swinv2-tiny-patch4-window8-256-dmae-humeda-DAV23

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: 0.6901
• Accuracy: 0.8118

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• 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: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
6.4493	1.0	17	1.5281	0.2941
5.7922	2.0	34	1.3176	0.3882
4.2502	3.0	51	1.2015	0.4353
3.2402	4.0	68	0.8902	0.7176
2.5386	5.0	85	0.6509	0.7765
2.0351	6.0	102	0.6759	0.7647
1.8225	7.0	119	0.6607	0.7765
1.4778	8.0	136	0.7162	0.7529
1.4076	9.0	153	0.9084	0.7294
1.2056	10.0	170	0.6901	0.8118
0.9552	11.0	187	0.9153	0.7765
0.9859	12.0	204	0.8694	0.7529
0.8309	13.0	221	0.7666	0.8
0.7722	14.0	238	0.9118	0.7529
0.7632	15.0	255	0.8953	0.7529
0.5868	16.0	272	0.9678	0.7529
0.6577	17.0	289	1.0503	0.7765
0.5816	18.0	306	1.0602	0.7294
0.6222	19.0	323	1.1543	0.7765
0.4861	20.0	340	0.9739	0.8118
0.4422	21.0	357	1.0354	0.8
0.506	22.0	374	1.1097	0.8118
0.3833	23.0	391	1.2009	0.7765
0.4574	24.0	408	1.1366	0.7765
0.4467	25.0	425	1.0601	0.8118
0.4451	26.0	442	1.0935	0.7765
0.4384	27.0	459	1.1617	0.7647
0.4321	28.0	476	1.1012	0.7765
0.4398	29.0	493	1.0825	0.7882
0.361	30.0	510	1.1127	0.7647
0.4428	31.0	527	1.2024	0.7529
0.451	32.0	544	1.1550	0.7647
0.403	33.0	561	1.1646	0.7765
0.3059	34.0	578	1.2442	0.7765
0.3022	35.0	595	1.1976	0.7765
0.319	36.0	612	1.1564	0.7765
0.3737	37.0	629	1.1857	0.7765
0.3063	37.6667	640	1.1930	0.7765

Framework versions

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