swin-tiny-patch4-window7-224-finetuned-eurosat_animals

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

• Loss: 0.0772
• Accuracy: 0.9817

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: 128
• eval_batch_size: 128
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 512
• 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
2.1783	0.98	37	2.0846	0.1985
1.6593	1.99	75	1.2526	0.7228
0.47	2.99	113	0.2017	0.9494
0.2459	4.0	151	0.1004	0.9728
0.1891	4.98	188	0.0877	0.9785
0.1638	5.99	226	0.0666	0.9831
0.154	6.99	264	0.0693	0.9803
0.1542	8.0	302	0.0646	0.9822
0.1305	8.98	339	0.0663	0.9822
0.1337	9.99	377	0.0593	0.9841
0.129	10.99	415	0.0593	0.9836
0.1179	12.0	453	0.0598	0.9836
0.1025	12.98	490	0.0636	0.9813
0.0993	13.99	528	0.0637	0.9817
0.1043	14.99	566	0.0578	0.9827
0.0996	16.0	604	0.0644	0.9831
0.0866	16.98	641	0.0813	0.9785
0.0879	17.99	679	0.0734	0.9813
0.0812	18.99	717	0.0639	0.9855
0.0864	20.0	755	0.0619	0.9860
0.086	20.98	792	0.0693	0.9794
0.0781	21.99	830	0.0638	0.9831
0.0826	22.99	868	0.0681	0.9822
0.074	24.0	906	0.0687	0.9817
0.0828	24.98	943	0.0738	0.9836
0.0727	25.99	981	0.0655	0.9827
0.0692	26.99	1019	0.0713	0.9836
0.0696	28.0	1057	0.0729	0.9817
0.0792	28.98	1094	0.0707	0.9836
0.0657	29.99	1132	0.0647	0.9827
0.0747	30.99	1170	0.0769	0.9808
0.0861	32.0	1208	0.0665	0.9841
0.0693	32.98	1245	0.0617	0.9850
0.0682	33.99	1283	0.0636	0.9855
0.0615	34.99	1321	0.0685	0.9841
0.0581	36.0	1359	0.0702	0.9822
0.0713	36.98	1396	0.0675	0.9855
0.0593	37.99	1434	0.0697	0.9827
0.0543	38.99	1472	0.0701	0.9831
0.0628	40.0	1510	0.0720	0.9799
0.0606	40.98	1547	0.0794	0.9808
0.0619	41.99	1585	0.0720	0.9827
0.0612	42.99	1623	0.0768	0.9813
0.0435	44.0	1661	0.0748	0.9831
0.0614	44.98	1698	0.0738	0.9841
0.0544	45.99	1736	0.0750	0.9822
0.0569	46.99	1774	0.0802	0.9803
0.0527	48.0	1812	0.0772	0.9831
0.0535	48.98	1849	0.0724	0.9831
0.063	49.99	1887	0.0736	0.9831
0.0534	50.99	1925	0.0767	0.9822
0.0515	52.0	1963	0.0736	0.9817
0.0522	52.98	2000	0.0739	0.9827
0.0474	53.99	2038	0.0687	0.9831
0.0515	54.99	2076	0.0675	0.9846
0.0558	56.0	2114	0.0676	0.9822
0.0461	56.98	2151	0.0714	0.9813
0.0532	57.99	2189	0.0753	0.9803
0.0539	58.99	2227	0.0826	0.9803
0.0428	60.0	2265	0.0785	0.9827
0.0361	60.98	2302	0.0821	0.9813
0.0515	61.99	2340	0.0813	0.9817
0.047	62.99	2378	0.0817	0.9813
0.046	64.0	2416	0.0765	0.9827
0.039	64.98	2453	0.0794	0.9827
0.0399	65.99	2491	0.0805	0.9822
0.0478	66.99	2529	0.0758	0.9817
0.0426	68.0	2567	0.0726	0.9831
0.0383	68.98	2604	0.0785	0.9827
0.0407	69.99	2642	0.0795	0.9813
0.0413	70.99	2680	0.0774	0.9822
0.0428	72.0	2718	0.0765	0.9827
0.0375	72.98	2755	0.0730	0.9836
0.0428	73.99	2793	0.0715	0.9841
0.0443	74.99	2831	0.0744	0.9841
0.0383	76.0	2869	0.0748	0.9822
0.0338	76.98	2906	0.0818	0.9808
0.0445	77.99	2944	0.0759	0.9817
0.0374	78.99	2982	0.0757	0.9827
0.0404	80.0	3020	0.0793	0.9813
0.0336	80.98	3057	0.0750	0.9827
0.0364	81.99	3095	0.0816	0.9813
0.0403	82.99	3133	0.0795	0.9822
0.0287	84.0	3171	0.0818	0.9803
0.0425	84.98	3208	0.0819	0.9803
0.0446	85.99	3246	0.0775	0.9813
0.0341	86.99	3284	0.0772	0.9803
0.0414	88.0	3322	0.0757	0.9813
0.0401	88.98	3359	0.0751	0.9822
0.0442	89.99	3397	0.0775	0.9827
0.046	90.99	3435	0.0794	0.9808
0.0378	92.0	3473	0.0773	0.9822
0.0343	92.98	3510	0.0759	0.9817
0.0348	93.99	3548	0.0766	0.9822
0.0409	94.99	3586	0.0773	0.9817
0.0389	96.0	3624	0.0770	0.9822
0.0362	96.98	3661	0.0773	0.9817
0.0317	97.99	3699	0.0772	0.9817
0.0244	98.01	3700	0.0772	0.9817

Framework versions

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