hushem_5x_deit_base_sgd_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: 1.2313
• Accuracy: 0.4889

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
1.4159	1.0	27	1.4534	0.1333
1.3715	2.0	54	1.4329	0.1333
1.349	3.0	81	1.4161	0.1556
1.3491	4.0	108	1.4025	0.1556
1.3037	5.0	135	1.3915	0.2
1.2977	6.0	162	1.3813	0.2444
1.2881	7.0	189	1.3728	0.2889
1.2735	8.0	216	1.3643	0.3333
1.2461	9.0	243	1.3568	0.3556
1.2282	10.0	270	1.3495	0.3556
1.2197	11.0	297	1.3426	0.3556
1.1815	12.0	324	1.3361	0.3778
1.1874	13.0	351	1.3296	0.3778
1.1512	14.0	378	1.3234	0.3778
1.169	15.0	405	1.3177	0.4
1.1635	16.0	432	1.3122	0.4
1.1212	17.0	459	1.3068	0.4
1.1132	18.0	486	1.3013	0.4
1.0934	19.0	513	1.2960	0.4
1.0783	20.0	540	1.2914	0.4
1.0674	21.0	567	1.2869	0.4
1.0564	22.0	594	1.2826	0.4222
1.0602	23.0	621	1.2784	0.4444
1.0292	24.0	648	1.2744	0.4667
1.0348	25.0	675	1.2706	0.4667
1.0373	26.0	702	1.2671	0.4667
1.0143	27.0	729	1.2638	0.4667
1.0044	28.0	756	1.2607	0.4667
0.9861	29.0	783	1.2578	0.4667
1.0112	30.0	810	1.2551	0.4667
0.9561	31.0	837	1.2525	0.4667
0.9839	32.0	864	1.2500	0.4667
0.9768	33.0	891	1.2477	0.4667
0.936	34.0	918	1.2456	0.4667
0.9571	35.0	945	1.2436	0.4667
0.9423	36.0	972	1.2418	0.4667
0.9413	37.0	999	1.2401	0.4667
0.9304	38.0	1026	1.2386	0.4889
0.9391	39.0	1053	1.2372	0.4889
0.9013	40.0	1080	1.2360	0.4889
0.9198	41.0	1107	1.2349	0.4889
0.9119	42.0	1134	1.2340	0.4889
0.9214	43.0	1161	1.2332	0.4889
0.8928	44.0	1188	1.2325	0.4889
0.9196	45.0	1215	1.2320	0.4889
0.906	46.0	1242	1.2316	0.4889
0.9098	47.0	1269	1.2314	0.4889
0.9113	48.0	1296	1.2313	0.4889
0.9534	49.0	1323	1.2313	0.4889
0.8999	50.0	1350	1.2313	0.4889

Framework versions

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