image_classification

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

• Loss: 1.2259
• Accuracy: 0.5625

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: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: polynomial
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.8751	1.0	20	1.7512	0.3
1.3825	2.0	40	1.4946	0.425
1.1532	3.0	60	1.3387	0.45
0.9865	4.0	80	1.3469	0.4562
0.8767	5.0	100	1.2275	0.55
0.7586	6.0	120	1.2560	0.5062
0.5985	7.0	140	1.2596	0.5062
0.5052	8.0	160	1.3010	0.5687
0.4243	9.0	180	1.2613	0.5563
0.387	10.0	200	1.2750	0.5312
0.3529	11.0	220	1.3103	0.55
0.218	12.0	240	1.1832	0.55
0.2428	13.0	260	1.2527	0.5563
0.2399	14.0	280	1.4836	0.5375
0.218	15.0	300	1.4056	0.4875
0.1784	16.0	320	1.3879	0.5563
0.2021	17.0	340	1.4346	0.5375
0.1342	18.0	360	1.4666	0.4813
0.1499	19.0	380	1.4104	0.5687
0.1032	20.0	400	1.5402	0.525
0.1214	21.0	420	1.4114	0.55
0.153	22.0	440	1.5887	0.525
0.1276	23.0	460	1.4588	0.5188
0.1114	24.0	480	1.4866	0.5312
0.1305	25.0	500	1.4203	0.5687

Framework versions

• Transformers 4.37.2
• Pytorch 2.1.0+cu121
• Datasets 2.17.0
• Tokenizers 0.15.1