smids_10x_deit_tiny_rms_0001_fold2
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: 1.7512
- Accuracy: 0.8469
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 |
|---|---|---|---|---|
| 0.7916 | 1.0 | 750 | 0.7607 | 0.5990 |
| 0.7137 | 2.0 | 1500 | 0.7292 | 0.6922 |
| 0.7389 | 3.0 | 2250 | 0.6487 | 0.6905 |
| 0.641 | 4.0 | 3000 | 0.6309 | 0.7055 |
| 0.6314 | 5.0 | 3750 | 0.6856 | 0.6905 |
| 0.5224 | 6.0 | 4500 | 0.5750 | 0.7554 |
| 0.5522 | 7.0 | 5250 | 0.5498 | 0.7537 |
| 0.5542 | 8.0 | 6000 | 0.5375 | 0.7587 |
| 0.5077 | 9.0 | 6750 | 0.5514 | 0.7604 |
| 0.4953 | 10.0 | 7500 | 0.5328 | 0.7870 |
| 0.4839 | 11.0 | 8250 | 0.5735 | 0.7454 |
| 0.5867 | 12.0 | 9000 | 0.5344 | 0.7820 |
| 0.5019 | 13.0 | 9750 | 0.5724 | 0.7720 |
| 0.4158 | 14.0 | 10500 | 0.5203 | 0.7870 |
| 0.338 | 15.0 | 11250 | 0.5452 | 0.7903 |
| 0.3872 | 16.0 | 12000 | 0.4927 | 0.8070 |
| 0.3271 | 17.0 | 12750 | 0.5792 | 0.8070 |
| 0.3473 | 18.0 | 13500 | 0.4754 | 0.8203 |
| 0.3721 | 19.0 | 14250 | 0.4949 | 0.7887 |
| 0.3629 | 20.0 | 15000 | 0.4870 | 0.8136 |
| 0.3126 | 21.0 | 15750 | 0.4639 | 0.8303 |
| 0.2547 | 22.0 | 16500 | 0.5296 | 0.8236 |
| 0.3002 | 23.0 | 17250 | 0.5506 | 0.8003 |
| 0.2834 | 24.0 | 18000 | 0.5134 | 0.8186 |
| 0.2447 | 25.0 | 18750 | 0.6771 | 0.8053 |
| 0.211 | 26.0 | 19500 | 0.5840 | 0.8120 |
| 0.2265 | 27.0 | 20250 | 0.6139 | 0.8369 |
| 0.2018 | 28.0 | 21000 | 0.7828 | 0.8203 |
| 0.1385 | 29.0 | 21750 | 0.7229 | 0.8186 |
| 0.1626 | 30.0 | 22500 | 0.6948 | 0.8369 |
| 0.1046 | 31.0 | 23250 | 0.7987 | 0.8419 |
| 0.0986 | 32.0 | 24000 | 0.6857 | 0.8552 |
| 0.1014 | 33.0 | 24750 | 0.9042 | 0.8303 |
| 0.0734 | 34.0 | 25500 | 0.9163 | 0.8403 |
| 0.03 | 35.0 | 26250 | 0.9836 | 0.8353 |
| 0.0303 | 36.0 | 27000 | 0.9935 | 0.8436 |
| 0.0569 | 37.0 | 27750 | 1.1088 | 0.8286 |
| 0.0188 | 38.0 | 28500 | 1.2743 | 0.8319 |
| 0.0192 | 39.0 | 29250 | 1.3350 | 0.8386 |
| 0.032 | 40.0 | 30000 | 1.3569 | 0.8419 |
| 0.01 | 41.0 | 30750 | 1.2223 | 0.8353 |
| 0.0249 | 42.0 | 31500 | 1.4712 | 0.8436 |
| 0.0031 | 43.0 | 32250 | 1.5420 | 0.8403 |
| 0.0008 | 44.0 | 33000 | 1.5564 | 0.8419 |
| 0.026 | 45.0 | 33750 | 1.5884 | 0.8502 |
| 0.0017 | 46.0 | 34500 | 1.7455 | 0.8386 |
| 0.0001 | 47.0 | 35250 | 1.7636 | 0.8353 |
| 0.0001 | 48.0 | 36000 | 1.7269 | 0.8453 |
| 0.0 | 49.0 | 36750 | 1.7457 | 0.8453 |
| 0.0 | 50.0 | 37500 | 1.7512 | 0.8469 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2