MLOps Regression using Abalone's Shell Data

Abalone

Introduction

This project aims to build an MLOps pipeline capable of training and deploying a regression model that determines an Abalone's age based on its characteristics. An Abalone is a shelled marine mollusc, consumed as food by a variety of cultures around the world.

About the dataset

The dataset is a collection of measurements from Abalone specimens and is available at kaggle: Abalone Dataset. Precisely determining the age of an Abalone specimen is a difficult and time consuming task, so being able to estimate it using machine learning methods is a desireable solution. Therefore, this dataset contains a little over 4000 entries with the length, height, weight etc of the Abalone. Below there is an extensive list of parameters from the dataset.

    - sex
    - length
    - diameter
    - height
    - whole weight
    - shucked weight
    - viscera weight
    - shell weight
    - rings

About the model

The model is a regression [...]

About the training

The training is done in two steps: [...]

About the deployment

The deployment is done in two steps: [...]

Technology Stack

• wandb
• keras
• juypter notebook

Getting Started

To get started, clone the repository and run the following command:

git clone https://github.com/lpdcalves/mlops-abalone-regression.git

And then run the following command:

conda env create -f environment.yml

After that, all you need to do is run the pipeline:

cd mlflow
mlflow run .