	0.3772	0.3772	nan	0.3772	0.3772
0.0056	585.0	2340	0.0668	0.2911	0.5822	0.5822	nan	0.5822	0.5822
0.0055	590.0	2360	0.0511	0.2152	0.4303	0.4303	nan	0.4303	0.4303
0.0027	595.0	2380	0.0570	0.2193	0.4386	0.4386	nan	0.4386	0.4386
0.0046	600.0	2400	0.0373	0.0786	0.1572	0.1572	nan	0.1572	0.1572
0.0035	605.0	2420	0.0468	0.0870	0.1741	0.1741	nan	0.1741	0.1741
0.0022	610.0	2440	0.0476	0.1029	0.2058	0.2058	nan	0.2058	0.2058
0.0058	615.0	2460	0.0418	0.1003	0.2007	0.2007	nan	0.2007	0.2007
0.003	620.0	2480	0.0520	0.2134	0.4269	0.4269	nan	0.4269	0.4269
0.0072	625.0	2500	0.0475	0.198	0.396	0.396	nan	0.396	0.396
0.0054	630.0	2520	0.0371	0.1370	0.2741	0.2741	nan	0.2741	0.2741
0.0033	635.0	2540	0.0635	0.132	0.264	0.264	nan	0.264	0.264
0.0076	640.0	2560	0.0540	0.1278	0.2556	0.2556	nan	0.2556	0.2556
0.0067	645.0	2580	0.0339	0.128	0.256	0.256	nan	0.256	0.256
0.0032	650.0	2600	0.0399	0.1341	0.2683	0.2683	nan	0.2683	0.2683
0.0024	655.0	2620	0.0467	0.108	0.216	0.216	nan	0.216	0.216
0.0032	660.0	2640	0.0472	0.1653	0.3306	0.3306	nan	0.3306	0.3306
0.006	665.0	2660	0.0540	0.1963	0.3927	0.3927	nan	0.3927	0.3927
0.0021	670.0	2680	0.0532	0.1886	0.3771	0.3771	nan	0.3771	0.3771
0.0037	675.0	2700	0.0661	0.2588	0.5177	0.5177	nan	0.5177	0.5177
0.0073	680.0	2720	0.0514	0.1327	0.2654	0.2654	nan	0.2654	0.2654
0.0038	685.0	2740	0.0596	0.2090	0.4179	0.4179	nan	0.4179	0.4179
0.0038	690.0	2760	0.0421	0.1279	0.2557	0.2557	nan	0.2557	0.2557
0.0036	695.0	2780	0.0626	0.1976	0.3952	0.3952	nan	0.3952	0.3952
0.003	700.0	2800	0.0525	0.212	0.424	0.424	nan	0.424	0.424
0.0046	705.0	2820	0.0515	0.274	0.548	0.548	nan	0.548	0.548
0.004	710.0	2840	0.0691	0.2222	0.4444	0.4444	nan	0.4444	0.4444
0.0029	715.0	2860	0.0495	0.1559	0.3117	0.3117	nan	0.3117	0.3117
0.0057	720.0	2880	0.0442	0.1541	0.3081	0.3081	nan	0.3081	0.3081
0.007	725.0	2900	0.0835	0.1562	0.3124	0.3124	nan	0.3124	0.3124
0.0023	730.0	2920	0.0686	0.1466	0.2931	0.2931	nan	0.2931	0.2931
0.0054	735.0	2940	0.0813	0.1685	0.3370	0.3370	nan	0.3370	0.3370
0.0035	740.0	2960	0.0679	0.1623	0.3247	0.3247	nan	0.3247	0.3247
0.0035	745.0	2980	0.0535	0.1502	0.3004	0.3004	nan	0.3004	0.3004
0.003	750.0	3000	0.0645	0.1266	0.2531	0.2531	nan	0.2531	0.2531
0.0059	755.0	3020	0.0617	0.1535	0.3070	0.3070	nan	0.3070	0.3070
0.004	760.0	3040	0.0698	0.1315	0.2630	0.2630	nan	0.2630	0.2630
0.0032	765.0	3060	0.0515	0.1148	0.2297	0.2297	nan	0.2297	0.2297
0.0031	770.0	3080	0.0574	0.1309	0.2618	0.2618	nan	0.2618	0.2618
0.0039	775.0	3100	0.0480	0.1624	0.3248	0.3248	nan	0.3248	0.3248
0.0064	780.0	3120	0.0490	0.1434	0.2869	0.2869	nan	0.2869	0.2869
0.0056	785.0	3140	0.0492	0.1514	0.3028	0.3028	nan	0.3028	0.3028
0.0018	790.0	3160	0.0703	0.1704	0.3408	0.3408	nan	0.3408	0.3408
0.0015	795.0	3180	0.0819	0.2045	0.4090	0.4090	nan	0.4090	0.4090
