smids_3x_deit_tiny_sgd_0001_fold2

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: 0.6082
• Accuracy: 0.7554

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
1.1785	1.0	225	1.1903	0.3511
1.1205	2.0	450	1.1198	0.3760
1.092	3.0	675	1.0810	0.4309
1.0637	4.0	900	1.0502	0.4626
1.0626	5.0	1125	1.0222	0.4842
0.9785	6.0	1350	0.9974	0.5008
0.9914	7.0	1575	0.9736	0.5175
0.9757	8.0	1800	0.9516	0.5408
0.9275	9.0	2025	0.9302	0.5541
0.9237	10.0	2250	0.9088	0.5624
0.9248	11.0	2475	0.8885	0.5691
0.8852	12.0	2700	0.8691	0.5957
0.8693	13.0	2925	0.8509	0.6073
0.8044	14.0	3150	0.8334	0.6156
0.8135	15.0	3375	0.8176	0.6389
0.8455	16.0	3600	0.8028	0.6506
0.8081	17.0	3825	0.7887	0.6539
0.75	18.0	4050	0.7752	0.6622
0.762	19.0	4275	0.7624	0.6722
0.779	20.0	4500	0.7498	0.6772
0.7954	21.0	4725	0.7380	0.6822
0.7639	22.0	4950	0.7272	0.6938
0.7811	23.0	5175	0.7165	0.7022
0.7218	24.0	5400	0.7064	0.7055
0.725	25.0	5625	0.6973	0.7105
0.6921	26.0	5850	0.6889	0.7105
0.7457	27.0	6075	0.6809	0.7221
0.6572	28.0	6300	0.6736	0.7271
0.6717	29.0	6525	0.6667	0.7288
0.6421	30.0	6750	0.6605	0.7354
0.6376	31.0	6975	0.6545	0.7421
0.6251	32.0	7200	0.6490	0.7438
0.6178	33.0	7425	0.6441	0.7488
0.6335	34.0	7650	0.6394	0.7504
0.6522	35.0	7875	0.6351	0.7521
0.6157	36.0	8100	0.6314	0.7521
0.5791	37.0	8325	0.6279	0.7521
0.596	38.0	8550	0.6248	0.7521
0.5791	39.0	8775	0.6218	0.7554
0.6017	40.0	9000	0.6193	0.7554
0.5375	41.0	9225	0.6170	0.7571
0.6415	42.0	9450	0.6151	0.7571
0.5804	43.0	9675	0.6133	0.7571
0.5964	44.0	9900	0.6118	0.7571
0.5945	45.0	10125	0.6106	0.7554
0.6415	46.0	10350	0.6096	0.7554
0.5566	47.0	10575	0.6089	0.7554
0.5708	48.0	10800	0.6085	0.7554
0.6262	49.0	11025	0.6082	0.7554
0.5519	50.0	11250	0.6082	0.7554

Framework versions

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