VS Code Extension for Kaggle

Simplify your Kaggle workflow: Update and manage datasets and kernels directly within VS Code.

Prerequisites

This extension leverages the kaggle command-line tool for interacting with Kaggle. You'll need to have the Kaggle CLI installed and configured on your machine.

Extension Settings

This extension contributes the following setting:

• fastkaggle.executablePath: Path to the kaggle executable. If the kaggle command is in your system's PATH, you can use this extension without modifying this setting.

Features

Update Datasets and Notebooks

• Shortcut: Ctrl+Shift+U.
• Command Palette: Cmd+Shift+P then select "Kaggle: Update Dataset/Notebook".

This command automatically detects the dataset-metadata.json or kernel-metadata.json file in the current file's directory or its parent directories. It then uses that directory as the working directory for executing the Kaggle update command.

Check Update Status

Quickly check the status of your updated datasets and notebooks by pressing Cmd+Shift+K. This is especially useful for monitoring long-running notebooks. ✨

Create Datasets and Notebook

• Command Palette: Cmd+Shift+P then select "Kaggle: New Dataset" or "Kaggle: New Notebook".

This command will generate metadata and a plain jupyter notebook. It helps you quickly create new notebooks and datasets.

Metadata Validation

The extension validates the JSON schema of kernel-metadata.json and dataset-metadata.json files, ensuring your Kaggle dataset and notebook definitions are correct.

Sidebar View for Datasets, Notebooks and Models

The extension provides a sidebar view for datasets, notebooks and models.

Right-click on a dataset or notebook in the sidebar to access a context menu, allowing you to check its status.

The status is displayed based on the output of the kaggle status subcommand.

List Competitions

Command Palette: Cmd+Shift+P then select "Kaggle: List Competitions".

Lists recent competitions sorted by deadline. This executes kaggle competitions list and displays the output.

Enjoy your Kaggle journey with FastKaggle!