Boya1_RMSProp_1-e5_20Epoch_swin-base-window7-224-in22k_fold5

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: 2.4142
• Accuracy: 0.6725

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
1.1865	1.0	924	1.1272	0.6132
0.9864	2.0	1848	1.0047	0.6555
0.7613	3.0	2772	0.9529	0.6720
0.5344	4.0	3696	0.9895	0.6780
0.5033	5.0	4620	1.0342	0.6685
0.5228	6.0	5544	1.0876	0.6709
0.4147	7.0	6468	1.1734	0.6690
0.2373	8.0	7392	1.3076	0.6617
0.2292	9.0	8316	1.4535	0.6579
0.1735	10.0	9240	1.5622	0.6625
0.1961	11.0	10164	1.6717	0.6663
0.1611	12.0	11088	1.8090	0.6706
0.0814	13.0	12012	1.9522	0.6633
0.0487	14.0	12936	2.0777	0.6649
0.0762	15.0	13860	2.1837	0.6628
0.1072	16.0	14784	2.2581	0.6663
0.0286	17.0	15708	2.3433	0.6671
0.0537	18.0	16632	2.3849	0.6679
0.0348	19.0	17556	2.4070	0.6739
0.0662	20.0	18480	2.4142	0.6725

Framework versions

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