This is the instruction for my implementation. You can run my code through 2 options: Kaggle or Colab4

• Kaggle: If you want to run all the steps (preprocess data, train, and evaluate)
• Google Colab: If you only want to evaluate the model on your dataset

My model "dogvscat.h5" is shared on Google Drive: 'https://drive.google.com/file/d/1hNVqzvvxysg7o8NZK3wqv8UsFd5VtUAF/view?usp=sharing'

• Access to my Kaggle notebook: https://www.kaggle.com/code/npk7264/dog-vs-cat-classification.

• This notebook contains all the steps: preprocessing data, training, and evaluation

• Select "Copy & Edit" to reimplement my code

• After that, "Run All" to run all scripts in my notebook

• To evaluate the model with your dataset, in the table of contents, there are:

  • "Test saved model on an image": this part will predict for only one image
  • "Test saved model on your testing dataset": this part is used for evaluation with the full dataset

• Download the notebook: https://github.com/npk7264/Dog-Vs-Cat-Classification/blob/main/evaluate_on_your_dataset.ipynb.

• This notebook is only used for the evaluation of your dataset because I train the model on Kaggle.

• Upload it to Google Colab: https://colab.research.google.com/

• In "Test on an image" part, replace "your_img_path" to predict your image

• In "Evaluate on your dataset", set up you dataset with structure: