swinv2-base-patch4-window8-256-finetuned-ind-17-imbalanced-aadhaarmask
This model is a fine-tuned version of microsoft/swinv2-base-patch4-window8-256 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.3463
- Accuracy: 0.8463
- Recall: 0.8463
- F1: 0.8464
- Precision: 0.8482
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Recall | F1 | Precision |
|---|---|---|---|---|---|---|---|
| No log | 0.9974 | 293 | 0.6222 | 0.7901 | 0.7901 | 0.7737 | 0.7747 |
| No log | 1.9983 | 587 | 0.4901 | 0.8063 | 0.8063 | 0.7998 | 0.8066 |
| No log | 2.9991 | 881 | 0.4374 | 0.8225 | 0.8225 | 0.8170 | 0.8356 |
| No log | 4.0 | 1175 | 0.4262 | 0.8340 | 0.8340 | 0.8270 | 0.8541 |
| No log | 4.9974 | 1468 | 0.4079 | 0.8310 | 0.8310 | 0.8290 | 0.8379 |
| No log | 5.9983 | 1762 | 0.4117 | 0.8370 | 0.8370 | 0.8361 | 0.8509 |
| No log | 6.9991 | 2056 | 0.3807 | 0.8370 | 0.8370 | 0.8361 | 0.8416 |
| No log | 8.0 | 2350 | 0.3419 | 0.8595 | 0.8595 | 0.8583 | 0.8609 |
| No log | 8.9974 | 2643 | 0.3628 | 0.8438 | 0.8438 | 0.8424 | 0.8448 |
| 0.4492 | 9.9745 | 2930 | 0.3638 | 0.8399 | 0.8399 | 0.8394 | 0.8410 |
Framework versions
- Transformers 4.40.1
- Pytorch 2.2.0a0+81ea7a4
- Datasets 2.19.0
- Tokenizers 0.19.1