smids_3x_deit_base_sgd_001_fold3

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: 0.2808
• Accuracy: 0.8867

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
0.957	1.0	225	0.9484	0.6067
0.7319	2.0	450	0.7399	0.7417
0.5675	3.0	675	0.6023	0.8
0.5071	4.0	900	0.5170	0.8183
0.4486	5.0	1125	0.4641	0.8317
0.3985	6.0	1350	0.4275	0.85
0.3723	7.0	1575	0.4010	0.8483
0.3723	8.0	1800	0.3821	0.8533
0.3669	9.0	2025	0.3688	0.8617
0.3526	10.0	2250	0.3586	0.8667
0.3131	11.0	2475	0.3476	0.8733
0.3182	12.0	2700	0.3438	0.8717
0.329	13.0	2925	0.3317	0.875
0.2936	14.0	3150	0.3264	0.8733
0.281	15.0	3375	0.3211	0.875
0.29	16.0	3600	0.3177	0.875
0.3306	17.0	3825	0.3133	0.8767
0.268	18.0	4050	0.3109	0.875
0.2813	19.0	4275	0.3073	0.8817
0.29	20.0	4500	0.3041	0.8833
0.2543	21.0	4725	0.3034	0.8833
0.2485	22.0	4950	0.3036	0.8833
0.2389	23.0	5175	0.2976	0.885
0.2585	24.0	5400	0.2973	0.885
0.2265	25.0	5625	0.2954	0.8867
0.2806	26.0	5850	0.2939	0.8867
0.2023	27.0	6075	0.2926	0.8867
0.2462	28.0	6300	0.2897	0.885
0.2516	29.0	6525	0.2898	0.8867
0.2125	30.0	6750	0.2889	0.8867
0.2359	31.0	6975	0.2881	0.8883
0.2277	32.0	7200	0.2865	0.8867
0.2555	33.0	7425	0.2855	0.8883
0.202	34.0	7650	0.2847	0.8883
0.2329	35.0	7875	0.2859	0.8883
0.262	36.0	8100	0.2851	0.8883
0.2072	37.0	8325	0.2831	0.8867
0.2187	38.0	8550	0.2836	0.8867
0.2311	39.0	8775	0.2822	0.8867
0.2221	40.0	9000	0.2824	0.8867
0.2051	41.0	9225	0.2822	0.8867
0.2267	42.0	9450	0.2815	0.8867
0.2187	43.0	9675	0.2815	0.8867
0.2138	44.0	9900	0.2815	0.8867
0.2129	45.0	10125	0.2813	0.8867
0.2034	46.0	10350	0.2810	0.8867
0.2287	47.0	10575	0.2808	0.8867
0.2051	48.0	10800	0.2807	0.8867
0.256	49.0	11025	0.2807	0.8867
0.2173	50.0	11250	0.2808	0.8867

Framework versions

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