smids_3x_deit_tiny_rms_001_fold3
This model is a fine-tuned version of facebook/deit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.7451
- Accuracy: 0.7017
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 |
|---|---|---|---|---|
| 1.0251 | 1.0 | 225 | 0.9264 | 0.5433 |
| 0.8762 | 2.0 | 450 | 0.9033 | 0.5383 |
| 0.9046 | 3.0 | 675 | 0.8597 | 0.52 |
| 0.8168 | 4.0 | 900 | 0.8502 | 0.5867 |
| 0.8182 | 5.0 | 1125 | 0.8396 | 0.5417 |
| 0.8366 | 6.0 | 1350 | 0.8032 | 0.64 |
| 0.7873 | 7.0 | 1575 | 0.7872 | 0.6367 |
| 0.8222 | 8.0 | 1800 | 0.7892 | 0.605 |
| 0.8096 | 9.0 | 2025 | 0.8074 | 0.63 |
| 0.7866 | 10.0 | 2250 | 0.8155 | 0.5667 |
| 0.7895 | 11.0 | 2475 | 0.7692 | 0.6433 |
| 0.7721 | 12.0 | 2700 | 0.8106 | 0.6017 |
| 0.781 | 13.0 | 2925 | 0.7742 | 0.6533 |
| 0.7888 | 14.0 | 3150 | 0.7929 | 0.6117 |
| 0.7617 | 15.0 | 3375 | 0.7600 | 0.6683 |
| 0.8324 | 16.0 | 3600 | 0.7701 | 0.6433 |
| 0.7598 | 17.0 | 3825 | 0.8095 | 0.6333 |
| 0.7476 | 18.0 | 4050 | 0.7803 | 0.6033 |
| 0.7071 | 19.0 | 4275 | 0.7505 | 0.6683 |
| 0.7193 | 20.0 | 4500 | 0.7784 | 0.6183 |
| 0.6927 | 21.0 | 4725 | 0.7879 | 0.6467 |
| 0.666 | 22.0 | 4950 | 0.7212 | 0.6967 |
| 0.6763 | 23.0 | 5175 | 0.7194 | 0.6833 |
| 0.6715 | 24.0 | 5400 | 0.7919 | 0.6367 |
| 0.7294 | 25.0 | 5625 | 0.7785 | 0.6733 |
| 0.6936 | 26.0 | 5850 | 0.7216 | 0.6983 |
| 0.6322 | 27.0 | 6075 | 0.8000 | 0.6833 |
| 0.6761 | 28.0 | 6300 | 0.7942 | 0.6183 |
| 0.688 | 29.0 | 6525 | 0.7281 | 0.6567 |
| 0.6228 | 30.0 | 6750 | 0.7332 | 0.6567 |
| 0.6366 | 31.0 | 6975 | 0.7601 | 0.6717 |
| 0.6176 | 32.0 | 7200 | 0.7157 | 0.6883 |
| 0.6636 | 33.0 | 7425 | 0.7555 | 0.6567 |
| 0.6315 | 34.0 | 7650 | 0.7242 | 0.665 |
| 0.5915 | 35.0 | 7875 | 0.6940 | 0.6783 |
| 0.6259 | 36.0 | 8100 | 0.6760 | 0.6917 |
| 0.6325 | 37.0 | 8325 | 0.6834 | 0.6967 |
| 0.5846 | 38.0 | 8550 | 0.7137 | 0.6733 |
| 0.6018 | 39.0 | 8775 | 0.6801 | 0.6933 |
| 0.5692 | 40.0 | 9000 | 0.6837 | 0.6883 |
| 0.5234 | 41.0 | 9225 | 0.6917 | 0.6833 |
| 0.5543 | 42.0 | 9450 | 0.6614 | 0.7017 |
| 0.5363 | 43.0 | 9675 | 0.6720 | 0.7017 |
| 0.5474 | 44.0 | 9900 | 0.6703 | 0.7067 |
| 0.5234 | 45.0 | 10125 | 0.7035 | 0.6983 |
| 0.4923 | 46.0 | 10350 | 0.7111 | 0.7017 |
| 0.5435 | 47.0 | 10575 | 0.6985 | 0.7133 |
| 0.4932 | 48.0 | 10800 | 0.7085 | 0.7133 |
| 0.486 | 49.0 | 11025 | 0.7485 | 0.7033 |
| 0.4701 | 50.0 | 11250 | 0.7451 | 0.7017 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2