YKXBCi/vit-base-patch16-224-in21k-ucSat

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on an unknown dataset. It achieves the following results on the evaluation set:

• Train Loss: 1.3216
• Train Accuracy: 0.9960
• Train Top-3-accuracy: 1.0
• Validation Loss: 1.3683
• Validation Accuracy: 0.9688
• Validation Top-3-accuracy: 0.9931
• Epoch: 4

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:

• optimizer: {'inner_optimizer': {'class_name': 'AdamWeightDecay', 'config': {'name': 'AdamWeightDecay', 'learning_rate': {'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 3e-05, 'decay_steps': 275, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.01}}, 'dynamic': True, 'initial_scale': 32768.0, 'dynamic_growth_steps': 2000}
• training_precision: mixed_float16

Training results

Train Loss	Train Accuracy	Train Top-3-accuracy	Validation Loss	Validation Accuracy	Validation Top-3-accuracy	Epoch
2.7376	0.5375	0.7284	2.3789	0.8958	0.9757	0
2.1030	0.9449	0.9972	1.8664	0.9479	0.9896	1
1.6719	0.9812	1.0	1.5763	0.9618	0.9931	2
1.4357	0.9926	1.0	1.4201	0.9688	0.9931	3
1.3216	0.9960	1.0	1.3683	0.9688	0.9931	4

Framework versions

• Transformers 4.18.0
• TensorFlow 2.6.0
• Datasets 2.1.0
• Tokenizers 0.12.1