smids_10x_deit_tiny_adamax_001_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: 1.5843
• Accuracy: 0.8717

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
0.3385	1.0	750	0.3848	0.84
0.2692	2.0	1500	0.3830	0.8633
0.2345	3.0	2250	0.4255	0.8617
0.1851	4.0	3000	0.4988	0.8517
0.1806	5.0	3750	0.5032	0.8433
0.1568	6.0	4500	0.5429	0.8633
0.0638	7.0	5250	0.6033	0.855
0.1397	8.0	6000	0.6990	0.845
0.1208	9.0	6750	0.6852	0.8483
0.0667	10.0	7500	0.8743	0.8383
0.0482	11.0	8250	0.7516	0.8667
0.0306	12.0	9000	0.8187	0.8783
0.0125	13.0	9750	0.8525	0.86
0.0512	14.0	10500	1.0441	0.8483
0.0023	15.0	11250	1.0562	0.85
0.0353	16.0	12000	1.1914	0.8583
0.0637	17.0	12750	1.1115	0.8667
0.025	18.0	13500	1.1677	0.865
0.0126	19.0	14250	1.0523	0.8833
0.0	20.0	15000	1.0935	0.8633
0.0359	21.0	15750	1.1791	0.8733
0.0003	22.0	16500	1.0630	0.87
0.0003	23.0	17250	1.0996	0.8667
0.0006	24.0	18000	1.0915	0.8817
0.0001	25.0	18750	1.1484	0.8617
0.0	26.0	19500	1.1656	0.875
0.0179	27.0	20250	1.2101	0.8717
0.0	28.0	21000	1.3179	0.86
0.0	29.0	21750	1.2425	0.8733
0.0	30.0	22500	1.3660	0.87
0.0	31.0	23250	1.3781	0.87
0.0	32.0	24000	1.4541	0.86
0.0003	33.0	24750	1.3447	0.8717
0.0	34.0	25500	1.3846	0.8633
0.0	35.0	26250	1.3907	0.8733
0.0	36.0	27000	1.4240	0.87
0.0	37.0	27750	1.3878	0.8717
0.0	38.0	28500	1.4082	0.87
0.0	39.0	29250	1.4530	0.8717
0.0	40.0	30000	1.4653	0.8717
0.0	41.0	30750	1.4878	0.87
0.0	42.0	31500	1.5011	0.8717
0.0	43.0	32250	1.5107	0.8717
0.0	44.0	33000	1.5209	0.8717
0.0	45.0	33750	1.5429	0.8717
0.0	46.0	34500	1.5577	0.8717
0.0	47.0	35250	1.5684	0.8717
0.0	48.0	36000	1.5772	0.8717
0.0	49.0	36750	1.5824	0.8717
0.0	50.0	37500	1.5843	0.8717

Framework versions

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