smids_10x_deit_base_sgd_0001_fold4

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: 0.3973
• Accuracy: 0.8417

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
1.0764	1.0	750	1.0715	0.4467
1.0016	2.0	1500	1.0200	0.565
0.9544	3.0	2250	0.9543	0.625
0.8923	4.0	3000	0.8833	0.665
0.8291	5.0	3750	0.8138	0.7317
0.7577	6.0	4500	0.7525	0.7483
0.6942	7.0	5250	0.6996	0.775
0.6967	8.0	6000	0.6552	0.785
0.6475	9.0	6750	0.6183	0.7967
0.5601	10.0	7500	0.5880	0.8017
0.5611	11.0	8250	0.5631	0.8033
0.5838	12.0	9000	0.5423	0.8117
0.4987	13.0	9750	0.5248	0.8133
0.5348	14.0	10500	0.5101	0.815
0.4457	15.0	11250	0.4972	0.8167
0.4244	16.0	12000	0.4863	0.82
0.394	17.0	12750	0.4768	0.8217
0.4347	18.0	13500	0.4683	0.8267
0.455	19.0	14250	0.4609	0.8283
0.4228	20.0	15000	0.4543	0.8317
0.3783	21.0	15750	0.4484	0.8333
0.4279	22.0	16500	0.4432	0.835
0.4396	23.0	17250	0.4385	0.8333
0.4808	24.0	18000	0.4342	0.8367
0.4208	25.0	18750	0.4304	0.8367
0.3784	26.0	19500	0.4269	0.8367
0.3843	27.0	20250	0.4238	0.8367
0.3813	28.0	21000	0.4210	0.8333
0.3843	29.0	21750	0.4183	0.835
0.3906	30.0	22500	0.4160	0.835
0.3475	31.0	23250	0.4138	0.8367
0.3808	32.0	24000	0.4118	0.8383
0.373	33.0	24750	0.4100	0.8383
0.4094	34.0	25500	0.4083	0.84
0.3294	35.0	26250	0.4068	0.84
0.3714	36.0	27000	0.4054	0.84
0.3219	37.0	27750	0.4042	0.84
0.3856	38.0	28500	0.4031	0.84
0.3967	39.0	29250	0.4021	0.84
0.3872	40.0	30000	0.4012	0.8417
0.3755	41.0	30750	0.4004	0.8417
0.3647	42.0	31500	0.3997	0.8417
0.4013	43.0	32250	0.3991	0.8417
0.3537	44.0	33000	0.3986	0.8417
0.3786	45.0	33750	0.3982	0.8417
0.344	46.0	34500	0.3978	0.8417
0.2971	47.0	35250	0.3976	0.8417
0.3345	48.0	36000	0.3974	0.8417
0.3788	49.0	36750	0.3974	0.8417
0.2837	50.0	37500	0.3973	0.8417

Framework versions

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