bvallelunga / lambda-sentiment

AWS lambda function + REST API demo for sentiment

Lambda Sentiment

Setup

Make sure you have the following installed in your env:

• Python 3.6
• Chalice

You will also need to have your AWS credential configured. If you have previously configured your machine to run boto3 (the AWS SDK for Python) or the AWS CLI then you can skip this section. Otherwise, you will need to do the following:

$ mkdir ~/.aws
$ cat >> ~/.aws/config
[default]
aws_access_key_id=YOUR_ACCESS_KEY_HERE
aws_secret_access_key=YOUR_SECRET_ACCESS_KEY
region=YOUR_REGION (such as us-west-2, us-west-1, etc)

Project Structure

app.py is the heart of the code, it is a flask-esque file that routes/handles incoming requests; this is where we would call ModelInterface to generate predictions for incoming requests.

chalicelib/ is where all the supplementary code files live i.e. the project files which are imported in app.py. For us, this is where the modelinterface.py for an app would live.

requirements.txt is the app's requirements file; nothing different here.

Deploying a Lambda Function

First, install the app's dependencies:

$ pip install -r requirements.txt

Next, create and deploy a lambda function:

$ chalice deploy
Creating deployment package.
Creating IAM role: sentiment-dev
Creating lambda function: sentiment-dev
Creating Rest API
Resources deployed:
  - Lambda ARN: arn:aws:lambda:us-east-1:017491206004:function:sentiment-dev
  - Rest API URL: https://s26c58spfj.execute-api.us-east-1.amazonaws.com/api/

You can now get predictions by sending Doppler requests to the REST API URL. Note, you can also find the ARN and REST API URL in the .chalice/deployed folder.