metadata
license: apache-2.0
task_categories:
- text-generation
language:
- ja
size_categories:
- 1M<n<10M
Descrption
This dataset is part of the OSCAR-2301 cleaned.
There are about 0.5b tokens counted by calm2 tokenizer.
NOTE
This dataset has not passed sentence end boundary determination or Perplexity Filtering, so there is room for improvement in quality.