imageclassification

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: 1.0955
• Accuracy: 0.6188

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	40	1.8595	0.3438
No log	2.0	80	1.5878	0.4062
No log	3.0	120	1.4659	0.4062
No log	4.0	160	1.3915	0.5
No log	5.0	200	1.2930	0.5563
No log	6.0	240	1.2829	0.5
No log	7.0	280	1.2438	0.525
No log	8.0	320	1.3132	0.5188
No log	9.0	360	1.3179	0.5125
No log	10.0	400	1.1849	0.575
No log	11.0	440	1.1756	0.5563
No log	12.0	480	1.1960	0.55
1.087	13.0	520	1.1721	0.575
1.087	14.0	560	1.1838	0.5875
1.087	15.0	600	1.1801	0.575
1.087	16.0	640	1.1278	0.6125
1.087	17.0	680	1.1835	0.5938
1.087	18.0	720	1.2308	0.525
1.087	19.0	760	1.2079	0.5875
1.087	20.0	800	1.1869	0.5813

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.0
• Tokenizers 0.15.2