hushem_40x_deit_small_sgd_00001_fold3

This model is a fine-tuned version of facebook/deit-small-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.3703
• Accuracy: 0.3256

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: 1e-05
• 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
1.8914	1.0	217	1.5805	0.2558
2.0178	2.0	434	1.5654	0.2558
2.0179	3.0	651	1.5510	0.2558
1.8888	4.0	868	1.5374	0.2558
1.872	5.0	1085	1.5245	0.2558
1.7831	6.0	1302	1.5124	0.2558
1.836	7.0	1519	1.5009	0.2558
1.8178	8.0	1736	1.4901	0.2558
1.7694	9.0	1953	1.4801	0.2326
1.7678	10.0	2170	1.4706	0.2326
1.659	11.0	2387	1.4618	0.2326
1.6239	12.0	2604	1.4536	0.2558
1.6882	13.0	2821	1.4460	0.2558
1.6748	14.0	3038	1.4391	0.2558
1.6892	15.0	3255	1.4327	0.2791
1.725	16.0	3472	1.4268	0.2791
1.6371	17.0	3689	1.4214	0.2791
1.6193	18.0	3906	1.4164	0.3256
1.6512	19.0	4123	1.4119	0.3256
1.6188	20.0	4340	1.4078	0.3256
1.643	21.0	4557	1.4041	0.3256
1.5803	22.0	4774	1.4006	0.3256
1.592	23.0	4991	1.3975	0.3256
1.5987	24.0	5208	1.3946	0.3256
1.566	25.0	5425	1.3921	0.3488
1.5574	26.0	5642	1.3897	0.3488
1.4978	27.0	5859	1.3876	0.3488
1.524	28.0	6076	1.3857	0.3488
1.5682	29.0	6293	1.3839	0.3488
1.5042	30.0	6510	1.3823	0.3488
1.5589	31.0	6727	1.3808	0.3023
1.5347	32.0	6944	1.3795	0.3023
1.5403	33.0	7161	1.3783	0.3023
1.5548	34.0	7378	1.3772	0.3023
1.5321	35.0	7595	1.3762	0.3023
1.5015	36.0	7812	1.3753	0.3023
1.4993	37.0	8029	1.3745	0.3023
1.4844	38.0	8246	1.3738	0.3023
1.5191	39.0	8463	1.3732	0.3023
1.515	40.0	8680	1.3726	0.3256
1.4957	41.0	8897	1.3721	0.3256
1.5585	42.0	9114	1.3717	0.3256
1.5037	43.0	9331	1.3713	0.3256
1.4828	44.0	9548	1.3710	0.3256
1.4967	45.0	9765	1.3708	0.3256
1.5387	46.0	9982	1.3706	0.3256
1.5118	47.0	10199	1.3705	0.3256
1.5073	48.0	10416	1.3704	0.3256
1.5166	49.0	10633	1.3703	0.3256
1.4994	50.0	10850	1.3703	0.3256

Framework versions

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