segformer-b0-finetuned-oldapp-oct-1

This model is a fine-tuned version of nvidia/mit-b0 on the PushkarA07/oldapptiles5 dataset. It achieves the following results on the evaluation set:

• Loss: 0.0968
• Mean Iou: 0.9990
• Mean Accuracy: 1.0
• Overall Accuracy: 1.0

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: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Mean Iou	Mean Accuracy	Overall Accuracy
0.6574	0.7143	10	0.6497	0.9990	1.0	1.0
0.6005	1.4286	20	0.5761	0.9990	1.0	1.0
0.5676	2.1429	30	0.4740	0.9990	1.0	1.0
0.6287	2.8571	40	0.4394	0.9990	1.0	1.0
0.5121	3.5714	50	0.4173	0.9990	1.0	1.0
0.4744	4.2857	60	0.3842	0.9990	1.0	1.0
0.4413	5.0	70	0.4107	0.9990	1.0	1.0
0.4134	5.7143	80	0.3737	0.9990	1.0	1.0
0.4139	6.4286	90	0.3424	0.9990	1.0	1.0
0.4097	7.1429	100	0.3248	0.9990	1.0	1.0
0.3645	7.8571	110	0.3218	0.9990	1.0	1.0
0.3287	8.5714	120	0.2928	0.9990	1.0	1.0
0.3113	9.2857	130	0.3021	0.9990	1.0	1.0
0.3085	10.0	140	0.2962	0.9990	1.0	1.0
0.2879	10.7143	150	0.2596	0.9990	1.0	1.0
0.2958	11.4286	160	0.2409	0.9990	1.0	1.0
0.2788	12.1429	170	0.2396	0.9990	1.0	1.0
0.2565	12.8571	180	0.2076	0.9990	1.0	1.0
0.2457	13.5714	190	0.2184	0.9990	1.0	1.0
0.2328	14.2857	200	0.1962	0.9990	1.0	1.0
0.1916	15.0	210	0.2003	0.9990	1.0	1.0
0.3277	15.7143	220	0.1875	0.9990	1.0	1.0
0.2053	16.4286	230	0.1718	0.9990	1.0	1.0
0.2555	17.1429	240	0.1571	0.9990	1.0	1.0
0.1863	17.8571	250	0.1546	0.9990	1.0	1.0
0.1944	18.5714	260	0.1503	0.9990	1.0	1.0
0.2652	19.2857	270	0.1456	0.9990	1.0	1.0
0.1614	20.0	280	0.1442	0.9990	1.0	1.0
0.139	20.7143	290	0.1413	0.9990	1.0	1.0
0.1631	21.4286	300	0.1308	0.9990	1.0	1.0
0.1988	22.1429	310	0.1256	0.9990	1.0	1.0
0.1294	22.8571	320	0.1190	0.9990	1.0	1.0
0.1174	23.5714	330	0.1185	0.9990	1.0	1.0
0.1287	24.2857	340	0.1251	0.9990	1.0	1.0
0.1322	25.0	350	0.1308	0.9990	1.0	1.0
0.1667	25.7143	360	0.1215	0.9990	1.0	1.0
0.1095	26.4286	370	0.1226	0.9990	1.0	1.0
0.1992	27.1429	380	0.1331	0.9990	1.0	1.0
0.1987	27.8571	390	0.1174	0.9990	1.0	1.0
0.1587	28.5714	400	0.1162	0.9990	1.0	1.0
0.1043	29.2857	410	0.1161	0.9990	1.0	1.0
0.1073	30.0	420	0.1112	0.9990	1.0	1.0
0.14	30.7143	430	0.1279	0.9990	1.0	1.0
0.1183	31.4286	440	0.1361	0.9990	1.0	1.0
0.1096	32.1429	450	0.1430	0.9990	1.0	1.0
0.0957	32.8571	460	0.1397	0.9990	1.0	1.0
0.1605	33.5714	470	0.1436	0.9990	1.0	1.0
0.0837	34.2857	480	0.1163	0.9990	1.0	1.0
0.1032	35.0	490	0.1223	0.9990	1.0	1.0
0.1815	35.7143	500	0.0829	0.9990	1.0	1.0
0.0762	36.4286	510	0.1128	0.9990	1.0	1.0
0.0754	37.1429	520	0.1676	0.9990	1.0	1.0
0.0883	37.8571	530	0.1639	0.9990	1.0	1.0
0.0721	38.5714	540	0.1843	0.9990	1.0	1.0
0.0703	39.2857	550	0.1493	0.9990	1.0	1.0
0.0766	40.0	560	0.1616	0.9990	1.0	1.0
0.0636	40.7143	570	0.1292	0.9990	1.0	1.0
0.0771	41.4286	580	0.1087	0.9990	1.0	1.0
0.1019	42.1429	590	0.1540	0.9990	1.0	1.0
0.0734	42.8571	600	0.1639	0.9990	1.0	1.0
0.0504	43.5714	610	0.1544	0.9990	1.0	1.0
0.0606	44.2857	620	0.1403	0.9990	1.0	1.0
0.0925	45.0	630	0.1664	0.9990	1.0	1.0
0.0584	45.7143	640	0.1589	0.9990	1.0	1.0
0.0662	46.4286	650	0.1696	0.9990	1.0	1.0
0.0537	47.1429	660	0.1487	0.9990	1.0	1.0
0.0772	47.8571	670	0.1688	0.9990	1.0	1.0
