google/vit-base-patch16-224-in21k-v2-finetuned

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: 1.0612
  • Accuracy: 0.7969

Model description

  • Model type: Language model
  • Language(s) (NLP): English
  • License: Apache 2.0
  • Related Model: google/vit-base-patch16-224-in21k
  • Original Checkpoints: google/vit-base-patch16-224-in21k
  • Resources for more information: Research paper

Intended uses & limitations

This model can be used to classify what type of food in the image provided.

Training and evaluation data

The model was trained on food101 dataset with 80:20 train-test-split.

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 5e-05
  • train_batch_size: 16
  • eval_batch_size: 16
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 64
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 3
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Accuracy
1.9201 1.0 947 1.9632 0.7297
1.2002 2.0 1894 1.2327 0.7805
0.9561 3.0 2841 1.0612 0.7969

Framework versions

  • Transformers 4.40.2
  • Pytorch 2.4.0+cu121
  • Datasets 2.21.0
  • Tokenizers 0.19.1
Downloads last month
72
Safetensors
Model size
85.9M params
Tensor type
F32
·
Inference Examples
Unable to determine this model's library. Check the docs .

Model tree for thuyentruong/food_classification_model

Finetuned
(1770)
this model

Dataset used to train thuyentruong/food_classification_model

Evaluation results