Boya1_RMSProp_1-e5_20Epoch_swin-base-window7-224-in22k_fold2

This model is a fine-tuned version of microsoft/swin-base-patch4-window7-224-in22k on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 2.4218
• Accuracy: 0.6454

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: 1e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• 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
1.1618	1.0	923	1.1839	0.5970
0.8976	2.0	1846	1.0699	0.6378
0.7736	3.0	2769	0.9583	0.6708
0.6954	4.0	3692	0.9868	0.6651
0.6308	5.0	4615	1.0373	0.6632
0.4596	6.0	5538	1.1537	0.6511
0.4024	7.0	6461	1.1814	0.6554
0.2437	8.0	7384	1.2764	0.65
0.2069	9.0	8307	1.4493	0.6457
0.1113	10.0	9230	1.5231	0.6497
0.1803	11.0	10153	1.6738	0.6414
0.1099	12.0	11076	1.7749	0.6473
0.1161	13.0	11999	1.9080	0.6473
0.1045	14.0	12922	2.0173	0.6505
0.085	15.0	13845	2.1608	0.6470
0.0137	16.0	14768	2.2375	0.6408
0.0385	17.0	15691	2.3465	0.6430
0.0121	18.0	16614	2.3696	0.6476
0.0316	19.0	17537	2.4233	0.6446
0.0511	20.0	18460	2.4218	0.6454

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0
• Datasets 2.14.6
• Tokenizers 0.14.1