hushem_1x_deit_tiny_sgd_lr001_fold2

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: 1.4182
• Accuracy: 0.3556

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	1.6125	0.1333
1.5852	2.0	12	1.5871	0.1333
1.5852	3.0	18	1.5653	0.1556
1.5028	4.0	24	1.5474	0.1556
1.4795	5.0	30	1.5322	0.1556
1.4795	6.0	36	1.5188	0.1556
1.4252	7.0	42	1.5071	0.1556
1.4252	8.0	48	1.4989	0.1556
1.3707	9.0	54	1.4901	0.1778
1.365	10.0	60	1.4824	0.2
1.365	11.0	66	1.4748	0.2
1.3235	12.0	72	1.4694	0.2444
1.3235	13.0	78	1.4635	0.2444
1.3233	14.0	84	1.4596	0.2444
1.2774	15.0	90	1.4554	0.2444
1.2774	16.0	96	1.4518	0.2444
1.2584	17.0	102	1.4482	0.2667
1.2584	18.0	108	1.4450	0.2667
1.2788	19.0	114	1.4423	0.2667
1.2388	20.0	120	1.4398	0.2667
1.2388	21.0	126	1.4370	0.2889
1.2317	22.0	132	1.4351	0.2667
1.2317	23.0	138	1.4327	0.2889
1.2286	24.0	144	1.4312	0.2889
1.2033	25.0	150	1.4298	0.2889
1.2033	26.0	156	1.4283	0.3111
1.1965	27.0	162	1.4267	0.3111
1.1965	28.0	168	1.4258	0.3111
1.1963	29.0	174	1.4246	0.3111
1.1946	30.0	180	1.4236	0.3111
1.1946	31.0	186	1.4227	0.3333
1.1805	32.0	192	1.4218	0.3556
1.1805	33.0	198	1.4211	0.3556
1.1439	34.0	204	1.4203	0.3556
1.1699	35.0	210	1.4197	0.3556
1.1699	36.0	216	1.4193	0.3556
1.156	37.0	222	1.4190	0.3556
1.156	38.0	228	1.4187	0.3556
1.1475	39.0	234	1.4185	0.3556
1.1517	40.0	240	1.4183	0.3556
1.1517	41.0	246	1.4182	0.3556
1.1468	42.0	252	1.4182	0.3556
1.1468	43.0	258	1.4182	0.3556
1.1597	44.0	264	1.4182	0.3556
1.1542	45.0	270	1.4182	0.3556
1.1542	46.0	276	1.4182	0.3556
1.1604	47.0	282	1.4182	0.3556
1.1604	48.0	288	1.4182	0.3556
1.1576	49.0	294	1.4182	0.3556
1.143	50.0	300	1.4182	0.3556

Framework versions

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