smids_10x_deit_tiny_rms_0001_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: 1.4754
- Accuracy: 0.83
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 |
|---|---|---|---|---|
| 0.8611 | 1.0 | 750 | 0.8101 | 0.59 |
| 0.7337 | 2.0 | 1500 | 0.7328 | 0.5933 |
| 0.6092 | 3.0 | 2250 | 0.7436 | 0.6133 |
| 0.4788 | 4.0 | 3000 | 0.6067 | 0.7117 |
| 0.5206 | 5.0 | 3750 | 0.5059 | 0.79 |
| 0.4534 | 6.0 | 4500 | 0.5164 | 0.7733 |
| 0.4278 | 7.0 | 5250 | 0.5196 | 0.7883 |
| 0.421 | 8.0 | 6000 | 0.4508 | 0.82 |
| 0.3903 | 9.0 | 6750 | 0.4527 | 0.8067 |
| 0.3955 | 10.0 | 7500 | 0.4845 | 0.7867 |
| 0.4005 | 11.0 | 8250 | 0.4149 | 0.835 |
| 0.3221 | 12.0 | 9000 | 0.4996 | 0.815 |
| 0.3587 | 13.0 | 9750 | 0.4119 | 0.8433 |
| 0.2736 | 14.0 | 10500 | 0.4909 | 0.85 |
| 0.2907 | 15.0 | 11250 | 0.4508 | 0.8217 |
| 0.2288 | 16.0 | 12000 | 0.5192 | 0.825 |
| 0.2378 | 17.0 | 12750 | 0.4692 | 0.8383 |
| 0.2297 | 18.0 | 13500 | 0.5883 | 0.8217 |
| 0.2065 | 19.0 | 14250 | 0.4741 | 0.8483 |
| 0.1668 | 20.0 | 15000 | 0.5264 | 0.82 |
| 0.1867 | 21.0 | 15750 | 0.5865 | 0.8217 |
| 0.1397 | 22.0 | 16500 | 0.5812 | 0.8283 |
| 0.15 | 23.0 | 17250 | 0.7165 | 0.83 |
| 0.083 | 24.0 | 18000 | 0.6760 | 0.8267 |
| 0.3309 | 25.0 | 18750 | 0.8204 | 0.79 |
| 0.6907 | 26.0 | 19500 | 0.6426 | 0.7217 |
| 0.4092 | 27.0 | 20250 | 0.7365 | 0.72 |
| 0.4562 | 28.0 | 21000 | 0.4704 | 0.7817 |
| 0.4588 | 29.0 | 21750 | 0.4367 | 0.7983 |
| 0.4089 | 30.0 | 22500 | 0.5011 | 0.7883 |
| 0.4005 | 31.0 | 23250 | 0.5056 | 0.805 |
| 0.2986 | 32.0 | 24000 | 0.4328 | 0.83 |
| 0.3319 | 33.0 | 24750 | 0.5242 | 0.805 |
| 0.2453 | 34.0 | 25500 | 0.4883 | 0.8383 |
| 0.2459 | 35.0 | 26250 | 0.4886 | 0.8233 |
| 0.2873 | 36.0 | 27000 | 0.5046 | 0.8217 |
| 0.1814 | 37.0 | 27750 | 0.5539 | 0.8133 |
| 0.2073 | 38.0 | 28500 | 0.5379 | 0.8233 |
| 0.2284 | 39.0 | 29250 | 0.5765 | 0.815 |
| 0.1577 | 40.0 | 30000 | 0.6100 | 0.8333 |
| 0.1325 | 41.0 | 30750 | 0.6730 | 0.8333 |
| 0.1239 | 42.0 | 31500 | 0.7561 | 0.8333 |
| 0.0741 | 43.0 | 32250 | 0.8428 | 0.8217 |
| 0.0999 | 44.0 | 33000 | 0.9214 | 0.8167 |
| 0.0863 | 45.0 | 33750 | 0.9203 | 0.8367 |
| 0.0855 | 46.0 | 34500 | 1.0656 | 0.83 |
| 0.064 | 47.0 | 35250 | 1.2523 | 0.815 |
| 0.0407 | 48.0 | 36000 | 1.2659 | 0.83 |
| 0.0522 | 49.0 | 36750 | 1.4174 | 0.835 |
| 0.0025 | 50.0 | 37500 | 1.4754 | 0.83 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2