hushem_40x_deit_small_adamax_00001_fold1
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: 1.3015
- Accuracy: 0.8444
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 |
|---|---|---|---|---|
| 0.238 | 1.0 | 215 | 0.7567 | 0.6667 |
| 0.0267 | 2.0 | 430 | 0.5445 | 0.7778 |
| 0.0051 | 3.0 | 645 | 0.6144 | 0.8 |
| 0.0012 | 4.0 | 860 | 0.6615 | 0.8 |
| 0.0005 | 5.0 | 1075 | 0.6553 | 0.8 |
| 0.0003 | 6.0 | 1290 | 0.6621 | 0.8222 |
| 0.0003 | 7.0 | 1505 | 0.6958 | 0.8222 |
| 0.0002 | 8.0 | 1720 | 0.7076 | 0.8222 |
| 0.0001 | 9.0 | 1935 | 0.7375 | 0.8222 |
| 0.0001 | 10.0 | 2150 | 0.7327 | 0.8222 |
| 0.0001 | 11.0 | 2365 | 0.7423 | 0.8222 |
| 0.0001 | 12.0 | 2580 | 0.7689 | 0.8 |
| 0.0001 | 13.0 | 2795 | 0.7876 | 0.8 |
| 0.0 | 14.0 | 3010 | 0.7990 | 0.8 |
| 0.0 | 15.0 | 3225 | 0.8203 | 0.8 |
| 0.0 | 16.0 | 3440 | 0.8447 | 0.8 |
| 0.0 | 17.0 | 3655 | 0.8558 | 0.8 |
| 0.0 | 18.0 | 3870 | 0.8774 | 0.8 |
| 0.0 | 19.0 | 4085 | 0.8896 | 0.8 |
| 0.0 | 20.0 | 4300 | 0.8965 | 0.8 |
| 0.0 | 21.0 | 4515 | 0.9254 | 0.8 |
| 0.0 | 22.0 | 4730 | 0.9318 | 0.8 |
| 0.0 | 23.0 | 4945 | 0.9571 | 0.8 |
| 0.0 | 24.0 | 5160 | 0.9711 | 0.8222 |
| 0.0 | 25.0 | 5375 | 0.9833 | 0.8222 |
| 0.0 | 26.0 | 5590 | 0.9915 | 0.8222 |
| 0.0 | 27.0 | 5805 | 1.0134 | 0.8222 |
| 0.0 | 28.0 | 6020 | 1.0327 | 0.8222 |
| 0.0 | 29.0 | 6235 | 1.0249 | 0.8222 |
| 0.0 | 30.0 | 6450 | 1.0679 | 0.8222 |
| 0.0 | 31.0 | 6665 | 1.0896 | 0.8222 |
| 0.0 | 32.0 | 6880 | 1.0990 | 0.8222 |
| 0.0 | 33.0 | 7095 | 1.1103 | 0.8222 |
| 0.0 | 34.0 | 7310 | 1.1167 | 0.8222 |
| 0.0 | 35.0 | 7525 | 1.1494 | 0.8222 |
| 0.0 | 36.0 | 7740 | 1.1474 | 0.8444 |
| 0.0 | 37.0 | 7955 | 1.1611 | 0.8444 |
| 0.0 | 38.0 | 8170 | 1.2104 | 0.8222 |
| 0.0 | 39.0 | 8385 | 1.1969 | 0.8444 |
| 0.0 | 40.0 | 8600 | 1.2127 | 0.8222 |
| 0.0 | 41.0 | 8815 | 1.2186 | 0.8444 |
| 0.0 | 42.0 | 9030 | 1.2356 | 0.8444 |
| 0.0 | 43.0 | 9245 | 1.2578 | 0.8444 |
| 0.0 | 44.0 | 9460 | 1.2543 | 0.8444 |
| 0.0 | 45.0 | 9675 | 1.2707 | 0.8222 |
| 0.0 | 46.0 | 9890 | 1.2807 | 0.8444 |
| 0.0 | 47.0 | 10105 | 1.2891 | 0.8444 |
| 0.0 | 48.0 | 10320 | 1.3057 | 0.8222 |
| 0.0 | 49.0 | 10535 | 1.3045 | 0.8444 |
| 0.0 | 50.0 | 10750 | 1.3015 | 0.8444 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2