vit-base-flowers102 / README.md
MaulikMadhavi's picture
trained oxford-flowers
1ad31c5
metadata
license: apache-2.0
base_model: google/vit-base-patch16-224-in21k
tags:
  - image-classification
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: vit-base-flowers102
    results: []

vit-base-flowers102

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the nelorth/oxford-flowers dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0770
  • Accuracy: 0.9853

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
  • num_epochs: 4
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy
2.5779 0.22 100 2.8895 0.7775
1.2226 0.45 200 1.5942 0.9255
0.606 0.67 300 0.8012 0.9529
0.3413 0.89 400 0.4845 0.9706
0.1571 1.11 500 0.2611 0.9814
0.1237 1.34 600 0.1691 0.9784
0.049 1.56 700 0.1146 0.9892
0.0763 1.78 800 0.1209 0.9863
0.0864 2.0 900 0.1223 0.9804
0.0786 2.23 1000 0.1075 0.9833
0.0269 2.45 1100 0.0919 0.9843
0.0178 2.67 1200 0.0795 0.9873
0.0165 2.9 1300 0.0727 0.9873
0.0144 3.12 1400 0.0784 0.9853
0.0138 3.34 1500 0.0759 0.9853
0.0135 3.56 1600 0.0737 0.9863
0.0123 3.79 1700 0.0770 0.9853

Framework versions

  • Transformers 4.35.2
  • Pytorch 2.1.0+cu121
  • Datasets 2.16.1
  • Tokenizers 0.15.0