hushem_40x_deit_tiny_adamax_001_fold3

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: 1.2217
• Accuracy: 0.8837

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
0.3354	1.0	217	0.9126	0.6744
0.1561	2.0	434	1.0121	0.6977
0.0656	3.0	651	1.0046	0.7442
0.07	4.0	868	0.8401	0.7674
0.0448	5.0	1085	1.1419	0.7907
0.0667	6.0	1302	1.2686	0.8372
0.0198	7.0	1519	1.1351	0.8372
0.009	8.0	1736	1.6617	0.8140
0.0113	9.0	1953	1.1646	0.8372
0.0284	10.0	2170	0.8325	0.9302
0.0093	11.0	2387	1.2543	0.8372
0.0217	12.0	2604	1.5838	0.8140
0.0059	13.0	2821	1.3486	0.8605
0.0271	14.0	3038	0.9639	0.8837
0.0039	15.0	3255	0.9754	0.9070
0.0	16.0	3472	1.4140	0.8140
0.0	17.0	3689	1.2390	0.8837
0.0011	18.0	3906	1.3250	0.8605
0.0004	19.0	4123	1.2826	0.8372
0.0002	20.0	4340	1.1795	0.8605
0.0	21.0	4557	0.9932	0.8837
0.0	22.0	4774	0.9999	0.8837
0.0	23.0	4991	1.0055	0.8837
0.0	24.0	5208	1.0099	0.8837
0.0	25.0	5425	1.0179	0.8837
0.0	26.0	5642	1.0241	0.8837
0.0	27.0	5859	1.0302	0.8837
0.0	28.0	6076	1.0378	0.8837
0.0	29.0	6293	1.0445	0.8837
0.0	30.0	6510	1.0519	0.8837
0.0	31.0	6727	1.0608	0.8837
0.0	32.0	6944	1.0697	0.8837
0.0	33.0	7161	1.0787	0.8837
0.0	34.0	7378	1.0879	0.8837
0.0	35.0	7595	1.0974	0.8837
0.0	36.0	7812	1.1074	0.8837
0.0	37.0	8029	1.1171	0.8837
0.0	38.0	8246	1.1279	0.8837
0.0	39.0	8463	1.1383	0.8837
0.0	40.0	8680	1.1481	0.8837
0.0	41.0	8897	1.1586	0.8837
0.0	42.0	9114	1.1687	0.8837
0.0	43.0	9331	1.1785	0.8837
0.0	44.0	9548	1.1883	0.8837
0.0	45.0	9765	1.1969	0.8837
0.0	46.0	9982	1.2048	0.8837
0.0	47.0	10199	1.2119	0.8837
0.0	48.0	10416	1.2174	0.8837
0.0	49.0	10633	1.2210	0.8837
0.0	50.0	10850	1.2217	0.8837

Framework versions

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