<!-- 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-HAM-10000-sharpened
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the ahishamm/HAM_db_sharpened dataset. It achieves the following results on the evaluation set:
- Loss: 0.4392
- Accuracy: 0.8529
- Recall: 0.8529
- F1: 0.8529
- Precision: 0.8529
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: 4
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Recall | F1 | Precision |
---|---|---|---|---|---|---|---|
0.7303 | 0.2 | 100 | 0.7828 | 0.7197 | 0.7197 | 0.7197 | 0.7197 |
0.7198 | 0.4 | 200 | 0.7519 | 0.7377 | 0.7377 | 0.7377 | 0.7377 |
0.7519 | 0.6 | 300 | 0.7125 | 0.7541 | 0.7541 | 0.7541 | 0.7541 |
0.6657 | 0.8 | 400 | 0.6623 | 0.7571 | 0.7571 | 0.7571 | 0.7571 |
0.5896 | 1.0 | 500 | 0.5964 | 0.7835 | 0.7835 | 0.7835 | 0.7835 |
0.515 | 1.2 | 600 | 0.5745 | 0.8015 | 0.8015 | 0.8015 | 0.8015 |
0.4318 | 1.4 | 700 | 0.5061 | 0.8200 | 0.8200 | 0.8200 | 0.8200 |
0.4299 | 1.6 | 800 | 0.5239 | 0.8075 | 0.8075 | 0.8075 | 0.8075 |
0.4793 | 1.8 | 900 | 0.5366 | 0.8125 | 0.8125 | 0.8125 | 0.8125 |
0.4202 | 2.0 | 1000 | 0.4882 | 0.8244 | 0.8244 | 0.8244 | 0.8244 |
0.2105 | 2.2 | 1100 | 0.5330 | 0.8234 | 0.8234 | 0.8234 | 0.8234 |
0.2597 | 2.4 | 1200 | 0.4604 | 0.8369 | 0.8369 | 0.8369 | 0.8369 |
0.2261 | 2.59 | 1300 | 0.4893 | 0.8409 | 0.8409 | 0.8409 | 0.8409 |
0.1853 | 2.79 | 1400 | 0.4793 | 0.8494 | 0.8494 | 0.8494 | 0.8494 |
0.1739 | 2.99 | 1500 | 0.4392 | 0.8529 | 0.8529 | 0.8529 | 0.8529 |
0.0629 | 3.19 | 1600 | 0.4941 | 0.8584 | 0.8584 | 0.8584 | 0.8584 |
0.0802 | 3.39 | 1700 | 0.4974 | 0.8613 | 0.8613 | 0.8613 | 0.8613 |
0.0712 | 3.59 | 1800 | 0.5416 | 0.8594 | 0.8594 | 0.8594 | 0.8594 |
0.0365 | 3.79 | 1900 | 0.5318 | 0.8574 | 0.8574 | 0.8574 | 0.8574 |
0.0591 | 3.99 | 2000 | 0.5344 | 0.8574 | 0.8574 | 0.8574 | 0.8574 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu118
- Datasets 2.13.1
- Tokenizers 0.13.3