hushem_40x_deit_base_sgd_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: 1.2151
• Accuracy: 0.4286

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
1.3918	1.0	219	1.4045	0.3095
1.3704	2.0	438	1.3956	0.3095
1.3491	3.0	657	1.3880	0.3333
1.3369	4.0	876	1.3811	0.3333
1.3406	5.0	1095	1.3747	0.3333
1.3171	6.0	1314	1.3686	0.3333
1.2982	7.0	1533	1.3628	0.3571
1.2896	8.0	1752	1.3571	0.3571
1.2549	9.0	1971	1.3513	0.3810
1.2384	10.0	2190	1.3457	0.4048
1.2507	11.0	2409	1.3401	0.4286
1.2362	12.0	2628	1.3346	0.4286
1.1966	13.0	2847	1.3293	0.4286
1.2279	14.0	3066	1.3240	0.4286
1.2136	15.0	3285	1.3188	0.4286
1.1856	16.0	3504	1.3138	0.4286
1.1941	17.0	3723	1.3088	0.4286
1.1805	18.0	3942	1.3039	0.4286
1.1554	19.0	4161	1.2991	0.4048
1.1709	20.0	4380	1.2943	0.4048
1.1523	21.0	4599	1.2895	0.4048
1.138	22.0	4818	1.2848	0.4048
1.0984	23.0	5037	1.2803	0.4048
1.1405	24.0	5256	1.2759	0.4048
1.1028	25.0	5475	1.2716	0.4286
1.1236	26.0	5694	1.2674	0.4286
1.0819	27.0	5913	1.2634	0.4286
1.1245	28.0	6132	1.2595	0.4286
1.0929	29.0	6351	1.2557	0.4286
1.0861	30.0	6570	1.2521	0.4048
1.082	31.0	6789	1.2486	0.4048
1.0826	32.0	7008	1.2452	0.4048
1.0889	33.0	7227	1.2420	0.4048
1.052	34.0	7446	1.2390	0.4286
1.056	35.0	7665	1.2361	0.4286
1.0391	36.0	7884	1.2333	0.4286
1.0236	37.0	8103	1.2307	0.4286
1.0474	38.0	8322	1.2283	0.4286
1.0069	39.0	8541	1.2261	0.4286
1.0443	40.0	8760	1.2242	0.4286
1.0711	41.0	8979	1.2223	0.4048
1.053	42.0	9198	1.2207	0.4286
1.0356	43.0	9417	1.2193	0.4286
1.0491	44.0	9636	1.2181	0.4286
0.9928	45.0	9855	1.2171	0.4286
1.0402	46.0	10074	1.2163	0.4286
1.0792	47.0	10293	1.2157	0.4286
1.0146	48.0	10512	1.2153	0.4286
1.0325	49.0	10731	1.2152	0.4286
1.0249	50.0	10950	1.2151	0.4286

Framework versions

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