hushem_5x_deit_small_adamax_001_fold5

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: 1.9978
• Accuracy: 0.7561

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
1.4105	1.0	28	1.2401	0.4878
1.1305	2.0	56	1.2429	0.4634
0.8883	3.0	84	1.0090	0.5366
0.7137	4.0	112	0.7064	0.7317
0.4682	5.0	140	0.8445	0.7073
0.5286	6.0	168	1.1652	0.5854
0.3742	7.0	196	0.6789	0.8049
0.3573	8.0	224	0.7278	0.7317
0.272	9.0	252	0.8849	0.7561
0.1472	10.0	280	1.2627	0.6829
0.1837	11.0	308	1.1712	0.7317
0.0937	12.0	336	1.4720	0.7073
0.1467	13.0	364	1.7992	0.6585
0.1394	14.0	392	1.7959	0.5854
0.0985	15.0	420	1.4497	0.7317
0.0548	16.0	448	1.4327	0.7561
0.0354	17.0	476	1.5157	0.7317
0.0897	18.0	504	1.9967	0.7561
0.0783	19.0	532	1.8000	0.6829
0.0872	20.0	560	2.1630	0.7073
0.0467	21.0	588	1.7971	0.7073
0.0024	22.0	616	1.1519	0.8049
0.001	23.0	644	1.4688	0.7805
0.0101	24.0	672	1.1822	0.8293
0.005	25.0	700	1.2237	0.8293
0.0001	26.0	728	1.7234	0.7317
0.0025	27.0	756	1.4712	0.7561
0.0001	28.0	784	2.0676	0.7805
0.0003	29.0	812	2.0101	0.7317
0.0	30.0	840	2.0010	0.7561
0.0	31.0	868	1.9976	0.7561
0.0	32.0	896	1.9954	0.7561
0.0	33.0	924	1.9948	0.7561
0.0	34.0	952	1.9948	0.7561
0.0	35.0	980	1.9948	0.7561
0.0	36.0	1008	1.9942	0.7561
0.0	37.0	1036	1.9944	0.7561
0.0	38.0	1064	1.9949	0.7561
0.0	39.0	1092	1.9951	0.7561
0.0	40.0	1120	1.9957	0.7561
0.0	41.0	1148	1.9963	0.7561
0.0	42.0	1176	1.9964	0.7561
0.0	43.0	1204	1.9969	0.7561
0.0	44.0	1232	1.9973	0.7561
0.0	45.0	1260	1.9974	0.7561
0.0	46.0	1288	1.9976	0.7561
0.0	47.0	1316	1.9978	0.7561
0.0	48.0	1344	1.9978	0.7561
0.0	49.0	1372	1.9978	0.7561
0.0	50.0	1400	1.9978	0.7561

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0