hushem_5x_deit_small_sgd_0001_fold2

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: 1.3896
• Accuracy: 0.3333

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
1.5183	1.0	27	1.5038	0.1778
1.502	2.0	54	1.4946	0.2
1.5109	3.0	81	1.4859	0.2222
1.5446	4.0	108	1.4781	0.2444
1.4687	5.0	135	1.4710	0.2444
1.4554	6.0	162	1.4641	0.2889
1.4113	7.0	189	1.4582	0.2889
1.4434	8.0	216	1.4525	0.2667
1.4243	9.0	243	1.4473	0.2667
1.4268	10.0	270	1.4425	0.2889
1.386	11.0	297	1.4382	0.2889
1.4235	12.0	324	1.4341	0.2667
1.4228	13.0	351	1.4304	0.2667
1.4091	14.0	378	1.4269	0.2889
1.4135	15.0	405	1.4239	0.2667
1.4228	16.0	432	1.4210	0.2889
1.4188	17.0	459	1.4184	0.2889
1.3824	18.0	486	1.4159	0.3333
1.3861	19.0	513	1.4136	0.3111
1.393	20.0	540	1.4115	0.3111
1.4051	21.0	567	1.4096	0.3111
1.373	22.0	594	1.4077	0.3333
1.3737	23.0	621	1.4060	0.3333
1.3668	24.0	648	1.4044	0.3556
1.362	25.0	675	1.4030	0.3556
1.3931	26.0	702	1.4016	0.3556
1.3504	27.0	729	1.4003	0.3556
1.3706	28.0	756	1.3992	0.3556
1.359	29.0	783	1.3981	0.3556
1.3774	30.0	810	1.3972	0.3556
1.3678	31.0	837	1.3963	0.3556
1.3418	32.0	864	1.3955	0.3556
1.3702	33.0	891	1.3947	0.3556
1.3589	34.0	918	1.3940	0.3556
1.3212	35.0	945	1.3933	0.3333
1.3648	36.0	972	1.3928	0.3333
1.3509	37.0	999	1.3922	0.3333
1.3506	38.0	1026	1.3917	0.3333
1.3673	39.0	1053	1.3913	0.3333
1.3657	40.0	1080	1.3910	0.3333
1.3651	41.0	1107	1.3906	0.3333
1.3688	42.0	1134	1.3904	0.3333
1.3871	43.0	1161	1.3901	0.3333
1.3307	44.0	1188	1.3899	0.3333
1.3505	45.0	1215	1.3898	0.3333
1.3367	46.0	1242	1.3897	0.3333
1.3605	47.0	1269	1.3896	0.3333
1.3556	48.0	1296	1.3896	0.3333
1.3876	49.0	1323	1.3896	0.3333
1.3357	50.0	1350	1.3896	0.3333

Framework versions

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