smids_5x_deit_small_rms_0001_fold2
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: 1.0513
- Accuracy: 0.8902
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 |
|---|---|---|---|---|
| 0.2658 | 1.0 | 375 | 0.3126 | 0.8819 |
| 0.1908 | 2.0 | 750 | 0.3283 | 0.8918 |
| 0.0825 | 3.0 | 1125 | 0.4361 | 0.8702 |
| 0.0963 | 4.0 | 1500 | 0.5286 | 0.8619 |
| 0.0865 | 5.0 | 1875 | 0.5193 | 0.8885 |
| 0.0848 | 6.0 | 2250 | 0.5531 | 0.8835 |
| 0.072 | 7.0 | 2625 | 0.6299 | 0.8669 |
| 0.0017 | 8.0 | 3000 | 0.6200 | 0.8819 |
| 0.0259 | 9.0 | 3375 | 0.6078 | 0.9002 |
| 0.0609 | 10.0 | 3750 | 0.5800 | 0.8918 |
| 0.045 | 11.0 | 4125 | 0.6268 | 0.8802 |
| 0.068 | 12.0 | 4500 | 0.5720 | 0.8869 |
| 0.0342 | 13.0 | 4875 | 0.7774 | 0.8652 |
| 0.007 | 14.0 | 5250 | 0.7661 | 0.8769 |
| 0.0136 | 15.0 | 5625 | 0.7670 | 0.8885 |
| 0.025 | 16.0 | 6000 | 0.9672 | 0.8752 |
| 0.0117 | 17.0 | 6375 | 0.6723 | 0.9018 |
| 0.01 | 18.0 | 6750 | 0.7201 | 0.8835 |
| 0.1057 | 19.0 | 7125 | 0.7988 | 0.8686 |
| 0.0079 | 20.0 | 7500 | 0.8529 | 0.8735 |
| 0.0114 | 21.0 | 7875 | 0.9574 | 0.8835 |
| 0.0141 | 22.0 | 8250 | 0.8344 | 0.8819 |
| 0.0006 | 23.0 | 8625 | 0.9308 | 0.8769 |
| 0.0008 | 24.0 | 9000 | 0.8418 | 0.8752 |
| 0.0001 | 25.0 | 9375 | 0.7076 | 0.8869 |
| 0.0123 | 26.0 | 9750 | 0.9006 | 0.8686 |
| 0.0003 | 27.0 | 10125 | 0.9386 | 0.8702 |
| 0.0081 | 28.0 | 10500 | 1.0332 | 0.8735 |
| 0.0 | 29.0 | 10875 | 0.9316 | 0.8752 |
| 0.0272 | 30.0 | 11250 | 0.9157 | 0.8835 |
| 0.0042 | 31.0 | 11625 | 0.9011 | 0.8752 |
| 0.0038 | 32.0 | 12000 | 0.9259 | 0.8769 |
| 0.0044 | 33.0 | 12375 | 0.9290 | 0.8869 |
| 0.0 | 34.0 | 12750 | 0.9423 | 0.8852 |
| 0.0 | 35.0 | 13125 | 0.8933 | 0.8902 |
| 0.0 | 36.0 | 13500 | 0.8976 | 0.8885 |
| 0.0 | 37.0 | 13875 | 0.8889 | 0.8835 |
| 0.0004 | 38.0 | 14250 | 1.0859 | 0.8802 |
| 0.0 | 39.0 | 14625 | 1.0992 | 0.8869 |
| 0.0031 | 40.0 | 15000 | 1.0003 | 0.8952 |
| 0.0 | 41.0 | 15375 | 1.0009 | 0.8985 |
| 0.0027 | 42.0 | 15750 | 1.0542 | 0.8885 |
| 0.0026 | 43.0 | 16125 | 1.0230 | 0.8852 |
| 0.0026 | 44.0 | 16500 | 1.0414 | 0.8885 |
| 0.0027 | 45.0 | 16875 | 1.0121 | 0.8885 |
| 0.0 | 46.0 | 17250 | 1.0455 | 0.8885 |
| 0.006 | 47.0 | 17625 | 1.0427 | 0.8918 |
| 0.0 | 48.0 | 18000 | 1.0481 | 0.8918 |
| 0.0024 | 49.0 | 18375 | 1.0499 | 0.8918 |
| 0.0022 | 50.0 | 18750 | 1.0513 | 0.8902 |
Framework versions
- Transformers 4.32.1
- Pytorch 2.1.0+cu121
- Datasets 2.12.0
- Tokenizers 0.13.2