<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
finetuned-pokemon
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the pokemon_images dataset. It achieves the following results on the evaluation set:
- Loss: 0.5483
- Accuracy: 0.8849
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: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 4.2294 | 0.18 | 100 | 4.2177 | 0.2408 |
| 3.3443 | 0.35 | 200 | 3.3392 | 0.4415 |
| 2.6829 | 0.53 | 300 | 2.5744 | 0.6848 |
| 2.1544 | 0.71 | 400 | 1.9695 | 0.7430 |
| 1.5132 | 0.88 | 500 | 1.4847 | 0.7761 |
| 0.8297 | 1.06 | 600 | 1.1503 | 0.7974 |
| 0.6333 | 1.23 | 700 | 0.8776 | 0.8349 |
| 0.4634 | 1.41 | 800 | 0.7625 | 0.8462 |
| 0.4646 | 1.59 | 900 | 0.7139 | 0.8293 |
| 0.4092 | 1.76 | 1000 | 0.6812 | 0.8368 |
| 0.4207 | 1.94 | 1100 | 0.6669 | 0.8330 |
| 0.2222 | 2.12 | 1200 | 0.6346 | 0.8449 |
| 0.201 | 2.29 | 1300 | 0.5713 | 0.8574 |
| 0.2236 | 2.47 | 1400 | 0.5948 | 0.8587 |
| 0.2694 | 2.65 | 1500 | 0.5563 | 0.8680 |
| 0.1744 | 2.82 | 1600 | 0.5784 | 0.8593 |
| 0.2472 | 3.0 | 1700 | 0.6019 | 0.8518 |
| 0.0779 | 3.17 | 1800 | 0.5899 | 0.8693 |
| 0.0966 | 3.35 | 1900 | 0.5861 | 0.8587 |
| 0.0962 | 3.53 | 2000 | 0.5981 | 0.8630 |
| 0.1396 | 3.7 | 2100 | 0.6345 | 0.8455 |
| 0.083 | 3.88 | 2200 | 0.5822 | 0.8668 |
| 0.0649 | 4.06 | 2300 | 0.6148 | 0.8530 |
| 0.0476 | 4.23 | 2400 | 0.5906 | 0.8687 |
| 0.1119 | 4.41 | 2500 | 0.5902 | 0.8649 |
| 0.0748 | 4.59 | 2600 | 0.5671 | 0.8780 |
| 0.0934 | 4.76 | 2700 | 0.5834 | 0.8724 |
| 0.0648 | 4.94 | 2800 | 0.6095 | 0.8580 |
| 0.051 | 5.11 | 2900 | 0.6209 | 0.8574 |
| 0.0341 | 5.29 | 3000 | 0.5671 | 0.8780 |
| 0.1553 | 5.47 | 3100 | 0.5838 | 0.8730 |
| 0.0425 | 5.64 | 3200 | 0.5808 | 0.8712 |
| 0.0269 | 5.82 | 3300 | 0.5734 | 0.8693 |
| 0.1104 | 6.0 | 3400 | 0.5823 | 0.8718 |
| 0.1136 | 6.17 | 3500 | 0.5698 | 0.8774 |
| 0.0276 | 6.35 | 3600 | 0.5996 | 0.8699 |
| 0.0408 | 6.53 | 3700 | 0.5642 | 0.8768 |
| 0.0147 | 6.7 | 3800 | 0.5913 | 0.8762 |
| 0.0395 | 6.88 | 3900 | 0.5838 | 0.8630 |
| 0.014 | 7.05 | 4000 | 0.5722 | 0.8762 |
| 0.025 | 7.23 | 4100 | 0.5749 | 0.8768 |
| 0.0536 | 7.41 | 4200 | 0.5693 | 0.8806 |
| 0.0602 | 7.58 | 4300 | 0.5483 | 0.8849 |
| 0.0618 | 7.76 | 4400 | 0.5597 | 0.8774 |
| 0.0447 | 7.94 | 4500 | 0.5539 | 0.8812 |
| 0.0811 | 8.11 | 4600 | 0.5692 | 0.8762 |
| 0.0514 | 8.29 | 4700 | 0.5665 | 0.8812 |
| 0.0635 | 8.47 | 4800 | 0.5737 | 0.8793 |
| 0.0474 | 8.64 | 4900 | 0.5817 | 0.8799 |
| 0.03 | 8.82 | 5000 | 0.5833 | 0.8774 |
| 0.0224 | 8.99 | 5100 | 0.5743 | 0.8818 |
| 0.0205 | 9.17 | 5200 | 0.5794 | 0.8831 |
| 0.0326 | 9.35 | 5300 | 0.5838 | 0.8806 |
| 0.0617 | 9.52 | 5400 | 0.5777 | 0.8812 |
| 0.019 | 9.7 | 5500 | 0.5781 | 0.8806 |
| 0.0412 | 9.88 | 5600 | 0.5797 | 0.8806 |
Framework versions
- Transformers 4.30.1
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
