smids_5x_deit_tiny_rms_00001_fold5
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: 0.8603
- Accuracy: 0.905
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 |
|---|---|---|---|---|
| 0.2338 | 1.0 | 375 | 0.3930 | 0.8433 |
| 0.1865 | 2.0 | 750 | 0.3259 | 0.8733 |
| 0.1356 | 3.0 | 1125 | 0.2805 | 0.9033 |
| 0.0896 | 4.0 | 1500 | 0.3878 | 0.88 |
| 0.0385 | 5.0 | 1875 | 0.4177 | 0.8883 |
| 0.0314 | 6.0 | 2250 | 0.4802 | 0.8967 |
| 0.0588 | 7.0 | 2625 | 0.6345 | 0.895 |
| 0.0141 | 8.0 | 3000 | 0.7091 | 0.9033 |
| 0.0524 | 9.0 | 3375 | 0.8142 | 0.8817 |
| 0.0425 | 10.0 | 3750 | 0.7582 | 0.8983 |
| 0.0006 | 11.0 | 4125 | 0.7258 | 0.9 |
| 0.0097 | 12.0 | 4500 | 0.7403 | 0.9 |
| 0.0104 | 13.0 | 4875 | 0.9310 | 0.89 |
| 0.0001 | 14.0 | 5250 | 0.7672 | 0.9 |
| 0.0 | 15.0 | 5625 | 0.9240 | 0.8917 |
| 0.0003 | 16.0 | 6000 | 0.8712 | 0.8983 |
| 0.0135 | 17.0 | 6375 | 0.7633 | 0.9033 |
| 0.0335 | 18.0 | 6750 | 1.0118 | 0.8917 |
| 0.0155 | 19.0 | 7125 | 0.8189 | 0.905 |
| 0.0 | 20.0 | 7500 | 0.8004 | 0.8983 |
| 0.0 | 21.0 | 7875 | 1.0772 | 0.88 |
| 0.0255 | 22.0 | 8250 | 0.7694 | 0.91 |
| 0.0019 | 23.0 | 8625 | 0.8682 | 0.8983 |
| 0.0 | 24.0 | 9000 | 0.8775 | 0.8933 |
| 0.0 | 25.0 | 9375 | 0.9259 | 0.9017 |
| 0.0 | 26.0 | 9750 | 0.8433 | 0.895 |
| 0.0119 | 27.0 | 10125 | 0.9223 | 0.8983 |
| 0.0 | 28.0 | 10500 | 0.7870 | 0.91 |
| 0.0 | 29.0 | 10875 | 0.9279 | 0.895 |
| 0.0131 | 30.0 | 11250 | 0.9531 | 0.8933 |
| 0.0 | 31.0 | 11625 | 0.8850 | 0.8967 |
| 0.0 | 32.0 | 12000 | 0.8772 | 0.8983 |
| 0.0 | 33.0 | 12375 | 0.8996 | 0.8917 |
| 0.0 | 34.0 | 12750 | 0.9022 | 0.8983 |
| 0.0 | 35.0 | 13125 | 0.8990 | 0.8933 |
| 0.0 | 36.0 | 13500 | 0.8690 | 0.9033 |
| 0.0 | 37.0 | 13875 | 0.8890 | 0.9 |
| 0.0071 | 38.0 | 14250 | 0.8769 | 0.9017 |
| 0.0 | 39.0 | 14625 | 0.8323 | 0.9067 |
| 0.0 | 40.0 | 15000 | 0.8920 | 0.9033 |
| 0.0 | 41.0 | 15375 | 0.8465 | 0.9083 |
| 0.0 | 42.0 | 15750 | 0.8536 | 0.905 |
| 0.0 | 43.0 | 16125 | 0.8497 | 0.905 |
| 0.0 | 44.0 | 16500 | 0.8492 | 0.905 |
| 0.0 | 45.0 | 16875 | 0.8481 | 0.9067 |
| 0.0 | 46.0 | 17250 | 0.8573 | 0.9067 |
| 0.0029 | 47.0 | 17625 | 0.8575 | 0.9067 |
| 0.0 | 48.0 | 18000 | 0.8605 | 0.905 |
| 0.0 | 49.0 | 18375 | 0.8627 | 0.905 |
| 0.0013 | 50.0 | 18750 | 0.8603 | 0.905 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2