hushem_1x_deit_tiny_sgd_001_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: 1.4913
• Accuracy: 0.1778

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	1.6461	0.2222
1.647	2.0	12	1.5827	0.2
1.647	3.0	18	1.5400	0.2
1.5111	4.0	24	1.5101	0.2
1.4472	5.0	30	1.4855	0.1778
1.4472	6.0	36	1.4711	0.1778
1.3765	7.0	42	1.4618	0.2
1.3765	8.0	48	1.4555	0.2
1.3363	9.0	54	1.4523	0.2222
1.3131	10.0	60	1.4505	0.2
1.3131	11.0	66	1.4495	0.2
1.2743	12.0	72	1.4504	0.2
1.2743	13.0	78	1.4505	0.2
1.2923	14.0	84	1.4516	0.2
1.2475	15.0	90	1.4529	0.2
1.2475	16.0	96	1.4558	0.2
1.2052	17.0	102	1.4591	0.1778
1.2052	18.0	108	1.4603	0.1778
1.2375	19.0	114	1.4628	0.1778
1.1665	20.0	120	1.4654	0.1778
1.1665	21.0	126	1.4668	0.1778
1.1508	22.0	132	1.4681	0.1778
1.1508	23.0	138	1.4710	0.1778
1.1615	24.0	144	1.4735	0.1778
1.1372	25.0	150	1.4742	0.1778
1.1372	26.0	156	1.4775	0.1778
1.1389	27.0	162	1.4787	0.1778
1.1389	28.0	168	1.4813	0.1778
1.1191	29.0	174	1.4821	0.1778
1.106	30.0	180	1.4844	0.1778
1.106	31.0	186	1.4853	0.1778
1.1156	32.0	192	1.4867	0.1778
1.1156	33.0	198	1.4872	0.1778
1.127	34.0	204	1.4879	0.1778
1.1055	35.0	210	1.4887	0.1778
1.1055	36.0	216	1.4895	0.1778
1.089	37.0	222	1.4902	0.1778
1.089	38.0	228	1.4907	0.1778
1.0605	39.0	234	1.4911	0.1778
1.0925	40.0	240	1.4913	0.1778
1.0925	41.0	246	1.4913	0.1778
1.1025	42.0	252	1.4913	0.1778
1.1025	43.0	258	1.4913	0.1778
1.1085	44.0	264	1.4913	0.1778
1.0909	45.0	270	1.4913	0.1778
1.0909	46.0	276	1.4913	0.1778
1.0889	47.0	282	1.4913	0.1778
1.0889	48.0	288	1.4913	0.1778
1.0611	49.0	294	1.4913	0.1778
1.1045	50.0	300	1.4913	0.1778

Framework versions

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