model_0-D2-SW-Tuned

This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.1783
• Accuracy: 0.9889
• Precision: 0.9889
• Sensitivity: 0.9667
• Specificity: 0.9933
• F1: 0.9889
• Auc: 0.9991
• Mcc: 0.96
• J Stat: 0.96
• Confusion Matrix: [[149, 1], [1, 29]]

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: 4.286661471431606e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.07159086635599052
• num_epochs: 7
• label_smoothing_factor: 0.0713211794743841
• weight_decay: 0.010753387469303224

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Sensitivity	Specificity	F1	Auc	Mcc	J Stat	Confusion Matrix
0.2816	1.0	376	0.2185	0.9722	0.9727	0.9333	0.98	0.9724	0.9929	0.9015	0.9133	[[147, 3], [2, 28]]
0.2998	2.0	752	0.2313	0.9611	0.9607	0.8667	0.98	0.9608	0.9947	0.8583	0.8467	[[147, 3], [4, 26]]
0.1991	3.0	1128	0.2396	0.9667	0.9662	0.8667	0.9867	0.9662	0.9951	0.8775	0.8533	[[148, 2], [4, 26]]
0.2611	4.0	1504	0.1923	0.9833	0.9837	0.9667	0.9867	0.9834	0.998	0.9410	0.9533	[[148, 2], [1, 29]]
0.2044	5.0	1880	0.2105	0.9722	0.9741	0.9667	0.9733	0.9727	0.9989	0.9054	0.94	[[146, 4], [1, 29]]
0.207	6.0	2256	0.1922	0.9833	0.9832	0.9333	0.9933	0.9832	0.9991	0.9394	0.9267	[[149, 1], [2, 28]]
0.1677	7.0	2632	0.1783	0.9889	0.9889	0.9667	0.9933	0.9889	0.9991	0.96	0.96	[[149, 1], [1, 29]]

Framework versions

• Transformers 4.57.1
• Pytorch 2.6.0+cu124
• Datasets 4.4.1
• Tokenizers 0.22.1