metadata
license: apache-2.0
library_name: peft
tags:
- generated_from_trainer
datasets:
- medmnist-v2
metrics:
- accuracy
- precision
- recall
- f1
base_model: google/vit-base-patch16-224-in21k
model-index:
- name: blood-vit-base-finetuned
results: []
blood-vit-base-finetuned
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the medmnist-v2 dataset. It achieves the following results on the evaluation set:
- Loss: 0.0692
- Accuracy: 0.9790
- Precision: 0.9772
- Recall: 0.9785
- F1: 0.9778
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: 0.005
- 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
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 |
|---|---|---|---|---|---|---|---|
| 0.4059 | 1.0 | 187 | 0.1878 | 0.9311 | 0.9132 | 0.9328 | 0.9201 |
| 0.3796 | 2.0 | 374 | 0.2729 | 0.9083 | 0.9131 | 0.8875 | 0.8861 |
| 0.424 | 3.0 | 561 | 0.3701 | 0.8668 | 0.8797 | 0.8520 | 0.8492 |
| 0.3141 | 4.0 | 748 | 0.1849 | 0.9381 | 0.9267 | 0.9336 | 0.9283 |
| 0.2553 | 5.0 | 935 | 0.1075 | 0.9644 | 0.9630 | 0.9612 | 0.9617 |
| 0.2686 | 6.0 | 1122 | 0.1679 | 0.9486 | 0.9561 | 0.9437 | 0.9489 |
| 0.2556 | 7.0 | 1309 | 0.0934 | 0.9661 | 0.9651 | 0.9599 | 0.9619 |
| 0.1777 | 8.0 | 1496 | 0.0835 | 0.9696 | 0.9697 | 0.9683 | 0.9686 |
| 0.1607 | 9.0 | 1683 | 0.0739 | 0.9772 | 0.9733 | 0.9792 | 0.9759 |
| 0.1898 | 10.0 | 1870 | 0.0627 | 0.9790 | 0.9764 | 0.9812 | 0.9786 |
Framework versions
- PEFT 0.9.0
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2