hushem_1x_deit_tiny_rms_0001_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: 3.4166
- Accuracy: 0.5556
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: 0.0001
- 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 |
|---|---|---|---|---|
| No log | 1.0 | 6 | 2.1314 | 0.2444 |
| 2.0481 | 2.0 | 12 | 1.5573 | 0.2444 |
| 2.0481 | 3.0 | 18 | 1.4598 | 0.2444 |
| 1.5099 | 4.0 | 24 | 1.4194 | 0.2444 |
| 1.4253 | 5.0 | 30 | 1.3528 | 0.2667 |
| 1.4253 | 6.0 | 36 | 1.6348 | 0.2444 |
| 1.3319 | 7.0 | 42 | 1.3901 | 0.4444 |
| 1.3319 | 8.0 | 48 | 1.3151 | 0.2889 |
| 1.2142 | 9.0 | 54 | 1.3395 | 0.3333 |
| 1.1416 | 10.0 | 60 | 1.4176 | 0.3556 |
| 1.1416 | 11.0 | 66 | 1.9072 | 0.2667 |
| 0.9889 | 12.0 | 72 | 1.7446 | 0.3111 |
| 0.9889 | 13.0 | 78 | 1.4748 | 0.3778 |
| 0.8552 | 14.0 | 84 | 1.7450 | 0.3778 |
| 0.6798 | 15.0 | 90 | 1.6042 | 0.4889 |
| 0.6798 | 16.0 | 96 | 1.5863 | 0.4222 |
| 0.563 | 17.0 | 102 | 1.9311 | 0.4 |
| 0.563 | 18.0 | 108 | 1.9509 | 0.4444 |
| 0.3845 | 19.0 | 114 | 2.1256 | 0.4667 |
| 0.2041 | 20.0 | 120 | 2.4131 | 0.4889 |
| 0.2041 | 21.0 | 126 | 2.1029 | 0.4667 |
| 0.1874 | 22.0 | 132 | 2.0412 | 0.5778 |
| 0.1874 | 23.0 | 138 | 2.4952 | 0.4889 |
| 0.0735 | 24.0 | 144 | 2.8992 | 0.4667 |
| 0.0229 | 25.0 | 150 | 2.7495 | 0.5556 |
| 0.0229 | 26.0 | 156 | 3.2879 | 0.4667 |
| 0.0293 | 27.0 | 162 | 3.1526 | 0.5111 |
| 0.0293 | 28.0 | 168 | 3.0123 | 0.5333 |
| 0.0023 | 29.0 | 174 | 3.0812 | 0.5556 |
| 0.0008 | 30.0 | 180 | 3.1384 | 0.5556 |
| 0.0008 | 31.0 | 186 | 3.2017 | 0.5556 |
| 0.0005 | 32.0 | 192 | 3.2443 | 0.5556 |
| 0.0005 | 33.0 | 198 | 3.2806 | 0.5556 |
| 0.0005 | 34.0 | 204 | 3.3167 | 0.5556 |
| 0.0004 | 35.0 | 210 | 3.3393 | 0.5556 |
| 0.0004 | 36.0 | 216 | 3.3662 | 0.5556 |
| 0.0004 | 37.0 | 222 | 3.3843 | 0.5556 |
| 0.0004 | 38.0 | 228 | 3.3970 | 0.5556 |
| 0.0003 | 39.0 | 234 | 3.4053 | 0.5556 |
| 0.0003 | 40.0 | 240 | 3.4123 | 0.5556 |
| 0.0003 | 41.0 | 246 | 3.4159 | 0.5556 |
| 0.0003 | 42.0 | 252 | 3.4166 | 0.5556 |
| 0.0003 | 43.0 | 258 | 3.4166 | 0.5556 |
| 0.0003 | 44.0 | 264 | 3.4166 | 0.5556 |
| 0.0003 | 45.0 | 270 | 3.4166 | 0.5556 |
| 0.0003 | 46.0 | 276 | 3.4166 | 0.5556 |
| 0.0003 | 47.0 | 282 | 3.4166 | 0.5556 |
| 0.0003 | 48.0 | 288 | 3.4166 | 0.5556 |
| 0.0003 | 49.0 | 294 | 3.4166 | 0.5556 |
| 0.0003 | 50.0 | 300 | 3.4166 | 0.5556 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1