Referencje:
Wyzwanie ARC
Użyj następującego polecenia, aby załadować ten zestaw danych do TFDS:
ds = tfds.load('huggingface:ai2_arc/ARC-Challenge')
- Opis :
A new dataset of 7,787 genuine grade-school level, multiple-choice science questions, assembled to encourage research in
advanced question-answering. The dataset is partitioned into a Challenge Set and an Easy Set, where the former contains
only questions answered incorrectly by both a retrieval-based algorithm and a word co-occurrence algorithm. We are also
including a corpus of over 14 million science sentences relevant to the task, and an implementation of three neural baseline models for this dataset. We pose ARC as a challenge to the community.
- Licencja : Brak znanej licencji
- Wersja : 1.0.0
- Podziały :
Podział | Przykłady |
---|---|
'test' | 1172 |
'train' | 1119 |
'validation' | 299 |
- Cechy :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"choices": {
"feature": {
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"label": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"answerKey": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}
ARC-łatwy
Użyj następującego polecenia, aby załadować ten zestaw danych do TFDS:
ds = tfds.load('huggingface:ai2_arc/ARC-Easy')
- Opis :
A new dataset of 7,787 genuine grade-school level, multiple-choice science questions, assembled to encourage research in
advanced question-answering. The dataset is partitioned into a Challenge Set and an Easy Set, where the former contains
only questions answered incorrectly by both a retrieval-based algorithm and a word co-occurrence algorithm. We are also
including a corpus of over 14 million science sentences relevant to the task, and an implementation of three neural baseline models for this dataset. We pose ARC as a challenge to the community.
- Licencja : Brak znanej licencji
- Wersja : 1.0.0
- Podziały :
Podział | Przykłady |
---|---|
'test' | 2376 |
'train' | 2251 |
'validation' | 570 |
- Cechy :
{
"id": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"question": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"choices": {
"feature": {
"text": {
"dtype": "string",
"id": null,
"_type": "Value"
},
"label": {
"dtype": "string",
"id": null,
"_type": "Value"
}
},
"length": -1,
"id": null,
"_type": "Sequence"
},
"answerKey": {
"dtype": "string",
"id": null,
"_type": "Value"
}
}