---
license: apache-2.0
base_model: google/vit-base-patch16-224-in21k
tags:
- generated_from_trainer
datasets:
- webdataset
metrics:
- accuracy
- f1
- precision
- recall
model-index:
- name: vit-base-patch16-224-in21k-finetuned_v2024-7-24-frost
  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: 0.9530973451327434
    - name: F1
      type: f1
      value: 0.8798185941043084
    - name: Precision
      type: precision
      value: 0.8858447488584474
    - name: Recall
      type: recall
      value: 0.8738738738738738
---

<!-- 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-patch16-224-in21k-finetuned_v2024-7-24-frost

This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the webdataset dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1391
- Accuracy: 0.9531
- F1: 0.8798
- Precision: 0.8858
- Recall: 0.8739

## 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
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 17
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch   | Step | Validation Loss | Accuracy | F1     | Precision | Recall |
|:-------------:|:-------:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
| 0.3281        | 1.5625  | 100  | 0.3177          | 0.9009   | 0.6957 | 0.8767    | 0.5766 |
| 0.2532        | 3.125   | 200  | 0.2424          | 0.9177   | 0.7832 | 0.8116    | 0.7568 |
| 0.1762        | 4.6875  | 300  | 0.1849          | 0.9407   | 0.8453 | 0.8673    | 0.8243 |
| 0.1525        | 6.25    | 400  | 0.1834          | 0.9257   | 0.8056 | 0.8286    | 0.7838 |
| 0.1447        | 7.8125  | 500  | 0.1612          | 0.9416   | 0.8472 | 0.8714    | 0.8243 |
| 0.1114        | 9.375   | 600  | 0.1522          | 0.9434   | 0.8545 | 0.8624    | 0.8468 |
| 0.1004        | 10.9375 | 700  | 0.1525          | 0.9451   | 0.8571 | 0.8774    | 0.8378 |
| 0.0831        | 12.5    | 800  | 0.1442          | 0.9513   | 0.8741 | 0.8884    | 0.8604 |
| 0.0654        | 14.0625 | 900  | 0.1378          | 0.9496   | 0.8690 | 0.8873    | 0.8514 |
| 0.0583        | 15.625  | 1000 | 0.1391          | 0.9531   | 0.8798 | 0.8858    | 0.8739 |


### Framework versions

- Transformers 4.42.4
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1