METEoRS's picture
update model card README.md
ea08269
metadata
license: apache-2.0
tags:
  - generated_from_trainer
datasets:
  - food101
metrics:
  - accuracy
model-index:
  - name: my_awesome_food_model
    results:
      - task:
          name: Image Classification
          type: image-classification
        dataset:
          name: food101
          type: food101
          config: default
          split: train[:20200]
          args: default
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.8853960396039604

my_awesome_food_model

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the food101 dataset. It achieves the following results on the evaluation set:

  • Loss: 0.4703
  • Accuracy: 0.8854

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: 8
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 16
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 10

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.4019 1.0 1010 1.3796 0.8156
0.6238 2.0 2020 0.6604 0.8448
0.3691 3.0 3030 0.5661 0.8522
0.3947 4.0 4040 0.5226 0.8614
0.3511 5.0 5050 0.5125 0.8644
0.2504 6.0 6060 0.5180 0.8656
0.1285 7.0 7070 0.5312 0.8668
0.2301 8.0 8080 0.4779 0.875
0.0844 9.0 9090 0.4823 0.8839
0.1189 10.0 10100 0.4703 0.8854

Framework versions

  • Transformers 4.29.2
  • Pytorch 2.0.1+cu118
  • Datasets 2.12.0
  • Tokenizers 0.13.3