pikachu_model
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.1405
- Accuracy: 0.9786
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: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 3.9745 | 1.0 | 70 | 3.8989 | 0.5574 |
| 3.0708 | 1.99 | 140 | 3.0319 | 0.8415 |
| 2.4196 | 2.99 | 210 | 2.4623 | 0.9225 |
| 1.9768 | 4.0 | 281 | 2.0344 | 0.9492 |
| 1.6809 | 5.0 | 351 | 1.7300 | 0.9715 |
| 1.4707 | 5.99 | 421 | 1.4962 | 0.9742 |
| 1.2854 | 6.99 | 491 | 1.3465 | 0.9724 |
| 1.1553 | 8.0 | 562 | 1.2592 | 0.9742 |
| 1.0859 | 9.0 | 632 | 1.1849 | 0.9724 |
| 1.0657 | 9.96 | 700 | 1.1405 | 0.9786 |
Framework versions
- Transformers 4.33.2
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3
- Downloads last month
- 0