vit-base-patch16-224-finetuned-main-gpu-30e-final

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: 0.0231
• Accuracy: 0.9940

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
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.5113	1.0	551	0.4745	0.7971
0.3409	2.0	1102	0.2697	0.8961
0.2675	3.0	1653	0.1611	0.9381
0.2092	4.0	2204	0.1176	0.9548
0.2008	5.0	2755	0.0889	0.9656
0.1555	6.0	3306	0.0666	0.9759
0.1614	7.0	3857	0.0576	0.9778
0.1518	8.0	4408	0.0517	0.9814
0.1231	9.0	4959	0.0528	0.9812
0.1076	10.0	5510	0.0426	0.9850
0.0953	11.0	6061	0.0634	0.9795
0.1097	12.0	6612	0.0398	0.9860
0.0763	13.0	7163	0.0348	0.9866
0.0895	14.0	7714	0.0341	0.9884
0.06	15.0	8265	0.0381	0.9883
0.0767	16.0	8816	0.0382	0.9875
0.0868	17.0	9367	0.0309	0.9898
0.091	18.0	9918	0.0339	0.9885
0.0817	19.0	10469	0.0243	0.9913
0.0641	20.0	11020	0.0286	0.9906
0.0703	21.0	11571	0.0314	0.9906
0.0642	22.0	12122	0.0261	0.9913
0.0695	23.0	12673	0.0260	0.9920
0.0664	24.0	13224	0.0241	0.9928
0.0552	25.0	13775	0.0258	0.9928
0.056	26.0	14326	0.0230	0.9939
0.0488	27.0	14877	0.0221	0.9936
0.0389	28.0	15428	0.0225	0.9930
0.0402	29.0	15979	0.0231	0.9940
0.0424	30.0	16530	0.0211	0.9939

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2