mini_eurosat

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on a EuroSat dataset with 100 image in each class. It achieves the following results on the evaluation set:

• Train Loss: 0.2701
• Train Accuracy: 0.9158
• Validation Loss: 0.3930
• Validation Accuracy: 0.9233
• Epoch: 4

Model description

More information needed

Intended uses & limitations

This is just a demo for learning purpose and should not be used in productions

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• optimizer: {'name': 'AdamWeightDecay', 'learning_rate': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 3e-05, 'decay_steps': 1065, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}
• training_precision: float32

Training results

Train Loss	Train Accuracy	Validation Loss	Validation Accuracy	Epoch
1.6612	0.4653	1.0561	0.6964	0
0.7501	0.7761	0.6024	0.8248	1
0.4255	0.8559	0.4709	0.8784	2
0.3095	0.8941	0.3980	0.9063	3
0.2701	0.9158	0.3930	0.9233	4

Framework versions

• Transformers 4.38.2
• TensorFlow 2.15.0
• Datasets 2.18.0
• Tokenizers 0.15.2