hushem_1x_deit_small_rms_00001_fold3

This model is a fine-tuned version of facebook/deit-small-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 0.3985
• Accuracy: 0.8372

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: 1e-05
• 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	1.1600	0.4419
1.1403	2.0	12	0.8608	0.6744
1.1403	3.0	18	0.6312	0.7907
0.44	4.0	24	0.5560	0.7442
0.137	5.0	30	0.6293	0.6977
0.137	6.0	36	0.5955	0.7442
0.03	7.0	42	0.4797	0.8372
0.03	8.0	48	0.3954	0.8140
0.0079	9.0	54	0.4175	0.8372
0.0043	10.0	60	0.4040	0.7907
0.0043	11.0	66	0.4128	0.8372
0.0029	12.0	72	0.4075	0.8372
0.0029	13.0	78	0.4003	0.8372
0.0022	14.0	84	0.3993	0.8140
0.0018	15.0	90	0.3966	0.8140
0.0018	16.0	96	0.4005	0.8372
0.0015	17.0	102	0.4011	0.8372
0.0015	18.0	108	0.3998	0.8372
0.0013	19.0	114	0.3985	0.8372
0.0012	20.0	120	0.3993	0.8140
0.0012	21.0	126	0.3972	0.8372
0.0011	22.0	132	0.4006	0.8605
0.0011	23.0	138	0.3962	0.8372
0.001	24.0	144	0.3991	0.8605
0.0009	25.0	150	0.3957	0.8140
0.0009	26.0	156	0.3974	0.8372
0.0008	27.0	162	0.3957	0.8372
0.0008	28.0	168	0.3962	0.8372
0.0008	29.0	174	0.3950	0.8372
0.0007	30.0	180	0.3967	0.8372
0.0007	31.0	186	0.3962	0.8372
0.0007	32.0	192	0.3971	0.8372
0.0007	33.0	198	0.3980	0.8372
0.0007	34.0	204	0.3974	0.8372
0.0006	35.0	210	0.3977	0.8372
0.0006	36.0	216	0.3977	0.8372
0.0006	37.0	222	0.3981	0.8372
0.0006	38.0	228	0.3981	0.8372
0.0006	39.0	234	0.3981	0.8372
0.0006	40.0	240	0.3984	0.8372
0.0006	41.0	246	0.3985	0.8372
0.0006	42.0	252	0.3985	0.8372
0.0006	43.0	258	0.3985	0.8372
0.0006	44.0	264	0.3985	0.8372
0.0006	45.0	270	0.3985	0.8372
0.0006	46.0	276	0.3985	0.8372
0.0006	47.0	282	0.3985	0.8372
0.0006	48.0	288	0.3985	0.8372
0.0006	49.0	294	0.3985	0.8372
0.0006	50.0	300	0.3985	0.8372

Framework versions

• Transformers 4.35.1
• Pytorch 2.1.0+cu118
• Datasets 2.14.7
• Tokenizers 0.14.1