Boya1_SGD_1e3_20Epoch_Swin-large-224_fold2

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

• Loss: 2.0270
• Accuracy: 0.3805

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: 0.001
• 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
2.4749	1.0	923	2.5397	0.1935
2.4343	2.0	1846	2.4492	0.1943
2.4893	3.0	2769	2.3908	0.1997
2.355	4.0	3692	2.3485	0.2208
2.2183	5.0	4615	2.3215	0.2484
2.2553	6.0	5538	2.2854	0.2741
2.2168	7.0	6461	2.2456	0.2970
2.2632	8.0	7384	2.2147	0.3143
2.1811	9.0	8307	2.1785	0.33
2.1273	10.0	9230	2.1568	0.3373
2.055	11.0	10153	2.1247	0.3449
2.0484	12.0	11076	2.1086	0.3549
2.0352	13.0	11999	2.0861	0.3638
2.0495	14.0	12922	2.0733	0.3646
2.094	15.0	13845	2.0546	0.3705
2.0506	16.0	14768	2.0465	0.3759
1.9902	17.0	15691	2.0401	0.3773
1.9714	18.0	16614	2.0324	0.3792
1.967	19.0	17537	2.0279	0.3808
2.1255	20.0	18460	2.0270	0.3805

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.1+cu121
• Datasets 2.21.0
• Tokenizers 0.13.2