vit_itri_46

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

• Loss: 0.0480
• Accuracy: 0.9910
• Precision: 0.9911
• Recall: 0.9910
• F1: 0.9907

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: 0.0001
• train_batch_size: 24
• eval_batch_size: 4
• seed: 42
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.3639	1.0	708	0.2151	0.9401	0.9413	0.9401	0.9345
0.1373	2.0	1416	0.1163	0.9602	0.9612	0.9602	0.9590
0.0879	3.0	2124	0.0978	0.9735	0.9731	0.9735	0.9728
0.0639	4.0	2832	0.1333	0.9650	0.9657	0.9650	0.9637
0.053	5.0	3540	0.0817	0.9777	0.9771	0.9777	0.9766
0.0391	6.0	4248	0.0876	0.9756	0.9775	0.9756	0.9758
0.0284	7.0	4956	0.0579	0.9857	0.9858	0.9857	0.9852
0.0198	8.0	5664	0.0821	0.9820	0.9827	0.9820	0.9818
0.0176	9.0	6372	0.0717	0.9862	0.9862	0.9862	0.9859
0.0143	10.0	7080	0.1047	0.9820	0.9816	0.9820	0.9810
0.0135	11.0	7788	0.0724	0.9862	0.9864	0.9862	0.9860
0.0093	12.0	8496	0.0869	0.9857	0.9858	0.9857	0.9852
0.0067	13.0	9204	0.0863	0.9846	0.9858	0.9846	0.9840
0.0059	14.0	9912	0.0603	0.9867	0.9869	0.9867	0.9865
0.0038	15.0	10620	0.0591	0.9889	0.9891	0.9889	0.9886
0.003	16.0	11328	0.0568	0.9894	0.9895	0.9894	0.9891
0.0029	17.0	12036	0.0589	0.9873	0.9873	0.9873	0.9869
0.0005	18.0	12744	0.0530	0.9910	0.9911	0.9910	0.9907
0.0003	19.0	13452	0.0453	0.9920	0.9921	0.9920	0.9918
0.0002	20.0	14160	0.0480	0.9910	0.9911	0.9910	0.9907

Framework versions

• Transformers 4.53.0.dev0
• Pytorch 2.7.1+cu126
• Datasets 3.6.0
• Tokenizers 0.21.1