Boya1_Adamax_1-e4_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: 3.5356
• Accuracy: 0.5569

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
1.3865	1.0	924	1.5601	0.4768
1.2182	2.0	1848	1.4517	0.4963
1.2424	3.0	2772	1.3040	0.5531
0.8291	4.0	3696	1.3092	0.5745
0.6764	5.0	4620	1.3977	0.5724
0.5779	6.0	5544	1.5087	0.5601
0.3166	7.0	6468	1.7036	0.5577
0.2404	8.0	7392	1.9068	0.5528
0.144	9.0	8316	2.1442	0.5547
0.165	10.0	9240	2.4839	0.5509
0.0646	11.0	10164	2.7042	0.5490
0.0029	12.0	11088	2.9034	0.5523
0.0317	13.0	12012	3.1091	0.5504
0.0012	14.0	12936	3.2476	0.5496
0.0008	15.0	13860	3.3162	0.5569
0.0005	16.0	14784	3.3879	0.5525
0.0003	17.0	15708	3.4370	0.5517
0.0007	18.0	16632	3.4907	0.5542
0.0003	19.0	17556	3.5171	0.5566
0.0003	20.0	18480	3.5356	0.5569

Framework versions

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