smids_10x_deit_tiny_rms_001_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: 2.1068
- Accuracy: 0.8020
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 |
|---|---|---|---|---|
| 0.8332 | 1.0 | 750 | 0.7567 | 0.6090 |
| 0.7395 | 2.0 | 1500 | 0.7599 | 0.6123 |
| 0.682 | 3.0 | 2250 | 0.6859 | 0.6905 |
| 0.725 | 4.0 | 3000 | 0.6463 | 0.7171 |
| 0.6632 | 5.0 | 3750 | 0.6560 | 0.7238 |
| 0.5777 | 6.0 | 4500 | 0.6347 | 0.7072 |
| 0.6357 | 7.0 | 5250 | 0.6141 | 0.7321 |
| 0.595 | 8.0 | 6000 | 0.6313 | 0.7121 |
| 0.5551 | 9.0 | 6750 | 0.6406 | 0.6955 |
| 0.5544 | 10.0 | 7500 | 0.5482 | 0.7720 |
| 0.5611 | 11.0 | 8250 | 0.5288 | 0.7704 |
| 0.6632 | 12.0 | 9000 | 0.5868 | 0.7537 |
| 0.5709 | 13.0 | 9750 | 0.6149 | 0.7288 |
| 0.4511 | 14.0 | 10500 | 0.4977 | 0.8020 |
| 0.4295 | 15.0 | 11250 | 0.5625 | 0.7770 |
| 0.4618 | 16.0 | 12000 | 0.5273 | 0.7837 |
| 0.4342 | 17.0 | 12750 | 0.5207 | 0.7804 |
| 0.4253 | 18.0 | 13500 | 0.5301 | 0.7720 |
| 0.4352 | 19.0 | 14250 | 0.5236 | 0.7754 |
| 0.418 | 20.0 | 15000 | 0.5318 | 0.7804 |
| 0.4496 | 21.0 | 15750 | 0.5216 | 0.7970 |
| 0.4003 | 22.0 | 16500 | 0.5391 | 0.7720 |
| 0.4411 | 23.0 | 17250 | 0.4904 | 0.8003 |
| 0.3266 | 24.0 | 18000 | 0.5436 | 0.7854 |
| 0.3733 | 25.0 | 18750 | 0.6780 | 0.7521 |
| 0.3536 | 26.0 | 19500 | 0.5100 | 0.8003 |
| 0.4154 | 27.0 | 20250 | 0.5545 | 0.8020 |
| 0.414 | 28.0 | 21000 | 0.5841 | 0.7937 |
| 0.3146 | 29.0 | 21750 | 0.5867 | 0.7887 |
| 0.3401 | 30.0 | 22500 | 0.5923 | 0.7987 |
| 0.2331 | 31.0 | 23250 | 0.6367 | 0.7837 |
| 0.238 | 32.0 | 24000 | 0.6276 | 0.8070 |
| 0.209 | 33.0 | 24750 | 0.6337 | 0.8070 |
| 0.2121 | 34.0 | 25500 | 0.6961 | 0.7854 |
| 0.2544 | 35.0 | 26250 | 0.7936 | 0.7870 |
| 0.2442 | 36.0 | 27000 | 0.7270 | 0.7970 |
| 0.2459 | 37.0 | 27750 | 0.7553 | 0.8020 |
| 0.1428 | 38.0 | 28500 | 0.8600 | 0.7987 |
| 0.0788 | 39.0 | 29250 | 0.9727 | 0.7937 |
| 0.1811 | 40.0 | 30000 | 1.0324 | 0.7937 |
| 0.1405 | 41.0 | 30750 | 1.0037 | 0.8103 |
| 0.1282 | 42.0 | 31500 | 1.1830 | 0.7937 |
| 0.0664 | 43.0 | 32250 | 1.2624 | 0.7970 |
| 0.04 | 44.0 | 33000 | 1.4942 | 0.7987 |
| 0.0582 | 45.0 | 33750 | 1.4631 | 0.8103 |
| 0.0738 | 46.0 | 34500 | 1.6687 | 0.8120 |
| 0.0282 | 47.0 | 35250 | 1.8321 | 0.8087 |
| 0.0021 | 48.0 | 36000 | 1.9181 | 0.8087 |
| 0.01 | 49.0 | 36750 | 2.0036 | 0.8037 |
| 0.0004 | 50.0 | 37500 | 2.1068 | 0.8020 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2