hushem_5x_deit_small_adamax_00001_fold4
This model is a fine-tuned version of facebook/deit-small-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.5172
- Accuracy: 0.8333
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 |
|---|---|---|---|---|
| 1.2786 | 1.0 | 28 | 1.2504 | 0.4286 |
| 1.0062 | 2.0 | 56 | 1.1218 | 0.4524 |
| 0.7412 | 3.0 | 84 | 0.9795 | 0.4762 |
| 0.5492 | 4.0 | 112 | 0.8870 | 0.6190 |
| 0.3717 | 5.0 | 140 | 0.7811 | 0.6667 |
| 0.3128 | 6.0 | 168 | 0.7165 | 0.7619 |
| 0.2014 | 7.0 | 196 | 0.6520 | 0.7143 |
| 0.169 | 8.0 | 224 | 0.6139 | 0.7143 |
| 0.0981 | 9.0 | 252 | 0.5693 | 0.7619 |
| 0.0593 | 10.0 | 280 | 0.5620 | 0.7381 |
| 0.0447 | 11.0 | 308 | 0.5082 | 0.8095 |
| 0.0256 | 12.0 | 336 | 0.4626 | 0.8333 |
| 0.0121 | 13.0 | 364 | 0.5351 | 0.7857 |
| 0.0073 | 14.0 | 392 | 0.4890 | 0.8333 |
| 0.0046 | 15.0 | 420 | 0.4933 | 0.8095 |
| 0.004 | 16.0 | 448 | 0.4823 | 0.8095 |
| 0.0033 | 17.0 | 476 | 0.4875 | 0.8095 |
| 0.0029 | 18.0 | 504 | 0.4984 | 0.8095 |
| 0.0023 | 19.0 | 532 | 0.4992 | 0.8095 |
| 0.0021 | 20.0 | 560 | 0.5018 | 0.8095 |
| 0.0018 | 21.0 | 588 | 0.4990 | 0.8095 |
| 0.0017 | 22.0 | 616 | 0.5045 | 0.8095 |
| 0.0014 | 23.0 | 644 | 0.5039 | 0.8095 |
| 0.0014 | 24.0 | 672 | 0.5120 | 0.8095 |
| 0.0013 | 25.0 | 700 | 0.5026 | 0.8095 |
| 0.0011 | 26.0 | 728 | 0.5060 | 0.8095 |
| 0.0012 | 27.0 | 756 | 0.5096 | 0.8095 |
| 0.0011 | 28.0 | 784 | 0.5088 | 0.8095 |
| 0.001 | 29.0 | 812 | 0.5017 | 0.8095 |
| 0.0009 | 30.0 | 840 | 0.5154 | 0.8095 |
| 0.001 | 31.0 | 868 | 0.5070 | 0.8095 |
| 0.0009 | 32.0 | 896 | 0.5093 | 0.8095 |
| 0.0009 | 33.0 | 924 | 0.5133 | 0.8095 |
| 0.0008 | 34.0 | 952 | 0.5115 | 0.8095 |
| 0.0008 | 35.0 | 980 | 0.5134 | 0.8095 |
| 0.0008 | 36.0 | 1008 | 0.5048 | 0.8333 |
| 0.0007 | 37.0 | 1036 | 0.5114 | 0.8095 |
| 0.0007 | 38.0 | 1064 | 0.5110 | 0.8333 |
| 0.0007 | 39.0 | 1092 | 0.5114 | 0.8333 |
| 0.0007 | 40.0 | 1120 | 0.5148 | 0.8333 |
| 0.0007 | 41.0 | 1148 | 0.5122 | 0.8333 |
| 0.0007 | 42.0 | 1176 | 0.5146 | 0.8333 |
| 0.0007 | 43.0 | 1204 | 0.5155 | 0.8333 |
| 0.0007 | 44.0 | 1232 | 0.5189 | 0.8333 |
| 0.0006 | 45.0 | 1260 | 0.5166 | 0.8333 |
| 0.0006 | 46.0 | 1288 | 0.5169 | 0.8333 |
| 0.0007 | 47.0 | 1316 | 0.5173 | 0.8333 |
| 0.0006 | 48.0 | 1344 | 0.5172 | 0.8333 |
| 0.0007 | 49.0 | 1372 | 0.5172 | 0.8333 |
| 0.0007 | 50.0 | 1400 | 0.5172 | 0.8333 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.15.0