swin-tiny-patch4-window7-224-finetuned-eurosat

This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.4560
• Accuracy: 0.8406

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: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.7738	1.0	13	1.3796	0.5845
1.2319	2.0	26	1.1322	0.6039
1.0355	3.0	39	0.9885	0.6232
0.7439	4.0	52	1.2022	0.6232
0.6792	5.0	65	0.7238	0.7246
0.6195	6.0	78	0.7041	0.7536
0.5151	7.0	91	0.6132	0.7826
0.556	8.0	104	0.6381	0.7488
0.4727	9.0	117	0.6127	0.7923
0.4879	10.0	130	0.4921	0.8551
0.436	11.0	143	0.5578	0.7923
0.3781	12.0	156	0.5095	0.8261
0.4201	13.0	169	0.5151	0.8454
0.3773	14.0	182	0.4612	0.8261
0.3611	15.0	195	0.5384	0.7971
0.3855	16.0	208	0.5267	0.8261
0.3926	17.0	221	0.4100	0.8647
0.3513	18.0	234	0.4508	0.8454
0.3389	19.0	247	0.4420	0.8502
0.3232	20.0	260	0.4560	0.8406

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.19.2
• Tokenizers 0.19.1