CodeTrans-MT-Large

Reported on 4 benchmarks across 4 tasks · 1 paper · 4 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing2 results

Code GenerationonCodeSearchNet - Java
Smoothed BLEU-4· 2021-04-06
21.87
SOTA
CodeTrans: Towards Cracking the Language of Silicon's Code Through Self-Supervised Deep Learning and High Performance Computing arXiv:2104.02443
Source Code SummarizationonSummarizing Source Code using a Neural Attention Model - C#
Smoothed BLEU-4· 2021-04-06
23.57
SOTA
CodeTrans: Towards Cracking the Language of Silicon's Code Through Self-Supervised Deep Learning and High Performance Computing arXiv:2104.02443

Adversarial1 result

Text GenerationonCodeSearchNet - Java
Smoothed BLEU-4· 2021-04-06
21.87
SOTA
CodeTrans: Towards Cracking the Language of Silicon's Code Through Self-Supervised Deep Learning and High Performance Computing arXiv:2104.02443

Computer Code1 result

Code Documentation GenerationonCodeSearchNet - Java
Smoothed BLEU-4· 2021-04-06
21.87
SOTA
CodeTrans: Towards Cracking the Language of Silicon's Code Through Self-Supervised Deep Learning and High Performance Computing arXiv:2104.02443