vit-base-patch16-224-RU5-10
This model is a fine-tuned version of google/vit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.8095
- Accuracy: 0.7333
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: 5.5e-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: 0.05
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
No log | 0.92 | 9 | 1.2939 | 0.4667 |
1.3501 | 1.95 | 19 | 1.1706 | 0.5833 |
1.2272 | 2.97 | 29 | 1.0594 | 0.6333 |
1.0941 | 4.0 | 39 | 0.9773 | 0.6 |
0.979 | 4.92 | 48 | 0.9142 | 0.6833 |
0.8694 | 5.95 | 58 | 0.8569 | 0.7 |
0.7662 | 6.97 | 68 | 0.8364 | 0.6833 |
0.7002 | 8.0 | 78 | 0.8071 | 0.7 |
0.6443 | 8.92 | 87 | 0.8095 | 0.7333 |
0.629 | 9.23 | 90 | 0.8134 | 0.7167 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu118
- Datasets 2.16.1
- Tokenizers 0.15.0
- Downloads last month
- 0
Finetuned from
Evaluation results
- Accuracy on imagefoldervalidation set self-reported0.733