hushem_1x_deit_base_rms_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: 1.3674
• Accuracy: 0.4878

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	5.8345	0.2683
4.8133	2.0	12	1.9738	0.2439
4.8133	3.0	18	1.6557	0.2439
2.3825	4.0	24	1.4419	0.2439
1.6511	5.0	30	1.5141	0.2439
1.6511	6.0	36	1.7332	0.2683
1.5506	7.0	42	1.4915	0.2439
1.5506	8.0	48	1.4901	0.2683
1.4941	9.0	54	1.4008	0.2683
1.5024	10.0	60	1.4017	0.2439
1.5024	11.0	66	1.4108	0.2683
1.6905	12.0	72	1.4762	0.2439
1.6905	13.0	78	1.4772	0.2439
1.4363	14.0	84	1.3917	0.3659
1.4324	15.0	90	1.3778	0.2439
1.4324	16.0	96	1.4917	0.2439
1.4176	17.0	102	1.8605	0.2439
1.4176	18.0	108	1.2587	0.4634
1.4153	19.0	114	1.3519	0.3171
1.363	20.0	120	1.2976	0.3902
1.363	21.0	126	1.7214	0.3902
1.2297	22.0	132	1.5932	0.3415
1.2297	23.0	138	1.0760	0.5122
1.1323	24.0	144	1.1518	0.4390
1.0463	25.0	150	1.1823	0.4146
1.0463	26.0	156	1.0632	0.4634
1.0497	27.0	162	1.1057	0.5122
1.0497	28.0	168	0.9873	0.4390
0.9597	29.0	174	1.0710	0.5122
1.0006	30.0	180	1.1482	0.4146
1.0006	31.0	186	1.1124	0.4634
0.934	32.0	192	1.1437	0.4146
0.934	33.0	198	1.1241	0.4390
0.8599	34.0	204	1.1438	0.4390
0.852	35.0	210	1.1783	0.4634
0.852	36.0	216	1.2807	0.4878
0.8357	37.0	222	1.2879	0.4878
0.8357	38.0	228	1.3101	0.4390
0.7932	39.0	234	1.2773	0.4878
0.7254	40.0	240	1.3480	0.4878
0.7254	41.0	246	1.3839	0.4878
0.7183	42.0	252	1.3674	0.4878
0.7183	43.0	258	1.3674	0.4878
0.6348	44.0	264	1.3674	0.4878
0.6561	45.0	270	1.3674	0.4878
0.6561	46.0	276	1.3674	0.4878
0.6538	47.0	282	1.3674	0.4878
0.6538	48.0	288	1.3674	0.4878
0.6489	49.0	294	1.3674	0.4878
0.6536	50.0	300	1.3674	0.4878

Framework versions

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