hushem_5x_deit_tiny_sgd_001_fold4

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: 1.0007
• Accuracy: 0.5238

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
1.4776	1.0	28	1.5069	0.2857
1.3802	2.0	56	1.4145	0.3571
1.3245	3.0	84	1.3548	0.3571
1.3009	4.0	112	1.3137	0.4286
1.2628	5.0	140	1.2827	0.4524
1.2443	6.0	168	1.2514	0.5238
1.1651	7.0	196	1.2261	0.5238
1.1485	8.0	224	1.2037	0.5238
1.1029	9.0	252	1.1805	0.5238
1.0945	10.0	280	1.1607	0.5238
1.1057	11.0	308	1.1451	0.5476
1.0601	12.0	336	1.1295	0.5476
1.0375	13.0	364	1.1248	0.5476
1.024	14.0	392	1.1065	0.5952
0.9777	15.0	420	1.0997	0.5952
0.9798	16.0	448	1.0984	0.5952
0.9759	17.0	476	1.0858	0.5952
0.9492	18.0	504	1.0744	0.5476
0.911	19.0	532	1.0716	0.5952
0.9409	20.0	560	1.0622	0.5476
0.8706	21.0	588	1.0578	0.5476
0.9232	22.0	616	1.0547	0.5952
0.8639	23.0	644	1.0468	0.5
0.9013	24.0	672	1.0442	0.5238
0.8242	25.0	700	1.0432	0.5238
0.8379	26.0	728	1.0386	0.5238
0.8656	27.0	756	1.0271	0.5238
0.8539	28.0	784	1.0232	0.5
0.831	29.0	812	1.0228	0.5238
0.7984	30.0	840	1.0256	0.5
0.8188	31.0	868	1.0204	0.5
0.8337	32.0	896	1.0202	0.5
0.7879	33.0	924	1.0178	0.5
0.7864	34.0	952	1.0219	0.5238
0.8414	35.0	980	1.0150	0.5238
0.8067	36.0	1008	1.0140	0.5238
0.7647	37.0	1036	1.0119	0.5238
0.7807	38.0	1064	1.0087	0.5238
0.7751	39.0	1092	1.0072	0.5238
0.7728	40.0	1120	1.0064	0.5238
0.7814	41.0	1148	1.0052	0.5238
0.7361	42.0	1176	1.0026	0.5238
0.7838	43.0	1204	1.0019	0.5238
0.7388	44.0	1232	1.0012	0.5238
0.7605	45.0	1260	1.0006	0.5238
0.7578	46.0	1288	1.0005	0.5238
0.7479	47.0	1316	1.0010	0.5238
0.7186	48.0	1344	1.0007	0.5238
0.7471	49.0	1372	1.0007	0.5238
0.7354	50.0	1400	1.0007	0.5238

Framework versions

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