hushem_1x_deit_base_rms_0001_fold4

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

• Loss: 0.7145
• Accuracy: 0.8095

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	1.4829	0.2619
1.8031	2.0	12	2.0915	0.2381
1.8031	3.0	18	1.4809	0.2619
1.5525	4.0	24	1.5704	0.2381
1.3693	5.0	30	1.0775	0.5476
1.3693	6.0	36	1.1325	0.7143
1.2011	7.0	42	2.3370	0.2381
1.2011	8.0	48	0.7067	0.6667
0.8987	9.0	54	0.5307	0.7619
0.4373	10.0	60	1.1564	0.6429
0.4373	11.0	66	0.6814	0.7619
0.23	12.0	72	0.7840	0.6905
0.23	13.0	78	1.0320	0.6429
0.0568	14.0	84	1.3932	0.6667
0.0426	15.0	90	0.6994	0.7619
0.0426	16.0	96	0.6428	0.8095
0.0009	17.0	102	0.6585	0.8095
0.0009	18.0	108	0.6712	0.8095
0.0005	19.0	114	0.6782	0.8095
0.0003	20.0	120	0.6859	0.8095
0.0003	21.0	126	0.6888	0.8095
0.0003	22.0	132	0.6932	0.8095
0.0003	23.0	138	0.6960	0.8095
0.0002	24.0	144	0.7001	0.8095
0.0002	25.0	150	0.7019	0.8095
0.0002	26.0	156	0.7021	0.8095
0.0002	27.0	162	0.7042	0.8095
0.0002	28.0	168	0.7068	0.8095
0.0002	29.0	174	0.7077	0.8095
0.0002	30.0	180	0.7092	0.8095
0.0002	31.0	186	0.7093	0.8095
0.0002	32.0	192	0.7100	0.8095
0.0002	33.0	198	0.7106	0.8095
0.0001	34.0	204	0.7108	0.8095
0.0001	35.0	210	0.7121	0.8095
0.0001	36.0	216	0.7125	0.8095
0.0001	37.0	222	0.7133	0.8095
0.0001	38.0	228	0.7137	0.8095
0.0001	39.0	234	0.7141	0.8095
0.0001	40.0	240	0.7143	0.8095
0.0001	41.0	246	0.7145	0.8095
0.0001	42.0	252	0.7145	0.8095
0.0001	43.0	258	0.7145	0.8095
0.0001	44.0	264	0.7145	0.8095
0.0001	45.0	270	0.7145	0.8095
0.0001	46.0	276	0.7145	0.8095
0.0001	47.0	282	0.7145	0.8095
0.0001	48.0	288	0.7145	0.8095
0.0001	49.0	294	0.7145	0.8095
0.0001	50.0	300	0.7145	0.8095

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0