hushem_1x_deit_base_rms_001_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: 1.2820
• Accuracy: 0.5714

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	6.4622	0.2381
4.4464	2.0	12	3.3632	0.2381
4.4464	3.0	18	1.7517	0.2619
2.081	4.0	24	1.7192	0.2381
1.6572	5.0	30	1.4331	0.2619
1.6572	6.0	36	1.7438	0.2381
1.515	7.0	42	1.5334	0.2381
1.515	8.0	48	1.4427	0.2619
1.4729	9.0	54	1.4737	0.2619
1.4733	10.0	60	1.3911	0.2381
1.4733	11.0	66	1.4837	0.2619
1.4345	12.0	72	1.3420	0.3333
1.4345	13.0	78	1.3532	0.3095
1.37	14.0	84	1.2042	0.5714
1.3487	15.0	90	1.2734	0.3095
1.3487	16.0	96	1.1311	0.4286
1.2658	17.0	102	1.1548	0.4762
1.2658	18.0	108	1.2031	0.3571
1.2896	19.0	114	1.2313	0.4762
1.2598	20.0	120	1.3330	0.3810
1.2598	21.0	126	1.1274	0.5238
1.2329	22.0	132	1.2033	0.5238
1.2329	23.0	138	1.1130	0.5
1.2013	24.0	144	1.1588	0.5
1.1821	25.0	150	1.3546	0.3810
1.1821	26.0	156	1.1188	0.4524
1.1779	27.0	162	1.1678	0.3571
1.1779	28.0	168	1.2401	0.3333
1.1114	29.0	174	1.0781	0.5476
1.1371	30.0	180	1.0969	0.5476
1.1371	31.0	186	1.2482	0.4762
1.0827	32.0	192	1.0695	0.5
1.0827	33.0	198	1.2349	0.4762
1.1051	34.0	204	1.1006	0.5238
1.0196	35.0	210	1.0684	0.5476
1.0196	36.0	216	0.9937	0.5238
1.0022	37.0	222	1.2962	0.5238
1.0022	38.0	228	1.0911	0.5476
0.9974	39.0	234	1.1681	0.5
0.9742	40.0	240	1.3925	0.5476
0.9742	41.0	246	1.2876	0.5714
0.9338	42.0	252	1.2820	0.5714
0.9338	43.0	258	1.2820	0.5714
0.8995	44.0	264	1.2820	0.5714
0.9368	45.0	270	1.2820	0.5714
0.9368	46.0	276	1.2820	0.5714
0.9236	47.0	282	1.2820	0.5714
0.9236	48.0	288	1.2820	0.5714
0.929	49.0	294	1.2820	0.5714
0.9198	50.0	300	1.2820	0.5714

Framework versions

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