hushem_5x_deit_small_sgd_0001_fold4

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: 1.3721
• Accuracy: 0.2619

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
1.5364	1.0	28	1.4759	0.2857
1.6032	2.0	56	1.4677	0.2857
1.5235	3.0	84	1.4598	0.2857
1.5363	4.0	112	1.4530	0.2857
1.4963	5.0	140	1.4466	0.2857
1.4798	6.0	168	1.4404	0.2857
1.4963	7.0	196	1.4349	0.2857
1.441	8.0	224	1.4297	0.3095
1.5032	9.0	252	1.4249	0.3095
1.4231	10.0	280	1.4205	0.3095
1.4482	11.0	308	1.4164	0.3095
1.4398	12.0	336	1.4127	0.3095
1.468	13.0	364	1.4093	0.3095
1.4278	14.0	392	1.4061	0.3095
1.4624	15.0	420	1.4032	0.3095
1.438	16.0	448	1.4004	0.2857
1.4401	17.0	476	1.3979	0.2857
1.416	18.0	504	1.3956	0.3095
1.4033	19.0	532	1.3934	0.3333
1.4123	20.0	560	1.3916	0.3333
1.4056	21.0	588	1.3899	0.3095
1.4089	22.0	616	1.3883	0.3333
1.3801	23.0	644	1.3868	0.3333
1.3733	24.0	672	1.3854	0.3095
1.3798	25.0	700	1.3840	0.3095
1.4051	26.0	728	1.3828	0.3095
1.4017	27.0	756	1.3817	0.3095
1.4006	28.0	784	1.3807	0.3095
1.368	29.0	812	1.3797	0.3095
1.3628	30.0	840	1.3788	0.3333
1.3803	31.0	868	1.3780	0.2619
1.3495	32.0	896	1.3773	0.2619
1.393	33.0	924	1.3766	0.2619
1.3379	34.0	952	1.3760	0.2619
1.3609	35.0	980	1.3754	0.2619
1.3521	36.0	1008	1.3748	0.2619
1.3648	37.0	1036	1.3744	0.2619
1.341	38.0	1064	1.3740	0.2619
1.3689	39.0	1092	1.3736	0.2619
1.3877	40.0	1120	1.3733	0.2619
1.4062	41.0	1148	1.3730	0.2619
1.3585	42.0	1176	1.3727	0.2619
1.3339	43.0	1204	1.3725	0.2619
1.3351	44.0	1232	1.3724	0.2619
1.3978	45.0	1260	1.3722	0.2619
1.3819	46.0	1288	1.3721	0.2619
1.3511	47.0	1316	1.3721	0.2619
1.3593	48.0	1344	1.3721	0.2619
1.3691	49.0	1372	1.3721	0.2619
1.3757	50.0	1400	1.3721	0.2619

Framework versions

• Transformers 4.35.2
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0