Dataset Viewer
Auto-converted to Parquet Duplicate
The dataset viewer is not available for this split.
Parquet error: Scan size limit exceeded: attempted to read 546393943 bytes, limit is 300000000 bytes Make sure that 1. the Parquet files contain a page index to enable random access without loading entire row groups2. otherwise use smaller row-group sizes when serializing the Parquet files
Error code:   TooBigContentError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

Swallow-Nemotron-Post-Training-Dataset-v1-ja-cpt

Dataset Overview

This dataset is a reformatted subset of the tokyotech-llm/Swallow-Nemotron-Post-Training-Dataset-v1 dataset, specifically derived from the v1-Ja-202601 subset. It was created to facilitate Continuous Pre-Training (CPT) by extracting only the text_gpt_oss field from the original data.

Dataset Statistics & Token Counts

The token counts for each category were calculated using the gigatoken library with the openai/gpt-oss-20b tokenizer.

Category File Count Total Tokens Token Size
Code 18 files 6,285,974,424 ~6.29 B
Math 19 files 6,927,925,779 ~6.93 B
STEM 24 files 3,205,527,028 ~3.21 B

License

This dataset is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0) available at https://creativecommons.org/licenses/by/4.0/legalcode.

Downloads last month
54