smids_10x_deit_tiny_adamax_00001_fold1
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: 0.8545
- Accuracy: 0.9115
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 |
|---|---|---|---|---|
| 0.2768 | 1.0 | 751 | 0.3271 | 0.8681 |
| 0.2559 | 2.0 | 1502 | 0.2686 | 0.8932 |
| 0.1723 | 3.0 | 2253 | 0.2752 | 0.8932 |
| 0.1343 | 4.0 | 3004 | 0.2784 | 0.8898 |
| 0.1389 | 5.0 | 3755 | 0.2896 | 0.8965 |
| 0.1143 | 6.0 | 4506 | 0.3456 | 0.8881 |
| 0.0797 | 7.0 | 5257 | 0.3441 | 0.8865 |
| 0.0583 | 8.0 | 6008 | 0.3964 | 0.8998 |
| 0.0472 | 9.0 | 6759 | 0.4458 | 0.8915 |
| 0.0757 | 10.0 | 7510 | 0.4767 | 0.8982 |
| 0.0191 | 11.0 | 8261 | 0.5147 | 0.8915 |
| 0.031 | 12.0 | 9012 | 0.5873 | 0.8898 |
| 0.0022 | 13.0 | 9763 | 0.6291 | 0.8982 |
| 0.0003 | 14.0 | 10514 | 0.6449 | 0.9048 |
| 0.0014 | 15.0 | 11265 | 0.6651 | 0.8982 |
| 0.0237 | 16.0 | 12016 | 0.7228 | 0.9015 |
| 0.0016 | 17.0 | 12767 | 0.7272 | 0.8948 |
| 0.0001 | 18.0 | 13518 | 0.7560 | 0.9032 |
| 0.0001 | 19.0 | 14269 | 0.7571 | 0.8982 |
| 0.0 | 20.0 | 15020 | 0.7689 | 0.9048 |
| 0.0 | 21.0 | 15771 | 0.7584 | 0.9048 |
| 0.0 | 22.0 | 16522 | 0.7967 | 0.9032 |
| 0.0 | 23.0 | 17273 | 0.7987 | 0.9065 |
| 0.0001 | 24.0 | 18024 | 0.8298 | 0.9065 |
| 0.0 | 25.0 | 18775 | 0.8022 | 0.9098 |
| 0.0 | 26.0 | 19526 | 0.8054 | 0.9098 |
| 0.0 | 27.0 | 20277 | 0.8124 | 0.9065 |
| 0.0 | 28.0 | 21028 | 0.8128 | 0.9082 |
| 0.0194 | 29.0 | 21779 | 0.8361 | 0.9015 |
| 0.0 | 30.0 | 22530 | 0.8316 | 0.9065 |
| 0.0 | 31.0 | 23281 | 0.8255 | 0.9132 |
| 0.0 | 32.0 | 24032 | 0.8225 | 0.9115 |
| 0.0 | 33.0 | 24783 | 0.8294 | 0.9098 |
| 0.0 | 34.0 | 25534 | 0.8377 | 0.9082 |
| 0.0 | 35.0 | 26285 | 0.8477 | 0.9032 |
| 0.0 | 36.0 | 27036 | 0.8439 | 0.9115 |
| 0.0 | 37.0 | 27787 | 0.8492 | 0.9065 |
| 0.0 | 38.0 | 28538 | 0.8435 | 0.9098 |
| 0.0 | 39.0 | 29289 | 0.8490 | 0.9098 |
| 0.0 | 40.0 | 30040 | 0.8482 | 0.9115 |
| 0.0 | 41.0 | 30791 | 0.8506 | 0.9065 |
| 0.0 | 42.0 | 31542 | 0.8515 | 0.9082 |
| 0.0 | 43.0 | 32293 | 0.8517 | 0.9115 |
| 0.0 | 44.0 | 33044 | 0.8525 | 0.9115 |
| 0.0 | 45.0 | 33795 | 0.8550 | 0.9048 |
| 0.0 | 46.0 | 34546 | 0.8557 | 0.9082 |
| 0.0 | 47.0 | 35297 | 0.8547 | 0.9115 |
| 0.0 | 48.0 | 36048 | 0.8545 | 0.9115 |
| 0.0 | 49.0 | 36799 | 0.8544 | 0.9115 |
| 0.0 | 50.0 | 37550 | 0.8545 | 0.9115 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2