smids_10x_deit_small_adamax_00001_fold2

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: 1.1874
• Accuracy: 0.8719

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
0.2546	1.0	750	0.2964	0.8885
0.1392	2.0	1500	0.2964	0.8935
0.1051	3.0	2250	0.3173	0.8802
0.0797	4.0	3000	0.3716	0.8802
0.0803	5.0	3750	0.4496	0.8769
0.0599	6.0	4500	0.5455	0.8769
0.0367	7.0	5250	0.6753	0.8686
0.0203	8.0	6000	0.7402	0.8752
0.0136	9.0	6750	0.8455	0.8686
0.0001	10.0	7500	0.8969	0.8686
0.0056	11.0	8250	0.9305	0.8769
0.0002	12.0	9000	0.9474	0.8752
0.0	13.0	9750	0.9957	0.8785
0.0	14.0	10500	1.0123	0.8769
0.0001	15.0	11250	0.9720	0.8835
0.0001	16.0	12000	1.0684	0.8785
0.0003	17.0	12750	1.1079	0.8752
0.0	18.0	13500	1.0971	0.8752
0.0	19.0	14250	1.0987	0.8735
0.0	20.0	15000	1.1190	0.8769
0.0	21.0	15750	1.1376	0.8686
0.0049	22.0	16500	1.1379	0.8686
0.0014	23.0	17250	1.1542	0.8752
0.0	24.0	18000	1.1536	0.8735
0.0	25.0	18750	1.1721	0.8719
0.0	26.0	19500	1.1498	0.8719
0.01	27.0	20250	1.1595	0.8719
0.0	28.0	21000	1.1250	0.8785
0.0	29.0	21750	1.1514	0.8686
0.0	30.0	22500	1.1182	0.8735
0.0	31.0	23250	1.1637	0.8752
0.0	32.0	24000	1.1726	0.8735
0.0	33.0	24750	1.1697	0.8719
0.0	34.0	25500	1.1588	0.8752
0.0	35.0	26250	1.1653	0.8702
0.0	36.0	27000	1.1669	0.8719
0.0141	37.0	27750	1.1767	0.8719
0.0	38.0	28500	1.1781	0.8719
0.0	39.0	29250	1.1951	0.8702
0.0	40.0	30000	1.1887	0.8702
0.0	41.0	30750	1.1872	0.8702
0.0	42.0	31500	1.1896	0.8702
0.0	43.0	32250	1.1930	0.8702
0.0	44.0	33000	1.1942	0.8702
0.0056	45.0	33750	1.1902	0.8702
0.0	46.0	34500	1.1880	0.8702
0.0	47.0	35250	1.1877	0.8702
0.0	48.0	36000	1.1882	0.8702
0.0	49.0	36750	1.1884	0.8702
0.0	50.0	37500	1.1874	0.8719

Framework versions

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