hushem_1x_deit_small_rms_0001_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: 1.6051
• Accuracy: 0.6279

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	1.6050	0.2326
1.8871	2.0	12	1.6526	0.2558
1.8871	3.0	18	1.5125	0.3488
1.4752	4.0	24	1.4964	0.2558
1.4078	5.0	30	1.4513	0.2558
1.4078	6.0	36	1.3471	0.2791
1.2466	7.0	42	1.9402	0.3023
1.2466	8.0	48	1.3364	0.3953
1.2272	9.0	54	1.3189	0.3721
1.0035	10.0	60	1.3183	0.4186
1.0035	11.0	66	1.1773	0.4419
0.7718	12.0	72	1.0313	0.5814
0.7718	13.0	78	1.0831	0.6279
0.4478	14.0	84	1.4838	0.5814
0.14	15.0	90	1.1904	0.6744
0.14	16.0	96	1.2473	0.6512
0.0804	17.0	102	1.4013	0.6977
0.0804	18.0	108	1.4032	0.6512
0.0101	19.0	114	1.4918	0.6977
0.0016	20.0	120	1.4874	0.6279
0.0016	21.0	126	1.4977	0.6279
0.0007	22.0	132	1.5076	0.6279
0.0007	23.0	138	1.5163	0.6512
0.0006	24.0	144	1.5254	0.6512
0.0005	25.0	150	1.5330	0.6512
0.0005	26.0	156	1.5401	0.6512
0.0004	27.0	162	1.5491	0.6279
0.0004	28.0	168	1.5572	0.6279
0.0004	29.0	174	1.5632	0.6279
0.0003	30.0	180	1.5688	0.6279
0.0003	31.0	186	1.5748	0.6279
0.0003	32.0	192	1.5796	0.6279
0.0003	33.0	198	1.5848	0.6279
0.0003	34.0	204	1.5896	0.6279
0.0003	35.0	210	1.5928	0.6279
0.0003	36.0	216	1.5963	0.6279
0.0003	37.0	222	1.5989	0.6279
0.0003	38.0	228	1.6012	0.6279
0.0003	39.0	234	1.6030	0.6279
0.0002	40.0	240	1.6043	0.6279
0.0002	41.0	246	1.6049	0.6279
0.0002	42.0	252	1.6051	0.6279
0.0002	43.0	258	1.6051	0.6279
0.0002	44.0	264	1.6051	0.6279
0.0002	45.0	270	1.6051	0.6279
0.0002	46.0	276	1.6051	0.6279
0.0002	47.0	282	1.6051	0.6279
0.0002	48.0	288	1.6051	0.6279
0.0002	49.0	294	1.6051	0.6279
0.0002	50.0	300	1.6051	0.6279

Framework versions

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