smids_1x_deit_small_sgd_00001_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: 1.0422
- Accuracy: 0.4983
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: 1e-05
- 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 |
|---|---|---|---|---|
| 1.0773 | 1.0 | 75 | 1.0761 | 0.4267 |
| 1.0923 | 2.0 | 150 | 1.0743 | 0.43 |
| 1.0827 | 3.0 | 225 | 1.0725 | 0.4317 |
| 1.0592 | 4.0 | 300 | 1.0709 | 0.4333 |
| 1.0688 | 5.0 | 375 | 1.0693 | 0.43 |
| 1.0722 | 6.0 | 450 | 1.0678 | 0.4317 |
| 1.0759 | 7.0 | 525 | 1.0664 | 0.4317 |
| 1.0583 | 8.0 | 600 | 1.0650 | 0.4333 |
| 1.0565 | 9.0 | 675 | 1.0637 | 0.4383 |
| 1.0589 | 10.0 | 750 | 1.0624 | 0.44 |
| 1.0608 | 11.0 | 825 | 1.0612 | 0.4483 |
| 1.0706 | 12.0 | 900 | 1.0600 | 0.4517 |
| 1.0517 | 13.0 | 975 | 1.0589 | 0.4567 |
| 1.0525 | 14.0 | 1050 | 1.0579 | 0.4567 |
| 1.0257 | 15.0 | 1125 | 1.0569 | 0.4583 |
| 1.0608 | 16.0 | 1200 | 1.0559 | 0.4617 |
| 1.0548 | 17.0 | 1275 | 1.0550 | 0.46 |
| 1.0482 | 18.0 | 1350 | 1.0541 | 0.4617 |
| 1.0606 | 19.0 | 1425 | 1.0533 | 0.4633 |
| 1.0832 | 20.0 | 1500 | 1.0524 | 0.4667 |
| 1.0387 | 21.0 | 1575 | 1.0517 | 0.4717 |
| 1.0524 | 22.0 | 1650 | 1.0510 | 0.4733 |
| 1.043 | 23.0 | 1725 | 1.0503 | 0.4733 |
| 1.0404 | 24.0 | 1800 | 1.0496 | 0.475 |
| 1.0507 | 25.0 | 1875 | 1.0490 | 0.4767 |
| 1.026 | 26.0 | 1950 | 1.0484 | 0.48 |
| 1.0409 | 27.0 | 2025 | 1.0478 | 0.48 |
| 1.0569 | 28.0 | 2100 | 1.0473 | 0.4867 |
| 1.0416 | 29.0 | 2175 | 1.0468 | 0.4867 |
| 1.0319 | 30.0 | 2250 | 1.0463 | 0.4867 |
| 1.0368 | 31.0 | 2325 | 1.0459 | 0.49 |
| 1.0498 | 32.0 | 2400 | 1.0455 | 0.4933 |
| 1.0315 | 33.0 | 2475 | 1.0451 | 0.4933 |
| 1.0281 | 34.0 | 2550 | 1.0447 | 0.49 |
| 1.0165 | 35.0 | 2625 | 1.0444 | 0.4917 |
| 1.0233 | 36.0 | 2700 | 1.0441 | 0.4933 |
| 1.0217 | 37.0 | 2775 | 1.0438 | 0.4967 |
| 1.0413 | 38.0 | 2850 | 1.0435 | 0.4967 |
| 1.0419 | 39.0 | 2925 | 1.0433 | 0.4967 |
| 1.0408 | 40.0 | 3000 | 1.0431 | 0.4967 |
| 1.0269 | 41.0 | 3075 | 1.0429 | 0.4983 |
| 1.0155 | 42.0 | 3150 | 1.0428 | 0.4983 |
| 1.0319 | 43.0 | 3225 | 1.0426 | 0.4983 |
| 1.015 | 44.0 | 3300 | 1.0425 | 0.4983 |
| 1.0304 | 45.0 | 3375 | 1.0424 | 0.4983 |
| 1.037 | 46.0 | 3450 | 1.0424 | 0.4983 |
| 1.0444 | 47.0 | 3525 | 1.0423 | 0.4983 |
| 1.0465 | 48.0 | 3600 | 1.0423 | 0.4983 |
| 1.0337 | 49.0 | 3675 | 1.0423 | 0.4983 |
| 1.0221 | 50.0 | 3750 | 1.0422 | 0.4983 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0