swinv2-tiny-patch4-window8-256-dmae-humeda-DAV38

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: 1.1368
• Accuracy: 0.6591

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: cosine
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	4	1.4929	0.4545
No log	2.0	8	1.2769	0.4545
4.8085	3.0	12	1.2009	0.5682
4.8085	4.0	16	1.0885	0.5114
4.8085	5.0	20	0.9977	0.6591
3.4418	6.0	24	0.9279	0.7045
3.4418	7.0	28	0.9311	0.6591
3.4418	8.0	32	0.9485	0.6591
2.4374	9.0	36	0.9279	0.6705
2.4374	10.0	40	0.9567	0.6705
2.4374	11.0	44	0.9832	0.6705
1.6575	12.0	48	0.9993	0.6705
1.6575	13.0	52	1.0360	0.6705
1.6575	14.0	56	1.0418	0.6591
1.1852	15.0	60	1.0619	0.6477
1.1852	16.0	64	1.0820	0.6705
1.1852	17.0	68	1.1155	0.6477
0.9213	18.0	72	1.1154	0.6591
0.9213	19.0	76	1.1240	0.6591
0.9213	20.0	80	1.1349	0.6591
0.7877	21.0	84	1.1367	0.6591
0.7877	22.0	88	1.1367	0.6591
0.7877	22.6154	90	1.1368	0.6591

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0