hushem_5x_deit_small_sgd_0001_fold5

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: 1.3702
• Accuracy: 0.2683

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
1.5914	1.0	28	1.5102	0.2195
1.502	2.0	56	1.4998	0.2439
1.5359	3.0	84	1.4897	0.2439
1.4953	4.0	112	1.4806	0.2195
1.505	5.0	140	1.4721	0.2439
1.5366	6.0	168	1.4645	0.2439
1.5251	7.0	196	1.4572	0.2439
1.4698	8.0	224	1.4506	0.2439
1.4915	9.0	252	1.4443	0.2439
1.4618	10.0	280	1.4384	0.2439
1.4473	11.0	308	1.4329	0.2439
1.4682	12.0	336	1.4279	0.2439
1.4426	13.0	364	1.4233	0.2439
1.4128	14.0	392	1.4190	0.2683
1.4363	15.0	420	1.4150	0.2683
1.4383	16.0	448	1.4113	0.2683
1.4168	17.0	476	1.4079	0.2683
1.4317	18.0	504	1.4047	0.2683
1.4208	19.0	532	1.4016	0.2927
1.4021	20.0	560	1.3989	0.2927
1.4325	21.0	588	1.3963	0.2927
1.4072	22.0	616	1.3940	0.2927
1.3729	23.0	644	1.3918	0.2927
1.3955	24.0	672	1.3898	0.2927
1.3868	25.0	700	1.3879	0.2927
1.3985	26.0	728	1.3861	0.2683
1.3854	27.0	756	1.3845	0.2683
1.3968	28.0	784	1.3830	0.2683
1.3689	29.0	812	1.3816	0.2683
1.4069	30.0	840	1.3803	0.2683
1.387	31.0	868	1.3791	0.2683
1.3786	32.0	896	1.3780	0.2683
1.3773	33.0	924	1.3769	0.2683
1.3779	34.0	952	1.3760	0.2683
1.3797	35.0	980	1.3751	0.2683
1.3671	36.0	1008	1.3744	0.2683
1.3638	37.0	1036	1.3737	0.2683
1.3614	38.0	1064	1.3731	0.2683
1.3646	39.0	1092	1.3725	0.2683
1.3609	40.0	1120	1.3720	0.2683
1.3899	41.0	1148	1.3716	0.2683
1.3896	42.0	1176	1.3712	0.2683
1.3725	43.0	1204	1.3709	0.2683
1.3896	44.0	1232	1.3706	0.2683
1.3695	45.0	1260	1.3704	0.2683
1.3698	46.0	1288	1.3703	0.2683
1.3813	47.0	1316	1.3702	0.2683
1.3636	48.0	1344	1.3702	0.2683
1.3528	49.0	1372	1.3702	0.2683
1.3747	50.0	1400	1.3702	0.2683

Framework versions

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