smids_10x_deit_base_sgd_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: 0.3787
• Accuracy: 0.835

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
1.0636	1.0	750	1.0738	0.4167
0.9993	2.0	1500	1.0277	0.5117
0.9435	3.0	2250	0.9677	0.5817
0.8744	4.0	3000	0.8995	0.6383
0.8256	5.0	3750	0.8300	0.705
0.7543	6.0	4500	0.7676	0.73
0.673	7.0	5250	0.7129	0.76
0.6641	8.0	6000	0.6669	0.7717
0.6662	9.0	6750	0.6286	0.7817
0.612	10.0	7500	0.5968	0.79
0.5807	11.0	8250	0.5703	0.7933
0.5436	12.0	9000	0.5477	0.8
0.5472	13.0	9750	0.5286	0.805
0.5287	14.0	10500	0.5121	0.8083
0.4928	15.0	11250	0.4975	0.8083
0.4187	16.0	12000	0.4850	0.8133
0.4686	17.0	12750	0.4741	0.8167
0.4337	18.0	13500	0.4641	0.82
0.4835	19.0	14250	0.4556	0.8217
0.4689	20.0	15000	0.4478	0.825
0.4014	21.0	15750	0.4407	0.8233
0.4524	22.0	16500	0.4346	0.8233
0.4363	23.0	17250	0.4289	0.825
0.4366	24.0	18000	0.4239	0.825
0.419	25.0	18750	0.4192	0.8267
0.4041	26.0	19500	0.4151	0.8283
0.334	27.0	20250	0.4112	0.8317
0.3991	28.0	21000	0.4077	0.8317
0.4162	29.0	21750	0.4044	0.83
0.3862	30.0	22500	0.4016	0.8317
0.3788	31.0	23250	0.3990	0.8333
0.3692	32.0	24000	0.3965	0.8333
0.3919	33.0	24750	0.3944	0.8333
0.3436	34.0	25500	0.3923	0.8317
0.385	35.0	26250	0.3904	0.8317
0.4009	36.0	27000	0.3887	0.835
0.3069	37.0	27750	0.3872	0.835
0.3924	38.0	28500	0.3858	0.835
0.3366	39.0	29250	0.3846	0.835
0.3431	40.0	30000	0.3835	0.8333
0.3539	41.0	30750	0.3825	0.8333
0.3975	42.0	31500	0.3816	0.8333
0.3795	43.0	32250	0.3809	0.835
0.36	44.0	33000	0.3803	0.835
0.3923	45.0	33750	0.3798	0.835
0.3388	46.0	34500	0.3793	0.835
0.3751	47.0	35250	0.3790	0.835
0.3764	48.0	36000	0.3788	0.835
0.3502	49.0	36750	0.3787	0.835
0.28	50.0	37500	0.3787	0.835

Framework versions

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