hushem_5x_deit_base_sgd_001_fold1

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: 1.2935
• Accuracy: 0.4222

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
1.3985	1.0	27	1.4731	0.2444
1.3797	2.0	54	1.4540	0.2667
1.3671	3.0	81	1.4399	0.3333
1.3529	4.0	108	1.4301	0.3333
1.3075	5.0	135	1.4206	0.3778
1.3006	6.0	162	1.4113	0.3778
1.2955	7.0	189	1.4036	0.3778
1.2684	8.0	216	1.3964	0.4
1.2547	9.0	243	1.3899	0.4
1.2309	10.0	270	1.3835	0.4
1.2188	11.0	297	1.3776	0.3778
1.1974	12.0	324	1.3722	0.3778
1.1972	13.0	351	1.3669	0.4
1.1775	14.0	378	1.3615	0.3778
1.1771	15.0	405	1.3571	0.3778
1.1595	16.0	432	1.3529	0.3778
1.11	17.0	459	1.3491	0.4
1.116	18.0	486	1.3456	0.4
1.0955	19.0	513	1.3420	0.4
1.0866	20.0	540	1.3386	0.4
1.0678	21.0	567	1.3355	0.4
1.0655	22.0	594	1.3327	0.4
1.0356	23.0	621	1.3298	0.4
1.0185	24.0	648	1.3265	0.3778
1.0437	25.0	675	1.3237	0.4
1.0442	26.0	702	1.3211	0.3778
1.028	27.0	729	1.3185	0.3778
1.0044	28.0	756	1.3165	0.3778
1.002	29.0	783	1.3148	0.4
0.9934	30.0	810	1.3131	0.4
0.9758	31.0	837	1.3109	0.4
0.9861	32.0	864	1.3087	0.4
0.9889	33.0	891	1.3069	0.4
0.9637	34.0	918	1.3052	0.4
0.9733	35.0	945	1.3034	0.4
0.9304	36.0	972	1.3021	0.4222
0.9586	37.0	999	1.3007	0.4222
0.9329	38.0	1026	1.2994	0.4222
0.918	39.0	1053	1.2983	0.4222
0.9142	40.0	1080	1.2972	0.4222
0.9236	41.0	1107	1.2963	0.4222
0.929	42.0	1134	1.2957	0.4222
0.9525	43.0	1161	1.2951	0.4222
0.8934	44.0	1188	1.2944	0.4222
0.9348	45.0	1215	1.2941	0.4222
0.9068	46.0	1242	1.2937	0.4222
0.9064	47.0	1269	1.2936	0.4222
0.9044	48.0	1296	1.2934	0.4222
0.9396	49.0	1323	1.2935	0.4222
0.894	50.0	1350	1.2935	0.4222

Framework versions

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