CrossCodeBench: Benchmarking Cross-Task Generalization of Source Code Models

no code implementations • 8 Feb 2023 • Changan Niu, Chuanyi Li, Vincent Ng, Bin Luo

Despite the recent advances showing that a model pre-trained on large-scale source code data is able to gain appreciable generalization capability, it still requires a sizeable amount of data on the target task for fine-tuning.

Benchmarking Few-Shot Learning +1

Deep Learning Meets Software Engineering: A Survey on Pre-Trained Models of Source Code

no code implementations • 24 May 2022 • Changan Niu, Chuanyi Li, Bin Luo, Vincent Ng

In particular, the development and use of pre-trained models of source code has enabled state-of-the-art results to be achieved on a wide variety of SE tasks.