smids_3x_deit_base_sgd_0001_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: 0.6429
• Accuracy: 0.77

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
1.0879	1.0	225	1.0991	0.3483
1.0983	2.0	450	1.0868	0.3783
1.0739	3.0	675	1.0747	0.4117
1.0339	4.0	900	1.0623	0.44
1.037	5.0	1125	1.0494	0.475
1.0134	6.0	1350	1.0355	0.4983
1.0062	7.0	1575	1.0204	0.5233
0.9728	8.0	1800	1.0042	0.5467
0.9628	9.0	2025	0.9873	0.56
0.9357	10.0	2250	0.9698	0.58
0.9397	11.0	2475	0.9521	0.6
0.916	12.0	2700	0.9337	0.6183
0.9116	13.0	2925	0.9154	0.6283
0.8855	14.0	3150	0.8973	0.6467
0.8322	15.0	3375	0.8797	0.67
0.8412	16.0	3600	0.8631	0.6883
0.8508	17.0	3825	0.8471	0.695
0.8144	18.0	4050	0.8319	0.7
0.8203	19.0	4275	0.8174	0.715
0.8033	20.0	4500	0.8038	0.7283
0.7838	21.0	4725	0.7909	0.7317
0.7999	22.0	4950	0.7790	0.73
0.7579	23.0	5175	0.7676	0.7383
0.7789	24.0	5400	0.7568	0.7433
0.7344	25.0	5625	0.7468	0.7467
0.7494	26.0	5850	0.7374	0.7517
0.7562	27.0	6075	0.7284	0.755
0.701	28.0	6300	0.7202	0.7583
0.7274	29.0	6525	0.7124	0.7617
0.666	30.0	6750	0.7052	0.7617
0.691	31.0	6975	0.6984	0.765
0.6782	32.0	7200	0.6921	0.765
0.7115	33.0	7425	0.6862	0.7667
0.6811	34.0	7650	0.6808	0.7717
0.6447	35.0	7875	0.6758	0.7717
0.6506	36.0	8100	0.6713	0.7717
0.6763	37.0	8325	0.6671	0.7717
0.6655	38.0	8550	0.6632	0.7733
0.6291	39.0	8775	0.6597	0.7767
0.6533	40.0	9000	0.6566	0.775
0.6426	41.0	9225	0.6538	0.7733
0.6744	42.0	9450	0.6514	0.7733
0.6449	43.0	9675	0.6493	0.775
0.6886	44.0	9900	0.6475	0.7733
0.6363	45.0	10125	0.6460	0.7733
0.6267	46.0	10350	0.6448	0.7717
0.6739	47.0	10575	0.6439	0.77
0.6555	48.0	10800	0.6433	0.7717
0.6557	49.0	11025	0.6430	0.77
0.6715	50.0	11250	0.6429	0.77

Framework versions

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