swinv2-tiny-patch4-window8-256-dmae-humeda-DAV7

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: 0.9241
  • Accuracy: 0.6923

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: 2e-05
  • train_batch_size: 64
  • eval_batch_size: 64
  • seed: 42
  • 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: 30

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.6286 1.0 12 1.5624 0.2692
1.5581 2.0 24 1.4706 0.5192
1.4708 3.0 36 1.2967 0.5385
1.3022 4.0 48 1.1556 0.4615
0.9735 5.0 60 1.0354 0.5385
0.7914 6.0 72 1.1014 0.4423
0.7376 7.0 84 1.1087 0.4808
0.692 8.0 96 0.9630 0.6346
0.6586 9.0 108 0.9429 0.6346
0.5799 10.0 120 0.9332 0.6346
0.5557 11.0 132 1.1712 0.5192
0.5233 12.0 144 1.0447 0.5577
0.4427 13.0 156 0.8928 0.6538
0.5043 14.0 168 1.0120 0.5962
0.4167 15.0 180 0.9241 0.6923
0.4601 16.0 192 0.8848 0.6538
0.4619 17.0 204 0.9239 0.6923
0.3822 18.0 216 0.9208 0.6731
0.3707 19.0 228 1.0374 0.5769
0.365 20.0 240 0.9900 0.6538
0.3412 21.0 252 1.0541 0.6538
0.3265 22.0 264 0.9913 0.6731
0.3096 23.0 276 1.0355 0.6346
0.3603 24.0 288 0.9986 0.6538
0.2924 25.0 300 1.0046 0.6731
0.3489 26.0 312 1.0560 0.6346
0.2974 27.0 324 1.0076 0.6538
0.2924 28.0 336 1.0164 0.6538
0.3369 29.0 348 1.0260 0.6346
0.2884 30.0 360 1.0293 0.6346

Framework versions

  • Transformers 4.47.1
  • Pytorch 2.5.1+cu121
  • Datasets 3.2.0
  • Tokenizers 0.21.0
Downloads last month
4
Safetensors
Model size
27.6M params
Tensor type
F32
ยท
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐Ÿ™‹ Ask for provider support

Model tree for RobertoSonic/swinv2-tiny-patch4-window8-256-dmae-humeda-DAV7

Finetuned
(137)
this model