Skip to content

A list of paper templates in the area of machine learning.

License

Notifications You must be signed in to change notification settings

daskol/typst-templates

Repository files navigation

Linting and testing

Typst: Templates

A curated list of paper templates in the area of machine learning.

Overview

Some conferences and journals in machine learning allow submissions in PDF without special requirement to use LaTeX. They also provides a template and an example paper in LaTeX. With official author instructions, these materials enable us to make our own template in Typst mark up language. We start with template for ICML and are going to add templates for other Core A* conferences and journals during next calendar years. The fields of research is below.

  • Artificial intelligence (4602).
  • Computer vision and multimedia computation (4603).
  • Data management and data science (4605).
  • Machine learning (4611).

The list of A* conferences of the fields given follows.

Additionally, we are going to provide templates for popular machine learning scientific journals as well.

Usage

You can use this template in the Typst WebApp by clicking Start from template on the dashboard and searching for a template (e.g. lucky-icml).

Alternatively, you can use the CLI to kick this project off using the command

typst init @preview/lucky-icml

Utilities

Colored Annotations

Typst of version 0.10.0 does not produce colored annotations. In order to mitigate the issue, we add a simple script to the repository. The script is plain and simple. One can use it as follows.

./colorize-annotations.py \
    neurips/example-paper.typst.pdf neurips/example-paper-colored.typst.pdf

It is written with PyMuPDF library and inserts colored annotation.

CI/CD

At the moment, CI/CD pipeline tries only to render all available templates with a specific version of typst compiler. In order to run CI/CD checks locally, one can use container image ghcr.io/daskol/typst-templates to run script render-templates.sh. End-to-end command for local check follows.

make image test