smids_5x_deit_base_adamax_0001_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: 0.9417
• Accuracy: 0.9067

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
0.2147	1.0	375	0.4092	0.8317
0.1408	2.0	750	0.3269	0.915
0.0228	3.0	1125	0.4567	0.9067
0.0111	4.0	1500	0.5788	0.9033
0.0156	5.0	1875	0.6062	0.9
0.0288	6.0	2250	0.6656	0.8917
0.0007	7.0	2625	0.6456	0.9017
0.0002	8.0	3000	0.6407	0.8917
0.0002	9.0	3375	0.6824	0.9083
0.0084	10.0	3750	0.6593	0.905
0.0203	11.0	4125	0.7617	0.9017
0.0033	12.0	4500	0.7022	0.9167
0.0	13.0	4875	0.8023	0.9033
0.0	14.0	5250	0.8062	0.9083
0.0	15.0	5625	0.8735	0.905
0.0293	16.0	6000	0.8124	0.9133
0.0	17.0	6375	0.8110	0.915
0.0	18.0	6750	0.7934	0.9167
0.0	19.0	7125	0.8257	0.9117
0.0	20.0	7500	0.8169	0.905
0.0	21.0	7875	0.7971	0.9167
0.0	22.0	8250	0.8206	0.905
0.0031	23.0	8625	0.8887	0.9067
0.0	24.0	9000	0.8570	0.91
0.0	25.0	9375	0.9027	0.9017
0.0	26.0	9750	0.8809	0.9067
0.0	27.0	10125	0.8772	0.9083
0.0	28.0	10500	0.8815	0.9083
0.0	29.0	10875	0.8462	0.91
0.0028	30.0	11250	0.8854	0.9083
0.0	31.0	11625	0.8584	0.9083
0.0	32.0	12000	0.8933	0.905
0.0	33.0	12375	0.8718	0.9083
0.0	34.0	12750	0.8798	0.9067
0.0	35.0	13125	0.8653	0.9083
0.0	36.0	13500	0.8742	0.9133
0.0	37.0	13875	0.8914	0.9083
0.0	38.0	14250	0.8921	0.91
0.0	39.0	14625	0.9001	0.9083
0.0025	40.0	15000	0.9101	0.9083
0.0	41.0	15375	0.9161	0.9067
0.0	42.0	15750	0.9182	0.9083
0.0	43.0	16125	0.9246	0.905
0.0	44.0	16500	0.9291	0.9083
0.0	45.0	16875	0.9302	0.9067
0.0	46.0	17250	0.9341	0.9067
0.0	47.0	17625	0.9378	0.9067
0.0	48.0	18000	0.9402	0.9067
0.0	49.0	18375	0.9417	0.9067
0.0	50.0	18750	0.9417	0.9067

Framework versions

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