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: 5.5031
• Accuracy: 0.0

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: 5e-05
• train_batch_size: 256
• eval_batch_size: 256
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 1024
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	1	0.6744	1.0
No log	2.0	2	0.7507	0.0
No log	3.0	3	0.9175	0.0
No log	4.0	4	1.1669	0.0
No log	5.0	5	1.4443	0.0
No log	6.0	6	1.7218	0.0
No log	7.0	7	2.0269	0.0
No log	8.0	8	2.3374	0.0
No log	9.0	9	2.6657	0.0
0.0781	10.0	10	2.9900	0.0
0.0781	11.0	11	3.2990	0.0
0.0781	12.0	12	3.5921	0.0
0.0781	13.0	13	3.8577	0.0
0.0781	14.0	14	4.1048	0.0
0.0781	15.0	15	4.3232	0.0
0.0781	16.0	16	4.5163	0.0
0.0781	17.0	17	4.6854	0.0
0.0781	18.0	18	4.8332	0.0
0.0781	19.0	19	4.9602	0.0
0.0003	20.0	20	5.0735	0.0
0.0003	21.0	21	5.1691	0.0
0.0003	22.0	22	5.2486	0.0
0.0003	23.0	23	5.3151	0.0
0.0003	24.0	24	5.3696	0.0
0.0003	25.0	25	5.4131	0.0
0.0003	26.0	26	5.4466	0.0
0.0003	27.0	27	5.4711	0.0
0.0003	28.0	28	5.4879	0.0
0.0003	29.0	29	5.4983	0.0
0.0	30.0	30	5.5031	0.0

Framework versions

• Transformers 4.33.3
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3