hushem_1x_deit_base_adamax_001_fold5

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

• Loss: 2.8386
• Accuracy: 0.6341

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
No log	1.0	6	1.4880	0.2683
1.545	2.0	12	1.4136	0.2439
1.545	3.0	18	1.3443	0.3171
1.396	4.0	24	1.1663	0.5122
1.3173	5.0	30	1.2019	0.4878
1.3173	6.0	36	1.2222	0.5122
1.3167	7.0	42	1.4763	0.2439
1.3167	8.0	48	1.1385	0.5610
1.2585	9.0	54	1.3584	0.3659
1.2419	10.0	60	1.0949	0.5122
1.2419	11.0	66	1.1100	0.4634
1.1714	12.0	72	1.2381	0.3902
1.1714	13.0	78	1.4043	0.4146
1.0593	14.0	84	1.1047	0.4878
1.0451	15.0	90	0.9907	0.4878
1.0451	16.0	96	1.3026	0.5122
0.8805	17.0	102	1.0082	0.6098
0.8805	18.0	108	1.1309	0.4634
0.8077	19.0	114	1.2367	0.5610
0.8096	20.0	120	1.4920	0.4878
0.8096	21.0	126	1.8018	0.4878
0.6582	22.0	132	1.5639	0.5854
0.6582	23.0	138	1.2712	0.4878
0.5106	24.0	144	1.1237	0.5854
0.4184	25.0	150	1.6831	0.5610
0.4184	26.0	156	2.0109	0.6098
0.2718	27.0	162	2.2516	0.6341
0.2718	28.0	168	2.0767	0.5610
0.1639	29.0	174	2.6167	0.5854
0.0535	30.0	180	2.8485	0.6341
0.0535	31.0	186	2.7124	0.6585
0.0454	32.0	192	2.8298	0.6585
0.0454	33.0	198	3.2241	0.6341
0.091	34.0	204	2.4575	0.5854
0.1109	35.0	210	3.7388	0.5610
0.1109	36.0	216	2.3707	0.7073
0.0834	37.0	222	2.5281	0.6341
0.0834	38.0	228	3.1120	0.6098
0.0051	39.0	234	2.7929	0.6341
0.0015	40.0	240	2.7025	0.6341
0.0015	41.0	246	2.8185	0.6341
0.0008	42.0	252	2.8386	0.6341
0.0008	43.0	258	2.8386	0.6341
0.0006	44.0	264	2.8386	0.6341
0.0007	45.0	270	2.8386	0.6341
0.0007	46.0	276	2.8386	0.6341
0.0007	47.0	282	2.8386	0.6341
0.0007	48.0	288	2.8386	0.6341
0.0006	49.0	294	2.8386	0.6341
0.0007	50.0	300	2.8386	0.6341

Framework versions

• Transformers 4.35.1
• Pytorch 2.1.0+cu118
• Datasets 2.14.7
• Tokenizers 0.14.1