hushem_40x_deit_tiny_sgd_00001_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: 1.4218
- Accuracy: 0.2444
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 |
|---|---|---|---|---|
| 1.4676 | 1.0 | 215 | 1.4979 | 0.2444 |
| 1.453 | 2.0 | 430 | 1.4941 | 0.2444 |
| 1.4514 | 3.0 | 645 | 1.4903 | 0.2444 |
| 1.4264 | 4.0 | 860 | 1.4866 | 0.2222 |
| 1.4845 | 5.0 | 1075 | 1.4831 | 0.2222 |
| 1.4049 | 6.0 | 1290 | 1.4797 | 0.2444 |
| 1.408 | 7.0 | 1505 | 1.4764 | 0.2444 |
| 1.4075 | 8.0 | 1720 | 1.4733 | 0.2444 |
| 1.4274 | 9.0 | 1935 | 1.4702 | 0.2444 |
| 1.4165 | 10.0 | 2150 | 1.4673 | 0.2444 |
| 1.3408 | 11.0 | 2365 | 1.4645 | 0.2444 |
| 1.387 | 12.0 | 2580 | 1.4617 | 0.2444 |
| 1.3966 | 13.0 | 2795 | 1.4591 | 0.2444 |
| 1.3631 | 14.0 | 3010 | 1.4566 | 0.2444 |
| 1.3966 | 15.0 | 3225 | 1.4542 | 0.2444 |
| 1.4085 | 16.0 | 3440 | 1.4520 | 0.2444 |
| 1.3593 | 17.0 | 3655 | 1.4498 | 0.2444 |
| 1.3872 | 18.0 | 3870 | 1.4477 | 0.2444 |
| 1.3857 | 19.0 | 4085 | 1.4457 | 0.2222 |
| 1.3961 | 20.0 | 4300 | 1.4439 | 0.2222 |
| 1.3725 | 21.0 | 4515 | 1.4421 | 0.2222 |
| 1.3634 | 22.0 | 4730 | 1.4405 | 0.2222 |
| 1.3404 | 23.0 | 4945 | 1.4389 | 0.2222 |
| 1.2947 | 24.0 | 5160 | 1.4374 | 0.2222 |
| 1.3286 | 25.0 | 5375 | 1.4360 | 0.2222 |
| 1.3597 | 26.0 | 5590 | 1.4346 | 0.2222 |
| 1.3935 | 27.0 | 5805 | 1.4334 | 0.2222 |
| 1.3126 | 28.0 | 6020 | 1.4322 | 0.2222 |
| 1.3862 | 29.0 | 6235 | 1.4311 | 0.2222 |
| 1.345 | 30.0 | 6450 | 1.4301 | 0.2222 |
| 1.3332 | 31.0 | 6665 | 1.4291 | 0.2222 |
| 1.3215 | 32.0 | 6880 | 1.4283 | 0.2222 |
| 1.3753 | 33.0 | 7095 | 1.4274 | 0.2222 |
| 1.3397 | 34.0 | 7310 | 1.4267 | 0.2222 |
| 1.3085 | 35.0 | 7525 | 1.4260 | 0.2222 |
| 1.3414 | 36.0 | 7740 | 1.4254 | 0.2222 |
| 1.3773 | 37.0 | 7955 | 1.4248 | 0.2222 |
| 1.2916 | 38.0 | 8170 | 1.4243 | 0.2222 |
| 1.2953 | 39.0 | 8385 | 1.4238 | 0.2222 |
| 1.3053 | 40.0 | 8600 | 1.4234 | 0.2222 |
| 1.3127 | 41.0 | 8815 | 1.4230 | 0.2222 |
| 1.2816 | 42.0 | 9030 | 1.4227 | 0.2222 |
| 1.3493 | 43.0 | 9245 | 1.4225 | 0.2222 |
| 1.3258 | 44.0 | 9460 | 1.4223 | 0.2222 |
| 1.3441 | 45.0 | 9675 | 1.4221 | 0.2444 |
| 1.2959 | 46.0 | 9890 | 1.4220 | 0.2444 |
| 1.3609 | 47.0 | 10105 | 1.4219 | 0.2444 |
| 1.276 | 48.0 | 10320 | 1.4219 | 0.2444 |
| 1.2931 | 49.0 | 10535 | 1.4218 | 0.2444 |
| 1.3383 | 50.0 | 10750 | 1.4218 | 0.2444 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2