smids_3x_deit_tiny_adamax_00001_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: 0.8922
• Accuracy: 0.8867

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
0.4176	1.0	225	0.4035	0.8483
0.2988	2.0	450	0.3067	0.8767
0.2655	3.0	675	0.2905	0.875
0.1842	4.0	900	0.2684	0.8983
0.1435	5.0	1125	0.2769	0.8933
0.1106	6.0	1350	0.2720	0.895
0.1509	7.0	1575	0.2967	0.8917
0.1529	8.0	1800	0.3180	0.8767
0.1311	9.0	2025	0.3248	0.89
0.0801	10.0	2250	0.3813	0.885
0.0435	11.0	2475	0.4094	0.8833
0.0973	12.0	2700	0.4656	0.88
0.0775	13.0	2925	0.4789	0.8917
0.0342	14.0	3150	0.5459	0.88
0.0207	15.0	3375	0.5599	0.8833
0.0139	16.0	3600	0.5932	0.8917
0.0015	17.0	3825	0.6480	0.88
0.0008	18.0	4050	0.6641	0.88
0.0269	19.0	4275	0.6876	0.885
0.0066	20.0	4500	0.7051	0.8883
0.0002	21.0	4725	0.7338	0.8883
0.0003	22.0	4950	0.7295	0.88
0.0053	23.0	5175	0.7640	0.8833
0.0118	24.0	5400	0.8006	0.8833
0.0002	25.0	5625	0.7995	0.885
0.0002	26.0	5850	0.8061	0.8833
0.0002	27.0	6075	0.8090	0.8817
0.0003	28.0	6300	0.8501	0.875
0.0137	29.0	6525	0.8643	0.8767
0.0001	30.0	6750	0.8347	0.885
0.0001	31.0	6975	0.8412	0.8867
0.0001	32.0	7200	0.8482	0.8867
0.0139	33.0	7425	0.8560	0.8833
0.0	34.0	7650	0.8490	0.8817
0.0026	35.0	7875	0.8633	0.8867
0.0	36.0	8100	0.8671	0.8883
0.0222	37.0	8325	0.8736	0.885
0.0	38.0	8550	0.8850	0.8783
0.0	39.0	8775	0.8799	0.8833
0.0	40.0	9000	0.8936	0.88
0.0	41.0	9225	0.8899	0.8817
0.0	42.0	9450	0.8900	0.885
0.0114	43.0	9675	0.8889	0.8833
0.0	44.0	9900	0.8840	0.8867
0.0	45.0	10125	0.8851	0.8867
0.0	46.0	10350	0.8906	0.885
0.0	47.0	10575	0.8900	0.885
0.0	48.0	10800	0.8911	0.8867
0.0	49.0	11025	0.8913	0.8867
0.0	50.0	11250	0.8922	0.8867

Framework versions

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