hushem_40x_beit_large_adamax_001_fold3

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

• Loss: 3.1862
• Accuracy: 0.7674

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
0.3341	1.0	217	1.0412	0.6512
0.1567	2.0	434	0.8679	0.7907
0.1261	3.0	651	0.9519	0.7907
0.0451	4.0	868	1.0258	0.7674
0.0494	5.0	1085	1.2986	0.8140
0.0356	6.0	1302	1.5169	0.7442
0.0142	7.0	1519	1.4785	0.7907
0.048	8.0	1736	1.4460	0.8140
0.0363	9.0	1953	1.0943	0.7674
0.0409	10.0	2170	1.6345	0.7907
0.0021	11.0	2387	1.2558	0.8140
0.0072	12.0	2604	1.1994	0.8372
0.0193	13.0	2821	1.2732	0.8372
0.0006	14.0	3038	1.5708	0.7674
0.0013	15.0	3255	1.1380	0.8837
0.0001	16.0	3472	1.3578	0.8837
0.0	17.0	3689	1.3940	0.8837
0.0	18.0	3906	1.4630	0.8605
0.0	19.0	4123	1.4804	0.8140
0.0	20.0	4340	1.5039	0.8372
0.0	21.0	4557	1.5153	0.8605
0.0	22.0	4774	1.6110	0.8372
0.0	23.0	4991	1.6351	0.8372
0.0	24.0	5208	1.6586	0.8372
0.0	25.0	5425	1.6837	0.8605
0.0	26.0	5642	2.1644	0.8140
0.0	27.0	5859	1.8231	0.8372
0.0	28.0	6076	1.8592	0.8837
0.0	29.0	6293	2.3766	0.7907
0.0004	30.0	6510	2.2802	0.7674
0.0	31.0	6727	2.0919	0.7907
0.0	32.0	6944	2.0989	0.7907
0.0	33.0	7161	2.1214	0.7907
0.0	34.0	7378	2.1583	0.7907
0.0	35.0	7595	2.1876	0.7907
0.0	36.0	7812	2.1795	0.7907
0.0007	37.0	8029	3.1536	0.7674
0.0	38.0	8246	3.0845	0.7674
0.0	39.0	8463	2.9748	0.7907
0.0	40.0	8680	2.9984	0.7907
0.0	41.0	8897	3.0029	0.7907
0.0	42.0	9114	3.0143	0.7907
0.0	43.0	9331	3.0354	0.7907
0.0	44.0	9548	3.0480	0.7907
0.0	45.0	9765	3.0564	0.7907
0.0	46.0	9982	3.1685	0.7674
0.0	47.0	10199	3.1763	0.7674
0.0	48.0	10416	3.1810	0.7674
0.0	49.0	10633	3.1846	0.7674
0.0	50.0	10850	3.1862	0.7674

Framework versions

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