بيجبينش

مراجع:

الملخص_السرد_الفهم

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/abstract_narrative_understanding')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 3000
'train' 2400
'validation' 600
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

مفارقات تاريخية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/anachronisms')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 230
'train' 184
'validation' 46
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التناظرية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/analogical_similarity')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 323
'train' 259
'validation' 64
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التحليلي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/analytic_entailment')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 70
'train' 54
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

علم الحساب

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/arithmetic')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 15023
'train' 12019
'validation' 3004
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

ascii_word_recognition

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/ascii_word_recognition')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 5000
'train' 4000
'validation' 1000
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التأليف_التحقق

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/authorship_verification')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 880
'train' 704
'validation' 176
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التصنيف التلقائي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/auto_categorization')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 328
'train' 263
'validation' 65
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التصحيح التلقائي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/auto_debugging')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 34
'train' 18
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

bbq_lite_json

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/bbq_lite_json')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 16076
'train' 12866
'validation' 3210
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

bridging_anaphora_resolution_barqa

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/bridging_anaphora_resolution_barqa')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 648
'train' 519
'validation' 129
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

حكم_سببي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/causal_judgment')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 190
'train' 152
'validation' 38
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

السبب_ والتأثير

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/cause_and_effect')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 153
'train' 123
'validation' 30
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

checkmate_in_one

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/checkmate_in_one')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 3498
'train' 2799
'validation' 699
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

chess_state_tracking

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/chess_state_tracking')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 6000
'train' 4800
'validation' 1200
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

chinese_remainder_theorem

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/chinese_remainder_theorem')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 500
'train' 400
'validation' 100
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

cifar10_classification

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/cifar10_classification')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 20000
'train' 16000
'validation' 4000
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

code_line_description

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/code_line_description')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 60
'train' 44
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

أسماء رمزية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/codenames')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 85
'train' 68
'validation' 17
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

اللون

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/color')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 4000
'train' 3200
'validation' 800
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

عروس

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/common_morpheme')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 50
'train' 34
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تركيبات_مفاهيمية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/conceptual_combinations')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 103
'train' 84
'validation' 19
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

conlang_translation

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/conlang_translation')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 164
'train' 132
'validation' 32
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

السياقية_المعلمية_المعرفة_التعارضات

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/contextual_parametric_knowledge_conflicts')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 17528
'train' 14023
'validation' 3505
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

كراش_بلوسوم

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/crash_blossom')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 38
'train' 22
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

crass_ai

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/crass_ai')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 44
'train' 28
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

cryobiology_spanish

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/cryobiology_spanish')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 146
'train' 117
'validation' 29
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

كريبتونيت

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/cryptonite')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 26157
'train' 20926
'validation' 5231
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

cs_algorithms

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/cs_algorithms')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1320
'train' 1056
'validation' 264
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

كشف_الرائحة القاتمة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/dark_humor_detection')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 80
'train' 64
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تاريخ_فهم

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/date_understanding')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 369
'train' 296
'validation' 73
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

disambiguation_qa

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/disambiguation_qa')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 258
'train' 207
'validation' 51
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

discourse_marker_prediction. التوقع

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/discourse_marker_prediction')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 857
'train' 686
'validation' 171
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

disfl_qa

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/disfl_qa')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 8000
'train' 6400
'validation' 1600
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

dyck_languages

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/dyck_languages')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1000
'train' 800
'validation' 200
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الابتدائية_الرياضية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/elementary_math_qa')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 38160
'train' 30531
'validation' 7629
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

فيلم emoji

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/emoji_movie')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 100
'train' 80
'validation' 20
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

emojis_emotion_ التنبؤ

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/emojis_emotion_prediction')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 131
'train' 105
'validation' 26
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الأحكام التجريبية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/empirical_judgments')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 99
'train' 80
'validation' 19
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الأمثال الإنجليزية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/english_proverbs')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 34
'train' 18
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

English_russian_proverbs

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/english_russian_proverbs')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 80
'train' 64
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

استلزم القطبية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/entailed_polarity')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 148
'train' 119
'validation' 29
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

استلزم_قطبية_هندية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/entailed_polarity_hindi')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 138
'train' 111
'validation' 27
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

المنطق المعرفي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/epistemic_reasoning')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 2000
'train' 1600
'validation' 400
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تقييم_المعلومات_الأساسية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/evaluating_information_essentiality')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 68
'train' 52
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

