smids_3x_deit_base_adamax_0001_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: 0.7119
• Accuracy: 0.9098

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
0.3221	1.0	226	0.2895	0.8915
0.1579	2.0	452	0.3905	0.8664
0.0813	3.0	678	0.6613	0.8397
0.039	4.0	904	0.5507	0.8915
0.0377	5.0	1130	0.4323	0.9015
0.0016	6.0	1356	0.4901	0.9032
0.0006	7.0	1582	0.4428	0.9182
0.0001	8.0	1808	0.4887	0.9165
0.0056	9.0	2034	0.5309	0.9132
0.0002	10.0	2260	0.5356	0.9115
0.0005	11.0	2486	0.6183	0.8998
0.0	12.0	2712	0.5569	0.9032
0.0218	13.0	2938	0.6429	0.9048
0.0047	14.0	3164	0.6294	0.9015
0.0001	15.0	3390	0.6584	0.8982
0.0	16.0	3616	0.7150	0.8915
0.003	17.0	3842	0.7055	0.9032
0.0	18.0	4068	0.7249	0.9032
0.0033	19.0	4294	0.7504	0.8998
0.0056	20.0	4520	0.7091	0.9032
0.0	21.0	4746	0.7020	0.9082
0.0027	22.0	4972	0.6539	0.9082
0.0024	23.0	5198	0.6895	0.9065
0.0	24.0	5424	0.7110	0.9048
0.0	25.0	5650	0.6678	0.9065
0.0	26.0	5876	0.6869	0.9048
0.0	27.0	6102	0.6975	0.9065
0.0	28.0	6328	0.7125	0.9082
0.0	29.0	6554	0.6594	0.9082
0.0	30.0	6780	0.6721	0.9115
0.0	31.0	7006	0.7233	0.9065
0.0	32.0	7232	0.7025	0.9082
0.0029	33.0	7458	0.7410	0.9032
0.0	34.0	7684	0.6915	0.9065
0.0	35.0	7910	0.6943	0.9032
0.0	36.0	8136	0.6848	0.9098
0.0	37.0	8362	0.6997	0.9082
0.0	38.0	8588	0.6917	0.9082
0.0	39.0	8814	0.7108	0.9065
0.0	40.0	9040	0.6898	0.9115
0.0026	41.0	9266	0.6995	0.9082
0.0027	42.0	9492	0.7042	0.9082
0.0	43.0	9718	0.7153	0.9082
0.0	44.0	9944	0.7075	0.9065
0.0	45.0	10170	0.7108	0.9065
0.0	46.0	10396	0.7067	0.9098
0.0	47.0	10622	0.7091	0.9098
0.0	48.0	10848	0.7104	0.9082
0.0	49.0	11074	0.7114	0.9098
0.0	50.0	11300	0.7119	0.9098

Framework versions

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