Boya1_SGD_1-e3_20Epoch_Deit-tiny-patch16_fold1

This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.8160
• Accuracy: 0.3951

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
2.4438	1.0	924	2.4927	0.1898
2.3969	2.0	1848	2.3384	0.2389
2.2609	3.0	2772	2.2168	0.2878
2.0421	4.0	3696	2.1285	0.3068
2.0227	5.0	4620	2.0634	0.3296
1.99	6.0	5544	2.0084	0.3397
1.9954	7.0	6468	1.9664	0.3549
2.0727	8.0	7392	1.9354	0.3652
2.0158	9.0	8316	1.9072	0.3704
1.8488	10.0	9240	1.8880	0.3750
1.8985	11.0	10164	1.8721	0.3790
1.7309	12.0	11088	1.8576	0.3812
1.8129	13.0	12012	1.8465	0.3899
1.7599	14.0	12936	1.8384	0.3866
1.7902	15.0	13860	1.8309	0.3894
1.7502	16.0	14784	1.8250	0.3932
1.7034	17.0	15708	1.8221	0.3934
1.8587	18.0	16632	1.8187	0.3940
1.8137	19.0	17556	1.8165	0.3942
1.9039	20.0	18480	1.8160	0.3951

Framework versions

• Transformers 4.40.1
• Pytorch 2.1.0
• Datasets 2.19.0
• Tokenizers 0.19.1