metadata
library_name: transformers
license: apache-2.0
base_model: google/vit-base-patch16-224
tags:
- generated_from_trainer
datasets:
- medmnist-v2
metrics:
- accuracy
- f1
model-index:
- name: ViT_bloodmnist
results:
- task:
name: Image Classification
type: image-classification
dataset:
name: medmnist-v2
type: medmnist-v2
config: bloodmnist
split: validation
args: bloodmnist
metrics:
- name: Accuracy
type: accuracy
value: 0.9748611517100263
- name: F1
type: f1
value: 0.97180354304681
ViT_bloodmnist
This model is a fine-tuned version of google/vit-base-patch16-224 on the medmnist-v2 dataset. It achieves the following results on the evaluation set:
- Loss: 0.0879
- Accuracy: 0.9749
- F1: 0.9718
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: 5e-05
- train_batch_size: 32
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
|---|---|---|---|---|---|
| 0.2747 | 1.0 | 374 | 0.0930 | 0.9696 | 0.9652 |
| 0.0955 | 2.0 | 748 | 0.0998 | 0.9702 | 0.9670 |
| 0.0405 | 3.0 | 1122 | 0.0812 | 0.9743 | 0.9725 |
| 0.0194 | 4.0 | 1496 | 0.0829 | 0.9796 | 0.9784 |
| 0.0081 | 5.0 | 1870 | 0.1328 | 0.9720 | 0.9696 |
| 0.0026 | 6.0 | 2244 | 0.1252 | 0.9743 | 0.9735 |
| 0.0004 | 7.0 | 2618 | 0.0997 | 0.9790 | 0.9778 |
| 0.0001 | 8.0 | 2992 | 0.1049 | 0.9784 | 0.9768 |
| 0.0001 | 9.0 | 3366 | 0.1072 | 0.9778 | 0.9761 |
| 0.0001 | 10.0 | 3740 | 0.1077 | 0.9778 | 0.9761 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.2
- Tokenizers 0.19.1