swinv2-small-patch4-window16-256

This model is a fine-tuned version of microsoft/swinv2-small-patch4-window16-256 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2761
• Accuracy: 1.0
• Precision: 1.0
• Sensitivity: 1.0
• Specificity: 1.0
• F1: 1.0
• Auc: 1.0
• Mcc: 1.0
• J Stat: 1.0
• Confusion Matrix: [[150, 0], [0, 30]]

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: 6.197470303061579e-05
• train_batch_size: 8
• eval_batch_size: 8
• 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: 0.006309410963953089
• num_epochs: 10
• label_smoothing_factor: 0.15696760193504863

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Sensitivity	Specificity	F1	Auc	Mcc	J Stat	Confusion Matrix
0.3854	1.0	188	0.3930	0.9261	0.9342	0.9462	0.9191	0.9279	0.9713	0.8235	0.8652	[[1022, 90], [21, 369]]
0.3667	2.0	376	0.3353	0.9660	0.9665	0.8872	0.9937	0.9655	0.9881	0.9109	0.8809	[[1105, 7], [44, 346]]
0.3275	3.0	564	0.3080	0.9827	0.9827	0.9513	0.9937	0.9826	0.9939	0.9547	0.9450	[[1105, 7], [19, 371]]
0.3221	4.0	752	0.3034	0.9847	0.9848	0.9487	0.9973	0.9846	0.9963	0.9600	0.9460	[[1109, 3], [20, 370]]
0.3039	5.0	940	0.2990	0.9860	0.9860	0.9590	0.9955	0.9860	0.9985	0.9635	0.9545	[[1107, 5], [16, 374]]
0.3032	6.0	1128	0.2912	0.9907	0.9907	0.9744	0.9964	0.9907	0.9995	0.9757	0.9708	[[1108, 4], [10, 380]]
0.2804	7.0	1316	0.2882	0.9913	0.9914	0.9744	0.9973	0.9913	0.9999	0.9774	0.9717	[[1109, 3], [10, 380]]
0.2941	8.0	1504	0.2824	0.9947	0.9947	0.9821	0.9991	0.9947	0.9999	0.9861	0.9812	[[1111, 1], [7, 383]]
0.2904	9.0	1692	0.2797	0.9960	0.9960	0.9872	0.9991	0.9960	1.0000	0.9896	0.9863	[[1111, 1], [5, 385]]
0.2754	10.0	1880	0.2788	0.9967	0.9967	0.9872	1.0	0.9967	1.0000	0.9913	0.9872	[[1112, 0], [5, 385]]

Framework versions

• Transformers 4.52.4
• Pytorch 2.6.0+cu124
• Datasets 4.0.0
• Tokenizers 0.21.2