hushem_40x_deit_small_adamax_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: 0.4380
- Accuracy: 0.9024
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 |
|---|---|---|---|---|
| 0.2763 | 1.0 | 220 | 0.4688 | 0.7561 |
| 0.0369 | 2.0 | 440 | 0.2296 | 0.9024 |
| 0.0054 | 3.0 | 660 | 0.3595 | 0.8537 |
| 0.0018 | 4.0 | 880 | 0.2500 | 0.9024 |
| 0.0006 | 5.0 | 1100 | 0.2205 | 0.9268 |
| 0.0003 | 6.0 | 1320 | 0.2552 | 0.9024 |
| 0.0003 | 7.0 | 1540 | 0.2520 | 0.9024 |
| 0.0002 | 8.0 | 1760 | 0.2923 | 0.8780 |
| 0.0001 | 9.0 | 1980 | 0.2754 | 0.8780 |
| 0.0001 | 10.0 | 2200 | 0.2945 | 0.8780 |
| 0.0001 | 11.0 | 2420 | 0.2841 | 0.9024 |
| 0.0001 | 12.0 | 2640 | 0.3077 | 0.9024 |
| 0.0001 | 13.0 | 2860 | 0.3076 | 0.8780 |
| 0.0 | 14.0 | 3080 | 0.3160 | 0.8780 |
| 0.0 | 15.0 | 3300 | 0.2918 | 0.9024 |
| 0.0 | 16.0 | 3520 | 0.3305 | 0.8780 |
| 0.0 | 17.0 | 3740 | 0.3206 | 0.8780 |
| 0.0 | 18.0 | 3960 | 0.3174 | 0.8780 |
| 0.0 | 19.0 | 4180 | 0.3189 | 0.8780 |
| 0.0 | 20.0 | 4400 | 0.3130 | 0.9024 |
| 0.0 | 21.0 | 4620 | 0.3383 | 0.8780 |
| 0.0 | 22.0 | 4840 | 0.3473 | 0.8780 |
| 0.0 | 23.0 | 5060 | 0.3548 | 0.8780 |
| 0.0 | 24.0 | 5280 | 0.3221 | 0.8780 |
| 0.0 | 25.0 | 5500 | 0.3554 | 0.8780 |
| 0.0 | 26.0 | 5720 | 0.3715 | 0.8780 |
| 0.0 | 27.0 | 5940 | 0.3690 | 0.8780 |
| 0.0 | 28.0 | 6160 | 0.3648 | 0.8780 |
| 0.0 | 29.0 | 6380 | 0.3806 | 0.8780 |
| 0.0 | 30.0 | 6600 | 0.3725 | 0.8780 |
| 0.0 | 31.0 | 6820 | 0.4022 | 0.8780 |
| 0.0 | 32.0 | 7040 | 0.3871 | 0.8780 |
| 0.0 | 33.0 | 7260 | 0.4133 | 0.8780 |
| 0.0 | 34.0 | 7480 | 0.4117 | 0.8780 |
| 0.0 | 35.0 | 7700 | 0.3832 | 0.8780 |
| 0.0 | 36.0 | 7920 | 0.3977 | 0.8780 |
| 0.0 | 37.0 | 8140 | 0.3959 | 0.8780 |
| 0.0 | 38.0 | 8360 | 0.4687 | 0.8780 |
| 0.0 | 39.0 | 8580 | 0.4404 | 0.8780 |
| 0.0 | 40.0 | 8800 | 0.3819 | 0.9024 |
| 0.0 | 41.0 | 9020 | 0.4514 | 0.8780 |
| 0.0 | 42.0 | 9240 | 0.4623 | 0.8780 |
| 0.0 | 43.0 | 9460 | 0.4136 | 0.9024 |
| 0.0 | 44.0 | 9680 | 0.4401 | 0.9024 |
| 0.0 | 45.0 | 9900 | 0.4714 | 0.9024 |
| 0.0 | 46.0 | 10120 | 0.4588 | 0.9024 |
| 0.0 | 47.0 | 10340 | 0.4584 | 0.9024 |
| 0.0 | 48.0 | 10560 | 0.4588 | 0.9024 |
| 0.0 | 49.0 | 10780 | 0.4430 | 0.9024 |
| 0.0 | 50.0 | 11000 | 0.4380 | 0.9024 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2