MAL-Character-Recog-Flask-Server

This project is used for hosting the DL pytorch model of malayalam character recognition. Made in Flask a python framework for backend. Really simple and neat.

Getting Started

Unix Bash (Linux, Mac, etc.):

$ export FLASK_APP=app
$ export FLASK_ENV=development
$ flask run

Windows CMD:

> set FLASK_APP=app
> set FLASK_ENV=development
> flask run

Windows PowerShell:

> $env:FLASK_APP = "app"
> $env:FLASK_ENV = "development"
> flask run

Prerequisites

FLASK
TORCH
TORCHVISION
NUMPY

Deployment

Currently deployed to heroku using gunicorn. Requirement text is provided with the versions of libraries used.

Built With

• FLASK - The web framework used
• TORCH - The DL model behind it
• HEROKU CLI - Used for CI in Heroku

License

This project is licensed under the MIT License - see the LICENSE.md file for details

Contributors ✨

Akhil Mohan

Acknowledgments

• Pytorch documentation
• Heroku simplicity loved it
• Some testing alpha beta