swin-tiny-patch4-window7-224-finetuned-mobile-eye-tracking-dataset-v2
This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.2836
- Accuracy: 0.8438
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: 23
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| No log | 1.0 | 9 | 1.2172 | 0.5 |
| 1.3679 | 2.0 | 18 | 0.8452 | 0.6562 |
| 1.0999 | 3.0 | 27 | 0.5728 | 0.7188 |
| 0.7024 | 4.0 | 36 | 0.6103 | 0.6953 |
| 0.5483 | 5.0 | 45 | 0.5153 | 0.75 |
| 0.511 | 6.0 | 54 | 0.5210 | 0.7344 |
| 0.4819 | 7.0 | 63 | 0.4082 | 0.7656 |
| 0.439 | 8.0 | 72 | 0.3823 | 0.8125 |
| 0.4162 | 9.0 | 81 | 0.4072 | 0.7578 |
| 0.411 | 10.0 | 90 | 0.3181 | 0.8672 |
| 0.411 | 11.0 | 99 | 0.3095 | 0.8594 |
| 0.3965 | 12.0 | 108 | 0.3924 | 0.8203 |
| 0.3734 | 13.0 | 117 | 0.3023 | 0.8359 |
| 0.3943 | 14.0 | 126 | 0.3604 | 0.8203 |
| 0.3535 | 15.0 | 135 | 0.3193 | 0.8203 |
| 0.3382 | 16.0 | 144 | 0.3052 | 0.8438 |
| 0.3276 | 17.0 | 153 | 0.3389 | 0.8594 |
| 0.3305 | 18.0 | 162 | 0.2985 | 0.8438 |
| 0.3197 | 19.0 | 171 | 0.3254 | 0.8359 |
| 0.3222 | 20.0 | 180 | 0.2836 | 0.8438 |
| 0.3222 | 21.0 | 189 | 0.3047 | 0.8359 |
| 0.3115 | 22.0 | 198 | 0.3102 | 0.8281 |
| 0.32 | 23.0 | 207 | 0.3079 | 0.8281 |
Framework versions
- Transformers 4.33.0
- Pytorch 2.6.0+cu124
- Datasets 3.5.0
- Tokenizers 0.13.3