smids_5x_deit_base_adamax_0001_fold5

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: 0.9013
• Accuracy: 0.9017

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
0.1558	1.0	375	0.3339	0.8717
0.0801	2.0	750	0.2674	0.9
0.064	3.0	1125	0.4397	0.9033
0.0238	4.0	1500	0.5225	0.875
0.0319	5.0	1875	0.5721	0.905
0.0056	6.0	2250	0.5081	0.9083
0.0107	7.0	2625	0.5806	0.91
0.0108	8.0	3000	0.6004	0.9067
0.0023	9.0	3375	0.7259	0.895
0.0005	10.0	3750	0.7347	0.9033
0.0001	11.0	4125	0.7841	0.8967
0.0	12.0	4500	0.7216	0.9167
0.0	13.0	4875	0.7364	0.9083
0.0053	14.0	5250	0.7059	0.9067
0.0001	15.0	5625	0.7607	0.9017
0.0	16.0	6000	0.7546	0.9083
0.0001	17.0	6375	0.7848	0.9033
0.0042	18.0	6750	0.7392	0.8983
0.0	19.0	7125	0.7453	0.9183
0.0	20.0	7500	0.8298	0.9067
0.0	21.0	7875	0.8069	0.9067
0.0038	22.0	8250	0.7995	0.9067
0.0	23.0	8625	0.8015	0.91
0.0	24.0	9000	0.8099	0.9067
0.0	25.0	9375	0.7950	0.9117
0.0	26.0	9750	0.8272	0.91
0.0	27.0	10125	0.7940	0.905
0.0	28.0	10500	0.8281	0.915
0.0	29.0	10875	0.8337	0.9067
0.0031	30.0	11250	0.8245	0.9067
0.0	31.0	11625	0.8597	0.9033
0.0	32.0	12000	0.8445	0.9067
0.0	33.0	12375	0.8424	0.9033
0.0	34.0	12750	0.8455	0.9017
0.0	35.0	13125	0.8539	0.9017
0.0	36.0	13500	0.8610	0.8967
0.0	37.0	13875	0.8681	0.905
0.0026	38.0	14250	0.8625	0.9017
0.0	39.0	14625	0.8694	0.9067
0.0	40.0	15000	0.8718	0.9
0.0	41.0	15375	0.8794	0.905
0.0	42.0	15750	0.8824	0.9
0.0	43.0	16125	0.8842	0.905
0.0	44.0	16500	0.8874	0.9017
0.0	45.0	16875	0.8897	0.9017
0.0	46.0	17250	0.8954	0.9017
0.0025	47.0	17625	0.8975	0.9017
0.0	48.0	18000	0.9000	0.9017
0.0	49.0	18375	0.9014	0.9017
0.0023	50.0	18750	0.9013	0.9017

Framework versions

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