Credits & Licenses

This dataset is a merged and standardized version of multiple public datasets. All original authors retain their rights under their respective licenses.

Sources

Original Dataset (of the OpenCodeInstruct 75K dataset above)

Name: NVIDIA OpenCodeInstruct
Author: NVIDIA
License: Creative Commons Attribution 4.0 (CC-BY 4.0)
License URL: https://creativecommons.org/licenses/by/4.0/
Source: https://huggingface.co/datasets/nvidia/OpenCodeInstruct

Relationship to the Original Dataset

OpenCodeInstruct 75K is derived from the NVIDIA OpenCodeInstruct dataset.
It consists of the first 75,000 rows extracted from the original dataset.
The underlying content of those rows was not modified during extraction.

Modifications in OpenCodeInstruct 75K

The dataset was limited to the first 75,000 rows of the original dataset.
No additional filtering or semantic modification was performed beyond this row limit.
The subset was redistributed separately under the BSD 3-Clause License.

Use in This Dataset

OpenCodeInstruct 75K is included as one of the sources used to build this final merged dataset.
While OpenCodeInstruct 75K itself contains 75,000 rows, the final merged dataset contains more rows, because it also incorporates the other datasets listed in this document.

Notes

All datasets were reformatted to a unified JSONL schema (instruction, response).
This project performs aggregation, subsetting, and schema standardization.
No claim of original authorship over the underlying data is made.
Attribution to the original datasets is preserved in accordance with their respective licenses.

License

This dataset is released under the BSD 3-Clause License. License URL: https://opensource.org/licenses/BSD-3-Clause