hushem_40x_deit_base_sgd_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: 0.2186
• Accuracy: 0.9286

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
1.2742	1.0	219	1.3421	0.4286
1.1364	2.0	438	1.2693	0.4048
0.9912	3.0	657	1.1701	0.5238
0.8292	4.0	876	1.0493	0.6429
0.7771	5.0	1095	0.9148	0.7143
0.6956	6.0	1314	0.8048	0.7381
0.519	7.0	1533	0.7062	0.8095
0.5042	8.0	1752	0.6401	0.7857
0.4397	9.0	1971	0.5785	0.8333
0.3933	10.0	2190	0.5338	0.8571
0.341	11.0	2409	0.4959	0.8810
0.3345	12.0	2628	0.4569	0.8810
0.2949	13.0	2847	0.4265	0.9048
0.2608	14.0	3066	0.3999	0.9286
0.2368	15.0	3285	0.3796	0.9286
0.2257	16.0	3504	0.3614	0.9286
0.232	17.0	3723	0.3430	0.9286
0.1928	18.0	3942	0.3249	0.9286
0.1804	19.0	4161	0.3144	0.9286
0.1542	20.0	4380	0.3019	0.9048
0.1333	21.0	4599	0.2915	0.9286
0.1333	22.0	4818	0.2894	0.9048
0.1178	23.0	5037	0.2746	0.9286
0.1098	24.0	5256	0.2771	0.9048
0.1099	25.0	5475	0.2649	0.9048
0.0836	26.0	5694	0.2732	0.9048
0.0751	27.0	5913	0.2625	0.9048
0.0745	28.0	6132	0.2608	0.9048
0.0826	29.0	6351	0.2526	0.9048
0.079	30.0	6570	0.2463	0.9286
0.0659	31.0	6789	0.2439	0.9048
0.0738	32.0	7008	0.2422	0.9286
0.0683	33.0	7227	0.2335	0.9286
0.0674	34.0	7446	0.2343	0.9048
0.0633	35.0	7665	0.2311	0.9048
0.0608	36.0	7884	0.2259	0.9286
0.0543	37.0	8103	0.2239	0.9286
0.0444	38.0	8322	0.2256	0.9286
0.0496	39.0	8541	0.2255	0.9286
0.0513	40.0	8760	0.2253	0.9286
0.0449	41.0	8979	0.2226	0.9286
0.0449	42.0	9198	0.2216	0.9286
0.0549	43.0	9417	0.2202	0.9286
0.0488	44.0	9636	0.2213	0.9286
0.0437	45.0	9855	0.2208	0.9286
0.0362	46.0	10074	0.2201	0.9286
0.0622	47.0	10293	0.2188	0.9286
0.0546	48.0	10512	0.2185	0.9286
0.0472	49.0	10731	0.2186	0.9286
0.0581	50.0	10950	0.2186	0.9286

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.0+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2