|
--- |
|
license: apache-2.0 |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
- f1 |
|
- precision |
|
- recall |
|
model-index: |
|
- name: vit-base-skin |
|
results: [] |
|
--- |
|
|
|
<!-- 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-skin |
|
|
|
This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.7037 |
|
- Accuracy: 0.8549 |
|
- F1: 0.8534 |
|
- Precision: 0.8536 |
|
- Recall: 0.8549 |
|
|
|
## 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: 6 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:| |
|
| 0.7904 | 0.16 | 100 | 0.7231 | 0.7772 | 0.7372 | 0.7397 | 0.7772 | |
|
| 0.7746 | 0.32 | 200 | 0.6372 | 0.7668 | 0.7147 | 0.7458 | 0.7668 | |
|
| 0.611 | 0.48 | 300 | 0.6455 | 0.7409 | 0.6930 | 0.7665 | 0.7409 | |
|
| 0.8406 | 0.64 | 400 | 0.6233 | 0.8031 | 0.8098 | 0.8433 | 0.8031 | |
|
| 0.587 | 0.8 | 500 | 0.6028 | 0.7513 | 0.7163 | 0.7659 | 0.7513 | |
|
| 0.5532 | 0.96 | 600 | 0.4689 | 0.8290 | 0.8090 | 0.8377 | 0.8290 | |
|
| 0.4326 | 1.12 | 700 | 0.4968 | 0.8290 | 0.8200 | 0.8368 | 0.8290 | |
|
| 0.4713 | 1.28 | 800 | 0.4973 | 0.8187 | 0.8222 | 0.8436 | 0.8187 | |
|
| 0.4333 | 1.44 | 900 | 0.5500 | 0.7720 | 0.7615 | 0.7705 | 0.7720 | |
|
| 0.441 | 1.6 | 1000 | 0.5518 | 0.8238 | 0.8398 | 0.8774 | 0.8238 | |
|
| 0.4172 | 1.76 | 1100 | 0.5608 | 0.8031 | 0.7802 | 0.8260 | 0.8031 | |
|
| 0.4062 | 1.92 | 1200 | 0.4730 | 0.8290 | 0.8312 | 0.8704 | 0.8290 | |
|
| 0.271 | 2.08 | 1300 | 0.4893 | 0.8031 | 0.8018 | 0.8164 | 0.8031 | |
|
| 0.2294 | 2.24 | 1400 | 0.4859 | 0.8342 | 0.8369 | 0.8442 | 0.8342 | |
|
| 0.2687 | 2.4 | 1500 | 0.4805 | 0.8394 | 0.8391 | 0.8424 | 0.8394 | |
|
| 0.2348 | 2.56 | 1600 | 0.4667 | 0.8497 | 0.8522 | 0.8567 | 0.8497 | |
|
| 0.2038 | 2.72 | 1700 | 0.5050 | 0.8135 | 0.8148 | 0.8222 | 0.8135 | |
|
| 0.2102 | 2.88 | 1800 | 0.4730 | 0.8497 | 0.8527 | 0.8695 | 0.8497 | |
|
| 0.0978 | 3.04 | 1900 | 0.4673 | 0.8446 | 0.8450 | 0.8508 | 0.8446 | |
|
| 0.104 | 3.19 | 2000 | 0.5348 | 0.8342 | 0.8274 | 0.8313 | 0.8342 | |
|
| 0.0562 | 3.35 | 2100 | 0.5748 | 0.8342 | 0.8264 | 0.8299 | 0.8342 | |
|
| 0.1443 | 3.51 | 2200 | 0.5903 | 0.8446 | 0.8432 | 0.8448 | 0.8446 | |
|
| 0.1245 | 3.67 | 2300 | 0.5773 | 0.8601 | 0.8627 | 0.8779 | 0.8601 | |
|
| 0.081 | 3.83 | 2400 | 0.6190 | 0.8394 | 0.8424 | 0.8487 | 0.8394 | |
|
| 0.1314 | 3.99 | 2500 | 0.6078 | 0.8549 | 0.8509 | 0.8506 | 0.8549 | |
|
| 0.0415 | 4.15 | 2600 | 0.7039 | 0.8290 | 0.8312 | 0.8358 | 0.8290 | |
|
| 0.0402 | 4.31 | 2700 | 0.7477 | 0.8238 | 0.8166 | 0.8179 | 0.8238 | |
|
| 0.0045 | 4.47 | 2800 | 0.7207 | 0.8497 | 0.8493 | 0.8539 | 0.8497 | |
|
| 0.0608 | 4.63 | 2900 | 0.7339 | 0.8342 | 0.8370 | 0.8469 | 0.8342 | |
|
| 0.0168 | 4.79 | 3000 | 0.7894 | 0.8290 | 0.8388 | 0.8539 | 0.8290 | |
|
| 0.0042 | 4.95 | 3100 | 0.7268 | 0.8601 | 0.8628 | 0.8681 | 0.8601 | |
|
| 0.0149 | 5.11 | 3200 | 0.7145 | 0.8601 | 0.8577 | 0.8600 | 0.8601 | |
|
| 0.0074 | 5.27 | 3300 | 0.7424 | 0.8342 | 0.8354 | 0.8380 | 0.8342 | |
|
| 0.0029 | 5.43 | 3400 | 0.7123 | 0.8653 | 0.8649 | 0.8686 | 0.8653 | |
|
| 0.0123 | 5.59 | 3500 | 0.7052 | 0.8653 | 0.8633 | 0.8632 | 0.8653 | |
|
| 0.0028 | 5.75 | 3600 | 0.7027 | 0.8601 | 0.8590 | 0.8601 | 0.8601 | |
|
| 0.0029 | 5.91 | 3700 | 0.7037 | 0.8549 | 0.8534 | 0.8536 | 0.8549 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.29.2 |
|
- Pytorch 1.13.1 |
|
- Datasets 2.14.5 |
|
- Tokenizers 0.13.3 |
|
|
