smids_5x_deit_tiny_adamax_0001_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: 0.9117
• Accuracy: 0.9067

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: 0.0001
• 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
0.1964	1.0	375	0.3187	0.86
0.1451	2.0	750	0.2705	0.905
0.0998	3.0	1125	0.3251	0.905
0.0219	4.0	1500	0.4834	0.89
0.034	5.0	1875	0.5396	0.8967
0.0222	6.0	2250	0.4814	0.9133
0.0006	7.0	2625	0.6569	0.9133
0.013	8.0	3000	0.5398	0.91
0.001	9.0	3375	0.6798	0.9117
0.0003	10.0	3750	0.7290	0.9083
0.0012	11.0	4125	0.7598	0.9067
0.0	12.0	4500	0.6815	0.9167
0.0001	13.0	4875	0.7765	0.895
0.0004	14.0	5250	0.6391	0.9117
0.0001	15.0	5625	0.8172	0.9067
0.003	16.0	6000	0.6833	0.9083
0.0	17.0	6375	0.7002	0.9133
0.0103	18.0	6750	0.7679	0.91
0.0	19.0	7125	0.8144	0.905
0.0	20.0	7500	0.8111	0.905
0.0	21.0	7875	0.8619	0.9
0.0063	22.0	8250	0.7672	0.9117
0.0001	23.0	8625	0.8118	0.905
0.0	24.0	9000	0.7951	0.9133
0.0	25.0	9375	0.8220	0.9033
0.0	26.0	9750	0.7685	0.915
0.0	27.0	10125	0.8895	0.9
0.0	28.0	10500	0.7796	0.9167
0.0	29.0	10875	0.8774	0.9083
0.0031	30.0	11250	0.8526	0.905
0.0	31.0	11625	0.9000	0.9
0.0	32.0	12000	0.8541	0.9067
0.0	33.0	12375	0.8916	0.905
0.0	34.0	12750	0.8844	0.905
0.0	35.0	13125	0.8624	0.9067
0.0	36.0	13500	0.8971	0.9017
0.0	37.0	13875	0.8894	0.9083
0.003	38.0	14250	0.8778	0.905
0.0	39.0	14625	0.8771	0.9083
0.0	40.0	15000	0.8911	0.905
0.0	41.0	15375	0.8755	0.91
0.0	42.0	15750	0.8923	0.905
0.0	43.0	16125	0.8874	0.9067
0.0	44.0	16500	0.8938	0.9067
0.0	45.0	16875	0.8983	0.9067
0.0	46.0	17250	0.9065	0.9067
0.0027	47.0	17625	0.9068	0.9067
0.0	48.0	18000	0.9127	0.905
0.0	49.0	18375	0.9142	0.905
0.0023	50.0	18750	0.9117	0.9067

Framework versions

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