참고자료:
크기가 큰
TFDS에 이 데이터세트를 로드하려면 다음 명령어를 사용하세요.
ds = tfds.load('huggingface:lccc/large')
- 설명 :
LCCC: Large-scale Cleaned Chinese Conversation corpus (LCCC) is a large corpus of Chinese conversations.
A rigorous data cleaning pipeline is designed to ensure the quality of the corpus.
This pipeline involves a set of rules and several classifier-based filters.
Noises such as offensive or sensitive words, special symbols, emojis,
grammatically incorrect sentences, and incoherent conversations are filtered.
- 라이센스 : MIT
- 버전 : 1.0.0
- 분할 :
나뉘다 | 예 |
---|---|
'train' | 12007759 |
- 특징 :
{
"dialog": [
{
"dtype": "string",
"id": null,
"_type": "Value"
}
]
}
베이스
TFDS에 이 데이터세트를 로드하려면 다음 명령어를 사용하세요.
ds = tfds.load('huggingface:lccc/base')
- 설명 :
LCCC: Large-scale Cleaned Chinese Conversation corpus (LCCC) is a large corpus of Chinese conversations.
A rigorous data cleaning pipeline is designed to ensure the quality of the corpus.
This pipeline involves a set of rules and several classifier-based filters.
Noises such as offensive or sensitive words, special symbols, emojis,
grammatically incorrect sentences, and incoherent conversations are filtered.
- 라이센스 : MIT
- 버전 : 1.0.0
- 분할 :
나뉘다 | 예 |
---|---|
'test' | 10000 |
'train' | 6820506 |
'validation' | 20000 |
- 특징 :
{
"dialog": [
{
"dtype": "string",
"id": null,
"_type": "Value"
}
]
}