smids_10x_deit_base_sgd_00001_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: 1.0053
• Accuracy: 0.5833

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: 1e-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: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.1184	1.0	750	1.1087	0.3417
1.0954	2.0	1500	1.1049	0.3567
1.0943	3.0	2250	1.1012	0.3733
1.0965	4.0	3000	1.0976	0.3883
1.097	5.0	3750	1.0941	0.3983
1.0859	6.0	4500	1.0907	0.41
1.0772	7.0	5250	1.0874	0.4133
1.0913	8.0	6000	1.0841	0.43
1.0722	9.0	6750	1.0809	0.4383
1.0755	10.0	7500	1.0778	0.445
1.0725	11.0	8250	1.0746	0.4467
1.0546	12.0	9000	1.0716	0.4517
1.0625	13.0	9750	1.0686	0.46
1.0581	14.0	10500	1.0656	0.4683
1.047	15.0	11250	1.0626	0.4733
1.0434	16.0	12000	1.0597	0.48
1.0251	17.0	12750	1.0568	0.4917
1.0453	18.0	13500	1.0540	0.4933
1.0513	19.0	14250	1.0512	0.5
1.0388	20.0	15000	1.0484	0.51
1.0333	21.0	15750	1.0457	0.52
1.0252	22.0	16500	1.0431	0.5267
1.0295	23.0	17250	1.0405	0.5283
1.0271	24.0	18000	1.0379	0.5317
1.0317	25.0	18750	1.0355	0.5333
1.0277	26.0	19500	1.0331	0.5367
1.0061	27.0	20250	1.0308	0.5467
1.0064	28.0	21000	1.0286	0.555
1.0287	29.0	21750	1.0265	0.56
1.0101	30.0	22500	1.0245	0.5633
1.0026	31.0	23250	1.0225	0.5667
1.0037	32.0	24000	1.0207	0.5667
1.0192	33.0	24750	1.0189	0.57
1.0129	34.0	25500	1.0173	0.57
1.007	35.0	26250	1.0157	0.57
0.9958	36.0	27000	1.0143	0.5717
1.0225	37.0	27750	1.0130	0.5733
1.0064	38.0	28500	1.0118	0.575
0.99	39.0	29250	1.0107	0.5733
0.9987	40.0	30000	1.0097	0.5767
1.0146	41.0	30750	1.0088	0.5817
0.9734	42.0	31500	1.0080	0.5817
1.0086	43.0	32250	1.0073	0.5817
0.9898	44.0	33000	1.0068	0.5833
0.9877	45.0	33750	1.0063	0.5833
0.9974	46.0	34500	1.0059	0.5833
0.9888	47.0	35250	1.0057	0.5833
0.9987	48.0	36000	1.0055	0.5833
1.01	49.0	36750	1.0054	0.5833
0.9819	50.0	37500	1.0053	0.5833

Framework versions

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