smids_5x_deit_base_adamax_001_fold3

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: 1.1005
• 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
0.307	1.0	375	0.3243	0.88
0.1951	2.0	750	0.2848	0.8967
0.1756	3.0	1125	0.3260	0.8767
0.1301	4.0	1500	0.3461	0.8933
0.1724	5.0	1875	0.3433	0.8783
0.1105	6.0	2250	0.5327	0.8517
0.105	7.0	2625	0.4495	0.89
0.1373	8.0	3000	0.3477	0.8933
0.0545	9.0	3375	0.5403	0.8767
0.026	10.0	3750	0.6392	0.8717
0.0547	11.0	4125	0.6160	0.875
0.0385	12.0	4500	0.5572	0.885
0.0376	13.0	4875	0.6146	0.8967
0.0031	14.0	5250	0.6509	0.8883
0.0185	15.0	5625	0.6515	0.885
0.0353	16.0	6000	0.7637	0.885
0.0052	17.0	6375	0.7211	0.8817
0.011	18.0	6750	0.5915	0.9067
0.0053	19.0	7125	0.6576	0.89
0.0044	20.0	7500	0.6728	0.8983
0.0003	21.0	7875	0.7362	0.8817
0.0001	22.0	8250	0.7370	0.8817
0.0265	23.0	8625	0.6954	0.895
0.0011	24.0	9000	0.7244	0.8883
0.0056	25.0	9375	0.7383	0.8917
0.0	26.0	9750	0.6944	0.9033
0.0001	27.0	10125	0.8581	0.8933
0.0002	28.0	10500	0.7732	0.8917
0.0001	29.0	10875	0.9540	0.8867
0.005	30.0	11250	0.8145	0.8933
0.0003	31.0	11625	0.8223	0.8967
0.0	32.0	12000	0.8225	0.89
0.0	33.0	12375	0.8479	0.8933
0.0	34.0	12750	0.8571	0.895
0.0	35.0	13125	0.9119	0.8917
0.0	36.0	13500	0.9029	0.8917
0.0	37.0	13875	0.9226	0.8967
0.0	38.0	14250	0.9083	0.895
0.0	39.0	14625	1.0048	0.8933
0.0026	40.0	15000	1.0018	0.8883
0.0	41.0	15375	1.0177	0.8917
0.0	42.0	15750	1.0273	0.8917
0.0	43.0	16125	1.0393	0.8933
0.0	44.0	16500	1.0649	0.895
0.0	45.0	16875	1.0825	0.8883
0.0	46.0	17250	1.0743	0.895
0.0	47.0	17625	1.0848	0.8917
0.0	48.0	18000	1.0902	0.8917
0.0	49.0	18375	1.0954	0.89
0.0	50.0	18750	1.1005	0.89

Framework versions

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