<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
masked-50-pretraining-20-epoch
This model is a fine-tuned version of koobear/cam16-no-train-mask-final-50-epochs on the image_folder dataset. It achieves the following results on the evaluation set:
- Loss: 0.7943
- Accuracy: 0.9444
- F1: 0.9458
- Precision: 0.9550
- Recall: 0.9444
- Balanced Acc: 0.9372
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: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall | Balanced Acc |
|---|---|---|---|---|---|---|---|---|
| 1.6244 | 1.0 | 3 | 1.5530 | 0.0694 | 0.0160 | 0.0105 | 0.0694 | 0.4 |
| 1.4855 | 2.0 | 6 | 1.4500 | 0.6389 | 0.7013 | 0.8772 | 0.6389 | 0.8093 |
| 1.3995 | 3.0 | 9 | 1.3828 | 0.8611 | 0.8683 | 0.8926 | 0.8611 | 0.8160 |
| 1.3087 | 4.0 | 12 | 1.3094 | 0.9028 | 0.9048 | 0.9130 | 0.9028 | 0.8287 |
| 1.231 | 5.0 | 15 | 1.2352 | 0.9306 | 0.9332 | 0.9421 | 0.9306 | 0.9330 |
| 1.1601 | 6.0 | 18 | 1.1765 | 0.9306 | 0.9332 | 0.9421 | 0.9306 | 0.9330 |
| 1.0896 | 7.0 | 21 | 1.1198 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
| 1.0311 | 8.0 | 24 | 1.0661 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
| 0.9698 | 9.0 | 27 | 1.0256 | 0.9306 | 0.9346 | 0.9504 | 0.9306 | 0.9330 |
| 0.9176 | 10.0 | 30 | 0.9865 | 0.9167 | 0.9211 | 0.9407 | 0.9167 | 0.9287 |
| 0.8757 | 11.0 | 33 | 0.9467 | 0.9306 | 0.9325 | 0.9453 | 0.9306 | 0.9330 |
| 0.824 | 12.0 | 36 | 0.9137 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
| 0.7869 | 13.0 | 39 | 0.8857 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
| 0.7569 | 14.0 | 42 | 0.8570 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
| 0.7276 | 15.0 | 45 | 0.8374 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
| 0.7092 | 16.0 | 48 | 0.8219 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
| 0.6889 | 17.0 | 51 | 0.8105 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
| 0.6725 | 18.0 | 54 | 0.8019 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
| 0.6617 | 19.0 | 57 | 0.7966 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
| 0.6575 | 20.0 | 60 | 0.7943 | 0.9444 | 0.9458 | 0.9550 | 0.9444 | 0.9372 |
Framework versions
- Transformers 4.33.0
- Pytorch 2.0.0
- Datasets 2.1.0
- Tokenizers 0.13.3
