This project utilizes the potential of machine learning to learn numerical patterns through drawing numbers on images to reproduce an Image Sorter. Machine learning was done through an IPython file, interpreted by a Notebook Jupyter, and the UI application was created using the Flask Framework.
- Python - Version 3.6.3 - It can be the major version of python or a minor version. You can easily mantain multiple versions from python with Pyenv library.
virtualenv --python={PATH_TO_PYTHON_3.6.3}/python.exe venv
virtualenv --python={PATH_TO_PYTHON_3.6.3}/python3.6 venv
\venv\Scripts\activate.bat
source venv/bin/activate
pip install -r requeriments.txt
First running all the MNIST file with Jupyter Notebook.
jupyter notebook
After that, close the jupyter notebook and run the commands:
set FLASK_APP=srv.py && cd digits_classifier && flask run
export FLASK_APP=srv.py && cd digits_classifier && flask run
Use the Pycharm IDE to manage your apps.
If you have an access from Intel AI Dev Cloud and want run this application on your virtual machine, only install the wget package and run all blocks from MNIST file. After that, copy all path from your local machine, make all the steps bellow.
Please create a new issue to buf fixes or contributing with us.
This project is licensed under the Apache License - see the LICENSE.md file for details