<!-- 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. -->
vit-base-cifar10
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the cifar10 dataset. It achieves the following results on the evaluation set:
- Loss: 2.3302
- Accuracy: 0.106
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
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.3324 | 1.0 | 664 | 2.3352 | 0.0967 |
| 2.3489 | 2.0 | 1328 | 2.3288 | 0.1049 |
| 2.4899 | 3.0 | 1992 | 2.4473 | 0.0989 |
| 2.479 | 4.0 | 2656 | 2.4894 | 0.1 |
| 2.4179 | 5.0 | 3320 | 2.4404 | 0.0947 |
| 2.3881 | 6.0 | 3984 | 2.3931 | 0.102 |
| 2.3597 | 7.0 | 4648 | 2.3744 | 0.0967 |
| 2.3721 | 8.0 | 5312 | 2.3667 | 0.0935 |
| 2.3456 | 9.0 | 5976 | 2.3495 | 0.1036 |
| 2.3361 | 10.0 | 6640 | 2.3473 | 0.1025 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.1+cu117
- Datasets 2.8.0
- Tokenizers 0.13.2
