swin-tiny-patch4-window7-224-finetuned-eurosat_DATA7_20240410

This model was trained from scratch on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.0817
• Accuracy: 0.9731

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: 5e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 256
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.98	1.0	182	1.0130	0.4827
0.843	2.0	365	0.8658	0.5978
0.742	3.0	548	0.7068	0.6848
0.69	4.0	731	0.6297	0.7254
0.6413	5.0	913	0.5716	0.7571
0.611	6.0	1096	0.5445	0.7671
0.5844	7.0	1279	0.5181	0.7802
0.5616	8.0	1462	0.4757	0.7985
0.5634	9.0	1644	0.5245	0.7896
0.5223	10.0	1827	0.4991	0.7902
0.4723	11.0	2010	0.4363	0.8216
0.4443	12.0	2193	0.3813	0.8403
0.4538	13.0	2375	0.3500	0.8574
0.4273	14.0	2558	0.3326	0.8624
0.4247	15.0	2741	0.3224	0.8684
0.4063	16.0	2924	0.3096	0.8707
0.367	17.0	3106	0.2713	0.8945
0.3605	18.0	3289	0.3160	0.8755
0.3475	19.0	3472	0.2559	0.8995
0.3262	20.0	3655	0.2437	0.9030
0.3218	21.0	3837	0.2343	0.9090
0.3125	22.0	4020	0.2267	0.9113
0.336	23.0	4203	0.2170	0.9138
0.2813	24.0	4386	0.2062	0.9199
0.2802	25.0	4568	0.1956	0.9196
0.2996	26.0	4751	0.1923	0.9244
0.2699	27.0	4934	0.1934	0.9273
0.2642	28.0	5117	0.1973	0.9242
0.2491	29.0	5299	0.1686	0.9394
0.2611	30.0	5482	0.1793	0.9326
0.2383	31.0	5665	0.1744	0.9332
0.2338	32.0	5848	0.1537	0.9448
0.2225	33.0	6030	0.1569	0.9405
0.2383	34.0	6213	0.1422	0.9480
0.2253	35.0	6396	0.1413	0.9455
0.2257	36.0	6579	0.1535	0.9442
0.2308	37.0	6761	0.1655	0.9423
0.2241	38.0	6944	0.1272	0.9530
0.2253	39.0	7127	0.1464	0.9440
0.1996	40.0	7310	0.1332	0.9527
0.225	41.0	7492	0.1311	0.9530
0.1918	42.0	7675	0.1546	0.9459
0.1937	43.0	7858	0.1388	0.9515
0.2043	44.0	8041	0.1185	0.9596
0.1802	45.0	8223	0.1195	0.9557
0.1821	46.0	8406	0.1152	0.9604
0.1712	47.0	8589	0.1273	0.9575
0.1865	48.0	8772	0.1209	0.9565
0.1706	49.0	8954	0.1057	0.9611
0.1817	50.0	9137	0.1114	0.9602
0.1753	51.0	9320	0.1114	0.9621
0.1826	52.0	9503	0.1055	0.9634
0.178	53.0	9685	0.1069	0.9644
0.1522	54.0	9868	0.1036	0.9625
0.171	55.0	10051	0.1068	0.9619
0.1656	56.0	10234	0.0925	0.9644
0.163	57.0	10416	0.0939	0.9677
0.1631	58.0	10599	0.1130	0.9592
0.169	59.0	10782	0.0907	0.9671
0.1491	60.0	10965	0.1055	0.9632
0.1572	61.0	11147	0.0940	0.9638
0.1617	62.0	11330	0.1008	0.9636
0.1584	63.0	11513	0.0989	0.9673
0.1597	64.0	11696	0.1026	0.9648
0.1466	65.0	11878	0.1008	0.9665
0.1468	66.0	12061	0.0947	0.9644
0.1562	67.0	12244	0.0864	0.9707
0.1589	68.0	12427	0.0980	0.9656
0.1505	69.0	12609	0.0908	0.9681
0.1497	70.0	12792	0.0879	0.9690
0.1362	71.0	12975	0.0864	0.9700
0.1418	72.0	13158	0.0949	0.9684
0.1345	73.0	13340	0.0994	0.9681
0.1333	74.0	13523	0.0859	0.9700
0.1414	75.0	13706	0.0912	0.9692
0.137	76.0	13889	0.0863	0.9719
0.1326	77.0	14071	0.0811	0.9707
0.1429	78.0	14254	0.0875	0.9690
0.1363	79.0	14437	0.0909	0.9690
0.1344	80.0	14620	0.0913	0.9692
0.1221	81.0	14802	0.0908	0.9706
0.1192	82.0	14985	0.0835	0.9715
0.1252	83.0	15168	0.0865	0.9711
0.1404	84.0	15351	0.0922	0.9700
0.124	85.0	15533	0.0845	0.9700
0.1278	86.0	15716	0.0859	0.9721
0.1271	87.0	15899	0.0835	0.9725
0.1254	88.0	16082	0.0843	0.9721
0.1363	89.0	16264	0.0852	0.9707
0.1144	90.0	16447	0.0846	0.9729
0.1217	91.0	16630	0.0822	0.9729
0.1185	92.0	16813	0.0818	0.9731
0.1095	93.0	16995	0.0825	0.9725
0.1181	94.0	17178	0.0811	0.9729
0.1191	95.0	17361	0.0839	0.9736
0.1107	96.0	17544	0.0825	0.9729
0.1093	97.0	17726	0.0825	0.9734
0.1187	98.0	17909	0.0811	0.9731
0.1302	99.0	18092	0.0823	0.9727
0.1146	99.59	18200	0.0817	0.9731

Framework versions

• Transformers 4.38.1
• Pytorch 2.1.2+cu121
• Datasets 2.17.1
• Tokenizers 0.15.2