smids_10x_deit_small_sgd_0001_fold2
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: 0.4097
- Accuracy: 0.8270
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: 0.0001
- 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 |
|---|---|---|---|---|
| 0.9968 | 1.0 | 750 | 1.0169 | 0.4659 |
| 0.9174 | 2.0 | 1500 | 0.9543 | 0.5308 |
| 0.8121 | 3.0 | 2250 | 0.8838 | 0.6273 |
| 0.7871 | 4.0 | 3000 | 0.8228 | 0.6522 |
| 0.691 | 5.0 | 3750 | 0.7665 | 0.6922 |
| 0.6733 | 6.0 | 4500 | 0.7184 | 0.7271 |
| 0.611 | 7.0 | 5250 | 0.6739 | 0.7488 |
| 0.5495 | 8.0 | 6000 | 0.6348 | 0.7537 |
| 0.5871 | 9.0 | 6750 | 0.6046 | 0.7587 |
| 0.5362 | 10.0 | 7500 | 0.5781 | 0.7754 |
| 0.5478 | 11.0 | 8250 | 0.5567 | 0.7754 |
| 0.5521 | 12.0 | 9000 | 0.5409 | 0.7804 |
| 0.475 | 13.0 | 9750 | 0.5265 | 0.7787 |
| 0.4124 | 14.0 | 10500 | 0.5147 | 0.7887 |
| 0.4689 | 15.0 | 11250 | 0.5048 | 0.7870 |
| 0.4042 | 16.0 | 12000 | 0.4956 | 0.7903 |
| 0.3787 | 17.0 | 12750 | 0.4873 | 0.7937 |
| 0.4203 | 18.0 | 13500 | 0.4799 | 0.7937 |
| 0.4173 | 19.0 | 14250 | 0.4729 | 0.7987 |
| 0.4444 | 20.0 | 15000 | 0.4676 | 0.8020 |
| 0.4225 | 21.0 | 15750 | 0.4619 | 0.8020 |
| 0.3886 | 22.0 | 16500 | 0.4572 | 0.8070 |
| 0.3882 | 23.0 | 17250 | 0.4523 | 0.8120 |
| 0.3793 | 24.0 | 18000 | 0.4484 | 0.8103 |
| 0.4027 | 25.0 | 18750 | 0.4443 | 0.8136 |
| 0.4864 | 26.0 | 19500 | 0.4411 | 0.8136 |
| 0.4229 | 27.0 | 20250 | 0.4378 | 0.8153 |
| 0.4258 | 28.0 | 21000 | 0.4349 | 0.8153 |
| 0.3905 | 29.0 | 21750 | 0.4322 | 0.8170 |
| 0.4099 | 30.0 | 22500 | 0.4297 | 0.8170 |
| 0.3721 | 31.0 | 23250 | 0.4276 | 0.8186 |
| 0.4104 | 32.0 | 24000 | 0.4255 | 0.8203 |
| 0.3815 | 33.0 | 24750 | 0.4237 | 0.8220 |
| 0.3966 | 34.0 | 25500 | 0.4218 | 0.8220 |
| 0.4057 | 35.0 | 26250 | 0.4202 | 0.8220 |
| 0.4004 | 36.0 | 27000 | 0.4187 | 0.8220 |
| 0.3921 | 37.0 | 27750 | 0.4174 | 0.8220 |
| 0.4046 | 38.0 | 28500 | 0.4161 | 0.8220 |
| 0.3819 | 39.0 | 29250 | 0.4149 | 0.8220 |
| 0.4626 | 40.0 | 30000 | 0.4139 | 0.8236 |
| 0.4062 | 41.0 | 30750 | 0.4130 | 0.8236 |
| 0.3793 | 42.0 | 31500 | 0.4123 | 0.8253 |
| 0.3246 | 43.0 | 32250 | 0.4116 | 0.8253 |
| 0.3382 | 44.0 | 33000 | 0.4110 | 0.8270 |
| 0.3636 | 45.0 | 33750 | 0.4106 | 0.8270 |
| 0.4008 | 46.0 | 34500 | 0.4102 | 0.8270 |
| 0.3708 | 47.0 | 35250 | 0.4099 | 0.8270 |
| 0.3436 | 48.0 | 36000 | 0.4098 | 0.8270 |
| 0.3738 | 49.0 | 36750 | 0.4097 | 0.8270 |
| 0.373 | 50.0 | 37500 | 0.4097 | 0.8270 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2