مدقق الحقائق

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/fact_checker')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 7154
'train' 5724
'validation' 1430
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الخيال

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/fantasy_reasoning')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 201
'train' 161
'validation' 40
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

قليل_shot_nlg

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/few_shot_nlg')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 153
'train' 123
'validation' 30
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الشكل_من_الكشف_الخطاب

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/figure_of_speech_detection')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 59
'train' 43
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الرسمية_الخطأ_المسلسل_التفويض

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/formal_fallacies_syllogisms_negation')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 14200
'train' 11360
'validation' 2840
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

جوهرة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/gem')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 14802
'train' 11845
'validation' 2957
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الجنس_ شاملة_الجنس_الألماني

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/gender_inclusive_sentences_german')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 200
'train' 160
'validation' 40
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

معرفة عامة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/general_knowledge')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 70
'train' 54
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الأشكال الهندسية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/geometric_shapes')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 359
'train' 288
'validation' 71
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الهدف_خطوة_ويكيشو

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/goal_step_wikihow')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 7053
'train' 5643
'validation' 1410
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

gre_reading_comprehension

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/gre_reading_comprehension')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 31
'train' 15
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

hhh_alignment

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/hhh_alignment')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 221
'train' 179
'validation' 42
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

hindi_question_answering

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/hindi_question_answering')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 6610
'train' 5288
'validation' 1322
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

المعرفة الهندوسية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/hindu_knowledge')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 175
'train' 140
'validation' 35
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

السمية hinglish_toxicity

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/hinglish_toxicity')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 200
'train' 160
'validation' 40
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

حواس_أعضاء_إنسان

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/human_organs_senses')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 42
'train' 26
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

مفرط

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/hyperbaton')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 50000
'train' 40000
'validation' 10000
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

Ident_math_theorems

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/identify_math_theorems')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 53
'train' 37
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

Ident_odd_metaphor

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/identify_odd_metaphor')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 47
'train' 31
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التضمينات

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/implicatures')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 492
'train' 394
'validation' 98
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

العلاقات الضمنية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/implicit_relations')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 85
'train' 68
'validation' 17
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

intent_recognition

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/intent_recognition')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 693
'train' 555
'validation' 138
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

international_phonetic_alphabet_nli

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/international_phonetic_alphabet_nli')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 126
'train' 101
'validation' 25
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الدولية_فونية_الأبجدية_ترجمة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/international_phonetic_alphabet_transliterate')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1003
'train' 803
'validation' 200
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تقاطع_ الهندسة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/intersect_geometry')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 249999
'train' 200000
'validation' 49999
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

السخرية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/irony_identification')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 99
'train' 80
'validation' 19
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

kanji_ascii

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/kanji_ascii')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1092
'train' 875
'validation' 217
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الكانادا

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/kannada')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 316
'train' 253
'validation' 63
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

key_value_maps

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/key_value_maps')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 101
'train' 80
'validation' 21
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

known_unknowns

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/known_unknowns')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 46
'train' 30
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

ألعاب_اللغة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/language_games')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 2128
'train' 1704
'validation' 424
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تحديد اللغة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/language_identification')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 10000
'train' 8000
'validation' 2000
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

linguistic_mappings

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/linguistic_mappings')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 15527
'train' 12426
'validation' 3101
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

linguistics_puzzles

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/linguistics_puzzles')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 2000
'train' 1600
'validation' 400
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

list_functions

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/list_functions')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 10750
'train' 8700
'validation' 2050
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

logic_grid_puzzle

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/logic_grid_puzzle')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1000
'train' 800
'validation' 200
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

منطق_أحراج

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/logical_args')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 32
'train' 16
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الاقتطاع المنطقي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/logical_deduction')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1500
'train' 1200
'validation' 300
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

كشف_خطأ_منطقي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/logical_fallacy_detection')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 2800
'train' 2240
'validation' 560
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تسلسل منطقي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/logical_sequence')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 39
'train' 23
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الاستنتاج الرياضي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/mathematical_induction')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 69
'train' 53
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

المصفوفة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/matrixshapes')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 4462
'train' 3570
'validation' 892
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

metaphor_boolean

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/metaphor_boolean')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 680
'train' 544
'validation' 136
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

استعارة_فهم

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/metaphor_understanding')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 234
'train' 188
'validation' 46
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

