beit-base-patch16-224-pt22k-finetuned-eurosat

This model is a fine-tuned version of microsoft/beit-base-patch16-224-pt22k on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9338
• Accuracy: 0.6667

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
No log	1.0	3	1.3657	0.3095
No log	2.0	6	1.1966	0.4286
No log	3.0	9	1.1076	0.4524
1.2696	4.0	12	1.0717	0.5714
1.2696	5.0	15	0.9948	0.5238
1.2696	6.0	18	1.0701	0.5
1.0945	7.0	21	0.9920	0.5
1.0945	8.0	24	0.9338	0.6667
1.0945	9.0	27	0.9605	0.5714
0.9538	10.0	30	0.9285	0.6190
0.9538	11.0	33	0.9113	0.5714
0.9538	12.0	36	0.8414	0.6190
0.9538	13.0	39	0.9422	0.5476
0.8646	14.0	42	0.8165	0.6429
0.8646	15.0	45	0.9582	0.5238
0.8646	16.0	48	0.8548	0.6190
0.8082	17.0	51	0.8568	0.6190
0.8082	18.0	54	0.8792	0.5476
0.8082	19.0	57	0.8819	0.5476
0.7731	20.0	60	0.8454	0.5714

Framework versions

• Transformers 4.30.0
• Pytorch 2.2.1+cu121
• Datasets 2.18.0
• Tokenizers 0.13.3