Skip to content

Dataset · google

CodeXGLUE code vulnerability detection dataset

google/code_x_glue_cc_defect_detection

Dataset for binary classification of source code for vulnerabilities.

What's inside

Contains 10K to 100K source code examples in the code language. The task is binary classification (0/1): identifying insecure code susceptible to resource leaks or DoS attacks. Data is taken from the Microsoft CodeXGLUE repository.

What you can do with it

  • Train code vulnerability detection models
  • Build AI-based static code analysis systems
Tasks
Text classification
Languages
Code
Size
10K–100K rows
Downloads this month
2K
Likes
29

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("google/code_x_glue_cc_defect_detection")