smids_3x_deit_small_rms_001_fold4
This model is a fine-tuned version of facebook/deit-small-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.1078
- Accuracy: 0.7533
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: 0.001
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.1005 | 1.0 | 225 | 0.9456 | 0.5067 |
| 0.8365 | 2.0 | 450 | 0.8727 | 0.46 |
| 0.7918 | 3.0 | 675 | 0.7967 | 0.555 |
| 0.7924 | 4.0 | 900 | 0.7472 | 0.62 |
| 0.8029 | 5.0 | 1125 | 0.7002 | 0.6433 |
| 0.7311 | 6.0 | 1350 | 0.6514 | 0.685 |
| 0.6939 | 7.0 | 1575 | 0.6643 | 0.7133 |
| 0.6634 | 8.0 | 1800 | 0.6358 | 0.715 |
| 0.6019 | 9.0 | 2025 | 0.6144 | 0.7067 |
| 0.6316 | 10.0 | 2250 | 0.6281 | 0.6967 |
| 0.6363 | 11.0 | 2475 | 0.6280 | 0.7233 |
| 0.645 | 12.0 | 2700 | 0.5846 | 0.7433 |
| 0.6675 | 13.0 | 2925 | 0.5877 | 0.7333 |
| 0.5165 | 14.0 | 3150 | 0.5728 | 0.75 |
| 0.52 | 15.0 | 3375 | 0.6706 | 0.73 |
| 0.5715 | 16.0 | 3600 | 0.6345 | 0.6983 |
| 0.4691 | 17.0 | 3825 | 0.6573 | 0.7167 |
| 0.5674 | 18.0 | 4050 | 0.5616 | 0.7633 |
| 0.5252 | 19.0 | 4275 | 0.5645 | 0.755 |
| 0.4602 | 20.0 | 4500 | 0.6017 | 0.755 |
| 0.4817 | 21.0 | 4725 | 0.5934 | 0.73 |
| 0.5616 | 22.0 | 4950 | 0.5845 | 0.7583 |
| 0.4883 | 23.0 | 5175 | 0.6144 | 0.7483 |
| 0.4781 | 24.0 | 5400 | 0.6001 | 0.7517 |
| 0.4229 | 25.0 | 5625 | 0.5953 | 0.77 |
| 0.4499 | 26.0 | 5850 | 0.6184 | 0.7767 |
| 0.4522 | 27.0 | 6075 | 0.6104 | 0.7433 |
| 0.4849 | 28.0 | 6300 | 0.6429 | 0.745 |
| 0.3967 | 29.0 | 6525 | 0.6127 | 0.7533 |
| 0.429 | 30.0 | 6750 | 0.6125 | 0.775 |
| 0.3969 | 31.0 | 6975 | 0.6078 | 0.7733 |
| 0.3935 | 32.0 | 7200 | 0.6333 | 0.7517 |
| 0.3619 | 33.0 | 7425 | 0.6571 | 0.795 |
| 0.4192 | 34.0 | 7650 | 0.6837 | 0.7833 |
| 0.4245 | 35.0 | 7875 | 0.6680 | 0.76 |
| 0.3704 | 36.0 | 8100 | 0.6540 | 0.78 |
| 0.393 | 37.0 | 8325 | 0.6566 | 0.7867 |
| 0.314 | 38.0 | 8550 | 0.7028 | 0.7783 |
| 0.3398 | 39.0 | 8775 | 0.7698 | 0.7667 |
| 0.3529 | 40.0 | 9000 | 0.7537 | 0.7633 |
| 0.3444 | 41.0 | 9225 | 0.8464 | 0.775 |
| 0.2933 | 42.0 | 9450 | 0.7973 | 0.7717 |
| 0.2384 | 43.0 | 9675 | 0.9278 | 0.765 |
| 0.3008 | 44.0 | 9900 | 0.9100 | 0.7767 |
| 0.2711 | 45.0 | 10125 | 0.9061 | 0.7683 |
| 0.3037 | 46.0 | 10350 | 0.9619 | 0.7783 |
| 0.2347 | 47.0 | 10575 | 1.0097 | 0.765 |
| 0.2672 | 48.0 | 10800 | 1.0493 | 0.75 |
| 0.2045 | 49.0 | 11025 | 1.0959 | 0.76 |
| 0.1823 | 50.0 | 11250 | 1.1078 | 0.7533 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2