hushem_40x_deit_tiny_rms_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: 2.2392
• Accuracy: 0.7907

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
1.246	1.0	217	1.4100	0.2558
0.9627	2.0	434	1.1709	0.5116
0.8333	3.0	651	1.1793	0.4186
0.7844	4.0	868	0.8077	0.6279
0.7035	5.0	1085	1.3327	0.5116
0.7976	6.0	1302	0.7941	0.6977
0.7352	7.0	1519	0.9909	0.6279
0.6247	8.0	1736	0.7281	0.6977
0.6212	9.0	1953	1.1902	0.6279
0.6647	10.0	2170	1.0897	0.5581
0.4763	11.0	2387	0.9383	0.6047
0.5076	12.0	2604	0.5861	0.7907
0.4573	13.0	2821	0.9438	0.5349
0.3857	14.0	3038	0.7991	0.6977
0.3919	15.0	3255	0.9377	0.6047
0.352	16.0	3472	1.0859	0.5814
0.3551	17.0	3689	1.2113	0.6744
0.3196	18.0	3906	1.2889	0.6279
0.2405	19.0	4123	0.9915	0.6512
0.2367	20.0	4340	1.6136	0.6279
0.2222	21.0	4557	1.4836	0.5814
0.1901	22.0	4774	1.0739	0.7209
0.173	23.0	4991	1.3956	0.6512
0.1711	24.0	5208	1.7072	0.6279
0.1027	25.0	5425	1.4657	0.6512
0.0952	26.0	5642	1.6372	0.6744
0.1462	27.0	5859	2.2566	0.5814
0.1003	28.0	6076	1.5093	0.6512
0.0764	29.0	6293	1.9318	0.6512
0.1025	30.0	6510	1.9630	0.6047
0.0702	31.0	6727	2.1273	0.6512
0.0313	32.0	6944	1.6171	0.7209
0.0732	33.0	7161	1.2147	0.7209
0.0384	34.0	7378	1.9804	0.7209
0.0177	35.0	7595	1.8221	0.6512
0.0098	36.0	7812	2.4941	0.6744
0.0407	37.0	8029	2.6063	0.6512
0.0798	38.0	8246	3.5391	0.5581
0.0022	39.0	8463	2.7971	0.6512
0.0004	40.0	8680	1.8602	0.7209
0.0547	41.0	8897	2.4427	0.6744
0.0003	42.0	9114	2.1061	0.6977
0.0003	43.0	9331	2.0283	0.5814
0.0017	44.0	9548	2.1926	0.6744
0.0001	45.0	9765	1.9704	0.7674
0.0	46.0	9982	2.2645	0.7442
0.0001	47.0	10199	2.3408	0.7674
0.0	48.0	10416	2.2312	0.7674
0.0	49.0	10633	2.2302	0.7907
0.0	50.0	10850	2.2392	0.7907

Framework versions

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