Skip to content

Dataset · nhblk123

Large Open Source Code Corpus from GitHub

nhblk123/helaxai_data_pluse

A GitHub-sourced collection of source code (100M to 1B records) for training code LLMs.

What's inside

The dataset card describes it as The Stack v3, the successor to The Stack v2, containing source code in text and tabular formats. The data was collected directly from GitHub for pretraining models with whole-repository context, with a size ranging from 100 million to 1 billion examples.

What you can do with it

  • Pre-training language models for code generation
  • Training models to handle full-repository context
  • Analyzing and gathering statistics on open source code across different programming languages
Tasks
Text generation
Languages
Code
Size
100M–1B rows
Downloads this month
26.4K
Likes
1

How to load it

from datasets import load_dataset

ds = load_dataset("nhblk123/helaxai_data_pluse")