robotstech/deploy_DL_project_with_fastapi

How to deploy a machine learning project with fastapi

Deploy a Deep learning project with fastapi

This documents shows how to deploy a Deep learning model using fast api. It will be using this MNIST-CNN Model.

Project setup

Before going through the steps make sure you have the following pre-installed

Prerequisite

1. Python 3.6+
2. Virtualenv

Make sure to download/clone this repository and navigate to the folder in yout terminal. Now follow the indtructions below

1. Create the virtual environment.

    virtualenv /path/to/venv --python=/path/to/python3

You can find out the path to your python3 interpreter with the command which python3.

2. Activate the environment and install dependencies.

   • Linux

       source /path/to/venv/bin/activate
       pip install -r requirements.txt
   

   • Windows

       ./path/to/venv/bin/activate
       pip install -r requirements.txt
   

3. Launch the service

    uvicorn main:app --workers 1 --host 0.0.0.0 --port 8008

Posting requests locally

When the service is running, try this link in your browser

    127.0.0.1:8008/docs

You can test the model with the Sample MNIST Image from How to Develop a CNN for MNIST Handwritten Digit Classification below using postman

1. Download this sample image for testing

2. Using Postman: upload image

3. Using Postman: send a Post Request

Refrences

1. MNIST CNN Model repo
2. Tutorial: How to deploy your ConvNet classifier with Keras and FastAPI article
3. How to Develop a CNN for MNIST Handwritten Digit Classification