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: 1.2066
• Accuracy: 0.5813

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: 2
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine_with_restarts
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
2.0741	1.0	20	2.0298	0.2687
1.9068	2.0	40	1.7590	0.425
1.6486	3.0	60	1.5578	0.4688
1.4978	4.0	80	1.4362	0.5375
1.3643	5.0	100	1.3577	0.5312
1.2724	6.0	120	1.3503	0.5437
1.1678	7.0	140	1.2626	0.575
1.074	8.0	160	1.2404	0.5813
1.0216	9.0	180	1.2679	0.5375
0.943	10.0	200	1.1997	0.6
0.9146	11.0	220	1.1864	0.5938
0.8716	12.0	240	1.2533	0.5437
0.8739	13.0	260	1.1740	0.5625
0.8903	14.0	280	1.2089	0.55
0.8424	15.0	300	1.2022	0.5625

Framework versions

• Transformers 4.42.4
• Pytorch 2.4.0+cu121
• Datasets 2.21.0
• Tokenizers 0.19.1