Edit model card

Food Classifier

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:

  • Evaluation loss: 0.7166455984115601
  • Accuracy: 0.8753663366336634

Model Details

A model that can detect 101 variety of food.

Model Description

  • Developed by: Dricz
  • Model type: Image classification
  • Language(s) (NLP): English
  • Finetuned from model: google/vit-base-patch16-224-in21k

Training Details

Training Data

  • training_loss: 1.7299224627936907
  • train_runtime: 3538
  • train_samples_per_second: 21.409
  • train_steps_per_second: 1.338
  • total_flos: 5.8752267138432e+18
  • train_loss: 1.7299224627936907
  • epoch: 1.0

Training Hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 5e-05
  • train_batch_size: 16
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 1

Evaluation

  • eval_loss: 0.7166455984115601
  • eval_accuracy: 0.8753663366336634
  • eval_runtime: 446.9362
  • eval_steps_per_second: 3.533
  • epoch: 1.0
Downloads last month
5
Safetensors
Model size
85.9M params
Tensor type
F32
·
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Dataset used to train Dricz/food-classifier-224