---
license: apache-2.0
language: code
tags:
- code
- python
- codesearchnet-format
dataset_info:
  features:
  - name: code
    dtype: string
  - name: docstring
    dtype: string
  - name: func_name
    dtype: string
  - name: language
    dtype: string
  - name: repo
    dtype: string
  - name: path
    dtype: string
  - name: url
    dtype: string
  - name: license
    dtype: string
  splits:
  - name: train
    num_bytes: 930859606.6885663
    num_examples: 605202
  - name: validation
    num_bytes: 16569909.301435407
    num_examples: 10728
  - name: test
    num_bytes: 18859879.313334443
    num_examples: 13115
  download_size: 305620969
  dataset_size: 966289395.3033361
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---

---
## 📘 Dataset Description

This dataset contains Python functions and their documentation comments (`docstring`) extracted from public GitHub repositories. Each function is paired with metadata such as the repository name, file path, original license, and GitHub URL.

Only code licensed under the following OSI-approved permissive licenses was included:

- MIT
- Apache License 2.0
- BSD 2-Clause
- BSD 3-Clause
- ISC

These licenses permit reuse and redistribution (including commercial use) provided that proper attribution and license terms are preserved.

---
## 📄 Licensing and Legal Notice

All code snippets in this dataset were collected from public GitHub repositories that explicitly declare their licenses as one of the following: **MIT, Apache-2.0, BSD-2-Clause, BSD-3-Clause, or ISC**.

To comply with license terms:

- Each sample retains its license type in the `license` field.
- Repository names and GitHub URLs are included to allow proper attribution.
- License texts are available in the `LICENSES/` directory.
- No code from repositories lacking a license or using GPL-style licenses was included.

> ⚠️ **Disclaimer**: While care has been taken to ensure license compliance, users are responsible for verifying license terms and ensuring proper attribution when using this dataset in downstream applications.

---
## Splits
- **train**: 605,202 examples
- **validation**: 10,728 examples
- **test**: 13,115 examples

**Total:** 629,045 examples