smids_3x_deit_small_sgd_001_fold5
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: 0.2955
- Accuracy: 0.875
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 |
|---|---|---|---|---|
| 0.8786 | 1.0 | 225 | 0.8329 | 0.6917 |
| 0.6756 | 2.0 | 450 | 0.6561 | 0.7467 |
| 0.5645 | 3.0 | 675 | 0.5574 | 0.7867 |
| 0.4671 | 4.0 | 900 | 0.4924 | 0.8067 |
| 0.3977 | 5.0 | 1125 | 0.4538 | 0.82 |
| 0.4177 | 6.0 | 1350 | 0.4235 | 0.8367 |
| 0.3878 | 7.0 | 1575 | 0.4039 | 0.8417 |
| 0.4378 | 8.0 | 1800 | 0.3874 | 0.8433 |
| 0.3622 | 9.0 | 2025 | 0.3772 | 0.8483 |
| 0.345 | 10.0 | 2250 | 0.3683 | 0.8517 |
| 0.3638 | 11.0 | 2475 | 0.3631 | 0.8533 |
| 0.3441 | 12.0 | 2700 | 0.3527 | 0.8583 |
| 0.3313 | 13.0 | 2925 | 0.3447 | 0.865 |
| 0.2901 | 14.0 | 3150 | 0.3405 | 0.8633 |
| 0.2288 | 15.0 | 3375 | 0.3333 | 0.865 |
| 0.3024 | 16.0 | 3600 | 0.3306 | 0.865 |
| 0.2544 | 17.0 | 3825 | 0.3278 | 0.8683 |
| 0.299 | 18.0 | 4050 | 0.3253 | 0.8667 |
| 0.2662 | 19.0 | 4275 | 0.3235 | 0.8667 |
| 0.2847 | 20.0 | 4500 | 0.3172 | 0.8683 |
| 0.2132 | 21.0 | 4725 | 0.3164 | 0.8667 |
| 0.2384 | 22.0 | 4950 | 0.3131 | 0.8717 |
| 0.2264 | 23.0 | 5175 | 0.3102 | 0.8733 |
| 0.2574 | 24.0 | 5400 | 0.3121 | 0.8667 |
| 0.2327 | 25.0 | 5625 | 0.3088 | 0.8683 |
| 0.2687 | 26.0 | 5850 | 0.3062 | 0.8667 |
| 0.28 | 27.0 | 6075 | 0.3048 | 0.8667 |
| 0.2544 | 28.0 | 6300 | 0.3033 | 0.8683 |
| 0.2339 | 29.0 | 6525 | 0.3018 | 0.87 |
| 0.2 | 30.0 | 6750 | 0.3023 | 0.8733 |
| 0.1716 | 31.0 | 6975 | 0.3008 | 0.8733 |
| 0.2152 | 32.0 | 7200 | 0.2995 | 0.8717 |
| 0.2129 | 33.0 | 7425 | 0.2994 | 0.8733 |
| 0.1758 | 34.0 | 7650 | 0.2988 | 0.875 |
| 0.1848 | 35.0 | 7875 | 0.3009 | 0.875 |
| 0.2108 | 36.0 | 8100 | 0.2991 | 0.875 |
| 0.2223 | 37.0 | 8325 | 0.2978 | 0.875 |
| 0.1689 | 38.0 | 8550 | 0.2975 | 0.8733 |
| 0.1768 | 39.0 | 8775 | 0.2974 | 0.8767 |
| 0.2093 | 40.0 | 9000 | 0.2965 | 0.8733 |
| 0.1994 | 41.0 | 9225 | 0.2966 | 0.8733 |
| 0.2309 | 42.0 | 9450 | 0.2956 | 0.8733 |
| 0.2412 | 43.0 | 9675 | 0.2974 | 0.8767 |
| 0.2229 | 44.0 | 9900 | 0.2958 | 0.875 |
| 0.2153 | 45.0 | 10125 | 0.2965 | 0.8767 |
| 0.1978 | 46.0 | 10350 | 0.2959 | 0.8767 |
| 0.2092 | 47.0 | 10575 | 0.2956 | 0.875 |
| 0.2126 | 48.0 | 10800 | 0.2958 | 0.875 |
| 0.2109 | 49.0 | 11025 | 0.2956 | 0.875 |
| 0.1728 | 50.0 | 11250 | 0.2955 | 0.875 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2