|
--- |
|
license: apache-2.0 |
|
base_model: google/vit-base-patch16-224 |
|
tags: |
|
- image-classification |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
- precision |
|
- recall |
|
- f1 |
|
model-index: |
|
- name: vit-beta1-0.85 |
|
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-beta1-0.85 |
|
|
|
This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the skin-cancer dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.5102 |
|
- Accuracy: 0.8558 |
|
- Precision: 0.8568 |
|
- Recall: 0.8558 |
|
- F1: 0.8553 |
|
|
|
## 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: 0.0001 |
|
- train_batch_size: 16 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.85,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_steps: 1733 |
|
- num_epochs: 100 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:| |
|
| 1.763 | 1.0 | 321 | 0.9505 | 0.6952 | 0.6346 | 0.6952 | 0.6202 | |
|
| 1.149 | 2.0 | 642 | 0.7147 | 0.7445 | 0.7457 | 0.7445 | 0.7230 | |
|
| 1.0452 | 3.0 | 963 | 0.6250 | 0.7573 | 0.7591 | 0.7573 | 0.7321 | |
|
| 1.0048 | 4.0 | 1284 | 0.5614 | 0.7784 | 0.7792 | 0.7784 | 0.7737 | |
|
| 0.931 | 5.0 | 1605 | 0.6082 | 0.7739 | 0.8020 | 0.7739 | 0.7823 | |
|
| 0.9808 | 6.0 | 1926 | 0.5542 | 0.7982 | 0.7984 | 0.7982 | 0.7951 | |
|
| 0.8908 | 7.0 | 2247 | 0.5957 | 0.7545 | 0.8202 | 0.7545 | 0.7709 | |
|
| 0.7747 | 8.0 | 2568 | 0.5766 | 0.7694 | 0.8155 | 0.7694 | 0.7836 | |
|
| 0.741 | 9.0 | 2889 | 0.5431 | 0.7996 | 0.8190 | 0.7996 | 0.8047 | |
|
| 0.7179 | 10.0 | 3210 | 0.5865 | 0.7774 | 0.8313 | 0.7774 | 0.7904 | |
|
| 0.6102 | 11.0 | 3531 | 0.5288 | 0.8096 | 0.8361 | 0.8096 | 0.8180 | |
|
| 0.574 | 12.0 | 3852 | 0.5991 | 0.7996 | 0.8332 | 0.7996 | 0.8096 | |
|
| 0.4515 | 13.0 | 4173 | 0.5890 | 0.8370 | 0.8334 | 0.8370 | 0.8293 | |
|
| 0.4629 | 14.0 | 4494 | 0.5573 | 0.8121 | 0.8463 | 0.8121 | 0.8205 | |
|
| 0.3927 | 15.0 | 4815 | 0.5279 | 0.8332 | 0.8506 | 0.8332 | 0.8357 | |
|
| 0.3535 | 16.0 | 5136 | 0.5364 | 0.8356 | 0.8494 | 0.8356 | 0.8405 | |
|
| 0.2635 | 17.0 | 5457 | 0.5475 | 0.8547 | 0.8626 | 0.8547 | 0.8532 | |
|
| 0.2493 | 18.0 | 5778 | 0.5102 | 0.8558 | 0.8568 | 0.8558 | 0.8553 | |
|
| 0.2125 | 19.0 | 6099 | 0.6120 | 0.8329 | 0.8623 | 0.8329 | 0.8418 | |
|
| 0.2179 | 20.0 | 6420 | 0.5721 | 0.8568 | 0.8563 | 0.8568 | 0.8563 | |
|
| 0.1598 | 21.0 | 6741 | 0.5503 | 0.8651 | 0.8623 | 0.8651 | 0.8633 | |
|
| 0.1194 | 22.0 | 7062 | 0.5829 | 0.8679 | 0.8672 | 0.8679 | 0.8669 | |
|
| 0.1245 | 23.0 | 7383 | 0.6138 | 0.8682 | 0.8632 | 0.8682 | 0.8629 | |
|
| 0.1239 | 24.0 | 7704 | 0.6136 | 0.8731 | 0.8703 | 0.8731 | 0.8695 | |
|
| 0.1159 | 25.0 | 8025 | 0.5931 | 0.8752 | 0.8724 | 0.8752 | 0.8726 | |
|
| 0.089 | 26.0 | 8346 | 0.5847 | 0.8776 | 0.8743 | 0.8776 | 0.8750 | |
|
| 0.1123 | 27.0 | 8667 | 0.5941 | 0.8752 | 0.8710 | 0.8752 | 0.8719 | |
|
| 0.0779 | 28.0 | 8988 | 0.6038 | 0.8766 | 0.8722 | 0.8766 | 0.8729 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.40.0.dev0 |
|
- Pytorch 2.2.1+cu121 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.15.2 |
|
|
