hushem_40x_deit_tiny_rms_0001_fold4

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

• Loss: 0.4037
• Accuracy: 0.9524

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
0.1838	1.0	219	0.4926	0.8571
0.0446	2.0	438	0.2754	0.9286
0.0295	3.0	657	0.9751	0.8810
0.0096	4.0	876	0.1123	0.9762
0.0055	5.0	1095	0.3687	0.9048
0.0033	6.0	1314	0.3076	0.9524
0.0283	7.0	1533	0.8089	0.8571
0.0044	8.0	1752	0.2435	0.9286
0.0018	9.0	1971	0.7038	0.8571
0.0191	10.0	2190	0.5242	0.9048
0.0001	11.0	2409	0.8130	0.9286
0.0007	12.0	2628	0.6030	0.9048
0.0189	13.0	2847	0.5406	0.9048
0.0002	14.0	3066	0.6774	0.8571
0.0018	15.0	3285	0.6982	0.9286
0.0001	16.0	3504	0.3877	0.9524
0.0008	17.0	3723	0.6996	0.8810
0.0	18.0	3942	0.5507	0.9286
0.0	19.0	4161	0.3796	0.9524
0.0001	20.0	4380	0.3967	0.9286
0.0	21.0	4599	0.4081	0.9286
0.0	22.0	4818	0.3898	0.9286
0.0	23.0	5037	0.3709	0.9286
0.0	24.0	5256	0.3640	0.9524
0.0	25.0	5475	0.3789	0.9524
0.0	26.0	5694	0.3987	0.9286
0.0	27.0	5913	0.4326	0.9286
0.0	28.0	6132	0.4566	0.9286
0.0	29.0	6351	0.4673	0.9286
0.0	30.0	6570	0.4642	0.9286
0.0	31.0	6789	0.4534	0.9286
0.0	32.0	7008	0.4388	0.9286
0.0	33.0	7227	0.4268	0.9286
0.0	34.0	7446	0.4182	0.9286
0.0	35.0	7665	0.4134	0.9286
0.0	36.0	7884	0.4102	0.9286
0.0	37.0	8103	0.4079	0.9286
0.0	38.0	8322	0.4066	0.9286
0.0	39.0	8541	0.4041	0.9286
0.0	40.0	8760	0.4048	0.9286
0.0	41.0	8979	0.4034	0.9524
0.0	42.0	9198	0.4032	0.9524
0.0	43.0	9417	0.4038	0.9524
0.0	44.0	9636	0.4040	0.9524
0.0	45.0	9855	0.4040	0.9524
0.0	46.0	10074	0.4038	0.9524
0.0	47.0	10293	0.4038	0.9524
0.0	48.0	10512	0.4039	0.9524
0.0	49.0	10731	0.4037	0.9524
0.0	50.0	10950	0.4037	0.9524

Framework versions

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