smids_3x_deit_tiny_sgd_00001_fold1
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: 1.0996
- Accuracy: 0.4057
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 |
|---|---|---|---|---|
| 1.3881 | 1.0 | 226 | 1.3111 | 0.3489 |
| 1.3128 | 2.0 | 452 | 1.2903 | 0.3523 |
| 1.3013 | 3.0 | 678 | 1.2711 | 0.3589 |
| 1.3217 | 4.0 | 904 | 1.2541 | 0.3656 |
| 1.2917 | 5.0 | 1130 | 1.2386 | 0.3639 |
| 1.3196 | 6.0 | 1356 | 1.2247 | 0.3656 |
| 1.2618 | 7.0 | 1582 | 1.2122 | 0.3673 |
| 1.2868 | 8.0 | 1808 | 1.2013 | 0.3689 |
| 1.2007 | 9.0 | 2034 | 1.1914 | 0.3790 |
| 1.1905 | 10.0 | 2260 | 1.1825 | 0.3856 |
| 1.2678 | 11.0 | 2486 | 1.1746 | 0.3823 |
| 1.1575 | 12.0 | 2712 | 1.1675 | 0.3856 |
| 1.1907 | 13.0 | 2938 | 1.1613 | 0.3840 |
| 1.2093 | 14.0 | 3164 | 1.1556 | 0.3840 |
| 1.2019 | 15.0 | 3390 | 1.1505 | 0.3756 |
| 1.1269 | 16.0 | 3616 | 1.1458 | 0.3756 |
| 1.2046 | 17.0 | 3842 | 1.1416 | 0.3790 |
| 1.1582 | 18.0 | 4068 | 1.1378 | 0.3740 |
| 1.1486 | 19.0 | 4294 | 1.1344 | 0.3806 |
| 1.1865 | 20.0 | 4520 | 1.1312 | 0.3773 |
| 1.1413 | 21.0 | 4746 | 1.1283 | 0.3773 |
| 1.1132 | 22.0 | 4972 | 1.1257 | 0.3856 |
| 1.1589 | 23.0 | 5198 | 1.1233 | 0.3873 |
| 1.1721 | 24.0 | 5424 | 1.1210 | 0.3856 |
| 1.1316 | 25.0 | 5650 | 1.1189 | 0.3856 |
| 1.1482 | 26.0 | 5876 | 1.1170 | 0.3957 |
| 1.126 | 27.0 | 6102 | 1.1153 | 0.4007 |
| 1.0926 | 28.0 | 6328 | 1.1137 | 0.4040 |
| 1.1041 | 29.0 | 6554 | 1.1121 | 0.4023 |
| 1.206 | 30.0 | 6780 | 1.1107 | 0.3973 |
| 1.1379 | 31.0 | 7006 | 1.1094 | 0.3940 |
| 1.1454 | 32.0 | 7232 | 1.1082 | 0.3990 |
| 1.1347 | 33.0 | 7458 | 1.1071 | 0.4040 |
| 1.0924 | 34.0 | 7684 | 1.1061 | 0.4057 |
| 1.0887 | 35.0 | 7910 | 1.1052 | 0.4057 |
| 1.1281 | 36.0 | 8136 | 1.1043 | 0.4057 |
| 1.1197 | 37.0 | 8362 | 1.1035 | 0.4057 |
| 1.0883 | 38.0 | 8588 | 1.1028 | 0.4090 |
| 1.1185 | 39.0 | 8814 | 1.1022 | 0.4090 |
| 1.1206 | 40.0 | 9040 | 1.1017 | 0.4090 |
| 1.1449 | 41.0 | 9266 | 1.1012 | 0.4073 |
| 1.0923 | 42.0 | 9492 | 1.1008 | 0.4057 |
| 1.1262 | 43.0 | 9718 | 1.1004 | 0.4073 |
| 1.1447 | 44.0 | 9944 | 1.1002 | 0.4040 |
| 1.11 | 45.0 | 10170 | 1.0999 | 0.4040 |
| 1.1348 | 46.0 | 10396 | 1.0998 | 0.4057 |
| 1.1134 | 47.0 | 10622 | 1.0997 | 0.4057 |
| 1.119 | 48.0 | 10848 | 1.0996 | 0.4057 |
| 1.1325 | 49.0 | 11074 | 1.0996 | 0.4057 |
| 1.1613 | 50.0 | 11300 | 1.0996 | 0.4057 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.1+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2