0.0042	800.0	3200	0.0774	0.2131	0.4262	0.4262	nan	0.4262	0.4262
0.0037	805.0	3220	0.0510	0.1390	0.2779	0.2779	nan	0.2779	0.2779
0.0032	810.0	3240	0.0628	0.1786	0.3571	0.3571	nan	0.3571	0.3571
0.0026	815.0	3260	0.0778	0.1746	0.3492	0.3492	nan	0.3492	0.3492
0.0022	820.0	3280	0.0600	0.1619	0.3237	0.3237	nan	0.3237	0.3237
0.0042	825.0	3300	0.0691	0.1792	0.3583	0.3583	nan	0.3583	0.3583
0.0032	830.0	3320	0.0831	0.2368	0.4737	0.4737	nan	0.4737	0.4737
0.0041	835.0	3340	0.0627	0.1748	0.3495	0.3495	nan	0.3495	0.3495
0.0024	840.0	3360	0.0598	0.1503	0.3007	0.3007	nan	0.3007	0.3007
0.0017	845.0	3380	0.0719	0.1588	0.3175	0.3175	nan	0.3175	0.3175
0.0044	850.0	3400	0.0526	0.1526	0.3051	0.3051	nan	0.3051	0.3051
0.0059	855.0	3420	0.0610	0.17	0.34	0.34	nan	0.34	0.34
0.0046	860.0	3440	0.0628	0.1697	0.3394	0.3394	nan	0.3394	0.3394
0.0058	865.0	3460	0.0507	0.1936	0.3872	0.3872	nan	0.3872	0.3872
0.0049	870.0	3480	0.0482	0.1659	0.3319	0.3319	nan	0.3319	0.3319
0.003	875.0	3500	0.0692	0.2345	0.4690	0.4690	nan	0.4690	0.4690
0.0048	880.0	3520	0.0525	0.1594	0.3189	0.3189	nan	0.3189	0.3189
0.003	885.0	3540	0.0564	0.2008	0.4015	0.4015	nan	0.4015	0.4015
0.004	890.0	3560	0.0428	0.1561	0.3123	0.3123	nan	0.3123	0.3123
0.0018	895.0	3580	0.0454	0.144	0.288	0.288	nan	0.288	0.288
0.0023	900.0	3600	0.0460	0.1642	0.3284	0.3284	nan	0.3284	0.3284
0.0043	905.0	3620	0.0513	0.2052	0.4103	0.4103	nan	0.4103	0.4103
0.0013	910.0	3640	0.0502	0.1739	0.3479	0.3479	nan	0.3479	0.3479
0.0057	915.0	3660	0.0440	0.1687	0.3374	0.3374	nan	0.3374	0.3374
0.006	920.0	3680	0.0483	0.1517	0.3034	0.3034	nan	0.3034	0.3034
0.0029	925.0	3700	0.0443	0.1550	0.3101	0.3101	nan	0.3101	0.3101
0.0044	930.0	3720	0.0506	0.1661	0.3323	0.3323	nan	0.3323	0.3323
0.0016	935.0	3740	0.0612	0.1957	0.3914	0.3914	nan	0.3914	0.3914
0.0026	940.0	3760	0.0525	0.1766	0.3531	0.3531	nan	0.3531	0.3531
0.005	945.0	3780	0.0551	0.2171	0.4342	0.4342	nan	0.4342	0.4342
0.0035	950.0	3800	0.0554	0.1841	0.3683	0.3683	nan	0.3683	0.3683
0.003	955.0	3820	0.0271	0.1519	0.3037	0.3037	nan	0.3037	0.3037
0.0054	960.0	3840	0.0493	0.1688	0.3375	0.3375	nan	0.3375	0.3375
0.0031	965.0	3860	0.0518	0.1751	0.3502	0.3502	nan	0.3502	0.3502
0.005	970.0	3880	0.0569	0.1903	0.3807	0.3807	nan	0.3807	0.3807
0.0023	975.0	3900	0.0498	0.1952	0.3903	0.3903	nan	0.3903	0.3903
0.0023	980.0	3920	0.0581	0.2254	0.4508	0.4508	nan	0.4508	0.4508
0.0014	985.0	3940	0.0361	0.1692	0.3383	0.3383	nan	0.3383	0.3383
0.0028	990.0	3960	0.0524	0.2134	0.4268	0.4268	nan	0.4268	0.4268
0.0039	995.0	3980	0.0586	0.2223	0.4446	0.4446	nan	0.4446	0.4446
0.0055	1000.0	4000	0.0579	0.2161	0.4321	0.4321	nan	0.4321	0.4321

Framework versions

—Transformers 4.49.0
—Pytorch 2.6.0+cu124
—Datasets 3.4.1
—Tokenizers 0.21.1