smids_3x_deit_tiny_adamax_001_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: 1.1177
- Accuracy: 0.8681
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 |
|---|---|---|---|---|
| 0.5567 | 1.0 | 226 | 0.5713 | 0.7462 |
| 0.3828 | 2.0 | 452 | 0.3656 | 0.8564 |
| 0.3502 | 3.0 | 678 | 0.3812 | 0.8381 |
| 0.276 | 4.0 | 904 | 0.4744 | 0.8280 |
| 0.2886 | 5.0 | 1130 | 0.4090 | 0.8447 |
| 0.2624 | 6.0 | 1356 | 0.3974 | 0.8598 |
| 0.1626 | 7.0 | 1582 | 0.4030 | 0.8781 |
| 0.1074 | 8.0 | 1808 | 0.4449 | 0.8631 |
| 0.184 | 9.0 | 2034 | 0.3919 | 0.8831 |
| 0.0836 | 10.0 | 2260 | 0.5392 | 0.8564 |
| 0.0539 | 11.0 | 2486 | 0.6608 | 0.8381 |
| 0.0511 | 12.0 | 2712 | 0.6349 | 0.8564 |
| 0.046 | 13.0 | 2938 | 0.6675 | 0.8848 |
| 0.054 | 14.0 | 3164 | 0.6371 | 0.8664 |
| 0.0362 | 15.0 | 3390 | 0.8363 | 0.8414 |
| 0.0153 | 16.0 | 3616 | 0.7151 | 0.8664 |
| 0.0583 | 17.0 | 3842 | 0.7946 | 0.8514 |
| 0.0315 | 18.0 | 4068 | 0.8456 | 0.8614 |
| 0.0078 | 19.0 | 4294 | 0.7428 | 0.8431 |
| 0.0034 | 20.0 | 4520 | 0.8571 | 0.8614 |
| 0.0384 | 21.0 | 4746 | 0.7674 | 0.8731 |
| 0.0465 | 22.0 | 4972 | 0.6983 | 0.8715 |
| 0.0099 | 23.0 | 5198 | 1.0003 | 0.8581 |
| 0.0034 | 24.0 | 5424 | 0.9037 | 0.8631 |
| 0.0001 | 25.0 | 5650 | 0.9614 | 0.8548 |
| 0.0004 | 26.0 | 5876 | 1.0237 | 0.8548 |
| 0.0094 | 27.0 | 6102 | 0.8996 | 0.8698 |
| 0.0001 | 28.0 | 6328 | 0.9361 | 0.8765 |
| 0.0 | 29.0 | 6554 | 1.0528 | 0.8614 |
| 0.0001 | 30.0 | 6780 | 0.9933 | 0.8614 |
| 0.0001 | 31.0 | 7006 | 1.0600 | 0.8614 |
| 0.0 | 32.0 | 7232 | 1.0559 | 0.8664 |
| 0.0031 | 33.0 | 7458 | 1.0415 | 0.8598 |
| 0.0 | 34.0 | 7684 | 1.1002 | 0.8648 |
| 0.0 | 35.0 | 7910 | 1.0102 | 0.8731 |
| 0.0 | 36.0 | 8136 | 1.0422 | 0.8731 |
| 0.0 | 37.0 | 8362 | 1.0448 | 0.8681 |
| 0.0 | 38.0 | 8588 | 1.0235 | 0.8698 |
| 0.0 | 39.0 | 8814 | 1.0543 | 0.8648 |
| 0.0 | 40.0 | 9040 | 1.0808 | 0.8681 |
| 0.0029 | 41.0 | 9266 | 1.0840 | 0.8698 |
| 0.0032 | 42.0 | 9492 | 1.0669 | 0.8698 |
| 0.0 | 43.0 | 9718 | 1.1155 | 0.8631 |
| 0.0 | 44.0 | 9944 | 1.0986 | 0.8698 |
| 0.0 | 45.0 | 10170 | 1.1093 | 0.8664 |
| 0.0 | 46.0 | 10396 | 1.1032 | 0.8664 |
| 0.0 | 47.0 | 10622 | 1.1140 | 0.8664 |
| 0.0 | 48.0 | 10848 | 1.1176 | 0.8681 |
| 0.0 | 49.0 | 11074 | 1.1190 | 0.8681 |
| 0.0 | 50.0 | 11300 | 1.1177 | 0.8681 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2