Beer Predictions

Overview

Project Description:

With the intention of predicting a particular Beer type, a Neural Network was developed, trained and deployed.

Project Objectives:

1. Design a Neural Network using PyTorch library.
2. Deploy the model behind an API endpoing using FastAPI library.
3. Deploy the app to the public using Heroku library.
4. Train, test, document, and publish.

Endpoints

/ The root of the directory, which displays all the key information and details about the project, including: Overview, Endpoints, Model Info and More Info.
/docs/ All of the detailed documentation. Read here first.
/health/ Check to ensure that the app is healthy and ready to run.
/beer/type/ Check to review the architecture of the model.
/beers/type/ Predict single Beer type, based on set input criteria.
/model/architecture/ Predict multiple Beer types, based on set input criteria.

Model Info

Input Parameters:
Param Type Validation
brewery_name str Must be a valid brewery name
review_aroma float Must be between 0 and 5
review_appearance float Must be between 0 and 5
review_palate float Must be between 0 and 5
review_taste float Must be between 0 and 5
Output Format:
Format Reason
str If the input params are all scalar, then model will result in a single str prediction.
list of str If the input params are a list, then the model will result in a list of str, who's length is the same as the number of input params.

More Info

Author Chris Mahoney
App BeerPrediction
Repo: BeerPrediction
Version 0.1.0
Published 7/Mar/2021
License MIT