swin-tiny-patch4-window7-224_rice-leaf-disease-augmented-v4_v5_fft
This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3980
- Accuracy: 0.9362
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 |
|---|---|---|---|---|
| 1.9379 | 0.5 | 64 | 1.5473 | 0.5336 |
| 1.0179 | 1.0 | 128 | 0.5529 | 0.8255 |
| 0.4389 | 1.5 | 192 | 0.3367 | 0.8792 |
| 0.2443 | 2.0 | 256 | 0.2354 | 0.9228 |
| 0.1151 | 2.5 | 320 | 0.2422 | 0.9329 |
| 0.0648 | 3.0 | 384 | 0.2351 | 0.9228 |
| 0.0325 | 3.5 | 448 | 0.3316 | 0.9094 |
| 0.0235 | 4.0 | 512 | 0.2389 | 0.9329 |
| 0.0148 | 4.5 | 576 | 0.2317 | 0.9362 |
| 0.0081 | 5.0 | 640 | 0.2005 | 0.9362 |
| 0.0054 | 5.5 | 704 | 0.2198 | 0.9396 |
| 0.0043 | 6.0 | 768 | 0.2171 | 0.9430 |
| 0.0038 | 6.5 | 832 | 0.2200 | 0.9396 |
| 0.003 | 7.0 | 896 | 0.2383 | 0.9362 |
| 0.0183 | 7.5 | 960 | 0.2142 | 0.9262 |
| 0.0126 | 8.0 | 1024 | 0.2931 | 0.9295 |
| 0.0093 | 8.5 | 1088 | 0.2718 | 0.9396 |
| 0.0035 | 9.0 | 1152 | 0.3975 | 0.9195 |
| 0.0019 | 9.5 | 1216 | 0.2839 | 0.9295 |
| 0.0015 | 10.0 | 1280 | 0.2764 | 0.9329 |
| 0.001 | 10.5 | 1344 | 0.3293 | 0.9195 |
| 0.0013 | 11.0 | 1408 | 0.3007 | 0.9295 |
| 0.0011 | 11.5 | 1472 | 0.3193 | 0.9362 |
| 0.0122 | 12.0 | 1536 | 0.2385 | 0.9463 |
| 0.0269 | 12.5 | 1600 | 0.3694 | 0.9295 |
| 0.0092 | 13.0 | 1664 | 0.3212 | 0.9295 |
| 0.0032 | 13.5 | 1728 | 0.2687 | 0.9362 |
| 0.0023 | 14.0 | 1792 | 0.4258 | 0.9128 |
| 0.0015 | 14.5 | 1856 | 0.3478 | 0.9329 |
| 0.0007 | 15.0 | 1920 | 0.3445 | 0.9329 |
| 0.0005 | 15.5 | 1984 | 0.3371 | 0.9329 |
| 0.001 | 16.0 | 2048 | 0.3407 | 0.9329 |
| 0.0079 | 16.5 | 2112 | 0.4630 | 0.9195 |
| 0.0098 | 17.0 | 2176 | 0.3260 | 0.9362 |
| 0.016 | 17.5 | 2240 | 0.5011 | 0.9195 |
| 0.0085 | 18.0 | 2304 | 0.3927 | 0.9362 |
| 0.0057 | 18.5 | 2368 | 0.3910 | 0.9295 |
| 0.0016 | 19.0 | 2432 | 0.3433 | 0.9262 |
| 0.001 | 19.5 | 2496 | 0.4334 | 0.9228 |
| 0.0007 | 20.0 | 2560 | 0.4074 | 0.9228 |
| 0.0005 | 20.5 | 2624 | 0.4064 | 0.9228 |
| 0.0018 | 21.0 | 2688 | 0.5353 | 0.9228 |
| 0.0097 | 21.5 | 2752 | 0.4666 | 0.9128 |
| 0.0127 | 22.0 | 2816 | 0.3886 | 0.9228 |
| 0.0036 | 22.5 | 2880 | 0.2585 | 0.9430 |
| 0.0036 | 23.0 | 2944 | 0.4433 | 0.9329 |
| 0.001 | 23.5 | 3008 | 0.4032 | 0.9329 |
| 0.0006 | 24.0 | 3072 | 0.4428 | 0.9295 |
| 0.0012 | 24.5 | 3136 | 0.4001 | 0.9228 |
| 0.0007 | 25.0 | 3200 | 0.3978 | 0.9262 |
| 0.0006 | 25.5 | 3264 | 0.5251 | 0.9161 |
| 0.0054 | 26.0 | 3328 | 0.5131 | 0.9228 |
| 0.005 | 26.5 | 3392 | 0.4757 | 0.9027 |
| 0.004 | 27.0 | 3456 | 0.4730 | 0.9195 |
| 0.0025 | 27.5 | 3520 | 0.4133 | 0.9228 |
| 0.0012 | 28.0 | 3584 | 0.4215 | 0.9228 |
| 0.0007 | 28.5 | 3648 | 0.4058 | 0.9329 |
| 0.0004 | 29.0 | 3712 | 0.4024 | 0.9262 |
| 0.0003 | 29.5 | 3776 | 0.3974 | 0.9362 |
| 0.0002 | 30.0 | 3840 | 0.3980 | 0.9362 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1+cu124
- Datasets 3.3.2
- Tokenizers 0.21.1