Research index
Type4Py: Practical Deep Similarity Learning-Based Type Inference for Python
Mir, Latoškinas, Proksch and 1 other
Zenodo (CERN European Organization for Nuclear Research)24 Jan 2022Open access
NobleIDark:/48914/w1/20260524/83D87BE2DOI10.5281/zenodo.5913787
Open access
Type4Py: Practical Deep Similarity Learning-Based Type Inference for Python
Mir, Latoškinas, Proksch and 1 other · Zenodo (CERN European Organization for Nuclear Research) · 2022
NobleIDark:/48914/w1/20260524/83D87BE2DOI10.5281/zenodo.5913787
Abstract
This contains artifacts for the Type4Py paper, which is accepted at the ICSE'22 technical track. Read README for installation and usage guide. The dataset folder contains the type-checked version of the ManyTypes4Py dataset to train Type4Py and re-produce the results of the paper. The implementation of Type4Py is also publicly available on a GitHub repository.