LCC Java is a Java code dataset released by Microsoft in 2023, originating from The Large Code Corpus (LCC) project, aimed at providing high-quality code corpora for tasks such as code understanding, code generation, and code completion.

The dataset is approximately 1.87 GB and contains 120,000 Java code samples, sourced from real Java source code in open-source projects, and is divided into training, validation, and test sets. Each sample mainly contains context code and the corresponding target code snippet, stored in JSON format, and can be used for supervised learning and sequence-to-sequence (Seq2Seq) training of code-related models.

Dataset Composition

The dataset includes the following three parts:

• Training set (train): 100,000 samples, approximately 1.61 GB.
• Validation set (validation): 10,000 samples, approximately 131 MB.
• Test set (test): 10,000 samples, approximately 129 MB.

Each sample mainly contains the following fields:

• context: Context code, providing background information about the code for the model.
• gt: Target code snippet, i.e., the code that the model needs to predict or generate.