smids_3x_deit_tiny_adamax_001_fold5

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: 0.9334
• Accuracy: 0.89

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
0.6273	1.0	225	0.4733	0.805
0.2977	2.0	450	0.3448	0.8733
0.3197	3.0	675	0.3620	0.88
0.2909	4.0	900	0.4019	0.8367
0.1808	5.0	1125	0.3972	0.8617
0.2005	6.0	1350	0.4622	0.8483
0.1408	7.0	1575	0.4696	0.8717
0.1502	8.0	1800	0.4332	0.8633
0.2129	9.0	2025	0.3925	0.8833
0.114	10.0	2250	0.4888	0.8733
0.1195	11.0	2475	0.4884	0.8717
0.1124	12.0	2700	0.4439	0.8733
0.0652	13.0	2925	0.6058	0.8633
0.0425	14.0	3150	0.5627	0.875
0.039	15.0	3375	0.5971	0.875
0.0795	16.0	3600	0.6169	0.8733
0.0453	17.0	3825	0.6835	0.875
0.0066	18.0	4050	0.7361	0.8783
0.1031	19.0	4275	0.7941	0.8717
0.0061	20.0	4500	0.7090	0.8733
0.0198	21.0	4725	0.7250	0.8783
0.0155	22.0	4950	0.7045	0.8917
0.023	23.0	5175	0.8046	0.8817
0.057	24.0	5400	0.7359	0.8817
0.0134	25.0	5625	0.7403	0.8867
0.0001	26.0	5850	0.7361	0.895
0.0011	27.0	6075	0.7945	0.8733
0.0014	28.0	6300	0.6928	0.8983
0.0133	29.0	6525	0.6789	0.895
0.0002	30.0	6750	0.7451	0.8967
0.0001	31.0	6975	0.7847	0.8867
0.0	32.0	7200	0.7580	0.8917
0.0043	33.0	7425	0.8908	0.8833
0.0	34.0	7650	0.7939	0.89
0.0034	35.0	7875	0.8753	0.8933
0.0	36.0	8100	0.8470	0.8867
0.0046	37.0	8325	0.9037	0.8867
0.0001	38.0	8550	0.8793	0.89
0.0	39.0	8775	0.8702	0.8917
0.0	40.0	9000	0.8835	0.8883
0.0	41.0	9225	0.9101	0.8883
0.0	42.0	9450	0.9070	0.8933
0.0037	43.0	9675	0.9025	0.8933
0.0	44.0	9900	0.9038	0.89
0.0	45.0	10125	0.9112	0.89
0.0	46.0	10350	0.9228	0.89
0.0	47.0	10575	0.9243	0.89
0.0	48.0	10800	0.9283	0.89
0.0	49.0	11025	0.9314	0.89
0.0	50.0	11250	0.9334	0.89

Framework versions

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