hushem_1x_deit_tiny_sgd_001_fold1

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: 1.3946
• Accuracy: 0.2667

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	1.6081	0.2889
1.6517	2.0	12	1.5532	0.3333
1.6517	3.0	18	1.5183	0.3111
1.5073	4.0	24	1.4941	0.2
1.4569	5.0	30	1.4762	0.1333
1.4569	6.0	36	1.4655	0.1333
1.377	7.0	42	1.4570	0.1333
1.377	8.0	48	1.4508	0.1333
1.3495	9.0	54	1.4443	0.1333
1.3234	10.0	60	1.4390	0.1333
1.3234	11.0	66	1.4339	0.1778
1.2813	12.0	72	1.4301	0.1778
1.2813	13.0	78	1.4257	0.2
1.3124	14.0	84	1.4223	0.2
1.2528	15.0	90	1.4195	0.2
1.2528	16.0	96	1.4170	0.2222
1.2252	17.0	102	1.4152	0.2
1.2252	18.0	108	1.4125	0.2222
1.2441	19.0	114	1.4108	0.2
1.1872	20.0	120	1.4088	0.2
1.1872	21.0	126	1.4068	0.2
1.1818	22.0	132	1.4052	0.2222
1.1818	23.0	138	1.4041	0.2
1.1835	24.0	144	1.4032	0.2222
1.1551	25.0	150	1.4021	0.2222
1.1551	26.0	156	1.4013	0.2222
1.1564	27.0	162	1.4008	0.2
1.1564	28.0	168	1.3999	0.2222
1.1662	29.0	174	1.3989	0.2222
1.116	30.0	180	1.3985	0.2222
1.116	31.0	186	1.3976	0.2444
1.153	32.0	192	1.3972	0.2444
1.153	33.0	198	1.3964	0.2444
1.1437	34.0	204	1.3958	0.2444
1.1259	35.0	210	1.3954	0.2444
1.1259	36.0	216	1.3954	0.2667
1.1125	37.0	222	1.3951	0.2667
1.1125	38.0	228	1.3951	0.2667
1.0816	39.0	234	1.3948	0.2667
1.1207	40.0	240	1.3948	0.2667
1.1207	41.0	246	1.3947	0.2667
1.1291	42.0	252	1.3946	0.2667
1.1291	43.0	258	1.3946	0.2667
1.1338	44.0	264	1.3946	0.2667
1.1093	45.0	270	1.3946	0.2667
1.1093	46.0	276	1.3946	0.2667
1.1123	47.0	282	1.3946	0.2667
1.1123	48.0	288	1.3946	0.2667
1.096	49.0	294	1.3946	0.2667
1.1328	50.0	300	1.3946	0.2667

Framework versions

• Transformers 4.35.0
• Pytorch 2.1.0+cu118
• Datasets 2.14.6
• Tokenizers 0.14.1