Getting Started

  1. Create a Weights and Biases account (not required but highly recommended)

  2. Get your data into the proper format. At the moment you should try to get all of your relevant data in either a single CSV file or multiple CSV files (train, validation, test). In either case all of the columns that you want to use should be present in a single CSV file.

  3. Visit one of our many tutorials

  4. Read Training models the complete guide

  5. Visit our Frequently Asked Questions Page for more info.

Documentation

We have two main sources of documentation

(1) Confluence (which if you are reading this you already here!): Contains high-level discussion and guides to using our framework end-to-end along with project plans. Start with confluence.

(2) ReadTheDocs: Contains detailed information on the framework internals and the Python API.

Time Series Forecasting vs Classification vs Anomaly Detection

Before getting started you want to determine what type of problem you have. See this Medium article for tips and tricks getting started.