deit-small-patch16-224_rice-leaf-disease-augmented-v4_v5_fft
This model is a fine-tuned version of facebook/deit-small-patch16-224 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3237
- Accuracy: 0.9430
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: 5e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine_with_restarts
- lr_scheduler_warmup_steps: 256
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.9877 | 0.5 | 64 | 1.6839 | 0.4530 |
| 1.1728 | 1.0 | 128 | 0.7806 | 0.7517 |
| 0.4995 | 1.5 | 192 | 0.4855 | 0.8389 |
| 0.2433 | 2.0 | 256 | 0.2821 | 0.9161 |
| 0.0745 | 2.5 | 320 | 0.3466 | 0.9060 |
| 0.0506 | 3.0 | 384 | 0.3206 | 0.9195 |
| 0.0163 | 3.5 | 448 | 0.2656 | 0.9195 |
| 0.0061 | 4.0 | 512 | 0.2853 | 0.9295 |
| 0.002 | 4.5 | 576 | 0.2004 | 0.9430 |
| 0.0009 | 5.0 | 640 | 0.2256 | 0.9396 |
| 0.0006 | 5.5 | 704 | 0.2412 | 0.9362 |
| 0.0006 | 6.0 | 768 | 0.2381 | 0.9362 |
| 0.0005 | 6.5 | 832 | 0.2384 | 0.9362 |
| 0.0005 | 7.0 | 896 | 0.2366 | 0.9430 |
| 0.0004 | 7.5 | 960 | 0.2608 | 0.9329 |
| 0.0003 | 8.0 | 1024 | 0.2530 | 0.9430 |
| 0.0002 | 8.5 | 1088 | 0.2587 | 0.9396 |
| 0.0002 | 9.0 | 1152 | 0.2602 | 0.9430 |
| 0.0002 | 9.5 | 1216 | 0.2671 | 0.9396 |
| 0.0002 | 10.0 | 1280 | 0.2640 | 0.9396 |
| 0.0002 | 10.5 | 1344 | 0.2637 | 0.9396 |
| 0.0002 | 11.0 | 1408 | 0.2644 | 0.9396 |
| 0.0002 | 11.5 | 1472 | 0.2677 | 0.9396 |
| 0.0001 | 12.0 | 1536 | 0.2767 | 0.9362 |
| 0.0001 | 12.5 | 1600 | 0.2761 | 0.9396 |
| 0.0001 | 13.0 | 1664 | 0.2833 | 0.9396 |
| 0.0001 | 13.5 | 1728 | 0.2810 | 0.9396 |
| 0.0001 | 14.0 | 1792 | 0.2839 | 0.9396 |
| 0.0001 | 14.5 | 1856 | 0.2860 | 0.9396 |
| 0.0001 | 15.0 | 1920 | 0.2846 | 0.9396 |
| 0.0001 | 15.5 | 1984 | 0.2850 | 0.9396 |
| 0.0001 | 16.0 | 2048 | 0.2850 | 0.9396 |
| 0.0001 | 16.5 | 2112 | 0.2854 | 0.9396 |
| 0.0001 | 17.0 | 2176 | 0.2950 | 0.9396 |
| 0.0001 | 17.5 | 2240 | 0.2976 | 0.9396 |
| 0.0001 | 18.0 | 2304 | 0.2972 | 0.9396 |
| 0.0001 | 18.5 | 2368 | 0.2984 | 0.9396 |
| 0.0001 | 19.0 | 2432 | 0.2980 | 0.9396 |
| 0.0001 | 19.5 | 2496 | 0.3003 | 0.9396 |
| 0.0001 | 20.0 | 2560 | 0.3002 | 0.9396 |
| 0.0001 | 20.5 | 2624 | 0.3001 | 0.9396 |
| 0.0001 | 21.0 | 2688 | 0.3030 | 0.9396 |
| 0.0 | 21.5 | 2752 | 0.3009 | 0.9396 |
| 0.0 | 22.0 | 2816 | 0.3065 | 0.9396 |
| 0.0 | 22.5 | 2880 | 0.3093 | 0.9396 |
| 0.0 | 23.0 | 2944 | 0.3082 | 0.9396 |
| 0.0 | 23.5 | 3008 | 0.3105 | 0.9396 |
| 0.0 | 24.0 | 3072 | 0.3112 | 0.9396 |
| 0.0 | 24.5 | 3136 | 0.3115 | 0.9396 |
| 0.0 | 25.0 | 3200 | 0.3117 | 0.9396 |
| 0.0 | 25.5 | 3264 | 0.3091 | 0.9396 |
| 0.0 | 26.0 | 3328 | 0.3159 | 0.9430 |
| 0.0 | 26.5 | 3392 | 0.3166 | 0.9430 |
| 0.0 | 27.0 | 3456 | 0.3231 | 0.9396 |
| 0.0 | 27.5 | 3520 | 0.3229 | 0.9396 |
| 0.0 | 28.0 | 3584 | 0.3223 | 0.9430 |
| 0.0 | 28.5 | 3648 | 0.3246 | 0.9396 |
| 0.0 | 29.0 | 3712 | 0.3229 | 0.9430 |
| 0.0 | 29.5 | 3776 | 0.3234 | 0.9430 |
| 0.0 | 30.0 | 3840 | 0.3237 | 0.9430 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1+cu124
- Datasets 3.3.2
- Tokenizers 0.21.1