hushem_1x_deit_tiny_rms_lr001_fold5
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: 1.0599
- Accuracy: 0.5366
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 |
|---|---|---|---|---|
| No log | 1.0 | 6 | 2.6067 | 0.2439 |
| 4.0909 | 2.0 | 12 | 1.8085 | 0.2439 |
| 4.0909 | 3.0 | 18 | 1.7809 | 0.2439 |
| 2.0948 | 4.0 | 24 | 1.7586 | 0.2439 |
| 1.6719 | 5.0 | 30 | 1.5135 | 0.2439 |
| 1.6719 | 6.0 | 36 | 1.7849 | 0.2683 |
| 1.5694 | 7.0 | 42 | 1.4636 | 0.3902 |
| 1.5694 | 8.0 | 48 | 1.4809 | 0.2683 |
| 1.519 | 9.0 | 54 | 1.3587 | 0.3415 |
| 1.5241 | 10.0 | 60 | 1.3823 | 0.2439 |
| 1.5241 | 11.0 | 66 | 1.3645 | 0.3415 |
| 1.4557 | 12.0 | 72 | 1.2525 | 0.3659 |
| 1.4557 | 13.0 | 78 | 1.2955 | 0.3171 |
| 1.3674 | 14.0 | 84 | 1.3174 | 0.3415 |
| 1.3868 | 15.0 | 90 | 1.2787 | 0.3415 |
| 1.3868 | 16.0 | 96 | 1.6408 | 0.2683 |
| 1.3152 | 17.0 | 102 | 1.2750 | 0.3171 |
| 1.3152 | 18.0 | 108 | 1.0560 | 0.5366 |
| 1.2693 | 19.0 | 114 | 1.3256 | 0.4878 |
| 1.2554 | 20.0 | 120 | 1.3190 | 0.3902 |
| 1.2554 | 21.0 | 126 | 1.2498 | 0.3902 |
| 1.1813 | 22.0 | 132 | 1.2514 | 0.3902 |
| 1.1813 | 23.0 | 138 | 1.0907 | 0.5366 |
| 1.1113 | 24.0 | 144 | 1.2821 | 0.3415 |
| 1.1728 | 25.0 | 150 | 1.1433 | 0.4878 |
| 1.1728 | 26.0 | 156 | 1.0143 | 0.5366 |
| 1.1037 | 27.0 | 162 | 0.9542 | 0.5854 |
| 1.1037 | 28.0 | 168 | 1.1443 | 0.5122 |
| 1.0914 | 29.0 | 174 | 1.0904 | 0.4878 |
| 1.1385 | 30.0 | 180 | 1.1995 | 0.4146 |
| 1.1385 | 31.0 | 186 | 0.9746 | 0.6098 |
| 1.0636 | 32.0 | 192 | 1.1104 | 0.4634 |
| 1.0636 | 33.0 | 198 | 0.9890 | 0.6098 |
| 1.0129 | 34.0 | 204 | 1.2113 | 0.3902 |
| 0.999 | 35.0 | 210 | 1.0001 | 0.6098 |
| 0.999 | 36.0 | 216 | 1.0972 | 0.5122 |
| 0.9802 | 37.0 | 222 | 1.1639 | 0.4390 |
| 0.9802 | 38.0 | 228 | 1.0730 | 0.5122 |
| 0.9625 | 39.0 | 234 | 1.0471 | 0.4878 |
| 0.9424 | 40.0 | 240 | 1.0692 | 0.5366 |
| 0.9424 | 41.0 | 246 | 1.0654 | 0.5366 |
| 0.9521 | 42.0 | 252 | 1.0599 | 0.5366 |
| 0.9521 | 43.0 | 258 | 1.0599 | 0.5366 |
| 0.9184 | 44.0 | 264 | 1.0599 | 0.5366 |
| 0.9335 | 45.0 | 270 | 1.0599 | 0.5366 |
| 0.9335 | 46.0 | 276 | 1.0599 | 0.5366 |
| 0.9251 | 47.0 | 282 | 1.0599 | 0.5366 |
| 0.9251 | 48.0 | 288 | 1.0599 | 0.5366 |
| 0.9168 | 49.0 | 294 | 1.0599 | 0.5366 |
| 0.8964 | 50.0 | 300 | 1.0599 | 0.5366 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1