smids_5x_deit_small_sgd_00001_fold4

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

• Loss: 0.9877
• Accuracy: 0.5167

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
1.0889	1.0	375	1.0662	0.4217
1.0663	2.0	750	1.0632	0.4233
1.047	3.0	1125	1.0602	0.43
1.0589	4.0	1500	1.0573	0.43
1.0463	5.0	1875	1.0544	0.4317
1.0388	6.0	2250	1.0516	0.4333
1.0109	7.0	2625	1.0487	0.435
1.0394	8.0	3000	1.0459	0.4383
1.0347	9.0	3375	1.0431	0.4417
1.0355	10.0	3750	1.0404	0.4467
1.059	11.0	4125	1.0377	0.4467
1.0235	12.0	4500	1.0352	0.445
1.0136	13.0	4875	1.0326	0.4467
1.0313	14.0	5250	1.0301	0.45
1.0046	15.0	5625	1.0277	0.4567
1.0138	16.0	6000	1.0253	0.4633
1.0055	17.0	6375	1.0230	0.465
0.998	18.0	6750	1.0207	0.4667
1.0178	19.0	7125	1.0186	0.4667
1.019	20.0	7500	1.0165	0.4717
0.9884	21.0	7875	1.0145	0.4783
1.0226	22.0	8250	1.0125	0.48
1.0239	23.0	8625	1.0106	0.4833
1.0151	24.0	9000	1.0088	0.49
0.997	25.0	9375	1.0071	0.49
0.9698	26.0	9750	1.0054	0.4917
0.958	27.0	10125	1.0038	0.495
1.0132	28.0	10500	1.0023	0.4933
0.9673	29.0	10875	1.0008	0.4983
0.9986	30.0	11250	0.9995	0.5
0.9881	31.0	11625	0.9982	0.505
1.0083	32.0	12000	0.9970	0.505
0.9851	33.0	12375	0.9959	0.5067
0.9949	34.0	12750	0.9949	0.5067
0.988	35.0	13125	0.9939	0.5083
1.0062	36.0	13500	0.9930	0.51
0.9899	37.0	13875	0.9922	0.5083
0.9951	38.0	14250	0.9914	0.51
1.0002	39.0	14625	0.9908	0.5133
0.9573	40.0	15000	0.9902	0.5133
0.9723	41.0	15375	0.9896	0.515
0.977	42.0	15750	0.9892	0.515
0.9762	43.0	16125	0.9888	0.515
0.9976	44.0	16500	0.9885	0.5167
0.965	45.0	16875	0.9882	0.5167
0.9904	46.0	17250	0.9880	0.5167
0.9962	47.0	17625	0.9879	0.5167
0.982	48.0	18000	0.9878	0.5167
0.9851	49.0	18375	0.9877	0.5167
0.9675	50.0	18750	0.9877	0.5167

Framework versions

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