Boya1_Adamax_1-e4_20Epoch_Deit-tiny-patch16_fold3

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: 3.5040
• Accuracy: 0.5695

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: 0.0001
• 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
1.3705	1.0	923	1.4925	0.4968
1.1741	2.0	1846	1.3247	0.5411
1.1089	3.0	2769	1.2524	0.5777
0.8912	4.0	3692	1.2699	0.5712
0.6118	5.0	4615	1.3695	0.5725
0.4514	6.0	5538	1.5162	0.5690
0.3342	7.0	6461	1.6732	0.5641
0.1558	8.0	7384	1.8402	0.5668
0.139	9.0	8307	2.0769	0.5676
0.0399	10.0	9230	2.4530	0.5582
0.0251	11.0	10153	2.6195	0.5630
0.0197	12.0	11076	2.8679	0.5598
0.0022	13.0	11999	3.0450	0.5593
0.0102	14.0	12922	3.1628	0.5614
0.0226	15.0	13845	3.2622	0.5655
0.0004	16.0	14768	3.3164	0.5668
0.0003	17.0	15691	3.3759	0.5703
0.0002	18.0	16614	3.4406	0.5687
0.0002	19.0	17537	3.4891	0.5695
0.0004	20.0	18460	3.5040	0.5695

Framework versions

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