deit-small-patch16-224-finetuned-eurosat

This model is a fine-tuned version of facebook/deit-small-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.6817
• Accuracy: 0.7978

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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.5174	0.9966	218	1.3672	0.5855
1.282	1.9977	437	1.1843	0.6260
1.117	2.9989	656	1.0301	0.6845
1.0176	4.0	875	0.9670	0.7070
0.9912	4.9966	1093	0.8551	0.7477
0.9458	5.9977	1312	0.8534	0.7392
0.8502	6.9989	1531	0.8049	0.7600
0.8954	8.0	1750	0.7716	0.7683
0.872	8.9966	1968	0.7443	0.7779
0.8186	9.9977	2187	0.7304	0.7835
0.747	10.9989	2406	0.7178	0.7911
0.6843	12.0	2625	0.7062	0.7925
0.7453	12.9966	2843	0.7031	0.7939
0.7472	13.9977	3062	0.6891	0.7965
0.7067	14.9486	3270	0.6817	0.7978

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1