smids_3x_deit_base_rms_001_fold4

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

• Loss: 0.6585
• Accuracy: 0.7683

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: 0.001
• 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: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.111	1.0	225	1.0986	0.34
0.9281	2.0	450	0.8238	0.535
0.8429	3.0	675	0.7715	0.5733
0.7817	4.0	900	0.9515	0.5183
0.8357	5.0	1125	0.7549	0.615
0.8459	6.0	1350	0.7437	0.6367
0.8489	7.0	1575	0.8096	0.6133
0.7642	8.0	1800	0.7237	0.6317
0.7593	9.0	2025	0.7493	0.6517
0.7519	10.0	2250	0.7058	0.6667
0.7342	11.0	2475	0.6982	0.6883
0.7801	12.0	2700	0.7132	0.6683
0.8045	13.0	2925	0.6860	0.6917
0.6831	14.0	3150	0.6911	0.68
0.695	15.0	3375	0.6849	0.6917
0.7337	16.0	3600	0.6867	0.6933
0.6786	17.0	3825	0.7826	0.65
0.7106	18.0	4050	0.7080	0.685
0.7255	19.0	4275	0.6333	0.7083
0.5816	20.0	4500	0.6578	0.71
0.5877	21.0	4725	0.6049	0.74
0.64	22.0	4950	0.6186	0.715
0.6573	23.0	5175	0.6447	0.7033
0.6339	24.0	5400	0.5686	0.77
0.593	25.0	5625	0.5930	0.7583
0.6542	26.0	5850	0.6005	0.74
0.605	27.0	6075	0.5866	0.75
0.5386	28.0	6300	0.5839	0.7417
0.5132	29.0	6525	0.5733	0.76
0.512	30.0	6750	0.5641	0.7617
0.5165	31.0	6975	0.5435	0.7783
0.526	32.0	7200	0.5563	0.765
0.4963	33.0	7425	0.6237	0.7433
0.5513	34.0	7650	0.5398	0.7733
0.5129	35.0	7875	0.5786	0.76
0.5453	36.0	8100	0.5387	0.765
0.5096	37.0	8325	0.5322	0.7833
0.4852	38.0	8550	0.5586	0.775
0.4638	39.0	8775	0.5742	0.755
0.5442	40.0	9000	0.5696	0.7617
0.4516	41.0	9225	0.5837	0.78
0.4465	42.0	9450	0.5694	0.7833
0.4118	43.0	9675	0.5778	0.775
0.436	44.0	9900	0.5795	0.7683
0.3828	45.0	10125	0.6308	0.7517
0.409	46.0	10350	0.5949	0.76
0.402	47.0	10575	0.6432	0.755
0.3606	48.0	10800	0.6291	0.7683
0.2902	49.0	11025	0.6518	0.7683
0.3362	50.0	11250	0.6585	0.7683

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.0+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2