image_classification

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.1877
• Accuracy: 0.625

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: 5e-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
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	40	1.8317	0.2938
No log	2.0	80	1.5647	0.4437
No log	3.0	120	1.4497	0.4938
No log	4.0	160	1.3529	0.5188
No log	5.0	200	1.2883	0.5125
No log	6.0	240	1.2861	0.5125
No log	7.0	280	1.2655	0.55
No log	8.0	320	1.2890	0.5125
No log	9.0	360	1.1955	0.575
No log	10.0	400	1.2180	0.5687
No log	11.0	440	1.2835	0.55
No log	12.0	480	1.2838	0.5188
1.0368	13.0	520	1.2168	0.5875
1.0368	14.0	560	1.1713	0.6312
1.0368	15.0	600	1.2222	0.5875
1.0368	16.0	640	1.3160	0.5563
1.0368	17.0	680	1.2512	0.6125
1.0368	18.0	720	1.3575	0.5563
1.0368	19.0	760	1.3514	0.5375
1.0368	20.0	800	1.3472	0.5625
1.0368	21.0	840	1.3449	0.5375
1.0368	22.0	880	1.3783	0.5375
1.0368	23.0	920	1.3240	0.575
1.0368	24.0	960	1.3391	0.5687
0.2885	25.0	1000	1.3723	0.55

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3