Dog Breed Classifier

Usage

Clone this repo.

$ git clone https://github.com/ysm0622/dog_breed_classifier.git

Download dog image dataset.

$ cd dog_breed_classifier
$ sh data/download_data.sh

Install dependencies.

Install Pipenv first, if you didn't installed it.

MacOS

$ brew install pipenv

Ubuntu

$ sudo apt install software-properties-common python-software-properties
$ sudo add-apt-repository ppa:pypa/ppa
$ sudo apt update
$ sudo apt install pipenv

Using pip

$ pip install pipenv

Then, install all dependencies using this command.

$ pipenv install --skip-lock

Dependencies

[packages]
tensorflow = "*"
pprint = "*"
"partial.py" = "*"
sklearn = "*"
scipy = "*"

Build TFRecords Data

$ pipenv shell
$ python src/data.py

It takes really really long time. (about 3-4 hours) shit...

Now you got *.tfrecords files under data/Records/.

Run

🔥 ing......

ysm-dev / dog_breed_classifier