smids_3x_deit_small_adamax_00001_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: 1.0772
• Accuracy: 0.87

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: 1e-05
• 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
0.3797	1.0	225	0.3792	0.8467
0.1834	2.0	450	0.3415	0.8583
0.1521	3.0	675	0.3364	0.8767
0.1673	4.0	900	0.3475	0.8733
0.1145	5.0	1125	0.3681	0.8733
0.0638	6.0	1350	0.4237	0.8683
0.0638	7.0	1575	0.4385	0.875
0.1043	8.0	1800	0.4695	0.88
0.0499	9.0	2025	0.5316	0.8783
0.0307	10.0	2250	0.6026	0.87
0.0636	11.0	2475	0.6654	0.87
0.0182	12.0	2700	0.6645	0.8767
0.0011	13.0	2925	0.7026	0.8717
0.0033	14.0	3150	0.7540	0.8717
0.0005	15.0	3375	0.7821	0.87
0.0003	16.0	3600	0.7783	0.8767
0.0002	17.0	3825	0.8172	0.8717
0.0057	18.0	4050	0.8281	0.8717
0.0001	19.0	4275	0.8404	0.8717
0.0001	20.0	4500	0.8657	0.8733
0.0001	21.0	4725	0.8953	0.875
0.0001	22.0	4950	0.9021	0.87
0.0001	23.0	5175	0.9104	0.8733
0.0001	24.0	5400	0.9154	0.8733
0.0001	25.0	5625	0.9415	0.87
0.0	26.0	5850	0.9539	0.8733
0.0	27.0	6075	0.9672	0.87
0.0001	28.0	6300	0.9843	0.8683
0.0	29.0	6525	0.9786	0.87
0.0	30.0	6750	0.9994	0.8683
0.0	31.0	6975	0.9988	0.8733
0.0	32.0	7200	1.0032	0.8717
0.0	33.0	7425	1.0174	0.87
0.0	34.0	7650	1.0208	0.87
0.0	35.0	7875	1.0273	0.8717
0.0	36.0	8100	1.0325	0.8717
0.0	37.0	8325	1.0347	0.87
0.0	38.0	8550	1.0403	0.87
0.0	39.0	8775	1.0555	0.87
0.0	40.0	9000	1.0534	0.87
0.0	41.0	9225	1.0588	0.87
0.0084	42.0	9450	1.0654	0.87
0.0	43.0	9675	1.0680	0.87
0.0079	44.0	9900	1.0688	0.8683
0.0	45.0	10125	1.0727	0.87
0.0	46.0	10350	1.0741	0.87
0.0	47.0	10575	1.0751	0.87
0.0	48.0	10800	1.0765	0.87
0.0	49.0	11025	1.0763	0.8667
0.0	50.0	11250	1.0772	0.87

Framework versions

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