hushem_5x_deit_base_sgd_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: 1.3613
• Accuracy: 0.3488

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
1.4194	1.0	28	1.3854	0.3721
1.4285	2.0	56	1.3841	0.3721
1.4355	3.0	84	1.3830	0.3721
1.435	4.0	112	1.3819	0.3721
1.3846	5.0	140	1.3808	0.3488
1.4335	6.0	168	1.3798	0.3488
1.4069	7.0	196	1.3789	0.3488
1.4319	8.0	224	1.3779	0.3488
1.4012	9.0	252	1.3769	0.3488
1.401	10.0	280	1.3761	0.3488
1.4013	11.0	308	1.3752	0.3488
1.4057	12.0	336	1.3744	0.3488
1.3918	13.0	364	1.3736	0.3488
1.3961	14.0	392	1.3729	0.3721
1.3654	15.0	420	1.3722	0.3721
1.3967	16.0	448	1.3715	0.3488
1.3921	17.0	476	1.3708	0.3488
1.3819	18.0	504	1.3702	0.3488
1.3847	19.0	532	1.3696	0.3488
1.402	20.0	560	1.3690	0.3488
1.3988	21.0	588	1.3684	0.3488
1.3796	22.0	616	1.3679	0.3488
1.3761	23.0	644	1.3674	0.3721
1.3729	24.0	672	1.3669	0.3721
1.3864	25.0	700	1.3664	0.3488
1.3808	26.0	728	1.3660	0.3488
1.3849	27.0	756	1.3656	0.3488
1.3863	28.0	784	1.3652	0.3488
1.3673	29.0	812	1.3648	0.3488
1.3797	30.0	840	1.3644	0.3488
1.3679	31.0	868	1.3641	0.3488
1.3844	32.0	896	1.3638	0.3488
1.3656	33.0	924	1.3634	0.3488
1.3701	34.0	952	1.3631	0.3488
1.363	35.0	980	1.3629	0.3488
1.386	36.0	1008	1.3627	0.3488
1.3733	37.0	1036	1.3624	0.3488
1.3666	38.0	1064	1.3622	0.3488
1.3489	39.0	1092	1.3620	0.3488
1.3854	40.0	1120	1.3619	0.3488
1.3831	41.0	1148	1.3617	0.3488
1.3627	42.0	1176	1.3616	0.3488
1.3753	43.0	1204	1.3615	0.3488
1.3596	44.0	1232	1.3614	0.3488
1.3572	45.0	1260	1.3614	0.3488
1.3678	46.0	1288	1.3613	0.3488
1.3744	47.0	1316	1.3613	0.3488
1.3693	48.0	1344	1.3613	0.3488
1.3589	49.0	1372	1.3613	0.3488
1.3662	50.0	1400	1.3613	0.3488

Framework versions

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