smids_5x_deit_tiny_adamax_001_fold3

This model is a fine-tuned version of facebook/deit-small-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.0439
• Accuracy: 0.905

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
0.3308	1.0	375	0.3353	0.875
0.2337	2.0	750	0.3320	0.8817
0.1696	3.0	1125	0.3479	0.8783
0.1669	4.0	1500	0.3755	0.8767
0.1864	5.0	1875	0.3099	0.8983
0.1212	6.0	2250	0.3912	0.91
0.1119	7.0	2625	0.4167	0.8817
0.1024	8.0	3000	0.4153	0.8733
0.0484	9.0	3375	0.5188	0.8733
0.0551	10.0	3750	0.6042	0.885
0.0468	11.0	4125	0.6570	0.8767
0.0137	12.0	4500	0.6069	0.8733
0.0177	13.0	4875	0.7091	0.8817
0.0201	14.0	5250	0.7010	0.89
0.0183	15.0	5625	0.6654	0.8867
0.0149	16.0	6000	0.7079	0.8883
0.0065	17.0	6375	0.6100	0.8933
0.001	18.0	6750	0.9491	0.8817
0.0034	19.0	7125	0.8269	0.8833
0.0213	20.0	7500	0.8028	0.8833
0.0137	21.0	7875	0.7227	0.8933
0.0	22.0	8250	0.8796	0.8917
0.0014	23.0	8625	0.8924	0.8733
0.0002	24.0	9000	0.6942	0.8917
0.0	25.0	9375	0.7445	0.89
0.0	26.0	9750	0.7840	0.885
0.0103	27.0	10125	0.7469	0.9033
0.0	28.0	10500	0.8867	0.8783
0.0	29.0	10875	0.8617	0.8867
0.003	30.0	11250	0.8295	0.8983
0.0008	31.0	11625	0.9061	0.895
0.0	32.0	12000	0.8630	0.8967
0.0	33.0	12375	0.8010	0.9017
0.0	34.0	12750	0.8248	0.8983
0.0	35.0	13125	0.8438	0.91
0.0	36.0	13500	0.9235	0.9
0.0	37.0	13875	0.8167	0.9083
0.0	38.0	14250	0.8531	0.9033
0.0	39.0	14625	0.9035	0.9067
0.0027	40.0	15000	0.9614	0.9017
0.0	41.0	15375	0.9740	0.9017
0.0	42.0	15750	0.9907	0.9
0.0	43.0	16125	0.9964	0.9033
0.0	44.0	16500	1.0084	0.9033
0.0	45.0	16875	1.0215	0.905
0.0	46.0	17250	1.0234	0.9017
0.0	47.0	17625	1.0315	0.905
0.0	48.0	18000	1.0372	0.905
0.0	49.0	18375	1.0413	0.905
0.0	50.0	18750	1.0439	0.905

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.1+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2