xcsr

参考文献:

X-CSQA-en

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-en')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-zh

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-zh')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-de

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-de')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-es

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-es')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-fr

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-fr')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-it

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-it')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-jap

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-jap')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-nl

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-nl')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-pl

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-pl')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-pt

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-pt')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-ru

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-ru')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-ar

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-ar')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-vi

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-vi')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-こんにちは

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-hi')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-sw

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-sw')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CSQA-ur

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CSQA-ur')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1074
'validation' 1000
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-en

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-en')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-zh

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-zh')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-de

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-de')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-es

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-es')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-fr

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-fr')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-it

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-it')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-jap

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-jap')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-nl

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-nl')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-pl

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-pl')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-pt

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-pt')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-ru

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-ru')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-ar

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-ar')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-vi

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-vi')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-hi

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-hi')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-sw

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-sw')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}

X-CODAH-ur

次のコマンドを使用して、このデータセットを TFDS にロードします。

ds = tfds.load('huggingface:xcsr/X-CODAH-ur')
  • 説明
To evaluate multi-lingual language models (ML-LMs) for commonsense reasoning in a cross-lingual zero-shot transfer setting (X-CSR), i.e., training in English and test in other languages, we create two benchmark datasets, namely X-CSQA and X-CODAH. Specifically, we automatically translate the original CSQA and CODAH datasets, which only have English versions, to 15 other languages, forming development and test sets for studying X-CSR. As our goal is to evaluate different ML-LMs in a unified evaluation protocol for X-CSR, we argue that such translated examples, although might contain noise, can serve as a starting benchmark for us to obtain meaningful analysis, before more human-translated datasets will be available in the future.
  • ライセンス: 不明なライセンス
  • バージョン: 1.1.0
  • 分割:
スプリット
'test' 1000
'validation' 300
  • 特徴
{
    "id": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "lang": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question_tag": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "question": {
        "feature": {
            "stem": {
                "dtype": "string",
                "id": null,
                "_type": "Value"
            },
            "choices": {
                "feature": {
                    "label": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    },
                    "text": {
                        "dtype": "string",
                        "id": null,
                        "_type": "Value"
                    }
                },
                "length": -1,
                "id": null,
                "_type": "Sequence"
            }
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "answerKey": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    }
}