smids_3x_deit_tiny_rms_00001_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: 0.9705
- Accuracy: 0.9033
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: 1e-05
- 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 |
|---|---|---|---|---|
| 0.3413 | 1.0 | 225 | 0.2808 | 0.8933 |
| 0.2287 | 2.0 | 450 | 0.2742 | 0.9 |
| 0.1591 | 3.0 | 675 | 0.3006 | 0.89 |
| 0.0862 | 4.0 | 900 | 0.3381 | 0.8883 |
| 0.087 | 5.0 | 1125 | 0.4066 | 0.8783 |
| 0.0686 | 6.0 | 1350 | 0.4702 | 0.885 |
| 0.0747 | 7.0 | 1575 | 0.5496 | 0.8933 |
| 0.0256 | 8.0 | 1800 | 0.6539 | 0.885 |
| 0.0458 | 9.0 | 2025 | 0.6793 | 0.8883 |
| 0.0397 | 10.0 | 2250 | 0.7901 | 0.88 |
| 0.0252 | 11.0 | 2475 | 0.7599 | 0.8967 |
| 0.0212 | 12.0 | 2700 | 0.8640 | 0.89 |
| 0.033 | 13.0 | 2925 | 0.8965 | 0.88 |
| 0.0005 | 14.0 | 3150 | 0.8332 | 0.895 |
| 0.0068 | 15.0 | 3375 | 0.9483 | 0.8717 |
| 0.0008 | 16.0 | 3600 | 1.0157 | 0.8783 |
| 0.0016 | 17.0 | 3825 | 0.8948 | 0.8867 |
| 0.0002 | 18.0 | 4050 | 0.8418 | 0.895 |
| 0.0 | 19.0 | 4275 | 0.8994 | 0.8983 |
| 0.0044 | 20.0 | 4500 | 0.8798 | 0.9083 |
| 0.0 | 21.0 | 4725 | 1.0795 | 0.8717 |
| 0.0009 | 22.0 | 4950 | 1.0744 | 0.885 |
| 0.0066 | 23.0 | 5175 | 0.9462 | 0.8883 |
| 0.0 | 24.0 | 5400 | 0.8715 | 0.9017 |
| 0.0088 | 25.0 | 5625 | 0.9553 | 0.8983 |
| 0.0 | 26.0 | 5850 | 0.9300 | 0.9 |
| 0.0271 | 27.0 | 6075 | 0.9362 | 0.8967 |
| 0.0 | 28.0 | 6300 | 0.9453 | 0.8967 |
| 0.0056 | 29.0 | 6525 | 1.0241 | 0.8917 |
| 0.0134 | 30.0 | 6750 | 0.9852 | 0.9017 |
| 0.0062 | 31.0 | 6975 | 0.9946 | 0.9017 |
| 0.0 | 32.0 | 7200 | 1.0433 | 0.89 |
| 0.0 | 33.0 | 7425 | 0.9693 | 0.8983 |
| 0.0 | 34.0 | 7650 | 0.9977 | 0.8983 |
| 0.0 | 35.0 | 7875 | 0.9577 | 0.9033 |
| 0.0 | 36.0 | 8100 | 0.9692 | 0.8967 |
| 0.0 | 37.0 | 8325 | 0.9675 | 0.8967 |
| 0.0 | 38.0 | 8550 | 0.9775 | 0.9 |
| 0.0 | 39.0 | 8775 | 0.9600 | 0.9 |
| 0.0 | 40.0 | 9000 | 0.9503 | 0.8983 |
| 0.0 | 41.0 | 9225 | 0.9634 | 0.895 |
| 0.0 | 42.0 | 9450 | 0.9624 | 0.9 |
| 0.0 | 43.0 | 9675 | 0.9737 | 0.9 |
| 0.0 | 44.0 | 9900 | 0.9743 | 0.9017 |
| 0.0 | 45.0 | 10125 | 0.9706 | 0.9 |
| 0.0 | 46.0 | 10350 | 0.9706 | 0.9 |
| 0.0019 | 47.0 | 10575 | 0.9708 | 0.9017 |
| 0.0 | 48.0 | 10800 | 0.9707 | 0.9017 |
| 0.0 | 49.0 | 11025 | 0.9710 | 0.9017 |
| 0.0 | 50.0 | 11250 | 0.9705 | 0.9033 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2