hkivancoral/hushem_1x_deit_tiny_sgd_lr0001_fold3

hushem_1x_deit_sgd_adamax_lr0001_fold3

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: 1.4906
• Accuracy: 0.3023

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: 0.0001
• 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	1.5410	0.3023
1.6256	2.0	12	1.5383	0.3023
1.6256	3.0	18	1.5355	0.3023
1.6209	4.0	24	1.5329	0.3023
1.6213	5.0	30	1.5304	0.3023
1.6213	6.0	36	1.5279	0.3023
1.6054	7.0	42	1.5256	0.3023
1.6054	8.0	48	1.5235	0.3023
1.622	9.0	54	1.5216	0.3023
1.5637	10.0	60	1.5195	0.3023
1.5637	11.0	66	1.5177	0.3023
1.5868	12.0	72	1.5160	0.3023
1.5868	13.0	78	1.5142	0.3023
1.5951	14.0	84	1.5125	0.3023
1.5787	15.0	90	1.5108	0.3023
1.5787	16.0	96	1.5091	0.3023
1.5727	17.0	102	1.5077	0.3023
1.5727	18.0	108	1.5063	0.3023
1.5858	19.0	114	1.5049	0.3023
1.5652	20.0	120	1.5036	0.3023
1.5652	21.0	126	1.5024	0.3023
1.5577	22.0	132	1.5012	0.3023
1.5577	23.0	138	1.5001	0.3023
1.5855	24.0	144	1.4991	0.3023
1.5594	25.0	150	1.4981	0.3023
1.5594	26.0	156	1.4972	0.3023
1.5496	27.0	162	1.4964	0.3023
1.5496	28.0	168	1.4956	0.3023
1.5543	29.0	174	1.4949	0.3023
1.5415	30.0	180	1.4943	0.3023
1.5415	31.0	186	1.4938	0.3023
1.5408	32.0	192	1.4932	0.3023
1.5408	33.0	198	1.4926	0.3023
1.5602	34.0	204	1.4922	0.3023
1.5429	35.0	210	1.4918	0.3023
1.5429	36.0	216	1.4914	0.3023
1.5494	37.0	222	1.4912	0.3023
1.5494	38.0	228	1.4909	0.3023
1.5361	39.0	234	1.4908	0.3023
1.5628	40.0	240	1.4906	0.3023
1.5628	41.0	246	1.4906	0.3023
1.5458	42.0	252	1.4906	0.3023
1.5458	43.0	258	1.4906	0.3023
1.5716	44.0	264	1.4906	0.3023
1.5384	45.0	270	1.4906	0.3023
1.5384	46.0	276	1.4906	0.3023
1.5475	47.0	282	1.4906	0.3023
1.5475	48.0	288	1.4906	0.3023
1.5338	49.0	294	1.4906	0.3023
1.5337	50.0	300	1.4906	0.3023

Framework versions

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