Dogs Breed Classification

1. Installation

Make sure you can use Kaggle API.

> git clone https://github.com/dungxibo123/DogBreedsClassification
> cd DogBreedsClassification
> pip install -r requirements.txt
> pip install -U --force-reinstall protobuf==3.13.0 altair==4.0.0 # Due to the confliction some where inside the streamlit package

Note: If you have a GPU device, make sure you have installed the PyTorch package with compatible cuda version.

2. Preparation

> pip install -U kaggle
> kaggle datasets download -d jessicali9530/stanford-dogs-dataset
> unzip -qq stanford-dogs-dataset.zip -d data/


# Preparing mock data (use can define the number of class that you want to copy to the mock folder at line 16 [0:<your desire number of classes]
> python setup.py

3. Running the streamlit

Simply run

> streamlit run app.py

Everything should be fine now. If it is not, contact me :'(