saltudelft / ml-typeinf-competition

ML-based Type Inference Competition

NOTE: Check out the ManyTypes4Py dataset for benchmarking! This repository is archived!

Intro

Optional static typing provides various benefits for dynamic programming languages (DPLs) such as early bug detection, better program analysis, improved IDE support, and so on. In recent years, researchers have been proposed machine learning-based type inference approaches for DPLs. However, the type inference task is still quite challenging, especially for user-defined and rare types. Motivated by this, we have created a competition for ML-based type inference for Python.

Dataset

ManyTypes4Py

Contribution Guide

Evaluation Ceriteria

Rules

Leaderboard

#	Team (Company/Institution)	F1-Score	Tool/Approach
1
2