hushem_40x_deit_base_rms_001_fold1
This model is a fine-tuned version of facebook/deit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 7.0637
- Accuracy: 0.5333
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 |
|---|---|---|---|---|
| 1.3874 | 1.0 | 215 | 1.5845 | 0.2667 |
| 1.2829 | 2.0 | 430 | 1.2221 | 0.4 |
| 0.7962 | 3.0 | 645 | 2.1065 | 0.4 |
| 0.7528 | 4.0 | 860 | 1.0651 | 0.5556 |
| 0.6029 | 5.0 | 1075 | 1.5642 | 0.4889 |
| 0.6246 | 6.0 | 1290 | 1.7962 | 0.4222 |
| 0.589 | 7.0 | 1505 | 1.4819 | 0.4444 |
| 0.6081 | 8.0 | 1720 | 1.4452 | 0.4222 |
| 0.4808 | 9.0 | 1935 | 1.4389 | 0.4444 |
| 0.4155 | 10.0 | 2150 | 1.7698 | 0.4667 |
| 0.4393 | 11.0 | 2365 | 1.4569 | 0.5778 |
| 0.4007 | 12.0 | 2580 | 2.1115 | 0.4 |
| 0.3758 | 13.0 | 2795 | 1.5230 | 0.5556 |
| 0.3244 | 14.0 | 3010 | 2.2901 | 0.4444 |
| 0.3063 | 15.0 | 3225 | 2.0129 | 0.4889 |
| 0.3072 | 16.0 | 3440 | 2.2969 | 0.5333 |
| 0.2444 | 17.0 | 3655 | 2.5054 | 0.4667 |
| 0.2293 | 18.0 | 3870 | 2.3449 | 0.4889 |
| 0.2391 | 19.0 | 4085 | 2.0401 | 0.6444 |
| 0.1843 | 20.0 | 4300 | 2.7271 | 0.5333 |
| 0.2073 | 21.0 | 4515 | 2.2599 | 0.4889 |
| 0.194 | 22.0 | 4730 | 3.1378 | 0.4444 |
| 0.2943 | 23.0 | 4945 | 2.7236 | 0.5333 |
| 0.2089 | 24.0 | 5160 | 2.5054 | 0.5778 |
| 0.2145 | 25.0 | 5375 | 3.8073 | 0.4667 |
| 0.1232 | 26.0 | 5590 | 3.5697 | 0.4889 |
| 0.1349 | 27.0 | 5805 | 3.5985 | 0.5333 |
| 0.1548 | 28.0 | 6020 | 3.0930 | 0.4889 |
| 0.0655 | 29.0 | 6235 | 4.3232 | 0.4889 |
| 0.1304 | 30.0 | 6450 | 3.6994 | 0.5333 |
| 0.0997 | 31.0 | 6665 | 3.7329 | 0.5333 |
| 0.0825 | 32.0 | 6880 | 3.4793 | 0.5333 |
| 0.154 | 33.0 | 7095 | 5.2562 | 0.4667 |
| 0.1206 | 34.0 | 7310 | 4.5299 | 0.4889 |
| 0.1019 | 35.0 | 7525 | 3.6522 | 0.5111 |
| 0.019 | 36.0 | 7740 | 3.9235 | 0.5333 |
| 0.0485 | 37.0 | 7955 | 4.7342 | 0.5556 |
| 0.0155 | 38.0 | 8170 | 4.4779 | 0.5778 |
| 0.0142 | 39.0 | 8385 | 4.2139 | 0.5556 |
| 0.0256 | 40.0 | 8600 | 5.0724 | 0.5333 |
| 0.0211 | 41.0 | 8815 | 4.8895 | 0.4889 |
| 0.019 | 42.0 | 9030 | 4.8291 | 0.5556 |
| 0.0047 | 43.0 | 9245 | 5.9102 | 0.5333 |
| 0.0027 | 44.0 | 9460 | 5.9480 | 0.5556 |
| 0.0009 | 45.0 | 9675 | 6.2260 | 0.5333 |
| 0.0008 | 46.0 | 9890 | 6.6029 | 0.5556 |
| 0.0001 | 47.0 | 10105 | 6.7925 | 0.5556 |
| 0.0001 | 48.0 | 10320 | 6.7039 | 0.5333 |
| 0.0 | 49.0 | 10535 | 7.0556 | 0.5333 |
| 0.0001 | 50.0 | 10750 | 7.0637 | 0.5333 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2