smids_5x_deit_tiny_sgd_001_fold5

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: 0.2848
• Accuracy: 0.8717

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
0.7235	1.0	375	0.7207	0.7267
0.5565	2.0	750	0.5409	0.7917
0.4119	3.0	1125	0.4652	0.825
0.4334	4.0	1500	0.4261	0.825
0.3887	5.0	1875	0.3995	0.8317
0.3453	6.0	2250	0.3796	0.8483
0.3126	7.0	2625	0.3666	0.8533
0.2904	8.0	3000	0.3555	0.8533
0.2918	9.0	3375	0.3426	0.8617
0.2742	10.0	3750	0.3359	0.8583
0.2643	11.0	4125	0.3302	0.865
0.2302	12.0	4500	0.3258	0.8617
0.2469	13.0	4875	0.3202	0.8633
0.2562	14.0	5250	0.3189	0.8683
0.29	15.0	5625	0.3116	0.87
0.1812	16.0	6000	0.3053	0.87
0.297	17.0	6375	0.3040	0.8733
0.2383	18.0	6750	0.3060	0.8733
0.2162	19.0	7125	0.3016	0.8683
0.1713	20.0	7500	0.3023	0.8717
0.1959	21.0	7875	0.2972	0.8767
0.1937	22.0	8250	0.2967	0.87
0.2843	23.0	8625	0.2932	0.8683
0.199	24.0	9000	0.2945	0.8717
0.2226	25.0	9375	0.2906	0.875
0.2091	26.0	9750	0.2867	0.8683
0.2068	27.0	10125	0.2936	0.8717
0.1892	28.0	10500	0.2828	0.8767
0.2214	29.0	10875	0.2884	0.8683
0.1748	30.0	11250	0.2926	0.8683
0.2363	31.0	11625	0.2864	0.8733
0.209	32.0	12000	0.2852	0.8733
0.2053	33.0	12375	0.2863	0.8683
0.2387	34.0	12750	0.2806	0.8767
0.1656	35.0	13125	0.2863	0.87
0.1491	36.0	13500	0.2870	0.8667
0.1628	37.0	13875	0.2841	0.87
0.1814	38.0	14250	0.2851	0.8683
0.1769	39.0	14625	0.2852	0.8667
0.1855	40.0	15000	0.2843	0.8717
0.185	41.0	15375	0.2824	0.8717
0.1482	42.0	15750	0.2829	0.8733
0.2112	43.0	16125	0.2835	0.875
0.2011	44.0	16500	0.2835	0.8733
0.1775	45.0	16875	0.2835	0.8717
0.1816	46.0	17250	0.2845	0.8717
0.1826	47.0	17625	0.2843	0.8717
0.1411	48.0	18000	0.2851	0.8683
0.1734	49.0	18375	0.2850	0.8717
0.1888	50.0	18750	0.2848	0.8717

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.1+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2