swin-tiny-patch4-window7-224-med-device-classification
This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.2927
- Accuracy: 0.75
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: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 30
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| No log | 1.0 | 1 | 1.6093 | 0.5 |
| No log | 2.0 | 2 | 1.6063 | 0.5 |
| No log | 3.0 | 3 | 1.6063 | 0.5 |
| No log | 4.0 | 4 | 1.6100 | 0.25 |
| No log | 5.0 | 5 | 1.6100 | 0.25 |
| No log | 6.0 | 6 | 1.6170 | 0.25 |
| No log | 7.0 | 7 | 1.6170 | 0.25 |
| No log | 8.0 | 8 | 1.5910 | 0.25 |
| No log | 9.0 | 9 | 1.5910 | 0.25 |
| 0.7949 | 10.0 | 10 | 1.5705 | 0.25 |
| 0.7949 | 11.0 | 11 | 1.5705 | 0.25 |
| 0.7949 | 12.0 | 12 | 1.5368 | 0.25 |
| 0.7949 | 13.0 | 13 | 1.5368 | 0.25 |
| 0.7949 | 14.0 | 14 | 1.4843 | 0.25 |
| 0.7949 | 15.0 | 15 | 1.4843 | 0.25 |
| 0.7949 | 16.0 | 16 | 1.4413 | 0.25 |
| 0.7949 | 17.0 | 17 | 1.4413 | 0.25 |
| 0.7949 | 18.0 | 18 | 1.4050 | 0.5 |
| 0.7949 | 19.0 | 19 | 1.4050 | 0.5 |
| 0.6509 | 20.0 | 20 | 1.3670 | 0.5 |
| 0.6509 | 21.0 | 21 | 1.3670 | 0.5 |
| 0.6509 | 22.0 | 22 | 1.3404 | 0.5 |
| 0.6509 | 23.0 | 23 | 1.3404 | 0.5 |
| 0.6509 | 24.0 | 24 | 1.3212 | 0.5 |
| 0.6509 | 25.0 | 25 | 1.3212 | 0.5 |
| 0.6509 | 26.0 | 26 | 1.3087 | 0.5 |
| 0.6509 | 27.0 | 27 | 1.3087 | 0.5 |
| 0.6509 | 28.0 | 28 | 1.2969 | 0.75 |
| 0.6509 | 29.0 | 29 | 1.2969 | 0.75 |
| 0.5774 | 30.0 | 30 | 1.2927 | 0.75 |
Framework versions
- Transformers 4.30.0
- Pytorch 2.2.1+cu121
- Datasets 2.19.0
- Tokenizers 0.13.3