hushem_40x_beit_large_adamax_001_fold5

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: 1.1906
• Accuracy: 0.8049

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
0.3728	1.0	220	0.2484	0.9024
0.2424	2.0	440	1.0593	0.7805
0.1221	3.0	660	0.9944	0.7317
0.0746	4.0	880	1.4179	0.7073
0.0501	5.0	1100	0.6557	0.8049
0.0914	6.0	1320	1.5051	0.7073
0.0408	7.0	1540	0.1238	0.9512
0.0281	8.0	1760	0.6572	0.8537
0.0024	9.0	1980	0.9478	0.8049
0.0097	10.0	2200	0.6899	0.8537
0.0507	11.0	2420	1.0591	0.8049
0.0001	12.0	2640	0.9070	0.8780
0.0056	13.0	2860	1.1233	0.7805
0.0168	14.0	3080	1.3279	0.8049
0.0205	15.0	3300	1.4696	0.8049
0.0004	16.0	3520	1.8691	0.7561
0.0001	17.0	3740	1.4193	0.8293
0.0029	18.0	3960	1.9471	0.8049
0.0	19.0	4180	1.9190	0.7317
0.0	20.0	4400	2.0689	0.7317
0.0021	21.0	4620	0.3369	0.9024
0.0001	22.0	4840	0.9862	0.8537
0.0001	23.0	5060	0.9863	0.8780
0.0118	24.0	5280	1.0405	0.8049
0.0016	25.0	5500	1.4400	0.7805
0.0379	26.0	5720	1.0773	0.8537
0.0	27.0	5940	0.9902	0.8537
0.0	28.0	6160	0.9125	0.8293
0.0	29.0	6380	0.8492	0.8293
0.0	30.0	6600	1.3170	0.8293
0.0	31.0	6820	1.3145	0.7805
0.0	32.0	7040	0.7274	0.8780
0.0	33.0	7260	0.7992	0.8780
0.0	34.0	7480	0.7001	0.9024
0.0	35.0	7700	0.7059	0.9024
0.0	36.0	7920	0.7509	0.9024
0.0	37.0	8140	0.7646	0.9024
0.0	38.0	8360	1.2149	0.8293
0.0	39.0	8580	1.2146	0.8293
0.0	40.0	8800	1.2180	0.8293
0.0	41.0	9020	1.1864	0.8049
0.0	42.0	9240	1.1736	0.8049
0.0	43.0	9460	1.1601	0.8049
0.0	44.0	9680	1.1683	0.8049
0.0	45.0	9900	1.1682	0.8049
0.0	46.0	10120	1.1690	0.8049
0.0	47.0	10340	1.1691	0.8049
0.0	48.0	10560	1.1738	0.8049
0.0	49.0	10780	1.1753	0.8049
0.0	50.0	11000	1.1906	0.8049

Framework versions

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