wmc_v2_swin-tiny-patch4-window7-224_base_wm811k_cls_contra_learning_1017_all_cls

This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0733
  • Accuracy: 0.9767
  • Precision: 0.8901
  • Recall: 0.9041
  • F1: 0.8937

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: 32
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 3
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy Precision Recall F1
0.3022 0.1079 100 0.8643 0.6787 0.0969 0.1188 0.1016
0.1973 0.2158 200 0.4558 0.8645 0.5013 0.4063 0.3734
0.1885 0.3237 300 0.2848 0.8865 0.5775 0.5620 0.5382
0.17 0.4316 400 0.2082 0.9122 0.7820 0.6967 0.6702
0.1565 0.5395 500 0.1775 0.9164 0.7701 0.7289 0.6901
0.1251 0.6474 600 0.1820 0.9265 0.6982 0.8022 0.6862
0.087 0.7553 700 0.1075 0.9688 0.8296 0.8707 0.8433
0.1474 0.8632 800 0.1504 0.9351 0.7887 0.8037 0.7687
0.1065 0.9711 900 0.1576 0.9262 0.8158 0.7454 0.7077
0.1262 1.0790 1000 0.1323 0.9475 0.8042 0.8150 0.7552
0.1223 1.1869 1100 0.1503 0.9340 0.7646 0.8377 0.7605
0.1433 1.2948 1200 0.0846 0.9739 0.8748 0.8704 0.8688
0.11 1.4028 1300 0.1043 0.9608 0.8525 0.8474 0.8353
0.0907 1.5107 1400 0.0814 0.9757 0.8717 0.8733 0.8619
0.0951 1.6186 1500 0.0899 0.9731 0.8459 0.9010 0.8669
0.1085 1.7265 1600 0.1138 0.9593 0.7984 0.8802 0.8262
0.0883 1.8344 1700 0.0794 0.9756 0.8803 0.8954 0.8832
0.0835 1.9423 1800 0.0853 0.9751 0.8444 0.9116 0.8706
0.0907 2.0502 1900 0.0992 0.9640 0.8266 0.8745 0.8319
0.091 2.1581 2000 0.0729 0.9779 0.8858 0.9061 0.8936
0.1216 2.2660 2100 0.0802 0.9743 0.8711 0.9100 0.8879
0.0866 2.3739 2200 0.0853 0.9742 0.8656 0.9099 0.8845
0.0815 2.4818 2300 0.0799 0.9735 0.8860 0.8924 0.8849
0.0767 2.5897 2400 0.0643 0.9803 0.8870 0.9182 0.9010
0.0809 2.6976 2500 0.0720 0.9779 0.8813 0.9173 0.8978
0.1004 2.8055 2600 0.0731 0.9780 0.8966 0.9072 0.8998
0.0607 2.9134 2700 0.0733 0.9767 0.8901 0.9041 0.8937

Framework versions

  • Transformers 4.45.2
  • Pytorch 2.4.1+cu121
  • Datasets 3.0.1
  • Tokenizers 0.20.1
Downloads last month
8
Safetensors
Model size
27.6M params
Tensor type
I64
·
F32
·
Inference API
Unable to determine this model’s pipeline type. Check the docs .

Model tree for Niraya666/wmc_v2_swin-tiny-patch4-window7-224_base_wm811k_cls_contra_learning_1017_all_cls

Finetuned
(480)
this model