hushem_5x_deit_tiny_sgd_0001_fold4

This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

• Loss: 1.3803
• Accuracy: 0.3810

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
1.509	1.0	28	1.6728	0.2857
1.4745	2.0	56	1.6494	0.2857
1.4631	3.0	84	1.6263	0.2857
1.4901	4.0	112	1.6063	0.2857
1.441	5.0	140	1.5891	0.2857
1.4757	6.0	168	1.5714	0.2857
1.4394	7.0	196	1.5560	0.2857
1.4378	8.0	224	1.5417	0.2857
1.4304	9.0	252	1.5286	0.2857
1.455	10.0	280	1.5162	0.2857
1.4518	11.0	308	1.5058	0.2857
1.4092	12.0	336	1.4955	0.3095
1.4226	13.0	364	1.4865	0.3095
1.3993	14.0	392	1.4783	0.3095
1.3967	15.0	420	1.4705	0.3095
1.411	16.0	448	1.4633	0.3095
1.4285	17.0	476	1.4567	0.3095
1.3979	18.0	504	1.4501	0.3333
1.3856	19.0	532	1.4445	0.3333
1.3836	20.0	560	1.4396	0.3333
1.3724	21.0	588	1.4346	0.3333
1.4133	22.0	616	1.4302	0.3333
1.3803	23.0	644	1.4261	0.3571
1.38	24.0	672	1.4220	0.3571
1.3524	25.0	700	1.4181	0.3571
1.3732	26.0	728	1.4145	0.3571
1.3766	27.0	756	1.4110	0.3571
1.3865	28.0	784	1.4081	0.3571
1.3436	29.0	812	1.4052	0.3571
1.3611	30.0	840	1.4024	0.3571
1.3712	31.0	868	1.3999	0.3571
1.3547	32.0	896	1.3975	0.3571
1.3736	33.0	924	1.3953	0.3571
1.3568	34.0	952	1.3933	0.3571
1.3531	35.0	980	1.3914	0.3571
1.3544	36.0	1008	1.3897	0.3571
1.3181	37.0	1036	1.3881	0.3571
1.3538	38.0	1064	1.3867	0.3571
1.3828	39.0	1092	1.3855	0.3810
1.3242	40.0	1120	1.3844	0.3810
1.3519	41.0	1148	1.3834	0.3810
1.3218	42.0	1176	1.3826	0.3810
1.3603	43.0	1204	1.3819	0.3810
1.3249	44.0	1232	1.3813	0.3810
1.3333	45.0	1260	1.3809	0.3810
1.3432	46.0	1288	1.3806	0.3810
1.333	47.0	1316	1.3804	0.3810
1.3303	48.0	1344	1.3803	0.3810
1.3729	49.0	1372	1.3803	0.3810
1.3281	50.0	1400	1.3803	0.3810

Framework versions

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