دقيقة_ألغاز_قا

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/minute_mysteries_qa')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 477
'train' 383
'validation' 94
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

المفاهيم الخاطئة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/misconceptions')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 219
'train' 176
'validation' 43
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

المفاهيم الخاطئة_الروسية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/misconceptions_russian')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 49
'train' 33
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

mnist_ascii

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/mnist_ascii')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 69984
'train' 55988
'validation' 13996
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تعديل_حساب

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/modified_arithmetic')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 6000
'train' 4800
'validation' 1200
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الجواز الأخلاقي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/moral_permissibility')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 342
'train' 274
'validation' 68
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

movie_dialog_same_or_different

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/movie_dialog_same_or_different')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 50000
'train' 40000
'validation' 10000
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

توصية الفيلم

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/movie_recommendation')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 500
'train' 400
'validation' 100
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

mult_data_wrangling

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/mult_data_wrangling')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 7854
'train' 6380
'validation' 1474
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

مولتييمو

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/multiemo')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1437281
'train' 1149873
'validation' 287408
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تعليمات_طبيعية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/natural_instructions')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 193250
'train' 154615
'validation' 38635
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/navigate')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1000
'train' 800
'validation' 200
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

nonsense_words_grammar

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/nonsense_words_grammar')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 50
'train' 34
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الروايات

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/novel_concepts')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 32
'train' 16
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

عد_الأشياء

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/object_counting')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1000
'train' 800
'validation' 200
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

فردي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/odd_one_out')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 86
'train' 69
'validation' 17
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

العاملين

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/operators')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 210
'train' 168
'validation' 42
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تجزئة الفقرة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/paragraph_segmentation')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 9000
'train' 7200
'validation' 1800
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

parsinlu_qa

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/parsinlu_qa')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1050
'train' 840
'validation' 210
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

parsinlu_reading_comprehension

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/parsinlu_reading_comprehension')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 518
'train' 415
'validation' 103
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

طيور البطريق في الجدول

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/penguins_in_a_table')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 149
'train' 120
'validation' 29
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

عناصر_ دورية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/periodic_elements')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 654
'train' 524
'validation' 130
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الأدب الفارسي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/persian_idioms')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 66
'train' 50
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

العبارات_العلاقة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/phrase_relatedness')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 100
'train' 80
'validation' 20
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الحدس الجسدي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/physical_intuition')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 81
'train' 65
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الفيزياء

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/physics')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 229
'train' 184
'validation' 45
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

أسئلة_فيزياء

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/physics_questions')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 54
'train' 38
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

play_dialog_same_or_different

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/play_dialog_same_or_different')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 3264
'train' 2612
'validation' 652
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

polish_sequence_labeling

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/polish_sequence_labeling')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 12812
'train' 10250
'validation' 2562
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الافتراضات_ as_nli

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/presuppositions_as_nli')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 735
'train' 588
'validation' 147
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

qa_wikidata

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/qa_wikidata')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 20321
'train' 16257
'validation' 4064
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

سؤال_الاختيار

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/question_selection')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1582
'train' 1266
'validation' 316
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

نص حقيقي أو مزيف

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/real_or_fake_text')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 15088
'train' 12072
'validation' 3016
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

المنطق_ حول_الألوان_الأشياء

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/reasoning_about_colored_objects')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 2000
'train' 1600
'validation' 400
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تكرار_نسخ_لوجيك

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/repeat_copy_logic')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 32
'train' 16
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

إعادة صياغة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/rephrase')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 78
'train' 62
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

riddle_sense

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/riddle_sense')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 49
'train' 33
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

ruin_names

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/ruin_names')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 448
'train' 359
'validation' 89
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

salient_translation_error_detection. الكشف عن الخطأ

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/salient_translation_error_detection')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 998
'train' 799
'validation' 199
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

Scientific_press_release

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/scientific_press_release')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 50
'train' 34
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التحليل_الدلالي_في_محتوى_النص

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/semantic_parsing_in_context_sparc')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1155
'train' 924
'validation' 231
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تحليل_عنصر_دلالي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/semantic_parsing_spider')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1034
'train' 828
'validation' 206
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

غموض الجملة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/sentence_ambiguity')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 60
'train' 44
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التشابه_التجريد

