metadata
dataset_info:
features:
- name: content
dtype: string
splits:
- name: train
num_bytes: 2068932282
num_examples: 30150987
download_size: 1530555311
dataset_size: 2068932282
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
This dataset contains raw Cantonese sentences obtained from Hong Kong-based local forum. Some sentences are processed and used in the Cantonese-Written Chinese Parallel Corpus Gen 3.