smids_5x_deit_base_adamax_0001_fold4

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: 1.3050
• Accuracy: 0.8833

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
0.2254	1.0	375	0.3701	0.855
0.1315	2.0	750	0.4134	0.86
0.0805	3.0	1125	0.5790	0.8933
0.0294	4.0	1500	0.6055	0.8917
0.0147	5.0	1875	0.8763	0.8667
0.0107	6.0	2250	0.7925	0.8817
0.0094	7.0	2625	0.8429	0.8833
0.0086	8.0	3000	0.8991	0.89
0.0002	9.0	3375	0.9026	0.8933
0.0003	10.0	3750	1.0478	0.8683
0.0026	11.0	4125	1.0371	0.8817
0.0	12.0	4500	1.0179	0.88
0.0	13.0	4875	1.0263	0.8733
0.0038	14.0	5250	1.0099	0.8783
0.0	15.0	5625	1.0357	0.875
0.0	16.0	6000	1.0401	0.8733
0.0	17.0	6375	1.0642	0.8767
0.0051	18.0	6750	1.0754	0.875
0.0	19.0	7125	1.0660	0.8767
0.0	20.0	7500	1.0944	0.8783
0.0	21.0	7875	1.1121	0.88
0.0	22.0	8250	1.0926	0.8817
0.0	23.0	8625	1.0773	0.8767
0.0	24.0	9000	1.1261	0.875
0.0	25.0	9375	1.1126	0.8833
0.0	26.0	9750	1.1400	0.8867
0.0	27.0	10125	1.1471	0.8833
0.0	28.0	10500	1.1463	0.8833
0.0	29.0	10875	1.1486	0.885
0.0	30.0	11250	1.1954	0.8783
0.0	31.0	11625	1.1951	0.88
0.0	32.0	12000	1.2025	0.8833
0.0	33.0	12375	1.2060	0.8783
0.0	34.0	12750	1.2510	0.88
0.0	35.0	13125	1.2394	0.885
0.0	36.0	13500	1.2452	0.885
0.0	37.0	13875	1.2431	0.885
0.0025	38.0	14250	1.2453	0.8833
0.0	39.0	14625	1.2570	0.8867
0.0	40.0	15000	1.2692	0.885
0.0	41.0	15375	1.2782	0.885
0.0	42.0	15750	1.2837	0.8833
0.0	43.0	16125	1.2874	0.885
0.0	44.0	16500	1.2939	0.8833
0.0	45.0	16875	1.2976	0.885
0.0	46.0	17250	1.3011	0.885
0.0	47.0	17625	1.3035	0.885
0.0	48.0	18000	1.3049	0.885
0.0	49.0	18375	1.3052	0.8833
0.0	50.0	18750	1.3050	0.8833

Framework versions

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