Add part 3 and some other updates

Author: aditya10

2-docker-nginx-gunicorn/README.md

@@ -1,7 +1,7 @@
 # Flask ML API with Docker, Gunicorn WSGI and Nginx
 Building upon the first phase, we dockerize the flask app, and use Gunicorn WSGI
 to achieve a much more stable and production-ready Flask API. We also use Nginx 
-to interface with 
+to interface with Gunicorn.
 
 The model is developed in [this tutorial](https://medium.com/technonerds/using-fastais-ulmfit-to-make-a-state-of-the-art-multi-label-text-classifier-bf54e2943e83).
 

2-docker-nginx-gunicorn/docker-compose.yml

@@ -15,7 +15,7 @@ services:
       - "5000:5000"
 
   nginx:
-    container_name: nginx_1
+    container_name: nginx
     restart: always
     build: ./nginx
     networks:

3-multiple-endpoints-blueprints/README.md

@@ -0,0 +1,42 @@
+# Managing multiple endpoints on Flask ML API
+Adding multiple endpoints to the project in part 2 would make the `app.py` file very large and unorganized. 
+So, we can use Blueprints - a built in Flask tool to organize endpoints into separate files.
+
+In part 2, we dockerized the flask API and we use Gunicorn WSGI
+to achieve a much more stable and production-ready Flask API. 
+We also use Nginx to interface with Gunicorn.
+
+The model is developed in [this tutorial](https://medium.com/technonerds/using-fastais-ulmfit-to-make-a-state-of-the-art-multi-label-text-classifier-bf54e2943e83).
+
+## Anatomy
+* Gunicorn:
+    * Gunicorn is a popular WSGI that works seamlessly with Flask.
+    * Flask needs a Web Server Gateway Interface (WSGI) to talk to a web server.
+    * Flask's built-in WSGI is not capable of handling production APIs, because it lacks security features and can only run one worker.
+    * In this project, Gunicorn will start automatically in the api Docker container with the following config (see Dockerfile):
+        ```
+        ["gunicorn", "-w", "3", "-b", ":5000", "-t", "360", "--reload", "api.wsgi:app"]
+        ```
+* Nginx:
+    * Nginx is a popular webserver platform. It is used to interface with gunicorn (on port 5000) and relay it to port 80.
+* Docker:
+    * Docker and docker-compose allow apps to be easily deployed. 
+    * Using docker, we containerize our API to work independently of the environment
+
+## How to run
+* Install `docker` and `docker-compose` for your specific platform (you can easily find instructions on Google)
+* Start the API using:
+    ```
+    docker-compose build
+    docker-compose up
+    ```
+    This will build and spin up a virtual container with the API.
+    * Building the continer for the first time will take a while, as it needs to install all requirements
+    * You can use `0.0.0.0:80` to interface with it.
+* You can now query `GET 0.0.0.0/classification` using [Postman](https://www.getpostman.com/) with a json body:
+    ``` 
+    {
+    	"text": "Your classification text goes here"
+    }
+    ```
+    It will return the associated category.

3-multiple-endpoints-blueprints/api/Dockerfile

@@ -0,0 +1,15 @@
+FROM python:3.6
+
+#update
+RUN apt-get update
+
+#install requirements
+COPY ./requirements.txt /tmp/requirements.txt
+WORKDIR /tmp
+RUN pip3 install -r requirements.txt
+
+#copy app
+COPY . /api
+WORKDIR /
+
+CMD ["gunicorn", "-w", "3", "-b", ":5000", "-t", "360", "--reload", "api.wsgi:app"]

3-multiple-endpoints-blueprints/api/__init__.py

@@ -0,0 +1,9 @@
+from flask import Flask
+from .endpoints.classification import classification_api
+
+app = Flask(__name__)
+app.register_blueprint(classification_api)
+
+
+if __name__ == '__main__':
+    app.run(host='0.0.0.0')