Datasets:
License:
metadata
license: apache-2.0
task_categories:
- translation
- text-generation
monolingual.tgz contains a vast collection of Ancient Chinese sentences.
trans.tgz combines vairous types of parallel corpora, with Ancient Chinese sentences aligned with Modern Chinese sentences.
finetune.tgz selects certain classical books as benchmarks, with Ancient Chinese sentences aligned with Modern Chinese sentences.
The data in trans.tgz and finetune.tgz does not overlap. More information can be found here RUCAIBox/Erya (github.com).