swinv2-tiny-patch4-window8-256-dmae-humeda-DAV18

This model is a fine-tuned version of microsoft/swinv2-tiny-patch4-window8-256 on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.4318
• Accuracy: 0.6731

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: 3e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 64
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.2
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
3.3548	1.0	22	1.5527	0.3077
3.2091	2.0	44	1.5476	0.3077
2.9355	3.0	66	1.3747	0.4808
2.3577	4.0	88	1.3003	0.4231
1.4766	5.0	110	1.1989	0.4808
1.3704	6.0	132	1.0006	0.6538
1.118	7.0	154	1.0605	0.6154
1.0178	8.0	176	0.9814	0.5962
0.8974	9.0	198	1.0096	0.6538
0.672	10.0	220	1.1785	0.6154
0.6533	11.0	242	1.3520	0.5577
0.4788	12.0	264	0.9816	0.75
0.5707	13.0	286	1.0714	0.6538
0.424	14.0	308	1.2733	0.6923
0.4466	15.0	330	1.1687	0.7115
0.4284	16.0	352	1.1879	0.6731
0.4528	17.0	374	1.1845	0.6346
0.3062	18.0	396	1.2126	0.7115
0.2953	19.0	418	1.4019	0.6154
0.3169	20.0	440	1.3913	0.6538
0.3235	21.0	462	1.1455	0.6346
0.3069	22.0	484	1.4152	0.6538
0.2402	23.0	506	1.1008	0.7115
0.2055	24.0	528	1.2535	0.6538
0.265	25.0	550	1.3737	0.6538
0.2097	26.0	572	1.4048	0.6923
0.2185	27.0	594	1.3243	0.6923
0.1778	28.0	616	1.3771	0.6346
0.1815	29.0	638	1.3688	0.5769
0.2201	30.0	660	1.3827	0.6346
0.1808	31.0	682	1.3749	0.6731
0.2013	32.0	704	1.4271	0.6538
0.1822	33.0	726	1.4023	0.6923
0.1602	34.0	748	1.3908	0.6731
0.1303	35.0	770	1.4396	0.6731
0.1848	36.0	792	1.4828	0.6923
0.0982	37.0	814	1.4231	0.6923
0.1267	38.0	836	1.4148	0.6731
0.1467	39.0	858	1.3743	0.6731
0.0891	40.0	880	1.4194	0.6923
0.1252	41.0	902	1.4187	0.6923
0.1189	42.0	924	1.4043	0.6346
0.1328	43.0	946	1.4175	0.6731
0.172	44.0	968	1.4249	0.6731
0.1324	45.0	990	1.4320	0.6731
0.1363	46.0	1012	1.4320	0.6731
0.125	47.0	1034	1.4318	0.6731
0.1122	47.7442	1050	1.4318	0.6731

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0