---
library_name: transformers
license: apache-2.0
base_model: google/vit-base-patch16-224
tags:
- generated_from_trainer
datasets:
- webdataset
metrics:
- accuracy
- f1
- precision
- recall
model-index:
- name: frost-vision-v2-google_vit-base-patch16-224-v2024-11-11
  results:
  - task:
      name: Image Classification
      type: image-classification
    dataset:
      name: webdataset
      type: webdataset
      config: default
      split: train
      args: default
    metrics:
    - name: Accuracy
      type: accuracy
      value: 0.9320422535211268
    - name: F1
      type: f1
      value: 0.8224471021159153
    - name: Precision
      type: precision
      value: 0.8171846435100548
    - name: Recall
      type: recall
      value: 0.8277777777777777
---

<!-- 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. -->

# frost-vision-v2-google_vit-base-patch16-224-v2024-11-11

This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the webdataset dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1658
- Accuracy: 0.9320
- F1: 0.8224
- Precision: 0.8172
- Recall: 0.8278

## 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: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 30
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Accuracy | F1     | Precision | Recall |
|:-------------:|:-------:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
| 0.3127        | 1.4085  | 100  | 0.2932          | 0.8940   | 0.6725 | 0.8153    | 0.5722 |
| 0.193         | 2.8169  | 200  | 0.2136          | 0.9190   | 0.7834 | 0.7969    | 0.7704 |
| 0.1503        | 4.2254  | 300  | 0.1815          | 0.9278   | 0.8100 | 0.8108    | 0.8093 |
| 0.1313        | 5.6338  | 400  | 0.1623          | 0.9327   | 0.8183 | 0.8415    | 0.7963 |
| 0.1166        | 7.0423  | 500  | 0.1658          | 0.9320   | 0.8224 | 0.8172    | 0.8278 |
| 0.093         | 8.4507  | 600  | 0.1606          | 0.9384   | 0.8405 | 0.8276    | 0.8537 |
| 0.0931        | 9.8592  | 700  | 0.1625          | 0.9366   | 0.8370 | 0.8191    | 0.8556 |
| 0.0733        | 11.2676 | 800  | 0.1714          | 0.9356   | 0.8310 | 0.8287    | 0.8333 |
| 0.0693        | 12.6761 | 900  | 0.1568          | 0.9398   | 0.8403 | 0.8475    | 0.8333 |
| 0.0615        | 14.0845 | 1000 | 0.1666          | 0.9342   | 0.8270 | 0.8262    | 0.8278 |
| 0.0562        | 15.4930 | 1100 | 0.1636          | 0.9394   | 0.8404 | 0.8420    | 0.8389 |
| 0.0507        | 16.9014 | 1200 | 0.1613          | 0.9401   | 0.8435 | 0.8388    | 0.8481 |
| 0.0552        | 18.3099 | 1300 | 0.1590          | 0.9412   | 0.8455 | 0.8447    | 0.8463 |
| 0.0439        | 19.7183 | 1400 | 0.1704          | 0.9394   | 0.8425 | 0.8333    | 0.8519 |
| 0.0367        | 21.1268 | 1500 | 0.1702          | 0.9426   | 0.8484 | 0.8523    | 0.8444 |
| 0.0424        | 22.5352 | 1600 | 0.1685          | 0.9394   | 0.8419 | 0.8358    | 0.8481 |
| 0.0306        | 23.9437 | 1700 | 0.1771          | 0.9380   | 0.8397 | 0.8262    | 0.8537 |
| 0.0352        | 25.3521 | 1800 | 0.1691          | 0.9401   | 0.8440 | 0.8364    | 0.8519 |
| 0.0323        | 26.7606 | 1900 | 0.1687          | 0.9426   | 0.8509 | 0.8409    | 0.8611 |
| 0.0297        | 28.1690 | 2000 | 0.1732          | 0.9401   | 0.8455 | 0.8304    | 0.8611 |
| 0.0229        | 29.5775 | 2100 | 0.1712          | 0.9412   | 0.8475 | 0.8360    | 0.8593 |


### Framework versions

- Transformers 4.44.2
- Pytorch 2.5.0+cu121
- Datasets 3.1.0
- Tokenizers 0.19.1