hushem_40x_deit_base_rms_001_fold2

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: 8.5594
• Accuracy: 0.5778

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: 0.001
• 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
1.1732	1.0	215	1.0108	0.4667
0.7763	2.0	430	1.2138	0.5333
0.7021	3.0	645	1.2446	0.4
0.6002	4.0	860	1.7707	0.4444
0.4988	5.0	1075	2.1116	0.4667
0.4269	6.0	1290	2.3849	0.5556
0.3366	7.0	1505	2.4322	0.5556
0.2961	8.0	1720	3.2646	0.5556
0.2377	9.0	1935	3.1438	0.5333
0.2435	10.0	2150	3.6031	0.5778
0.2593	11.0	2365	3.5951	0.4889
0.1482	12.0	2580	3.8372	0.5111
0.1871	13.0	2795	3.7490	0.6222
0.1246	14.0	3010	3.7977	0.5333
0.166	15.0	3225	3.7321	0.5778
0.1672	16.0	3440	4.6413	0.4889
0.1752	17.0	3655	4.9330	0.5556
0.1214	18.0	3870	4.3615	0.5556
0.0488	19.0	4085	4.4231	0.5111
0.1336	20.0	4300	4.4451	0.5778
0.1002	21.0	4515	3.7455	0.5778
0.0734	22.0	4730	4.4970	0.5556
0.0322	23.0	4945	4.8990	0.5333
0.214	24.0	5160	5.1865	0.5778
0.1242	25.0	5375	5.0088	0.5333
0.0033	26.0	5590	4.9606	0.5556
0.0333	27.0	5805	4.4063	0.5778
0.0592	28.0	6020	4.1719	0.5556
0.0444	29.0	6235	6.2342	0.5111
0.0039	30.0	6450	5.9834	0.5333
0.003	31.0	6665	6.2329	0.5333
0.0008	32.0	6880	6.2499	0.6
0.1078	33.0	7095	5.2542	0.6222
0.0258	34.0	7310	6.7980	0.4889
0.0052	35.0	7525	6.6849	0.5333
0.0003	36.0	7740	6.1342	0.5556
0.0005	37.0	7955	5.4920	0.5778
0.0004	38.0	8170	5.3684	0.5778
0.0148	39.0	8385	5.3551	0.5556
0.0054	40.0	8600	7.4300	0.5111
0.0	41.0	8815	6.8539	0.5556
0.0	42.0	9030	6.8688	0.5556
0.0	43.0	9245	7.1702	0.5778
0.0	44.0	9460	7.4631	0.5778
0.0	45.0	9675	7.7338	0.5778
0.0	46.0	9890	7.9825	0.5778
0.0	47.0	10105	8.2172	0.5778
0.0	48.0	10320	8.4047	0.5778
0.0	49.0	10535	8.5267	0.5778
0.0	50.0	10750	8.5594	0.5778

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.0+cu121
• Datasets 2.12.0
• Tokenizers 0.13.2