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: 1.2727
• Accuracy: 0.5312

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
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• 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
2.0804	1.0	10	2.0714	0.1625
2.0428	2.0	20	2.0324	0.2313
1.9463	3.0	30	1.8978	0.3438
1.7768	4.0	40	1.7234	0.375
1.6163	5.0	50	1.6029	0.4188
1.509	6.0	60	1.5122	0.5
1.4118	7.0	70	1.4839	0.4375
1.3381	8.0	80	1.4268	0.475
1.2653	9.0	90	1.4095	0.4813
1.1979	10.0	100	1.3504	0.5375
1.1219	11.0	110	1.3293	0.4875
1.0858	12.0	120	1.3023	0.4875
1.0214	13.0	130	1.3063	0.5188
1.0085	14.0	140	1.3306	0.5312
0.9615	15.0	150	1.2838	0.5
0.9277	16.0	160	1.3073	0.5125
0.898	17.0	170	1.2606	0.5437
0.8747	18.0	180	1.3116	0.5437
0.8657	19.0	190	1.3171	0.5375
0.8462	20.0	200	1.2619	0.525

Framework versions

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