VGG-pet-Vision

About

VGG-pet-Vision is a project that focuses on classifying images of dogs and cats using the VGG-16 convolutional neural network architecture. The goal is to build a machine learning model that accurately distinguishes between images of dogs and cats.

Project Structure

Models

The trained machine learning model is saved in the artifacts/ directory.

Trained Model: artifacts/model_vgg16.pth

Data

The project uses the Kaggle Dogs vs. Cats dataset, which contains images of dogs and cats for training and testing the model.

Pipelines

1. Data Preprocessing:

   • src/pipelines/data_preprocessing.py: Script for preprocessing the raw image data, including resizing, normalization, and augmentation.

2. Model Training:

   • src/pipelines/model_training.py: Script for training the VGG-16 model using the preprocessed data.

3. Prediction:

   • src/pipelines/prediction.py: Script for making predictions using the trained model on new images.

Usage

To train the model:

python src/pipelines/model_training.py