hushem_1x_deit_tiny_sgd_lr001_fold5
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: 1.3032
- Accuracy: 0.3902
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: 0.001
- 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 | 1.6123 | 0.1220 |
| 1.5785 | 2.0 | 12 | 1.5813 | 0.1463 |
| 1.5785 | 3.0 | 18 | 1.5542 | 0.1463 |
| 1.5395 | 4.0 | 24 | 1.5297 | 0.1463 |
| 1.4749 | 5.0 | 30 | 1.5083 | 0.1951 |
| 1.4749 | 6.0 | 36 | 1.4884 | 0.1951 |
| 1.4296 | 7.0 | 42 | 1.4718 | 0.1707 |
| 1.4296 | 8.0 | 48 | 1.4578 | 0.1463 |
| 1.4059 | 9.0 | 54 | 1.4447 | 0.1463 |
| 1.3876 | 10.0 | 60 | 1.4316 | 0.2195 |
| 1.3876 | 11.0 | 66 | 1.4209 | 0.2195 |
| 1.3523 | 12.0 | 72 | 1.4102 | 0.2195 |
| 1.3523 | 13.0 | 78 | 1.4009 | 0.2439 |
| 1.3412 | 14.0 | 84 | 1.3926 | 0.2439 |
| 1.3216 | 15.0 | 90 | 1.3847 | 0.2927 |
| 1.3216 | 16.0 | 96 | 1.3782 | 0.3171 |
| 1.2923 | 17.0 | 102 | 1.3713 | 0.3415 |
| 1.2923 | 18.0 | 108 | 1.3652 | 0.3415 |
| 1.305 | 19.0 | 114 | 1.3592 | 0.3415 |
| 1.2722 | 20.0 | 120 | 1.3536 | 0.3415 |
| 1.2722 | 21.0 | 126 | 1.3490 | 0.3659 |
| 1.2479 | 22.0 | 132 | 1.3441 | 0.3659 |
| 1.2479 | 23.0 | 138 | 1.3399 | 0.3659 |
| 1.2818 | 24.0 | 144 | 1.3360 | 0.3659 |
| 1.2363 | 25.0 | 150 | 1.3318 | 0.3659 |
| 1.2363 | 26.0 | 156 | 1.3281 | 0.3659 |
| 1.2375 | 27.0 | 162 | 1.3249 | 0.3659 |
| 1.2375 | 28.0 | 168 | 1.3220 | 0.3659 |
| 1.2164 | 29.0 | 174 | 1.3194 | 0.3659 |
| 1.2359 | 30.0 | 180 | 1.3171 | 0.3902 |
| 1.2359 | 31.0 | 186 | 1.3148 | 0.3902 |
| 1.2121 | 32.0 | 192 | 1.3127 | 0.3902 |
| 1.2121 | 33.0 | 198 | 1.3110 | 0.3902 |
| 1.2131 | 34.0 | 204 | 1.3092 | 0.3902 |
| 1.1973 | 35.0 | 210 | 1.3077 | 0.3902 |
| 1.1973 | 36.0 | 216 | 1.3064 | 0.3902 |
| 1.1836 | 37.0 | 222 | 1.3054 | 0.3902 |
| 1.1836 | 38.0 | 228 | 1.3046 | 0.3902 |
| 1.2087 | 39.0 | 234 | 1.3039 | 0.3902 |
| 1.2019 | 40.0 | 240 | 1.3035 | 0.3902 |
| 1.2019 | 41.0 | 246 | 1.3033 | 0.3902 |
| 1.2033 | 42.0 | 252 | 1.3032 | 0.3902 |
| 1.2033 | 43.0 | 258 | 1.3032 | 0.3902 |
| 1.1754 | 44.0 | 264 | 1.3032 | 0.3902 |
| 1.1907 | 45.0 | 270 | 1.3032 | 0.3902 |
| 1.1907 | 46.0 | 276 | 1.3032 | 0.3902 |
| 1.2082 | 47.0 | 282 | 1.3032 | 0.3902 |
| 1.2082 | 48.0 | 288 | 1.3032 | 0.3902 |
| 1.1699 | 49.0 | 294 | 1.3032 | 0.3902 |
| 1.2038 | 50.0 | 300 | 1.3032 | 0.3902 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.1.0+cu118
- Datasets 2.14.6
- Tokenizers 0.14.1