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/similarities_abstraction')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 76
'train' 60
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

simp_turing_concept

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/simp_turing_concept')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 6390
'train' 5112
'validation' 1278
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

simple_arithmetic_json

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/simple_arithmetic_json')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 30
'train' 14
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

simple_arithmetic_json_multiple_choice

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/simple_arithmetic_json_multiple_choice')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 8
'train' 0
'validation' 0
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

simple_arithmetic_json_subtasks

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/simple_arithmetic_json_subtasks')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 30
'train' 15
'validation' 15
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

simple_arithmetic_multiple_targets_json

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/simple_arithmetic_multiple_targets_json')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 10
'train' 0
'validation' 0
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

أسئلة بسيطة_أخلاقية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/simple_ethical_questions')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 115
'train' 92
'validation' 23
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

simple_text_editing

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/simple_text_editing')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 47
'train' 31
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

يشخر

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/snarks')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 181
'train' 145
'validation' 36
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

Social_iqa

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/social_iqa')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1935
'train' 1548
'validation' 387
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

دعم اجتماعي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/social_support')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 897
'train' 718
'validation' 179
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التفاهم الرياضي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/sports_understanding')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 986
'train' 789
'validation' 197
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

قصص_غريبة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/strange_stories')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 174
'train' 140
'validation' 34
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

Strategyqa

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/strategyqa')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 2289
'train' 1832
'validation' 457
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

معلومات كافية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/sufficient_information')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 39
'train' 23
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

انتحار القرص

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/suicide_risk')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 40
'train' 24
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

امثال سواحيلية انجليزية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/swahili_english_proverbs')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 153
'train' 123
'validation' 30
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

swedish_to_german_proverbs

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/swedish_to_german_proverbs')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 72
'train' 56
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

رمز_تفسير

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/symbol_interpretation')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 990
'train' 795
'validation' 195
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

النتائج الزمنية

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/temporal_sequences')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1000
'train' 800
'validation' 200
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

توتر

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/tense')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 286
'train' 229
'validation' 57
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الوقت

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/timedial')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 2550
'train' 2040
'validation' 510
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

موضعي_تحدث

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/topical_chat')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 22295
'train' 17836
'validation' 4459
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تتبع_الكائنات_المجزأة

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/tracking_shuffled_objects')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 3750
'train' 3000
'validation' 750
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

الفهم

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/understanding_fables')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 189
'train' 152
'validation' 37
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التراجع عن التغيير

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/undo_permutation')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 300
'train' 240
'validation' 60
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تحويل_وحدات

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/unit_conversion')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 23936
'train' 19151
'validation' 4785
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تفسير_الوحدات

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/unit_interpretation')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 100
'train' 80
'validation' 20
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

unnatural_in_context_learning

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/unnatural_in_context_learning')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 73420
'train' 58736
'validation' 14684
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

التحقق من صحة الفيتامين

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/vitaminc_fact_verification')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 54668
'train' 43735
'validation' 10933
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

what_is_the_tao

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/what_is_the_tao')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 36
'train' 20
'validation' 16
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

أي ويكي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/which_wiki_edit')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 571
'train' 457
'validation' 114
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

وينووي

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/winowhy')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 2862
'train' 2290
'validation' 572
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تصنيف_الكلمات

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/word_sorting')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 1900
'train' 1520
'validation' 380
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}

تجميع الكلمات

استخدم الأمر التالي لتحميل مجموعة البيانات هذه في TFDS:

ds = tfds.load('huggingface:bigbench/word_unscrambling')
  • الوصف :
The Beyond the Imitation Game Benchmark (BIG-bench) is a collaborative benchmark intended to
probe large language models, and extrapolate their future capabilities.
  • الترخيص : ترخيص Apache 2.0
  • الإصدار : 0.0.0
  • الانقسامات :
انشق، مزق أمثلة
'default' 8917
'train' 7134
'validation' 1783
  • الميزات :
{
    "idx": {
        "dtype": "int32",
        "id": null,
        "_type": "Value"
    },
    "inputs": {
        "dtype": "string",
        "id": null,
        "_type": "Value"
    },
    "targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_targets": {
        "feature": {
            "dtype": "string",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    },
    "multiple_choice_scores": {
        "feature": {
            "dtype": "int32",
            "id": null,
            "_type": "Value"
        },
        "length": -1,
        "id": null,
        "_type": "Sequence"
    }
}