metadata
license: cc0-1.0
language:
- zh
pretty_name: Cantonese-Written Chinese Parallel Dataset
task_categories:
- translation
size_categories:
- 100K<n<1M
Cantonese-Written Chinese Parallel Dataset with roughly 1 million pairs of Cantonese and Written Chinese sentences, in JSONL format. The sentence topics include news, novels and daily conversations. Translations are done by Microsoft Azure Translate API, and manually corrected problems by me. Sentences are also manually verified by me one by one.
The dataset could be extended if there are more resources, in terms of manpower and money.