Boya1_RMSProp_1-e5_20Epoch_swin-base-window7-224-in22k_fold1

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: 2.3289
• Accuracy: 0.6551

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
1.1406	1.0	924	1.1487	0.6011
0.8906	2.0	1848	1.0575	0.6472
0.9616	3.0	2772	0.9518	0.6657
0.6331	4.0	3696	1.0209	0.6611
0.6302	5.0	4620	1.0093	0.6749
0.5342	6.0	5544	1.1249	0.6608
0.408	7.0	6468	1.1207	0.6676
0.2471	8.0	7392	1.2523	0.6649
0.2109	9.0	8316	1.3584	0.6583
0.1506	10.0	9240	1.5006	0.6502
0.1883	11.0	10164	1.5942	0.6545
0.1463	12.0	11088	1.7342	0.6488
0.1095	13.0	12012	1.8730	0.6526
0.0854	14.0	12936	1.9291	0.6526
0.0397	15.0	13860	2.0830	0.6559
0.0432	16.0	14784	2.1495	0.6540
0.0261	17.0	15708	2.2222	0.6488
0.0435	18.0	16632	2.2862	0.6556
0.0983	19.0	17556	2.3223	0.6551
0.0738	20.0	18480	2.3289	0.6551

Framework versions

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