beit-base-patch16-224-hasta-65-fold3
This model is a fine-tuned version of microsoft/beit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.9494
- Accuracy: 0.6111
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 100
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| No log | 0.5714 | 1 | 1.2972 | 0.3333 |
| No log | 1.7143 | 3 | 1.1961 | 0.4167 |
| No log | 2.8571 | 5 | 1.5408 | 0.2778 |
| No log | 4.0 | 7 | 1.1727 | 0.3611 |
| No log | 4.5714 | 8 | 1.1418 | 0.3611 |
| 1.2077 | 5.7143 | 10 | 1.2295 | 0.2778 |
| 1.2077 | 6.8571 | 12 | 1.1358 | 0.4167 |
| 1.2077 | 8.0 | 14 | 1.1396 | 0.4444 |
| 1.2077 | 8.5714 | 15 | 1.1827 | 0.3611 |
| 1.2077 | 9.7143 | 17 | 1.0719 | 0.3889 |
| 1.2077 | 10.8571 | 19 | 1.0911 | 0.4167 |
| 1.004 | 12.0 | 21 | 1.0003 | 0.5278 |
| 1.004 | 12.5714 | 22 | 0.9457 | 0.5 |
| 1.004 | 13.7143 | 24 | 1.0182 | 0.5 |
| 1.004 | 14.8571 | 26 | 0.9877 | 0.5 |
| 1.004 | 16.0 | 28 | 0.9537 | 0.5278 |
| 1.004 | 16.5714 | 29 | 0.9310 | 0.5833 |
| 0.8983 | 17.7143 | 31 | 0.8763 | 0.5278 |
| 0.8983 | 18.8571 | 33 | 0.9626 | 0.5556 |
| 0.8983 | 20.0 | 35 | 1.0252 | 0.5833 |
| 0.8983 | 20.5714 | 36 | 1.0288 | 0.5278 |
| 0.8983 | 21.7143 | 38 | 1.0807 | 0.4722 |
| 0.7347 | 22.8571 | 40 | 0.9995 | 0.4444 |
| 0.7347 | 24.0 | 42 | 1.3380 | 0.4722 |
| 0.7347 | 24.5714 | 43 | 1.3829 | 0.4722 |
| 0.7347 | 25.7143 | 45 | 1.0370 | 0.4722 |
| 0.7347 | 26.8571 | 47 | 1.0372 | 0.5278 |
| 0.7347 | 28.0 | 49 | 0.9478 | 0.5278 |
| 0.6583 | 28.5714 | 50 | 0.9328 | 0.5833 |
| 0.6583 | 29.7143 | 52 | 1.0592 | 0.5278 |
| 0.6583 | 30.8571 | 54 | 0.9613 | 0.5278 |
| 0.6583 | 32.0 | 56 | 0.8315 | 0.5556 |
| 0.6583 | 32.5714 | 57 | 0.8749 | 0.5278 |
| 0.6583 | 33.7143 | 59 | 1.0390 | 0.5556 |
| 0.5633 | 34.8571 | 61 | 0.9446 | 0.5833 |
| 0.5633 | 36.0 | 63 | 0.9586 | 0.5833 |
| 0.5633 | 36.5714 | 64 | 0.9877 | 0.5833 |
| 0.5633 | 37.7143 | 66 | 1.0395 | 0.5833 |
| 0.5633 | 38.8571 | 68 | 0.9494 | 0.6111 |
| 0.4905 | 40.0 | 70 | 1.0124 | 0.5833 |
| 0.4905 | 40.5714 | 71 | 1.1537 | 0.5556 |
| 0.4905 | 41.7143 | 73 | 1.2837 | 0.5833 |
| 0.4905 | 42.8571 | 75 | 1.0821 | 0.5556 |
| 0.4905 | 44.0 | 77 | 1.0084 | 0.5556 |
| 0.4905 | 44.5714 | 78 | 1.0433 | 0.5556 |
| 0.4609 | 45.7143 | 80 | 1.2119 | 0.5556 |
| 0.4609 | 46.8571 | 82 | 1.2976 | 0.5556 |
| 0.4609 | 48.0 | 84 | 1.1444 | 0.5833 |
| 0.4609 | 48.5714 | 85 | 1.0444 | 0.5833 |
| 0.4609 | 49.7143 | 87 | 0.9338 | 0.5278 |
| 0.4609 | 50.8571 | 89 | 0.9501 | 0.5833 |
| 0.4106 | 52.0 | 91 | 1.0081 | 0.5833 |
| 0.4106 | 52.5714 | 92 | 1.0580 | 0.5833 |
| 0.4106 | 53.7143 | 94 | 1.1188 | 0.5833 |
| 0.4106 | 54.8571 | 96 | 1.1291 | 0.5833 |
| 0.4106 | 56.0 | 98 | 1.1156 | 0.6111 |
| 0.4106 | 56.5714 | 99 | 1.1099 | 0.6111 |
| 0.3784 | 57.1429 | 100 | 1.1064 | 0.6111 |
Framework versions
- Transformers 4.41.0
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1