0.0529	48.5714	680	0.1637	0.9990	1.0	1.0
0.0538	49.2857	690	0.1573	0.9990	1.0	1.0
0.045	50.0	700	0.1661	0.9990	1.0	1.0
0.0588	50.7143	710	0.1824	0.9990	1.0	1.0
0.0482	51.4286	720	0.1653	0.9990	1.0	1.0
0.0811	52.1429	730	0.1579	0.9990	1.0	1.0
-0.0544	52.8571	740	0.1355	0.9990	1.0	1.0
0.0463	53.5714	750	0.1514	0.9990	1.0	1.0
0.0465	54.2857	760	0.1259	0.9990	1.0	1.0
0.0798	55.0	770	0.1504	0.9990	1.0	1.0
-0.042	55.7143	780	0.1638	0.9990	1.0	1.0
-0.2192	56.4286	790	0.1666	0.9990	1.0	1.0
0.0365	57.1429	800	0.1834	0.9990	1.0	1.0
0.0765	57.8571	810	0.1456	0.9990	1.0	1.0
0.0552	58.5714	820	0.1491	0.9990	1.0	1.0
0.0375	59.2857	830	0.1515	0.9990	1.0	1.0
0.0499	60.0	840	0.1082	0.9990	1.0	1.0
0.0787	60.7143	850	0.1422	0.9990	1.0	1.0
0.0562	61.4286	860	0.1337	0.9990	1.0	1.0
0.0623	62.1429	870	0.1399	0.9990	1.0	1.0
0.0966	62.8571	880	0.1412	0.9990	1.0	1.0
0.0811	63.5714	890	0.1311	0.9990	1.0	1.0
0.0496	64.2857	900	0.1591	0.9990	1.0	1.0
0.0447	65.0	910	0.1587	0.9990	1.0	1.0
0.0345	65.7143	920	0.1637	0.9990	1.0	1.0
0.0637	66.4286	930	0.1427	0.9990	1.0	1.0
0.0644	67.1429	940	0.1611	0.9990	1.0	1.0
0.0779	67.8571	950	0.1566	0.9990	1.0	1.0
0.0417	68.5714	960	0.1488	0.9990	1.0	1.0
0.0969	69.2857	970	0.1577	0.9990	1.0	1.0
0.0452	70.0	980	0.1166	0.9990	1.0	1.0
0.0373	70.7143	990	0.1429	0.9990	1.0	1.0
0.0438	71.4286	1000	0.1425	0.9990	1.0	1.0
0.0606	72.1429	1010	0.1238	0.9990	1.0	1.0
0.0389	72.8571	1020	0.1284	0.9990	1.0	1.0
0.0402	73.5714	1030	0.1350	0.9990	1.0	1.0
0.0349	74.2857	1040	0.1583	0.9990	1.0	1.0
0.031	75.0	1050	0.1563	0.9990	1.0	1.0
0.0501	75.7143	1060	0.1501	0.9990	1.0	1.0
0.0412	76.4286	1070	0.1417	0.9990	1.0	1.0
0.0532	77.1429	1080	0.1456	0.9990	1.0	1.0
0.0378	77.8571	1090	0.1059	0.9990	1.0	1.0
-0.3927	78.5714	1100	0.1200	0.9990	1.0	1.0
0.0499	79.2857	1110	0.1396	0.9990	1.0	1.0
0.0501	80.0	1120	0.1277	0.9990	1.0	1.0
0.0408	80.7143	1130	0.1494	0.9990	1.0	1.0
0.0369	81.4286	1140	0.1394	0.9990	1.0	1.0
0.0014	82.1429	1150	0.1306	0.9990	1.0	1.0
0.0359	82.8571	1160	0.1557	0.9990	1.0	1.0
-0.4227	83.5714	1170	0.1380	0.9990	1.0	1.0
0.0307	84.2857	1180	0.1351	0.9990	1.0	1.0
0.0433	85.0	1190	0.1379	0.9990	1.0	1.0
0.0407	85.7143	1200	0.1346	0.9990	1.0	1.0
0.0247	86.4286	1210	0.1572	0.9990	1.0	1.0
0.0498	87.1429	1220	0.1398	0.9990	1.0	1.0
0.0399	87.8571	1230	0.1261	0.9990	1.0	1.0
0.0354	88.5714	1240	0.0936	0.9990	1.0	1.0
0.0336	89.2857	1250	0.1343	0.9990	1.0	1.0
0.0291	90.0	1260	0.1410	0.9990	1.0	1.0
0.0379	90.7143	1270	0.1520	0.9990	1.0	1.0
0.0331	91.4286	1280	0.1490	0.9990	1.0	1.0
0.034	92.1429	1290	0.1449	0.9990	1.0	1.0
0.0315	92.8571	1300	0.1447	0.9990	1.0	1.0
0.0474	93.5714	1310	0.0865	0.9990	1.0	1.0
0.0295	94.2857	1320	0.1292	0.9990	1.0	1.0
0.0347	95.0	1330	0.1097	0.9990	1.0	1.0
0.0315	95.7143	1340	0.1264	0.9990	1.0	1.0
0.0427	96.4286	1350	0.1458	0.9990	1.0	1.0
0.0072	97.1429	1360	0.1381	0.9990	1.0	1.0
0.0481	97.8571	1370	0.1120	0.9990	1.0	1.0
0.0312	98.5714	1380	0.1331	0.9990	1.0	1.0
0.0617	99.2857	1390	0.1368	0.9990	1.0	1.0
-0.4803	100.0	1400	0.0968	0.9990	1.0	1.0

Framework versions

• Transformers 4.44.2
• Pytorch 2.5.0+cu121
• Datasets 3.1.0
• Tokenizers 0.19.1