dinov2-base_rice-leaf-disease-augmented-v4_v5_pft
This model is a fine-tuned version of facebook/dinov2-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.2563
- Accuracy: 0.9295
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: 0.0003
- 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 |
|---|---|---|---|---|
| 2.283 | 0.5 | 64 | 1.9783 | 0.2013 |
| 1.4862 | 1.0 | 128 | 1.1465 | 0.6745 |
| 0.9264 | 1.5 | 192 | 0.7663 | 0.7550 |
| 0.6628 | 2.0 | 256 | 0.6171 | 0.8020 |
| 0.5289 | 2.5 | 320 | 0.5400 | 0.8322 |
| 0.4616 | 3.0 | 384 | 0.5054 | 0.8322 |
| 0.4022 | 3.5 | 448 | 0.4499 | 0.8490 |
| 0.3653 | 4.0 | 512 | 0.4136 | 0.8691 |
| 0.3432 | 4.5 | 576 | 0.4172 | 0.8691 |
| 0.3168 | 5.0 | 640 | 0.3760 | 0.8826 |
| 0.3124 | 5.5 | 704 | 0.3938 | 0.8658 |
| 0.2992 | 6.0 | 768 | 0.3811 | 0.8859 |
| 0.3062 | 6.5 | 832 | 0.3866 | 0.8826 |
| 0.2951 | 7.0 | 896 | 0.3926 | 0.8725 |
| 0.2967 | 7.5 | 960 | 0.3657 | 0.8758 |
| 0.2635 | 8.0 | 1024 | 0.3501 | 0.8859 |
| 0.2474 | 8.5 | 1088 | 0.3510 | 0.8993 |
| 0.2452 | 9.0 | 1152 | 0.3402 | 0.8960 |
| 0.2185 | 9.5 | 1216 | 0.3353 | 0.8926 |
| 0.2283 | 10.0 | 1280 | 0.3332 | 0.9060 |
| 0.2138 | 10.5 | 1344 | 0.3235 | 0.9027 |
| 0.2143 | 11.0 | 1408 | 0.3267 | 0.8993 |
| 0.2156 | 11.5 | 1472 | 0.3379 | 0.8926 |
| 0.2125 | 12.0 | 1536 | 0.3256 | 0.8826 |
| 0.1998 | 12.5 | 1600 | 0.3135 | 0.8926 |
| 0.1961 | 13.0 | 1664 | 0.3031 | 0.8993 |
| 0.1818 | 13.5 | 1728 | 0.2855 | 0.9228 |
| 0.1788 | 14.0 | 1792 | 0.2977 | 0.9094 |
| 0.1709 | 14.5 | 1856 | 0.3101 | 0.9060 |
| 0.1652 | 15.0 | 1920 | 0.2969 | 0.9128 |
| 0.1578 | 15.5 | 1984 | 0.2916 | 0.9161 |
| 0.1661 | 16.0 | 2048 | 0.2904 | 0.9195 |
| 0.1727 | 16.5 | 2112 | 0.2854 | 0.9128 |
| 0.1617 | 17.0 | 2176 | 0.2802 | 0.9128 |
| 0.1519 | 17.5 | 2240 | 0.2954 | 0.8993 |
| 0.1491 | 18.0 | 2304 | 0.2812 | 0.9094 |
| 0.1366 | 18.5 | 2368 | 0.2877 | 0.9060 |
| 0.1424 | 19.0 | 2432 | 0.2791 | 0.9195 |
| 0.1297 | 19.5 | 2496 | 0.2810 | 0.9161 |
| 0.1325 | 20.0 | 2560 | 0.2787 | 0.9161 |
| 0.1309 | 20.5 | 2624 | 0.2764 | 0.9195 |
| 0.1298 | 21.0 | 2688 | 0.2715 | 0.9161 |
| 0.1334 | 21.5 | 2752 | 0.2768 | 0.9094 |
| 0.1268 | 22.0 | 2816 | 0.2712 | 0.9128 |
| 0.1169 | 22.5 | 2880 | 0.2739 | 0.9195 |
| 0.1203 | 23.0 | 2944 | 0.2607 | 0.9262 |
| 0.1091 | 23.5 | 3008 | 0.2703 | 0.9161 |
| 0.1123 | 24.0 | 3072 | 0.2611 | 0.9262 |
| 0.1056 | 24.5 | 3136 | 0.2606 | 0.9262 |
| 0.1061 | 25.0 | 3200 | 0.2630 | 0.9262 |
| 0.112 | 25.5 | 3264 | 0.2899 | 0.9094 |
| 0.1076 | 26.0 | 3328 | 0.2749 | 0.9195 |
| 0.105 | 26.5 | 3392 | 0.2570 | 0.9329 |
| 0.1048 | 27.0 | 3456 | 0.2654 | 0.9295 |
| 0.098 | 27.5 | 3520 | 0.2593 | 0.9262 |
| 0.0951 | 28.0 | 3584 | 0.2591 | 0.9295 |
| 0.0915 | 28.5 | 3648 | 0.2553 | 0.9262 |
| 0.0907 | 29.0 | 3712 | 0.2568 | 0.9295 |
| 0.0885 | 29.5 | 3776 | 0.2582 | 0.9295 |
| 0.087 | 30.0 | 3840 | 0.2563 | 0.9295 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1+cu124
- Datasets 3.3.2
- Tokenizers 0.21.1