hushem_40x_deit_tiny_sgd_00001_fold5

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: 1.5613
• Accuracy: 0.1951

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
1.4842	1.0	220	1.6446	0.1951
1.5035	2.0	440	1.6414	0.1951
1.4995	3.0	660	1.6381	0.1951
1.5021	4.0	880	1.6349	0.1707
1.454	5.0	1100	1.6317	0.1707
1.4629	6.0	1320	1.6285	0.1707
1.4161	7.0	1540	1.6253	0.1707
1.4101	8.0	1760	1.6223	0.1707
1.4149	9.0	1980	1.6192	0.1707
1.4443	10.0	2200	1.6162	0.1707
1.4163	11.0	2420	1.6133	0.1707
1.4351	12.0	2640	1.6104	0.1707
1.4104	13.0	2860	1.6076	0.1707
1.3915	14.0	3080	1.6048	0.1707
1.4251	15.0	3300	1.6022	0.1707
1.4091	16.0	3520	1.5996	0.1951
1.384	17.0	3740	1.5971	0.1951
1.3979	18.0	3960	1.5947	0.1951
1.3842	19.0	4180	1.5923	0.1951
1.3555	20.0	4400	1.5900	0.1951
1.3519	21.0	4620	1.5879	0.1951
1.3873	22.0	4840	1.5859	0.1951
1.3791	23.0	5060	1.5839	0.1951
1.3799	24.0	5280	1.5820	0.1951
1.3568	25.0	5500	1.5802	0.1951
1.369	26.0	5720	1.5786	0.1951
1.3732	27.0	5940	1.5770	0.1951
1.3491	28.0	6160	1.5754	0.1951
1.3457	29.0	6380	1.5740	0.1951
1.3169	30.0	6600	1.5726	0.1951
1.3748	31.0	6820	1.5714	0.1951
1.3384	32.0	7040	1.5702	0.1951
1.3281	33.0	7260	1.5691	0.1951
1.3359	34.0	7480	1.5681	0.1951
1.3414	35.0	7700	1.5671	0.1951
1.3339	36.0	7920	1.5662	0.1951
1.3778	37.0	8140	1.5654	0.1951
1.3669	38.0	8360	1.5647	0.1951
1.3509	39.0	8580	1.5641	0.1951
1.3269	40.0	8800	1.5635	0.1951
1.3717	41.0	9020	1.5630	0.1951
1.3455	42.0	9240	1.5626	0.1951
1.3737	43.0	9460	1.5622	0.1951
1.3166	44.0	9680	1.5619	0.1951
1.3504	45.0	9900	1.5617	0.1951
1.3509	46.0	10120	1.5615	0.1951
1.3526	47.0	10340	1.5614	0.1951
1.3222	48.0	10560	1.5613	0.1951
1.3165	49.0	10780	1.5613	0.1951
1.3501	50.0	11000	1.5613	0.1951

Framework versions

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