swinv2-tiny-patch4-window8-256-dmae-humeda-DAV21

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: 0.8631
• Accuracy: 0.7308

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: 42

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	0.9302	10	1.5510	0.3462
6.5929	1.9302	20	1.4802	0.2692
5.6252	2.9302	30	1.1115	0.4038
3.874	3.9302	40	0.9996	0.5577
2.7168	4.9302	50	0.8436	0.6538
2.2435	5.9302	60	0.9320	0.6154
2.2435	6.9302	70	0.8412	0.6346
1.9334	7.9302	80	0.8622	0.6731
1.6303	8.9302	90	0.9152	0.7115
1.2748	9.9302	100	0.9721	0.6731
1.0945	10.9302	110	1.0827	0.6538
0.8395	11.9302	120	0.9153	0.7115
0.8395	12.9302	130	0.8631	0.7308
0.8587	13.9302	140	1.1039	0.6538
0.8574	14.9302	150	1.0463	0.6923
0.7096	15.9302	160	0.9991	0.7115
0.6606	16.9302	170	1.0519	0.6731
0.5513	17.9302	180	1.0865	0.7115
0.5513	18.9302	190	1.1140	0.6731
0.61	19.9302	200	1.0290	0.6731
0.5278	20.9302	210	1.1003	0.6923
0.4639	21.9302	220	1.2472	0.6538
0.4719	22.9302	230	1.1546	0.6923
0.4212	23.9302	240	1.1084	0.7308
0.4212	24.9302	250	1.2953	0.6731
0.4109	25.9302	260	1.1868	0.7308
0.4236	26.9302	270	1.2560	0.6346
0.3638	27.9302	280	1.2161	0.7115
0.3944	28.9302	290	1.1582	0.7308
0.3621	29.9302	300	1.2993	0.6923
0.3621	30.9302	310	1.1401	0.7115
0.3203	31.9302	320	1.3228	0.7115
0.3014	32.9302	330	1.2813	0.6923
0.3464	33.9302	340	1.4768	0.6538
0.2891	34.9302	350	1.2304	0.7308
0.3153	35.9302	360	1.3096	0.6923
0.3153	36.9302	370	1.3565	0.7115
0.2762	37.9302	380	1.2931	0.6923
0.3191	38.9302	390	1.2441	0.7308
0.3009	39.9302	400	1.2110	0.7308
0.2645	40.9302	410	1.2433	0.7115
0.2497	41.9302	420	1.2461	0.6923

Framework versions

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