smids_5x_deit_base_adamax_001_fold5

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: 0.8666
• 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: 0.001
• 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
0.3384	1.0	375	0.4257	0.8267
0.3243	2.0	750	0.3051	0.8883
0.2315	3.0	1125	0.3393	0.8783
0.1699	4.0	1500	0.4297	0.8583
0.1105	5.0	1875	0.3821	0.8983
0.1049	6.0	2250	0.3824	0.895
0.0625	7.0	2625	0.5340	0.8967
0.0706	8.0	3000	0.5827	0.8783
0.039	9.0	3375	0.4159	0.895
0.0887	10.0	3750	0.4518	0.905
0.042	11.0	4125	0.4385	0.91
0.0677	12.0	4500	0.5266	0.8983
0.0355	13.0	4875	0.4982	0.8883
0.0188	14.0	5250	0.5825	0.9083
0.0091	15.0	5625	0.4685	0.915
0.0008	16.0	6000	0.6661	0.8983
0.026	17.0	6375	0.5630	0.9
0.0121	18.0	6750	0.6999	0.8967
0.0069	19.0	7125	0.5495	0.9083
0.0011	20.0	7500	0.6260	0.9033
0.0026	21.0	7875	0.6616	0.91
0.0056	22.0	8250	0.6236	0.915
0.0072	23.0	8625	0.7060	0.905
0.0005	24.0	9000	0.7311	0.9067
0.0	25.0	9375	0.7450	0.91
0.0001	26.0	9750	0.7238	0.91
0.0019	27.0	10125	0.7673	0.9
0.0007	28.0	10500	0.7394	0.91
0.0072	29.0	10875	0.7457	0.91
0.0069	30.0	11250	0.9604	0.8883
0.0	31.0	11625	0.7446	0.91
0.0	32.0	12000	0.7855	0.905
0.0	33.0	12375	0.7691	0.905
0.0	34.0	12750	0.7719	0.9067
0.0	35.0	13125	0.7976	0.9017
0.0	36.0	13500	0.8067	0.9033
0.0	37.0	13875	0.7973	0.9067
0.0041	38.0	14250	0.8120	0.9067
0.0	39.0	14625	0.8149	0.9067
0.0	40.0	15000	0.7879	0.9067
0.0	41.0	15375	0.8013	0.9067
0.0	42.0	15750	0.8079	0.905
0.0	43.0	16125	0.8212	0.9017
0.0	44.0	16500	0.8180	0.905
0.0	45.0	16875	0.8381	0.9067
0.0	46.0	17250	0.8519	0.905
0.003	47.0	17625	0.8539	0.9067
0.0	48.0	18000	0.8604	0.9083
0.0	49.0	18375	0.8650	0.9067
0.0021	50.0	18750	0.8666	0.9067

Framework versions

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