Getting started with DVC

In this tutorial, we are going to master the principal commands and create a data pipeline with DVC.

Detailed description of the project

The article with the explanations is Getting started with DVC

Project Structure

• data/: contains all the data
  • raw_data/: contains original data
  • processed_data/: contained processed data
• src: contains the following scripts
  • preprocess.py: Python script to preprocess the dataset
  • split.py: Python script to split pre-processed data into training and test sets
  • train.py: Python script to train catboost model, save artifact and performances

Set up the project

1. Clone the template branch:

git clone --branch template https://github.com/eugeniaring/dvc-getting-started-guide.git

2. Create a virtual environment:

python3 -m venv venv
source venv/bin/activate

3. Install requirements

pip install -r requirements.txt