smids_10x_deit_tiny_adamax_0001_fold4

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: 1.1802
• Accuracy: 0.8917

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
0.2287	1.0	750	0.3453	0.8783
0.1774	2.0	1500	0.3819	0.8783
0.1288	3.0	2250	0.5077	0.8617
0.0591	4.0	3000	0.5258	0.8767
0.013	5.0	3750	0.7962	0.87
0.0441	6.0	4500	0.7798	0.875
0.0022	7.0	5250	0.8865	0.8783
0.0168	8.0	6000	0.9982	0.8817
0.0002	9.0	6750	0.9825	0.8833
0.012	10.0	7500	0.9837	0.8883
0.0139	11.0	8250	1.0185	0.88
0.0283	12.0	9000	1.0469	0.8767
0.0013	13.0	9750	1.1375	0.885
0.0051	14.0	10500	1.1468	0.8817
0.0	15.0	11250	1.1486	0.875
0.0211	16.0	12000	1.0421	0.8867
0.0	17.0	12750	1.1215	0.8783
0.0	18.0	13500	1.1501	0.8917
0.0001	19.0	14250	1.2352	0.88
0.0002	20.0	15000	1.2860	0.8883
0.0	21.0	15750	1.1704	0.8833
0.0	22.0	16500	1.0833	0.8933
0.0	23.0	17250	1.1109	0.8933
0.0	24.0	18000	1.1424	0.8933
0.0	25.0	18750	1.0812	0.89
0.0	26.0	19500	1.1046	0.8917
0.0	27.0	20250	1.1453	0.8883
0.0	28.0	21000	1.1203	0.885
0.0	29.0	21750	1.1015	0.8933
0.0	30.0	22500	1.1212	0.8967
0.0	31.0	23250	1.1480	0.8883
0.0	32.0	24000	1.1454	0.8833
0.0	33.0	24750	1.1314	0.8867
0.0	34.0	25500	1.1208	0.885
0.0	35.0	26250	1.1448	0.8833
0.0	36.0	27000	1.1486	0.8833
0.0	37.0	27750	1.1572	0.885
0.0	38.0	28500	1.1406	0.8867
0.0	39.0	29250	1.1768	0.89
0.0	40.0	30000	1.1690	0.885
0.0	41.0	30750	1.1715	0.8883
0.0	42.0	31500	1.1720	0.89
0.0	43.0	32250	1.1654	0.8917
0.0	44.0	33000	1.1692	0.8917
0.0	45.0	33750	1.1750	0.8917
0.0	46.0	34500	1.1770	0.8917
0.0	47.0	35250	1.1783	0.8917
0.0	48.0	36000	1.1786	0.8917
0.0	49.0	36750	1.1796	0.8917
0.0	50.0	37500	1.1802	0.8917

Framework versions

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