smids_10x_deit_tiny_sgd_001_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: 0.3073
- Accuracy: 0.8902
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: 0.001
- 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 |
|---|---|---|---|---|
| 0.5172 | 1.0 | 750 | 0.5682 | 0.7687 |
| 0.3347 | 2.0 | 1500 | 0.4469 | 0.8153 |
| 0.3177 | 3.0 | 2250 | 0.3953 | 0.8469 |
| 0.3776 | 4.0 | 3000 | 0.3688 | 0.8502 |
| 0.2886 | 5.0 | 3750 | 0.3556 | 0.8519 |
| 0.2396 | 6.0 | 4500 | 0.3328 | 0.8502 |
| 0.2545 | 7.0 | 5250 | 0.3237 | 0.8586 |
| 0.2435 | 8.0 | 6000 | 0.3188 | 0.8569 |
| 0.2366 | 9.0 | 6750 | 0.3065 | 0.8686 |
| 0.232 | 10.0 | 7500 | 0.3041 | 0.8652 |
| 0.2399 | 11.0 | 8250 | 0.2971 | 0.8785 |
| 0.2717 | 12.0 | 9000 | 0.2941 | 0.8769 |
| 0.2579 | 13.0 | 9750 | 0.2863 | 0.8852 |
| 0.1661 | 14.0 | 10500 | 0.2895 | 0.8802 |
| 0.1655 | 15.0 | 11250 | 0.2865 | 0.8785 |
| 0.1921 | 16.0 | 12000 | 0.2897 | 0.8802 |
| 0.1525 | 17.0 | 12750 | 0.2854 | 0.8835 |
| 0.1653 | 18.0 | 13500 | 0.2861 | 0.8819 |
| 0.1849 | 19.0 | 14250 | 0.2939 | 0.8702 |
| 0.1923 | 20.0 | 15000 | 0.2850 | 0.8835 |
| 0.1967 | 21.0 | 15750 | 0.2874 | 0.8802 |
| 0.1373 | 22.0 | 16500 | 0.2916 | 0.8802 |
| 0.1229 | 23.0 | 17250 | 0.2891 | 0.8869 |
| 0.1054 | 24.0 | 18000 | 0.2911 | 0.8802 |
| 0.1456 | 25.0 | 18750 | 0.2869 | 0.8869 |
| 0.2052 | 26.0 | 19500 | 0.2987 | 0.8835 |
| 0.1723 | 27.0 | 20250 | 0.2918 | 0.8835 |
| 0.1277 | 28.0 | 21000 | 0.2937 | 0.8902 |
| 0.1569 | 29.0 | 21750 | 0.2956 | 0.8902 |
| 0.1514 | 30.0 | 22500 | 0.2954 | 0.8885 |
| 0.1603 | 31.0 | 23250 | 0.2954 | 0.8902 |
| 0.1428 | 32.0 | 24000 | 0.2940 | 0.8918 |
| 0.1564 | 33.0 | 24750 | 0.3002 | 0.8835 |
| 0.1386 | 34.0 | 25500 | 0.3023 | 0.8852 |
| 0.1564 | 35.0 | 26250 | 0.2982 | 0.8869 |
| 0.183 | 36.0 | 27000 | 0.3004 | 0.8885 |
| 0.1456 | 37.0 | 27750 | 0.3058 | 0.8869 |
| 0.1394 | 38.0 | 28500 | 0.3047 | 0.8869 |
| 0.121 | 39.0 | 29250 | 0.3021 | 0.8902 |
| 0.1192 | 40.0 | 30000 | 0.3035 | 0.8852 |
| 0.1706 | 41.0 | 30750 | 0.3048 | 0.8918 |
| 0.1421 | 42.0 | 31500 | 0.3036 | 0.8885 |
| 0.1223 | 43.0 | 32250 | 0.3066 | 0.8852 |
| 0.1116 | 44.0 | 33000 | 0.3060 | 0.8885 |
| 0.1122 | 45.0 | 33750 | 0.3075 | 0.8885 |
| 0.1411 | 46.0 | 34500 | 0.3066 | 0.8885 |
| 0.1644 | 47.0 | 35250 | 0.3072 | 0.8885 |
| 0.0953 | 48.0 | 36000 | 0.3070 | 0.8902 |
| 0.1109 | 49.0 | 36750 | 0.3072 | 0.8902 |
| 0.1061 | 50.0 | 37500 | 0.3073 | 0.8902 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2