hushem_40x_deit_small_rms_001_fold4

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: 2.0611
• Accuracy: 0.8571

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
1.2113	1.0	219	1.1711	0.4286
1.0563	2.0	438	1.0188	0.5476
1.1492	3.0	657	1.0982	0.5
0.8355	4.0	876	0.9249	0.4524
0.9277	5.0	1095	0.9715	0.4762
0.9059	6.0	1314	0.8917	0.5476
0.697	7.0	1533	1.5323	0.5
0.7359	8.0	1752	0.7730	0.6190
0.6186	9.0	1971	0.7734	0.6190
0.5814	10.0	2190	0.7874	0.7143
0.5808	11.0	2409	0.5974	0.7619
0.5808	12.0	2628	0.7519	0.7381
0.5465	13.0	2847	0.4863	0.8095
0.4671	14.0	3066	0.6575	0.6905
0.5228	15.0	3285	0.6495	0.7143
0.4327	16.0	3504	0.7075	0.7619
0.3283	17.0	3723	0.6356	0.7381
0.3853	18.0	3942	0.5432	0.7857
0.3803	19.0	4161	0.9396	0.7857
0.3493	20.0	4380	0.8015	0.7143
0.3953	21.0	4599	0.7074	0.7619
0.3223	22.0	4818	1.0523	0.6667
0.2414	23.0	5037	1.0911	0.6667
0.2219	24.0	5256	1.1394	0.6905
0.2892	25.0	5475	0.7116	0.7619
0.2739	26.0	5694	1.1234	0.7143
0.2207	27.0	5913	0.8565	0.7857
0.1354	28.0	6132	1.1975	0.7381
0.2042	29.0	6351	0.8634	0.7619
0.152	30.0	6570	0.8119	0.7857
0.1453	31.0	6789	0.8364	0.7381
0.1714	32.0	7008	1.2193	0.8095
0.1106	33.0	7227	1.0792	0.7619
0.1157	34.0	7446	0.9831	0.7619
0.096	35.0	7665	1.1093	0.8095
0.0452	36.0	7884	0.9133	0.8095
0.0552	37.0	8103	1.3044	0.8095
0.0539	38.0	8322	0.9892	0.8095
0.041	39.0	8541	1.1780	0.8571
0.0165	40.0	8760	1.3517	0.8333
0.0361	41.0	8979	1.5071	0.8333
0.046	42.0	9198	1.2679	0.8571
0.0477	43.0	9417	1.7256	0.8333
0.0088	44.0	9636	1.2515	0.8333
0.0208	45.0	9855	1.8769	0.8571
0.0012	46.0	10074	1.9828	0.8333
0.0014	47.0	10293	1.9685	0.8571
0.0001	48.0	10512	1.7583	0.8571
0.0001	49.0	10731	2.0944	0.8571
0.0001	50.0	10950	2.0611	0.8571

Framework versions

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